Human cytomegalovirus (HCMV), a member of the Betaherpesvirinae subfamily, is a ubiquitous pathogen that infects approximately 60 to 80% of the adult population worldwide (Boeckh M, et al, 2011). It generally causes asymptomatic latent infection in immunocompetent populations, but leads to serious illness and even death among immunocompromised populations, such as transplant recipients, infants with an immature immune system, and patients infected with human immunodeficiency virus (HIV) (Lazzarotto T, et al, 2011; Sung H, et al, 2010). Thus, HCMV infection is an important public-health problem. There are some anti-viral drugs currently used to treat HCMV infection in clinical practice, but their curative rates are not satisfactory, and there are issues such as severe side effects and emergence of drug-resistant HCMV strains (Griffiths PD, 2002; Manley K, et al., 2011). The most effective preventive measure against HCMV infection will therefore be vaccination; unfortunately, more than 30 years of research efforts have not yielded a usable HCMV vaccine (Schleiss MR, et al, 2005). Hence, the development of a CMV vaccine has been assigned the highest priority by the US Institute of Medicine (Arvin AM, et al, 2004; Zhong J, et al, 2007).
HCMV has broad cell tropism, and the pathway for HCMV entry into cells is very complex, involving interactions of multiple viral envelope glycoproteins and a series of cell receptors (Britt WJ, et al, 1996). The HCMV envelope mainly contains three glycoprotein complexes (gC Ⅰ, Ⅱ, and Ⅲ), which are all necessary for virus infection (Wang D, et al, 2005). Of these, the gC Ⅱ complex, consisting of the gM (also known as gpUL100) and gN (gpUL73) proteins, is the most abundant glycoprotein complex. The gM protein accounts for 10% of the net weight of the virus, and studies have shown that knockout of either the gM or gN genes is a lethal mutation (Britt WJ, et al., 2004; Krzyzaniak M, et al., 2007). The gC Ⅱ complex binds to the heparan sulfate proteoglycan and is inferred to play a role in the initial steps of virus entry into cells. In addition, the gM protein also participates in virus replication and assembly (Krzyzaniak M et al, 2007), and the carboxyl terminal of the gN protein plays a key role in virus envelopment (Mach M, et al, 2007). It has been reported that natural HCMV infection can induce antibody responses against the gC Ⅱ complex (Mach M, et al, 2000; Shimamura M et al, 2006), and that monoclonal antibodies against gM or gN are capable of neutralizing HCMV infection, indicating that, similar to the gB and gH proteins, the gM and gN proteins are also major target antigens in anti-viral immune responses.
Currently, HCMV vaccine development is focusing mainly on the gC I antigen gB, and the related vaccine has entered clinical Phase Ⅲ trials. However, clinical studies have shown that the protection offered by the gB protein is only about 50% (Pass RF, et al, 2009), therefore, it is necessary for research and development to be carried out into new antigens that give greater protection. As not only is the gM/gN complex highly conserved in the Herpesviridae family and capable of inducing neutralizing antibody (NAb) response, but it is also the most abundant glycoprotein of HCMV and has become an attractive candidate vaccine antigen. Shen et al. reported that HCMV gM and gN DNA vaccines induced neutralizing antibodies that, in in vitro tests, demonstrated neutralizing activity against multiple HCMV strains (Shen S, et al., 2007). However, as CMV infection has strict species specificity, no animal model is available for studying the mechanisms of HCMV infection and immunity, and so far no in vivo protective study using gM-gN as vaccine antigens has been reported. Mice infected with murine cytomegalovirus (MCMV) has been the most commonly used animal model for simulating HCMV infection (Brune W, et al, 2001; Qureshi, MH, et al, 2005). So far there have been few reports describing the gM/gN proteins of MCMV. The M100 and M73 open reading frames of the MCMV genome encode the homologs of gM and gN, respectively. MCMV gM is speculated to comprise eight transmembrane domains and four N-linked glycosylation sites. Anthony et al. found that MCMV gM was completely conserved between six MCMV strains, suggesting that gM is an antigenically conserved protein (Scalzo A A, et al., 1995).
Generation of a DNA vaccine has long been an important direction for CMV vaccine research and development. Currently, a number of DNA vaccines have entered into clinical trials, such as VCL-CB01 (Phase Ⅱ) and VCL -CT02 (Phase Ⅰ) (both Vical Inc., San Diego, CA, USA), in which the protective antigens selected are gB/pp65 and gB/pp65/IE1, respectively. In this study, we prepared gM and gN DNA vaccines based on sequences of the MCMV Smith strain, and the two DNA vaccines were then tested separately and in combination in a lethal MCMV infection mouse model. The results showed that gM-gN antigens have good immunogenicity, and co-immunization with gM and gN DNA vaccines was able to provide the mice with complete protection against a lethal MCMV infection.
To test the efficacy of the gM/gN DNA vaccines, BALB/c mice were divided into seven groups and vaccinated with vector pDNA, pgB, pgM, pgN, or a combination of pgM and pgN at a dose of 50 μg for each DNA. Three weeks after the last immunization, the mice were challenged with a lethal dose of SG-MCMV (5×LD50). On day 5 post-challenge, three mice from each group were euthanased, and the spleens removed aseptically for measurement of viral titer. The remaining 10 mice from each group were monitored for 21 days for weight loss and survival.
As shown in Table 1, all mice in the negative control group G (3×pV50) died within the first week after challenge, mice in the positive control group A (3×pgB50) had a survival rate of 50%, and mice immunized with pgM (group B: 3×pgM50) or pgN (group C: 3×pgN50) alone had a survival rate of 30% or 20%, respectively. For the co-immunization groups, mice immunized three times with the high dose (group D: 3×pgM/pgN50) were fully protected, and their survival rate was 100%, whereas reducing the number of co-immunizations by one-third (group E: 2×pgM/pgN50) or the dose by one-half (group F: 3×pgM/pgN25) resulted in the survival rate dropping to 70% and 30%, respectively. In addition, mice in the negative control group showed clear signs of infection after challenge, presenting as lethargy, piloerection, anorexia, and emaciation. The first of the negative control mice died on day 3 and all died within 1 week after viral challenge, whereas the mice in the immunized groups survived for longer than those in the negative control group (Fig. 1A).
Table 1. Protection provided by gM, gN, or gM + gN DNA vaccines against lethal SG-MCMV challenge in mice after immunizationa
Figure 1. Survival (A) and body-weight losses (B) after lethal challenge in mice immunized with pgM/pgN DNAs. Mice were immunized with pgB, pgM, or pgN, respectively, or co-immunized with gM/gN pDNAs at a dose of 50 μg. Three weeks later, mice were infected with a lethal dose of SG-MCMV. Survival and body-weight changes of the mice were monitored 21 days post-challenge. Data points represent mean ± SD of each group in(B).
When the residual virus load in the spleen on day 5 post-challenge was compared, the titers of groups A, B, D, E, and F were significantly lower than that in the negative control group G, while the titer of group C (3×pgN50) was not significantly different from that of the control group. Further, when the immunized groups were compared, the pgM/pgN co-immunization groups had lower viral titers than the pgM or pgN group, and viral load decreased as the immunization dose increased. In particular, viral titer in group D (3×pgM/pgN50) was about 100 times lower than that in the negative control group G (Table 1), indicating that the immune responses induced by pgM/pgN coimmunization were able to effectively clear the virus in mice.
After challenge, the body weight of the mice changed in parallel with the change in other health indicators. Mice in the negative control group had the highest weight loss, nearly 30%. Mice in the single-antigen immunization groups showed clear symptoms of infection after challenge, and their weight loss was obvious but slightly less than the negative control group. Mice in the pgM/pgN co-immunization groups had milder symptoms of infection and less weight loss than mice in the single-antigen immunization groups, indicating that co-immunization provided mice with better protection (Fig. 1B).
Two weeks after the last vaccination, serum samples were collected for determination of the NAb levels elicited by pgM/pgN immunization. As shown in Fig. 2, no NAb could be detected in sera from the negative control group. For the single-antigen groups, the relative NAb titer levels in immune sera were: pgN < pgM < pgB. NAb titer levels in pgM/pgN groups were higher than that in the pgM or pgN group, and increased in line with the immunization times and the doses. NAb titers in groups D (3×pgM/pgN50) and E (2×pgM/pgN50) were significantly higher than that in the pgM or pgN groups, indicating that co-immunization of pgM and pgN induced stronger NAb response than immunization with each antigen alone.
Figure 2. Neutralization titers of sera obtained from pgM/pgN immunized mice. Mice were vaccinated with gB, gM, or gN alone, or a mixture of gM/gN pDNAs at a dose of 50 μg. Sera were collected 2 weeks later and tested. Results were the highest serum dilutions at which 50% reduction in viral titer was achieved. Values represent the means ± SD of each group. aSignificant difference (p < 0.05) versus the mice in the negative control group. bSignificant difference (p < 0.05) versus the mice in gM or gN groups.
Immunoblot assay was performed to determine the specificity of the gM or gN immune sera. 293T cells were transfected with pgM or pgN alone or in combination (pgM/pgN). Cell lysates from MCMV-infected 3T3 cells were used as positive control. The polyclonal antiserum against gM reacted with proteins of approximately 86 kDa and 32 kDa in 293T cells transfected with pgM or pgM/pgN (Fig. 3A, lanes 2 and 3), and the lysates from MCMV-infected 3T3 cells also showed bands of a similar molecular weight (Fig. 3A, lane 5). Anti-gM sera also recognized a band with a molecular weight of about 125 kDa in transfected cells, along with the 86 kDa band mentioned above, which may represent the aggregates of gM protein, as gM is highly hydrophobic and insoluble, and readily forms high-molecular-weight aggregates (Fig. 3A, lanes 2 and 3). Fig. 3B showed that anti-gN sera identified a band of around 48 kDa from 293T cells transfected with pgN or pgM/pgN (Fig. 3B, lanes 2 and 3), similar to the band in infected 3T3 cells (Fig. 3B, lane 5). Anti-gN sera also reacted with a protein of around 15 kDa in transfected cells, which may be the gN form without post-translational modification (Fig. 3B, lanes 2 and 3). This was consistent with the previous finding that HCMV gN is a highly glycosylated envelope protein which is translated into an 18-kDa protein before being posttranslationally processed (Mach M, et al, 2000). When anti-gM/gN polyclonal sera were used, the 48 and 32 kDa bands were detected simultaneously in pgM/pgN-cotransfected 293T or MCMV-infected 3T3 cells (Fig. 3C, lanes 1 and 3). According to previous studies, HCMV gM and gN are able to form a complex with a molecular weight of 70 to 100 kDa (Mach M, et al, 2000). In the current study, lysates from MCMV pgM/pgN co-transfected 293T cells also showed some bands of between 70 and 100 kDa in size, similar to the bands in MCMV-infected 3T3 cells. The western blot results suggested that both the gM and gN DNA vaccines were able to express the proteins in mammalian cells, and that the anti-sera from gM or gN pDNA immunized mice had fine specificity. These anti-sera were also tested in immunofluorescence assays in MCMV-infected cells. As shown in Fig. 4, MCMV-infected 3T3 cells showed clear morphological changes, with the cell shape changing from the typical fibroblast shape to expanded round cells, the nucleus changing from oval to irregular, and some adjacent cells congregating and forming syncytia. Detection was first performed using anti-sera from pgM or pgN immunized mice, respectively, and this showed that in infected cells, gM or gN proteins were mainly distributed in the juxtanuclear regions rather than throughout the entire cytoplasm. A similar result was obtained with anti-sera from the pgM/pgN co-immunization group; the gM/gN complex proteins were mainly distributed in the juxtanuclear regions and not in the nucleus. The immunofluorescence results also indicated that the gM/gN immune sera had good specificity.
Figure 3. Immunoblotting analysis of transiently expressed gM and gN proteins with antisera to gM and gN. 293T: cell lysates of 293T cells transfected either with the vector DNA, pgM, or pgN, or co-transfected with pgM/pgN. 3T3: cell lysates of 3T3 cells infected (+) or not infected (-) with MCMV Smith strain. Lysate proteins were resolved on 10% SDS-PAGE gels, transferred to PVDF membranes, and immunoblotted with (A) polyclonal anti-gM, (B) anti-gN or (C) anti-gM/gN sera.
Figure 4. Confocal microscope analyses of 3T3 cells infected with MCMV Smith strain. Infected 3T3 cells were harvested 3 days post-infection and then fixed, permeabilized, and stained for immunofluorescence. Anti-gM/gN was detected with FITCconjugated anti-mouse IgG. Nuclei were stained with Hoechst 33258. Specific immunofluorescence was observed with a confocal laser scanning microscope. (Scale bar =10 μm).
To further define whether immunization with pgM/pgN can induce cellular immune responses, an ELISPOT assay was conducted. Cell-mediated immunity (CMI) to DNA vaccines was measured by quantifying the number of IFN-γ secreting splenocytes in immunized mice. Mice were inoculated three times with pgM, pgN, or pgM/pgN at a dose of 50 μg each. Data are presented as the average number of spots in triplicate stimulant wells.
The stimulant selected was inactivated MCMV virions, as gM is the most abundant virus glycoprotein and is present on the virus envelope in the form of gM-gN complex. As shown in Fig. 5, immunization with the pgM or pgN vaccine was able to induce some level of cellular immune response, and pgM induced a stronger response than pgN. For mice co-immunized with pgM/pgN, the number of specific IFN-γ-secreting splenocytes was significantly higher than that in mice immunized with pgM or pgN alone, being nearly four times higher than in pgM-immunized mice, thus indicating that pgM/pgN co-immunization was able to induce stronger cellular immune responses. In the negative control group, stimulation of splenocytes with inactivated MCMV resulted in only a few non-specific spots (spots ≤10 per 106 cells), which was similar to the background value of the ELISPOT plates (spleen cells without antigen stimulation). When splenocytes were stimulated with concanavalin, the non-specific IFN-γ positive spots were up to 2000 per 106 cells (data not shown).
Figure 5. Cellular immune responses of mice vaccinated with gM/gN DNA vaccines. Mice were immunized with gM or gN or a mixture of gM and gN pDNAs at a dose of 50 μg. The control group was inoculated with vector plasmid. Splenocytes were isolated 2 weeks later and stimulated in vitro with 10 μg/mL of the inactivated MCMV. The results represent the geometrical means ± SD of each group. a Significant difference (p < 0.05) versus the mice in control group. b Significant difference (p < 0.05) versus the mice in gM or gN groups.