Search

Histone deacetylase 7 (HDAC7) belongs to the class II HDAC family and plays a pivotal role in the maintenance of endothelium integrity. There are 8 splicing variants in mouse HDAC7 mRNAs. Within the 5’ terminal non-coding area of some variants, there exist some short open reading frames (sORFs). Whether these sORFs can be translated and whether the resulting peptides play roles in cellular physiology remain unclear. In this study, we demonstrated that one sORF encoding a 7-amino-acid (7-aa) peptide could be translated in vascular progenitor cells (VPCs). Importantly, this 7-aa peptide (7A) could transfer the phosphate group from the phosphorylated Ser393 site of MEKK1 to the Thr145 site of 14-3-3γ protein. The phosphorylated 7A (7Ap) could then directly phosphorylate 14-3-3γ protein in a cell-free, in-gel buffer system. The adjacent histidine and proline residues are essential for the phosphate group reception and transfer. In vitro functional analyses revealed that 7A and 7Ap increased VPC self-renewal and migration and enhanced vascular endothelial growth factor (VEGF)-induced VPC migration and differentiation toward the endothelial cell (EC) lineage, in which MEKK1 and 14-3-3γ served as the upstream kinase and downstream effector, respectively. Knockdown of either MEKK1 or 14-3-3γ attenuated VEGF-induced VPC migration and differentiation. Exogenous 7Ap could rescue the effect of VEGF on the MEKK1 siRNA-transfected VPCs but not on the 14-3-3γ siRNA-transfected VPCs. In vivo studies revealed that 7A, especially 7Ap, induced capillary vessel formation in Matrigel plug assays, increased re endothelialization and suppressed neointima formation in the femoral artery injury model, and promoted foot blood perfusion recovery in the hindlimb ischemia model by increasing Sca1+ cell niche formation. These results indicate that the sORFs within the non-coding area can be translated and that 7A may play an important role in cellular processes, such as proliferation, migration and differentiation, by acting as a phosphorylation carrier.

Introduction Endothelial-mesenchymal transition (EndMT) is a process where endothelial cells (ECs) undergo dramatic cytoskeletal remodelling and gene expression changes, leading to a more motile mesenchymal phenotype that can significantly contribute to cardiac fibrosis. Histone deacetylase 3 (HDAC3), a class I HDAC, is essential in the maintenance of endothelial homeostasis. Our previous study indicates that mouse HDAC3 mRNA can undergo unconventional splicing, and that the HDAC3-alpha (HDAC3a) spliced isoform promotes EndMT in human aortic endothelial cells (HAECs). This project aims to elucidate the role of HDAC3a in mediating EndMT and its underlying mechanisms in the development of pressure-overload-induced cardiac fibrosis. Methods and results Our previous study reveals that HDAC3a promotes EndMT in HAECs via activation of transforming growth factor-beta 2, which could possibly be mediated through increased protease enzyme activity. Therefore HAECs overexpressed with HDAC3a via adenoviral gene transfer (Ad-HDAC3a) have undergone a protease array, which revealed that cells overexpressed with HDAC3a have an elevated level of a disintegrin and metalloproteinase with thrombospondin motifs 1 (ADAMTS1). This was further confirmed with increased mRNA and protein expression levels of ADAMTS1 in Ad-HDAC3a-infected HAECs using RT-qPCR and Western blotting analysis respectively (both n = 4, P < 0.05). A mouse model of pressure overload-induced cardiac fibrosis was established by transverse aortic constriction (TAC), and cardiac fibrosis was confirmed by increased Col1a1 expression seen using RT-qPCR and Picrosirius red staining in heart tissue sections from mice 7 d post-TAC. Immunofluorescent staining detected CD31+/alpha-SMA+ (indicative of EndMT) and HDAC3a+/CD31+ cells in heart tissues from TAC mice but not in those from sham-operated mice. There were significant increases in mRNA expression of HDAC3a and ADAMTS1, together with EndMT transcription factors Snai1, Snai2 and Twist1 in heart tissues from 7 d post-TAC mice compared to those from 7 d post-sham mice (n = 7-– 8, P < 0.05). Elevated HDAC3a and ADAMTS1 protein expression were also detected in the TAC heart tissues (n = 5, P < 0.05). Conclusion These results indicate an association between the HDAC3 splicing isoform HDAC3a, ADAMTS1 and EndMT during TAC-induced cardiac fibrosis. Further investigation of the underlying mechanisms mediated by HDAC3a and ADAMTS1 in cardiac fibrosis is warranted.

The generation of induced pluripotent stem (iPS) cells is a fascinated tool for regenerative medicine. However, the main restriction for iPS cell application is a risk of tumour development. In this study we established a new method to generate a partially induced stem (PiPS) cells by transferring four reprogramming factors (OCT4, SOX2, KLF4 and c-MYC) to human fibroblasts. PiPS cells did not form tumours in vivo, and have abilities to specifically differentiate into different cell lineages, such as neurons, adipocytes, osteocytes, chondrocytes as well as endothelial cells in response to define media and culture conditions. The PiPS-derived endothelial cells expressed a panel of endothelial markers at gene and protein levels and they formed vascular tubes in vitro and in vivo Matrigel plaque assays. To clarify the mechanism of PiPS cell differentiation into endothelial lineage, data from a series of experiments indicate a novel gene SETSIP was crucial in endothelial differentiation. SETSIP was expressed in parallel with endothelial genes, and its overexpression induced endothelial marker expression, while its downregulation by shRNA resulted in suppression of these genes. Interesting, this protein was translocated to the cell nucleus during endothelial differentiation, while ChIP assays showed that it bound direct to VE-cadherin promoter. Functionally, PiPS-derived endothelial cells displayed clearly endothelial properties in two animal models. When seeded on decellularised vessels in a special constructed bioreactor and implanted in SCID mice, the cells displayed well attachment, stabilisation, patency and typical vascular structure. Also, these cells showed endothelial engraftment to form typical vascular architecture when they were injected into the infarcted tissues in an ischaemia model. Thus, we developed a new method to generate PiPS cells from human fibroblasts that can differentiate into a variety of cell types without tumour risk, in which endothelial cells can be produced via OCT4-VEGF-SETSIP pathway useful for regenerating damaged tissue and vessels in vivo.

The generation of induced pluripotent stem (iPS) cells from somatic cells can be a useful tool for regenerative medicine. The artificial nature of iPS cells raises concerns whether iPS and embryonic stem (ES) cells, are molecularly and functionally comparable. In this study we generated iPS cells using a genomic integration free method, and using the advance of proteomics, we aimed to elucidate any differences in the protein profile between iPS and ES cells, in terms of pluripotency and differentiation potential to vascular cell lineages. The results demonstrated that 180 proteins were differentially expressed, in which 66% showed a decreased expression pattern in iPS cells. Functional classification analysis revealed that these proteins were associated with mRNA processing, energy metabolism, cytoskeleton, proliferation, and differentiation. Further experiments demonstrated that iPS cells have a decreased proliferation capacity when compared to ES cells. Importantly, iPS cells also displayed a greater potential to differentiate into smooth muscle and endothelial cell lineages, when compared to ES cells. When iPS and ES cells were seeded on collagen IV-coated dishes and cultured with differentiation media, MEM supplemented with 10% serum, iPS cells displayed a differentiation morphology as early as day 3 in comparison to ES cells, and real-time PCR data confirmed that iPS express smooth muscle cell markers such as SMA, calponin, and SM22 in higher levels. When vascular endothelial growth factor (VEGF) was added to differentiation media, iPS cells expressed endothelial markers such as CD31, CD144, Flk-1 and Flt-1 as early as day 3 on differentiation, while ES cells showed limited expression of these markers during the early stages of differentiation. Therefore, identifying differentially regulated protein expression between the two cell types and the pathways they influence, will allow us to elucidate the mechanisms that lead to vascular cell differentiation, thus making iPS technology a successful tool for regenerative medicine and treatment of vascular diseases.

Histone deacetylases (HDACs) play a crucial role in transcriptional regulation through modulation of chromatin structure. The class I HDAC, HDAC3, is involved in maintaining endothelial cell integrity.1 Sustained activation of the x-box binding protein 1 (XBP1), an endoplasmic reticulum stress response transcription factor, results in the development of atherosclerosis in apoE−/− mice.2 HDAC3 and XBP1 are similarly expressed in the bifurcation regions of the aorta. In this study, we investigated whether cross-talk existed between HDAC3 and XBP1, and its role in the maintenance of endothelial cell integrity. Our study demonstrated that disturbed flow upregulated HDAC3 and unspliced XBP1 (XBP1u) protein production through the KDR/PI3K/Akt pathway. Knockdown of XBP1 by shRNA lentiviral transfection ablated disturbed flow-induced HDAC3 upregulation. Similarly to HDAC3, overexpression of XBP1u by adenoviral gene transfer increased Akt phosphorylation at Serine473 and haem oxygenase 1 gene transcription, which showed a protective role in hydrogen peroxide-induced apoptosis of endothelial cells. Co-immunoprecipitation assays demonstrated that HDAC3 physically associates with XBP1u and this could be enhanced by disturbed flow and VEGF treatment. The use of truncated HDAC3 constructs demonstrated that XBP1 binds to the central section of HDAC3. Further experiments indicated that overexpression of XBP1u increased the binding of HDAC3 to IRE1α, Akt and PI3K, especially after VEGF treatment. In contrast, sustained activation of spliced XBP1 decreased HDAC3 protein production through transcriptional suppression, leading to endothelial apoptosis. These results suggest that XBP1u protects endothelial cells from oxidative stress by interacting with HDAC3. Targeting this interaction may provide novel therapeutic strategies for vascular disease via maintaining endothelial integrity.

An initial step in the development of atherosclerosis is endothelial cell dysfunction. Our previous study has shown that sustained activation of X-box binding protein 1 (XBP1) splicing, a key signal transducer in endoplasmic reticulum stress response, leads to endothelial apoptosis and atherosclerosis development (Zeng et al, PNAS 2009). Autophagy is characterised as a survival response as well as a pathway culminating in cell death. In this study, we aimed to investigate whether XBP1 splicing could also activate autophagy in endothelial cells, and its role in the cell survival and apoptosis. XBP1 splicing was introduced to human umbilical vein endothelial cells by adenoviral gene transfer, and autophagic vesicle formation was observed 48 and 72 h postinfection. Autophagy gene expression such as Beclin 1, LC3-β, NIX, ATGF-tv1, BNIP2, BNIP3, ATG5 and ATG12 was induced by XBP1 splicing as detected at RNA and protein levels. Endothelial cells transfected with the LC3β-GFP construct showed a high GFP expression of autophagic vesicle formation only in the presence of spliced XBP1 compared with the non-infected cells. Moreover, our results showed that endostatin, a well-characterised endogenous inhibitor of angiogenesis that induces apoptosis, activated autophagy genes 4 h after treatment, through XBP1 splicing in an IRE1-α-dependent manner. Knockdown of XBP1 or IRE1α by ShRNA in endothelial cells ablated endostatin-induced autophagy responsive genes expression and enhanced the apoptotic effect of endostatin. These results suggest that activation of XBP1 splicing can trigger an autophagic survival signal pathway, providing novel insights into the mechanisms involved in endothelial survival.

The X-box binding protein 1 (XBP1) is an endoplasmic reticulum stress response transcription factor. Our previous study showed that sustained activation of XBP1 splicing led to atherosclerosis development (Zeng et al. PNAS 2009). However, the function of XBP1 expression and splicing in endothelial cells and angiogenesis remains unclear. To study this issue, we generated XBP1 knockout mice by deletion of exons 1 and 2 of the gene. XBP1-null (XBP1−/−) embryos at E12.5 showed growth retardation and pale colouration phenotype. The average body weight of XBP1−/−embryos was 40% less than that of wild-type (XBP1+/+) animals. The blood vessel density in XBP1−/− embryos was significantly reduced, owing to a smaller number of CD31+ and Flk1+ cells. In in vitro culture of whole embryonic cells, XBP1−/− cells grew significantly more slowly and lost response to VEGF stimulation. To study the mechanism of XBP1-mediated cell growth, human umbilical vein endothelial cells were treated with VEGF that transiently activated IRE1α phosphorylation at Ser724. The mRNA movement inhibitor, cycloheximide, ablated VEGF-induced IRE1α phosphorylation and XBP1 splicing. Co-immunoprecipitation assay showed that there was interaction among KDR, IRE1α and unspliced XBP1 (XBP1u), which could be increased by VEGF treatment. Further experiments demonstrated that the C-terminal region of KDR and the kinase domain of IRE1α are responsible for their interactions. MTT and BrdU incorporation assays indicate that transient activation of XBP1 splicing increased while long-term activation decreased endothelial survival and proliferation. Knockdown of XBP1 or IRE1α ablated VEGF-induced proliferation in endothelial cells. Immunofluorescent staining and TOP Flash reporter assay showed that overexpression of XBP1s increased β-catenin translocation into the nucleus. Thus, this study demonstrated for the first time that XBP1 is crucial for endothelial growth and angiogenesis, in which VEGF-stimulated IRE1α/XBP1 splicing system and interaction with β-catenin are key elements, indicating a potential target of XBP1 for protecting endothelial integrity.

Histone deacetylase 3 (HDAC3), a member of the class I histone deacetylases, is known to have a crucial role in endothelial cell differentiation12 and maintenance of endothelial integrity in response to disturbed flow.3 In this study, we investigated the function of HDAC3 in endothelial protection from inflammation, and the underlying mechanism. The inflammatory mediator lipopolysaccharides (LPS) induced HDAC3 and galectin-9 production in a similar pattern in human umbilical vein endothelial cells. Overexpression of HDAC3 by adenoviral gene transfer increased galectin-9 expression. Pharmacological inhibition of HDAC activity with a pan-HDAC inhibitor (TSA) or HDAC3 specific inhibitor (apicidin) reduced the baseline and LPS-induced galectin-9 expression. In addition, knockdown of HDAC3 through shRNA lentiviral transfection abolished the baseline and LPS-induced galectin-9 expression. Similar results were observed on interferon γ (IFNγ)-induced galectin-9 expression. To explore the underlying mechanism, the interaction of HDAC3 with galectin-9 upstream signal pathway phosphoinositol-3-kinase (PI3K)/signal transmission and transducer 3 (Stat3)/interferon response factor (IRF3) was assessed. Co-immunoprecipitation assay showed that HDAC3 formed a complex with PI3K/Stat3/IRF3, which was enhanced by LPS and IFNγ treatment. Using truncated forms of HDAC3, it was shown that the C-terminal of HDAC3 was responsible for the formation of the complex. Furthermore, venous administration of LPS or IFNγ to mice increased HDAC3 expression and binding to the above-mentioned proteins, leading to galectin-9 expression in the aorta. These results suggest that HDAC3 may protect endothelial cell from inflammation through galectin-9 expression, which may have an impact on preventing vascular inflammation related to the development of atherosclerosis.

Objective Protein kinase C (PKC) δ functions as a signal transducer mediating several essential functions of cell proliferation and apoptosis. However, the effect of PKCδ on neointimal formation is currently unknown. In this study, we used a vascular injury model in PKCδ knockout mice to investigate the role of PKCδ in the lesion development and underlying mechanism. Methods and results 3 weeks after wire injury of femoral arteries, neointimal lesions were significantly increased in PKCδ−/− mice compared with wild-type animals. Immunohistochemical staining revealed that total numbers of smooth muscle cells and macrophages in the lesions were markedly elevated without the alterations of the ratio between these two types of cells. To further study the mechanisms of PKCδ-mediated increase in the lesion, an in vivo endothelial migration model was established to evaluate endothelial wound healing after wire injury. Data indicate that re-endothelialisation of the injured vessel was markedly delayed in PKCδ−/− mice that coincided with more severe intimal hyperplasia. When endothelial cells were cultivated from cardiac tissues of PKCδ−/− and PKCδ+/+ mice, the ability of cell migration was significantly reduced in PKCδ−/− mice, but no difference in proliferation and apoptosis was found. This altered endothelial migration was also verified by PKCδ inhibitor and siRNA techniques in wild-type cells. Interestingly, vasohibin-1, an anti-migration protein, was elevated in endothelial cells derived from PKCδ-deficient mice, which was identified largely owing to delayed protein degradation mediated by PKCδ. Downregulation of vasohibin-1 restored the migration rate of PKCδ−/− endothelial cells to a level similar to that in wild-type cells. Conclusion Our data provide the first evidence that PKCδ-enhanced neointima formation occurs mainly due to delayed re-endothelialisation, which is mediated by increased cellular vasohibin-1 that is regulated by PKCδ.