Glycoprotein N of human cytomegalovirus protects the virus from neutralizing antibodies

Abstract

Herpes viruses persist in the infected host and are transmitted between hosts in the presence of a fully functional humoral immune response, suggesting that they can evade neutralization by antiviral antibodies. Human cytomegalovirus (HCMV) encodes a number of polymorphic highly glycosylated virion glycoproteins (g), including the essential envelope glycoprotein, gN. We have tested the hypothesis that glycosylation of gN contributes to resistance of the virus to neutralizing antibodies. Recombinant viruses carrying deletions in serine/threonine rich sequences within the glycosylated surface domain of gN were constructed in the genetic background of HCMV strain AD169. The deletions had no influence on the formation of the gM/gN complex and in vitro replication of the respective viruses compared to the parent virus. The gN-truncated viruses were significantly more susceptible to neutralization by a gN-specific monoclonal antibody and in addition by a number of gB- and gH-specific monoclonal antibodies. Sera from individuals previously infected with HCMV also more efficiently neutralized gN-truncated viruses. Immunization of mice with viruses that expressed the truncated forms of gN resulted in significantly higher serum neutralizing antibody titers against the homologous strain that was accompanied by increased antibody titers against known neutralizing epitopes on gB and gH. Importantly, neutralization activity of sera from animals immunized with gN-truncated virus did not exhibit enhanced neutralizing activity against the parental wild type virus carrying the fully glycosylated wild type gN. Our results indicate that the extensive glycosylation of gN could represent a potentially important mechanism by which HCMV neutralization by a number of different antibody reactivities can be inhibited.

Figure 1. Amino acid sequences of the parent gN protein and the different gN deletions.

The amino acid sequence of the gN strain AD169 (residues 1–101) is shown in the top row. Deletions of amino acids are indicated by dots for the truncated molecules. The putative signal sequence (residues 1–21, http://www.cbs.dtu.dk/services/SignalP/) and part of the membrane anchor domain (starting at residue 93, http://www.cbs.dtu.dk/services/TMHMM/) are indicated by bold letters.

Figure 2. Truncations in gN do not prevent gM/gN complex formation.

Cos7 cells were transfected with the indicated plasmids, and protein expression was assayed by reactivity with mab 14-16A (anti-gM/gN), anti-myc (detection of gN) and anti-TGN46 (trans-Golgi apparatus). Reactivity was detected with the appropriate secondary antibodies. The appearance of yellow in the merged pictures indicates colocalization of proteins. In the merge panel cell nuclei are also stained blue.

Figure 3. Characterization of the gN-truncated viruses.

A) Replication kinetics of gN-truncated viruses. Fibroblasts were infected with identical infectious doses of the different viruses at day zero. At the indicated days post infection supernatants from the infected cultures were harvested and virus titer was determined. B) gM/gN complex formation and intracellular localization in infected fibroblasts. Human fibroblasts were infected with the indicated recombinant viruses and protein expression was assayed 72 h later in indirect immunofluorescence analysis by reactivity with murine mabs specific for gN (14-16A), gM (IMP91-3/1) and gB (human mab C23). The appearance of yellow in the merged pictures indicates colocalization of signals. In the merge panel cell nuclei are also stained blue. C) Analysis of the gM/gN complex in extracellular virus particles. Lysates from gradient purified extracellular virions were subjected to western blot analysis under reducing (left panel) and non-reducing conditions (right panel). The western blot under reducing conditions was developed using antibodies against the major capsid protein (MCP) plus antibody 27–287, recognizing the gp58 part of gB. The analysis under non-reducing conditions was stained using a gM/gN affinity-purified human polyclonal antibody preparation or the gB-specific mab 27–287. D) ELISA analysis of glycoproteins in extracellular virus particles. Plates were coated with lysates from gradient-purified extracellular virions at a concentration of 500 ng/well. Individual proteins were detected by mabs specific for gB (27–287), gH (14-4B), the major capsid protein (MCP), gN (14-16A) and the appropriate secondary antibodies. E) Inhibition of infection by heparin. Heparin was added to extracellular virus for 1 h at the indicated concentrations. The mixture was added to fibroblasts and allowed to infect for 4 h at 37°C. The virus/heparin mixture was removed and percent of infected cells was determined 24 h later by indirect immunofluorescence using an antibody specific for IE-1.

Figure 4. Monoclonal antibodies to gB, gH and gN neutralize the gN-truncated viruses more efficiently than the parent virus.

Virus and antibody were incubated for 1 h at 37°C and the mixture was added to HFF monolayers. Cells were washed 4 h later and percent infected cells was determined at 24 h post infection by indirect immunofluorescence using an antibody specific for IE-1. Target proteins and the corresponding mab are indicated. Every mab was tested at least two times with similar results. Data were analyzed by 2way ANOVA and Bonferroni posttests. * p<0.05, **p<0.01, ***p<0.001.

Figure 5. A fraction of polyvalent human sera neutralize the gN-truncated viruses more efficiently than the parent virus.

Virus and antibody were incubated for 1 h at 37°C and the mixture was added to HFF monolayers. Cells were washed 4 h later and percent infected cells was determined at 24 h post infection by indirect immunofluorescence using an antibody specific for IE-1. Serum identification is shown above the respective graph. Shown are representative results from at least 3 independent analyses. Data were analyzed by 2way ANOVA and Bonferroni posttests. * p<0.05, **p<0.01, ***p<0.001.

Figure 6. Mechanistic aspects of neutralization.

A) Parent virus and gN-truncated viruses adsorb comparably to fibroblasts. Virus (m.o.i. 0.5) was incubated with the indicated antibodies (5 µg/ml) for 1 h at 37°C and cooled to 4°C. The virus/antibody mixture was added to HFF and incubated for 1 h at 4°C. Lysates were prepared and processed for quantitative real time PCR analysis. The virus only sample was set to 100% and used to calculate the remaining samples. B) Post-adsorption neutralization is enhanced for gN-truncated viruses. HFF were adsorbed with virus at a m.o.i. of 0.2 at 4°C for 1 h. Antibody at the indicated concentrations was added and the culture was shifted to 37°C. Extent of infection was analyzed 24 h later by indirect immunofluorescence using an antibody for IE-1. C) gN-truncated viruses are neutralized with faster kinetic than the parent virus. Virus and antibody were mixed at 4°C and added to precooled HFF monolayer cells. The mixture was incubated for the indicated time at 37°C and the percent infected cells was determined at 24 h post infection by indirect immunofluorescence using an antibody specific for IE-1.

Figure 7. Murine sera raised against the gN-truncated viruses show increased neutralization capacity for the gN-truncated viruses.

At an interval of 4 weeks, Balb/c mice were twice immunized i.p. with 5 µg each of gradient purified RVAD169, RVgN-41sig and RVgN-61sig, respectively. Four weeks after the second immunization, animals were boosted with 2 µg virions i.v. and the serum was collected 2 weeks later. Each pool consisted of sera from 3 animals. A) ELISA assay of the pooled sera using gradient purified virus as antigen (left panel) and purified recombinant gB (right panel). Mock: Serum pool from control mice immunized with PBS. 27–287: gB specific mab. B) Neutralization capacity of the serum pools against wild type and the gN-truncated viruses. The serum pools were tested twice with similar results.

Figure 8. Sera from gN-truncated viruses have different antigen specificity.

The individual serum pools were analyzed in an ELISA for reactivity against antigenic domains on gB (AD1, AD2 and AD4), gH (AP86) and the tegument protein pp150 (pp150). Control antibodies for antigen coating of the ELISA plates included 27–287 (gB-AD1, light grey), SA4 (AD86, middle grey) and C23 (gB-AD2, dark grey).