Recombinant Human Cytomegalovirus Expressing an Analog-Sensitive Kinase pUL97 as Novel Tool for Functional Analyses

Abstract

The human cytomegalovirus (HCMV) is a member of the beta-herpesvirus family and inflicts life-long latent infections in its hosts. HCMV has been shown to manipulate and dysregulate many cellular processes. One major interactor with the cellular host is the viral kinase pUL97. The UL97 gene is essential for viral replication, and kinase-deficient mutants of pUL97 display a severe replication defect. Recently, another group established an analog-sensitive version of the pUL97 protein. This mutant kinase can be treated with a non-hydrolysable ATP analog, thereby inhibiting its kinase function. This process is reversible by removing the ATP analog by media change. We introduced this mutant version of the pUL97 protein into the laboratory strain Ad169 of HCMV, BADwt, creating a BAD-UL97-as1 viral mutant. This mutant virus replicated normally in infected cells in the absence of the ATP analog and maintained its ability to phosphorylate its cellular substrates. However, when treated with the ATP analog, BAD-UL97-as1 displayed a defect in the production of intra- and extracellular viral DNA and in the production of viral progeny. Furthermore, in the presence of 3MB-PP1, a well-established substrate of pUL97 was no longer hyperphosphorylated. This effect was detectable as early as 4 h post treatment, which allows for studies on pUL97 without the complication of low viral titers. Nevertheless, we observed off-target effects of 3MB-PP1 on several cellular processes, which should be considered with this approach.

Keywords: BACmid-derived recombinant HCMV; analog-sensitive pUL97 variant; functional analyses; human cytomegalovirus (HCMV); pUL97-specific inhibitors; protein kinase pUL97; viral kinase activity.

Figure 1

Construction of BAD-UL97-as1. The triplet cat coding for histidin on position 411 within the UL97 gene was replaced by ggg coding for glycin using the BAC recombination technique described in [16].

Figure 2

Replication kinetics of BADwt and BAD-UL97-as1. (A) Cells were infected with BADwt or BAD-UL97-as1 with 4 genomes/cell. Samples were taken and viral DNA was extracted from infected cells and the cell supernatant at the indicated time points. The intracellular and extracellular viral genome copies were quantified by quantitative PCR analysis. (B) HFF were infected with 4 genomes per cell with either BADwt or BAD-UL97-as1. After 6 and 8 days post infection, supernatants were collected and used to analyze the level of viral infectivity by serial dilution on indicator cell cultures, using staining for the immediate-early protein 1 (IE1) of HCMV at two days after inoculation. Bars represent the mean of eight technical replicates and the standard deviation is indicated by error bars.

Figure 3

Replication kinetics of BADwt and BAD-UL97-as1 after treatment with 3MB-PP1. Cells were treated with either 40 µM of 3MB-PP1 or with DMSO one day before infection. The next day, cells were infected with BADwt or BAD-UL97-as1 with 4 genomes/cell. Samples were taken and viral DNA was extracted from infected cells and the cell supernatant at the indicated time points. The intracellular and extracellular viral genome copies were quantified by quantitative PCR analysis.

Figure 4

Phosphorylation of the retinoblastoma protein (Rb). Cells were pre-treated with DMSO, 3MB-PP1 or MBV one day before infection. The next day, cells were infected with BADwt or BAD-UL97-as1. Cells were harvested and lysed at 5 d.p.i. and analyzed in an SDS-PAGE and Western blot using a phospho-specific antibody for residues 807/811 of Rb. Antibodies against full protein Rb, the viral proteins pp28 and pUL97 were probed for reference.

Figure 5

(A) BAD-UL97-as1 infected cells (10 genomes/cell) were harvested at 5 d.p.i. 40 µM of 3MB-PP1 was added at either 5 days, 2 days, 1 day or 4 h before harvest. The samples were analyzed by SDS-Page and Western blot analysis, using a specific antibody against the phosphosite 807/811 of Rb. The quantification was performed by measuring the protein intensity of pRb and Rb using Image Studio Lite Version 5.2.5. The ratios (pRb807/811/Rb) in dependence of 3MB-PP1 are shown. Thereby the corresponding DMSO sample was indicated as 1 (n = 2). (B) BAD-UL97-as1 (2 genomes/cell) infected HFFs were treated with 3MB-PP1 at different time points during infection. At 5 days, 2 days, 1 day or 4 h before harvest, the inhibitor was added. After 5 days post-infection, the supernatant was collected, DNA was isolated and the viral DNA of three technical replicates was quantified by TaqMan-PCR. The means ± standard deviations of each sample are shown.

Figure 6

Dysregulated protein groups after 3MB-PP1 treatment in uninfected HFF. Pathway and process enrichment analysis of proteins up-regulated after 3MB-PP1 treatment (log2 ≥ 0.7 and p-value ≤ 0.05) (A). The top 20 clusters with their representative enriched terms are plotted according to their p-value in log base 10 (Log10 (P)). Enrichment analysis was performed with Metascape [30]. Pathway and process enrichment analysis of proteins down-regulated after 3MB-PP1 treatment (log2 ≤ −0.7 and p-value ≤ 0.05) (B). The top 20 clusters with their representative enriched terms are plotted according to their p-value in log base 10 (Log10 (P)). Enrichment analysis was performed with Metascape [30]. Pathway results are shown with the number of proteins found in the dataset and computed FDR for pathway enrichment (FDR < 0.001).

Figure 7

HFFs were cultivated in T175 flasks and used for infection with parental HCMV BADwt or recombinant BAD-UL97-as1 at a MOI of 0.5 for 4 d. An optional treatment with 40 µM of 3MB-PP1 (+, 3MB-PP1; −, DMSO solvent controls) was performed starting 4 h prior to sample collection and again during cell lysis. Total cell lysates were prepared and used for cyclin B1-specific CoIP with the indicated antibodies (see black boxes; Fc fragment was used as a negative control) under the continuous presence of 3MB. (A,B) show two independently produced biological replicates of this CoIP experiment, in order to illustrate the range of variability of individual signal strengths of detected protein bands. The CoIP samples were subjected to SDS-PAGE/Wb analysis (CoIP) and additional control stainings were performed to verify successful immunoprecipitation (cyclin IP control) and protein expression levels (lysate controls).