Thrombocytopenia induced by the histone deacetylase inhibitor abexinostat involves p53-dependent and -independent mechanisms.

Abexinostat is a pan histone deacetylase inhibitor (HDACi) that demonstrates efficacy in malignancy treatment. Like other HDACi, this drug induces a profound thrombocytopenia whose mechanism is only partially understood. We have analyzed its effect at doses reached in patient plasma on in vitro megakaryopoiesis derived from human CD34(+) cells. When added at day 0 in culture, abexinostat inhibited CFU-MK growth, megakaryocyte (MK) proliferation and differentiation. These effects required only a short incubation period. Decreased proliferation was due to induction of apoptosis and was not related to a defect in TPO/MPL/JAK2/STAT signaling. When added later (day 8), the compound induced a dose-dependent decrease (up to 10-fold) in proplatelet (PPT) formation. Gene profiling from MK revealed a silencing in the expression of DNA repair genes with a marked RAD51 decrease at protein level. DNA double-strand breaks were increased as attested by elevated γH2AX phosphorylation level. Moreover, ATM was phosphorylated leading to p53 stabilization and increased BAX and p21 expression. The use of a p53 shRNA rescued apoptosis, and only partially the defect in PPT formation. These results suggest that HDACi induces a thrombocytopenia by a p53-dependent mechanism along MK differentiation and a p53-dependent and -independent mechanism for PPT formation.

Influence of homocysteine on matrix metalloproteinase-2: activation and activity.

Increased levels of the physiological amino acid homocysteine (Hcy) are considered a risk factor for vascular disease. Hyperhomocysteinemia causes an intense remodelling of the extracellular matrix in arterial walls, particularly an elastolysis involving metalloproteinases. We investigated the activation of the latent elastolytic metalloproteinase proMMP-2 (72 kDa) by Hcy. Hcy was proved to exert a dual effect, activating proMMP-2 at low molar ratio (MR 10:1) and inhibiting active MMP2 at high molar ratio (MR > 1000:1). Methionine and the disulphide homocystine did not activate nor inhibit MMP-2, showing that the activation as well as the inhibition requires the thiol group to be free. The activation of proMMP-2 by Hcy is in accordance with the "cysteine-switch" mechanism, but occurs without further autoproteolysis of the enzyme molecule. In contrast with Hcy, the other physiological thiol compounds cysteine and reduced glutathione did not activate proMMP-2. These results suggest that the direct activation of proMMP2 by Hcy could be one of the mechanisms involved in the extracellular matrix deterioration in hyperhomocysteinemia-associated arteriosclerosis.

Hyperhomocysteinemia induces elastolysis in minipig arteries: structural consequences, arterial site specificity and effect of captopril-hydrochlorothiazide.

Hyperhomocysteinemia is a risk factor for arterial diseases, and the deterioration of the arterial elastic structures is one of the possible mechanisms underlying this epidemiological association. The aim of this paper is to quantitatively characterize such structural alterations and to explore their causes in a previous model of dietary induced mild hyperhomocysteinemia in minipigs. After four months, both a morphodensitometrical analysis of the elastic structure and a biochemical analysis of elastin and elastase activities were performed on the infrarenal abdominal aorta (IRAA) and the proximal left interventricular coronary artery (LIVCA) of control (C), hyperhomocysteinemic (H) and captopril-hydrochlorothiazide (Cp-Htz, 25 + 12.5 mg/d)-treated (H+/-Cp) minipigs (n = 8/group). Hyperhomocysteinemia was found to induce an increase in parietal elastolytic metalloproteinase activities. It resulted in opening and enlargement of fenestrae through the medial elastic laminae and in a decrease in medial elastin content (p < 10(-3)), expressed as well as volume density (%) as weight concentration (microg elastin/mg dry tissue). The thickness of the media and its basic lamellar organization was unchanged. The reduction in volume density was more dramatic in LIVCA (H: 4.7 +/- 0.9 vs C: 8.8 +/- 2.4), where it was evenly distributed within the media, than in IRAA (H: 6.7 +/- 1.1 vs C: 9.3 +/- 1.2), where the deep medial layers were less affected. Cp-Htz partly prevented the hyperhomocysteinemia-induced reduction of the medial elastin content in LIVCA (5.7 +/- 1.2) and IRAA (7.9 +/- 1.4). This effect, occurring in the subintimal layers of the media in both arteries but not in the deeper layers, resulted in a less beneficial effect in LIVCA than in IRAA. This result parallels the moderate beneficial therapeutic effect of ACE inhibitors against coronary atherosclerosis in humans. This paper reports for the first time a quantitative analysis of the arterial site-dependent deterioration of the elastic structure caused by mild hyperhomocysteinemia and the involvement of metalloproteinases in this process. These results confirm that the plaque-independent damage to elastic structure previously described in hyperhomocysteinemic-atherosclerotic minipigs was mainly due to homocysteine. This highlights that the metalloproteinase-related elastolysis and the subsequent structural deterioration is one of the major events underlying the epidemiological association between mild hyperhomocysteinemia and arterial diseases.