Encyclopedia
Please note this is an old version of this entry, which may differ significantly from the current revision.
Subjects:
Oncology
Ovarian cancer (OC) is the fifth leading cause of women’s death from cancers. The high mortality rate is attributed to the late presence of the disease and the lack of modern diagnostic tools, including molecular biomarkers. Moreover, OC is a highly heterogeneous disease, which contributes to early treatment failure. Thus, exploring OC molecular mechanisms could significantly enhance the understanding of the disease and provide new treatment options. Chromatin remodeling complexes (CRCs) are ATP-dependent molecular machines responsible for chromatin reorganization and involved in many DNA-related processes, including transcriptional regulation, replication, and reparation. Dysregulation of chromatin remodeling machinery may be related to cancer development and chemoresistance in OC. Some forms of OC and other gynecologic diseases have been associated with mutations in specific CRC genes. Most notably, ARID1A in endometriosis-related OC, SMARCA4, and SMARCB1 in hypercalcemic type small cell ovarian carcinoma (SCCOHT), ACTL6A, CHRAC1, RSF1 amplification in high-grade serous OC.
- chromatin remodeling complexes
- ovarian cancer
- ARID1A
1. Introduction
1.1. Ovarian Cancer Types and Mutations
Ovarian cancer (OC) is the third most common gynecologic malignancy and the second deadliest oncogynecologic disease in the world [1]. It is estimated that just in the United States, more than 12.8 thousand women will die from OC this year [2]. Early-stage OC presents a good diagnosis and a more than 90% 5-year survival rate. However, about 80% of OC patients are diagnosed with stage III-IV disease, where survival, even after therapy, is only 30% [3]. The high OC frequency and mortality rates are attributed to the lack of specific symptoms and insufficiency of the main OC blood biomarker CA125 to reduce mortality. Due to insufficient prognostic power, the CA125 serum biomarker is recommended as a prognostic tool only in high-grade ovarian cancer (HGSOC) patients [4]. Currently, there are no sensitive molecular biomarkers of clinical value for OC screening and patient management. Generally, after diagnosis, OC patients are treated by bilateral salpingo-oophorectomy (surgical removal of ovaries and fallopian tubes) with platinum or taxane therapy. Other therapies are available only after chemoresistance development which almost inevitably occurs in most HGSOC cases [5].
One of the culprits of OC chemoresistance and early therapy failure is the heterogeneity of the disease. Histologically, OCs are subdivided into epithelial and non-epithelial tumors. The non-epithelial OC only comprises 10% of OC cases. These are malignant germ cell tumors (e.g., small cell carcinoma of the ovary) that are usually present in preadolescent women, and sex cord-stromal tumors (granulosa cell tumors, fibroma, Sertoli-Leydig cell tumors, and others) that are more common and typically affect postmenopausal women [6]. Due to its rarity, non-epithelial tumors are rarely studied, and most research focuses on epithelial OC.
Epithelial type of OC is further subdivided into five major types according to histologic differences: the endometrioid, clear cell, mucinous, low-grade, and high-grade serous OCs [7]. According to the dualistic ovarian cancer model, epithelial OCs are grouped into two major types based on the site of origin of the tumor [8]. Type II tumors comprise mainly HGSOC but can also include carcinosarcomas and undifferentiated carcinomas. Unique for its high frequency of TP53 mutations and chromosomal instability, these tumors arise from the intraepithelial carcinoma in the fallopian tube. However, other origins, such as solid pseudoendometrioid transition tumors or borderline or low-grade serous OC, are possible. All other epithelial tumors (low-grade serous, mucinous, clear cell, borderline, seromucinous, and endometrioid) are considered less of a concern due to their slow-growing nature. Although borderline, mucinous, and seromucinous tumor origin is unclear, the endometroid and clear cell tumors are thought to originate from endometriosis and are similar to endometrial cancers [9]. Albeit highly heterogeneous, mutations in PIK3CA, PTEN, KRAS, BRAF, CTNNB1, and ARID1A genes are frequent amongst type I epithelial OCs and are considered potential OC biomarkers [10].
Regardless of type, a third of OC cases are related to homologous recombination gene insufficiency [5]. About 23% of OC cases are inherited. 10–15% of inherited cases are associated with BRCA1 and BRCA2 germline mutations [11,12]. The rest of the inherited OC cases are caused by TP53 mutations due to patients with the Li-Fraumeni syndrome [13]. The second gene group, most often mutated in OC, are the genes encoding proteins that are involved in chromatin remodeling complexes (CRCs) which are crucial for transcription, DNA repair, and most other cell day-to-day fluctuations in chromatin availability states.
1.2. Chromatin Remodelers and Their Functions in Human Cancer
Inside every human cell’s nucleus fits the genome comprising three billion DNA base pairs, almost two meters in length. This is achieved by winding the DNA around histone protein complexes forming nucleosomes that are the building blocks of the chromatin nucleoprotein structure. Each nucleosome has an octamer core made of H2A, H2B, H3, and H4 histone protein pairs and fits 147 bp of DNA. The nucleosomes are connected with linker DNA of 10–60 bp in length [14] and form the chromatin complex, a functionally dynamic structure. During interphase, chromatin hierarchically organizes and compartmentalizes DNA into active (euchromatin) and silent (heterochromatin) regions in the nucleus [15]. The distinct chromatin states are achieved through epigenetic modifications of the DNA, histone N-terminal tails, and nucleosome management done by ATP-dependent chromatin remodeling protein complexes. CRCs aid in nucleosome assembly and maintain dynamic chromatin activity through sliding, rearranging, evicting, and modifying nucleosomes. CRCs achieve these complex functions using their ATPase domain to translocate DNA in regard to histones [16]. Chromatin remodelers are involved in most essential cell processes, most notably regulating RNA pol II activity, gene silencing, homologous recombination (HR), and DNA repair [17].
There are four conserved families of ATP-dependent chromatin remodelers in mammals: chromodomain helicase DNA binding (CHD), imitation switch (ISWI), inositol requiring 80 (INO80), and switch/sucrose non-fermenting (SWI/SNF) [16]. The multiprotein complexes that remodel chromatin all share a common core ATPase-translocase subunit from the RNA/DNA helicase superfamily 2 (Snf2), but differ in accessory proteins that read epigenetic modifications and regulate the enzymatic activity of the complexes. SWI/SNF complexes have bromodomains capable of reading acetylated lysine on histone tails and function as chromatin accessibility regulators by sliding and evicting nucleosomes. ISWI reader domain (C-terminal HAND-SANT-SLIDE) binds to unmodified H4 tails and promotes chromatin assembly as well as proper nucleosome spacing. CHD remodelers are known to have similar functions to ISWI, although they contain two tandem chromodomains that read H3K4me3 active chromatin marks and are also known to bind to AT-rich DNA motifs. INO80 family remodelers are known to act in recombination and DNA replication as it is capable of binding to specialized DNA structures. In addition to nucleosome spacing, the INO80 family facilitates histone exchange by replacing nucleosome dimers and regulating histone variants [18].
Traditionally, SWI/SNF chromatin remodelers are associated with increasing chromatin accessibility, INO80 mainly acts as a chromatin repressor, and ISWI and CHD are most often described as histone chaperones in chromatin assembly [19]. However, these broad classifications are not fixed, as particular chromatin remodelers can be classified into two groups based on the chromatin regions they occupy. BRG1 (SWI/SNF), SNF2H (ISWI), CHD3, and CHD4 proteins are mainly found in active chromatin regions, while BRM (SWI/SNF), INO80, SNF2L (ISWI), CHD1 are associated with inactive chromatin [20].
2. Alterations of Chromatin Remodeler Complexes in Ovarian Cancer
2.1. ARID1A Alterations in Ovarian Cancer
About 25% of human cancers harbor mutations in at least one of 29 genes encoding SWI/SNF proteins [41]. Among them, ARID1A (BAF250a, B120, C1orf4, Osa1) is the most frequently altered. ARID1A mutations are particularly prevalent in gynecologic cancers (found in 10–60% of ovarian and endometrioid carcinoma cases) [42] and pre-malignant gynecological lesions, especially of endometrioid origin [43]. The highest ARID1A mutation rates are found in ovarian clear cell carcinoma (OCCC) and endometroid ovarian carcinoma suggesting its role in type I OC development and the potential for ARID1A’s use as a gynecologic cancer biomarker.
None of the previous research found any sufficient correlation between ARID1A protein expression loss and clinical features such as age, FIGO grade, and disease survival [60]. In a subset of microsatellite stable (MSS) endometrial cancer cases, ARID1A’s loss has been associated with a better prognosis and indicated as a potential prognostic biomarker [61]. Further investigation is highly needed to determine ARID1A’s potential as a gynecologic cancer biomarker.
Usually, ARID1A alterations are loss-of-function mutations such as large deletions, frameshift, or nonsense mutations that lead to the largest BAF complex subunit loss and inactivation of the SWI/SNF complex. However, ARID1A mutations are not sufficient to promote tumorigenesis alone. Tumors with ARID1A loss often harbor mutations of the PI3K/AKT pathway, such as inactivating PTEN or activating PIK3CA mutations. In mice models, separate ARID1A or PI3K/AKT pathway mutations promoted ovarian hyperplasia, while concomitant inactivation of ARID1A together with PIK3CA activating mutations induced tumors, similar to ovarian clear cell carcinoma (OCCC) [62].
ARID1A downregulation due to inactivating mutations has many downstream effects through epigenetic mark displacement, deregulating gene expression, and reduced protein interactions, primarily affecting DNA damage repair and signaling pathways.
SWI/SNF complex regulates active H3K27ac histone mark at enhancer and promoter regions through interaction with p300/CBR histone acetyltransferase [63]. The acetylation of +1 nucleosome may be required for normal RNA polymerase II (RNAPII) pausing, a crucial step in effective transcription initiation [64]. Thus, ARID1A’s downregulation results in dysregulated expression of at least 99 target genes [82]. The impairment of ARID1A’s function due to inactivating mutations is not fully compensated by its homolog ARID1B. Most notably, one of the affected genes is tumor suppressor TP53 [64]. ARID1A and TP53 mutations are mutually exclusive in gynecologic malignancies, however, p53 is indirectly regulated by ARID1A. ARID1A is an HDAC6 deacetylase transcriptional repressor. Thus, the loss of ARID1A leads to HDAC6 reactivation, which causes p53 lysine 120 residue deacetylation and apoptosis inhibition [65].
ARID1A’s deficiency has also been linked with telomere impairment. In ARID1A-deficient cells, topoisomerase IIα(TOP2A) cannot interact with the SWI/SNF ATPase BRG1 (SMARCA4), which it needs to resolve catenanes that develop during replication and transcription [69].
ARID1A’s mutational status in OC cells is also associated with reactive oxygen species (ROS) formation. ARID1A knockdown in OC cell lines increased intracellular ROS levels, as well as increased oxidative stress marker 8-hydrixyguanosine levels in OCCC patients with low ARID1A expression. Moreover, ARID1A-mutated cell lines were sensitive to ROS inductor elesclomol [76].
Telomere defects and DNA damage caused by ARID1A insufficiency leads to increased reliance on double-stranded DNA break (DSB) repair mechanisms [68]. Inactivation or loss of ARID1A or SWI/SNF complexes increases cell sensitivity to cisplatin and UV treatment through the impairment of multiple DSB repair mechanisms [74]. Most notably, ARID1A is involved in HR through DNA end processing (RPA and RAD51 loading), ATR activation, and G2-M cell-cycle arrest maintenance [72]. BAF factors, particularly ARID1A/B, are required for KU70/KU80 protein recruitment to DSB sites during NHEJ [73] as well as XPA accumulation at UV damage sites during NER (nucleotide excision repair) [74].
In ovarian and endometrial cancers, ARID1A mutations often co-occur with PI3K/AKT pathway gene KRAS, PIK3CA, and PTEN alterations [77]. Additionally, ARID1A’s loss activates the pathway by ANXA1 (AKT activator) upregulation and PIK3IP1 (PI3K inhibitor) expression downregulation. In particular, PIK3IP1 is regulated by ARID1A, suppressing EZH2 methyltransferase that inhibits PIK3IP1 expression through the H3K27me3 epigenetic mark [83].
Besides its role in SWI/SNF, ARID members have also been reported to co-precipitate with members of E3 ubiquitin ligase. Specifically, ARID1B is linked with elongin C (EloC) through BC box motif and with the addition of cullin 2 and ROC1 form E3 ubiquitin ligase that targets H2B histone explicitly at lysine 120 for monoubiquitination. ARID1B is a paralog that shares most of its sequence with ARID1A. Thus, a similar E3 ubiquitin ligase complex may be formed with ARID1A too. Mutation and depletion of the ARID domain result in decreased ubiquitination of H2B, similar to VHL mutations, causing decreased ubiquitination of HIF1α in clear cell renal cell carcinoma (ccRCC) [80]. The two genes are often mutated together in ccRCC. The reduced H2B ubiquitination is responsible for reduced H3 histone lysine 79 di-methylation and decreased gene expression [84].
2.2. Other SWI/SNF Alterations in Ovarian Cancer
Although ARID1A is the most frequently mutated SWI/SNF subunit coding gene, other SWI/SNF members, albeit less often, are also altered in OC. In particular, up to 90% of a rare form of OC, small cell ovarian carcinoma, hypercalcemic type (SCCOHT), a rhabdoid-like tumor has germline or somatic mutations of SWI/SNF ATPase domain BRG1 coding gene SMARCA4 [90]. Just like ARID1A, the SMARCA4 is not only involved in gene expression through the primary SWI/SNF function as chromatin remodeler (SMARCA4 mutations cause hyperactivation of PRC2 repressive complex) [43], but it also directly participates in DNA repair. BRG1 (SMARCA4) is recruited to DNA damage response (DDR) sites by PARP1 and is activated by deacetylation by SIRT1 to open the affected chromatin region in preparation for HR [91]. Moreover, BRG1 has been associated with γ-H2AX (a histone mark that indicates damaged DNA) formation, thus likely to promote DSB [92].
Meanwhile, other forms of OC, such as undifferentiated uterine or ovarian carcinoma and OCCC, are found to harbor alterations in other SMARC genes [93]. Most notably, other SMARC gene mutations are found in OCCC: SMARCA4 in up to 5%, SMARCA1 at 2%, SMARCA2 at 1%, and SMARCC1 at 2% [94]. Although SMARCA2 mutations are rare, most SCCOHT cases present with dual inactivation of both BGR1 and BRM, though, unlike SMARCA4, SMARCA2 is silenced epigenetically [95]. Similar to other rhabdoid tumors, SCCOHT may also harbor SWI/SNF core subunit SMARCB1 (also known as INI1/SNF5/BAF47) mutations instead of SMARCA4 [90].
2.3. ISWI Alterations in Ovarian Cancer
Similarly, to SWI/SNF, ISWI ATPase domains are also encoded by SMARC genes that frequent alterations in OC. Most notably, ISWI ATPase SNF2L (SMARCA1) mutations have been linked with OCCC [94]. Although SNF2H is the dominant ISWI ATPase, SNF2L has been described as a regulator of specific gene expression during differentiation. Crucially, SNF2L has been associated with ovarian development and meiotic progression of germ cells during its differentiation [100]. SNF2L interacts with progesterone receptor and steroidogenic acute regulatory protein in ovarian granulosa cells to promote differentiation of the ovary [101]. SMARCA1 deletion has been associated with increased apoptosis through caspase activator Apaf-1 expression upregulation [102] and enhanced proliferation and migration due to WNT signaling regulation in various cancer cells [103].
Depending on which ATPase it binds, ISWI protein remodeling and spacing factor 1 (coded by RSF1, also known as BAZ1A, HBXAP), forms either RSF-1 or RSF-5 CRC [27]. Immunohistochemical analysis showed that ISWI ATPase SNF2H (SMARCA5) is overexpressed in OC tissues together with its binding partner RSF1 which is also often upregulated in various tumors, including OC [104]. RSF1 is an essential interphase centromere protein that maintains chromosome stability and protein homeostasis but also has roles in DSB and transcription regulator functions through its interactions with HDAC1, CENP-A, ATM, SNF2H, cyclin E1, CBP, NF-κB, BubR1, and many other cancer-related proteins [105]. The RSF-1 complex protein expression correlates with cancer stage and poor clinical outcomes in OC patients [106]. The upregulation of RSF-1 and SMARCA5 expression leads to increased double-stranded breaks (DSB) and subsequent DDR [107].
2.4. CHD Family Alterations in Ovarian Cancer
CHD5 is the main CHD family protein linked with cancer [34]. Although mutations in the CHD5 gene are detected in OC, CHD5 is also downregulated by increased promoter methylation in OC patient samples [113]. In one study analyzing CHD gene promoter methylation in various cancer cell lines, out of all CHD genes, CHD5 promoter was methylated the most often, although the OC cell line (MDAH2774) showed no increase in methylation [114]. CHD5 polymorphism (rs9434741) has also been associated with endometriosis, an OC precursor [115], however, the clinical significance of this variant is not known. In alignment with CHD5 increased mutation and methylation, CHD5 mRNA expression was found to be downregulated by at least 2-fold in 41% (32/72) of invasive epithelial ovarian carcinomas in comparison with 12 controls and correlated with shorter disease-free survival times [116].
2.5. INO80 in Ovarian Cancer
Although several INO80 subunits are overexpressed in melanoma [122], cervical [123], and non-small cell lung cancer [124], according to TCGA data analysis [125,126] the only INO80 gene significantly amplified in OC patients is ACT6LA, shared with SWI/SNF family remodelers. ACT6LA amplification affects multiple CRCs and causes platinum resistance in OC [99]. However, other INO80 family domains are found to be significantly altered in OC cell models. TRRAP (transformation/transcription domain-associated protein) — a TIP60 remodeler/histone acetyltransferase complex adaptor protein is overexpressed in OC A2780 sphere cultures and was found to govern stemness marker NANOG expression and OC cell proliferation [127]. Yin yang 1 (YY1) — a non-conserved human INO80 family subunit [38] in OC cell line models, has been associated with chemoresistance through upregulation of lncRNA PART1, which targets miR-512-3p and causes CHRAC1 (another CRC gene) upregulation and cisplatin-resistant OC cell proliferation and migration [111]. In OC patients and cell lines YY1 itself was found to be suppressed by miR-381, which is downregulated in OC tissues [128].
This entry is adapted from the peer-reviewed paper 10.3390/ijms232213670
This entry is offline, you can click here to edit this entry!
