The Human Cytomegalovirus UL116 Gene Encodes an Envelope Glycoprotein Forming a Complex with gH Independently from gL

Cell lines, plasmids, and viruses.

A bacterial artificial chromosome (BAC) containing the genome of the HCMV TR strain was obtained from Oregon Health Science University. TR, a clinical HCMV strain derived from an ocular vitreous fluid sample from a patient with HIV disease (49), was cloned into a BAC after limited passage in fibroblasts (50). HCMV strain TR and recombinant UL116-Flag-TR virus were propagated in human foreskin fibroblasts (HFF-1; ATCC SCRC-1041) grown in minimal essential medium (Invitrogen) supplemented with 10% fetal calf serum, glutamine (100 mg/liter), and gentamicin (350 mg/liter). Virions were isolated by glycerol-tartrate gradient centrifugation as previously described (51). HEK293T cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, glutamine, and gentamicin. Lipofectamine 2000 (Invitrogen) was used to transfect HEK293T cells. Human codon-optimized UL116, gH (UL75), gL (UL115), and gB (UL55) HCMV genes based on the TR strain sequence were synthesized by GeneArt and cloned into the pcDNA3.1(−)/myc-His C vector (Invitrogen) in frame with C-terminal myc and the six-histidine tag sequences. Single-point mutations were introduced with the QuikChange mutagenesis kit (Stratagene), resulting in the mono-tag versions and the tag-less versions of these genes.

Generation of UL116 polyclonal antiserum.

Mouse antiserum recognizing UL116 was developed in-house. Briefly, His-tagged peptides encompassing the amino acid regions of residues 32 to 265 of UL116 from the TR strain were produced in Escherichia coli and purified through metal ion affinity chromatography. Purified peptides were used to immunize mice.

Construction and generation of UL116-Flag TR BAC.

Insertion of the Flag tag at the C terminus of the UL116 ORF in the TR BAC was achieved using the two-step red-mediated recombination method (52). The primer pair used to amplify the kanamycin insertion cassette was 5′-TTC GGC GCC AAC TGG CTC CTT ACC GTC ACA CTC TCA TCG TGC CGC AGA CTG ATT ACA AGG ATG ACG ACG ATA AG-3′ and 5′-TAT CAC CGG TCC AGG TGA GAA AGA GAA GCC GCA ATC CGG GCG GCG GCA CAT CA CTT ATC GTC GTC ATC CTT GTA ATC AGT CTG CGG CAC GAT GAG CAA CCA AAT TAA ACCA ATT CTG ATT TAG-3′, where the underlined base pairs encode the Flag peptide.

Reconstitution of infectious virus.

To reconstitute the virus, 2 μg of the BAC-HCMV DNA and 1 μg of a pp71 expression plasmid were transfected into MRC-5 cells by electroporation. Culture medium was changed 24 h later, and the cells were split and cultured until the appearance of cytopathic effect. The virus stock was prepared by harvesting cell-free culture supernatant when extensive cytopathic effect was visible.

Purification of HCMV TR virions.

HCMV particles in cell supernatants were separated into virion, dense-body, and noninfectious enveloped particle (NIEP) fractions by positive density/negative viscosity gradient centrifugation as described previously (53). Particle concentrations in the preparations were estimated by counting negatively stained samples by EM in relation to a standard concentration of latex beads. To separate virion envelope proteins from capsid and tegument proteins, 10⁸ particles were mixed 1:1 with envelope stripping buffer (2% Nonidet-P40 in phosphate-buffered saline [PBS]) and incubated for 15 min at 4°C. Particles were pelleted (12,000 × g for 5 min at 4°C), and the soluble envelope fraction was harvested. The insoluble capsid/tegument material was washed twice with envelope-stripping buffer and once in PBS before being solubilized in SDS-PAGE sample buffer.

Flow cytometry.

For the detection of membrane-exposed UL116, HEK293T cells transiently transfected with vectors coding for UL116, gH, gB, and empty vector were trypsin detached 48 h posttransfection, incubated for 30 min at room temperature (RT) with Live/Dead Aqua (Invitrogen), diluted 1:400 in PBS, and incubated with different dilutions of anti-UL116 polyclonal mouse sera for 60 min on ice. After three washes in PBS, the Alexa Fluor 647-conjugated goat anti-mouse secondary antibody was added at a 1:200 dilution and incubated for 30 min on ice. A total of 10⁴ cells were analyzed for each histogram using FACSCanto II (Becton Dickinson, Heidelberg, Germany). Experiments were performed in triplicate for statistical consistency; means and standard deviations were analyzed and plotted using GraphPad Prism software.

Immunogold EM.

Purified virions were air dried to the surface of Formvar-coated EM grids. The grids were treated with mouse anti-Flag antibody (F3165; Sigma) for 4 h at RT, washed 3 times with PBS, and incubated with goat anti-mouse antibody conjugated to 5-nm gold particles for 1 h at RT. After further washing in PBS, the grids were negatively stained with phosphotungstic acid and subjected to EM analysis.

Purification of gH/UL116 and gH/gL-C144A Fab complexes.

The gH/UL116 complex was purified from supernatants of HEK293 cells doubly transfected with plasmids encoding UL116 and gH containing a double Strep-tag and 6His tag at their C termini, respectively (UL116-strep and gH-His). The medium then was applied to StrepTrap HP 1-ml columns (GE Life Sciences), and the complex was eluted with the same buffer containing 2.5 mM desthiobiotin. The gH/gL complex harboring the gL-C144A mutation was purified from supernatants of HEK293 cells doubly transfected with plasmids encoding gH-His and untagged gL-C144A as previously described (54). Purified gH/gL-C144A or gH/UL116 was mixed at 1:1.2 with either 3G16 or MSL-109 Fabs. The ternary complexes were isolated by size exclusion chromatography (SEC).

MALS analysis.

Purified gH/UL116 complex, at two different dilutions, was analyzed by SEC on a Superose-6 10/30 column (GE Healthcare) using 25 mM Tris, pH 8.0, 150 mM NaCl as a running buffer. The SEC system was coupled in-line with the following calibrated detection systems: (i) HP1 1050 Hewlett-Packard UV detector (Norwalk, CT); (ii) MiniDawn Treos multiangle light scattering (MALS) detector (Wyatt Corporation, CA); (iii) quasielastic light scattering detector (Wyatt Corporation, CA); and (iv) Optilab T-reX refractive index (RI) detector (Wyatt Corporation). Briefly, these combined measurements allow both the absolute molar mass of an eluting glycoprotein and the individual contributions made by protein and carbohydrate components to be determined (55–57). The carbohydrate contribution to the molar mass can be assessed because glycans contribute directly to the RI signal but not the UV signal.

Competition between gL and UL116 for gH binding and formation of higher-order complexes.

HEK293T cells were transfected with DNA encoding gH-His and either gL, UL116-strep, or both. The samples involving UL116 were purified by Strep-tag affinity, while the gH/gL complex was purified by Ni affinity as described above. The elution fractions were run on SDS-PAGE with previously purified gH/gL and gH/UL116 as controls. To detect higher-order complex formation, HEK293T cells were transfected with DNA encoding soluble gH extracellular domain in four different combinations: (i) gH/gL/UL128/strep-UL130/UL131, (ii) gH/strep-UL116/UL128/UL130/UL131, (iii) gH/gL/His-gO, and (iv) gH/strep-UL116/gO. The formation of pentamer in supernatant was assessed by Western blotting with anti-UL128 antibody, whereas anti-His was used to reveal secreted gO. The samples involving UL116 were purified by Strep-tag affinity, and associated proteins were revealed by Coomassie staining of the gel.

Negative-staining electron microscopy and single-particle analysis.

Five microliters of purified gH/UL116/3G16, gH/UL116/MSL-109, or gH/gL-C144A/3G16 samples (approximately 10 ng) were placed onto a freshly glow-discharged holey carbon grid covered with a thin layer of continuous carbon. The grid was stained with sequential 75-μl drops of a freshly prepared 2% (wt/vol) uranyl formate solution. Samples were imaged using a Tecnai Spirit T12 transmission electron microscope operating at 120 keV at a nominal magnification of ×49,000 (1.57 Å/pixel at the detector level) using a defocus range of −0.8 to −1.2 μm. Images were recorded under low-dose conditions on a Gatan 4,096- by 4,096-pixel charge-coupled-device (CCD) camera. Particles were manually picked using e2boxer (EMAN2) (58) and extracted using a 224-pixel box size. The two data sets were band-pass filtered with a 200-Å high-pass cutoff and a 20-Å low-pass cutoff. For two-dimensional (2D) classification, we used iterative multivariate statistical analysis (MSA) and multireference alignment (MRA) in Imagic (59). 2D classes from gH/gL and gH/UL116 Fab complexes were aligned and compared by cross-correlation using the SPIDER AP SH command.

UL116 expression kinetics and glycosidase digestion.

Infected cells were harvested at the indicated time points and lysed in radioimmunoprecipitation assay (RIPA) buffer (Sigma) according to the manufacturer's protocol. Late-phase protein expression was inhibited by the use of phosphonoacetic acid (PAA; Sigma-Aldrich). A total of 250 μg/ml of PAA was added to the medium at the time of infection and maintained throughout infection. For deglycosylation treatments, 20 μg of protein extract was incubated with 2.5 μl of endoglycosidase H (Endo H; NEB), 2.5 μl peptide-N-glycosidase F (PNGase F; NEB), or buffer only for 3 h at 37°C according to the manufacturer's protocol. Samples were analyzed by Western immunoblotting.

SDS-PAGE and immunoblotting.

Proteins were separated by SDS-PAGE on 10 to 15% polyacrylamide gels under standard conditions. Proteins were transferred to nitrocellulose membranes, and membranes were blocked with PBS containing 0.1% Tween 20 and 5% powdered milk. Incubations with primary antibodies and sera, diluted in PBS–0.1% Tween 20 (PBST), were performed at RT for 1 h, followed by 3 washes in PBST before incubation with horseradish peroxidase-conjugated secondary antibodies (PerkinElmer) for 1 h. Membranes were washed 5 times in PBST prior to detection with an enhanced chemiluminescence detection system (Pierce).

Immunofluorescence.

Cells were plated on glass coverslips and infected with HCMV. At day 7 postinfection, cells were fixed in 4% paraformaldehyde, permeabilized with 0.5% Triton X-100, preblocked with HCMV seronegative human IgG, and incubated with primary antibody for 1 h at RT. Following washing in PBS, secondary antibodies were incubated for 1 h at 37°C, washed again, and mounted with 4′,6-diamidino-2-phenylindole (DAPI) ProLong safe stain mounting media (Invitrogen). Primary antibodies were rabbit anti-Flag (Sigma), mouse anti-Flag (Sigma), sheep anti-human TGN 46 (Serotec), mouse monoclonal antibody (MAb) anti-pp28 (Abcam), mouse anti-gH (AbD Serotech), and rabbit anti-gL (OHSU). Secondary antibodies were Alexa Fluor 488-, 568-, and 647-conjugated goat anti-mouse and anti-rabbit (Invitrogen). The intracellular locations of antibody-tagged proteins were examined under laser illumination in a Zeiss LMS 700 confocal microscope, and images were captured using ZEN 2009 software. Image analysis was performed with Fiji (NIH) and Imaris 8.1 (Bitplane). Deconvolution was performed using Autoquant X3 (MediaCybernetics) with fixed point spread function (PSF) modeling.

Immunoprecipitations.

HFF-1 cells were infected with HCMV TR and UL116-Flag-TR. Protein expression was allowed to proceed for 72 h, and then cells were washed in 1× PBS and lysed with a lysis buffer (25 mM Tris, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 5% glycerol, pH 7.4) in the presence of protease inhibitors (Complete EDTA-free; Roche). Five hundred micrograms of total protein was incubated overnight at 4°C with 5 μg of the anti-gH MSL-109 monoclonal antibody. Complexes were immunoprecipitated using Protein G Dynabeads (Invitrogen) according to the manufacturer's protocol. The beads were washed four times in lysis buffer and then boiled for 5 min in SDS-PAGE loading buffer. Proteins were separated on SDS-PAGE, and immunoblotting was performed as described above. A similar procedure was adopted for immunoprecipitation from supernatants of transiently transfected HEK293T cells. Complexes were captured in parallel experiments with covalently linked anti-His antibody magnetic beads (GenScript) and anti-c-myc magnetic beads (Pierce) to avoid background signals in the eluted materials.

Primary structure of UL116 gene product.

In all sequenced HCMV genomes, the UL116 gene is located in the unique-long (UL) region between the UL115 (gL) and UL117 genes on the antisense coding strand (Fig. 1A). The UL116 mRNA was previously shown to arise in the true-late stage of AD169 infection as part of the UL119-UL115 transcription unit, but the gene product was never analyzed (60). The UL116 gene is not conserved among alpha and gamma herpesviruses; thus, it does not belong to the core herpesvirus genes (61). Orthologous genes are present in cytomegaloviruses infecting nonhuman primates (62), mouse, rat, and guinea pig (63, 64). The UL116 ORF from the HCMV TR strain is predicted to encode a 313-amino-acid glycoprotein comprising a signal peptide (amino acid positions 1 to 24) (Fig. 1B, SP) and a threonine-rich domain (amino acids 27 to 157), with the resulting polypeptide backbone predicted to have a molecular mass of 34.2 kDa. Fourteen consensus sites for N-linked glycosylation are predicted, whereas the O-glycosylation acceptor potential sites are very numerous (>70). Moreover, UL116 lacks membrane anchor sequences and therefore is expected to be a secreted protein. A multisequence alignment of UL116 from clinical and laboratory-adapted HCMV strains reveals 98% sequence identity (Fig. 1B), suggesting a conserved functional role.

Kinetics of UL116 expression during HCMV replication.

To investigate the expression kinetics of UL116 during productive HCMV infection, we generated a recombinant HCMV, derived from the BAC clone of the TR strain, with a Flag tag fused to the UL116 C-terminal end. The reconstituted TR UL116-Flag virus, which showed growth kinetics identical to those of the parental virus (Fig. 2A), was used to infect HFF-1 cells. Extracts were prepared from infected cells at different time points, ranging from 4 to 120 h postinfection (hpi), and the expression kinetics of UL116 were monitored by immunoblotting using an anti-Flag antibody. UL116 migrated as two species, a faster-migrating protein of 76 kDa and a slower-migrating protein of about 125 kDa (Fig. 2B). The latter appeared to be the mature product of the 76-kDa species, as indicated by the increased accumulation of the 125-kDa species from 72 through 120 hpi compared to that of the 76-kDa band (Fig. 2B). As a control, expression of the immediate-early protein IE1 pp72 (UL123) and the late phosphoprotein pp28 (UL99) were monitored in the same samples (Fig. 2B). The detection of UL116 expression paralleled that of pp28, a late-phase marker (65). Consistent with the observed kinetic pattern, the inhibition of viral DNA synthesis, a prerequisite for late gene expression, with PAA resulted in the disappearance of the UL116 bands (Fig. 2C).

The high number of putative N- and O-linked glycosylation sites predicted on UL116 and the difference between the apparent molecular mass observed (125 kDa) by immunoblotting and the calculated molecular mass (34.2 kDa) of the polypeptide backbone suggested that UL116 undergoes an extensive posttranslational glycosylation process. The trafficking of glycoproteins to the Golgi apparatus is associated with the processing of N-linked oligosaccharides to complex oligosaccharides that are resistant to the action of Endo H (66). Endo H completely cleaves only endoplasmic reticulum (ER)-type N-linked carbohydrates, while it does not trim hybrid and complex high-mannose N-linked carbohydrates that form when the glycoproteins reach the Golgi apparatus. PNGase F is able to digest both ER and Golgi-type high-mannose carbohydrate moieties of glycoproteins (67). To analyze the glycosylation of UL116 during the virus life cycle, HFF-1 cells were infected with TR UL116-Flag for 3 days, followed by glycosidase digestion of the cell extract. As shown in Fig. 2D, digestion with Endo H had several effects: (i) the 125-kDa band was only slightly affected, and its apparent molecular mass was reduced by about 10 kDa; (ii) the 75-kDa band collapsed to ∼38 kDa, a value very close to the predicted mass based on sequence (Fig. 2C, compare lanes 1 and 2). PNGase F digestion caused the disappearance of the faster-migrating 76-kDa band that migrated to approximately 35 kDa, while the apparent mass of the 125-kDa band was reduced at ∼78 kDa (Fig. 2D, compare lanes 1 and 3). These results are consistent with the presence of an immature species entirely modified with ER-type glycan and of a mature form carrying mostly Golgi-type carbohydrates.

Confocal microscopy analysis of HCMV-infected human fibroblasts reveals colocalization of UL116 with the viral assembly complex.

Confocal microscopy was used to investigate the localization of UL116 in HCMV-infected cells. An anti-Flag antibody was employed together with antibodies specific for cellular compartments to trace the subcellular distribution of UL116. Human fibroblasts were infected with TR UL116-Flag and at 72 hpi were fixed and stained for confocal analysis. Among the cellular markers used, UL116 showed colocalization with the trans-Golgi marker TGN 46 (Fig. 3A, upper). Based on the cytoplasmic compartmentalization of UL116, we asked whether the structural HCMV proteins colocalized with UL116. We found a distinct colocalization of UL116 with tegument phosphoprotein pp28 and gH in infected HFF-1 cells 72 h after infection with TR UL116-Flag (Fig. 3A and B, respectively, bottom). Image deconvolution and 3D reconstruction allowed us to better define the colocalization surface between UL116 and the glycoprotein gH. As shown in Fig. 3B, both proteins strongly accumulate in a cytoplasmic perinuclear inclusion and in smaller vesicles scattered in its close proximity. pp28 and gH previously were reported to localize with other tegument or envelope proteins at the virus assembly complex (AC) site during the late phase of the infectious cycle (68, 69) and to be acquired by the nascent virion.

These data are consistent with the trafficking of UL116 to the site of viral AC and suggest that UL116 can associate with other envelope glycoproteins.

UL116 is a component of the HCMV virion envelope.

To verify if UL116 was incorporated into the virion and to establish its localization, we purified the recombinant TR UL116-Flag viral particles from the supernatant of infected HFF-1 cells and performed both Western blotting and immuno-EM analysis. TR UL116-Flag particles purified by positive density/negative viscosity glycerol/tartrate gradient were extracted with detergent to separate envelope and tegument fractions (51), which were probed by immunoblotting with different antibodies (Fig. 4A). pp65 was used as a marker of the tegument fraction, and the nonstructural HCMV protein IE1 was used to exclude samples that were contaminated by cellular extracts. As expected, the envelope proteins gB and gL were detected by specific antibodies in the envelope fraction (Fig. 4A). Notably, probing with the anti-Flag antibody showed that UL116 was present exclusively in the envelope fraction and was not detectable in the pp65-containing tegument fraction.

To confirm the presence of UL116 on the viral envelope, we performed immunoelectron microscopy on the purified TR UL116-Flag particles using a wild-type TR viral preparation as a control. The 10-nm gold-labeled anti-Flag secondary antibody displayed distinct labeling of the envelope (Fig. 4B). Taken together, these results are consistent with the localization of UL116 on the surface of the HCMV envelope.

Cotransfection of recombinant UL116 and gH.

To better characterize the UL116 gene product, we generated a recombinant myc-His tagged version of the gene that was codon optimized for efficient expression and cloned in a eukaryotic expression vector. Surprisingly, transfection in MRC-5 or HEK293T cells did not result in secretion of the recombinant product or its transport to the cell surface (Fig. 5A). Confocal analysis of MRC-5-transfected cells showed that UL116 colocalized almost completely with protein disulfide-isomerase (PDI), a marker of the endoplasmic reticulum (Fig. 5A). We hypothesized that correct localization of UL116 can be achieved following association with one or more of the major HCMV envelope glycoproteins (70). To investigate this possibility, we performed binary cotransfection experiments in HEK293T cells of TR-UL116 with well-characterized HCMV envelope glycoproteins and analyzed the membrane localization of UL116 by cytofluorimetric assay on nonpermeabilized cells with a polyclonal mouse UL116 antisera and a fluorescein isothiocyanate (FITC)-conjugated secondary anti-mouse antibody. To exclude possible nonspecific detection, we included single-gene-transfected cells as negative controls. The results from a representative experiment using gH/UL116 and gB/UL116 are shown in Fig. 5B. A strong UL116 plasma membrane signal was observed only in gH/UL116-coexpressing cells. The cotransfection of UL116 with other HCMV proteins, such as gO and UL10, did not allow the transport of UL116 on the plasma membrane (unpublished results). These data are consistent with UL116 depending on the formation of a heterodimeric complex with gH for translocation to the cell surface.

To test the formation of a gH/UL116 complex, we generated eukaryotic expression vectors carrying singly tagged gH and UL116 versions of the codon-optimized genes. We first performed coimmunoprecipitation experiments on extracts of HEK293T cells coexpressing gH-myc/UL116-His or gB/UL116-His. Extracts from single-gene transfection with gH-myc and gB were used as controls. Coimmunoprecipitations were performed with an anti-His antibody to detect proteins associated with UL116 and by anti-myc to reveal species associated with gH. Each immunoprecipitated sample was divided into three aliquots and treated for Western blotting using anti-His (for UL116), anti-myc (for gH), and anti-gB as probes. Pulldown of UL116 resulted in association with gH only and not with gB (Fig. 5C, bottom, lanes 5 and 10, respectively). Conversely, immunoprecipitation of gH resulted in the coimmunoprecipitation of UL116 (Fig. 5C, middle, lane 1). To confirm the formation of the gH-UL116 complex and to exclude potential artifacts induced by the overexpression system, we decided to test the interaction between the two proteins in HCMV-infected cells. To this aim, we performed coimmunoprecipitation experiments using the anti-gH human monoclonal antibody MSL-109. Extracts from TR UL116-Flag-infected HFF-1 cells at 96 hpi were prepared and submitted to anti-gH immunoprecipitation. Cell lysate from wild-type-TR-infected HFF-1 cells was used as a negative control. The gH-associated proteins were separated by SDS-PAGE, and as a positive control, gL was probed by Western blotting. Anti-Flag antibody was used as the probe to reveal the presence of UL116. As expected, gL was detected in extracts from cells infected with both viruses (Fig. 5D). Notably, a clear signal for UL116-Flag was detected in extracts from cells infected with TR UL116-Flag but not when wild-type TR virus was used for infection (Fig. 5D).

Finally, HEK293T cells were transfected with plasmids encoding a soluble form of gH and UL116 containing a Strep-tag at the C terminus. HEK293T cell supernatants were loaded on a Strep-Tactin resin, and the eluted material was analyzed using SEC, multiangle light scattering, and SDS-PAGE. SEC and SDS-PAGE revealed that gH and UL116 form a stable complex that migrates as a single peak. MALS analysis suggested that the mass of the gH/UL116 complex is 161 kDa (i.e., glycans account for ∼37% of the total mass of the complex), consistent with a 1-to-1 stoichiometry (Fig. 6A). The heterodimer is not disulfide linked and dissociates in single subunits in SDS-PAGE under nonreducing conditions (Fig. 6B).

These results demonstrate the formation of a gH/UL116 complex in transfected cells that does not require the presence of gL to form. Moreover, we also were able to confirm the formation of the complex during a productive HCMV infection.

gL competes with UL116 for gH binding in transfected cells.

To assess if gH, gL, and UL116 can form a ternary complex, we transfected HEK293T cells with a construct corresponding to the soluble gH extracellular domain together with either gL or UL116-strep or both gL and UL116-strep. Western blot analysis of the cell supernatants with an anti-gL antibody showed the formation of monomers and homodimers of the gH/gL heterodimers in cells transfected with either gH and gL or gH, gL, and UL116 (Fig. 7A). The formation of disulfide-linked monomers and homodimers of soluble gH/gL is consistent with what has been described previously (35, 54, 71). The Strep-tag at the C terminus of UL116 allowed us to isolate gH/UL116-containing complexes from cell supernatants. SDS-PAGE analysis of the eluted fraction revealed that when gH, gL, and UL116 are transfected together, only a very minor amount of gH/UL116 complex forms and no ternary complex could be detected (Fig. 7B). Together, these data suggest that gL competes with UL116 for gH binding. It has to be noted that gH and gL form disulfide-linked heterodimers, and perhaps once this complex forms UL116 is not able to displace gL.

The gL subunit is not only disulfide linked to gH but also is engaged in disulfide bond formation with gO, in the gH/gL/gO complex, or with UL128 in the pentamer (54). Although the absence of gL from the gH/UL116 dimer suggested that neither gO nor UL128 can associate with this heterodimer, we sought to verify this hypothesis. To assess if UL116 could replace gL as a scaffold for the formation of these complexes, we transfected HEK293 cells with constructs encoding the soluble extracellular gH domain with the following combinations: (i) gH/gL/UL128/strep-UL130/UL131; (ii) gH/strep-UL116/UL128/UL130/UL131; (iii) gH/gL/His-gO; and (iv) gH/strep-UL116/His-gO. Secreted proteins then were detected by Western blotting, and Strep-tag pulldown was used to reveal subunits associated with UL116. As can be seen in Fig. 7B and C, UL128 is secreted only when gL is present and not when UL116 is expressed. Similarly, gO does not form a complex when UL116 replaces gL (Fig. 7D). Thus, the association of UL116 with gH competes with the formation of both pentamer and gO subunits.

Negative-stain EM analysis of the purified UL116/gH complex.

We previously used single-particle negative-stain EM to characterize gH/gL, gH/gL/gO, and Pentamer complexes bound to neutralizing antibodies (54, 72). Our data revealed that HCMV gH/gL has a boot-shaped structure similar to that of HSV-2 gH/gL (73) and that gO and the UL128/UL130/UL131A subunits bind to the N-terminal end of gH/gL in gH/gL/gO and Pentamer, respectively. We also engineered a gL mutant, C144A, which prevents gH/gL homodimerization.

Here, we extended the EM analysis to the purified gH/UL116 complex bound to either MSL-109 or 3G16 antibody (41, 72). The trimeric gH/UL116/3G16 and gH/UL116/MSL-109 complexes were analyzed by negative-stain EM and single-particle analysis. 2D class averages were generated and compared to the corresponding 2D class averages for the gH/gL-C144A/3G16 complex (Fig. 8). This analysis revealed that the gH/UL116 complex has a boot-shaped structure similar to that of gH/gL. 3G16 and MSL-109 Fabs bind to the C-terminal domain and to the “heel” of the gH molecule, respectively, as previously described for gH/gL (54, 72). Furthermore, the EM analysis and binding data (unpublished results) suggest that UL116 occupies a position similar to that of gL at the N-terminal end of gH (Fig. 8) and that the presence of UL116 does not affect the 3G16 and MSL-109 binding sites.