PML Alternative Splice Products Differentially Regulate HAdV Productive Infection

Abstract

Promyelocytic leukemia nuclear bodies (PML-NBs) were considered to maintain antiviral capacity, as these spherical complexes are antagonized by viruses. Actual work provides evidence, that PML-NB-associated factors might also be beneficial for distinct viral processes indicating why genomes and replication centers of nuclear replicating viruses are often found juxtaposed to PML-NBs. Several early HAdV proteins target PML-NBs, such as E4orf3 that promotes redistribution into track-like structures. PML-associated dependency factors that enhance viral gene expression, such as Sp100A remain in the nuclear tracks while restrictive factors, such as Daxx, are inhibited by either proteasomal degradation or relocalization to repress antiviral functions. Here, we did a comprehensive analysis of nuclear PML isoforms during HAdV infection. Our results show cell line specific differences as PML isoforms differentially regulate productive HAdV replication and progeny production. Here, we identified PML-II as a dependency factor that supports viral progeny production, while PML-III and PML-IV suppress viral replication. In contrast, we identified PML-I as a positive regulator and PML-V as a restrictive factor during HAdV infection. Solely PML-VI was shown to repress adenoviral progeny production in both model systems. We showed for the first time, that HAdV can reorganize PML-NBs that contain PML isoforms other then PML-II. Intriguingly, HAdV was not able to fully disrupt PML-NBs composed out of the PML isoforms that inhibit viral replication, while PML-NBs composed out of PML isoforms with beneficial influence on the virus formed tracks in all examined cells. In sum, our findings clearly illustrate the crucial role of PML-track formation in efficient viral replication. IMPORTANCE Actual work provides evidence that PML-NB-associated factors might also be beneficial for distinct viral processes indicating why genomes and replication centers of nuclear replicating viruses are often found juxtaposed to PML-NBs. Alternatively spliced PML isoforms I-VII are expressed from one single pml gene containing nine exons and their transcription is tightly controlled and stimulated by interferons and p53. Several early HAdV proteins target PML-NBs, such as E4orf3, promoting redistribution into track-like structures. Our comprehensive studies indicate a diverging role of PML isoforms throughout the course of productive HAdV infection in either stably transformed human lung (H1299) or liver (HepG2) cells, in which we observed a multivalent regulation of HAdV by all six PML isoforms. PML-I and PML-II support HAdV-mediated track formation and efficient formation of viral replication centers, thus promoting HAdV productive infection. Simultaneously, PML-III, -IV,-V, and -VI antagonize viral gene expression and particle production.

Keywords: HAdV; PML-NB; SUMO; human adenovirus.

FIG 1

HAdV-wt promotes PML isoform expression. (A) HepG2 shPML and EYFP-PML.n cells and (B) H1299 shPML and EYFP-PML.n cells were infected with HAdV-wt at a multiplicity of 50 FFU/cell or 20 FFU/cell, respectively. Cell viability was determined using the CellTiter-Blue Cell viability system and measured using a Tecan Infinite 200M plate reader 25 h and 48 h p.i. xy charts represent average values and standard deviations based on three independent experiments. (C) HepG2 shPML and EYFP-PML.n cells and (F) H1299 shPML and EYFP-PML.n cells were infected with HAdV-wt at a multiplicity of 50 FFU/cell or 20 FFU/cell, respectively. (C, F) Cells were harvested 24 h p.i. and total-cell lysates were prepared. Total-cell lysates were separated by SDS-PAGE and subjected to immunoblot analyses using pAb rabbit α-PML (anti-PML), MAb mouse B6-8 (anti-E2A), and MAb mouse AC-15 (anti-β-actin). Molecular weights in kDa are indicated on the left, relevant proteins on the right. Densitometric analyses of PML isoform levels in HepG2 shPML/EYFP-PML.n (D) and in H1299 shPML/EYFP-PML.n (G), quantified with ImageJ (version 1.51f) software and normalized to respective anti-β-actin levels. Bar charts represent average values and standard deviations based on three independent experiments for HepG2 cells and four independent experiments for H1299 cells. Statistically significant differences were assessed using two-way ANOVA and Šidák correction test with the GraphPad Prism5 software. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. Calculated expressions levels of PML isoforms in HAdV-wt infected cells were subtracted from the corresponding mock samples to illustrate PML protein level regulation by HAdV-wt. Bar charts represent delta values of the single PML isoforms in HepG2 (E) and H1299 (H) cells.

FIG 2

PML differentially affect HAdV-wt replication. (A) HepG2 shPML/EYFP-PML.n cells and (B, C, D, E) H1299 shPML/EYFP-PML.n cells were infected with HAdV-wt at a multiplicity of 50 FFU/cell or 20 FFU/cell, respectively, and harvested 24 h p.i. (A, B) Virus particles were isolated from infected cells. and viral progeny production was determined using quantitative immunofluorescence staining of E2A/DBP in HEK293 cells. Bar charts represent average values and standard deviations based on two independent experiments. Statistically significant differences were assessed using one-way ANOVA and Dunnett’s test with the GraphPad Prism5 software. *, P ≤ 0.05; **, P ≤ 0.01; ***, P≤ 0.001; ****, P ≤ 0.0001. (C, D, E) Total mRNA was isolated using TRIzol, reverse transcribed and analyzed by RT-qPCR using primers specific for HAdV E1A (C, D) and hexon (E). The data was normalized to the respective RPL30 mRNA levels. Bar charts represent average values and standard deviations based on two independent experiments measured in technical duplicates. Statistically significant differences were assessed using an unpaired Student’s t test with the GraphPad Prism9 software. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 3

Expression of early viral proteins and the cement protein pVI differs after expression of single PML isoforms. HepG2 shPML/EYFP-PML.n and H1299 shPML/EYFP-PML.n cells were infected with HAdV-wt at a multiplicity of 50 FFU/cell or 20 FFU/cell, respectively, and harvested 24h p.i. Total-cell lysates were prepared and proteins were separated using SDS-PAGE and detected via immunoblotting using M73 (anti-E1A), 2A6 (anti-E1B-55K), B6-8 (anti-E2A), RSA3 (anti-E4orf6), 6A11 (anti-E4orf3), anti-pVI, anti-IVa2, and AC-15 (anti-β-actin). Molecular weights in kDa are indicated on the left, relevant proteins on the right. (B) Densitometric analyses of detected protein levels quantified with ImageJ (version 1.51f) software and normalized to respective α-β-actin steady state levels. Bar charts represent average values and standard deviations based on three independent experiments in HepG2 cells and on four independent experiments for H1299 cells for (A) E1A, (B) E1B-55K, (C) E2A, (D) E4orf6, (E) E4orf6/7, (F) E4orf3, (G) pVI, and (H) IVa2. Statistically significant differences were assessed using an unpaired Student's t test with the GraphPad Prism9 software. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 4

PML isoforms modulate late viral protein production. HepG2 shPML/EYFP-PML.n and H1299 shPML/EYFP-PML.n cells were infected with HAdV-wt at a multiplicity of 50 FFU/cell or 20 FFU/cell, respectively, and harvested 24h p.i. Total-cell lysates were prepared and proteins were separated using SDS-PAGE and detected via immunoblotting using anti-L4-33K, anti-L4-100K, L133 (anti-Capsid) and AC-15 (anti-β-actin). Densitometric analyses of detected protein levels quantified with ImageJ (version 1.51f) software and normalized to respective α-β-actin steady state levels. Bar charts represent average values and standard deviations based on three independent experiments in HepG2 cells and on four independent experiments for H1299 cells for (A) L4-33K, (B) L4-100K, (C) Capsid and (D) Capsid-associated. Statistically significant differences were assessed using an unpaired Student’s t test with the GraphPad Prism9 software. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 5

PML track and viral replication center formation differs among PML isoforms in HepG2 cells. (A, B) HepG2 shPML/EYFP-PML.n cells were either mock infected (A) or infected with HAdV-wt at a multiplicity of 50 FFU/cell (B). Cells were fixed with 4% paraformaldehyde (PFA) 24 h p.i. and double-labeled with MAb rabbit anti-GFP (PML isoforms) and MAb mouse B6-8 (anti-E2A). Primary Abs were detected with Alexa Fluor-488 anti-mouse (anti-GFP; EYFP-PML.n, green)- and Alexa Fluor-647 (E2A, red)-conjugated secondary Abs. Nuclear staining was performed using DAPI (4’,6-diamidino-2-phenylindole). Anti-GFP (green; panels b, f, j, n, r, v, z) and anti-E2A (red; panels c, g, k, o, s, w, a1) staining patterns representative of at least 50 analyzed cells are shown. Overlays of single images (merge) are shown in panels d, h, l, p, t, x, b1. (C, D) The formation of HAdV replication centers was monitored in at least 50 infected HepG2 shPML/EYFP-PML.n cells. (C) The bar chart represents the percentage of infected cells with replication centers as opposed to infected cells containing only diffuse E2A distribution. P values were calculated employing a Fisher exact test using the R language and environment for statistical computing. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. (D) Pie charts represent percentage of infected HepG2 shPML/EYFP-PML.n cells containing specific type of replication centers. (E) PML-NB distribution into tracks was monitored in at least 50 infected HepG2 shPML/EYFP-PML.n cells. The bar chart shows the percentage of infected HepG2 shPML/EYFP-PML.n cells that were reorganized (tracks) and the percentage of infected HepG2 shPML/EYFP-PML.n cells without reorganization of PML-NBs (dots).

FIG 6

Disruption of PML-NBs and establishment of viral replication centers is regulated by PML isoform expression in H1299 cells. (A, B) H1299 shPML/EYFP-PML.n cells were either mock infected (A) or infected with HAdV-wt at a multiplicity of 20 FFU/cell (B). Cells were fixed with 4% paraformaldehyde (PFA) 24 h p.i. and double-labeled with MAb rabbit α-GFP (EYFP-PML.n) and MAb mouse B6-8 (anti-E2A). Primary Abs were detected with Alexa Fluor-488 α-mouse (α-GFP; EYFP-PML.n, green)- and Alexa Fluor-647 (E2A, red)-conjugated secondary Abs. Nuclear staining was performed using DAPI (4’,6-diamidino-2-phenylindole). Anti-GFP (green; panels b, f, j, n, r, v, z) and anti-E2A (red; panels c, g, k, o, s, w, a1) staining patterns representative of at least 50 analyzed cells are shown. Overlays of single images (merge) are shown in panels d, h, l, p, t, x, b1. (C, D) The formation of HAdV replication centers was monitored in at least 50 infected H1299 shPML/EYFP-PML.n cells. (C) The bar chart represents the percentage of infected cells with replication centers as opposed to infected cells containing only diffuse E2A distribution. P values were calculated employing a Fisher exact test using the R language and environment for statistical computing. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. (D) Pie charts represent percentage of infected H1299 shPML/EYFP-PML.n cells containing specific type of replication centers. (E) PML-NB distribution into tracks was monitored in at least 50 infected H1299 shPML/EYFP-PML.n cells. The bar chart shows the percentage of infected H1299 shPML/EYFP-PML.n cells that were reorganized (tracks) and the percentage of infected H1299 shPML/EYFP-PML.n cells without reorganization of PML-NBs (dots).