Detection of gH/gL/gO expression in vitro

The gH, gL, gO-expressing plasmids were prepared as described in material and methods section. To detect their expression in vitro, gH, gL, and gO pDNAs were co-transfected into 293T cells and Western blot was performed. As shown in Fig. 1, three proteins were effectively expressed in the co-transfected cells. The gH protein was ~86 kDa, gL was ~42 kDa, and the ~125 kDa band might be the highly glycosylated mature gO protein, which were consistent with the results of single gene transfection (data not shown). Moreover, a ~250 kDa protein band was detected, which should be the heterotrimer formed by gH, gL and gO proteins, and the size was consistent with the complex detected by Western blot in MCMV infected 3T3 cells. We further performed immunofluorescence assays either with or without permeabilization on pgH/pgL/pgO transfected or MCMV infected 3T3 cells to detect the expression and localization of three proteins. In permeabilized cells (facilitating antibody to enter the cell cytoplasm and nucleus), as shown in Fig. 2A, gH, gL and gO expressed from transfection of a corresponding single plasmid was distributed in the cytoplasm; and gH was also localized on nuclear membrane. When the three proteins were co-expressed, they also mainly localized in the cytoplasm and were not present in the nucleus. For virus infected 3T3 cells (Fig. 2B), three proteins were mainly localized in juxtanuclear regions and were not present in the nucleus, which was consistent with the immunofluorescence observations of HCMV infected cells reported by Theiler et al [18]. In addition, staining of nonpermeabilized cells transfected with pgH revealed fluorescent staining was mainly concentrated at the periphery of cells, indicating the presence of cell surface-localized gH protein. Meanwhile, staining of nonpermeabilized cells transfected with single gL or single gO pDNA failed to detect any specific fluorescence. In contrast, when co-transfecting with gH/gL/gO pDNAs, a fraction of gL and gO were detected to be membrane-localized (Fig. 2C). These results indicated that gH is a transmembrane protein and a substantial fraction were readily surface-localized in transfected cells, while gL and gO may display the membrane-localization pattern when interacting with gH by forming the heterotrimer complex (gH/gL/gO). With respect to MCMV infected cells, a substantial fraction of gH, gL and gO were localized in the cell surface (Fig. 2D). Taken together, these results indicated that pgH/pgL/pgO co-transfection yielded effective expression of the proteins, and they shared the same intracellular localization and could form the gCIII complex.

Protection provided by pgH/pgL/pgO immunization

BALB/c mice were randomly divided into 9 groups. The groups A-D were respectively immunized with 50 μg pgB (positive control), pgH, pgL and pgO; the group I received empty vector DNA (negative control). The groups E, F, G and H were respectively co-immunized with pgH/pgL, pgL/pgO, pgH/pgO and pgH/pgL/pgO at a dose of 50 μg DNA total. Mice were immunized three times at three-week intervals, and were challenged with a lethal dose of SG-MCMV. The protective effects of the DNA vaccines were evaluated comprehensively using infection symptoms, weight loss, residual spleen viral loads and survival.

After the lethal challenge, symptoms were the most evident in the negative control group (group I), such as piloerection, lethargy, anorexia, arching back, emaciation, and all mice died within one week of infection. In comparison, the vaccine groups had fewer deaths and no further death occurred after one week (Fig. 3A). For the positive control, the pgB immunized mice had 50% survival. When the four viral antigens were compared for protection against the challenge, there were no significant differences among them. The survival of mice received pgH, pgL, or pgO single DNA immunization was 40%, 20% and 50%, respectively, indicating gH and gO provided similar immune protection as gB while gL was less protective (Table 1). The combination of two genes, gH/gL (Group E), gL/gO (group F), and gH/gO (group G) with 50 μg dose total and three immunizations gave survival of 80%, 40% and 60%, respectively, which showed no significant differences among the two gene co-immunization groups. The best, and the only full protection, was seen in mice given gH/gL/gO three genes co-immunization (group H).

Virus load in the challenged mice was measured by determining spleen virus titer. Very high virus titer was seen in mice in the negative control group, while virus titers were more or less lower in the vaccine groups. The titers in the co-immunization groups were lower than those of the single gene immunization groups. The most remarkable difference was found with the gH/gL/gO co-immunized mice (group H), as its spleen virus titer was about 10⁴-fold lower than the control (group I), indicating that immune responses induced by gH/gL/gO co-immunization could effectively clear virus in vivo (Table 1).

Body weight of the challenged mice changed along their health situation. Weight loss was most marked 5–7 days post-infection, for the negative control mice, maximal weight loss reached near 30% (Fig. 3B). The gH, gL, or gO single gene immunized mice had obvious symptoms and weight losses after challenge. Mice in the higher survival groups had less weight losses. In particular, mice given gH/gL/gO co-immunization (group H) showed little signs of infection and their weight losses were significantly less than other immunized groups (Fig. 3B). The above results demonstrated that immunization with a single gH or gO DNA vaccine could offer mice some protection, and co-immunization with all three genes of gCIII (gH/gL/gO) markedly improved the protective efficacy and provided mice full protection.

Neutralizing antibody response to pgH/pgL/pgO immunization

Neutralizing antibodies (Nab) in sera of immunized mice were detected by plaque reduction assay. As shown in Fig. 4, NAb titer was undetectable in the negative control group. The NAb titer levels in the pgB, pgH, pgO single gene immunization groups were similar and the level in the gL group was slightly lower. The NAb levels in the co-immunization groups were higher than those in the single immunization groups. At the immunization dose of 50 μg, the NAb level in gH/gL/gO co-immunization group (group H) was 100+ fold higher than that in the gH single immunization (group B) and the difference was very significant. Also NAb titer in the gH/gL/gO co-immunization group was significantly higher than those from the two gene co-immunization groups. Therefore, co-immunization with gH/gL/gO could induce the best neutralizing antibody responses.

Cell-mediated immune response to immunization

Cell-mediated immune (CMI) responses to pgH/pgL/pgO immunizations were assessed by measuring IFN-γ secretion in mouse splenocytes. BALB/c mice were immunized with gH, gL, gO or a mixture of three pDNAs at a dosage of 50 μg for three times. Data were presented as the average number of spots in triplicate stimulant wells.

As indicated in Fig. 5, the number of specific IFN-γ secreting splenocytes in the gH, gO and gH/gL/gO immunized mice was significantly higher than that of the control group (p<0.05), while no specific response was detected with the two stimulating peptides for the gL immunization group. Also CMI responses in the gH/gL/gO co-immunization group was significantly higher than that in the single gene immunization groups. For the control group, only a small number of non-specific spots were detected by polypeptides (number of spots ≤ 10/10⁶ cells), which was on the same level as the background value of the Elispot plate (un-stimulated splenocytes). The number of positive spots obtained with concanavalin stimulation was as high as 2000/10⁶ cells (data not shown). The above results demonstrated that gH and gO DNA vaccine could elicit a certain level of CMI response and gH/gL/gO co-immunization could elicit a highly effective CMI response.

Dose-dependent protective efficacy of pgH/pgL/pgO co-immunization

As described above, pgH/pgL/pgO co-immunization could elicit excellent immune protection. To determine the minimum dose of pgH/pgL/pgO for providing mice full protection, mice were immunized with one of the three doses, 50 μg (16.7 μg each), 10 μg (3.3 μg each) and 3 μg (1 μg each) for two or three times, respectively, and then challenged with a lethal dose of SG-MCMV.

As seen from Table 2, the best protection was seen in mice immunized with pgH/pgL/pgO at the 50 μg dose. Twice immunization could already offer 100% protection. For the 10 μg dose group, three immunizations also provided mice full protection, and two immunizations provided 80% protection. The protection was lower in the 3 μg dose group in which three immunizations could provide mice 50% protection. The neutralizing antibody level in the pgH/pgL/pgO co-immunization groups increased as dose and frequency of immunization. Both the 50 and 10 μg dose groups had significantly higher neutralizing antibody level compared with the corresponding 3 μg dose groups. Spleen virus titers showed a clear decreasing trend as immunization dose and frequency increased. The spleen virus titers in all but the 3 μg dose immunization groups were significantly lower than that in the negative control group. Meanwhile, the 50 and 10 μg dose groups had significantly lower virus loads than that in the 3 μg dose groups. Three immunizations of the gH, gL or gO single gene immunization at 50 μg dose provided maximal of 50% protection (Table 1), whereas three gene co-immunization at 3.3 μg each provided 100% protection (Table 2), suggesting that pgH/pgL/pgO co-immunization could induced more stronger immune protection.

Passive immunization with anti-gH/gL/gO sera

As antisera from pgH/pgL/pgO co-immunized mice showed excellent neutralizing activity in vitro, passive immunization was carried out to test whether anti-gH/gL/gO sera could protect mice from lethal MCMV infection. Antisera were prepared from mice immunized three times with 50 μg pgH/pgL/pgO DNA vaccine, and the control sera were prepared from mice immunized with the empty vector. The two types of sera were administered to 10 naïve mice via tail vein injection at 300 μl per mouse. Within 24 hours, these mice were challenged with a lethal dose of SG-MCMV. As shown in Table 3, all the mice in the positive control group (pgH/pgL/pgO immunization) survived the lethal challenge and showed no significant signs of infection; all the mice received with the anti-gH/gL/gO sera also obtained 100% protection and only showed slight signs of infection. In contrast, mice received with the control sera were all died. This result demonstrated that the antibody response elicited by pgH/pgL/pgO co-immunization plays important roles in immune protection. This experiment also indicated that passive immunization with anti-gH/gL/gO sera could be used to prevent and treat CMV infection.

Ethics statement

Six- to eight-week-old female BALB/c mice were purchased from the Center for Disease Control and Prevention in Hubei Province, China. They were bred in specific-pathogen-free conditions in the Animal Resource Center at the Wuhan Institute of Virology, Chinese Academy of Sciences. According to the experiment procedures, mice were housed in individually ventilated cages with free access to food and water in a temperature-controlled and humidity-controlled room maintained under filtered positive-pressure ventilation under a 12 hour light/12 hour dark cycle. All experiments involving animals have been reviewed and approved by the Animal Care Committee of Wuhan Institute of Virology (Permit Number: WIVA04201201), in accordance with the animal ethics guidelines of the Chinese National Health and Medical Research Council (NHMRC). All efforts were made to minimize animal suffering.

Virus and cells

MCMV Smith strain was used and propagated in NIH 3T3 cells. 3T3 and 293T cells were cultured in Minimal Essential Medium (MEM) containing 10% fetal calf serum (FCS). High virulent salivary gland-derived MCMV (SG-MCMV), propagated in BALB/c mice by 10 serial in vivo passages for virulence enhancement, was used in challenge experiments. Challenge was performed with 5 × LD₅₀ virus stock.

Plasmid construction and peptides

DNA vaccines were constructed by cloning the complete open reading frame of gH (M75), gL (M115) and gO (M74) gene from MCMV Smith strain into the eukaryotic expression vector pcDNA3.1. The corresponding primers for cloning gH, gL or gO were as follows:

• gH-F: 5'-CGGGATCCATGAAGTTGTCATTAATACTCTCCATCG-3',
• gH-R: 5'-GCTCTAGATTATCTTTTTTGCCGGCACAGCCGGTAC-3';
• gL-F: 5'-CGGGATCCATGATGCCTTTATTATTGCTCATACTG-3',
• gL-R: 5'-CGGAATTCACGGTCTCTTTCGTTGATATTGAGGG-3';
• gO-F: 5'-CGGAATTCATGAACCCCTTATTACTCATGTCG-3',
• gO-R: 5'-TTGGGCCCTCA GACACGGCTAAAGGATATTGAG-3'.

We also constructed gB pDNA as a positive control, which only encoded the extracellular domain of gB. All constructs were verified by sequencing in full. The plasmids were propagated in Escherichia coli DH5α bacteria and purified using NucleoBond Xtra kit (MACHEREY-NAGEL GmbH & Co. KG).

The peptide TGPSVRALME for gH, the peptide VSPPVLSVLV and LHDDGPIRPDPYRF for gL and the peptide TVVGPVNVTTLYK for gO, which were used for IFN-γ Enzyme Linked Immunospot assay (ELISPOT), were synthesized by GenScript Co., Ltd.

Immunization and challenge

BALB/c mice were vaccinated three times, 3 weeks apart, with empty vector, gB, gH, gL, gO or co-immunized with gH/gL, gH/gO, gL/gO or gH/gL/gO pDNAs at a dosage of 50 μg (25 μg each in the mixture of two pDNAs, 16.7 μg each in the mixture of three pDNAs) by injection into the right quadriceps muscle. Then in vivo electroporation was administered as described by Aihara and Miyazaki [36]. Negative control mice were inoculated with the empty vector pcDNA3.1 by electroporation.

For lethal challenge experiments, the more virulent SG-MCMV was used. The 50% lethal dose (LD50) of SG-MCMV stock was approximately 10⁵ PFU in BALB/c mice. At 3 weeks postimmunization, mice were challenged with a lethal dose (5 × LD₅₀, 200 μl / mouse) of SG-MCMV by intraperitoneal injection. This infection could cause systemic virus replication in mice and death of all unvaccinated mice within one week after the challenge. The mice were weighed and checked every day in order to monitor weight loss, apparent physical condition (bristled hair and wounded skin) and behaviour. Then the mice were humanely euthanized via cervical dislocation after chloroform (inhalation excess) in all cases in order to minimize or avoid animal suffering.

Immunoblotting analysis

293T cells were plated into six-well plates. At 24 h after plating, cells were transfected with vector plasmid or co-transfected with pgH/pgL/pgO using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. The cells were harvested and lysed 48 h after transfection. For infection study, 3T3 confluent cells were infected with MCMV (m.o.i = 1) for 1.5 h at 37°C. Unadsorbed virus was removed and infection medium (MEM containing 2% FBS) was added. Cells were collected 3 days postinfection and subsequently lysed. Cell lysates were separated using non-reducing SDS-PAGE (10% gel), blotted onto PVDF membranes and immunoblotted with polyclonal antisera to gH/gL/gO. The polyclonal antisera used in the experiment were obtained from immune serum of mice. The anti-gH sera, anti-gL sera and anti-gO sera derived from the corresponding pgH, pgL and pgO immunized mice, respectively. The anti-gH/gL/gO sera were taken from three plasmids pgH/pgL/pgO co-immunized mice.

Indirect immunofluorescence assay (IIFA)

Immunofluorescence assays were performed in permeabilized and nonpermeabilized cells, respectively. 3T3 cells were seeded on glass coverslips in tissue culture dishes. After 24 h, cells were infected with MCMV at an m.o.i. of 0.1. Cells were collected 3 days postinfection. For transfection, after overnight growth to reach 70% confluence, the 3T3 cells were transfected with gH, gL, gO or co-transfected with gH/gL/gO pDNAs using Lipofectamine 2000 (Invitrogen), and the cells were harvested 48 h later. For permeabilized cell staining, all the cell samples were washed with PBS, fixed with 4% paraformaldehyde, permeabilized with 0.25% Triton X-100, blocked with 5% non-fat milk, and stained with polyclonal antisera to gH, gL or gO. Next, the fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG (Millipore) secondary antibodies were added and then stained with Hoechst 33258 for 10 min. For nonpermeabilized cell surface staining, unfixed cells were washed and pre-blocked with blocking buffer and then incubated with primary antibody (anti-gH, gL or gO serum) in blocking buffer for 1 hour at room temperature to detect surface-localized viral glycoproteins. After primary antibody incubation and washing, trasfected or infected 3T3 cells were fixed with 4% paraformaldehyde and washed before adding secondary antibody FITC-conjugated anti-mouse IgG. Fluorescent image analysis was performed on a Leica laser scanning confocal microscope.

Detection of MCMV-specific neutralizing antibody

Mouse serum samples were collected 3 weeks after the last immunization. Serum samples were stored at -20°C until use. Neutralizing antibody directed against MCMV were determined by a plaque reduction assay as described before [37]. Briefly, decomplemented sera (30 μl) were serially diluted 2-fold with MEM. Each dilution was mixed with 100 PFU MCMV in 30 μl of MEM and were then incubated 1 hour at 4°C and 1 hour at 37°C. The mixture was layered onto 3T3 monolayers and PFU were calculated by the standard plaque assay. A neutralization titer was expressed as the highest serum dilution required to achieve a 50% reduction in the number of plaques.

IFN-γ ELISPOT assay

Mice were immunized with gH, gL, gO, or co-immunized with gH/gL/gO pDNAs thrice at a dosage of 50 μg by electroporation. Two weeks after the third immunization, splenocytes were isolated for ELISPOT assays. According to the instruction manual (U-CyTech, Netherlands), immunospot plates (Millipore, Bedford, MA) were coated with rat anti-mouse IFN-γ mAb, incubated at 4°C overnight and then blocked with 200 μl of blocking solution R. Subsequently, 2×10⁵ lymphocytes were added to the wells in triplicate, stimulated with 10 μg/ml of corresponding gH, gL, gO peptides or gH/gL/gO polypeptides mixture (for co-immunization group). After 18 hours, the lymphocytes were discarded and biotin-labeled anti-mouse IFN-γ Ab antibody was added to each well and incubated at 37°C for 1 h. Next, diluted Streptavidin-HRP conjugate solution was added and incubated at room temperature for 2 hours. Finally, the plates were treated with 100μl of AEC substrate solution and incubated at room temperature for 20 min in the dark. The reaction was stopped by washing with dematerialized water. Spots were quantified by an ELISPOT reader (Bioreader 4000; Bio-sys, Germany).

Titration of MCMV in spleen

Five days postchallenge, 3 mice from each group were humanely euthanized with chloroform (inhalation excess) and then bled from the heart with a syringe. After bleeding, the spleens was taken out and homogenized in 1:10 (w/v) volume of MEM containing 10% calf serum. The homogenized fluids were centrifuged and the supernatants stored in aliquots at -80°C. Viral loads were determined using a plaque-forming cell assay as described before [12].

Passive serum transfer and virus challenge

Two groups of 20 mice were immunized thrice with 50 μg gH/gL/gO pDNAs or empty vector pcDNA3.1. Two weeks after the third immunization, 40 mice were humanely euthanized with chloroform (inhalation excess) and blood was obtained by heart puncture. Naive BALB/c mice were passively immunized with the pooled anti-gH/gL/gO sera or control sera by tail vein injection in a volume of 300 μl, respectively. Then 24 hours later, all the mice were challenged with lethal dose (5 × LD₅₀) of SG-MCMV. Body weight losses and survival rates of mice were determined 5 days and 21 days post-challenge, respectively. Then, the mice were humanely euthanized via cervical dislocation after chloroform (inhalation excess) in all cases in order to minimize or avoid animal suffering.

Statistical analyses

The experimental results were evaluated by One-Way ANOVA (SPSS 17.0 software for Windows). The difference was considered statistically significant when P-value was less than 0.05. Fisher’s exact test was used to compare survival rates of mice in experimental and control groups.