Establishment of a Cell-Based Assay for Screening of Compounds Inhibiting Very Early Events in the Cytomegalovirus Replication Cycle and Characterization of a Compound Identified Using the Assay

Abstract

To simplify the detection of infectious human cytomegalovirus (HCMV), we generated a cell line that produced luciferase in a dose-dependent manner upon HCMV infection. Using this cell line, we identified anti-HCMV compounds from a diverse library of 9,600 compounds. One of them, 1-(3,5-dichloro-4-pyridyl)piperidine-4-carboxamide (DPPC), was effective against HCMV (Towne strain) infection of human lung fibroblast cells at a 50% effective concentration of 2.5 μM. DPPC also inhibited the growth of clinical HCMV isolates and guinea pig and mouse cytomegaloviruses. Experiments using various time frames for treatment of the cells with DPPC demonstrated that DPPC was effective during the first 24 h after HCMV infection. DPPC treatment decreased not only viral DNA replication but also IE1 and IE2 expression at mRNA and protein levels in the HCMV-infected cells. However, DPPC did not inhibit the attachment of HCMV particles to the cell surface. DPPC is a unique compound that targets the very early phase of cytomegalovirus infection, probably by disrupting a pathway that is important after viral entry but before immediate-early gene expression.

Human cytomegalovirus (CMV) (HCMV) is associated with significant morbidity in immunocompromised individuals, including transplant patients and human immunodeficiency virus (HIV)-positive individuals (35). Although prophylaxis with available anti-HCMV drugs, such as ganciclovir (GCV), effectively reduces the risk of CMV diseases and associated mortality of transplant patients, side effects, development of resistant strains, and late-onset diseases have necessitated a search for alternative compounds. In addition to nucleoside analogs, several new types of therapeutic compounds have been under investigation (6, 11, 18, 46). HCMV is also the major viral cause of birth defects and hearing impairment (35). Consistent with prior descriptions, we recently found that 15% of cases of severe sensorineural hearing loss were ascribed to congenital CMV infection and that half of them were late onset (32). Significant progress in the methodologies for universal CMV screening has enabled early identification of newborns with congenital infection (3, 31), which may lead to new treatment options with antiviral agents (21). However, since the side effects associated with currently available anti-HCMV drugs limit the treatment of newborns who are congenitally infected with HCMV but have no symptoms at birth, it is also critical to develop new anti-HCMV drugs for this purpose.

For screening and evaluation of novel antiviral compounds, a plaque reduction assay, which is made laborious and time-consuming by the slow growth of HCMV in tissue culture, has been used routinely as a standard assay. To reduce the inconvenience, several modified high-throughput assays, including the use of recombinant viruses expressing a marker gene product, colorimetric assays, and flow cytometric assays, have been described (1, 5, 13, 27, 28, 39). Previously we established reporter cell lines for human herpesvirus 8 (HHV-8) and varicella-zoster virus (VZV) and demonstrated their practical uses (17, 25, 47). In this study, we generated a similar cell line for HCMV. Since the promoter used for the establishment of the cell line is activated by HCMV immediate-early (IE) proteins, the cell line is suitable for high-throughput screening of new antiviral compounds that may inhibit the very early phase of infection. We also characterized one of the anti-HCMV compounds identified using this cell-based assay.

MATERIALS AND METHODS

Cells and viruses.

HLF (human lung fibroblast; Centers for Disease Control and Prevention Cell Culture Core Facility), U373MG (human glioma cell line; ATCC HTB17), Vero (ATCC CCL-81), MeWo (human melanoma cell line) (14), 293T (36), and NIH 3T3 (ATCC CRL-1658) cells were grown in Dulbecco's modified Eagle medium (Gibco-BRL) supplemented with 10% fetal bovine serum (FBS) (HyClone), 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco-BRL). MeWo cells were used for infection with VZV (P-Oka). MV9G, a VZV reporter cell line, was used for titration of VZV (47). A telomerase-immortalized human fibroblast cell line, hTERT-BJ1 (Invitrogen), was grown in Dulbecco's modified Eagle medium-medium 199 (4:1) supplemented with 10% FBS and infected with the HCMV Towne and RC256 (41) strains and with several HCMV clinical isolates. The clinical isolates were obtained by the minimum passage of HLF or hTERT-BJ1 cells that were inoculated with urine specimens collected from various sources, including postnatally infected children (C4, C7, C10, and C41), newborns with congenital CMV infection (N42 and C49), and HIV-infected patients (544, 569, 689, 724, and 751). Culture supernatants containing HCMV were used as virus stocks, and the supernatants were passed through a 0.45-μm-pore-size filter, if necessary, to completely remove any membrane-associated virus particles. Guinea pig lung fibroblast (ATCC CCL-158) cells were grown in F12 medium supplemented with 10% FBS and used for infection with guinea pig CMV (GPCMV) (strain 22122; ATCC VR-682). The mouse CMV (MCMV) Smith strain was cultured on NIH 3T3. Molt3 cells were cultured and infected with the HHV-6B Z29 strain as described previously (16).

Chemicals.

Heparan sulfate, condroitin sulfate B, GCV, nocodazole, suramin, roscovitine (Sigma), and a random library of 9,600 chemical compounds (Maybridge) were used in this study. Most chemicals were dissolved in dimethyl sulfoxide (DMSO).

Plasmid construction and mutagenesis.

Nucleocapsid DNA of HCMV strain AD169 was prepared as described previously (43). DNA fragments containing HCMV promoters were amplified from the nucleocapsid DNA by PCR with native PfuI DNA polymerase (Stratagene) and cloned into a reporter plasmid, pGL3-Basic (Promega), that encodes firefly luciferase. The regions of HCMV promoters analyzed in this study are as follows: TRL4, 184,676 to 184,944; TRL7, 181,685 to 182,145; UL21.5, 26,070 to 27,092; UL54, 81,440 to 80,996; UL57, 91,123 to 90,302; UL69, 100,871 to 100,435; UL83, 121,377 to 120,978; UL86, 128,756 to 123,327; UL98, 142,245 to 142,666; UL99, 142,307 to 144,010; UL112-113, 160,162 to 160,549 (coordinates are based on the AD169 sequence [9], GenBank accession no. X17403). Reporter plasmids containing truncated forms of the TRL4 promoter were constructed similarly, and one of them, pGL3-TRL4C, containing the HCMV (AD169) sequences from position 184,765 to 184,944, was used for the generation of reporter cell lines.

Transient-transfection assay.

Subconfluent U373MG cells in 96-well plates were transfected with a mixture of 125 ng of a reporter plasmid and 4 ng of a Renilla luciferase-expressing plasmid (pRL-CMV; Promega) as an internal control, by using FuGENE 6 (Roche). After 24 h, the cells were infected with HCMV (Towne) at a multiplicity of infection (MOI) of 0.01 to 0.05. At 46 h postinfection, the cells were washed once with phosphate-buffered saline (PBS); 25 μl of PBS and 25 μl of luciferase reagent solution were added to each well. After incubation for 30 to 40 min, 30-μl portions were transferred to black 96-well plates and both firefly and Renilla luciferase activities were obtained by a chemiluminescence assay reaction (Dual-Glo luciferase assay system; Promega) followed by measurement of relative light units (RLU) with a luminometer (JNR AB2300; Atto, Japan).

Generation of reporter cell line for HCMV.

U373MG cells were transfected with pGL-TRL4C along with pCMV-Script (Stratagene), a plasmid encoding the G418-resistant gene, at a 50:1 ratio. The reporter cell clone U4C-48 was chosen from 50 G418-resistant clones based on its high luciferase activity induced by HCMV infection together with its low background activity in the absence of infection.

Reporter cell assays.

The following two assay procedures were used to detect viral infection in reporter cells. (i) U4C-48 cells (2.8 × 10⁴ cells/well) in 96-well plates were inoculated with cell-free virus, centrifuged at 600 × g for 30 min, incubated at 37°C for 1.5 h, and then cultured for the appropriate time period after removal of the inoculums. Luciferase activities of U4C-48 cells were tested as described above. One HCMV (Towne) stock, for which the PFU titer was determined by plaque assay, was diluted serially (5- or 10-fold) and used for the infection of the reporter cells to obtain a standard curve. Titers of the virus stocks were estimated based on the standard curve and expressed as reporter assay units, which should be theoretically equivalent to PFU. (ii) hTERT-BJ1 cells (2.5 × 10⁴ cells/well) in 96-well plates were infected with HCMV at an MOI of 0.1 to 0.3 and cultured in the absence or presence of inhibitors. As quantitative references for the evaluation of antiviral compounds, hTERT-BJ1 cells in several additional wells were infected with twofold serial dilutions of HCMV and cultured in the absence of inhibitors. Two days later, U4C-48 cells (2 × 10⁴ cells/well) were added to each well and cocultured for an additional day. Luciferase activities of the cocultured cells were then measured.

Screening of anti-HCMV compounds.

Chemical compounds (10 mM) dissolved in DMSO were diluted to 80 μM with medium in 96-well 0.5-ml assay blocks (Coster). Then, 50 μl of the diluted compounds (final concentration, 20 μM) and 2,000 PFU of HCMV (Towne) in 50 μl of medium were added to U4C-48 cells (2.8 × 10⁴ cells/well in 100 μl of medium) in 96-well plates. Eighty compounds were examined in one plate along with four control wells. In the control wells, U4C-48 cells were uninfected or infected with 2,000 PFU, 1,000 PFU, or 500 PFU of HCMV, respectively, in the absence of the chemical compounds. The plates were centrifuged at 600 × g for 30 min and then cultured for 46 to 48 h. In each experiment, around 10 96-well plates were tested. Compounds that inhibited luciferase activities to a level less than that of the cells infected with 500 PFU of HCMV were evaluated further for anti-CMV activities and for cell viability.

Cell viabilities were measured using a commercial assay kit (CellTiter-Glo; Promega). Briefly, cells at 20 to 25% confluence in 96-well plates were cultured for 2 to 4 days in the presence of twofold serially diluted chemical compounds or an equivalent volume of DMSO. Culture supernatants were removed, and the cells were rinsed once with medium. Twenty-five microliters of medium, followed by 25 μl of luciferase reaction reagent, was added to each well. Thirty minutes later, 30 μl of the contents in each well was transferred to black 96-well plates to measure RLU in a luminometer. Experiments were done in triplicate. The 50% cytotoxicity concentrations were calculated by plotting log drug concentrations against obtained RLU.

Plaque reduction assays.

Semiconfluent cells in 12-well plates were inoculated with 50 PFU or 100 PFU of virus per well and centrifuged at 600 × g for 30 min. After 1 to 2 h of incubation at 37°C, the inoculums were replaced with medium containing twofold serial dilutions of the compounds, 1% methylcellulose, and 4% FBS. After incubation (4 to 5 days for MCMV, 6 to 7 days for GPCMV, and 10 to 12 days for HCMV), the medium was removed and the cells were stained with 2% crystal violet in 10% formalin before being washed with water. Plaques were counted under a dissecting scope. The assays were performed in triplicate. The 50% effective concentrations (EC₅₀) were calculated by linear regression from the plots of log drug concentrations against percentage reduction in plaque numbers at each antiviral concentration compared to that of the drug-free control.

Immunostaining and X-Gal staining of HCMV-infected cells.

Infected cells were fixed with 3.7% formalin, rinsed with PBS, and incubated with a 1:500 dilution of anti-HCMV IE monoclonal antibody (MAb810; Chemicon International) at 37°C for 1 h. After three washes with PBS, the cells were incubated with horseradish peroxidase-conjugated anti-mouse immunoglobulin G antibody (N-Histofine Simple Stain MAXPO, Nichirei, Japan) at 37°C for 1 h. Positive HCMV foci were visualized with DAB (Roche). 5-Bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) staining was done as described previously (17).

TCID50.

Semiconfluent hTERT-BJ1 cells in 96-well plates were inoculated with fourfold serial dilutions of virus stocks and centrifuged at 600 × g for 30 min. Four wells were used for each dilution of each stock. After incubation for 12 to 14 days, the medium was removed and the cells were stained. Wells containing infected cells were counted under a dissecting scope, and 50% tissue culture infectious doses (TCID₅₀) were calculated by the Reed and Muench method.

IFA and immunoblotting.

Immunofluorescence assay (IFA) and immunoblotting were done as previously described (24). The following monoclonal antibodies were used for detection of viral antigens: anti-HCMV IE1/2, 810; anti-HCMV IE1, 6E1, and IE2, 12E2 (Vancouver Biotech); anti-HCMV ICP22, 6F12, anti-HCMV pp65, 3A12, and anti-HCMV pp28, 5C3 (abcam).

Quantitative PCR assays.

DNA samples were prepared from the HCMV-infected cells using a commercial kit (QIAamp DNA mini kit; Qiagen). Quantitative PCR (qPCR) assays for HCMV UL83 and human albumin genes were described previously (32). Dilutions of a plasmid encoding the entire UL83 open reading frame of HCMV (AD169 strain) and dilutions of cellular DNA extracted from a human B-cell line were used as quantitative standards for the HCMV and human albumin genes, respectively.

Total RNA samples were prepared from the infected cells using a commercial kit (RNeasy Plus mini kit; Qiagen). Quantitative reverse transcription (RT)-PCR was done using a commercial kit (One-step RT-PCR master mix; Applied Biosciences) with the primers and probes targeting the exon junctions of the IE1, IE2, and glucose-6-phosphate dehydrogenase (G6PD) genes (38). The amounts of IE1 and IE2 RNAs were expressed as their threshold cycle (C_T) numbers after normalization to the G6PD level, i.e., normalized C_T (IE1 or IE2 in each sample) = measured C_T (IE1 or IE2 in each sample) − [C_T (G6PD in each sample) − average of C_T (G6PD)].

RESULTS

Establishment and characterization of CMV reporter cell line.

Both the characterization of several HCMV promoters and the analysis of global gene expression in HCMV-infected cells have been reported previously (8, 29). Based on those studies, DNA fragments containing 11 well-characterized HCMV promoters were amplified from HCMV (AD169), cloned into a luciferase reporter vector, and then analyzed for their responses to HCMV infection in transient-transfection experiments in two cell lines permissive to HCMV growth, a telomerase-immortalized fibroblast cell line, hTERT-BJ1, and a glioma cell line, U373MG. Among the promoters examined, the TRL4 promoter, which regulates a highly abundant RNA (β2.7), was the most strongly activated in both cell lines (Fig. 1A). Progressive truncation of the TRL4 promoter allowed mapping of its minimal essential region for transactivation to a 98-bp-long sequence located at bp −125 to −28 relative to the transcription start site (data not shown), which is consistent with the results reported by Klucher et al. (23). Our CMV reporter cell line was generated by transfecting U373MG cells with a plasmid encoding the luciferase gene driven under the TRL4 promoter (−125 to +56) and a plasmid encoding the G418 resistance gene. The cell clone U4C-48 was selected based on its strong activation of luciferase expression after HCMV infection. The luciferase activity of U4C-48 cells was detectable as early as 16 h after infection and plateaued by 45 h postinfection (p.i.) (data not shown). Infection of U4C-48 cells with HCMV induced luciferase activities in a dose-dependent manner with a detection limit of less than 100 PFU (Fig. 1B). No luciferase activity was detected in U4C-48 cells cocultured with cells infected with VZV (P-Oka) or HHV-6 (Z29) or in U4C-48 cells directly infected with cell-free VZV (data not shown), indicating that the reporter cells responded specifically to HCMV. As reported previously (15), infection efficiency was enhanced by centrifugation at 600 × g for 30 min (data not shown). Addition of polybrene during the attachment phase reduced the efficiency of infection (data not shown), which differed from the results observed for an HHV-8 reporter cell line (17).

FIG. 1.

Establishment and characterization of a CMV reporter cell line. (A) Selection of an HCMV promoter for establishment of an HCMV reporter cell line. DNA fragments containing the HCMV promoters for the indicated open reading frames were amplified by PCR and cloned into the luciferase reporter plasmid pGL3-Basic (Promega). U373MG glioma cells were transfected with the plasmids and then infected with HCMV at 24 h after transfection. Luciferase activities were measured at 24 h p.i. Means and SDs in RLU from triplicate experiments are shown. (B) U4C-48 cells (2.8 × 10⁴ cells/well) in 96-well plates were infected with serial dilutions of HCMV (Towne) and cultured for 45 h prior to measurement of luciferase activity. Means and SDs in RLU from triplicate experiments are shown.

Evaluation of known antiviral compounds by reporter cell assay.

As a proof of concept, the effects of well-known inhibitors of HCMV infection were assessed with the newly established reporter cell line (Fig. 2). In the presence and absence of inhibitors, the reporter cells were infected with HCMV and cultured for 45 h, and then luciferase activities of the cells were measured. As previously observed (2), heparan sulfate and suramin, a symmetrical polysulfate naphthylamine derivative of urea, inhibited HCMV infectivity in U4C-48 cells. In contrast, GCV treatment decreased luciferase activity only slightly in U4C-48 cells infected directly with cell-free virus; this was not unexpected, since the reporter cell assay depends on the activation of the TRL4 promoter, which mediates its regulation as an early gene. As reported previously by others (23), the TRL4 promoter is activated by IE2 in transient-transfection experiments and this activation is not blocked by the inhibitors of herpesvirus DNA polymerase, such as phosphonoacetic acid and GCV (data not shown). To evaluate inhibitors of DNA synthesis and the later stages of infection, a cell line permissive to HCMV infection, hTERT-BJ1, was infected with HCMV, cultured in the presence of the inhibitors, and then cocultured with U4C-48 cells. Using this method, the EC₅₀ of GCV against HCMV infection in hTERT-BJ1 cells was estimated to be 3 μM.

FIG. 2.

Use of the CMV reporter cell assay for antiviral evaluation. (A) U4C-48 cells (2.8 × 10⁴ cells/well) in 96-well plates were infected with 2,000 PFU of HCMV (Towne) and cultured for 45 h in the presence of the indicated amount of inhibitors. The obtained RLU, i.e., luciferase activities, in the cells without treatment were used as a 100% reference, and the luciferase activities of the cells treated with inhibitors were compared with this reference in each experiment. (B) hTERT-BJ1 cells (2.5 × 10⁴ cells/well) in 96-well plates were infected with 7 × 10³ PFU of HCMV (Towne) and cultured for 2 days in the absence or presence of the indicated concentration of GCV. Uninfected U4C-48 cells (2 × 10⁴ cells/well) were then overlaid on the infected cells and cocultured for an additional day. Means and SDs in RLU, obtained in triplicate wells, are shown.

Screening of novel anti-HCMV compounds by reporter cell assay.

To identify a compound(s) that inhibits the early phase of HCMV infection, U4C-48 cells were directly infected with HCMV in the presence of 20 μM of the respective chemical compounds. Three hundred compounds that decreased the luciferase activities in the reporter assay were identified from the 9,600 compounds tested. Approximately 200 compounds cytotoxic at 20 μM and a further 50 compounds weakly cytotoxic at 50 μM were excluded. Then, 11 compounds were chosen for further analysis based on their strong anti-HCMV activities in the reporter cell assay. Candidates were analyzed for their anti-HCMV activity by the immunostaining of HCMV-infected cells. One candidate showed no anti-HCMV activity in this way, suggesting that it is a luciferase inhibitor. One of the remaining candidates, 1-(3,5-dichloro-4-pyridyl) piperidine-4-carboxamide (DPPC) (Fig. 3A), was analyzed further.

FIG. 3.

(A) Chemical structure of DPPC. (B) Comparison of EC₅₀ of DPPC and GCV against HCMV (Towne), GPCMV, and MCMV, determined by plaque reduction assays. Means and SDs of the EC50 obtained in triplicate experiments are shown. (C) Comparison of the reporter cell assay with the immunostaining assay. U4C-48 cells (2.8 × 10⁴ cells/well) in 96-well plates were infected with 27 viral stocks and cultured for 46 h. Luciferase activities in each well were measured as described in Materials and Methods. Reporter assay-based viral titer units (RAU) of viral stocks (x axis) were obtained by comparing RLU measurements with a standard curve, such as the one shown in Fig. 1B, that was prepared with serial dilutions of a well-characterized HCMV (Towne) stock. Infectious units (IU) of virus stocks (y axis) were determined by counting immunostained foci of HCMV-infected cells. Each symbol represents one virus stock. (D) Correlation of TCID₅₀ titers with the reporter assay-based viral titer units that were obtained as described above. (E) The reporter cells were infected with five virus stocks (one laboratory strain and four clinical isolates) and cultured in the presence of the indicated concentrations of DPPC for 2 days to measure luciferase activities. The experiments were done in triplicate. Means of RLU at each concentration for the five virus stocks are shown. SDs are shown only for Towne to simplify the figure.

Effectiveness of DPPC against human, guinea pig, and mouse CMVs.

The EC₅₀ of DPPC against HCMV (Towne), GPCMV, and MCMV were determined by plaque reduction assays with HFL, guinea pig lung fibroblasts, and NIH 3T3 cells, respectively, and compared with those of GCV (Fig. 3B). The EC₅₀ of GCV were consistent with those in previous reports (for example, see references 10, 37, and 48). DPPC inhibited the growth of those CMVs as effectively as did GCV. Since the 50% cytotoxicity concentration of DPPC against hTERT-BJ1 and HFL cells was 100 μM, the selective index was more than 10. DPPC did not induce any detectable morphological changes in hTERT-BJ1 and HFL cells (data not shown). The EC₅₀ of DPPC was determined to be 3.5 μM by counting X-Gal-stained hTERT-BJ1 cells infected with HCMV RC256 (a recombinant Towne strain expressing β-galactosidase) in the presence of various concentrations of DPPC (data not shown).

To demonstrate effectiveness of DPPC not only against HCMV laboratory strains but also against clinical isolates, we verified that the reporter cell assay measured the viral titers of clinical isolates accurately, since the plaque assay is too laborious. For this purpose, we compared the titers estimated by the reporter cell assay with those determined by the immunostaining assay (Fig. 3C) and with the TCID₅₀ (Fig. 3D). Virus stocks used for comparison included 3 stocks of 2 laboratory strains (AD169 and Towne) and 24 stocks of 13 low-passage clinical isolates. The virus titers determined by the three assays were well correlated. As shown in Fig. 3E, the reporter cell assay demonstrated that DPPC had similar inhibitory effects on the laboratory strains and the clinical isolates. Based on the concentration of DPPC that was needed to decrease the luciferase activities to the same level as that in cells infected with half the amount of virus inoculums, the EC₅₀ of DPPC against the HCMV Towne strain in the reporter cells was estimated to be 5 μM and those against clinical isolates ranged from 5 μM to 10 μM.

DPPC treatment results in inhibition of HCMV DNA replication.

To identify the target of DPPC, we initially examined whether DPPC resulted in inhibition of HCMV DNA replication. hTERT-BJ1 cells were infected with HCMV at either a low (Fig. 4A) or high (Fig. 4B) MOI, and copy numbers of HCMV and cellular genes were determined by qPCR assays. Irrespective of the MOI, DPPC treatment decreased the HCMV DNA copy numbers/cell as efficiently as did GCV, indicating that DPPC inhibits a process coincident with or prior to viral DNA replication.

FIG. 4.

Effect of DPPC on HCMV DNA replication. hTERT-BJ1 cells were infected with HCMV at an MOI of 0.05 (A) or at an MOI of 3.0 (B) and cultured for the indicated durations in the presence of DMSO or 20 μM or 5 μM of DPPC or GCV. Copy numbers for HCMV UL83 and human albumin genes were obtained by qPCR assays. HCMV genome copy numbers divided by albumin gene copy numbers were expressed as HCMV copy numbers per cell. Means and SDs of the copy numbers obtained in triplicate experiments are shown.

DPPC was effective during the first 24 h of infection.

To further elucidate which step in the HCMV infection cycle was inhibited by DPPC, the reporter cells were infected with HCMV and treated with DPPC in various ways by changing starting and ending points of the DPPC treatments (Fig. 5A and B). DPPC was effective when present during the first 24 h of infection. To exclude the possibility that a longer treatment time rather than the time frame of infection affected the outcome, the infected cells were treated only for 24 h, starting at different time points (Fig. 5C). Even under such a condition, the earlier-phase treatment was still found to be effective. Importantly, the addition of DPPC at 6 h or at 12 h p.i. partially inhibited HCMV infection, suggesting that DPPC inhibits infection after entry.

FIG. 5.

hTERT-BJ1 cells (3 × 10⁴ cells/well) were plated in 96-well plates. The next day, the cells were infected with 2 × 10³ PFU of HCMV (Towne) and treated with 20 μM DPPC or DMSO. The periods of DPPC treatment are indicated by bars under each panel. For each set of conditions, luciferase activities were measured for three wells at 48 h after infection; means and SDs are plotted. (A) DPPC (or DMSO) was added to the cultures at the indicated time (h) after infection and kept in the medium until cells were harvested at 48 h p.i. Luciferase activities were plotted as a function of the starting point of the treatment. (B) DPPC (or DMSO) was added to the cultures from the beginning, and the medium containing DPPC (or DMSO) was replaced with medium without DPPC (or DMSO) at the indicated time (h) after infection. Luciferase activities were plotted as a function of the time of medium replacement. (C) DPPC (or DMSO) was added to the cultures at the indicated time (h) after infection, and the medium containing DPPC (or DMSO) was replaced with medium without DPPC (or DMSO) at 24 h after addition of DPPC or DMSO. Luciferase activities were plotted as a function of the time of the treatment initiation. The results of the treatments marked with *1 and *2 are also shown in panels B and A, respectively. In addition to the experiment shown here, we performed two similar preliminary experiments with a single measurement for each set of conditions and one experiment with duplicated measurements and obtained similar results.

DPPC decreases IE gene expression.

Next, we examined the effects of DPPC on viral gene expression. Immunoblotting and IFA analyses demonstrated that DPPC decreased the expression of immediate-early (IE1/IE2) and early (e.g., ICP22) HCMV gene products (Fig. 6). The decrease in IE1 and IE2 expression was confirmed by monoclonal antibodies 6E1 and 12E2, which recognize only IE1 and IE2, respectively (data not shown).

FIG. 6.

Effects of DPPC on IE gene expression. (A) hTERT-BJ1 cells were uninfected (lane 1) or infected with HCMV (lanes 2 and 3) at an MOI of 1.25 and cultured in the absence (lanes 1 and 2) or in the presence (lane 3) of 20 μM DPPC for 48 h. Cell lysates were separated in 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Viral antigens were detected by monoclonal antibody against ICP22 (α-ICP22). (B) hTERT-BJ1 cells were uninfected (lane 1) or infected with HCMV (lanes 2 to 4) at an MOI of 1.25 and at an MOI of 0.25 (lanes 5 to 7) and cultured in the absence (lanes 1, 2, and 5) or in the presence of 5 μM (lanes 3 and 6) or 20 μM (lanes 4 and 7) DPPC for 24 h. Cell lysates were separated in 10% SDS-PAGE. Viral antigens were detected by monoclonal antibody against IE1/IE2. (C) hTERT-BJ1 cells cultured in eight-well Lab-Tech chamber slides (Nunc) were infected with HCMV and cultured in the presence of 10 μM or 40 μM DPPC (or DMSO) for 24 h. IE1/IE2 antigens were detected by IFA with monoclonal antibody against IE1/IE2.

To clarify the mechanism underlying the decrease in IE1/IE2 gene products, the expression levels of IE1 and IE2 mRNA were compared with that of G6PD RNA by qRT-PCR (Fig. 7). Roscovitine, which inhibits HCMV IE gene expression (38), was used as a control. The mean and standard deviation (SD) of the threshold cycle numbers for G6PD were 22.0 ± 0.4, indicating that the expression level of G6PD is constant irrespective of treatment with drugs, whereas DPPC treatment decreased both IE1 and IE2 mRNA levels as did roscovitine treatment.

FIG. 7.

Effects of DPPC on IE1 and IE2 transcript levels. hTERT-BJ1 cells (4 × 10⁵ cells/T12.5 flask) were infected with HCMV at an MOI of 0.05 and cultured in the presence of DMSO, 15 μM roscovitine (Rosco), or 20 μM DPPC for the indicated periods. The cells were washed with PBS and stored at −80°C after the removal of PBS. Total RNA samples were purified from the cells, and the amounts of IE1, IE2, and G6PD transcripts were measured by quantitative RT-PCR assays. Expression levels of IE1 and IE2 RNAs are shown as C_T numbers after normalization to G6PD. The means and SDs from triplicate experiments are shown.

No effect of DPPC on attachment of virus particles.

To examine whether DPPC inhibits HCMV entry into cells, we measured the amount of HCMV DNA in the cells 2 h p.i. (Fig. 8). Incubation of the infected cells in the presence of heparan sulfate, which blocks virion attachment, but not in the presence of DPPC decreased the amount of HCMV in the cells, indicating that the attachment of virus particles to the cell surface was not inhibited by DPPC.

FIG. 8.

Effect of DPPC on the entry process. hTERT-BJ1 cells that were pretreated with DMSO, 20 μM DPPC, or 1 mg/ml heparan sulfate (HS) were infected with HCMV, incubated at 4°C for 2 h, rinsed twice with PBS to remove unbound virions, and then cultured at 37°C for 2 h. The drugs were present throughout the processes. The cells were rinsed with PBS and then treated with 5 mg/ml protease K and 1 mg/ml DNase I in PBS containing 5 mM MgCl₂ at 37°C for 30 min. The cells were harvested by centrifugation. The amount of HCMV DNA per cell was measured by qPCR assays for HCMV and cellular genes. Means and SDs of the DNA amounts obtained in triplicate experiments are shown.

DISCUSSION

In this study, we established a cell-based assay that was useful for the screening of compounds inhibiting very early events in the HCMV replication cycle. Detection of HIV through the use of reporter cell lines based on the transactivation of the long terminal repeat, as with the MAGI assay (22), has enhanced not only basic studies but also the clinical diagnosis of HIV. Similar cell lines have been described for several other viruses (34). We and others generated reporter cell lines for herpesviruses, including herpes simplex virus, B virus, HHV-8, and VZV (7, 17, 34, 42, 47). Two reporter cell lines for HCMV that depend on the activation of the UL54 (DNA polymerase) promoter were also published previously (13, 45). However, since our transient-transfection experiments in this study and genomewide DNA microarray analyses (8) clearly demonstrated that the TRL4 promoter is severalfold stronger than the UL54 promoter, our cell line may have an advantage in terms of sensitivity. Although the use of green fluorescent protein as a reporter in one of the cell lines makes it possible to obtain real-time results, it requires flow cytometry or an imaging device for quantitative detection. As shown in this study, the use of reagents for slow chemiluminescence and a 96-well-format luminometer allow high-throughput analysis. One limitation of our cell line is that it cannot replace plaque reduction assays for exact evaluation of antiviral drugs that inhibit a process(es) after transactivation of the reporter promoter, including DNA replication and the maturation of virus particles, although by varying the methodology, the assay still can be made useful for the screening of such drug candidates from a library of compounds.

Using the reporter cell-based assay, we identified several anti-HCMV compounds from a chemical library and characterized the antiviral activities of one of the compounds, DPPC. To our knowledge, this is the first report of the antiviral properties of DPPC. We found that DPPC inhibited a step prior to IE gene expression but probably after viral entry. That DPPC had no effect on the attachment of virus particles to the cell surface was concluded based on detection of HCMV DNA in cells incubated in the presence of DPPC but not in the presence of heparan sulfate for 2 h after infection. However, due to the lack of direct proof that the treatment with protease K and DNase I completely removed virus particles from the cell surface, we cannot conclude that DPPC has no effect on the penetration step of viral entry. The fact that the addition of DPPC at 6 h or at 12 h p.i. still partially inhibited CMV infection in the reporter cell assay suggests that DPPC has no effect on viral entry. To demonstrate that DPPC does not block entry of biologically active CMV into cells, it is necessary to perform additional experiments, such as those performed for the characterization of an inhibitor of HCMV-specific fusion (18). Some compounds are known to inhibit the capsid transport process after entry. For example, chemicals that depolymerize microtubules, such as nocodazole, partially block the transport of herpesvirus capsids to the nucleus (33, 40). Such inhibitors induce some morphological changes in cells, but we did not observe any such changes in DPPC-treated cells. Inhibitors of diacylglycerol/phorbol ester-dependent protein kinases C prevent MCMV replication at a very early stage of infection (26). Cdk inhibitors, such as roscovitine, also inhibited the replication of herpes simplex virus 1, VZV, and HCMV at a very early step of infection (19, 38, 44). The viral replication steps that were blocked by the protein kinase C and Cdk inhibitors remain unclear. Some CMV mutants exhibiting a deficiency in the very early phase of infection have been described (4, 12, 30). DPPC may be a good tool for dissecting the processes involved in the postentry phase of infection. In vitro selection of HCMV strains resistant to DPPC is under way.

DPPC did not show any cytotoxic effects, including the inhibition of cell growth or induction of morphological changes, in cells cultured at 80 μM. In parallel screening and cytotoxicity assays, the selective index was >10, indicating its potential for further characterization. Since the mechanism of action of DPPC is different from that of polymerase inhibitors, it would be worthwhile to examine whether isolates that are resistant to GCV are susceptible to DPPC. DPPC has a low calculated logP value (1.03), suggesting a high bioavailability, and the Actelion Property Explorer (www.actelion.com/uninet/www/www_main_p.nsf/Content/Technologies+Property+Explorer) found no toxicity risks. Detailed structure-activity relationship studies are under way.

We demonstrated that DPPC inhibited not only human but also animal CMVs in cell cultures, which makes it possible to evaluate its antiviral effects, pharmacokinetics, and safety with animal models. We recently reported that infection of pregnant guinea pigs with GPCMV caused CMV-associated labyrinthitis in their offspring (20). Evaluation of the inhibitory effects of DPPC in this model is a goal of our preclinical studies.