Neutralizing Antibodies Limit Cell-Associated Spread of Human Cytomegalovirus in Epithelial Cells and Fibroblasts

Abstract

Human cytomegalovirus (HCMV) can cause severe clinical disease in immunocompromised individuals, such as allograft recipients and infants infected in utero. Neutralizing activity of antibodies, measured as the ability to prevent the entry of cell-free virus, has been correlated with the reduction in HCMV transmission and the severity of HCMV-associated disease. However, in vivo HCMV amplification may occur mainly via cell-to-cell spread. Thus, quantifying the inhibition of cell-to-cell transmission could be important in the evaluation of therapeutic antibodies and/or humoral responses to infection or immunization. Here, we established a quantitative plaque reduction assay, which allowed for the measurement of the capacity of antibodies to limit HCMV spread in vitro. Using an automated fluorescence spot reader, infection progression was assayed by the expansion of viral plaques during the course of infection with various GFP-expressing viruses. We found that in contrast to non-neutralizing monoclonal antibodies (mAbs), neutralizing mAbs against both glycoprotein B and H (gB and gH) could significantly inhibit viral plaque expansion of different HCMV strains and was equally efficient in fibroblasts as in epithelial cells. In contrast, an anti-pentamer mAb was active only in epithelial cells. Taken together, our data demonstrate that specific anti-HCMV mAbs can significantly limit cell-associated virus spread in vitro.

Keywords: antibodies; cell-associated spread; cell-to-cell spread; glycoproteins; herpesviruses; human cytomegalovirus (HCMV); neutralization; plaque size reduction.

Figure 1

Plaque reduction assay (PRA) for automated quantification of cell–associated HCMV spread. HFFs in 96-well plates were infected with 100 PFU/well of HCMV strain Towne. After incubation for 24 h, the medium was replaced by agarose overlay medium containing the indicated (a–c) inhibitors (20 µM GCV, 2 µM LMV) or (d–f) antibodies (50 µg/mL) or BSA (50 µg/mL), respectively. Starting at 4 dpi, images of the whole 96-well were captured by a Fluorospot reader each day for at least 18 dpi and used for automated quantification of the mean plaque size (1E-3 Sq.mm) of all fluorescent spots detected per well as described in Materials and Methods. All experiments were performed in triplicate. (a,d) Representative images taken by the Fluorospot reader at the indicated dpi following treatment of HCMV-infected HFFs with the (a) HCMV inhibitors or (d) antibodies. (a) The magnifications demonstrate the size expansion of individual plaques over the course of the infection (4 dpi versus 10 dpi). (a,d) The numbers in the upper left corner represent the calculated mean plaque sizes (1E-3 Sq.mm) of the corresponding wells. (d) In the lower right corner, the total number of fluorescent spot counts is indicated. (b,e) Time course analyses of the mean plaque sizes of HCMV-infected cells treated with or without (b) HCMV inhibitors or (e) antibodies. nt mAbs are colored in green, nnt and negative control mAbs are shown in red. (c,f) Bar graphs illustrating the mean plaque sizes at the time points post infection at which the maximum mean plaque size of (c) the inhibitor control (10 dpi) or (f) no antibody control (9 dpi) was reached. The dashed lines indicate the mean plaque sizes of the (c) no inhibitor or (f) no antibody controls at 4 dpi, which reflects the initial fluorescent spot size of a single round infection. Statistical analysis was performed by ordinary one-way analysis of variance (ANOVA). n.s., not significant; **, p  ≤  0.01; ***, p  ≤  0.001; ****, p  ≤ 0.0001; p values refer to antibodies vs. no antibody control.

Figure 2

HCMV spread in fibroblasts is sensitive to inhibition by anti-gB and anti-gH antibodies. HFFs in 96-well plates were infected with 100 PFU/well of HCMV strain Towne. After incubation for 24 h, the medium was replaced by agarose overlay medium with or without the indicated (a,b) anti-gH antibodies (50 µg/mL), (c,d) anti-gB antibodies (50 µg/mL), anti-gB and anti-gH antibody combinations (25 µg of each antibody to a final concentration of 50 µg/mL), or (e,f) anti-gB and anti-gH antibody dilutions (50 µg, 25 µg, 12.5 µg, and 6.25 µg), respectively. Starting at 4 dpi, whole 96-well images were captured each day for at least up to 18 dpi and used for automated quantification of the mean plaque size (1E-3 Sq.mm) of all individual fluorescent spots detected per well as described in Materials and Methods. All experiments were performed in triplicate. (a,c,e) Time course analyses of the mean plaque sizes of HCMV-infected cells treated with or without (a) anti-gH antibodies, (c) anti-gB antibodies or anti-gB and anti-gH antibody combinations, or (e) anti-gB and anti-gH antibody dilutions. (a) The nt anti-gH mAbs are colored in green, the strain-specific anti-gH mAb 2B10 which does not neutralize Towne is shown in red. (b,d,f) Bar graphs of the mean plaque sizes at the time points post-infection with maximum mean plaque sizes of the no antibody controls ((b), 9 dpi; (d), 10 dpi; (f), 11 dpi). The dashed lines indicate the mean plaque sizes of the no antibody control at 4 dpi, illustrating the initial fluorescent spot size of a single-round infection. Statistical analysis was performed by ordinary one-way analysis of variance (ANOVA). n.s., not significant; *, p  ≤  0.1; **, p  ≤  0.01; ***, p  ≤  0.001; ****, p  ≤  0.0001; p values refer to antibodies vs. no antibody control.

Figure 3

Anti-gH and -gB antibodies limit HCMV spread in fibroblasts in a strain-independent manner. HFFs in 96-well plates were infected with 100 PFU/well of different HCMV strains ((a,b) TB40/E; (c,d) TB40/E-del pentamer; (e,f) TR). After incubation for 24 h, the medium was replaced by agarose overlay medium with or without the indicated antibodies (50 µg/mL). Starting at 4 dpi, whole 96-well images were captured each day for at least up to 19 dpi and used for automated quantification of the mean plaque size (1E-3 Sq.mm) of all individual fluorescent spots detected per well as described in Materials and Methods. All experiments were performed in triplicate. (a,c,e) Time course analyses of the mean plaque sizes of HFF cells infected with the HCMV strains TB40/E (a), TB40/E-del pentamer (e), or TR (c). nt mAbs are colored in green, nnt and negative control mAbs are shown in red. (b,d,f) Bar graphs of the mean plaque sizes at the time points post-infection with maximum mean plaque sizes of the no antibody controls ((b), 9 dpi; (d), 10 dpi; (f), 11 dpi). The dashed lines indicate the mean plaque sizes of the no antibody controls at 4 dpi, which is indicative of the initial fluorescent spot size of a single-round infection. Statistical analysis was performed by ordinary one-way analysis of variance (ANOVA). n.s., not significant; *, p  ≤  0.1; **, p  ≤  0.01; ***, p  ≤  0.001; ****, p  ≤  0.0001; p values refer to antibodies vs. no antibody control.

Figure 4

Anti-gB and anti-gH mAbs are equally efficient in limiting HCMV spread in fibroblasts and epithelial cells. ARPE-19 (ARPE) cells seeded in 96-well plates were infected with 100 PFU/well of HCMV strains (a,b) TB40/E and (c,d) TR, respectively. After incubation for 24 h, the medium was replaced by agarose overlay medium with or without the indicated antibodies (50 µg/mL). Starting at 4 dpi, whole 96-well images were captured each day for at least up to 21 dpi and used for automated quantification of the mean plaque size (1E-3 Sq.mm) of all individual fluorescent spots detected per well as described in Materials and Methods. All experiments were performed in triplicate. (a,c) Time course analyses of the mean plaque sizes of ARPE cells infected with the HCMV strains TB40/E (a) or TR (c). nt mAbs are colored in green, nnt and negative control mAbs are shown in red. (b,d) Bar graphs illustrating the mean plaque sizes at the time point of maximum mean plaque size of the no antibody controls (10 dpi). The dashed lines indicate the mean plaque sizes of the no antibody control at 4 dpi, which is indicative of the initial fluorescent spot size of a single-round infection. Statistical analysis was performed by ordinary one-way analysis of variance (ANOVA). n.s., not significant; *, p  ≤  0.1; **, p  ≤  0.01; ***, p  ≤  0.001; ****, p  ≤  0.0001; p values refer to antibodies vs. no antibody control. (e,f) Comparison of the capacity of the indicated mAbs to reduce the mean plaque size (given in % relative to the no antibody control after subtraction of the basic fluorescence mean spot size = mean plaque size of the no antibody control at 4 dpi) following infection of HFF as well as ARPE cells with TB40/E (e) and TR (f), respectively. Red, no significant reduction in mean plaque size (<30%); yellow, significant reduction in mean plaque size (30–49%); green, highly significant reduction in mean plaque size (>50%); n.d., not determined.