This study identifies a role for lncRNA MEG8 in the regulation of angiogenesis. Loss of MEG8 in ECs is accompanied by impaired angiogenic sprouting and proliferation. Furthermore, TFPI2 expression was induced following loss of MEG8. Mechanistically, MEG8 was found to interact with EZH2, which is part of the PRC2 complex. Accordingly, silencing of MEG8 resulted in a reduction of repressive H3K27me3 mark at the TFPI2 promoter. Silencing of TFPI2 rescued sprouting capacity of MEG8 deficient cells.
MEG8 is found in the 14q32 cluster in humans and 12F1 in mice. Recently, Sui et al. showed that downregulation of Meg8, also known as Rian in mice, inhibited cell viability and angiogenesis in mouse brain microvascular ECs, in accordance with our findings in a HUVEC model [20]. Voellenkle et al. showed an upregulation of MEG8 expression in HUVECs exposed to hypoxia [30]. We observed an induction of MEG8 expression in ISHD patients. Taken together, these results point to a possible protective role for MEG8 in ischemia. We hypothesize that MEG8 is upregulated during ischemia to contribute to cell survival and angiogenesis. A study by Zhang et al. identified a role for MEG8 in VSMC proliferation and migration through targeting PPARα. Contrary to findings in the endothelium, inhibition of MEG8 improved proliferation following PPARα. Enhanced expression of MEG8 was found to repress proliferation and migration of VSMCs. [19]. These results hint to a cell-specific function of MEG8.
Our results show a reduction in sprouting and proliferation (Fig. 1C-D), but not migration (Fig. 1E) after loss of MEG8. Extension of the sprouts, mediated by stalk cell proliferation, is therefore more likely to be disturbed when MEG8 is silenced. Sui et al. show reduced Vegf expression following loss of Meg8 in a mouse model which would result in impaired angiogenesis [20]. This mechanism does not seem to play a role in our human angiogenesis model since stimulation of spheroids with VEGF did not rescue sprouting following MEG8 knockdown. This would suggest that, in human ECs, MEG8 regulates angiogenesis independently of VEGF expression. Interestingly, MEG3, a lncRNA upstream of MEG8 in the 14q32 cluster has been shown to play a role in angiogenesis as well. Contrary to Meg8, inhibition of Meg3 was shown to induce angiogenesis in mice [16, 31]. Mechanistically, Meg3 was shown to negatively regulate the Notch pathway. Chen et al. showed differential expression of Notch genes (Notch2, Notch3, Hes1) in murine hepatic stellate cells after Meg8 knockdown [32]. We did not observe any change in Notch gene expression following loss of MEG8 in the human endothelium. These findings would suggest these two lncRNAs from the same cluster function in different cellular pathways.
MEG8 was found to localize to the chromatin (Fig. 3A), which suggested that MEG8 could be involved in epigenetic regulation of transcription, as has been shown previously for many nuclear lncRNAs [9]. RNA sequencing data suggested that MEG8 regulates expression of specific genes such as TFPI2 rather than global gene transcription (Fig. 2A-B). TFPI2 was selected as an interesting target since it has been suggested to play an inhibitory role in angiogenesis. Overexpression of TFPI2 was found to inhibit capillary formation in vitro [26–28]. Modulation of angiogenic growth factor levels has been proposed as a mechanism for this anti-angiogenic effect [27]. Interestingly, TFPI2 was found to be repressed in invasive cancers [33, 34].
We then sought to elucidate the underlying mechanism. MEG8 was found to interact with EZH2 (Fig. 3C) in primary ECs and there was a reduction in the repressive H3K27me3 mark at the TFPI2 promoter after MEG8 silencing (Fig. 3D). These results suggest MEG8 mediates EZH2 recruitment and subsequent H3K27 trimethylation of specific genomic regions.
We hypothesized that inhibition of TFPI2 induction by MEG8 silencing could rescue the effect of MEG8 silencing. Indeed, silencing of TFPI2 rescued sprouting following loss of MEG8 (Fig. 4A). These results suggest that MEG8 regulates angiogenic sprouting at least in part through regulation of TFPI2 expression. In addition, we measured the rate of proliferation, as this contributes to sprout extension. Knockdown of MEG8 resulted in impaired sprouting, we therefore asked whether TFPI2 was involved in regulation of proliferation as well. In previous studies, TFPI2 was shown to be induced by VEGF in the endothelium. Modulation of growth factor levels was proposed as a mechanism for the anti-angiogenic effects of TFPI2 [26]. TFPI2 was observed in turn to inhibit VEGF induced proliferation, suggesting a negative feedback loop to control cellular turnover [26, 35]. This mechanism appears less likely to play a role in our angiogenesis model since stimulation of spheroids with VEGF did not rescue sprouting. In our model, additional silencing of TFPI2 did not rescue proliferation rates in HUVECs after silencing of MEG8 (Fig. 4B-C). This result suggests that the effect of TFPI2 on sprouting is independent of proliferation. The degradation of the basement membrane is a crucial first step in the sprouting phase of angiogenesis [36]. TFPI2 is deposited in the extracellular matrix, inhibits a range of proteases such as MMPs and protects the matrix from degradation. TFPI2 thereby prevents tumour invasion and metastasis. [37]. We hypothesize that MEG8 could play a role in regulating TFPI2 protein levels and its deposition in the extracellular matrix. Remodelling of the extracellular matrix is potentially mediated through the MEG8/TFPI2 axis, and thereby contributes to angiogenesis.
Taken together, our study shows that MEG8 is regulated in ISHD and controls angiogenesis via epigenetic regulation of TFPI2 expression. Our study further highlights the MEG8/TFPI2 axis as potential therapeutic approach to improve angiogenesis in ISHD.