Novel Real-Time Monitoring System for Human Cytomegalovirus- Infected Cells In Vitro That Uses a Green Fluorescent Protein-PML-Expressing Cell Line

Virus, cell culture, and plasmids.HCMV strains (Towne, Davis, and other clinical isolates) were propagated and titrated in human embryonic lung (HEL) fibroblasts as described previously (26). For cell-associated clinical isolates, the number of plaque-forming cells (PFCs) present in an infected culture was estimated according to previously described methods (19). The following clinically isolated virus strains were used: strains 01, 03, and 05 were isolated from patients with congenital HCMV infection; strains 02 and 04 were from patients with bone marrow transplantation and kidney transplantation (21), respectively; strain 06 was from a patient with lung transplantation whose clinical course will be described elsewhere; strains 07 and 08 were isolated from a patient with CMV retinitis and encephalitis, as reported previously (13, 32); and strain 09 was from a patient with AIDS (14).

The full-length cDNA of human p180 (27) was inserted into the pcDNA4 HisMax vector (Invitrogen) to generate pcDNAp180-54R. The expression plasmid for IE1 was described previously (17).

Establishment of the SE/15 cell line.Chinese hamster ovary (CHO) cells have been reported to be as sensitive to HCMV as human fibroblasts regarding initial entry and IE1 expression (25). Since HCMV infection frequently caused detachment of CHO cells and this detachment was efficiently avoided by human p180 overexpression (unpublished observation), we generated new cell lines coexpressing GFP-PML and the human host cell factor p180 (27) by transfection of CHO/GFP-PML cells (17) with pcDNAp180-54R using Fugene 6. Cells were selected with 100 μg/ml zeocin and 100 μg/ml G418. Coexpression of GFP-PML and p180 was confirmed by immunofluorescence microscopy and Western blotting (not shown). Transfectants containing abnormally large GFP-PML structures appeared to exhibit difficulties in keeping the structures stably; thus, we carried out limited dilution to carefully select a suitable clone with respect to the stability and morphogenesis of the GFP-PML bodies, and we finally cloned a cell line, SE/15, and used these cells throughout this study.

Inoculation of SE/15 cells.SE/15 cells were seeded in 24- or 96-well tissue culture plates precoated with 10 μg/ml fibronectin (GIBCO/BRL) at a cell density of 1 × 10⁵ or 0.2 × 10⁵ cells/well, respectively. After 24 h of culture, HCMV was inoculated at a multiplicity of infection (MOI) of 3 and the cells were incubated in F12-Dulbecco modified Eagle medium containing 5% fetal bovine serum (FBS) for 18 h at 37°C unless otherwise described.

Quantification of infected SE/15 cells with an NPB fluorescence assay.At various times postinfection (p.i.), HCMV- or mock-infected SE/15 cells were washed with phosphate-buffered saline (PBS) and gently extracted with 250 or 30 μl of NPB (50 mM Tris-HCl buffer containing 0.15% NP-40, 150 mM NaCl, 15 mM MgCl₂, and 5 mM EDTA) for 24- or 96-well plates, respectively, at 4°C for 15 to 20 min. If necessary, the solutions were centrifuged briefly and the supernatants were transferred to a μClear Fluorescence Black P 96-well plate (Greiner). The fluorescence intensity was measured with a CytoFluor multiwell plate reader (series 4000; Perseptive Biosystems) with an excitation wavelength of 485 nm and an emission wavelength of 530 nm. HCMV infectivity was expressed as the ratio of the net fluorescence intensity of the NPB extract to the fluorescence intensity of the total cell lysate prior to extraction, where the net fluorescence intensity was the NPB fluorescence intensity of infected cells after subtraction of the NPB fluorescence intensity of mock-infected cells. The total cell lysates prior to extraction were prepared in a parallel experiment by scraping of cells, followed by lysis with NPB in a tube.

Quantification of infected cell numbers by a GFP-PML imaging assay.For measurement of infected cell numbers, fluorescence signals were captured with an Olympus fluorescence microscope (model IX71) equipped with a cooled charge-coupled-device camera (DP50; Olympus) using a fixed data collection time of 0.2 s. Imaging analysis was performed with the public-domain software Image J (http://rsb.info.nih.gov/ij ) to obtain a series of data sets of particle numbers of each particle size. To evaluate the total cell numbers of uninfected SE/15 cells, a correlation coefficient was calculated from the numbers of GFP-PML dots and DAPI (4′,6′-diamidino-2-phenylindole)-stained nuclei. Under these conditions, GFP-PML dots that were too small were not counted, to avoid capturing signals that were unable to be distinguished from electrical noise, and hence the mean number of GFP-PML dots/nucleus was estimated to be less than 2. In the present study, we used the following formula for evaluating the total number of cells bearing GFP-PML bodies: total number of cells = 0.632 × total number of GFP-PML dots. HCMV infectivity was expressed as the ratio of the number of nuclei possessing diffuse GFP-PML to the total number of SE/15 cells harboring GFP-PML.

Susceptibility tests against antiviral agents.The susceptibilities to antiviral compounds were measured by the following methods.

(i) NPB fluorescence assay.For heparin inhibition assays, HCMV (Towne) was pretreated with heparin (0 to 200 μg/ml) for 1 h at 37°C and inoculated into SE/15 cells. For antibody neutralization assays, HCMV (Towne) was pretreated with monoclonal antibodies (MAbs) against HCMV gB or gH (Goodwin Institute), or a control mouse immunoglobulin G (IgG), at a final concentration of 25 or 100 μg/ml in the absence of complement for 1 h at 37°C. Quantification of the NPB fluorescence intensity was performed according to the standard protocol described above. Infectivity was expressed as the percentage of infectivity at each concentration of a drug to that in the absence of the same drug.

(ii) Coculture of clinical isolates and GFP-PML imaging assays.Since SE/15 cells do not support a full replication cycle of HCMV, we performed coculturing of SE/15 cells with infected HEL cells for susceptibility tests. HCMV clinical isolates were cultured in HEL cells in 24-well plates (1 × 10⁵ cells/well) in the presence of various concentrations of antiviral drugs for 48 h. The infected cells were trypsinized, briefly centrifuged, and mixed with a suspension of SE/15 cells. The ratio of cocultured, infected HEL cells to SE/15 cells was determined to be 1:2 through our preliminary study to define the optimum results. Next, the cells were seeded in triplicate onto fibronectin-precoated μClear Fluorescence Black P 96-well plates. After another 48 h of incubation in the presence of various concentrations of antiviral drugs, infectivity was measured with a GFP-PML imaging assay. (For a schematic diagram of the protocol, see Fig. 6A.) Curve fitting and 50% inhibitory concentration (IC₅₀) calculations were performed with the PRISM4 program (GraphPad Inc.).

(iii) Plaque reduction assay.HEL monolayers in 60-mm dishes were infected with HCMV by inoculating approximately 100 to 200 PFU of each virus inoculum. The cells were then overlaid with medium containing 1% methylcellulose, 2% FBS, and various concentrations of drugs. Assay cultures were incubated for 10 to 14 days, and the plaques were counted after fixation with 5% formalin in PBS and staining with 0.03% methylene blue. Triplicate samples were prepared for each drug concentration.

Antibody and immunofluorescence microscopy.The following MAbs were used: anti-HCMV IE1 (MAB810; Chemicon); anti-PML (PG-M3; Santa Cruz Biotechnology); and anti-calnexin (Transduction Laboratories). Normal mouse IgG was obtained from Santa Cruz Biotechnology. The cells were fixed with paraformaldehyde, permeabilized with 0.1% Triton X-100, and then incubated with the antibodies as described previously (27). Alexa Fluor 568-conjugated goat anti-mouse IgG was used as secondary antibody. Cells were imaged with a Zeiss LSM410 confocal microscopy system.

SE/15 cells have functional PML bodies labeled with GFP.To generate a simple monitoring system for HCMV-infected cells in vitro, we have cloned a CHO cell line, SE/15, stably expressing both GFP-PML protein and p180 (see Materials and Methods). SE/15 cells contained a number of small nuclear foci labeled with GFP (Fig. 1A) that were also costained for PML (data not shown). Upon HCMV infection, infected cells positive for IE1 exhibited a diffuse GFP pattern throughout the nucleus (Fig. 1B), the intensity of which appeared to be higher than that of IE1 (compare panels B1 and B2). The unique appearance of nuclei containing dispersed GFP-PML, which was never observed in mock-infected cells, began at 3 to 4 h p.i., and the number of these nuclei continued to increase until 10 to 16 h p.i. These data indicate that SE/15 cells have functional GFP-PML bodies whose morphology is specifically changed upon HCMV infection. Thus, visual assessment under fluorescence microscopy is sufficient for sensitive detection of infected cells without the necessity for any further procedures such as fixation and staining. To verify the specificity of HCMV-induced GFP-PML dispersion, HSV-1 (strain F) was inoculated onto SE/15 cells. No morphological change of GFP-PML caused by the virus was observed, mainly because SE/15 cells do not support initial entry of HSV-1 (data not shown). To establish a quantification method for infected SE/15 cells, we applied two strategies, namely, measurement of the GFP fluorescence intensity and direct counting of the infected nuclei by imaging analysis.

• Open in new tab
• Download powerpoint

FIG. 1.

Fluorescence microscopic images of SE/15 cells showing aberrant localization of GFP-PML in HCMV-infected cells. SE/15 cells were mock-infected (A1 to A3) or infected with Towne (MOI = 3) (B1 to B3), followed by staining for IE1 at 18 h p.i. (A1 and B1) GFP signals. (A2 and B2) IE1 staining. (A3 and B3) Merged images. Bars, 25 μm.

Diffuse GFP-PML in HCMV-infected cells is efficiently extracted with NPB.In normal cells, it has been reported that PML protein is present as an insoluble form in the nucleus, probably as multiprotein complexes anchored to the nuclear matrix, and that SUMOylation of PML protein is required for this distribution (23, 34). Since de-SUMOylation of PML protein occurs in HCMV-infected cells (22), it is supposed that the diffuse PML in the infected nuclei would be soluble in NP-40 buffer. To address this idea, the cells were gently treated with NPB. As expected, most of the GFP signals in the HCMV-infected nuclei disappeared after NPB treatment under fluorescence microscopy, while the small GFP-PML dots in mock-infected cells remained unchanged (Fig. 2A). Western blotting analyses confirmed effective extraction of GFP-PML into NPB from the infected cells but not from mock-infected cells (Fig. 2B). Thus, GFP-PML protein in SE/15 cells retains characteristics similar to PML protein expressed in HCMV-infected fibroblasts. Identical results were obtained after IE1 transfection, instead of HCMV infection, confirming that expression of IE1 alone is sufficient to induce this change (data not shown). These data indicate that NPB extraction-based fluorescence measurements in SE/15 cells allow quantification of HCMV-infected cells.

• Open in new tab
• Download powerpoint

FIG. 2.

GFP-PML is efficiently extracted into NPB following HCMV infection. SE/15 cells were mock-infected or infected with HCMV as described in the legend to Fig. 1, followed by mild extraction by NPB, which permeabilized the cells but did not lyse them. (A) Fluorescence microscopic images of GFP before (a and b) and after (c and d) NPB treatment are shown for mock-infected (a and c) and HCMV-infected (b and d) cells. (B) Western blotting data of total cell lysates (lanes 1 and 2) and NPB extracts (lanes 3 and 4) from HCMV-infected (lanes 1 and 3) and mock-infected (lanes 2 and 4) cells are shown, respectively. Calnexin was used as a loading control.

Quantification of HCMV-infected cells by GFP fluorescence after NPB extraction.To establish a sensitive assay, we first optimized the conditions for extraction. Through systematic screening of various conditions, the optimum conditions were determined to be as follows. At various times p.i., mock- and HCMV-infected cells were washed with PBS and gently extracted with 50 mM Tris-HCl buffer (pH 7.4) containing 0.15% NP-40, 150 mM NaCl, 15 mM MgCl₂, and 5 mM EDTA at 4°C for 20 min. The HCMV infectivity was expressed as the ratio of the net fluorescence intensity of the NPB extract to the fluorescence intensity of the total cell lysate, where the net fluorescence intensity was the NPB fluorescence intensity of infected cells after subtraction of the NPB fluorescence intensity of mock-infected cells. Due to the distinct background fluorescence of mock-infected cells, which probably arises through cell debris and non-GFP fluorescent background, subtraction was necessary for accurate estimations. By this protocol, infected cells at various times p.i. were estimated as shown in Fig. 3A. Consistent with the microscopic observations, the net NPB fluorescence was detectable as early as 4 h p.i., increased until 10 to 12 h p.i., and then almost reached a plateau. The correlation with the virus titer was tested with serially diluted HCMV inoculums in 12- or 96-well plates (Fig. 3B), and the results revealed a good correlation between the NPB fluorescence and the input virus over a diverse range. On the 12-well plate scale, HCMV-infected cells at an MOI of 0.4 were measurable. The NPB fluorescence was well correlated with IE1 expression levels in blotting analyses (Fig. 3C).

• Open in new tab
• Download powerpoint

FIG. 3.

Correlation of the NPB fluorescence assay with IE1 expression levels. (A) SE/15 cells were infected with Towne (MOI = 3) in a 24-well plate. At various times p.i., the standard NPB fluorescence assay was performed as described in the text. (B) SE/15 cells were infected, as described for panel A, in a 96- or 12-well plate at a MOI of 0.001 to 10. At 18 h p.i., the standard NPB fluorescence assay was performed. (C) Western blot analysis of total cell lysates from the experiment shown in panel B. Data are means ± standard deviations from three experiments.

Application of the NPB fluorescence assay to inhibition experiments with antiviral agents.Next, we tested whether our assay, termed the NPB fluorescence assay, can be applied to determination of the IC₅₀ values of heparin in inhibition experiments. As shown in Fig. 4A, HCMV infectivity estimated by the NPB fluorescence assay was well fitted with a sigmoid curve, and the IC₅₀ was estimated to be about 10 μg/ml, similar to previously reported values (24). Again Western blotting analysis revealed that the net fluorescence intensity of NPB extracts was well correlated with the expression levels of IE1 (Fig. 4B). Furthermore, the neutralizing abilities of MAbs against HCMV gB or gH were evaluated in the absence of complement (Fig. 4C). The assay was also feasible in 96-well plates (data not shown). These data suggest that our NPB fluorescence assay provides a simple and rapid procedure for evaluating inhibitors of the initial replication phase without the use of any immunological reagents within 10 to 16 h.

• Open in new tab
• Download powerpoint

FIG. 4.

The NPB fluorescence assay is applicable for determining the effects of infectivity inhibitors. (A and B) HCMV was preincubated for 1 h at 37°C with various concentrations of heparin, an entry process inhibitor of the virion into the cells. SE/15 cells were infected (MOI = 1) in a 24-well plate for 18 h, followed by the standard NPB fluorescence assay as described in the text. (C) HCMV was preincubated with a MAb against gB or gH for 1 h at 37°C and assayed as described for panel A. Data are means ± standard deviations from three experiments. Asterisks indicate a statistically significant difference between HCMV- and mock-infected cells by Student's t test (**, P < 0.01; *,P< 0.05).

Quantification of HCMV-infected cells with a GFP-PML imaging assay.Next, we investigated the application of our system to drug susceptibility testing of low-titer HCMV clinical isolates. Since our preliminary study showed that a higher sensitivity was required for assaying clinical isolates, we developed an imaging assay for this purpose rather than using the NPB fluorescence assay.

Since GFP-PML is dispersed throughout the nuclei of HCMV-infected SE/15 cells and the GFP-PML areas are markedly larger than the GFP-PML dots of uninfected cells (Fig. 1 or 2), imaging analysis of fluorescence micrographs enables direct counting of the number of infected nuclei. A typical profiling pattern by particle size of mock-infected cells gave one peak representing the small GFP-PML foci (Fig. 5A, left). The total number of SE/15 cell nuclei was evaluated from the total number of GFP dots by using this pattern, as described in Materials and Methods. In contrast, a second broad peak appeared in the profiling pattern of HCMV-infected cells, corresponding to the sizes of nuclei bearing a diffuse GFP signal (Fig. 5A, right). Therefore, HCMV infectivity was expressed as the ratio of the number of nuclei possessing a diffuse GFP signal to the total number of SE/15 cell nuclei. Using this procedure, termed the GFP-PML imaging assay, we titrated various HCMV strains, including clinical isolates, and compared the results with the number of PFCs determined in HEL cells. Figure 5B shows representative graphs of the well-correlated relationship observed between the infectivity estimated by the GFP-PML imaging assay and the numbers of PFCs, thereby confirming the feasibility of the GFP-PML imaging assay. For all of the strains tested, the correlation coefficients were >0.90.

• Open in new tab
• Download powerpoint

FIG. 5.

HCMV infectivity estimated by the GFP-PML imaging assay is correlated with the number of PFCs. (A) Profile patterns of the particle sizes of mock-infected (left) and Towne-infected (right) cells. (B) Typical graphs showing the correlation between the HCMV infectivity of clinical isolates estimated by the GFP-PML imaging assay and the number of PFCs determined in HEL cells. Data are means ± standard deviations from three experiments.

Application of the GFP-PML imaging assay to drug susceptibility testing of clinical isolates.We performed drug susceptibility tests of various HCMV strains by using the GFP-PML imaging assay. Since SE/15 cells do not support a full replication cycle, we cocultured SE/15 cells with infected HEL cells which had been cultured in the presence or absence of antiviral agents for 48 h prior to coculturing (Fig. 6A). At 96 h p.i., the HCMV infectivities at various concentrations of drugs were measured to evaluate the IC₅₀ values from their inhibition curves. Figure 6B shows representative graphs of a cell-associated laboratory strain (Davis) and a clinical isolate (01) for ganciclovir (GCV) or foscarnet (FOS). Subsequently, we compared the IC₅₀ values obtained for the nine clinical isolates (Table 1), including four GCV-resistant strains, with the values obtained with a conventional plaque reduction assay. With the GFP-PML imaging assay, the IC₅₀ values of GCV for the four resistant viruses were estimated to be >9 μM, which were higher than those of the sensitive strains (<5 μM). In contrast, the IC₅₀ values of FOS for all the viruses were estimated to be <200 μM. The IC₅₀ values determined by the GFP-PML imaging assay were relatively higher but still remained comparable to those obtained with the plaque reduction assay. These results suggest that the GFP-PML imaging assay is sensitive enough to distinguish GCV-resistant viruses from GCV-sensitive viruses. The IC₅₀ values of these isolates were also determined by a standard plaque reduction assay using cell-free viruses, and the results correlated well with the data from the GFP-PML imaging assay (Table 1).

• Open in new tab
• Download powerpoint

FIG. 6.

Drug susceptibility testing of clinical isolates by the GFP-PML imaging assay. (A) Schematic representation of the protocol for drug susceptibility testing of clinical isolates with the coculture system. (B) Data for drug susceptibility testing of the Davis strain and clinical isolate #01 for GCV and FOS using cell-associated viruses. Data are means ± standard deviations from three experiments.