Phase Separation of Epstein-Barr Virus EBNA2 and Its Coactivator EBNALP Controls Gene Expression

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.

Keywords: EBNA2; EBNALP; Epstein-Barr virus; phase separation; superenhancer.

FIG 1

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.

FIG 2

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.

FIG 3

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.

FIG 3

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.

FIG 4

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.

FIG 5

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.

FIG 6

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).

FIG 7

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.

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.