The broad notion that other lncRNAs might guide epigenetic factors in regulating
gene expression took hold following a study on Homeobox gene regulation [10].
In this study, a lncRNA transcribed from the HOXC locus, dubbed HOTAIR, was
demonstrated to recruit PRC2 in trans to the HOXD locus to mediate H3K27me3
deposition and locus silencing [10]. Biochemical confirmation of the reported
role of HOTAIR in establishing repressive chromatin at the HOXD locus came
when HOTAIR was shown to interact not only with PRC2 but also with H3K4
demethylase complex LSD1-CoREST to form an epigenetic silencing scaffold at
the HOXD locus [11] (Fig. 1a). Two independent global RNA immunoprecipitation
studies demonstrated that PRC2 binds to a substantial fraction of lincRNAs and
transcripts genome-wide [12, 13]. These findings consolidated the hypothesis of a
global regulatory interface between epigenetic machinery and lncRNAs responsible
for the dynamic modulation of epigenetic landscapes. However, the centrality of
PRC2 in the gene regulatory function of HOTAIR has been disputed. A recent study
found that HOTAIR could mediate reporter silencing independently of PRC2 [14].
Perhaps, as seems to be the case for Xist, PRC2 functions more as a sensor of
the altered chromatin terrain of the newly HOTAIR-silenced locus and proceeds
to enrich it with H3K27me3 to compact the surrounding chromatin and thereby
stabilise and maintain its transcriptional repression.
Numerous other lncRNAs have been shown to manipulate Polycomb activity
in mediating gene repression. ANRIL, an antisense transcript of the INK4b(p15)ARF(p14)-INK4a(p16) tumour suppressor gene cluster, mediates cis recruitment
of both PRC1 and PRC2 to the locus, which epigenetically silence it via local
enrichment of H3K27me3 and H2AK199ub [15, 16]. KCNQ1OT1, a non-coding
transcript of the imprinted Kcnq1 locus, coordinates lineage-specific transcriptional
silencing of the 200 kb proximal region by scaffolding both PRC2 and H3K9 histone
methyltransferase (HMT) G9a and enriching the locus with their corresponding
repressive marks H3K9me3 and H3K27me3 [17].
LncRNAs can also guide the epigenetic writers of DNA methylation as well
as those of histone marks. For example, in the context of reprogramming
pre-induced pluripotent stem cells into iPSCs, lincRNA-p21 has been shown to
interact not only with H3K9 histone methyltransferase (HMT) but also with DNA
methyltransferase 1 (DNMT1), resulting in increasing H3K9me3 deposition and
CpG island hypermethylation of the target promoters. LincRNA-p21 thus acts
as a barrier to induced pluripotency [18]. It has also been proposed that, in addition
to its role in scaffolding PRC2 and G9a, KCNQ1OT1 interacts with DNMT1 to
direct CpG island methylation of the paternal Kcnq1 locus [19].
As well as recruiting and scaffolding epigenetic modification complexes,
lncRNAs can also modulate their activity by directly or indirectly inducing structural
changes in epigenetic editor proteins. In a classic example, numerous DNA damageinduced ncRNAs transcribed from upstream of the CCND1 gene recruit and allosterically activate TLS (translocated in liposarcoma) to the locus, which then inhibits
locally bound histone acetyltransferase (HAT) CBP/p300, preventing the establishment of the histone acetylation marks necessary for the epigenetic activation of
CCND1 transcription [20]. By contrast, enhancer RNAs (eRNAs) have recently
Altered Long Non-coding RNA Expression in Cancer: Potential Biomarkers. . .
405
Précédent

- 409/569

Suivant