Neutralizing Antibodies to Human Cytomegalovirus Recombinant Proteins Reduce Infection in an Ex Vivo Model of Developing Human Placentas

Human cytomegalovirus (HCMV) is the leading viral cause of congenital disease and permanent birth defects worldwide. Although the development of an effective vaccine is a public health priority, no vaccines are approved. Among the major antigenic targets are glycoproteins in the virion envelope, including gB, which facilitates cellular entry, and the pentameric complex (gH/gL/pUL128-131), required for the infection of specialized cell types. In this study, sera from rabbits immunized with the recombinant pentameric complex were tested for their ability to neutralize infection of epithelial cells, fibroblasts, and primary placental cell types. Sera from rhesus macaques immunized with recombinant gB or gB plus pentameric complex were tested for HCMV neutralizing activity on both cultured cells and cell column cytotrophoblasts in first-trimester chorionic villus explants. Sera from rabbits immunized with the pentameric complex potently blocked infection by pathogenic viral strains in amniotic epithelial cells and cytotrophoblasts but were less effective in fibroblasts and trophoblast progenitor cells. Sera from rhesus macaques immunized with the pentameric complex and gB more strongly reduced infection in fibroblasts, epithelial cells, and chorionic villus explants than sera from immunization with gB alone. These results suggest that the pentameric complex and gB together elicit antibodies that could have potential as prophylactic vaccine antigens.


Introduction
Human cytomegalovirus (HCMV) is the most common viral infection transmitted during pregnancy and is a cause of severe disease outcomes in newborns [1][2][3]. Permanent birth defects are most frequent following primary maternal infection in the first trimester of gestation. These infections can lead to intrauterine growth restriction (IUGR) and fetal impairment from viral replication in the placenta and developing organs or hypoxia associated with placental pathology and brain damage [1,2,[4][5][6][7][8][9]. Protection against maternal-fetal transmission depends on developing neutralizing antibodies to the viral glycoproteins and on cell-mediated immunity. Delayed T-cell responses, reduced antibodies to key viral proteins, and low antibody avidity after primary infection are associated with an increased risk of transmission [1,[10][11][12][13][14].

Virus Stocks and Microneutralization Assays in MRC-5 and ARPE-19 Cells
HCMV neutralizing assays were performed in MRC-5 and ARPE-19 cells using strains AD169 (American Type Culture Collection, Manassas, VA, USA), UxCA (kindly provided by Dr. Stuart Adler, Virginia Commonwealth University), and VR1814. Virus stocks were prepared in MRC-5 as follows: Cells were infected with VR1814, harvested along with the supernatant, and flash-frozen when infection reached~90%. After thawing, cell debris was removed by centrifugation, and the supernatant was aliquoted and stored at −80 • C. Virus titers were measured in a microtitration assay by infecting MRC-5 cells and counting HCMV immediate-early (IE1)-positive cells using a CTL-ImmunoSpot ® S6 Analyzer (Cellular Technology, Cleveland, OH, USA). A high-throughput quantitative micro-neutralization assay was performed on both MRC-5 and ARPE-19 cells in 96-well plates. The input virus (approximately 500 PFU/well) was mixed with media containing heat-inactivated serum or serial 3-fold dilutions of antisera and incubated for 1 h at 37 • C. The mixtures were then added to cells for 24 h to allow virus adsorption. HCMV-infected cells were quantified by counting immunostained cells with an anti-HCMV IE1 antibody (MilliporeSigma, Burlington, MA, USA), and an Alexa Fluor 488-conjugated goat antimouse Ig (H+L) secondary antibody (Thermal Fisher Scientific, Waltham, MA, USA). Counting was performed using a CTL-ImmunoSpot S6 Analyzer. The serum dilution that produced 50% inhibition for a given test sample was calculated by interpolation using Microsoft Excel.

Virus Neutralization Assays in Primary Cells and Anchoring Villus Explants
For assays using rabbit sera, the following cell types were seeded in 96-well plates and infected at the MOIs indicated: CTBs (MOI 0.08-0.1), AmEpCs (MOI 3.8), TBPCs (MOI 0.1), and HPFs (MOI 0.6). The stock virus was preincubated with media alone (no serum control), heat-inactivated preimmune serum (negative serum), or Cytogam, an HCMV hyperimmune globulin preparation (CSL Behring LLC, King of Prussia, PA, USA), at the indicated final dilutions for 1 h at 37 • C with moderate agitation. Virus-antibody mixtures were added to cells for 2 h in duplicate wells, and cells were washed and further incubated in the medium. At 2 dpi, cells were fixed, permeabilized, and immunostained with a mAb against HCMV IE1 (MilliporeSigma, Burlington, MA, USA). Infected cells were counted, and the percentage of cells infected at each dilution was determined relative to the average number of infected cells in the no-antibody control wells (n = 6). Fifty percent neutralizing titers were calculated using Microsoft Excel by the method of Reed and Muench. For assays using rhesus sera, ARPE-19 cells and AmEpCs were seeded on coverslips in 24-well plates. Virus MOI was adjusted to give a total of 600 to 800 infected cells/well by an untreated virus. Percentages of infected cells and 50% neutralization titers were determined as above.
For assays in anchoring villus explants, chorionic villi were dissected from human placentas (8-, 12-, and 14-week gestation), and 20 to 30 villus explants from each placenta were plated on Millicell-CM inserts (0.4 µm pore size, MilliporeSigma, Rocklin, CA, USA) coated with Matrigel in DMEM/F12 (1:1) with 10% FBS, penicillin/streptomycin, and amino acids, conditions that enable the differentiation of villus CTBs into invasive cells [18]. At 20 h after attachment to Matrigel, villus explants were infected overnight with VR1814 Vaccines 2022, 10, 1074 5 of 18 (3-4 × 10 6 PFU), preincubated with control and test sera as described above, washed, and cultured for two more days before fixation and frozen embedding. For quantification of infection in developing cell column CTBs , 1-3 sections of each explant were immunostained  with mAb 7D3 and a mAb against HCMV IE1 (MilliporeSigma, Burlington, MA, USA), and all cell columns were imaged on a Leica DMi8 microscope with a Leica DFC9000GT camera controlled by Leica Application Suite X software. Images were imported into Adobe Photoshop and cell columns were identified by cytokeratin staining and cell morphology. Cell column nuclei [DAPI (4 ,6-diamidino-2-phenylindole) channel] were quantified in ImageJ using the particle analysis function, and HCMV IE1-positive nuclei were counted manually. The aggregate percentage of cell column CTBs infected in each condition from a placenta (1-2 explants per condition) was determined by combining numbers from all cell columns analyzed.

Immunofluorescence and Imaging
Cells grown on coverslips were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. Frozen tissues were cut into 5 µm sections. For double and triple immunostaining, after blocking with 3-5% normal serum matching the secondary antibody source for cells and BSA plus 3-5% normal serum for explants, cells or tissue sections were simultaneously incubated with primary antibodies from different species followed by incubation with secondary antibodies labeled with fluorescein isothiocyanate (FITC) or rhodamine red-X (RRX) (Jackson ImmunoResearch, West Grove, PA, USA). Nuclei were stained with DAPI (Vector Laboratories, Burlingame, CA, USA). Alternatively, cells and tissues were incubated with primary antibodies against cellular proteins overnight, followed by incubation with secondary antibodies, then stained with antibodies to HCMV proteins. Images were obtained using a Leica DMi8 microscope as above. Images of whole explant sections were taken on a Leica M125 stereomicroscope equipped with a Leica MC170HD camera.

Anti-Pentamer Antibodies from Immunized Rabbits Potently Block Virus Entry into Fibroblasts and Epithelial Cells
To determine whether sera from rabbits immunized with recombinant HCMV pentamer (RBT pentamer) contain neutralizing antibodies, we measured neutralizing titers of sera against infection by different HCMV strains on ARPE-19 epithelial cells and MRC-5 fibroblasts ( Figure 1). Marked differences in serum neutralizing titers were found between these two cell types. Anti-pentamer antisera blocked infection by clinical strains (VR1814 and UxcA) more potently in ARPE-19 cells than in MRC-5 cells. The average 50% neutralization titers (NT50) in epithelial cells were generally 1-2 logs higher than those in fibroblasts. In addition, the kinetics of neutralizing antibody activities were different between the two cell lines. Neutralizing titers on MRC-5 cells were almost undetectable at week four after initial immunization but increased dramatically by week six, two weeks after the first boost, and sustained similar levels thereafter. For ARPE-19 cells, neutralizing titers were several-fold greater at week four than for MRC-5 cells and continued to increase after the first and second boosts, reaching NT50 values~10 5 at week nine, one week after the third boost.
fibroblasts. In addition, the kinetics of neutralizing antibody activities were different between the two cell lines. Neutralizing titers on MRC-5 cells were almost undetectable at week four after initial immunization but increased dramatically by week six, two weeks after the first boost, and sustained similar levels thereafter. For ARPE-19 cells, neutralizing titers were several-fold greater at week four than for MRC-5 cells and continued to increase after the first and second boosts, reaching NT50 values ~10 5 at week nine, one week after the third boost. Figure 1. Anti-pentamer rabbit immune sera neutralize HCMV infection of MRC-5 and ARPE-19 cells. Sera were collected at multiple time points from rabbits immunized with RBT pentamer, serially diluted, and tested for their ability to neutralize infection by the laboratory HCMV strain AD169 and the clinical strains VR1814 and UxcA on both MRC5 (fibroblast) and ARPE-19 (epithelial) cells. Fifty percent neutralization titers (NT50s) were determined for each serum sample and plotted as shown. Initial immunization (week 0) was followed by booster shots at weeks 4 and 8, as indicated by arrows. Bars = mean NT50 values, dots = values for individual serum samples.

Neutralizing Activities of Anti-Pentamer Rabbit Sera on Placental Cells
In these experiments, we determined whether RBT pentamer-specific rabbit sera protected against the infection of primary and limited-passage human placental cell types, including HPFs, TBPCs, and AmEpCs. The neutralizing activities of sera from five rabbits at six weeks post-immunization were compared with the neutralizing activities of HIG (Cytogam). All sera exhibited substantially lower neutralizing activities in HPFs ( Figure  2A) than in MRC-5 cells (Figure 1), with NT50 values well below 10 2 , more than one log lower than the values measured in MRC5 cells. Serum #4 had very low neutralizing activity, even at the 1:10 dilution (~32% inhibition; NT50 < 10 1 ). Cytogam exhibited higher neutralizing activity in these cells, with an NT50 of 3 × 10 2 . Negative control sera showed no appreciable neutralizing activity at the 1:10 dilution (not shown).
Next, we measured neutralizing activities on multipotent TBPCs-precursors of mature syncytiotrophoblasts and CTBs [52]. TBPCs are fully permissive for HCMV infection, but viral entry is independent of the pentameric complex, based on their susceptibility to a UL131A deletion mutant and the finding that anti-gB and anti-gH/gL reduce infection cells. Sera were collected at multiple time points from rabbits immunized with RBT pentamer, serially diluted, and tested for their ability to neutralize infection by the laboratory HCMV strain AD169 and the clinical strains VR1814 and UxcA on both MRC5 (fibroblast) and ARPE-19 (epithelial) cells. Fifty percent neutralization titers (NT50s) were determined for each serum sample and plotted as shown. Initial immunization (week 0) was followed by booster shots at weeks 4 and 8, as indicated by arrows. Bars = mean NT50 values, dots = values for individual serum samples.

Neutralizing Activities of Anti-Pentamer Rabbit Sera on Placental Cells
In these experiments, we determined whether RBT pentamer-specific rabbit sera protected against the infection of primary and limited-passage human placental cell types, including HPFs, TBPCs, and AmEpCs. The neutralizing activities of sera from five rabbits at six weeks post-immunization were compared with the neutralizing activities of HIG (Cytogam). All sera exhibited substantially lower neutralizing activities in HPFs ( Figure 2A) than in MRC-5 cells (Figure 1), with NT50 values well below 10 2 , more than one log lower than the values measured in MRC5 cells. Serum #4 had very low neutralizing activity, even at the 1:10 dilution (~32% inhibition; NT50 < 10 1 ). Cytogam exhibited higher neutralizing activity in these cells, with an NT50 of 3 × 10 2 . Negative control sera showed no appreciable neutralizing activity at the 1:10 dilution (not shown).
Next, we measured neutralizing activities on multipotent TBPCs-precursors of mature syncytiotrophoblasts and CTBs [52]. TBPCs are fully permissive for HCMV infection, but viral entry is independent of the pentameric complex, based on their susceptibility to a UL131A deletion mutant and the finding that anti-gB and anti-gH/gL reduce infection whereas anti-pentamer mAbs targeting pUL130-131 do not [42,43]. All sera had low to moderate neutralizing activities on TBPCs, with NT50 values between 2.9 × 10 2 and 3.4 × 10 3 , 2-4-fold higher than in HPFs ( Figure 2B). Cytogam was more potent, with an NT50 value of 7.9 × 10 3 . Neither of the negative control sera tested exhibited measurable neutralizing activity (not shown).
Next, we carried out neutralizing assays using primary AmEpCs isolated from amniotic membranes. AmEpCs are self-renewing, with stem cell characteristics [56] that support persistent HCMV infection [15]. In cases of congenital infection, HCMV-positive amniotic fluid facilitates diagnosis of transmission [1,9,40,57,58]. Consistent with an earlier study, all sera strongly neutralized infection in AmEpCs in a dose-dependent manner [43]. The two most potent sera (#2 and #3) reduced infection by~97% and~85%, respectively, at a 1:100,000 dilution, with NT50 values both exceeding 10 5 ( Figure 2C). NT50 values for the other three sera ranged from 6.4 × 10 3 to 1.6 × 10 4 . Overall, NT50 values were 2-4 orders of magnitude higher on AmEpCs than on HPFs and TBPCs. By comparison, the NT50 value for Cytogam was somewhat below those of sera #2 and #3. Both negative control sera tested exhibited slight neutralizing activity (~11% inhibition of infection) at the highest concentration (1:10 dilution; not shown). Interestingly, relative activities of individual immune sera were comparable between HPFs and TBPCs, but they did not fully predict their relative activities on AmEpCs.

Anti-Pentamer Rabbit Sera Block Viral Entry into Primary CTBs
We next measured neutralizing activities on primary CTBs isolated from three midgestation placentas. After isolation, CTBs initially have an epithelial-like phenotype but subsequently differentiate when cultured on Matrigel, switching to an endothelial-like phenotype [59,60]. All sera neutralized infection dose-dependently, but potencies varied both among sera, as above, but also significantly by cell donor, both in the average magnitude of neutralization and in the relative potencies of individual sera ( Figure 2D-F).
To assess neutralizing activities in the tissue environment, we performed assays on anchoring villus explants from a placenta at 14 weeks of gestation. In villus explants, CTBs differentiate and invade the Matrigel substrate prior to infection of anchoring villi. We reported that VR1814 replicates in differentiating CTBs in proximal cell columns and reduces outgrowth [61]. mAbs to pUL128-131 of the pentamer potently neutralize infection in cell column CTBs of anchoring villi at~100-fold lower concentrations than mAbs with targeting specificities for gB and gH [43]. Explants were infected with VR1814-serum mixtures or virus alone and fixed and frozen at 3 dpi, and fixed-frozen sections were immunostained for HCMV IE1 and CK, a CTB marker. Based on the results of primary CTBs from second-trimester placentas, serum dilutions up to 1:1000 were used for explant studies. This dilution achieves 28-63% inhibition of virus infection in primary CTBs ( Figure 2D-F). Total and infected CTBs were counted in 371 columns, and the aggregate percentage of infected cells was determined for each condition. In explants infected with VR1814 alone,~13% of 7D3-positive CTBs were infected, while negative sera exhibited no neutralizing effect (15-30% infected). All three sera tested (#2, #3, and #5) suppressed infection by >99% at the 1:1000 dilution (not shown).

Sera from Rhesus Macaques Immunized with gB and Pentamer Block HCMV Infection of Epithelial Cells and Fibroblasts
To evaluate the ability of gB and pentamer to elicit neutralizing antibodies against VR1814 infection in a mammalian model, rhesus macaques were immunized with either recombinant gB produced in mammalian cells (Sino gB) or with recombinant gB and pentamer, both produced in insect cells (RBT gB and RBT pentamer). The gB was not stabilized in the prefusion conformation. Neutralization titers of immunized animals were determined by a microneutralization assay as described (Materials and Methods). All animals in this study were seropositive for rhesus CMV (RhCMV) and had low preexisting titers of cross-neutralizing antibodies against HCMV. As shown in Figure 3A, compared with the sera from week zero (preimmune), sera from week six (after prime and boost) in animals immunized with gB alone showed little increase in neutralization titers in fibroblasts and only a slightly greater increase in neutralization titers in ARPE-19 cells. In contrast, sera from animals immunized with both gB and pentamer achieved~10-fold higher titers in fibroblasts, and~100-fold higher titers in epithelial cells ( Figure 3B). with the sera from week zero (preimmune), sera from week six (after prime and boost) in animals immunized with gB alone showed little increase in neutralization titers in fibroblasts and only a slightly greater increase in neutralization titers in ARPE-19 cells. In contrast, sera from animals immunized with both gB and pentamer achieved ~10-fold higher titers in fibroblasts, and ~100-fold higher titers in epithelial cells ( Figure 3B). . Fifty percent neutralization titers (NT50s) are shown for sera from each of five animals in each group (dots), as well as mean NT50 for each condition (bars). (C) Neutralization of VR1814 infection of AmEpCs from a 38.6-week-gestation placenta by sera collected at six weeks post-immunization from rhesus macaques immunized with RBT gB + RBT pentamer. NT50 values calculated from the mean neutralization curves (n = 2) are shown for each sample.
Next, we tested the efficacy of the sera (six weeks post-immunization) from rhesus macaques immunized with a combination of gB and the pentamer in neutralizing infection in primary AmEpCs from placentas at delivery (38.6 weeks). All sera at the 1:100 dilution almost completely abolished infection ( Figure 3C). NT50 values ranged from 3.1 to 8.6 × 10 3 , with A9N041 exhibiting the greatest activity and A9N045 the lowest.

Antibodies from Rhesus Macaques Immunized with gB and Pentamer Protect against Infection of CTB Cell Columns in Anchoring Villus Explants
We next assessed the neutralizing activities of selected sera from animals immunized with gB and pentamer on anchoring villus explants from placentas at 8 and 12 weeks of Next, we tested the efficacy of the sera (six weeks post-immunization) from rhesus macaques immunized with a combination of gB and the pentamer in neutralizing infection in primary AmEpCs from placentas at delivery (38.6 weeks). All sera at the 1:100 dilution almost completely abolished infection ( Figure 3C). NT50 values ranged from 3.1 to 8.6 × 10 3 , with A9N041 exhibiting the greatest activity and A9N045 the lowest.

Antibodies from Rhesus Macaques Immunized with gB and Pentamer Protect against Infection of CTB Cell Columns in Anchoring Villus Explants
We next assessed the neutralizing activities of selected sera from animals immunized with gB and pentamer on anchoring villus explants from placentas at 8 and 12 weeks of gestation. Up to three independent sections of each of 38 explants were immunostained for CTBs and HCMV IE proteins, and both the number of CTBs and number of infected CTBs were counted in every cell column identified. A total of 376 cell columns from 95 tissue sections were examined, and the aggregate percentages of infected cells were determined for each explant and condition (Figures 4-6). In the 8-week-gestation explants infected with VR1814 incubated with preimmune serum (negative serum), 22% of 7D3-positive CTBs were also positive for HCMV IE proteins, whereas preincubation with immune sera from all three animals suppressed infection in a dose-dependent manner ( Figure 4A). Relative to VR1814 treated with negative serum, all immune sera suppressed infection of cell column CTBs at dilutions of 1:10 (0-1.2% infected), 1:100 (2.4-12% infected), and 1:1000 (13-15% infected). Similarly, in the 12-week-gestation explants infected with VR1814 treated with negative serum, 27% of CTBs in cell columns were infected ( Figure 4B), and all immune sera exhibited high neutralizing activities at dilutions of 1:10 (0.4-0.7% infected), 1:100 (1.3-1.7% infected), and 1:1000 (5.3-11.7% infected). Representative images of tissue sections stained for CTBs and HCMV IE proteins from control and treated explants are shown in Figure 4 (eight-week gestation) and Figure 5 (12-week gestation).

Discussion
In the present study, we showed that animals immunized with the pentameric complex and gB produce neutralizing antibodies that reduce infection by pathogenic HCMV strains of a broad range of cell types and an ex vivo model of developing early-gestation human placentas. Sera from pentamer-immunized rabbits had neutralizing titers that potently blocked infection by pathogenic viral strains in fibroblast and epithelial cell lines, primary and low-passage placental TBPCs, AmEpCs, CTBs, and differentiating CTBs in anchoring villus explants. Sera from naturally RhCMV infected rhesus macaques subsequently immunized with HCMV gB blocked infection on fibroblast and epithelial cell lines; sera from naturally RhCMV-infected rhesus macaques subsequently immunized with HCMV gB and pentamer had increased neutralizing titers in the cell lines and blocked infection of primary AmEpCs and differentiating CTBs in anchoring villus explants. These results suggest that the combination of the pentameric complex and gB elicits antibodies that provide broad neutralizing activity to the different cell types and thus could reduce viral transmission. These antigens have potential as the basis for a prophylactic vaccine, though the independent role of gB in the neutralization of the virus in placental-derived cells was not demonstrated by these experiments.
Rabbits immunized with a soluble pentamer had higher titers of HCMV neutralizing antibodies on ARPE-19 cells and lower titers on MRC-5 cells and blocked infection of

Discussion
In the present study, we showed that animals immunized with the pentameric complex and gB produce neutralizing antibodies that reduce infection by pathogenic HCMV strains of a broad range of cell types and an ex vivo model of developing early-gestation human placentas. Sera from pentamer-immunized rabbits had neutralizing titers that potently blocked infection by pathogenic viral strains in fibroblast and epithelial cell lines, primary and low-passage placental TBPCs, AmEpCs, CTBs, and differentiating CTBs in anchoring villus explants. Sera from naturally RhCMV infected rhesus macaques subsequently immunized with HCMV gB blocked infection on fibroblast and epithelial cell lines; sera from naturally RhCMV-infected rhesus macaques subsequently immunized with HCMV gB and pentamer had increased neutralizing titers in the cell lines and blocked infection of primary AmEpCs and differentiating CTBs in anchoring villus explants. These results suggest that the combination of the pentameric complex and gB elicits antibodies that provide broad neutralizing activity to the different cell types and thus could reduce viral transmission. These antigens have potential as the basis for a prophylactic vaccine, though the independent role of gB in the neutralization of the virus in placental-derived cells was not demonstrated by these experiments.
Rabbits immunized with a soluble pentamer had higher titers of HCMV neutralizing antibodies on ARPE-19 cells and lower titers on MRC-5 cells and blocked infection of primary placental cell types, most potently AmEpCs. Interestingly, the relative potencies of the sera varied across cell types and among different isolates of CTBs. These results suggest important differences between cells from different tissues and between cells from different individuals. The results underscore the importance of evaluating vaccine-elicited antibodies across a range of primary placental cell types and, when feasible, across multiple donors to accurately model vaccine effectiveness in human populations. Importantly, the rabbit immune sera were able to inhibit infection of both primary CTBs from three different placentas and differentiate CTBs in cell columns of anchoring villus explants. These cells are likely to be the most important conduits of maternal-fetal transmission.
Ongoing efforts to determine naturally protective immune factors have identified maternal immune responses associated with reduced risk of transmission in cases of congenital HCMV infection. These include the early development of neutralizing antibodies to the HCMV glycoprotein pentamer complex [10,62] and increased antibody avidity [8,9,33,36,63,64]. The pentamer has been extensively studied as a vaccine candidate, as the target of the most potent neutralizing antibodies and critical for infection of specialized human cells [28,45,62]. Analysis of sera from HCMV seropositive women showed that neutralizing activities on epithelial cells were considerably higher than on fibroblasts [65]. Antibodies to the pentamer were the majority of the anti-HCMV neutralizing antibody response in HIG [40], and depletion of the pentamer significantly decreased neutralizing activity on epithelial cells [41].
Although the species-specific characteristics of CMV make preclinical testing of HCMV vaccine candidates in animal models a challenge for predicting clinical outcomes, the immunization of rhesus macaques with all five HCMV pentamer subunits resulted in high and sustained neutralizing titers that blocked epithelial and endothelial cell infection and inhibited entry into epithelial cells and fibroblasts [45,46,49]. These results are consistent with the neutralizing antibody responses induced by HCMV during natural infection [27,30,66]. Vaccine-derived HCMV pentamer-specific neutralizing antibodies in animal sera potently prevent virus transfer from endothelial cells to leukocytes and spread in epithelial cells [46], important mechanisms of virus dissemination in vivo, as well as infection of CTBs [43,45]. Inasmuch as CTBs are key targets for infection of the human placenta [18,67], these antibodies could be protective against vertical transmission [19,67,68].
We previously reported that mAbs specific to the pUL128-131 portion of the HCMV pentamer produced by B-cell clones isolated from healthy seropositive donors potently inhibited infection of CTBs and cell-cell spread in anchoring villi [43]. Furthermore, a replication-defective HCMV vaccine elicited neutralizing antibodies as well as CD4 + and CD8 + T cell responses to multiple viral antigens in rabbits and rhesus macaques [69]. The vaccine also induced long-lived memory B cells at frequencies comparable to those seen in seropositive subjects [70]. Together, our studies provide in vitro and ex vivo evidence for the importance of the pentamer for preventing virus spread in developing human placentas.
Immunization with HCMV gB has been demonstrated to reduce horizontal transmission of HCMV [71]. In phase II trials, a monomeric recombinant gB subunit vaccine candidate with M59 adjuvant demonstrated~50% efficacy in seronegative women of reproductive age against HCMV primary infection [71,72] and significantly reduced viremia and shortened antiviral treatment in solid organ transplant recipients [73]. Thus, eliciting antibody responses to multiple epitopes of gB and the pentamer may contribute to potential vaccine-mediated protection against congenital transmission. The natural conformation of gB within the viral envelope is a trimer, and some prefusion gB must be present on the virion surface to mediate fusion of the viral envelope and a cellular membrane during entry. In the present study, we immunized rhesus macaques with a form of recombinant gB that was trimeric but not stabilized in the prefusion conformation [74]. Therefore, this gB immunogen is not likely to have the conformation of gB on virions or the surfaces of HCMV-infected cells.
There is evidence that cellular immune responses that include the early development of HCMV-specific CD4 + and CD8 + T cells can contribute to reducing the transmission of the virus from pregnant women to the fetus after primary infection in the first trimester [11,12]. In human pregnancy, the basal decidua of the uterine wall contains immune cells, including natural killer cells, macrophages, and T cell subsets, which could protect the placenta and fetus [75]. We reported that HCMV replicates in the tissue environ-ment at the uterine-placental interface that serves as a local source of virus spread to the anchoring placenta [19,75]. Specifically, the virus replicates in decidual stromal cells and epithelial cells of endometrial glands. The effector memory CD8 + T cells expand and cluster at sites of infection and adjacent to the epithelium of glands and produce IFN-γ, suggesting a central role in controlling early responses to infection in the decidua of seropositive women. Interestingly, long-lived "memory-like" NK cells have been identified in HCMV-infected individuals, but little is known about their generation and density in tissue [76]. These cells exhibit preferential expansion in response to HCMV-infected cells in an antibodydependent manner, suggesting they could contribute to the control of viral replication in tissue. Intramuscular vaccination followed by electroporation of mice with enhanced DNA plasmid vectors encoding gH/gL or UL128/UL130/UL131A induced robust CD8 + T cell responses [77]. Furthermore, mice immunized with lipid nanoparticles encapsulating mRNA encoding the pentameric complex and gB and a second mRNA vaccine expressing the immunodominant CMV T cell antigen pp65 elicited broad and durable neutralizing antibodies as well as strong pentameric complex-specific T cell responses. These results suggest that the addition of pentamer to a T cell antigen generates a more robust combined humoral and cell-mediated immune response that may confer greater protection [78].
In summary, we have shown that the immunization of animals with the HCMV pentamer and combined immunization with pentamer and gB elicit robust neutralizing activity against infection of a broad array of human cell types, including CTBs and other placental cells that may enable vertical transmission. The recent success of RNA vaccines against the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) spike glycoprotein to prevent COVID-19 disease [79], which elicit neutralizing titers comparable to natural infection and cellular immune responses with specific CD8 + and CD4 + T cell expansion [80,81], has raised interest in using this technology in vaccines against additional targets. Thus, HCMV pentamer and gB antigens, whether produced as subunits or encoded by formulated RNA, hold promise for a new generation of HCMV vaccine candidates.