1000/1000
Hot
Most Recent
Enzymes, such as histone methyltransferases and demethylases, histone acetyltransferases and deacetylases, and DNA methyltransferases are known as epigenetic modifiers that are often implicated in tumorigenesis and disease. One of the best-studied chromatin-based mechanism is X chromosome inactivation (XCI), a process that establishes facultative heterochromatin on only one X chromosome in females and establishes the right dosage of gene expression. The specificity factor for this process is the long non-coding RNA Xinactivespecifictranscript (Xist), which is upregulated from one X chromosome in female cells. Subsequently, Xist is bound by the corepressor SHARP/SPEN, recruiting and/or activating histone deacetylases (HDACs), leading to the loss of active chromatin marks such as H3K27ac. In addition, polycomb complexes PRC1 and PRC2 establish wide-spread accumulation of H3K27me3 and H2AK119ub1 chromatin marks. The lack of active marks and establishment of repressive marks set the stage for DNA methyltransferases (DNMTs) to stably silence the X chromosome.
Less than 2% of the genome is transcribed in protein-encoding mRNAs; however, most of it is actively transcribed, which suggests that a fraction produces non-coding RNAs (ncRNAs). ncRNAs are classified based on their size in small ncRNAs (<200 bp) and long ncRNAs (>200 bp, also referred to as lncRNAs) [1][2]. In this review, we focus on lncRNAs.
lncRNAs can be classified based on their genomic localization [3] as well as on their cellular distribution [4]. It is proposed that lncRNAs are organized in secondary and tertiary structures [5] that may offer binding surfaces for proteins containing RNA-recognition motives (RRMs). lncRNAs are capable of interacting with coactivators or corepressors of transcription, recruiting them to specific genes or genomic regions [6][7][8][9]. In addition, lncRNAs are also able to regulate alternative splicing events by interacting with splicing factors [6][10].
Several lncRNAs have been associated with variety of diseases. Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) was found to be upregulated in renal cell carcinoma (RCC), gastric cancer (GC), gallbladder cancer (GBC), colorectal cancer (CRC), multiple myeloma, clear cell renal cell carcinoma (ccRCC), and glioma, as well as in osteosarcoma [11][12][13][14][15][16][17][18], and it has been proposed as a molecular marker therein [14][15][16][19]. The lncRNA imprinted H19 gene is maternally expressed and strongly downregulated directly after birth [20][21][22]. It was shown that H19 is strongly upregulated in gastric cancer [23][24][25], similarly to several other lncRNAs, such as PVT1 oncogene (PVT1), gastric carcinoma high expressed transcript 1 (GHET1), antisense ncRNA in the INK4 locus (ANRIL), SPRY4 intronic transcript 1 (SPRY4-IT1), and the already mentioned MALAT1 [18][26][27][28][29][30]. H19 is also upregulated in other cancer types, such as esophageal cancer, CRC and lung cancer [25]. Another example is represented by homeobox (HOX) transcript antisense RNA (HOTAIR), which is upregulated in hepatocellular carcinoma [31], in colorectal cancer [32], in gastric cancer [33], and pancreatic cancer [34].
In this review, we will focus our attention on X inactive specific transcript (Xist; XIST in human), a lncRNA whose main function is to inactivate one X chromosome in female cells to achieve dosage compensation between males (XY) and females (XX) (see below). Recent studies highlighted its frequent deregulation in cancer. XIST is responsible for silencing several genes, and the observation that the X-linked oncogenes ARAF-1 and ETS-like 1 (ELK-1) are overexpressed in tumors with multiple active X chromosomes [35] suggests that the deregulation of XIST may be associated with cancer. Several studies observed defective X chromosome inactivation (XCI) in breast and basal-like cancer and linked the deregulation of the X chromosome to breast cancer (BC) [36][37][38][39][40][41][42], to ovarian cancer [43], as well as to cancers in patients affected by Klinefelter syndrome [44]. This deregulation is usually given by a loss of XIST as result of disappearance of the inactive X chromosome (Xi) and amplification of the active one (Xa) [37][38][40][43][44].
The gathered knowledge of these studies suggest that lncRNAs are important mediators of pathological conditions and they may, in the future, serve as potential therapeutic targets.
XCI serves as a powerful paradigm to study chromatin dynamics at a chromosomal scale. XCI co-evolved with the mammalian sex chromosomes as a mechanism to equalize the dosage of X-encoded genes between male XY and female XX cells. The central player in this process is Xist, which was discovered as the first functional lncRNA in mammals, being upregulated from the future Xi, coating the Xi in cis, thereby recruiting chromatin remodelers directly and indirectly rendering the X chromosome inactive. Xist is located on the X chromosome and it is surrounded by several other lncRNA-encoding genes, including Tsix, just proximal to Xist (Jpx), and five prime to XistT (Ftx), which, in mouse, have been shown to be involved in Xist regulation through different mechanisms, including transcriptional interference, RNA-mediated recruitment of chromatin remodelers, and through transcription co-activation [45][46][47][48]. Xist encodes a 17 kb lncRNA (19 kb in human) that contains six repeat structures that play a crucial, sometimes redundant, role in Xist-mediated silencing as well as localization [49]. So far, most of the functional studies have been performed in mouse where deletions of the most 5′ located repeat A led to a silencing phenotype despite the fact that Xist spreading was unaffected. Several studies indicated that SHARP [SMRT (silencing mediator for retinoid or thyroid hormone receptors) and HDACs (histone deacetylases)-associated repressor protein], encoded by the SPEN (split ends) gene [also called SHARP or Mint (Msx2-interacting nuclear target protein)], is a crucial factor in the X inactivation process through interacting with the A repeat sequence and recruitment of several repressor complex members, such as nuclear receptor corepressor (NCoR), SMRT, and nucleosome remodeling deacetylase (NuRD) complexes [50][51][52][53][54][55] (see Table 1).
Table 1. Proteins and complexes involved in the regulation of X chromosome inactivation (XCI). The “Disease(s)” column indicates diseases caused by mutations in the XCI related genes/proteins described in this table. The functional link between these mutations and XCI remains to be investigated.
ACC: Adenoid cystic carcinoma; AEBP2: Adipocyte enhancer-binding protein 2; AKS: Au–Kline syndrome; AML: Acute myeloid leukemia; ASDs: Autism spectrum disorders; BC: Breast cancer; CC: Colon cancer; CBX2/4/6/7/8: Chromobox homolog 2/4/6/7/8; DD: Developmental disorder; DLBCL: Diffuse large B-cell lymphoma; DNMT3B: DNA methyltransferase 3B; DS-AMKL: Acute megakaryoblastic leukemia associated with Down syndrome; DSD: Disorders of sex development; EED: Embryonic ectoderm development; ESCC: Esophageal squamous cell carcinoma; ETP-ALL: Early T-cell precursor acute lymphoblastic leukaemia Early T-cell precursor acute lymphoblastic leukaemia; EZH2: Enhancer of zeste 2; FL: Follicular lymphoma; FSHD: Facioscapulohumeral dystrophy; GC: Gastric cancer; GPS2: G-protein pathway suppressor 2; HCC: Hepatocellular carcinoma; HD: Hirschsprung disease; HDAC3: Histone deacetylase 3; hnRNPK: Heterogeneous nuclear ribonucleoprotein K; ICF: Immunodeficiency, centromeric instability and facial anomalies; ID: Intellectual disability; JARID2: Jumanji and AT-rich interaction domain-containing 2; KLS: Kabuki-like syndrome; KS: Kabuki syndrome; MB: Medulloblastoma; MCL: Mantle cell lymphoma; MDS: Myelodysplastic syndromes; MF: Mycosis fungoides; MPN: myeloproliferative neoplasm; NCoR1/2: Nuclear receptor corepressor; ncPRC1: non-canonical PRC1 complex; NDDs: Neurodevelopmental disorders; OMZL: Ocular marginal zone lymphoma; OS: Okamoto syndrome; OSS: Osteosarcoma; PASC: Pancreatic adenosquamous carcinoma; PCGF1-6: PcG ring finger 1-6; PCGF3/5: PcG ring finger 3/5; PHC1-3: Polyhomeotic homolog 1-3; PM: Primary microcephaly; PR: Prostate cancer; PRC1: Polycomb repressive complex 1; PRC2: Polycomb repressive complex 2; PS: Pierpont syndrome; RBBP4/7: Retinoblastoma binding protein 4/7; RING1A/B: Really interesting new gene 1A/B; RYBP/YAF: RING1 And YY1 Binding Protein/YY1-associated factor; SCMH1/L2: Sex comb on midleg homolog 1/L2; SCZ: Schizophrenia; SHARP: SMRT (silencing mediator for retinoid or thyroid hormone receptors) and HDACs (histone deacetylases)-associated repressor protein; SMZL: Splenic marginal zone lymphoma; SUZ12: Suppressor of zeste 12; T-ALL: T-cell acute lymphoblastic leukemia; T-PLL: T-cell prolymphocytic leukemia; TBL1: Transducin β-like protein 1; TBLR1: Transducing β-like 1 (TBL1)-related protein; XCI: X chromosome inactivation; Xist: X inactive specific transcript; * NCoR2 is also known as SMRT (silencing mediator for retinoid or thyroid hormone receptors); ** EZH2 is also known as KMT6A (lysine (K) methyltransferase 6A).
SHARP is transiently enriched at the promoters and enhancers of genes that are subject to XCI and it recruits NCoR/SMRT complexes that contain HDACs, leading to histone deacetylation [55]. SHARP localization also shows overlap with NuRD complex members predominantly at promoters, and its action is only required during the initiation phase of XCI, as removal of SHARP after Xi is established has no effect [55][140]). As a consequence of the action of SHARP and its associated protein complexes, promoters and enhancers are deacetylated in a stepwise manner, paving the way for the action of the polycomb group (PcG) protein repressive complexes PRC1 and PRC2 that play a crucial role in the establishment and maintenance of the silent state of the Xi. PRC1 is a large multi-protein complex that is recruited to Xist through heterogeneous nuclear ribonucleoprotein K (hnRNPK) that acts as a bridge between PRC1 and Xist Repeat B and, to a lesser extent, Repeat C [65][66][77]. PRC1-directed deposition of monoubiquitination of K119 of histone H2A (H2A119ub1) is mediated by the core PRC1 complex member really interesting new gene 1 isoform A or B (RING1A/B) and, in turn, is recognized by PRC2 subunit jumanji and AT-rich interaction domain-containing 2 (JARID2) facilitating trimethylation on K27 of histone H3 (H3K27me3) by the enhancer of zeste 2/lysine (K) methyltransferase 6A (EZH2/KMT6A) [141][142][143]. Subsequently, PRC1 and PRC2 recruitment is re-enforced through the recruitment of PRC1 that recognizes the trimethylation of K27 of histone H3 (H3K27me3) through chromobox-containing protein (CBX), which further promotes H2AK119ub1 deposition, facilitating the spreading of silencing [144][145][146]. At a later stage of the XCI process, de novo DNA methyltransferases (DNMTs) are recruited to lock in the silent state through the deposition of DNA methylation at promoters and CpG islands (CGI). These studies highlight the concerted action of chromatin readers and writers directing the right order of epigenetic events required to establish the Xi that is propagated through a near infinite number of cell divisions.
The complete loss or alteration of the Xi is frequently observed in breast and ovarian cancers, amongst other types of cancer [147][148]. Initial studies showed that Xist/XIST RNA is essential for the initiation and establishment of XCI during development, but dispensable to maintain the Xi in female somatic cells [149][150]. Even so, more recent studies making use of more sensitive techniques detect the reactivation of X-linked genes upon nearly complete or partial Xist/XIST depletion. The human X chromosome codes for more than 900 coding genes [151], including several tumor suppressor genes and oncogenes [152][153]. Thus, gene dosage changes that are caused by potential reactivation or silencing of X-linked genes could be detrimental. So far, only one well documented study in mice revealed a clear causal relationship between Xist deletion in the hematopoietic lineage and high penetrance hematopoietic cancer [154].
In human, the absence of the Xi (Barr body) in female cancer cells and presence of multiple Xa’s have been frequently associated with different forms of cancer, such as breast cancer [38][40][44]. However, these events are primarily attributed to the loss of the Xi and duplication of the Xa due to chromosome segregation errors (see Figure 1) [38][40][44].
Epigenetic alterations that are caused by epigenetic erosion of the Xi have also been described. These erosion events affect histone modification, deposition, and DNA methylation, leading to the reactivation of X-linked genes in breast cancer cell lines and primary tumors [155]. Moreover, the Xi in female cancer genomes has been shown to accumulate more mutations than the autosomes in various cancer types, including medulloblastoma, breast cancer, glioblastoma, and acute myeloid leukemia (AML) [156]. Interestingly, recent studies suggest that high XIST expression levels correlate with a poor survival in various types of cancer [157]. Some of these studies propose that XIST acts as a competing endogenous RNA (ceRNA) [158][159], by depleting microRNAs. As a consequence, specific RNA targets cannot be degraded, which may lead to the dysregulation of downstream genes [160][161]. So far, both epigenetic and genetic changes have been observed in relation to the Xi of cancer cells, but whether these alterations are driving events that give a selective advantage to cancer cells is under debate. Nevertheless, evidence suggests that the Xi epigenetic status and XIST expression levels are potential cancer biomarkers as a readout for genomic instability or epigenomic changes. Therefore, understanding the factors and mechanisms that render and maintain the X chromosome inactive, both during embryonic development and in somatic cells during the maintenance phase of XCI, is of crucial importance.
The regulation of the X chromosome is controlled by chromatin modifiers that build up heterochromatin formation by deacetylating and methylating histone tails, finally leading to the DNA methylation of regulatory CpG islands (see Figure 2).
Specific enzymes that play a central role in XCI are HDACs, the PRC1 and PRC2 complexes, and DNMTs (see Table 1). Recently, the SHARP protein has been identified as a direct Xist interactor. This protein bridges Xist to HDACs allowing for histone deacetylation at the X chromosome.