An Essential Viral Transcription Activator Modulates Chromatin Dynamics

Abstract

Although ICP4 is the only essential transcription activator of herpes simplex virus 1 (HSV-1), its mechanisms of action are still only partially understood. We and others propose a model in which HSV-1 genomes are chromatinized as a cellular defense to inhibit HSV-1 transcription. To counteract silencing, HSV-1 would have evolved proteins that prevent or destabilize chromatinization to activate transcription. These proteins should act as HSV-1 transcription activators. We have shown that HSV-1 genomes are organized in highly dynamic nucleosomes and that histone dynamics increase in cells infected with wild type HSV-1. We now show that whereas HSV-1 mutants encoding no functional ICP0 or VP16 partially enhanced histone dynamics, mutants encoding no functional ICP4 did so only minimally. Transient expression of ICP4 was sufficient to enhance histone dynamics in the absence of other HSV-1 proteins or HSV-1 DNA. The dynamics of H3.1 were increased in cells expressing ICP4 to a greater extent than those of H3.3. The dynamics of H2B were increased in cells expressing ICP4, whereas those of canonical H2A were not. ICP4 preferentially targets silencing H3.1 and may also target the silencing H2A variants. In infected cells, histone dynamics were increased in the viral replication compartments, where ICP4 localizes. These results suggest a mechanism whereby ICP4 activates transcription by disrupting, or preventing the formation of, stable silencing nucleosomes on HSV-1 genomes.

Fig 1. The dynamics of linker and core histones are only minimally altered in the absence of functional ICP4.

U2OS (A) or Vero (B) cells were transfected with plasmids expressing GFP fused to H2A, H2B, H3.1, H3.3, H4, or H1.2. Transfected cells were mock infected or infected with 30 plaque forming units (PFU) per cell of HSV-1 strain n12 and histone dynamics were examined from 4 to 5 or 7 to 8 hours post infection (hpi) (4 hpi or 7 hpi, respectively) by FRAP. Frequency distribution plots showing the percentage of free GFP-H2A, -H2B, -H3.1, -H3.3, -H4, or -H1.2 per individual mock- (dashed line) or n12 (solid line) infected cell at 4 or 7 hpi. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 20 cells from at least 3 independent experiments, except for U2OS H2A (n = 20 cells from 2 independent experiments).

Fig 2. Core and linker histone dynamics during infection with wild-type or mutant HSV-1 strains defective in VP16, ICP0 or ICP4.

U2OS or Vero cells were transfected with plasmids expressing GFP fused to H2A, H2B, H3.1, H3.3, H4, or H1.2. Transfected cells were mock-infected or infected with 30 PFU per cell of HSV-1 strain KOS, n212, KM110, or n12, or 6 PFU per cell of strain KOS (U2OS cells). Histone dynamics were evaluated from 4 to 5 (4 hpi) or 7 to 8 (7 hpi) hpi by FRAP. A), B) Bar graphs showing the average levels of free GFP-H2A, -H2B, -H3.1, -H3.3, -H4, or -H1.2 relative to those in mock-infected cells (set at 1.0) at 4 or 7 hpi. C), D) Bar graphs showing the average initial rates of normalized fluorescence recovery (core histones) or the average T₅₀ (H1.2) relative to those in mock-infected cells (set at 1.0) at 4 or 7 hpi. Error bars, SEM. (•) Data for GFP-H2B, -H3.1, -H3.3, and -H1.2 for HSV-1 strains KOS, n212, and KM110 (Vero and U2OS) and GFP-H4 for HSV-1 strain KOS (Vero), included for comparison, are from previously published experiments [–40].

Fig 3. Functional ICP4 enhances histone dynamics during n12 infection.

n-33 cells transfected with plasmids expressing GFP fused to H2A, H2B, H3.3, H4, or H1.2 were mock infected or infected with 30 PFU per cell of HSV-1 strain KOS (☐) or n12 (■). Histone dynamics were evaluated from 4 to 5 (4 hpi) or 7 to 8 (7 hpi) hpi by FRAP. A) Bar graphs showing the average levels of free GFP-H2A, -H2B, -H3.3, -H4, or -H1.2 in KOS- or n12- infected cells relative to those in mock-infected cells (set at 1.0) at 4 or 7 hpi. B) Bar graphs showing the average initial rates of normalized fluorescence recovery (core histones) or the average T₅₀ (H1.2) in KOS- or n12- infected cells relative to those in mock-infected cells (set at 1.0) at 4 or 7 hpi. Error bars, SEM. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 15 cells from at least 2 independent experiments, except GFP-H2A and -H4 n ≥ 8 cells from 1 experiment.

Fig 4. The dynamics of H2B and H4, representative of each histone dimer, are enhanced in cells transiently expressing ICP4.

Vero cells were co-transfected with plasmids expressing GFP-H2B or -H4 and RFP-ICP4 or RFP, such that approximately half of the cells expressing detectable levels of GFP-histone also express detectable levels of RFP. A), D) Average fluorescence recovery curves for GFP-H2B and -H4, respectively, for cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. B), E) Bar graphs showing the average levels of free GFP-H2B or -H4, respectively, in cells expressing detectable levels of RFP-ICP4 or RFP relative to those in cells expressing undetectable levels of RFP-ICP4 or RFP, respectively. C), F) Bar graphs showing average GFP-H2B or -H4 slow exchange rate in cells expressing detectable levels of RFP-ICP4 or RFP relative to those in cells expressing undetectable levels of RFP-ICP4 or RFP. Error bars, SEM. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 15 cells from at least 3 independent experiments.

Fig 5. H3.1 dynamics are enhanced more than those of H3.3 in Vero cells transiently expressing ICP4.

Vero cells were co-transfected with plasmids expressing GFP-H3.1 or -H3.3 and RFP-ICP4 or RFP, such that approximately half of the cells expressing detectable levels of GFP also express detectable levels of RFP. A), C) Average fluorescence recovery curves for GFP-H3.3 and -H3.1, respectively, for cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. B), D) Bar graphs showing the average levels of free GFP-H3.3 or -H3.1, respectively, in cells expressing detectable levels of RFP-ICP4 or RFP relative to cells expressing undetectable levels of RFP-ICP4 or RFP, respectively. E) Frequency distribution of the free pool of GFP-H3.3 in cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. F) Frequency distribution of the free pool of GFP-H3.3 in cells expressing detectable (dark red line) or undetectable (green line) levels of RFP. G) Frequency distribution of the free pool of H3.1 in cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. H) Frequency distribution of the free pool of GFP-H3.1 in cells expressing detectable (dark red line) or undetectable (green line) levels of RFP. I) Representative images of fluorescent nuclei expressing GFP-H3.1 and detectable or undetectable levels of RFP-ICP4, immediately prior to (T = 0) or after (T = 1) photobleaching, or 200 seconds later. Error bars, SEM. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 15 cells from at least 3 independent experiments.

Fig 6. H3.1 dynamics are enhanced more than those of H3.3 in U2OS cells transiently expressing ICP4.

U2OS cells were co-transfected with plasmids expressing GFP-H3.1 or -H3.3 and RFP-ICP4 or RFP such that approximately half of the cells expressing detectable levels of GFP also express detectable levels of RFP. A), C) Average fluorescence recovery curves for GFP-H3.3 and -H3.1, respectively, for cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. B), D) Bar graphs showing the average levels of free GFP-H3.3 or -H3.1, respectively, in cells expressing detectable levels of RFP-ICP4 or RFP relative to those in cells expressing undetectable levels of RFP-ICP4 or RFP. E) Frequency distribution of the free pool of GFP-H3.3 in cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. F) Frequency distribution curve of the free pool of GFP-H3.3 in cells expressing detectable (dark red line) or undetectable (green line) levels of RFP. G) Frequency distribution of the free pool of GFP-H3.1 in cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. H) Frequency distribution curve of the free pool of GFP-H3.1 in cells expressing detectable (dark red line) or undetectable (green line) levels of RFP. I) Representative images of fluorescent nuclei expressing GFP-H3.1 and detectable or undetectable levels of RFP-ICP4, immediately prior to (T = 0) or 1 second after (T = 1) photobleaching, or 200 seconds later. Error bars, SEM. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 15 cells from at least 3 independent experiments.

Fig 7. The dynamics of H1.2, but not those of canonical H2A, are enhanced in cells transiently expressing ICP4.

Vero cells were co-transfected with plasmids expressing GFP-H2A or -H1.2 and RFP-ICP4 or RFP such that approximately half of the cells expressing detectable levels of GFP also express detectable levels of RFP. A), C) Average fluorescence recovery curves for GFP-H2A and -H1.2, respectively, for cells expressing detectable (red line) or undetectable (green line) levels of RFP-ICP4. B), D) Bar graphs showing the average levels of free GFP-H2A or -H1.2, respectively, in cells expressing detectable levels of RFP-ICP4 or RFP relative to those in cells expressing undetectable levels of RFP-ICP4 or RFP, respectively. E) Bar graphs showing the average T₅₀ of GFP-H1.2 in cells expressing detectable levels of RFP-ICP4 or RFP relative to those in cells expressing undetectable levels of RFP-ICP4 or RFP. Error bars, SEM. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 15 cells from at least 3 independent experiments.

Fig 8. The dynamics of no histone is altered in cells transiently expressing the truncated, non-functional, ICP4 mutant n12.

Vero cells were co-transfected with plasmids expressing GFP-histones and RFP-n12 or RFP such that approximately half of the cells expressing detectable levels of GFP also express detectable levels of RFP-n12. A-E), G) Bar graphs showing average levels of free GFP-H1.2, -H2A, -H2B, -H3.3, -H4, and -H3.1 in cells expressing detectable levels of RFP-n12 or RFP relative to those in cells expressing undetectable levels of RFP-n12 or RFP, respectively. F) Average fluorescence recovery curves for GFP-H3.1 in cells expressing detectable (orange line) or undetectable (green line) levels of RFP-n12. H) Distribution curve of the free pool of GFP-H3.1 in cells expressing detectable (orange line) or undetectable (green line) levels of RFP-n12. I) Distribution curve of the free pool of GFP-H3.1 in cells expressing detectable (dark red line) or undetectable (green line) levels of RFP. Error bars, SEM. **, P < 0.01; *, P < 0.05; n.s., not significant. n ≥ 15 cells from at least 3 independent experiments.

Fig 9. The majority of ICP4 localizes in the replication compartments with a small pool of histones.

Digital fluorescent micrographs showing Vero cells expressing GFP-H2B, infected with 30 PFU of HSV-1 strain KOS, and stained with anti-ICP4 antibodies. Cells were fixed at 7 hpi and immunostained for ICP4. Single channel and merged images are shown. Note the presence of a small pool of GFP-H2B in the replication compartments, similar localization has already been reported for GFP-H1.2, -H2B, -H3.1, and -H3.3 [–40].

Fig 10. Fluorescence recovery of GFP-tagged histones in HSV-1 replication compartments or cellular chromatin.

A) Fluorescence micrographs of the nucleus of an HSV-1 infected Vero cell expressing GFP-H1.2 and undergoing FRAP at 7 to 8 hpi. Left and middle micrographs, immediately prior (pre) or after photobleaching (T = 0), respectively; right micrograph, 60 seconds after photobleaching. White downward arrowhead and white circle, replication compartment region to be photobleached; black upward arrowhead and black circle, cellular chromatin region to be photobleached. Note the presence of a small pool of H1.2 in the replication compartments. B) Line graphs presenting the average ± SEM fluorescence recovery curves of histones GFP-H3.1 (n = 10), GFP-H3.3 (n = 13), GFP-H4 (n = 14), GFP-H2A (n = 10), GFP-H2B (n = 13), and GFP-H1.2 (n = 10) in the replication compartments or cellular chromatin at 7 to 8 h of infection with HSV-1, strain KOS. C) Dot plot presenting the free pools in the replication compartments or the cellular chromatin in each individual cell. Solid lines, same cells.