ABSTRACT
Biological macromolecule condensates formed by liquid-liquid phase separation (LLPS) have been discovered in recent years to be prevalent in biology. These condensates are involved in diverse processes, including the regulation of gene expression. LLPS of proteins have been found in animal, plant, and bacterial species but have scarcely been identified in viral proteins. Here, we discovered that Epstein-Barr virus (EBV) EBNA2 and EBNALP form nuclear puncta that exhibit properties of liquid-like condensates (or droplets), which are enriched in superenhancers of MYC and Runx3. EBNA2 and EBNALP are transcription factors, and the expression of their target genes is suppressed by chemicals that perturb LLPS. Intrinsically disordered regions (IDRs) of EBNA2 and EBNALP can form phase-separated droplets, and specific proline residues of EBNA2 and EBNALP contribute to droplet formation. These findings offer a foundation for understanding the mechanism by which LLPS, previously determined to be related to the organization of P bodies, membraneless organelles, nucleolus homeostasis, and cell signaling, plays a key role in EBV-host interactions and is involved in regulating host gene expression. This work suggests a novel anti-EBV strategy where developing appropriate drugs of interfering LLPS can be used to destroy the function of the EBV’s transcription factors.
IMPORTANCE Protein condensates can be assembled via liquid-liquid phase separation (LLPS), a process involving the concentration of molecules in a confined liquid-like compartment. LLPS allows for the compartmentalization and sequestration of materials and can be harnessed as a sensitive strategy for responding to small changes in the environment. This study identified the Epstein-Barr virus (EBV) proteins EBNA2 and EBNALP, which mediate virus and cellular gene transcription, as transcription factors that can form liquid-like condensates at superenhancer sites of MYC and Runx3. This study discovered the first identified LLPS of EBV proteins and emphasized the importance of LLPS in controlling host gene expression.
INTRODUCTION
Epstein-Barr virus (EBV) is the first identified human tumor virus isolated from African Burkitt’s lymphoma samples (1). It belongs to the γ-subfamily of herpesvirus and infects approximately 95% of the world’s population. EBV infection is associated with diverse cancers, including nasopharyngeal carcinoma, gastric cancers, and lymphoma (2–6). EBV is the most powerful transforming agent of human cells. Its ability to immortalize resting B cells in vitro results in the generation of immortal lymphoblastoid cell lines (LCLs) in which the virus persists in its latent form (7). EBNA2 and EBNALP are two EBV-encoded transcription factors expressed in newly infected cells (8). The coexpression of EBNA2 and EBNALP contributes to EBV-mediated B-cell growth transformation by driving resting B cells into the cell cycle (9).
EBNA2 is an EBV master transcription factor that modulates EBV latent gene transcription and enhances the expression of a considerable number of cellular genes that control cell growth and survival (10). EBNA2 itself cannot bind to DNA directly, and it is tethered to DNA through interaction with cellular transcription factors, such as RBPJ and ZNF143, to target viral and cellular gene regulatory elements (11–13). EBNA2 activates gene transcription through interactions between its C-terminal transactivation domain and histone acetyl transferases (11, 14, 15). EBNALP cooperates with EBNA2 to activate virus and cellular gene transcription (16, 17) through removing NCOR and RBPJ repressive complexes from promoters, enhancers, and matrix-associated deacetylase bodies (18). Although EBNA2 binding sites are close to gene promoters and enhancer elements in the viral and cellular genome (13, 19), recent studies found that EBNA2 activates MYC and Runx3 transcription from distal enhancers hundreds of kilobases upstream of their transcription start site through specific superenhancer elements, which are large clusters of transcriptional enhancers that drive expression of genes that define cell identity (20–23). Sequencing by chromatin immunoprecipitation (ChIP-seq) also reported that EBNALP converges at EBV superenhancer sites that have high and broad H3K4me1 and H3K27ac signals, which is a characteristic of superenhancers (24). Approximately 29% of EBNALP superenhancer sites, including MYC and Runx3, colocalize with EBNA2 (23, 25).
Liquid-liquid phase separation (LLPS) is a well-known phenomenon in polymer chemistry and is a physical process occurring spontaneously when single-phase molecular complexes separate into a dense phase and a dilute phase that then stably coexist (26). Increasing evidence showed that LLPS or biomolecular condensation underlies the formation of membraneless organelles, such as nucleolus, nuclear speckles, cajal bodies, stress granules, and processing bodies (P-bodies), which provide a mechanism to compartmentalize and concentrate biochemical reactions and at the same time protect the cell from the harmful effects (26–29). Some conditions are necessary for LLPS, such as hydrophobic and electrostatic interactions driven by intrinsically disordered regions (IDRs) in proteins and weak multivalent interactions (30, 31). In addition to membraneless organelles, LLPS recently was reported to explain the mechanisms of transcriptional regulation, such as superenhancers (32–34).
The dense assembly of superenhancers has been shown to exhibit dynamic transitions of formation and dissolution due to LLPS (33, 35, 36). EBNA2 and its coactivator EBNALP have been shown to interact with a number of transcription factors to form high-density assemblies or aggregates at active superenhancers. Here, we provide experimental evidence that EBNA2 and EBNALP undergo LLPS mediated by their IDRs to form liquid-like condensates at the superenhancer sites, and LLPS facilitates the transcription activities of EBNA2 and EBNALP. LLPS of proteins has been found in animal, plant, and bacterial species but has scarcely been identified in viral proteins. The current study identified LLPS of two EBV proteins and emphasized the importance of LLPS on control host gene expression.
RESULTS
EBNA2 and EBNALP form nuclear puncta in cells.Superenhancers have been shown to exhibit dynamic transitions because they are occupied by high densities of transcription factors, including EBNA2 and EBNALP, which converge at numerous EBV superenhancer sites (20, 23, 34). We speculated whether EBNA2 and EBNALP are components of nuclear condensates. To investigate the aggregation behavior of EBNA2 and EBNALP under physiological conditions, enhanced green fluorescent protein (EGFP)-tagged EBNA2 and EBNALP were transfected into HEK293 cells, and both its expression and localization were detected using fluorescence microscopy. We found that EGFP-tagged EBNA2 and EBNALP form lots of puncta in the nucleus, and, interestingly, the size of these puncta varies (Fig. 1). This result indicates that EBNA2 and EBNALP are components of the nuclear puncta within the nuclei, and large puncta most likely are fused by small puncta.
EBNA2 and EBNALP form nuclear puncta in cells. Immunofluorescence imaging of HEK293 cells after transfection of the EGFP-EBNA2 and EGFP-EBNALP plasmids. Fluorescence signal is shown alone (left) and merged with DAPI stain (right). HEK293 cells were transfected with EGFP-ID3 as a negative control. Scale bar, 5 μm.
EBNA2 and EBNALP nuclear puncta exhibit dynamic properties.To examine whether the large puncta of EBNA2 and EBNALP are fused by small puncta and exhibit features characteristic of dynamic liquid-like condensates, such as internal dynamical reorganization and rapid exchange kinetics, we assessed the internal order within the liquid-like condensates by means of fluorescence recovery after photobleaching (FRAP) measurements (26, 27, 29). We performed FRAP experiments in HEK293 cells transfected with EGFP-EBNA2 or EGFP-EBNALP. After photobleaching, both EGFP-EBNA2 and EGFP-EBNALP puncta recovered fluorescence on a time scale of seconds (Fig. 2A and B; see also Movies S1 and S2 in the supplemental material). ATP has well-characterized roles in providing energy for biochemical reactions within cells and has been reported to be involved in promoting condensate fluidity by its hydrotropic activity or driving energy-dependent processes (37, 38). EGFP-EBNA2 or EGFP-EBNALP puncta were barely effectively recovered after deleting the cellular ATP by glucose deprivation and antimycin A treatment in the FRAP experiment (Fig. 2C and D). To further confirm the biophysical properties of EBNA2 and EBNALP puncta, we treated HEK293 cells expressing EGFP-EBNA2 or EGFP-EBNALP with 1,6-hexanediol, which is an aliphatic alcohol known to disrupt liquid-like condensates by disturbing the hydrophobic interactions (39). We found that 1,6-hexanediol significantly reduced the number of EBNA2 or EBNALP puncta (Fig. 2E). These results demonstrate that EBNA2 and EBNALP puncta exhibit dynamic liquid-like properties in cells and are consistent with previously reported LLPS condensates.
EBNA2 and EBNALP puncta exhibit dynamic properties in cells. (A) Representative images of FRAP experiment of HEK293 cells that had been transfected with EGFP-EBNALP expression vector. (Left) The yellow circle highlights the puncta undergoing targeted bleaching. (Right) Quantification of FRAP fluorescence intensity data for EGFP-EBNALP puncta. Data were normalized to 100% by maximal fluorescence intensity. Values shown are means ± SD (n = 5). The main images were extracted from Movie S1. (B, left) HEK293 cells transfected with EGFP-EBNA2 expression vector. (Right) Quantification of FRAP data for EGFP-EBNA2 puncta (n = 5). The main images were extracted from Movie S2. (C, left) Representative images of FRAP experiment of EGFP-EBNALP in HEK293 cells upon ATP depletion. (Right) Quantification of FRAP data of EGFP-EBNA2 upon ATP depletion (n = 5). (D, left) Representative images of FRAP experiment of EGFP-EBNA2 in HEK293 cells upon ATP depletion. (Right) Quantification of FRAP data for EGFP-EBNALP puncta upon ATP depletion (n = 5). (E, left) Representative live images of EGFP-EBNA2 or EGFP-EBNALP in HEK293 cells before and after treatment with 3% 1,6-hexanediol for 15 s. (Right) Quantification of data for EGFP-EBNA2 or EGFP-EBNALP puncta (n = 5). Scale bar, 5 μm.
IDRs of EBNA2 and EBNALP participate in phase separate in cells.Intrinsically disordered regions (IDRs) are conformationally flexible, facilitating interactions with multiple partners through intramolecular and intermolecular mechanisms and structurally heterogeneous protein domains that encode diverse protein functions (40, 41). Recent studies showed that low-complexity IDRs drive LLPS (42–45). We found that EBNA2 and EBNALP contain large IDRs by using PONDR (http://www.pondr.com/) and IUPred2 (https://iupred2a.elte.hu/plot), two algorithms that predict IDRs in proteins based on their amino acid sequence (Fig. 3A and B). Interestingly, EBNALP protein has seven segments with the same low-complexity amino acid repeat sequences. To investigate whether the IDRs of EBNA2 or EBNALP promote phase separation in cells, the long EBNA2 IDRs (residues 145 to 441) and six-segment truncated EBNALP IDRs (residues 59 to 124, 59 to 190, 59 to 256, 59 to 322, 59 to 388, and 59 to 354) were fused to EGFP vectors and named EBNA2-1 and EBNALP-1 to EBNALP-6, respectively (Fig. 3C), and then were transfected into HEK293 cells. EBNA2-1 can form obvious nuclear puncta (Fig. 3D); FRAP experiments further showed that when bleaching was performed on the nuclear puncta, the fluorescence of EBNA2-1 was efficiently recovered on a time scale of seconds (Fig. 3E and F and Movie S3). Moreover, we found that the longer the truncated mutant of EBNALP, the stronger its ability to form nuclear puncta. Similarly, FRAP experiments showed that the longer the truncated mutant of EBNALP, the stronger the recovery ability after bleaching (Fig. 3G and Movies S4 to S9). These data reveal that the IDRs of EBNA2 or EBNALP mediate self-organization into macroscopic dynamic spherical structures in live cells that can respond rapidly to changes in the cellular environment.
IDRs of EBNA2 and EBNALP involved in phase separation in cells. (A and B) Predictions of IDRs of EBNA2 (A) and EBNALP (B) using PONDR (http://www.pondr.com/) and IUPred2 (https://iupred2a.elte.hu/plot) algorithms based on their amino acid positions and sequences. The orange bar designates the IDR under prediction. (C) Schematic diagram of a series of EGFP-tagged EBNA2 and EBNALP IDR truncated mutants. The orange bar indicates IDRs of EBNA2 (EBNA2-1) and truncated EBNALP (EBNALP-1 to EBNALP-6) shown in panels A and B. (D) Immunofluorescence monitored the nuclear puncta formation and the subcellular localization of EBNA2-1 in HEK293 cells. (E) FRAP of EBNA2-1 nuclear puncta. Time 0 indicates the time of the photobleaching pulse. The yellow circle highlights the puncta undergoing targeted bleaching. The main images were extracted from Movie S3. (F) Plot showing the time course of the recovery after photobleaching EBNA2-1 nuclear puncta. (G, left) Time 0 indicates the time of the photobleaching pulse. The yellow circle highlights the puncta undergoing targeted bleaching. The main images were extracted from Movies S4 to S9. (Right) Plot showing the time course of the recovery after photobleaching EBNALP-1 to EBNALP-6 nuclear puncta. Data are presented as means ± SD (n =5). Scale bar, 5 μm.
IDRs of EBNA2 and EBNALP drive phase separation in vitro.So far, our data indicate that EBNA2 and EBNALP are a component of liquid-like compartments in vivo that form by LLPS. Previous studies illustrated that the purified IDRs of some proteins are involved in phase-separated droplet formation in vitro (42, 46–48). To investigate whether EBNA2 and EBNALP are able to phase separate on their own, we studied the behavior of recombinant GFP-tagged EBNA2-1 and EBNALP-1 proteins. Freshly prepared solutions of the recombinant GFP-tagged EBNA2-1 and EBNALP-1 fusion proteins were added to buffer (pH 7.2); we observed a rapid increase of turbidity when 10% polyethylene glycol (PEG) 10000 (a molecular crowding agent) was added to the buffer, strongly suggesting LLPS (Fig. 4A). The size of phase-separated droplets had been reported to correlate with the concentration of protein components in the system (43). We next tested the ability to form droplets of EBNA2-1 and EBNALP-1 at various concentrations, ranging from 1 μM to 15 μM. We found that EBNA2-1 and EBNALP-1 formed droplets with concentration-dependent size distributions (Fig. 4B). We also found that 1,6-hexanediol significantly reduced the turbidity of EBNA2 or EBNALP droplets by treating EBNA2-1 and EBNALP-1 droplets with 1,6-hexanediol (Fig. 4C). To further reveal the biophysical properties of these liquid-like droplets, we tested the effect of various salt concentrations on the droplet formation ability. We found that the size and number of droplets were significantly suppressed as NaCl concentration increased from 50 mM to 300 mM (Fig. 4D). The combined observations demonstrate that IDRs of EBNA2 and EBNALP are prone to phase separation in vitro, and electrostatic interactions contribute to droplet formation.
IDRs of EBNA2 and EBNALP phase separate in vitro. (A) Visualization of turbidity associated with droplet formation. Tubes containing EGFP, EBNA2-1, or EBNALP-1 (final concentrations, 10 μM) in the presence (+) or absence (−) of PEG 10000 (final concentrations, 10%) are shown. (B) Representative images of droplet formation at different protein concentrations. EGFP, EBNA2-1, or EBNALP-1 was added to droplet formation buffer to final concentrations as indicated. (C) Turbidity in the absence and presence of 3% 1,6-hexanediol. Experiments were performed at 37°C in the presence of 10% PEG. Error bars represent SD (n = 5). (D) Representative images of droplet formation at different salt concentrations. EBNA2-1 or EBNALP-1 was added to droplet formation buffer to achieve 10 μM protein concentration with the indicated final NaCl concentration. Scale bar, 10 μm.
Proline residue is necessary for EBNA2 and EBNALP phase separation.Specific amino acid residues of proteins have been shown to be involved in LLPS (49). Amino acid content analysis of EBNA2 and EBNALP revealed that the characteristic hallmark of the EBNA2 proteins is proline rich, and EBNALP contains a striking compositional bias for proline, glycine, and arginine (Fig. 5A). To test the role of these rich amino acid residues, we replaced all these rich amino acid residues in EBNA2 and EBNALP with alanine residues and investigated the ability of this mutant to form liquid-like puncta in vivo. We first predicted the effect of these mutations on the IDRs of EBNA2 and EBNALP using IUPred2 software and found that the P-to-A mutant significantly destroyed the IDRs of EBNA2 and EBNALP, while R-to-A and G-to-A mutants had less influence on EBNALP (Fig. 5B and C). As we expected, the P-to-A mutants of EBNA2 and EBNALP were hardly capable of forming liquid-like puncta, whereas the R-to-A and G-to-A mutants of EBNALP still have the ability to form liquid-like puncta according to immunofluorescence (Fig. 5D to F). We also investigated the ability of the mutated EBNA2 and EBNALP proteins to form phase-separated droplets in vitro. We observed the EBNA2 P-to-A or EBNALP P-to-A mutant was almost incapable of forming phase-separated droplets under conditions where wild-type EBNA2 and EBNALP and the EBNALP R-to-A or EBNALP G-to-A mutant readily formed droplets (Fig. 5G and H), indicating that the proline residue in EBNA2 and EBNALP is necessary for liquid-like nuclear formation of puncta.
Proline residues are necessary for EBNA2 and EBNALP phase separation. (A) Heatmap analyzing the amino acid composition and position of EBNA2 (upper) and EBNALP (lower) proteins. Each row represents information for a single amino acid. The length of the row corresponds to the length of EBNA2 and EBNALP proteins. (B and C) Predictions of IDRs of EBNA2 mutating all prolines to alanines (B) and EBNALP mutating all prolines (upper), arginines (middle), or glycines (lower) to alanine (C) using IUPred2 (https://iupred2a.elte.hu/plot) algorithms. (D and E) Immunofluorescence detection of the effect of EBNA2 (D) or EBNALP (E) mutation on nuclear puncta formation in HEK293 cells. (F) Quantification of the nuclear puncta of EBNA2 and EBNALP, presented in Fig. 1, and the nuclear puncta of EBNA2 P-to-A, EBNALP P-to-A, EBNALP R-to-A, and EBNALP G-to-A mutants, presented in panels D and E. (G and H) Mutating all prolines to alanine of EBNA2 (G) and all prolines (upper), arginines (middle), or glycines (lower) to alanine of EBNALP (H) disrupts phase separation. Representative images of wild-type EBNA2 and EBNALP or all mutants using droplet formation assay. Error bars represent SD. Scale bar, 5 μm.
EBNA2 and EBNALP droplets control transcription for genes.Recent studies discovered that phase separation drives gene transcription through concentrated factors within transcriptional condensates enriched at enhancer-rich gene clusters (32, 34, 50, 51). EBV EBNA2 and EBNALP proteins control host cellular gene transcription through long-range enhancer-promoter looping to activate key oncogenes and inactivate tumor suppressor genes by assembling within some specific superenhancer elements in LCLs (18, 19, 23, 52). Thus, we speculate that the LLPS capacity of EBNA2 and EBNALP is associated with gene transcription. To determine the effect of LLPS on the target genes of EBNA2 and EBNALP, we treated three EBV-positive cell lines, GM12878, Raji, and primary B cells (primary B cells were infected with EBV at a multiplicity of infection [MOI] of 50 for 36 h), with 1,6-hexanediol and caused a significant reduction in the expression of MYC, Runx3, CR2, and CCR7 (Fig. 6A). Previous ChIP-seq results have confirmed that EBNA2 binds to −428 and −525 kb superenhancer elements upstream of the MYC transcription start site and a −98 kb superenhancer element upstream of the Runx3 transcription start site (20, 23). To determine the effect of 1,6-hexanediol on EBNA2 occupancy at the enhancers, we generated a series of primers corresponding to the putative MYC enhancers (MYC428-1 to -3 and MYC525-1 to -3) and Runx3 enhancers (Runx3-1 to -3) (Fig. 6B and C), and ChIP was performed with antibodies against EBNA2 in untreated or 1,6-hexanediol-treated GM12878 cells. 1,6-Hexanediol treatment caused an obvious reduction of EBNA2 binding in all three enhancers (Fig. 6D and E). We performed immunofluorescence for EBNA2 together with DNA-fluorescent in situ hybridization (FISH) for the genomic region containing the MYC and Runx3 superenhancers in GM12878 cells. We found that EBNA2 puncta consistently overlapped the DNA-FISH foci (Fig. 6F). We also investigated the effect of proline residues on EBNA2 transcription activity. We found that an EBNA2 P-to-A mutant caused a significant reduction in the expression of MYC, Runx3, CR2, and CCR7 compared to that of wild-type EBNA2, although this was not the case for Frizzled2 (a non-EBNA2 target gene used as a negative control here) (Fig. 7A). Similarly, ChIP was performed with antibodies against FLAG in transfected wild-type EBNA2 or the EBNA2 P-to-A mutant in BJAB cells. The EBNA2 P-to-A mutant caused an obvious reduction of EBNA2 binding in MYC or Runx3 enhancers compared to that of wild-type EBNA2 (Fig. 7B and C). These results suggest that EBNA2 forms condensates at superenhancers that control gene transcription.
1,6-Hexanediol disrupts EBNA2 transcription activity. (A) GM1278, Raji, and primary B cells were stimulated with 1,6-hexanediol for 2 h, and cell lysate was assayed for mRNA expression levels of MYC, Runx3, CR2, and CCR7 genes, all of which are EBNA2 target genes. (B and C) Schematic diagram of the putative MYC (B) and Runx3 (C) superenhancer region truncated forms and their interaction with EBNA2. (D) ChIP-PCR analysis for detection of EBNA2 binding to the MYC or Runx3 superenhancer sites in GM12878 cells in treatment with vehicle or 1,6-hexanediol for 2 h, with Frizzled2 promoter sites as a negative control. (E) Quantification of band intensities of the PCR products. (F) Colocalization between EBNA2 and the MYC and Runx3 superenhancer locus by immunofluorescence (IF) and DNA-FISH in fixed GM12878 cells. Separate images of the indicated IF and FISH are shown, along with an image showing the merged channels (overlapping signal in yellow).
Proline residues are necessary for EBNA2 transcription activity. (A) BJAB cells were transfected with FLAG-EBNA2 or FLAG-EBNA2 P-to-A mutant, and cell lysate was assayed for mRNA expression levels of MYC, Runx3, CR2, CCR7 (EBNA2 target genes), and Frizzled2 (non-EBNA2 target genes). (B) ChIP-PCR analysis for detection of EBNA2 binding to the MYC or Runx3 superenhancer sites in BJAB cells in the transfection with FLAG-EBNA2 or FLAG-EBNA2 P-to-A. (C) Quantification of band intensities of the PCR products. Frizzled2 promoter sites are shown as a negative control.
DISCUSSION
Phase separation in biology has exploded onto the stage in the past 10 years, and evidence is now mounting that LLPS underlies the formation of dynamics of many membraneless organelles and is involved in major cellular processes (49). In eukaryotic cells, thousands of proteins containing IDRs have been reported to be involved in the assembly of phase-separated macromolecules, such as FUS, hnRNPA, PML, Nucleolin, DDX4, etc. (42, 53–55), but such LLPS-forming proteins have not yet been identified in EBV. Actually, a very limited number of viral proteins had been identified with LLPS characterizes. Superenhancers are clusters of transcriptional enhancers clustered by binding of major transcription factors to drive the transcription of genes (56). Superenhancers have been proposed to form non-membrane-bound condensates by LLPS (33). Previous studies verified that EBV EBNA2 and EBNALP are two key components of superenhancers that regulate the transcription of key target genes (20, 22, 23, 25). Here, we demonstrate that transcription factors EBNA2 and EBNALP function by forming LLPS nuclear condensates to drive superenhancer-mediated transcription (Fig. 8).
Model depicting the main molecular mechanisms of EBNA2 and EBNALP condensation at superenhancers to activate genes by LLPS. In this model, EBNA2 and EBNALP recruit other coactivators and transcription factors, forming phase-separated condensates at enhancer sites to drive gene activation. These are driven in part by the interactions of IDRs.
Intrinsically disordered regions or proteins (IDRs or IDPs) are ubiquitous in proteomes, with ∼25% to 30% of eukaryotic proteins being mostly disordered, and serve in a range of cellular functions, including cellular signaling, gene expression, transport, enzyme function, and flexibility and interaction promiscuity (57, 58). IDP misfunction and aggregation are also associated with several diseases, including neurodegenerative diseases and cancer (59–62). In recent years, IDPs have been shown to be important players in cellular LLPS (27, 29, 63). EBNA2 and EBNALP contain large IDRs; in particular, EBNALP contains seven segments of repeated IDRs. These IDRs of both EBNA2 and EBNALP are sufficient to facilitate and form phase-separated droplets in living cells and in vitro. Moreover, we found that the longer the IDRs of EBNALP, the stronger its ability to form phase-separated condensates, which makes it likely that these IDRs mediate multivalent interactions, thereby facilitating LLPS, and it also means that the longer IDRs can provide a more effective scaffold to enhance interactions between IDRs and multiple partners of phase-separate condensates.
Amino acid sequences of IDRs/IDPs are biased and they are enriched in polar and charged residues, resulting in depletion in order-promoting residues Trp, Tyr, Phe, Ile, Leu, Val, Cys, and Asn and enrichment in disorder-promoting residues Ala, Arg, Gly, Gln, Ser, Glu, Lys, and Pro (64–66). Previous studies indicated that LLPS is governed by some specific amino acid residues to control multivalent interactions, such as glycine residues enhancing the fluidity while glutamine and serine residues promote hardening (67). Proline is a residue associated with IDP and polypeptide expansion (68). Proline-rich motifs of some proteins have been verified to play a vital role in LLPS (31, 69). Our studies show that proline residues are enriched in EBNA2 and EBNALP (27.3% for EBNA2 and 17.6% for EBNALP; the average occurrence rate for proline in the vertebrate proteomes is 5% [69]), and mutation of proline to alanine severely attenuates the LLPS of EBNA2 and EBNALP. Although EBNALP also contains a large proportion of glycine and arginine, which are involved in maintaining liquidity and governing the saturation concentration of phase separation (67), mutating them to alanine has little effect on the LLPS of EBNALP. These results suggest that the proline residues are essential for EBNA2 and EBNALP to form phase-separated droplets.
Posttranslational modifications play important roles in protein structures and functions at different levels by covalently adding various chemical groups to amino acid side chains, removing various chemical groups, or via enzymatic cleavage of peptide bonds (70). The assembly of components within many LLPS systems is electrostatically driven (63). Therefore, posttranslational modifications that alter the charge characteristic of amino acids within IDRs of proteins provide a means to regulate their multivalent interactions and phase separation behavior. Deregulation of posttranslational modifications is commonly associated with the development of various pathological conditions by affecting disorder-based LLPS (71–73). The importance of electrostatic interactions has been demonstrated by the phase separation behavior of DDX4 and hnRNPA1 (42, 74, 75), whose abilities to form phase-separated condensates are strongly affected by the salt concentration of the buffer. EBNA2 and EBNALP contain a relatively high proportion of polar amino acids or electrically charged amino acids, such as arginine and glycine, especially for EBNALP. When we increased the NaCl concentration, liquid-like droplets of EBNA2 and EBNALP were destroyed. Methylation of arginine residues in the RG or RGG domain of FUS/TLS, EWS, TAF15, hnRNPs, and DDX4 also has been reported to increase the LLPS threshold (42, 76). Interestingly, both EBNA2 and EBNALP contain some proportion of RG motifs. Whether these RG motifs are regulated by arginine methylation and, thus, affect LLPS of EBNA2 and EBNALP requires further study.
EBV is a causative agent of infectious mononucleosis and is associated with 200,000 new cases of cancer and 140,000 deaths annually (77). Finding a suitable treatment strategy for the diseases caused by EBV infection would be of major public health benefit. LLPS has displayed a crucial role in many diseases, such as neurodegenerative disorders and cancers (60, 78). EBV transcription factors or coactivators, including EBNA2 and EBNALP, can alter the host genome organization to modulate key oncogenes and tumor suppressors. Thus, understanding how these transcription factors or coactivators function in reorganizing the host genome will likely allow identification of novel therapeutics. Our study is the first to show that EBV has LLPS proteins; EBNA2 and EBNALP (as two key components of superenhancers) regulate the expression of key target genes through LLPS. This study provides the possibility that EBV reorganizes the host genome organization through LLPS, causing EBV-related diseases. Therefore, the development of novel therapeutic strategies to control cellular LLPS could be instrumental for the treatment of those EBV-related diseases.
MATERIALS AND METHODS
Cell culture and plasmid construction.Raji cells (EBV-positive B lymphoma cell line), GM12878 cells (EBV-positive B lymphoma cell line), BJAB cells (EBV-negative B lymphoma cell line), and peripheral blood mononuclear cells (PBMCs) were isolated from whole blood of two healthy donors by Ficoll centrifugation. Primary B cells were isolated from PBMCs with CD19 microbeads (Miltenyl Biotec, Bergisch Gladbach, Germany) and were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FBS). HEK293 cells were cultured in Dulbecco’s modified Eagle medium with 1 g/liter glucose and 10% FBS. All cell lines were obtained from the ATCC. The cell lines tested negative for mycoplasma contamination. All cell lines were authenticated by short tandem repeat profiling prior to use.
EGFP-EBNA2 plasmid was a kind gift from Ersheng Kuang at Sun Yat-Sen University. EBNALP plasmid was purchased from Addgene (37959). DNA fragments encoding FLAG-EBNA2, FLAG-EBNALP, EGFP-EBNALP-1, EGFP-EBNALP-2, EGFP-EBNALP-3, EGFP-EBNALP-4, EGFP-EBNALP-5, EGFP-EBNALP-6, and EGFP-EBNA2-1 were generated by PCR and cloned into an EGFP-tagged (pEGFP-C1) or FLAG-tagged empty vector. EBNA2 and EBNALP mutants were synthesized (Genescript, China) and inserted into the same base vector as that described above. All plasmids were verified by sequencing.
Plasmid transfection.Cells were plated at a density of 5 × 105 cells per well in 6-well plates and grown to about 60% confluence. The indicated plasmids then were transfected by Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions.
EBV preparation and infection.EBV was prepared in B958 cells as previously described (79). Briefly, EBV was grown and produced in a sufficient supply of B958 cells. Cell supernatants were concentrated at 50,000 × g at 4°C and resuspended in fresh FBS-free RPMI 1640. To determine the MOI of EBV, a DNA quantitative fluorescence diagnostic kit (Sansure Biotech, Changsha, China) was used according to the manufacturer’s protocol. In this study, we used EBV at an MOI of 50 to infect primary B cells.
Protein expression and purification.The low-complexity disordered regions of EBNA2 (residues 145 to 441) and EBNALP (59 to 124) were amplified and cloned into pET-15b-based expression vector with N-terminal 6× His and C-terminal GFP tags (Sino Biological, Beijing, China, and Cusabio, Wuhan, China). All proteins were expressed in Escherichia coli BL21(DE3) cells in LB medium at 16°C. Recombinant proteins were purified through an immobilized nickel-nitrilotriacetic acid affinity column (Ni-NTA Fastflow; Qiagen). The His tag of each protein was cleaved by incubating with thrombin (T4648; Sigma) at room temperature. Protein purity was verified by gel electrophoresis, and protein was combined and dialyzed with a buffer containing 50 mM Tris, pH 7.2, 200 mM NaCl, 10% glycerol, and 1 mM dithiothreitol (DTT).
RT-qPCR.Total RNA was extracted using TRIzol reagent (Invitrogen). For gene expression analyses, cDNA was synthesized from 2 μg total RNA with a reverse transcription (RT) kit (Thermo Fisher) in a 20-μl reaction mixture according to the manufacturer’s protocol. The mRNA level was evaluated by quantitative PCR (qPCR) using a SYBR green real-time qPCR kit (TaKaRa). Gene expression was normalized to the housekeeping gene GAPDH. The expression of each gene was quantified by measuring cycle threshold (CT) values, and the 2–ΔΔCT method was used to calculate relative changes in gene expression. The primers used for qPCR were the following: EBNA2 forward, 5′-TCTGCCACCTGCAACACTAA-3′; reverse, 5′-GTCTGGCACATGCAAGACA-3′; EBNALP forward, 5′-TCCCCTCGGACAGCTCCTA-3′; reverse, 5′-CCGCTTACCACCTCCTCTTCT-3′; CR2 forward, 5′-TGGAACCTGGGATAAACCTGC-3′; reverse, 5′-GACTTGTTTCCGTTCATGGAGA-3′; Runx3 forward, 5′-AGGCAATGACGAGAACTACTCC-3′; reverse, 5′-CGAAGGTCGTTGAACCTGG-3′; MYC forward, 5′-GGCTCCTGGCAAAAGGTCA-3′; reverse, 5′-CTGCGTAGTTGTGCTGATGT-3′; CCR7 forward, 5′-TGAGGTCACGGACGATTACAT-3′; reverse, 5′-GTAGGCCCACGAAACAAATGAT-3′; Frizzled2 forward, 5′-GTGCCATCCTATCTCAGCTACA-3′; reverse, 5′-CTGCATGTCTACCAAGTACGTG-3′.
Western blotting.Western blotting was performed as previously described (80). In brief, cells were lysed in radioimmunoprecipitation assay (RIPA) buffer containing a protease inhibitor mixture (Selleck) and incubated on a shaker at 4°C for 15 min. Lysates (50 μg of protein) were subjected to SDS-PAGE, and the separated bands were transferred to polyvinylidene difluoride (PVDF) (Millipore) and probed with antibodies against EBNA2 (MABE8; Millipore) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; D110016, Sangon Biotech, Shanghai, China). PVDF then was incubated with species-specific horseradish peroxidase (HRP)-conjugated secondary antibodies, and the immunoreactive bands were visualized by enhanced chemiluminescence (ECL; Millipore).
Immunofluorescence confocal microscopy.Immunofluorescence was performed as previously described, with some modifications (81). Cells transfected with the corresponding plasmid with the EGFP tag were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 1 h and permeabilized with 0.5% Triton X-100. Cells were washed three times with PBS, and then 4′,6-diamidino-2-phenylindole (DAPI) was used to stain nuclei for 5 min at room temperature in the dark. The confocal images were acquired using a Leica TCS SP8 confocal laser scanning microscope.
DNA-FISH combined with immunofluorescence.Main procedures for DNA-FISH were according to a previous publication (34). Briefly, the hybridization process was the following. The slides were predegenerated at 75°C for 3 min. The probe was heated at 78°C for 5 min to denature, followed by preannealing at 37°C for 1 h. The slides then were incubated with the hybridization mixture overnight at 37°C for 14 h. Following hybridization, the slides were washed with formamide-SSC (SSC is 0.15 M NaCl plus 0.015 M sodium citrate) buffer at different concentrations. After washing, drying, and mounting, the slides were examined under confocal microscopy. DNA-FISH probes were designed and generated by Qiagen to target MYC and Runx3 superenhancer. The MYC design region was GRCh37 chr8:128208000 to 128210000, and the Runx3 design region was GRCh37 chr1:25349261 to 25350000.
ATP depletion of cells.ATP depletion of cells was performed as previously described, with some modifications (82). In short, exponentially growing HEK293 cells were transfected with the indicated plasmid and cultured with glucose-free medium (DMEM; 11966025; Gibco) containing 20 mM 2-deoxyglucose (2-DG; inhibition of glycolysis) and 10 mM antimycin A (inhibition of mitochondrial ATP production) for 4 h. Recovery of cells from energy depletion was done by replacing energy depletion medium with medium containing 2% glucose.
Live imaging.Cells were grown on 2-cm glass dishes (NEST), and before imaging, cell culture medium was replaced with phenol red-free medium and imaged using a Leica TCS SP8 confocal microscope. Cells were imaged on a heated plate (37°C). Higher-resolution imaging was acquired with a Leica TCS SP8 lightning confocal microscope equipped with Leica hybrid detectors (HyD) using a 63×/1.4 oil objective.
ChIP.ChIP assays were performed as described previously (83). Briefly, GM12878 cells were fixed with 1% formaldehyde and incubated for 10 min at 37°C to generate DNA-protein cross-links and then cracked by SDS lysis buffer for 10 min on ice, followed by ultrasonic treatment to shear DNA into fragments of 200 to 400 bp and immunoprecipitated with anti EBNA2 (MABE8; Millipore) or with IgG (control) (7074S; Cell Signaling Technology). Antibody-bound complexes were precipitated with protein A/G-Sepharose beads. The DNA fragments in the immunoprecipitated complexes were released by reversing the cross-links at 65°C for 5 h, and purified DNA was analyzed by PCR and agarose gel electrophoresis. PCR and qPCR were performed using enhancer-specific primers for MYC or Runx3 and promoter-specific primers for Frizzled2 with amplification of the EBNA2-binding regions. The primers used for ChIP-qPCR are the following: MYC428-1 forward, 5′-CCACACTGATGCCTACTACACTGAA-3′; reverse, 5′-CACATTGTAACCCATAAACCATCAT-3′; MYC428-2 forward, 5′-AGCCCTGCCACTAACCGCTAACCAC-3′; reverse, 5′-TCCATCCATCCATCCATCCATCCAT-3′; MYC428-3 forward, 5′-TTTCTCTTATCCACTACTGAACCCT-3′; reverse, 5′-AATCTAGGAATAAGCCTGATCAGTG-3′; MYC525-1 forward, 5′-CACTCTGATCACTAGTTTCACCCTC-3′; reverse, 5′-AGAAGTCATGGATCAGTCCTATTGG-3′; MYC525-2 forward, 5′-CATGGCCCACAAGGTAGTAGTTCTG-3′; reverse, 5′-CAGATGTGGTGGAGCTTGTTGACTT-3′; MYC525-3 forward, 5′-GGGGTGAGTAGCGGGTGAAGAAATC-3′; reverse, 5′-AAAGAGGTATGTTCCAAGAGCCCCC-3′; Runx3-1 forward, 5′-CACCCATGGGGCCTGAGTTCTCATC-3′; reverse, 5′-CCGTCCCTGTGCTGAAGAAGATGAA-3′; Runx3-2 forward, 5′-TCTGCTCGCTCTGGTGACTCAT-3′; reverse, 5′-CCAGTCAGGAAGGCTCACTCTAA -3′; Runx3-3 forward, 5′-TGGCCATTCTCAACACCCAGCTCA-3′; reverse, 5′-GCATCTGTGTGTGTTCCGCAGTTTA-3′; Frizzled2-1 forward, 5′-CTACGTGTAAAGTGAAGTGAAAC-3′; reverse, 5′-CGGGTGGGGACTCTTTAAACCCTTG-3′; Frizzled2-2 forward, 5′-CAGGAGTTGATCCATCATCCAGTAT-3′; reverse, 5′-GCGGGAGGAAACTGGGGGGTTGT-3′; Frizzled2-3 forward, 5′-TGCACAATTAAGATTTGCCGCAGAA-3′; reverse, 5′-AATTGCCCCGTAGCCAACACAAAGG-3′.
In vitro droplet assay.Recombinant IDRs of EBNA2 or EBNALP fusion proteins were concentrated and desalted to an appropriate protein concentration by using Amicon Ultra centrifugal filters (Millipore). Recombinant protein was added to solutions at various concentrations with corresponding final salt and PEG 10000 in buffer (50 mM Tris, pH 7.2, 10% glycerol, and 1 mM DTT). The fusion proteins were loaded onto a homemade chamber slides and then imaged with a Leica TCS SP8 confocal microscope with a 63×/1.4 oil immersion objective.
FRAP assay.Fluorescence recovery after photobleaching (FRAP) was performed using a Leica TCS SP8 confocal microscope with 2.4-mW laser intensity for bleaching at room temperature, 63×/1.4 oil immersion objective, and photomultiplier tube detector. Considering the size of the droplets, different lenses and regions of interest (ROI) were chosen for FRAP experiments in buffer or cells. In each FRAP experiment, fluorescence intensities of another droplet of about the same size as the one used for photobleaching was recorded for fluorescence intensity correction, and a third region in the background with a similar size was also recorded for background signal subtraction. EGFP-labeled proteins were photobleached by 488-nm laser beams at 100% power. Each FRAP experiment involved 2 prebleach frames followed by 1 bleach frame, and recovery was monitored over 100 s. Five droplets were bleached at a time for each experiment. Each data point represented the averaged signal from five ROI with similar sizes. The intensity at the prebleach point was normalized to 100%. Data were expressed as means ± standard deviations (SD).
Statistical analysis.Statistical analysis was determined by independent t test or analysis of variance using SPSS17.0 and GraphPad Prism. Significance parameters were set at a P value of <0.05.
ACKNOWLEDGMENTS
We thank Ersheng Kuang for providing EBNA2 plasmid, Ya Cao and Jianhong Lu for providing reagents, and Jun Zhou and Jinru Xie (Xiangya Hospital, Central South University) for providing confocal microscopy technical assistance.
This work was supported by National Natural Science Foundation of China grants (81672889 and 81874170); China 111 Project Grant (111-2-12); Hunan Province Science and Technology Project Grant (2016JC2035); Graduate Research and Innovation Projects of Hunan Province (CX20190144); and the National College Students’ Innovation and Entrepreneurship Training Program of China (GS201910533237).
Q.P., Q.Y., and J.M. designed and conceived the experiments. Q.P., L.W., Z.Q., J.W., Xiang Zheng, L.W., Xiaoyue Zhang, Xuemei Zhang, C.L., Z.L., and Y.W. performed the experiments and analyzed the data. Q.P., G.L., Q.Y., and J.M. analyzed data. Q.P., Q.Y., and J.M. wrote the paper.
We have no conflicts of interest to declare.
FOOTNOTES
- Received 15 October 2019.
- Accepted 3 January 2020.
- Accepted manuscript posted online 15 January 2020.
Supplemental material is available online only.
- Copyright © 2020 American Society for Microbiology.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
