Regulation of chromatin structure and function: insights into the histone chaperone FACT

ABSTRACT

In eukaryotic cells, changes in chromatin accessibility are necessary for chromatin to maintain its highly dynamic nature at different times during the cell cycle. Histone chaperones interact with histones and regulate chromatin dynamics. Facilitates chromatin transcription (FACT) is an important histone chaperone that plays crucial roles during various cellular processes. Here, we analyze the structural characteristics of FACT, discuss how FACT regulates nucleosome/chromatin reorganization and summarize possible functions of FACT in transcription, replication, and DNA repair. The possible involvement of FACT in cell fate determination is also discussed.

Abbreviations: FACT: facilitates chromatin transcription, Spt16: suppressor of Ty16, SSRP1: structure-specific recognition protein-1, NTD: N-terminal domain, DD: dimerization domain, MD: middle domain, CTD: C-terminus domain, IDD: internal intrinsically disordered domain, HMG: high mobility group, CID: C-terminal intrinsically disordered domain, Nhp6: non-histone chromosomal protein 6, RNAPII: RNA polymerase II, CK2: casein kinase 2, AID: acidic inner disorder, PIC: pre-initiation complex, IR: ionizing radiation, DDSB: DNA double-strand break, PARlation: poly ADP-ribosylation, BER: base-excision repair, UVSSA: UV-stimulated scaffold protein A, HR: homologous recombination, CAF-1: chromatin assembly factor 1, Asf1: anti-silencing factor 1, Rtt106: regulator of Ty1 transposition protein 106, H3K56ac: H3K56 acetylation, KD: knock down, SETD2: SET domain containing 2, H3K36me3: trimethylation of lysine36 in histone H3, H2Bub: H2B ubiquitination, iPSCs: induced pluripotent stem cells, ESC: embryonic stem cell, H3K4me3: trimethylation of lysine 4 on histone H3 protein subunit, CHD1: chromodomain protein

Graphical Abstract

KEYWORDS:

• Histone chaperone
• FACT
• chromatin dynamics
• cell fate

Background

Chromatin is a stable and highly dynamic nuclear protein complex. The fundamental unit of chromatin in eukaryotes is nucleosome, which consists of a histone octamer wrapped by approximately 146 bp of DNA fragments [1]. The dense structure of nucleosomes undeniably protects DNA from damage and at the same time acts as a barrier during cellular processes such as transcription, replication, and DNA repair. Thus, efficient chromatin remodeling is required in all chromatin-related processes [2–4]. Chromatin remodeling refers to a wide range of chromatin changes induced by chromatin-remodeling factors interacting with nucleosomes, which can be achieved by various means, including (i) ATP-dependent remodeling complexes, (ii) histone posttranslational modifications, (iii) histone chaperones, and (iv) the interplay of all three [5–9]. Histone chaperones play crucial roles in the correct remodeling of nucleosomes, first proposed by Ron Laskey [10] to describe an acidic nuclear protein in Xenopus laevis egg extracts, representing a group of proteins that prevent incorrect ionic interactions between histones and DNA. Histone chaperones participate in nucleosome assembly/disassembly without using ATP, and they are not permanent components of nucleosome structures [11]. Among them, facilitates chromatin transcription (FACT) is a highly conserved histone chaperone that was initially discovered to remove histone H2A-H2B dimer during transcription [12,13]. In fact, FACT is capable of binding to all core histones [14] and has been verified to be involved in important cellular processes, including transcription, DNA damage repair, DNA replication, and cell fate determination [15–17].

The structural features of FACT

Human (h) FACT is a heterodimer consisting of two subunits: Suppressor of Ty16 (Spt16) and Structure-Specific Recognition Protein-1 (SSRP1) [18,19]. The two subunits of hFACT are interdependent, and the stability of hFACT depends on the presence of the SSRP1 mRNA [20].

Spt16 is the large subunit of FACT, and has several domains with histone-binding activity: an N-Terminal Domain (NTD), a Dimerization Domain (DD), a Middle Domain (MD), and a C-Terminal Domain (CTD) (Figure 1) [21–23]. Spt16 is very well conserved in yeast and humans (Figure 1). The NTD of Spt16 is an inactive amino-peptidase-like motif [24]. It has been shown that the NTD has very weak affinity for the H2A-H2B dimer but interacts preferentially with the H3-H4 tetramer [24–26]. The DD of Spt16 interacts with the NTD/DD of SSRP1, and this interaction is crucial for formation of the FACT complex [27]. The MD of Spt16 is essential for recognition of histones H2A-H2B by FACT [28]. Moreover, Spt16 MD also contributes to FACT interaction with H3-H4 tetramers [28,29] and helps dissociate Spt16 from chromatin at the end of transcription [27]. The acidic region of the Spt16 CTD is a highly conserved domain that binds explicitly to the H2A-H2B dimer [30]. The CTD is involved in nucleosome reorganization, and its absence inhibits the activity of FACT [29,31].

Figure 1. Domain organization of hFACT and yFACT. The heterodimeric hFACT complex is composed of the hSpt16 (cyan) and SSRP1 (green) subunits, through an interaction (gray) between the DD domain of hSPT16 (cyan) and the NTD/DD domain of SSRP1. In yeast, ySpt16 (cyan) and Pob3 (blue) subunits form the yFACT complex. (3D structures are from the UniProt website)

hSSRP1 consists of five different domains: the NTD/DD, MD, internal intrinsically disorder domain (IDD),high mobility group (HMG) domain, and C-terminal Intrinsically disordered (CID) domain (Figure 1) [22,32]. The NTD/DD contains a pleckstrin homology (PH) motif that interacts with Spt16 [28]. The MD of SSRP1 does not bind histones, but binds DNA and functions like HMG [33]. The HMG domain has DNA binding activity [34–36] and stabilizes nucleosome structure by maintaining H3-H4 tetramer binding to DNA [37]. Mouse embryos lacking the HMG domain die before E5.5, suggesting that the HMG domain is indispensable for early embryonic development [36]. Unlike the HMG domain, which binds to bent or kinked DNA, the CID domain of hSSRP1 binds to Z-DNA [38,39]. It has been shown that both the CTD and HMG domains are essential for the nucleosome stabilizing activity of FACT [40].

The homologous protein of SSRP1 in yeast is yPob3 [41,42], which contains the NTD/DD, MD and CTD [22] and forms the yFACT complex with ySpt16 (Figure 1). yPob3 lacks the HMG domain; however, non-histone chromosomal protein 6 (Nhp6) encodes an HMG1-like motif and provides HMG1-like function for yFACT in yeast [59606443], enabling yFACT to remodel nucleosomes. It was reported that Nhp6 does not interact directly with yFACT but binds to DNA at both ends of the nucleosomes and bends the structure sharply, which induces a change in nucleosomes accessibility and provides access for yFACT binding [44–46]. In the absence of Nhp6, Spt16 binds to H2B and disrupts the interaction between the histone octamer and DNA, and removes the H2A-H2B dimer by gradually unwrapping/rewrapping DNA from the surface of the octamer [47]. In line with this evidence, the presence of Nhp6 prevents loss of H2A-H2B during FACT-dependent nucleosome remodeling [48].

Functions of FACT in chromatin homeostasis

FACT maintains chromatin homeostasis through its dual functions of assembling and disassembling nucleosomes [14,37]. It has been reported that while Spt16 displaces the H2A-H2B dimer and disassembles nucleosome structure, SSRP1 maintains the H3-H4 tetramer on DNA and promotes deposition of the H2A-H2B dimer [37].

Recent reports have shed new light on the mechanistic insight of FACT in maintaining chromatin homeostasis [14]. FACT seems to preferentially bind histone H2A-H2B, and the affinity between H2A-H2B and FACT is approximately 20-fold higher than that of FACT and H3-H4. The interaction between the CTD of Spt16 and H2A-H2B activates the DNA-binding activity of FACT [30]. In an in vitro nucleosome assembly assay, FACT does not bind to (H3-H4)₂ tetrasome in the absence of H2A–H2B dimer. However, when prebound with H2A–H2B, FACT binds to tetrasomes efficiently and forms intermediate complexes (FACT sub-nucleosome) [49]. Thus, it was inferred from the experiment that FACT interacts with H2A-H2B dimer to form H2A-H2B-FACT and uses its DNA-binding activity to dock onto the (H3-H4)₂-DNA complex to form the FACT-subnucleosome [49]. The structure of the FACT-subnucleosome looks like a unicycle: (H3-H4)₂ is the central part of the wheel; the H2A-H2B dimer is similar to a pedal, and FACT is similar to a frame across the wheel [14,49]. When the FACT-subnucleosome meets the second H2A-H2B, the MD of SSRP1 undergoes structural changes to facilitate docking of the second H2A-H2B onto the histone core and forms nucleosomes [14]. During transcription or DNA replication, DNA temporarily leaves its interaction interface with the (H3-H4)₂ tetramer. At the same time, the MD of Spt16 rotates a certain angle (23°) to change the binding modes with the (H3-H4)₂ tetramer. The rotation of the Spt16 MD makes its pleckstrin homology domain clash with the H2A docking domain and detach H2A-H2B from the nucleosome. However, the CTD tethers and maintains H2A-H2B nearby to ensure that all components of nucleosomes are intact during the passage of polymerases [14,50].

FACT in transcription

Transcriptional initiation requires a Pre-Initiation Complex (PIC) on promoter to facilitate DNA-binding activity of RNAPII [51]. FACT not only regulates recruitment of PIC components at promoters to facilitate assembly of the PIC but also reorganize nucleosomes to promote accessibility of transcription start sites (TSS) [52,53]. ChIP-seq analysis in mESCs indicated that both of the FACT subunits were strongly associated with actively expressed genes and enriched around the TSS [54]. Moreover, Spt16 was predominantly enriched in the open reading frame (ORF) in yeast [55,56]. Meanwhile, FACT shows the same localization and dynamics with the elongating RNAPII and has been proposed to promote transcriptional elongation [12,18,53,57,61–]. Genome-wide profiling data indicated that FACT is important for RNAPII-mediated transcription in vivo and likely removes nucleosome blocks ahead of RNAPII during transcription and assembles them after RNAPII passage [62]. However, high-resolution mapping of RNAPII traversal of nucleosomes demonstrated that RNAPII could efficiently overcome nucleosome barriers on individual nucleosomes or di-nucleosome substrates in vitro [63,65–]. Two types of working models, the “dimer eviction model” (Figure 2a) and the “non-eviction model” (Figure 2b), have been proposed to explain the probable mechanism of FACT-mediated nucleosome reorganization during transcription [55,66,68–]. Of note, both of the models are RNAPII-dependent, although FACT is also found at genomic regions transcribed by RNAPI and RNAPIII [69]. In the “dimer eviction model,” FACT actively removes H2A–H2B from nucleosomes to enhance DNA accessibility. In support of this view, FACT cannot promote transcription when histones are cross-linked to prevent H2A–H2B dimers from being evicted [12,14]. In contrast, recent reports support the “non-eviction model”, where FACT interacts with nucleosomes and increases chromatin accessibility without evicting the H2A–H2B dimer. In accordance with this model, it was found that yFACT promotes transcriptional activation of the GAL1-10 promoter without significant dimer loss [67]. The loss of H2A-H2B dimers, especially H2B, seems to have a great impact on FACT-dependent transcription [31]. Thus, it is likely that FACT binds to the H2A-H2B dimer to ensure the integrity of the nucleosome [70]. Indeed, FACT has been recently proposed to interact with the H2A-H2B dimer through its CTD to ensure nucleosome survival during transcription [31,50]. RNAPII binds to promoter regions and uncoils nucleosome DNA, thereby exposing the DNA-binding surface of the proximal H2A-H2B dimer and enabling its interaction with FACT. In this way, FACT retains the proximal H2A-H2B dimer and maintains the integrity of the nucleosome during transcription. After RNAPII passage, the DNA recoils on the surface of the proximal H2A-H2B dimer, which moves FACT away from the proximal dimer and promotes its interaction with the distal dimer. Taken together, accumulating evidence suggests that FACT causes the instantaneous displacement of H2A-H2B from nucleosome to ensure successful passage of RNAPII during transcriptional elongation [30,49]. FACT also suppresses cryptic transcription buried within the gene body and ensures transcriptional fidelity by maintaining chromatin stability [16,53,71]. The evidence that the absence of FACT is usually accompanied by the loss of nucleosomes after transcription further supports that FACT safeguards the survival of nucleosomes during transcription [14,72,74–]. However, presence of FACT is also accompanied by an occasional eviction of H2A-H2B dimers from nucleosomes, which might due to the transient displacement of H2A-H2B dimer from the nucleosome.

Figure 2. Working models of FACT function. (a) The “dimer eviction model”, in which FACT actively removes H2A–H2B from nucleosomes to enhance DNA accessibility. (b) The “non-eviction model”, in which FACT makes the nucleosome structure more relaxed and dynamic without the eviction of H2A–H2B

Consistent with the “non-eviction model”, recent reports suggested that FACT cannot bind to intact nucleosomes; Instead, an unstable nucleosome structure during transcription promotes FACT recruitment [29,48,49,68,75–78]. In support of this view, curaxins, an anti-cancer drug that destabilizes nucleosomes, disrupts the co-localization of FACT and RNAPII, and promotes redistribution of FACT into unstable nucleosomes area in non-transcribed regions [55,76,79]. It should be noted that FACT may also regulate transcriptional activation in eukaryotic cells. In Drosophila S2 cells, depletion of FACT releases RNAPII from transcription pause sites and engages in productive elongation, indicating that FACT is able to maintain RNAPII at transcription pause sites and fine-tunes transcriptional activation [80]. Furthermore, FACT knock down (KD) significantly affects gene expression profile in non-small cell lung cancer, and different subsets of genes were affected by SSRP1 and Spt16 KD, indicating that FACT subunits have independent roles in regulating gene expression [81]. Of note, ATP-dependent chromatin-remodeling complexes and histone chaperones were reported to act in concert with FACT during transcription. In yeast, the chromatin-remodeling complex SWI/SNF and the histone chaperone Asf1 work together at heat shock gene promoters during transcription initiation, and yFACT participates in subsequent nucleosome reassembly [73]. For transcriptional elongation, H3K4me3 places the chromatin-remodeling factor CHD1 near the TSS, which in turn recruits FACT to promote transcriptional elongation [82,83].

FACT is also involved in RNAPI and RNAPIII mediated transcription; however, the underlying mechanism is less clear. It was reported that FACT interacts with RNAPI directly and the Spt16 subunit could deposit H2A-H2B dimers onto coding regions of the ribosomal DNA [69,84]. Interestingly, FACT also acts as a chaperone for histone H2A.Z during RNAPIII mediated tDNA transcription and mutation of the subunits (Spt16 and Pob3) increased tDNA transcription in yeast [85].

FACT in DNA damage repair

Ionizing radiation (IR), ultraviolet (UV) light and oxidative stress could cause DNA damage [86–88], and numerous studies suggest that FACT is also involved in DNA damage recognition and repair processes. The nucleosome barrier needs to be removed during the early DNA damage repair process. Poly-ADP-Ribosylation (PARlation) promotes removal of histones following DNA Double-Strand Breaks (DDSBs), and FACT is involved in this process [89]. Inhibition of PARlation significantly reduces recruitment of FACT, suggesting that FACT accumulation at DNA damage site is regulated by PARlation. In this process, the C-terminal region of SSRP1 plays a crucial role in recognizing PARylated histones and may serve as a signal for FACT to locate to the DNA damage sites [89]. During UV-induced DNA damage response, the HMG domain of SSRP1 recognizes DNA bending or kinking and recruits SSRP1 to the site of damage [90]. Moreover, accumulating evidence suggests that FACT is the primary regulator of histone H2A.X exchange during DNA damage repair [91]. H2A.X is a histone H2A variant in eukaryotes that contributes to chromatin remodeling during DNA damage repair [92]. H2A.X responds to DDSBs by rapid phosphorylation on the highly conserved serine 139 (γH2A.X) at its C-terminus, and γH2A.X acts as a signal to recruit DNA repair and DNA damage checkpoint proteins, such as INO80, MDC1, and TP53BP1 [93–95]. FACT not only promotes the deposition of newly synthesized H2A.X to nucleosomes [96,97] but also participates in exchange of histone variant H2A.X with canonical H2A in nucleosomes during DNA damage repair. Notably, FACT is responsible for exchanging H2A.X-H2B with H2A-H2B dimer but not with the H2A.Z-H2B dimer [96].

Base excision repair (BER) is an essential and universal DNA repair pathway. However, the presence of nucleosomes strongly interferes with the efficiency of BER [98–100]. FACT participates in BER through its “co-remodeling” activity [101] and acts in concert with Remodeling the Structure of Chromatin (RSC) to facilitate the removal of uracil from nucleoside DNA, indicating that FACT could promote the repair of DNA damage in the initial step of BER [101]. In transcription-coupled nucleotide excision repair (TC-NER), Spt16 has been shown to interact with UV-stimulated scaffold protein A (UVSSA) and promote recovery of subsequent transcription [102].

Homologous Recombination (HR) is another important DNA damage repair mechanism [103]. SSRP1 has been confirmed to interact with HR key repair protein Rad54 both in vitro and in vivo. SSRP1 KD leads to increased HR events resulting in more DNA damage at stalled replication sites. Consistently, overexpression of SSRP1 inhibits spontaneous and Hydroxyurea-induced HR events, indicating that SSRP1 plays an essential role in the HR repair pathway [104]. Studies demonstrated that Spt16 interacts with E3 ubiquitin ligase RNF20 during DNA damage and promotes chromatin remodeling during HR repair through histone H2B ubiquitylation [105–107].

Taken together, the findings suggest that FACT is involved in various DNA damage and repair responses (Figure 3). FACT interacts with DNA repair proteins and chromatin-remodeling factors at DNA damage sites to create a suitable chromatin environment. However, the exact mechanism of FACT-mediated chromatin remodeling during DNA damage responses remains to be determined.

Figure 3. The role of FACT in DNA damage repair. (a) FACT and RSC work together to promote DNA damage repair during the initial steps of the base excision repair process. (b) FACT acts as an early factor in UV-induced DNA damage and recruits UVSSA during the transcription-coupled nucleotide excision repair (TC-NER) process. (c) FACT participates in the exchange of histone variant H2A.X with conventional H2A during the DDSB repair process. (d) FACT promotes the homologous recombination repair process

FACT in DNA replication

In the process of DNA replication, the nucleosome on the parental DNA in front of the replication fork needs to be disassembled, and the two newly synthesized daughter DNA strands behind the replication fork need to be assembled into chromatin by parental and newly synthesized histones. This process is called DNA replication-coupled nucleosome assembly and is regulated by various histone chaperones [108]. Early genetic research showed that yFACT interacts with various DNA replication factors and participates in DNA replication [22,109–112]. For example, (i) yFACT binds to DNA polymerase α [113,114]; (ii) yFACT interacts with the DNA-replicating protein RPA1 (binds and stabilizes single-stranded DNA intermediates during DNA replication) [22]; and (iii) Pob3 is associated with a variety of replication factors (POL1, CTF4, DNA2, and CHL12) [109] and travels through chromatin with the replisome complex. Recently, it was reported that ubiquitination of Spt16 or histone H2B recruits FACT to the DNA replication origin [115,116]. After recruitment, FACT acts as an important partner for MiniChromosome Maintenance (MCM), which is the catalytic core of replication helicases in eukaryotes, and the instability of the FACT-MCM complex results in defective DNA replication initiation [44,115,117,118]. Since MCM has helicase activity only in the context of naked DNA, but not in the chromatin context, the presence of FACT increases chromatin accessibility and promotes MCM-mediated DNA unwinding on nucleosome templates [117,119]. Furthermore, FACT also increases the speed of replication, suggesting that FACT may promote the progression of replication forks through nucleosome reorganization [70,116,119–121].

As the replication fork progresses, both parental and newly synthesized histones are deposited to the leading and lagging strands [108]. A recent finding suggests that FACT collaborates with Chromatin Assembly Factor 1 (CAF-1) and Regulator of Ty1 Transposition Protein 106 (Rtt106) to participate in new histone deposition. Rtt106 transfers the newly synthesized H3-H4 to FACT, and FACT deposits the H3-H4 onto nucleosomes [120]. The histone modification H3K56ac facilitates this process. Meanwhile, the complex formed by FACT and MCM2 prevents loss of parental histones by retaining parental histones around the replication fork [121]. Collectively, FACT regulates replication initiation and progression through its chromatin-remodeling activity. Meanwhile, FACT is required for the deposition of newly synthesized histones and participates in the recycling process of the parental histones during replication-coupled nucleosome assembly.

FACT and posttranslational modifications

The highly dynamic nature of chromatin is regulated by a complex network consisting of chromatin remodelers, histone chaperones, and posttranslational modifications [122,123]. Among them, histone chaperone FACT has been shown to work in concert with various posttranslational modifications to regulate chromatin dynamics. In particular, H2B Ubiquitylation (H2Bub) and Spt16 have been shown to interplay in each other’s regulation during transcription. While Spt16 regulates the formation of H2Bub, H2Bub is necessary for the stable accumulation of Spt16 at gene-coding regions [124]. Earlier in vitro experiments suggested that RNAPII halts at the first nucleosome upon transcription initiation and recruits FACT, which in turn recruits ubiquitination machinery to induce H2Bub formation [58]. A recent study reported that the binding affinity of FACT to H2A-H2B dimers was reduced approximately 20% in the presence of ubiquitination in vitro, and the loss of H2Bub stabilized interaction between FACT and histones in vivo, suggesting that H2Bub reduces FACT and histone-binding affinity. The acidic path residues on histones are important for FACT binding, and one possibility is that the ubiquitination of H2B could mask certain FACT binding epitopes and block its access to FACT, therefore decreasing the affinity between FACT and the H2A-H2B dimer [125,126]. Moreover, H2Bub enhances the release of histones from FACT and fine-tunes chaperone-histone binding dynamics to facilitate deposition of histones on DNA during nucleosome reassembly following transcription [124,125]. FACT, in turn, is required for the deubiquitinating enzyme Ubp10 to cleave H2Bub from nucleosomes [115,116]. In addition, an increased level of H3K36me3 leads to enhanced recruitment of FACT to the chromatin template, thereby promoting recovery of chromatin structure after RNAPII passage [127]. In contrast, H3K56ac has been shown to impair the ability of FACT to assemble nucleosomes [128]. Of note, FACT also maintains the stability of epigenetic markers by recycling local histones, and its absence results in widespread changes in histone markers during transcription [80,129,130].

Beyond histone modifications, posttranslational modifications of FACT add another mechanistic layer for regulating its activity. Phosphorylation of SSRP1, especially at serine 510, by Casein Kinase 2 (CK2) reduces DNA binding activity of FACT [131]. Phosphorylation also occurs at the IDD of SSRP1, which weakens the binding between DNA and the HMG domain of SSRP1 and inhibits DNA-binding activity of FACT [23]. However, phosphorylation of the C-terminal Acidic Intrinsically Disordered (AID) segment promotes the binding of FACT and nucleosomes [132]. In yeast, Rtt101-mediated ubiquitination of Spt16 is crucial for the interaction between FACT and MCM [115,133]. In contrast, the E3 ubiquitin ligase UBR5-mediated ubiquitination of Spt16 inhibits FACT-dependent transcriptional elongation in human cells [134,135]. Interestingly, PARlation of Spt16, induced by PARP1, inhibits H2A.X exchanging activity of FACT [96].

FACT and cell fate determination

Cell fate determination is mostly regulated by expression of cell type-specific Transcription Factors (TFs) during differentiation [136,137] and requires a permissive chromatin environment to ensure efficient binding of the TFs [138]. In this context, multiple histone chaperones, including FACT, have been shown to regulate cell identity by reshaping chromatin structure [139].

FACT is essential for genomic reprogramming of induced pluripotent stem cells (iPSCs) [140,141] and has been suggested to interact directly with Oct4, a pioneer transcription factor that is essential for iPSC induction [142–144]. Consistently, FACT is abundantly expressed in undifferentiated cells [145]. In addition, FACT is responsible for removing histone H3 after nucleosome demethylation process mediated by Jumonji domain containing 1 C (Jmjd1c) and Oct4 [146]. Collectively, these evidences suggest that by regulating chromatin accessibility, FACT helps TFs to reset gene expression profiles during cell fate determination.

In contrast to abovementioned function, FACT is also able to maintain expression of specific genes during cell differentiation. SSRP1 KD significantly affects activation of the Wnt signaling pathway during differentiation of mesenchymal stem cells to osteoblasts, and depletion of Spt16 affects the anterior pharynx development in C. elegans embryos [147,148]. However, FACT may also act as a negative regulator and restrict expression of other cell-specific genes to protect cell identity [140,141]. In support of this view, FACT prevents mis-activation of transcription start sites in mouse ESCs by regulating structure of the promoter regions [54]. FACT and Tripartite Motif Containing 33 (TRIM33) form a complex on distal regulatory regions in macrophages and restrict transcription of nearby genes [149].

Conclusions and perspective

The histone chaperone FACT is essential for regulating cellular processes involving chromatin, such as transcription, replication and repair, where it interacts directly with nucleosomes and maintains chromatin homeostasis. Unique motifs and domains of the FACT subunits provide structural properties that are conducive to its interaction with histones, especially with the H2A-H2B dimer. While the CTD of the Spt16 subunit is indispensable for the FACT interaction with the H2A-H2B dimer, the MD of Spt16 is essential for FACT interaction with H3-H4 tetramer and is required for recognition of the H2A-H2B dimer in the nucleosome [28,29]. It is likely that FACT dissociates H2A-H2B from nucleosomes but keeps them nearby to maintain relative integrity of nucleosomes during transcriptional elongation. Furthermore, FACT is able to safeguard new histone deposition during DNA replication and fine-tunes histone variants exchange in the DNA damage repair process. Of note, FACT functions together with ATP-dependent chromatin remodelers to reorganize or remodel chromatin architecture during various cellular processes.

Since its discovery in 1999, more than 20 years of studies have deepened our understanding regarding the functions of FACT and the underlying molecular mechanisms. However, many outstanding questions regarding the exact role of FACT in different DNA-templated cellular processes remain to be explored. One is that if the basic role of FACT is to simply maintain a pool of stably folded H2A-H2B dimer, whether and how do other factors regulate its different functions in different cellular contexts? H2B ubiquitination plays an important role in regulating FACT activity, but the interplay between FACT and other posttranslational modifications remain mostly elusive. FACT complex genes in C. elegans include two Ssrp1 orthologs that have redundant functions in embryonic development. However, they are each uniquely required for normal oocyte development and fertility, suggesting that SSRP1 is likely to exert Spt16-independent functions in germ line development [148,150]. Thus, it will be interesting to elucidate the specific role of FACT during the development of mammalian oocytes and preimplantation embryos.

Emerging evidence suggests a novel role of FACT in cell fate determination and differentiation. Expression of FACT in undifferentiated cells is significantly higher than that in somatic cells. However, the higher expression of FACT is not associated with cell proliferation [145,151,152] and decreases with cell differentiation. In line with this, most cells in adult tissues do not have detectable FACT expression at protein level. This line of evidence collectively suggest that high FACT expression is associated with stem cells or less-differentiated cells, whereas low FACT level correlates with differentiated cell status [145]. Notably, FACT is also abundantly expressed in tumor cell, and down regulation of FACT expression greatly suppresses tumor cell growth [152]. Therefore, it is likely that FACT complex also plays important roles in cell fate decisions, and the function and molecular mechanism in the context of totipotency maintenance and carcinogenic transformation need to be explored further in the future.

Availability of data and materials

Not applicable.

Acknowledgments

We thank all members of the Nashun laboratory for stimulating discussions and apologize to all colleagues whose work could not be cited due to the space constraints.

Disclosure statement

The authors declare no conflict of interest.

Additional information

Funding

This study was supported by the National Natural Science Foundation of China (31760335, 31970759), Inner Mongolia University High-Level Talents Award to Buhe Nashun (21400-5165159) and the Science and Technology Major Project of Inner Mongolia Autonomous Region of China to the State Key Laboratory of Reproductive Regulation and Breeding of Grassland Livestock (2020ZD0008).