The roles of mutated SWI/SNF complexes in the initiation and development of hepatocellular carcinoma and its regulatory effect on the immune system: A review

Abstract Hepatocellular carcinoma (HCC) is a primary liver malignancy with a high global prevalence and a dismal prognosis. Studies are urgently needed to examine the molecular pathogenesis and biological characteristics of HCC. Chromatin remodelling, an integral component of the DNA damage response, protects against DNA damage‐induced genome instability and tumorigenesis by triggering the signalling events that activate the interconnected DNA repair pathways. The SWI/SNF complexes are one of the most extensively investigated adenosine triphosphate‐dependent chromatin remodelling complexes, and mutations in genes encoding SWI/SNF subunits are frequently observed in various human cancers, including HCC. The mutated SWI/SNF complex subunits exert dual functions by accelerating or inhibiting HCC initiation and progression. Furthermore, the abnormal SWI/SNF complexes influence the transcription of interferon‐stimulated genes, as well as the differentiation, activation and recruitment of several immune cell types. In addition, they exhibit synergistic effects with immune checkpoint inhibitors in the treatment of diverse tumour types. Therefore, understanding the mutations and deficiencies of the SMI/SNF complexes, together with the associated functional mechanisms, may provide a novel strategy to treat HCC through targeting the related genes or modulating the tumour microenvironment.


| BACKG ROU N D
Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related deaths worldwide, afflicting approximately 800 000 people annually. 1 The pathogenesis of HCC is extremely complicated, involving processes such as cell cycle regulation and signal transduction, and it reflects the functions and interactions of multiple genes at multiple steps. 2 HCC is currently treated by surgical resection and chemotherapy, but the mortality rate of this cancer remains high. In recent years, the use of immune checkpoint inhibitors (ICIs), such as ipilimumab (the inhibitor of cytotoxic T-lymphocyte antigen 4 [CTLA4]) and nivolumab (the inhibitor of programmed death-1 [PD-1]), have demonstrated survival benefits for HCC, which reveals that immune status is closely related to HCC progression. 3,4 Therefore, it is important to further explore the characteristics of HCC and to develop novel therapies for treating this cancer.
The mating-type switch/sucrose non-fermenting (SWI/SNF) complexes, which are capable of regulating gene transcription through induced genome instability and tumorigenesis by triggering the signalling events that activate the interconnected DNA repair pathways. The SWI/SNF complexes are one of the most extensively investigated adenosine triphosphate-dependent chromatin remodelling complexes, and mutations in genes encoding SWI/SNF subunits are frequently observed in various human cancers, including HCC. The mutated SWI/ SNF complex subunits exert dual functions by accelerating or inhibiting HCC initiation and progression. Furthermore, the abnormal SWI/SNF complexes influence the transcription of interferon-stimulated genes, as well as the differentiation, activation and recruitment of several immune cell types. In addition, they exhibit synergistic effects with immune checkpoint inhibitors in the treatment of diverse tumour types. Therefore, understanding the mutations and deficiencies of the SMI/SNF complexes, together with the associated functional mechanisms, may provide a novel strategy to treat HCC through targeting the related genes or modulating the tumour microenvironment.
adenosine triphosphate (ATP)-dependent nucleosome remodelling, have been shown to play a widespread role in carcinogenesis. 5 The SWI/SNF complexes are macromolecular complexes comprising [12][13][14][15] subunits, including a catalytic ATPase subunit, SWI/SNF related, matrix associated, actin-dependent regulator of chromatin, subfamily a, member 4 (SMARCA4)/brahma-related gene 1 (BRG1) or SMARCA2/ brahma (BRM); and several core subunits, such as SMARCB1/SNF5/ INI1/BAF47 or SMARCC1/BAF155. Other subunits, such as AT-rich interaction domain 1A (ARID1A) and ARID1B, are the mutually exclusive components of the BRG1-associated factor (BAF) complexes, while polybromo 1 (PBRM1) and ARID2 are specific for the polybromo BAF (PBAF) complexes. 6 Recently, a third SWI/SNF complex called non-canonical BAF (ncBAF, also termed GBAF) has been identified, which contains glioma tumour suppressor candidate region gene 1 (GLTSCR1)/ GLTSCR1-like (GLTSCR1L) subunits instead of the ARID-domain containing proteins. 7 Global genomic analyses suggest that the SWI/SNF complexes are associated with a mutation rate of 20% among all human tumours. 8 Frequent inactivating mutations in SWI/SNF subunits, such as ARID1A, ARID1B, ARID2, PBRM1 and SMARCA4, are repeatedly detected in numerous cancers (Figures 1 and 2). [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23] The mutation proportion and types of different SWI/SNF complex subunits in several HCC data sets are summarized in Table 1. In addition, increasing evidence indicates that the SWI/SNF complexes directly interact with numerous important proteins, such as beta-catenin, to modulate cancer formation. 24 The SWI/SNF complexes not only exert a vital part in transcription activation but also participate in transcriptional repression. 25 Furthermore, an increasing number of studies reveal that the ATPdependent chromatin remodelling complexes, like SWI/SNF, exhibit significant impacts on human immunodeficiency virus type 1 (HIV-1), 26 autoimmune reactivity 27 and hypersensitivity. 28 Nonetheless, the role of SWI/SNF complexes in HCC occurrence and development as well as the immune system has not yet been fully elucidated. This review integrates the available data on the role of aberrant SWI/SNF complexes during HCC progression and examines its regulatory effects on several immune cell types and ICI therapies.

| AB ERR ANT BAF COMPLE XE S PL AY D IVER S E ROLE S IN H CC DE VELOPMENT
In mammals, BAF complexes are composed of a single central ATPase, either SMARCA4/BRG1 or SMARCA2/BRM, and several BRG-/ F I G U R E 1 High-frequency gene mutation of BAF in different tumours. The sources are shown in brackets. BAF, BRG1-associated factor F I G U R E 2 High-frequency gene mutation of PBAF in different tumours. The sources are shown in brackets. PBAF, polybromo BAF BRM-associated factors. 29 In addition to the subunits homologous to those in Drosophila or yeast, several other subunits appear to be dedicated to vertebrate or mammalian complexes, including SS18/ SS18L1, BCL7A/B/C and BCL11/A/B. BAF complexes have been implicated in embryonic development. 30 The presence of the BAF complex on chromatin correlates with that of enhancers, 31 and its activity regulates a variety of important biological processes ranging from self-renewal and pluripotency in embryonic stem cells 32 to cardiac development. 33 A previous study revealed that cancers with a perturbation in BAF components were dependent on ncBAF, 7 highlighting a synthetic lethal relationship between the two distinct SWI/SNF complexes as opposed to paralogs of the same complex.
Research on the role of ncBAF in the development of HCC is still rare, although BAF subunits are frequently disrupted in a variety of different malignancies including HCC.

| SMARCA4/BRG1
SMARCA4 is located in chromosome 19p13.2, and its protein functions as the central catalytic component in the SWI/SNF complex.
Specifically, the SMARCA4 protein comprises multiple domains, including an evolutionarily conserved catalytic ATPase domain, a conserved C-terminal bromodomain, AT-hook motif and the less characterized N-terminal region, all of which play critical roles in modified histone protein recognition, DNA binding or SWI/ SNF recruitment. [34][35][36][37] SMARCA4 has been demonstrated to interact with diverse nuclear proteins involved in various cellular processes, such as transcriptional regulation, cell cycle control, proliferation, DNA repair and recombination. 38 However, the role of SMARCA4 in cancer occurrence and progression remains unclear. It is worth noting that SMARCA4 has been proposed to act as a tumour suppressor gene through diverse and c-Myc. 43,44 Paradoxically, SMARCA4 is also reported to be mu- In the context of HCC, Zhong et al 53 showed that the SMARCA4 single nucleotide polymorphism rs11879293, which was located in the intron between exon 1 and exon 2 of SMARCA4 and was essential for the interaction between SMARCA4 and SS18L1/CREST, showed a remarkable association with a reduced risk of HCC, as verified in stage 2 combined analysis. The mechanism underlying this association may be that rs11879293 acts as an intronic enhancer to alter the pathway, as well as the lipopolysaccharide (LPS)-induced pro-inflammatory mediators IL-6 and C-X-C motif chemokine 8 (CXCL8), may be involved in hepatocarcinogenesis. 54,55 In more recent studies, when copy number analysis was used in combination with expression profiling to identify cancer-associated mutations, the results revealed that SMARCA4 is up-regulated in HCC and that its level is markedly correlated with cancer progression among HCC patients. 56 Additionally, the nuclear expression of SMARCA4 also predicts the early recurrence of HCC in affected patients. Furthermore, SMARCA4, which is found to facilitate S-phase entry and attenuate apoptosis, also promotes cell proliferation through the up-regulation of SMAD6. On the other hand, it has been shown that SMARCA4 aggravates liver fibrosis through regulating the activation of hepatic stellate cells (HSCs) via the transforming growth factor beta (TGFβ)/SMAD signalling pathway. 57

| SMARCA2/BRM
SMARCA2, a homolog with 75% identity to SMARCA4, also regulates chromatin structure, but it is mutually exclusive of SMARCA4 in the SWI/SNF complexes. 59  and SMARCA2 also show differential expression patterns during development. 62 SMARCA4 tends to be highly expressed in proliferating cells, whereas SMARCA2 is mainly expressed in cells that cycle slowly (such as stem cells) and non-cycling differentiated cells. 63 In contrast, using data from loss-of-function screening of 165 cancer cell lines, Boris et al identified that SMARCA2 was an essential gene in SMARCA4 mutant cancer cell lines. They note that SMARCA4 inactivation resulted in greater incorporation of non-essential SMARCA2 subunits into the SWI/SNF complexes, revealing a role for SMARCA2 in oncogenesis induced by SMARCA4 loss, and they identified that ATPase and bromodomain-containing SMARCA2 may serve as potential therapeutic targets in these cancers. The hypothetical models depicting the function of residual SMARCA2-SWI/SNF complexes in SMARCA4-mutant cancers are exhibited in Figure 3.
Such findings suggest that the relationship between SMARCA2 and SMARCA4 may be complicated and regulated by different mechanisms; in addition, the specific role of mutated SMARCA2 in HCC may be transformed due to changes in cancer cell characteristics and the surrounding environment. Nevertheless, more studies investigating the association of SMARCA2 with SMARCA4 and the role of the former in HCC development should be performed.

| SMARCB1/BAF47
SMARCB1, a core subunit of the SWI/SNF complex, has been reported to modulate cell proliferation and apoptosis. 64 and inhibits HCC cell proliferation and migration by suppressing the lncRNA MVIH. These findings revealed the potential association of ARID1A with lncRNAs and also supported a role for ARID1A expression in HCC development. 83 In addition, the nuclear expression of p53 or beta-catenin also showed an inverse correlation with altered

| ARID1B/BAF250B
ARID1B, an isoform that is mutually exclusive with ARID1A in the SWI/SNF complexes and that is involved in regulating transcription and multiple downstream cellular processes, has been recently identified to be the primary mutant gene in various cancers. 85,86 Previous studies have shown that ARID1B gene mutations are responsible for neurodevelopmental retardation, intellectual disability, growth delay and dysmorphic features. 87 Interestingly, high ARID1B expression is associated with poor outcomes in bladder urothelial carcinoma and it also predicts the benefits of adjuvant chemotherapy for this cancer. 86

| BAF60a and SMARCC1/BAF155
Hepatic BAF60a, which serves as a linker between the SWI/SNF core

| CONTRIBUTI ON S TO ON COG ENE S IS
Although the potential mechanisms through which SWI/SNF complexes inhibit or promote HCC have been described in the different subsections above, insight into the specific mechanisms underlying tumour suppression or promotion and the reasons for the differing cancer spectrum associated with each subunit is still in its infancy.
In addition to a role in transcriptional regulation, several avenues of research have linked the SWI/SNF complex to DNA repair including nucleotide excision repair, double-strand break repair and DNA decatenation. [106][107][108] This begs the question as to whether the tumour suppressor activity of the complex arises via a role in controlling transcriptional programmes or whether it is derived from a role for the complex in protecting genome integrity. Mutations in some subunits, such as those in ARID2 mentioned earlier, may cause hepatocarcinogenesis associated with DNA damage and repair, while deletions in other subunits, such as SMARCB1, may not cause cancer via defects in DNA repair but rather due to epigenetic alterations such as disruption of chromatin-based contributions to the control of cell fate. 58 Of particular interest is the finding that mutations in more than one SWI/SNF subunit gene can occur in primary tumours, perhaps reflecting both haploinsufficiency and compound heterozygous effects. 109,110 Whether the mutations of the SWI/SNF subunit genes drive or repress tumours, especially HCC, via an epigenetic mechanism or a genetic mechanism remains to be determined.

| TARG E TING TUMOUR S WITH BAF/ PBAF DEFI CIEN CIE S
One anticancer drug discovery strategy, which reveals great promise in targeting cancer cells with genetic mutations, is the exploitation of synthetic lethality. 111  cence. Recent studies also reported that inhibition of enhancer of zeste homolog 2 (EZH2), the catalytic subunit of PRC2, caused synthetic lethality in ARID1A-mutated cancers. 113 Unfortunately, epigenetic drugs (such as EZH2 antagonists) did not exhibit synthetic lethality in ARID2-disrupted HCC, while exposure of ultraviolet irradiation and chemical compounds regulating the DNA repair system will be the effective treatment. 100 Additionally, the loss of function of BAF/PBAF subunits may lead to increased Polycomb activity and inhibition of Polycomb chromatin silencing may therefore be beneficial for patients with HCC bearing BAF/PBAF deficiencies. 58 Nevertheless, the effectiveness of these approaches may depend greatly on the downstream consequences of BAF/PBAF dysfunction within each cell type. partially activated by persistent and severe inflammation, this activation was insufficient to stop the inflammation. As mentioned above, SMARCA4 is more inclined to accelerate tumour development in the context of HCC and we speculate that this may be partly due to its effect in promoting immune tolerance.

| Roles of the SWI/SNF complexes in modulating the immune system
With respect to the other subunits, RNAi screening revealed that ARID1A knock-down promotes the resistance of Jurkat leukaemia cells to Fas (CD95)-mediated apoptosis, a cell-killing mechanism adopted by T cells and NK cells. 119,120 The association between ARID1A mutations and cancer subtypes with profound lymphocytic infiltration raises the possibility that ARID1A mutations may boost the ability of cancer cells to escape from immune surveillance. In addition, a study on ovarian cancer using differential expression analysis discovered that SMARCE1 mRNA levels are closely correlated with the number of intra-tumoral CD8 + cells. 121  and STATs. 122,123 The lack of an IFN-γ signature is suggested to be correlated with resistance to immunotherapy. 124 In addition to enhancing tumour cell immunogenicity through up-regulating the expression of tumour antigen processing and presentation, ISGs also promote programmed cell death and anti-proliferative effects. 125 Moreover, STAT proteins that bind to ISG promoters are regulated by the multi-protein complexes responsible for chromatin remod- It has recently been suggested that EZH2 inhibition enhances ICI efficacy in melanoma mouse models, and together, these findings pave the way for clinical trials that integrate small-molecule inhibitors of the chromatin remodelling pathways with ICIs as a novel synergistic treatment combination.

| Mutated SWI/SNF complexes shape the response of immunotherapy in different tumours
As suggested in recent literature, compared with SWI/SNF wildtype colorectal cancer (CRC), tumours harbouring SWI/SNF gene mutations show dramatically higher rates of microsatellite instability (MSI)-high, tumour mutational burden (TMB)-high and PD-L1 positivity, exhibiting a close correlation with the immune profile. 130 It has also been demonstrated that this subgroup of CRCs may possess a higher capacity to respond to monoclonal antibodies targeting PD-1 and PD-L1, further elucidating the relationship between the mutated SWI/SNF complexes and immunotherapy. On the other hand, mutation or low expression of the chromatin remodelling genes, including SMARCA2 and PBRM1, is reported to be associated with larger neoantigen burdens and a greater amount of activated CD8 + T-cell infiltration in human non-small-cell lung carcinoma. 131,132 Tumours with low PBRM1 expression also show a favourable prognosis and are connected to the increase in cytotoxic lymphocytes in their TME. 133 Furthermore, ARID1A deficiency is correlated with a microsatellite instability genomic signature, a F I G U R E 5 Loss of PBAF could improve the expression of ISGs. Loss of PBAF is proposed to alter chromatin structure so that IFN-g response elements in ISG promoters are more accessible to transcription factors, increasing their expression. When PBAF is intact, it might cooperate with EZH2 to modify chromatin and reduce the accessibility to IFN-g response elements. PBAF, polybromo BAF; ISGs, interferonstimulated genes predominant C>T mutation pattern and an increased mutation load across multiple human cancer types. 134

| Correlation of the SWI/SNF complexes with immunity in HCC
To date, only a few studies have reported the influence of the mutated SWI/SNF complexes on the immune microenvironment, as well as synergistic effects with immunotherapy among HCC patients. termates. Moreover, the synthesis of pro-inflammatory mediators was also down-regulated in primary hepatocytes isolated from the Smarca4-deficient mice relative to the WT mice, resulting in reduced macrophage chemotaxis. 137 The mechanism by which SMARCA4 contributes to the transcription of pro-inflammatory mediators is possibly through the regulation of the interaction between NF-κB and its co-factor myocardin-related transcription factor A (MRTF-A).
Non-alcoholic steatohepatitis (NASH) has been recognized as a major catalyst of HCC, and the hepatocyte-specific deletion of SMARCA4 relieves MCD-induced NASH in mice. [138][139][140] Nonetheless, more in-depth research is warranted to shed more light on how the regulation of aberrant SWI/SNF complexes affects the immune microenvironment of HCC and to lay a foundation for the novel immunotherapies. and multi-agent therapies are required due to the heterogeneity of HCC. Identifying such treatments will gain traction from the highthroughput screening of target molecules and NGS-based stratification of patients, to identify and explore novel, effective and more personalized therapeutic approaches.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R S ' CO NTR I B UTI O N S
XS, BH and JL created the idea for the review. BH performed the selection of literature, drafted the manuscript and prepared the figures. XY and JL revised the manuscript. All authors read and approved the final manuscript.

CO N S E NT FO R PU B LI C ATI O N
All authors agree to submit for consideration for publication in the journal.