Characterization of Recombinant Human Cytomegaloviruses Encoding IE1 Mutants L174P and 1-382 Reveals that Viral Targeting of PML Bodies Perturbs both Intrinsic and Innate Immune Responses

Oligonucleotides and plasmids.

The oligonucleotide primers used for this study were purchased from Biomers GmbH (Ulm, Germany) or Biomol GmbH (Hamburg, Germany) and are listed in Table 1. The eukaryotic expression plasmid pHM3260 encoding IE1-L174P was generated by site-directed mutagenesis of plasmid pHM494, which encodes wild-type IE1 (47), using primers 5′IE1mut-L174P and 3′IE1mut-L174P. Expression plasmids encoding IE1 and IE1-L174P fused to a C-terminal CLIP-tag were obtained by PCR amplification of the respective IE1 sequence using pHM494 or pHM3260 as the template along with primer pair 5′IE1_NheI and 3′IE1_EcoRI. Subsequently, PCR products were inserted into MCS1 of pCLIP_f (New England BioLabs, Frankfurt, Germany) via NheI and EcoRI. The prokaryotic expression plasmid encoding glutathione S-transferase (GST)-tagged IE1 was described elsewhere (6). The prokaryotic expression plasmid encoding GST-tagged IE1-L174P was constructed by PCR using pHM3260 as the template and primers 5′IE1-GEX-Bam and 3′IE1_Xho1, followed by ligation into pGEX-6P-1 (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) via BamHI and XhoI. The lentiviral pLVX-shRNA1-based vectors encoding control short hairpin RNA (shRNA) and PML shRNA were described previously (48). Lentiviral vectors that are based on pLVX-Tight-Puro and encode IE1 or IE1 1-382, both fused to a FLAG-tag, were generated by PCR amplification with primer pairs 5′Flag_IE1_BamHI and 3′IE1_EcoRI_stop or 5′Flag_IE1_BamHI and 3′IE1aa382_EcoRI along with pHM494 as the template, followed by insertion into pLVX-Tight-Puro via BamHI and EcoRI.

Cell culture, infections, and transfections.

Primary human foreskin fibroblasts (HFFs) were prepared from human foreskin tissue and were maintained in Eagle's minimal essential medium (Gibco/BRL, Eggenstein, Germany) supplemented with 5% fetal calf serum. HFFs with a small interfering RNA-mediated knockdown of PML (siPML) and control HFFs (siC) were cultured in Eagle's minimal essential medium (Gibco/BRL, Eggenstein, Germany) supplemented with 7% fetal calf serum and 5 μg/ml puromycin. HFFs with inducible expression of IE1 or IE1 1-382 were cultured in Eagle's minimal essential medium (Gibco/BRL, Eggenstein, Germany) supplemented with 10% tetracycline-free fetal bovine serum (Clontech, Palo Alto, CA, USA), 5 μg/ml puromycin, and 500 μg/ml Geneticin. Infection experiments were performed with the HCMV laboratory strain AD169 and recombinant viruses AD169/IE1 1-382 (6), AD169/IE1-L174P, and AD169ΔIE1. For this purpose, HFF cells were seeded into six-well dishes at a density of 3 × 10⁵ cells/well. One day later, cells were incubated with 1 ml viral supernatant and, after 1.5 h, provided with fresh medium before they were used for subsequent analyses.

HEK293T cells were cultivated in Dulbecco's minimal essential medium (Gibco/BRL, Eggenstein, Germany) containing 10% fetal calf serum. Transfection of HEK293T cells was performed by applying the standard calcium phosphate precipitation method. For this, HEK293T cells were seeded into six-well dishes (6 × 10⁵ cells/well) 1 day before transfection. For each transfection reaction, 1 to 2 μg of plasmid DNA was used. At about 16 h after transfection, the cells were washed two times with phosphate-buffered saline (PBS) and provided with fresh medium before they were used for further analyses.

Generation of stable HFF cells by lentivirus transduction.

In order to establish HFFs stably expressing the siRNAs siC and siPML, the Lenti-X shRNA system was utilized (Clontech, Palo Alto, CA, USA). For generation of replication-deficient lentiviruses, HEK293T cells were seeded in 10 cm dishes (5 × 10⁶ cells/dish) and cotransfected with pLVX-shRNA1 vectors containing control shRNA or PML shRNA and the Lenti-X HTX packaging mix (Clontech, Palo Alto, CA, USA) using the Lipofectamine 2000 reagent (Invitrogen, Karlsruhe, Germany). Viral supernatants were harvested 48 h after transfection, cleared by centrifugation, filtered, and stored at −80°C. Primary HFFs were incubated for 24 h with lentivirus supernatants in the presence of 7.5 μg/ml Polybrene (Sigma-Aldrich, Deisenhofen, Germany), and stably transduced cell populations were selected by adding 5 μg/ml puromycin to the cell culture medium.

For generation of HFFs with inducible expression of IE1 or IE1 1-382, the Lenti-X Tet-on Advanced Inducible Expression System was utilized (Clontech, Palo Alto, CA, USA). For generation of replication-deficient lentiviruses expressing the tetracycline-controlled transactivator rtTA-Advanced, HEK293T cells were cotransfected with the pLVX-Tet-on Advanced vector and the Lenti-X HTX packaging mix (Clontech, Palo Alto, CA, USA). For generation of lentiviruses expressing IE1 or IE1 1-382 under the control of the inducible _pTight promoter, HEK293T cells were cotransfected with pLVX-Tight-Puro vectors containing IE1 or IE1 1-382 and the Lenti-X HTX packaging mix (Clontech, Palo Alto, CA, USA). Viral supernatants were harvested 48 h after transfection, cleared by centrifugation, filtered, and stored at −80°C. Afterwards, primary HFFs were successively transduced with lentiviruses derived from the pLVX-Tet-On Advanced vector and lentiviruses derived from pLVX-Tight-Puro-IE1/IE1 1-382 vectors, followed by selection with 500 μg/ml Geneticin and 5 μg/ml puromycin, respectively.

Generation and reconstitution of recombinant cytomegaloviruses.

The wild-type HCMV bacterial artificial chromosome (BAC) HB15, which is based on the laboratory strain AD169, was manipulated by recombination-based genetic engineering in order to obtain AD169/IE1-L174P and AD169ΔIE1. AD169/IE1-L174P was generated by mutating the leucine residue at position 174 (CTG) in exon 4 of IE1 to proline (CCG). For this purpose, we utilized the two-step red-mediated recombination technique described by Tischer et al. (49), which uses the kanamycin gene as a first selection marker. The linear recombination fragment was generated by PCR using primers 5′BAC_short and 3′BAC-L174P and pEPkanS as the template DNA (kindly provided by K. Osterrieder, Berlin, Germany). The PCR product was treated with DpnI, gel purified, and subjected to a second round of PCR amplification using primers 5′BAC-L174P and 3′BAC-L174P_short. For homologous recombination, the PCR fragment was transformed into Escherichia coli strain GS1783 (a gift of M. Mach, Erlangen) already harboring HB15, and two-step bacteriophage λ red-mediated recombination was performed as described elsewhere (49). To identify positive transformants, the bacteria were plated on agar plates containing kanamycin (first recombination) or chloramphenicol and 1% arabinose (second recombination) and incubated at 32°C for 2 days. BAC DNA was purified from bacterial colonies growing on these plates, and successful recombination was confirmed by PCR, restriction enzyme digestion, and direct sequencing.

AD169ΔhIE1 was generated by deleting exon 4 of IE1 as described previously (6). Briefly, a linear recombination fragment, comprising a kanamycin resistance marker and 5′ and 3′ genomic sequences, was generated by PCR amplification using pKD13 as the template and primers 5′Intron3/pKD13 and 3′Exon4/pkd13. This fragment was used for electroporation of competent Escherichia coli strain DH10B harboring HB15, and homologous recombination was performed as described previously (50). The integrity of the resulting recombinant BAC was confirmed by PCR, restriction enzyme digestion, and direct sequencing.

HFFs stably expressing IE1, which were generated as described elsewhere (6), were utilized for reconstitution of AD169/IE1-L174P and AD169ΔIE1. The cells were seeded in six-well dishes (3 × 10⁵ cells/well) and cotransfected with purified BAC DNA (1 μg), pp71 expression plasmid pCB6-pp71 (0.5 μg) and a vector encoding the Cre recombinase (0.5 μg) using FuGENE6 transfection reagent (Promega, Mannheim, Germany). Transfected HFFs were propagated until viral plaque formation was observed, and the supernatants from these cultures were used for further virus propagation in IE1-expressing HFFs.

Titration of virus stocks.

Titers of wild-type AD169 were determined by IE1 fluorescence. For this purpose, HFFs were infected with various dilutions of virus stocks. After 24 h of incubation, cells were fixed and stained with monoclonal antibody p63-27, directed against IE1 (51). Subsequently, the number of IE1-positive cells was determined by quantifying IE1 expression in >200 cells using ImageJ software. Results were used to calculate viral titers, expressed as IE protein-forming units (IEU). To ensure infection with equal genome copy numbers of wild-type and recombinant viruses, AD169/IE1-L174P and AD169ΔIE1 were titrated via determination of genome equivalents in virus stocks and in infected cells by quantitative real-time PCR. Samples of virus stocks were subjected to treatment with proteinase K, and TaqMan-PCR was performed as described below in order to determine the amount of the gB gene corresponding to HCMV genome equivalents. For determination of genome copies having entered the cells, HFFs were infected with various dilutions of AD169 and recombinant viruses. Sixteen hours later, viral and cellular DNA was extracted from infected cells using the DNeasy blood and tissue kit (Qiagen, Hilden, Germany) and subjected to quantitative real-time PCR (TaqMan-PCR) using HCMV gB-specific primers and primers specific for the cellular albumin gene as internal control. Results were used to calculate the number of intracellular HCMV genomes resulting from infection with 1 ml of input virus.

Multistep growth curve analysis.

HFF cells were seeded into six-well dishes at a density of 3 × 10⁵ cells/well and infected the following day with wild-type AD169 at a multiplicity of infection (MOI) of 0.1 and equivalent genome copy numbers of AD169/IE1-L174P and AD169ΔIE1. Triplicate samples of infected cell supernatants were harvested at 2, 4, 6, 8, 10, and 13 days after inoculation and lysed by proteinase K treatment. Afterwards, quantitative real-time PCR (TaqMan-PCR) of the HCMV gB gene was conducted as described below to analyze the amount of viral genome copies in the supernatants.

Quantitative TaqMan real-time PCR.

For quantitative real-time PCR (TaqMan-PCR), an Applied Biosystems 7500 real-time PCR system was utilized (Applied Biosystems, Foster City, CA, USA) together with the corresponding software SDS (sequence detection system) (52). To quantify viral DNA, a sequence region within the HCMV gB gene region was amplified using primers 5′gB_forw and 3′gB_rev along with the fluorescence labeled probe CMV gB FAM/TAMRA, also directed against the gB gene locus. When intracellular genomes where examined, quantification of cellular albumin genes was conducted in parallel using primers 5′Alb and 3′Alb together with a fluorescence-labeled probe, Alb FAM/TAMRA (6-carboxyfluorescein/6-carboxytetramethylrhodamine). For determination of reference C_T values (cycle threshold), serial dilutions of the respective standards (10⁷ to 10¹ DNA molecules of gB or albumin) were analyzed by PCRs in parallel. The 20-μl reaction mixture contained 5 μl sample or standard DNA solution together with 10 μl 2× TaqMan PCR Mastermix (Applied Biosystems, Foster City, CA, USA), 1.5 μl of each primer (5 μM stock solution), 0.4 μl of probe (10 μM stock solution), and 1.6 μl of H₂O. The thermal cycling conditions consisted of two initial steps of 2 min at 50°C and 10 min at 95°C followed by 40 amplification cycles (15 s at 95°C, 1 min at 60°C). The viral genome copy numbers and albumin copy numbers were subsequently calculated using the sample-specific C_T value when set into relation to the standard serial dilutions.

RNA isolation and quantitative SYBR green reverse transcription-PCR (qRT-PCR).

Total RNAs from 3 × 10⁵ to 6 × 10⁵ infected or transduced HFFs were isolated in triplicate using TRIzol reagent according to the manufacturer's instructions (Invitrogen, Karlsruhe, Germany). To exclude any DNA contaminations, 10 units of RNase-free DNase I (Roche, Mannheim, Germany) together with 20 units of RNaseOUT RNase Inhibitor (Invitrogen, Karlsruhe, Germany) were added to each sample and incubated at 37°C for 45 min, followed by DNase I inactivation at 75°C for 10 min. RNAs were further purified by phenol extraction and ethanol precipitation, and the RNA pellets were resuspended in RNase-free H₂O. Afterwards, first-strand cDNAs were synthesized from 1 μg of total RNA using oligo(dT)₁₈ and random hexamer primers provided by the Maxima First Strand cDNA synthesis kit (Thermo Scientific, Waltham, MA, USA). First-strand cDNAs were diluted 4-fold with DNase-free water, and 2 μl was used as the template for quantitative real-time PCR mixtures containing 12.5 μl Maxima SYBR green/Rox qPCR master mix (Thermo Scientific, Waltham, MA, USA), 10.35 μl nuclease-free H₂O, and 0.15 μl of the following primer pairs: CMV 5′ (IE1) and CMV 3′ (IE1), PML-fw and PML-rev, MX1-fw and MX1-rev, TNFSF10-fw and TNFSF10-rev, OASL-fw and OASL-rev, IFIT3/4-fw and IFIT3/4-rev, CCL8-fw and CCL8-rev, IDO1-fw and IDO1-rev, 5′hISG54 and 3′hISG54, HLA-DRA-fw and HLA-DRA-rev, and UIMC-fw and UIMC-rev. For normalization of the qPCR values, GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was amplified as a housekeeping gene using primers huGAPDH_TM_for and huGAPDH_tm_rev. SYBR green PCR was conducted in 96-well plates compatible with the ABI Prism 7500 real-time PCR system (Applied Biosystems, Foster City, CA, USA) together with the corresponding software SDS (sequence detection system). The initial denaturation step (95°C, 10 min) was subsequently followed by 40 cycles of denaturation (95°C, 15 s), primer annealing (60°C, 35 s), and strand elongation, followed by a dissociation stage to ascertain the specificity of the reactions. Data were analyzed using the Applied Biosystems SDS software, and relative quantification of mRNA levels was performed.

CLIP labeling.

For labeling of CLIP-tagged IE1 proteins, transfected HEK293T cells were incubated with 3 μM red fluorescent substrate CLIP-Cell TMR STAR in cell culture medium (Dulbecco's minimal essential medium supplemented with 10% fetal calf serum) for 30 min at 37°C. Afterwards, the cells were washed three times with cell culture medium and incubated for another 30 min at 37°C before they were washed again. The cells were pooled and reseeded into 12-well plates to ensure equal expression and labeling efficiency in every sample. Cell lysates were prepared immediately (0 h) or at later times (2 h to 30 h) by resuspending cell pellets in 2× sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer, boiling at 95°C for 5 min, and sonication. Labeled proteins were separated on SDS-containing 8% polyacrylamide gels and visualized by in-gel fluorescence scanning using an INTAS Advanced Fluorescence Imager (INTAS, Göttingen, Germany).

Western blotting, indirect immunofluorescence, and antibodies.

For Western blot analysis, total lysates from transfected or infected cells were prepared in a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer, separated on SDS-containing 8 to 12% polyacrylamide gels, and transferred to nitrocellulose membranes. Chemoluminescence detection was performed according to the manufacturer's instructions (ECL Detection kit; Amersham Pharmacia Biotech Europe, Freiburg, Germany).

Indirect immunofluorescence analysis of infected HFF cells was performed by fixation of infected cells with 4% paraformaldehyde and fluorescence staining as described elsewhere (6). Samples were analyzed using a Leica TCS SP5 confocal microscope with 488-nm, 543-nm, and 633-nm laser lines, scanning each channel separately under image capture conditions that eliminated channel overlap. The images were exported, processed with Adobe Photoshop CS5, and assembled using CorelDraw X5.

The following monoclonal antibodies were used for immunofluorescence and Western blot analyses: α-IE1 p63-27 (51), α-UL44 BS510 (kindly provided by B. Plachter, Mainz, Germany), α-MCP 28-4 (53), α-FLAG 1804 (Sigma-Aldrich, Deisenhofen, Germany), and α-β-actin AC-15 (Sigma-Aldrich, Deisenhofen, Germany). The following polyclonal antibodies were used for immunofluorescence and Western blot analyses: α-IE2 pHM178 (produced by immunizing rabbits with prokaryotically expressed protein); α-PML #4 (a kind gift from P. Hemmerich, Jena, Germany), used for immunofluorescence analysis; α-PML A301-167A (Bethyl Laboratories, Montgomery, TX, USA) in combination with α-PML A301-168A (Bethyl Laboratories), applied for Western blot analysis; α-Sp100 GH3 (kindly provided by H. Will, Hamburg, Germany); and α-STAT2 H190, obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Alexa Fluor 488-/555-conjugated secondary antibodies for indirect immunofluorescence experiments were purchased from Life technologies (Karlsruhe, Germany), and horseradish peroxidase-conjugated anti-mouse/anti-rabbit secondary antibodies for Western blot analysis were obtained from Dianova (Hamburg, Germany).

Recombinant protein production and purification.

Recombinant IE1 and IE1-L174P were expressed in E. coli and purified as described elsewhere (6). Briefly, LB medium was inoculated with E. coli colonies freshly transformed with pGEX-6P-1-based expression plasmids encoding GST-tagged IE1 or IE1-L174P. Cell cultures were grown at either 30°C or 37°C, and expression of recombinant proteins was induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) for 4 to 6 h. Cell pellets were resuspended in phosphate buffer and lysed by sonication. Protein purification was achieved by GST affinity chromatography, proteolytic cleavage with PreScission protease, a second GST affinity chromatography, and size exclusion chromatography using a Superdex 200 prepgrade column (GE Healthcare Bio-Sciences AB, Uppsala, Sweden).

Bioinformatic analyses.

Modeling of the L174P mutation was performed with SwissModel, and the lowest-energy conformer of P174 was selected (54). WhatCheck was used for structure analysis and VMD for visualization (55, 56).

Insertion of L174P mutation into IE1 of HCMV results in a severe growth defect comparable to a deletion of IE1.

After entry of HCMV genomes into the nucleus, the immediate early protein IE1 disrupts the subnuclear, antiviral structure ND10 in order to initiate lytic HCMV infection. Previous studies have demonstrated that a leucine-to-proline mutation (L174P) within the globular core domain of IE1 (IE1_CORE) abolishes the interaction with the major ND10 protein PML and, consequently, abrogates the dispersal of ND10 in transient-transfection assays (38–40). To investigate the impact of this mutation on lytic HCMV infection, we generated a recombinant HCMV expressing IE1-L174P, referred to as AD169/IE1-L174P. As illustrated in Fig. 1A, the L174P mutation was introduced into the IE1 gene of HCMV laboratory strain AD169 using a two-step, markerless BAC recombination technique (49). In addition, a recombinant virus harboring a deletion of IE1 (AD169ΔIE1), which was generated as previously described (6), was included in this study, allowing comparative analysis of replication defects caused by IE1 mutation and deletion. Since it is known that IE1-deleted viruses display a nonlinear relationship between the multiplicity of infection (MOI) and the extent of virus growth (3, 8), titration of the newly generated viruses was performed by real-time PCR quantification of genome equivalents in viral supernatants and in infected cells. Figure 1B shows that considerably more viral genomes were detected in virus stocks of AD169/IE1-L174P and AD169ΔIE1 than in those of wild-type AD169, and a similar ratio of genome copies was determined in infected primary human foreskin fibroblast (HFF) cells. In contrast, lytic replication was initiated in significantly more cells infected with wild-type AD169 than in cells infected with AD169/IE1-L174P or AD169ΔIE1 using the same amount of input virus, illustrating the silencing of viral genomes in the absence of the ND10-antagonistic activity of IE1 (Fig. 1C). Next, we analyzed the growth kinetics of the recombinant viruses by performing multistep growth curve analysis. For this purpose, HFF cells were infected either with wild-type AD169 at MOIs of 0.1, 4, or 10, respectively, as determined by IE1 fluorescence, or with equal genome copy numbers of AD169/IE1-L174P and AD169ΔIE1. Aliquots of cell supernatants were collected at various times after infection and subjected to quantitative real-time PCR to determine the amount of the gB gene corresponding to HCMV genome equivalents. At an MOI of 0.1, AD169/IE1-L174P displayed a strongly reduced growth, which was comparable to the reduction caused by deletion of the IE1 gene (Fig. 1D). Release of mutant progeny virions was not detectable until 6 days postinfection (dpi) and, compared to the release of wild-type virions, was 3,000-fold to 7,000-fold decreased on the following days. At an MOI of 4 (Fig. 1E) or 10 (Fig. 1F), growth of AD169/IE1-L174P was still significantly reduced compared to wild-type virus. Furthermore, AD169/IE1-L174P exhibited a slightly increased replication in comparison to the IE1-deleted virus, suggesting residual growth-promoting functions of IE1-L174P under high-MOI conditions. The observed growth reduction of AD169/IE1-L174P underlines the importance of the PML-antagonistic function of IE1 for lytic replication; however, the severity of the defect observed at low MOI was surprising to us since IE1 was shown to promote replication through both interfering with PML-mediated intrinsic immunity via IE1_CORE and inhibiting interferon (IFN) signaling via its disordered C-terminal domain.

Recombinant AD169 encoding IE1-L174P displays reduced viral protein expression under low- and high-MOI conditions.

In order to examine the observed growth defect of AD169/IE1-L174P more precisely, viral protein accumulation was monitored during one round of replication. To this end, HFF cells were infected either at low MOI (0.1) (Fig. 2A) or high MOI (2) (Fig. 2B) with wild-type AD169 and equivalent genome copy numbers of AD169/IE1-L174P and AD169ΔIE1. Total cell lysates were prepared at different times after infection and expression patterns of viral immediate early (IE1, IE2 – p86, p60, p40), early (UL44), and late proteins (MCP) were analyzed by Western blotting. Intriguingly, hardly any IE1-L174P was present throughout the replication cycle (Fig. 2A and B, top panels). Only initially after high-multiplicity infection, comparable levels of IE1-L174P and wild-type IE1 were detected (Fig. 2B, top panels, 8 hpi), but the abundance of IE1-L174P declined as infection progressed (Fig. 2B, top panels, 24 hpi to 96 hpi). This loss of IE1-L174P during infection provides a plausible explanation for the similar growth characteristics of AD169/IE1-L174P and the IE1 deletion virus AD169ΔIE1. Similar to IE1, levels of the second major IE protein IE2p86 were comparable initially upon infection (Fig. 2A, second panels from the top, 24 hpi; Fig. 2B, second panels from the top, 8 hpi). However, expression of IE2p86 and the smaller isoforms IE2p60 and IE2p40 was strongly reduced at later times after infection with AD169/IE1-L174P and AD169ΔIE1 compared to wild-type AD169. Furthermore, no expression of early and late proteins was observed until 96 h after low-multiplicity infection with both of the recombinant viruses (Fig. 2A, third and fourth panels from the top). This is consistent with the delayed and reduced virus release determined in growth curve analysis (Fig. 1D). Notably, although IE1 was shown to be dispensable during high-multiplicity infections with mutants derived from the HCMV laboratory strain Towne (3), we detected a substantial attenuation of viral early and late gene expression in cells infected at a high MOI with AD169/IE1-L174P and AD169ΔIE1 (Fig. 2B, third and fourth panels from the top). We conclude that the low levels of IE1-L174P during AD169/IE1-L174P infection, which may be caused by transcriptional dysregulation or reduced protein stability, lead to a defect in viral gene expression—under low- and high-MOI conditions—that is similar to the defect caused by IE1 deletion.

IE1-L174P has a reduced half-life compared to wild-type IE1.

Since IE1 has been suggested to positively regulate its own promoter and, moreover, transactivation activities of IE1 were mapped to IE1_CORE containing leucine 174 (40, 57), the question arose whether the low protein levels of IE1-L174P may result from impaired transcription. To test this, we measured IE1 mRNA levels at 8 h (Fig. 3A) and 48 h (Fig. 3B) after high-MOI infection with wild-type AD169 or recombinant viruses. Quantitative RT-PCR analysis revealed slightly enhanced expression of IE1-L174P early after infection (Fig. 3A) and similar mRNA levels of IE1-L174P and wild-type IE1 at 48 hpi (Fig. 3B), indicating that the loss of IE1-L174P during infection is not due to deregulation at the transcriptional level. Therefore, we speculated that a leucine-to-proline mutation within the highly alpha-helical IE1_CORE domain may result in misfolding and degradation of the IE1 protein. To investigate the stability of IE1-L174P in living cells, we transfected HEK293T cells with expression constructs for wild-type IE1 or IE1-L174P and analyzed the protein levels at different times after addition of the protein synthesis inhibitor cycloheximide (CHX). While the protein level of wild-type IE1 remained almost unchanged after 30 h of CHX treatment (Fig. 3C, upper panels), a clear decline of IE1-L174P was observed over time (Fig. 3C, lower panels) indicating that mutation of leucine 174 to proline reduces the half-life of the IE1 protein. Since CHX does not fully inhibit de novo protein synthesis and since its toxicity can lead to erroneous results, we next utilized a pulse-chase assay based on CLIP-tag technology (New England BioLabs). For this, wild-type IE1 and IE1-L174P fused to a C-terminal CLIP-tag were transiently expressed in HEK293T cells, followed by labeling with the red fluorescent substrate CLIP-Cell TMR-Star. At sequential time points, the cells were harvested and the amount of labeled IE1-CLIP and IE1-L174P-CLIP was determined by in-gel fluorescence scanning of SDS-PAGE gels. Consistent with the results from CHX experiments, no significant reduction of IE1 was observed within 30 h after the labeling period, demonstrating a high stability of the wild-type protein (Fig. 3D, top panel, upper box). In contrast, labeled IE1-L174P was rapidly degraded (Fig. 3D, top panel, lower box) and displayed an approximate half-life of 5 h, as determined by quantification of the fluorescent signal (Fig. 3D, lower panel [graph]). These findings strongly point toward a negative effect of the leucine-to-proline mutation on the structural integrity and, thereby, on the stability of IE1.

L174P mutation affects the structural integrity of the IE1 fold.

To evaluate possible effects of L174P mutation on the IE1 fold, we performed bioinformatic analysis based on the recently determined crystal structure of IE1_CORE. IE1_CORE consists of 11 α-helices and adopts an elongated, femur-shaped fold that can be divided into an N-terminal head region and a C-terminal head region interconnected by a stalk (6). The L174P mutation is located within helix 5 (H5) of the C-terminal head region and most likely disrupts the helical structure of H5 by multiple unfavorable structural effects. Due to the missing amide proton, the hydrogen bond interaction with C170 is lost and, additionally, clashes with the oxygen atom of C170 backbone are caused by the carbon atoms of the proline imidazole ring (Fig. 4A and B; Table 2. Moreover, L174 forms an extensive hydrophobic packing with side chains of adjacent residues and with residues of neighboring helices, namely, I336, M339, and I343 of H10 and V316 of H9 (Fig. 4C). Upon mutation to proline, most of these contacts are lost due to missing side chain atoms, thus destabilizing the tight helix packing of H4, H5, H9, and H10 in the C-terminal head region and presumably resulting in misfolding of the entire IE1_CORE (Fig. 4D). In order to experimentally confirm the perturbation of the IE1 fold by L174P mutation, we produced recombinant proteins in bacteria and purified them by affinity chromatography and gel filtration. While gel filtration chromatography of wild-type IE1 resulted in one symmetric main peak, IE1-L174P eluted with the void volume, which contains proteins and protein aggregates with a molecular mass of more than 600 kDa (Fig. 4E), indicating an aggregated state of IE1-L174P. The aggregation of IE1-L174P impeded further investigation of the protein fold by additional methods but allowed the conclusion that the leucine-to-proline mutation within H5 indeed affects the structural integrity of the IE1_CORE fold, hereby inducing rapid degradation of IE1-L174P in living cells.

Effects of IE1 mutants IE1-L174P and IE1 1-382 on the inhibition of interferon signaling during viral infection.

Prior studies have demonstrated that L174P mutation in IE1 prevents targeting, de-SUMOylation, and dispersal of PML in transfection assays. To confirm this finding in HCMV-infected HFF cells, we next analyzed the subcellular localization of IE1 and PML after infection with wild-type AD169 or AD169/IE1-L174P. As illustrated in Fig. 5A, infection of HFFs with wild-type virus resulted in the typical, transient colocalization of IE1 and PML (Fig. 5A, panels e to h) followed by disruption of ND10 integrity within the first hours of infection (Fig. 5A, panels i to p). In contrast, IE1-L174P neither localized to ND10 nor induced dispersal of ND10 foci during infection (Fig. 5B, panels f, j, n). Consistently, Western blot analysis revealed that IE1-L174P fails to induce depletion of the SUMOylated forms of PML (Fig. 5C, upper panels) and of Sp100 (Fig. 5C, middle panels) but produces a similar pattern of the various PML and Sp100 species as detected in AD169ΔIE1-infected cells.

Since it has been shown that PML binding by IE1_CORE occurs in a manner independent of the interaction with signal and transducer of transcription (STAT) proteins by the disordered C terminus of IE1 (5, 41), we next explored the influence of IE1-L174P on STAT-mediated IFN signaling during infection. To address this issue, mRNA levels of interferon-stimulated genes (ISGs) were determined in HFF cells infected with high multiplicities of wild-type AD169 and equivalent genome copy numbers of AD169/IE1-L174P and AD169ΔIE1 for 48 h. We observed that transcription of PML, which is known to be IFN inducible, was slightly elevated in AD169-infected cells compared to uninfected cells but highly upregulated after infection with IE1-deleted virus AD169ΔIE1 (Fig. 5D, upper left panel). Infection with AD169/IE1-L174P resulted in an intermediate effect (Fig. 5D, upper left panel, and Fig. 5C, 48 hpi, top panel). Induction of other IFN-α/β-stimulated genes was also efficiently suppressed by wild-type AD169 but not by AD169ΔIE1 (Fig. 5D, upper right panel and lower panels), being in accordance with the described role of IE1 as an antagonist of type I IFN-dependent STAT signaling (42). Again, IE1-L174P displayed an intermediate inhibition of ISG upregulation, although the disordered C terminus harboring the STAT interaction motif is not affected by L174P mutation. This intermediate effect may arise from the low levels of IE1-L174P during infection. However, since PML recently emerged as a positive regulator of interferon signaling (44–46), it is also possible that the loss of PML binding caused by L174P mutation accounts for the inefficient suppression of the interferon response.

In order to obtain further evidence for a role of IE1-mediated PML modification in attenuation of ISG induction during viral infection we directly compared wild-type AD169 with the recombinant viruses AD169ΔIE1, AD169/IE1-L174P, and AD169/IE1 1-382 (Fig. 6). AD169/IE1 1-382 expresses a C-terminally truncated IE1 protein lacking the STAT-binding region and has previously been demonstrated to efficiently antagonize ND10-mediated intrinsic immunity during viral infection (6). As shown in Fig. 6A, IE1 1-382 is expressed at levels similar to those of wild-type IE1 at 48 h postinfection while the abundance of IE1-L174P was significantly reduced (Fig. 6A, top panel). Furthermore, while AD169/IE1 1-382 induced PML deSUMOylation, this was not the case for AD169/IE1-L174P (Fig. 6A, middle panel). When mRNA levels of ISGs were determined in HFF cells infected with the various viruses, we observed that AD169/IE1 1-382 efficiently suppressed ISG induction (Fig. 6B). For most investigated ISGs, the inhibitory effect exerted by IE1 1-382 was only slightly reduced compared to that of wild-type IE1, suggesting that PML modification by IE1 is of major importance for the inhibition of interferon signaling during viral infection.

PML positively regulates type I and type II IFN signaling.

Recent evidence suggested a positive effect of PML on transcription of IFN-γ-stimulated genes and an involvement of one specific PML isoform in IFN-β production (44–46); however, the exact role of PML is far from being understood. To further characterize the influence of PML on interferon signaling, we measured ISG induction in HFFs with an shRNA-mediated depletion of PML (siPML) in comparison to that observed in control cells (siC). As shown in Fig. 7, siPML cells exhibited a downregulation of all PML variants under unstimulated and IFN-stimulated conditions, as assessed by Western blot (Fig. 7A) and qRT-PCR (Fig. 7B) analysis. Intriguingly, stimulation of PML-depleted cells with IFN-α, IFN-β, or IFN-γ resulted in a less efficient upregulation of ISG expression than stimulation of control cells (Fig. 7C). This was true for ISGs that are predominantly induced by type I IFNs, for example TNFSF10, OASL, or IFIT3, and also for ISGs primarily induced by type II IFNs, like HLA-DRA (Fig. 7C). Expression levels of other genes, such as UIMC, did not show major changes, thus excluding a generally reduced transcription in siPML cells. Figure 7D summarizes the effects of IFN-α, IFN-β, and IFN-γ stimulation in siPML cells compared to siC cells and highlights the positive role of PML in both type I and type II IFN-induced gene expression. Taken together, our results indicate that PML not only is encoded by an ISG but also contributes to the interferon response by acting as a coregulator of ISG induction.

The PML-binding domain of IE1 suppresses IFN-induced expression of specific ISGs.

Having shown that PML positively regulates ISG expression, we wanted to further confirm that IE1 influences IFN signaling not only through the interaction with STAT proteins but also through its interaction with PML. To address this issue, we utilized HFF cells with doxycycline-inducible expression of wild-type IE1 or truncated IE1 1-382, which contains the PML-binding domain IE1_CORE but lacks the STAT-binding region. Western blot analysis of PML levels in the absence and presence of doxycycline revealed an efficient deSUMOylation of PML by wild-type IE1 and IE1 1-382 under IFN-induced conditions, confirming the expression of active IE1 proteins (Fig. 8A). Furthermore, indirect immunofluorescence experiments were performed in order to investigate the previously described recruitment of STAT2 by IE1 (5, 42). While full-length IE1 exhibited a distinct nuclear colocalization with STAT2 (Fig. 8B, panels a to e), this could not be observed for IE1 1-382, thus confirming that deletion of the C-terminal STAT-binding region abrogates the interaction with STAT2 (Fig. 8B, panels f to j). Next, the newly generated HFFs were stimulated with IFN-β and ISG induction was analyzed in the absence and presence of IE1 proteins. As illustrated in Fig. 8C, wild-type IE1 suppressed induction of type I IFN-responsive genes (Fig. 8C, left panel) but induced expression of HLA-DRA, an ISG usually controlled by IFN-γ (Fig. 8C, right panel). These opposed effects on type I and type II IFN-responsive genes are consistent with results from previous studies and were shown to depend on binding of STAT2 and STAT1, respectively (41–43). Intriguingly, expression of IE1 1-382, which binds PML but is no longer able to interact with STAT proteins, still had an influence on ISG induction. We detected a significant reduction of HLA-DRA mRNA levels (Fig. 8D), while expression of type I IFN-controlled genes was diminished to various degrees (Fig. 8D). To sum up, although the different extents of inhibition suggest complex underlying mechanisms that will require further studies to become fully elucidated, our data provide evidence for an inhibition of ISG induction by the PML interaction domain of IE1. We conclude that IE1 targets PML in order to manipulate intrinsic as well as innate immune responses.