Isolation and characterization of a noninfectious virion-like particle released from cells infected with human strains of cytomegalovirus

Isolation and characterization of a noninfectious virion-like particle released from cells infected with human strains of cytomegalovirus

VIROLOGY 130,118-133 (1983) Isolation and Characterization of a Noninfectious Virion-like Particle Released from Cells Infected with Human Strains ...

6MB Sizes 0 Downloads 5 Views

VIROLOGY

130,118-133

(1983)

Isolation and Characterization of a Noninfectious Virion-like Particle Released from Cells Infected with Human Strains of Cytomegalovirus* ALICE L&artment

IRMIERE

AND

WADE

GIBSON’

of Pharmmobgg and Experimental Therapeutics, The Johns Hopkins University School of Medicine, 725 North Wolfe Street, Baltimore, Maryland 21205 Received February

1, 1983; accepted June 7, 1983

Three types of virus particles have been recovered from the culture medium of human foreskin fibroblasts infected with human strains of cytomegalovirus (HCMV). Two of these, virions and dense bodies, are routinely observed and have been described by others. The third, produced in lesser amounts, has not been previously characterized. This particle, separable from virions by rate-velocity sedimentation, is morphologically distinguished from them only by core structure. Radiolabeling and biological assays have established that these particles, like dense bodies, lack DNA and are not infectious. Based on these properties, we have designated this virion-like structure as a noninfectious enveloped particle (NIEP). Comparisons of the protein constituents of these three particles has shown that dense bodies have the simplest composition. Approximately 95% of their protein mass is represented by a 69,000 Da (69K) matrix-like protein. While dense bodies appear to have a normal complement of virion glycoproteins, they completely lack other predominant virion species. The protein compositions of virions and NIEPs are more complex than that of dense bodies, and are distinguished from one another by the presence in NIEPs of a 35,000 Da (35K) protein absent from the two other particles. Biosynthetic radiolabeling and cell fractionation experiments have demonstrated that this 35K protein is produced only in infected cells, is phosphorylated and partitions with the nuclear fraction. These and other results suggest that this protein is the HCMV counterpart of the previously described B-capsid proteins VP22a of herpes simplex and 3’7K of CMV (strain Colburn). NIEPs are produced by all HCMV strains examined and have not been observed in preparations of herpes simplex virus- or Old World monkey CMV-infected cells. Although this particle is generally present in much lower amounts than virions, strain AD169 overproduces NIEPs by approximately lo-fold. We have also found that the additional NIEP protein of AD169 has an apparently larger size (i.e., 36K) than the corresponding protein of other strains. The correlation between AD169 NIEP overproduction and its altered protein suggests that the two may be causally related.

micrographs of thin-sectioned virus particles reveal an architectural organization Human strains of cytomegalovirus consisting of an inner DNA-containing core (HCMV) are distinguished by having the surrounded by an icosahedral capsid, an largest genome (approximate A!& 1.5 X 10’) apposed tegument layer, and an outermost of any known virus that replicates in the bilaminar membrane envelope (Kanich and cell nucleus (Kilpatrick and Huang, 197’7; Craighead, 1972; Watson, 1973; Iwasaki et DeMarchi et al, 19’78; Geelen et al, 1978). ah, 1973; Roizman and Furlong, 1974). Like other members of the herpesvirus Other types of virus particles have been family, cytomegalovirus (CMV) produces observed in thin-sectioned preparations of a complex, multilayered virion. Electron HCMV-infected cells. These include several types of intranuclear capsids having typ*This article is dedicated to the memory of Dr. ical herpesvirus morphology; cytoplasmic Wallace P. Rowe. dense bodies, enveloped and nonenveloped i Author to whom requests for reprints should be addressed. nucleocapsids; and perhaps more related INTRODUCTION

0042-6822/83 $3.00 Copyright All rights

0 1983 by Academic Press, Inc. of reproduction in any form reserved.

118

HCMV

VIRION-LIKE

to this report, enveloped cytoplasmic virion-like particles having an altered core structure as described by Ruebner et al. (1965) and Kanich and Craighead (1972) (Luse and Smith, 1958; Ruebner et al, 1965; McGavran and Smith, 1965; Patrizi et aL, 1965; Becker et ah, 1965). We are interested in the structure and biochemical composition of CMV virus particles as a means of gaining insight into the assembly pathway by which they are produced. Such information has provided valuable leads in assessing the function of specific proteins during the assembly and envelopment of strain Colburn CMV (simian-like isolate) (Gibson, 1981). In an attempt to extend this approach to human strains of CMV, we initially sought to examine the protein composition of purified virions. During the early stages of the study, however, we noticed an unexpected band above the virion position in our preparative gradients. This paper presents the results of experiments to compare the biological, physical, and biochemical properties of the particles in this previously undescribed band, with those of virions and dense bodies. These comparisons show that this particle (i) is architecturally similar to virions, but distinguished by its core structure; (ii) contains no DNA and is consequently noninfectious; and (iii) has all of the proteins present in virions, and one additional, abundant species. In view of these characteristics, we have designed this virion-like structure as a noninfectious enveloped particle (NIEP). In these studies we have utilized the positive density-negative viscosity gradient system of Barzilai et aL, (1972), as adapted for HCMV purification by Talbot and Almeida (1977). Because this procedure gives good separation of virions and dense bodies, we also discuss our results concerning the protein composition, DNA content, and infectivity of these two virus particles, as they relate to earlier descriptions (Sarov and Abady, 1975; Fiala et al., 1976; Stinski, 1976; Kim et al, 1976; Gupta et al, 1977; Siqueira-Linhares et aZ.,1981).2 ’ A preliminary report of this work was presented at the Sixth Herpesvirus Workshop held at Cold Spring Harbor Laboratory, August 31 to September 5, 1982.

119

PARTICLE MATERIALS

AND

METHODS

Cells and wirua Procedures for human foreskin fibroblast (HFF) culture, infection with CMV, titration of infectivity, and cell fractionation have been previously described (Gibson 1981). The sources for human cytomegalovirus strains Davis, Towne, AD169, 751, and Charcot-MarieTooth (CMT); the simian strain CSG; and strain Colburn, are described elsewhere (Weiner and Gibson, 1981; Gibson, 1983). Particle pur$cation. Particles were recovered from the infected-cell culture medium as follows. Infected cells were scraped from the culture vessel (e.g., 6-cm plastic petri dish; 32-0~ glass bottle) into the medium and collected by low-speed centrifugation (i.e., 1500 g, 4’, 10 min). The time of cell harvest varied with the experiment, but was generally 5 to 8 days after infection. The resulting cell-free medium was layered onto potassium tartrate-glycerol gradients prepared in 0.05 MTris-HCl, pH 7.4, 0.10 M NaCl (TN buffer) (Barzilai et al, 1972; Talbot and Almeida, 1977), and subjected to centrifugation (40,000 rpm, 4”, 15 min) using a Beckman SW41 rotor and Sorvall OTD-50 ultracentrifuge set on slow acceleration and braking modes. Particle bands were located in the gradient by their light-scattering properties and collected by aspiration, using a 23-gauge needle, through the wall of the centrifuge tube. Further purification, generally unnecessary for NIEPs and virions, could be achieved by an additional rate-velocity sedimentation step (as above), and/or banding the particles to equilibrium (i.e., 18 hr) using the same gradient and centrifugation conditions. As shown in Fig. 1, it proved convenient to subject dense bodies to equilibrium centrifugation since the resulting band (gradient “B”) was considerably narrower than the broad zone observed after rate-velocity sedimentation (gradient “A”). It is our experience that virions and dense bodies both band low in the gradient following equilibrium centrifugation, and are not well separated by this step. We have had reasons to explore modifications of the potassium tartrate-glycerol

120

IRMIERE

AND

GIBSON

amino acids (5 &i/ml, New England Nuclear, NEC 445) in medium containing l/10 the normal amounts of the labeled amino acids, and 2% fetal calf serum; (iii) [32P]orthophosphate (200 &i/ml, Amersham, PBS.llA); or (iv) r5S]methionine (5 &i/ml, Amersham, SJ.204) in DMEM lacking methionine (GIBCO No. 80-0184) and supplemented with 2% fetal calf serum.

Gradient fractionation and scintillation counting. Radiolabeled CMV particles were

FIG. 1. Separation of extracellular HCMV particles by centrifugation. (A) Clarified medium from AD169infected HFF cells was layered above a glycerol-tartrate gradient and centrifuged (40,000 rpm, 4’, 15 min) in a Beckman SW41 rotor. Illumination from the top of the gradient revealed two sharp light-scattering bands designated as noninfectious enveloped particles (NIEPs) and virions (Vir.) and a broad area containing dense bodies (DB). (B) Broad dense body zone was collected from the first gradient, diluted with TN buffer, and banded to equilibrium in a similar gradient (40,000 rpm, 4”, 18 hr). Illumination from the top revealed a narrower band, varying in tinge from reddish at the top to brownish at the bottom.

gradient system, and have obtained essentially the same separations using: (i) sodium tartrate (slightly lower solubility) in place of the potassium salt, to avoid potassium-SDS precipitation when samples are to be subjected directly to SDS-polyacrylamide gel analysis; (ii) sodium phosphate buffer (0.04 h4, pH 7.4) in place of Tris-HCl buffer to eliminate the presence of extraneous polyamines, characteristic of Tris; (iii) 18-ml gradients formed in 38-ml tubes, to accomodate larger volumes, and subjected to centrifugation (25,000 rpm, 4”, 40 min; Sorvall AH627 or Beckman SW27 rotor); and (iv) gradients formed from the bottom in nonwettable tubes in place of gradients formed from the top in wettable nitrocellulose tubes (no longer manufactured). Radiolabeling. CMV-infected fibroblasts growing in 6-cm-diameter petri dishes were radiolabeled by removing the culture medium 48 hr after infection, and replacing it with 3 to 4 ml of one of the following: (i) C3H]thymidine (10 &i/ml, Amersham, TRK.418); (ii) a mixture of 12 14C-labeled

separated in potassium tartrate-glycerol gradients as described above. An ISCO density gradient fractionator (No. 185) equipped with an absorptimeter was used to fractionate (0.18-ml aliquots) the gradients and monitor absorbance at 280 nm. Radioactivity in the resulting fractions was measured by trichloracetic acid (TCA) precipitation followed by scintillation spectrometry, as follows. A lo- or 50-~1 aliquot (rH]thymidineor 14C-amino acidlabeled samples, respectively) of each was placed on moistened glass microfiber filters (2.4-cm circle, Whatman GFC) and dried at 80”. The filters were then submerged in a 10% w/v TCA solution at 0”, rinsed sequentially under vacuum with 5% w/v TCA, then 70% v/v ethanol at 0”, and dried at 80”. The filters were placed in glass vials containing 3 ml of Spectrafluor (Amersham/Searle) scintillation cocktail prepared in toluene, and radioactivity was measured (10 min/sample) using a Searle Analytic ISOCAP/300 scintillation spectrometer. Electron microscopy. Particles were concentrated for electron microscopy by pelleting (35,000 rpm, 4”, 3 hr; Sorvall AH650 rotor), and then resuspended in 10 ~1 of TN buffer. Five microliters of the resulting suspension was placed on a 400-mesh parlodion-treated copper grid, coated with a thin layer of carbon. The grid was then washed for 30 set using TN buffer, and stained for 30 set with 1% uranyl formate, pH 4.5 (Bijlenga et al., 1976). Excess stain was removed by touching a piece of filter paper to the grid edge. An alternate procedure of concentrating the particles by vacuum dialysis against 0.05 1M Tris-HCl, pH 7.0, worked well and gave similar results for NIEP and virion preparations, but was found unsatisfactory for dense bodies

HCMV

VIRION-LIKE

due to the formation of electron dense crystals on the grid. Micrographs of negatively stained particles were taken with a Zeiss EM 10A electron microscope using an accelerating voltage of 80. The microscope magnification was calibrated with a 14,600 lines/cm crossed-carbon grating. Thin sections were prepared from particles recovered from glycerol-tartrate gradients, that were fixed, pelleted (35,000 rpm, 4”, 3 hr; Beckman SW50.1 rotor), osmicated, and dehydrated, using a Beckman Airfuge (100,000 Q, 27”, 5 min) to recover them following each resuspension. The final pellet was embedded, thin sectioned, stained with lead citrate and uranyl acetate, and examined using an electron microscope, as described elsewhere (Wolinsky et aL, 1974). Dimensions of ultrastructural features of these particles were calculated from measurements made on enlarged electronmicrographs. Perpendicular measurements were taken from at least eight examples of each particle type. Electrophwesis. In preparation for electrophoresis, particles recovered from gradients were concentrated to dryness by vacuum dialysis against TN buffer in a 75,000 M, exclusion limit colloidon bag (No. UH100/2A, Schleicher and Schuell, Keene, N. H.). Concentrated particles were recovered by rinsing the bag twice with 50 ~1 of solubilizing solution (i.e., 2% sodium dodecyl sulfate (SDS), 10% beta-mercaptoethanol, 10% glycerol, 50 mM Tris-HCl, pH 7.0, and 0.005% bromophenol blue), heated to 100” for 3 min and stored at -80’. SDSpolyacrylamide gel electrophoresis was done essentially according to Laemmli (1970), however, the ratio of methylenebisacrylamide:acrylamide was increased to 1.09:2&O (“high bis” gels) in order to separate the major capsid protein from the basic phosphoprotein. Stacking gels were 5% acrylamide with a crosslinker ratio of 0.735:2&O (methylenebisacrylamide:acrylamide). All resolving gels shown were lo%, “high bis” acrylamide, except that in the right-hand panel of Fig. 10, which was 10% and contained the normal amount of crosslinker. Since the AD169 36K intracellular protein is better resolved from host cell proteins with Pierce (No. 28365) SDS, we

121

PARTICLE

have used it throughout this study. After electrophoresis, gels containing [35S]methionine were processed for fluorography (Bonner and Laskey, 1974); dried onto Whatman 3MM paper at 65” instead of 100” to avoid shattering the more highly crosslinked gel; and exposed to Kodak XAR film at -70” (Laskey and Mills, 1975). Gels containing 3zP-labeled material were also dried in this manner, but fluorographic enhancement was achieved using a calcium tungstate intensifying screen (Cronex Lighting Plus, DuPont) at -70” (Laskey and Mills, 1977). Gels containing unlabeled proteins were stained with Coomassie Brilliant Blue stain (CBB) (Fairbanks et aL, 1971) and scanned at 540 nm using an EC 910 transmission densitometer (E-C Apparatus Corp., St. Petersburg, Fla.) prior to drying. Further details of electrophoresis conditions have been described elsewhere (Gibson, 1981). Protein molecular weights were determined by the method of Weber and Osborn (1969) using the following mixture of standards (gift of Dan Kiehart): myosin 200K, beta-galactosidase 116K, phosphorylase b 94K; bovine serum albumin 68K; catalase 60K; tubulin doublet 55K; actin 43K; aldolase 40K; carbonic anhydrase 29K. RESULTS

The initial experiments described in this report were done using the AD169 strain of human cytomegalovirus (HCMV). Other human and simian (SCMV) strains were used for comparison in the last two experiments. Separatim of extracellular CMV particles. The positive density/negative viscosity gradient system originally described by Barzilai et al, (1972) was used, essentially as reported by Talbot and Almeida (1977), to recover particles from the culture medium of human foreskin fibroblast (HFF) cells infected with strain AD169 HCMV. Three light-scattering bands were visible following rate-velocity centrifugation (Fig. 1A); the upper two sedimented as tight zones, while the lower was diffuse. Subsequent equilibrium centrifugation of the lower band material resulted in a markedly narrower zone (Fig. 1B). Incandescent il-

122

IRMIERE

AND

lumination of gradients containing the rebanded material showed the two upper bands to be white/opalescent in color, while the lower band varied in tinge from reddish at the top to brownish at the bottom, as noted by Talbot and Almeida (1977). Based on these and other characteristics described below (i.e., morphology and infectivity) it has been established that the two lower bands are comprised of virions and dense bodies, respectively. The uppermost band has not been reported previously, and is shown here to contain an unusual virionlike structure designated as noninfectious enveloped particle (NIEP). Before presenting further results, it is worthwhile reviewing our evidence that NIEPs are not artifacts peculiar to our culture system or isolation procedure. First, during the course of this work we have varied the following culture conditions without noticeable effect on NIEP production; capped 32-0~ glass bottles in place of 6-cm plastic petri dishes; phosphate-buffered medium (Leibovitz, GIBCO No. 320-1415), and media depleted in glucose (GIBCO No. 430-1600), methionine (GIBCO No. 80-0134) or other amino acids, in place of high-glucose DMEM (GIBCO No. 430-2100); newborn calf serum instead of fetal calf serum; and a 33” incubation temperature instead of 37”. Second, we have obtained essentially the same results using: WI-38 (Flow Laboratories) cells; high or low multiplicities of infection; frozen, continuous passage or freshly plaqueisolated virus stocks; and the vaccine strain of AD169, kindly provided through Merck, Sharp, and Dohme by Dr. H. Stern. And third, we have altered the particle isolation procedure in the following ways with essentially no effect on the amount of NIEPs relative to virions: substituted different salts and buffers in the glycerol-tartrate gradients as described under Materials and Methods; used 1550% (w/v) sucrose gradients prepared in either Tris or phosphate buffer; harvested the cells earlier (e.g., 5 days after infection) and much later (e.g., up to 1 month after infection) than usual (e.g., 7-10 days after infection); and used nonclarified, as well as high-speed (8500 g, 10 min, 4”) or low-speed (1500 g, 10 min, 4”) clarified media preparations.

GIBSON

Particle morpho&y. Comparisons of NIEPs, virions, and dense bodies by electron microscopy showed that all of the particles have a similar surface appearance but differ in their internal structure. Negative staining did not distinguish between NIEPs and virions taken directly from rate-velocity gradients, since virtually all of the particles in each preparation had the same diameter (i.e., 230 nm) and excluded stain (i.e., had an intact envelope). However, when pelleted before negative staining, approximately 40% of the NIEPs and 60% of the virions became penetrable to the stain. Such particles exhibited a noticeable difference in their internal capsid structure (compare Figs. 2A and B). Virions had a densely stained center, sharply delineated by the capsid, while NIEPs showed a stain-excluding core structure with a central electron-dense area. The capsid diameter of NIEPs, as determined from these negatively stained preparations, was slightly smaller than that of virions (98 vs 116 nm), and the NIEP core structure was estimated to be 55 nm in diameter. Particles containing two capsids, but surrounded by one envelope (Figs. 3A and B), were occasionally seen in preparations of negatively stained virions. Intact capsids devoid of envelope were never observed. Partially disrupted capsids, however, were often seen and allowed a more detailed observation of the CMV capsid ultrastructure (Figs. 3C-H). Capsid remnants showing sixfold symmetry (Figs. 3C, F, E, H) and infrequently fivefold symmetry (Figs. 3D, G) were also observed. Based on measurements made from particles in Figs. 2B and 3A, C, D, E, the HCMV capsomere: (i) is approximately 20 nm in length and 15 nm in diameter, (ii) has a channel about 3 nm in diameter that is open at least at the exterior end, (iii) possesses favored cleavage planes along the axis of the cylinder that give rise to broken semicircular appearing structures (Fig. 3E, see arrows), and (iv) has short spicule-like protrusions extending out symmetrically and resulting in a pinwheel appearance of the capsomere viewed end on (Figs. 3C, E). Dense bodies that had not been pelleted also excluded stain, but were distinguish-

HCMV

VIRION-LIKE

PARTICLE

123

FIG. 2. Negatively stained extracellular HCMV particles recovered from a glycerol-tartrate gradient. NIEPs (A), virions (B), and dense bodies (C) were collected from a gradient, pelleted, resuspended in TN buffer, adsorbed to grids, and stained with 1% many1 formate, as described under Materials and Methods. Particles in lower left-hand corners of A and B are nonpenetrated. Arrows indicate limiting membrane structure; bars represent 100 nm.

able from NIEPs and virions by their generally larger size, which ranged from about 250 to 600 nm. Pelleted dense bodies exhibited some positive staining with uranyl formate (Fig. 2C). No internal structure was ever observed in negatively stained dense bodies. To confirm the internal difference between NIEPs and virions indicated by neg-

ative staining, and to determine whether dense bodies are hollow or solid spheres, preparations of the three particle types were examined by electron microscopy following thin-sectioning and staining. Figure 4 shows that only NIEPs and virions contain capsids, while dense bodies were found to be solid spheres homogenously filled with a granular appearing material. As

124

IRMIERE

AND

GIBSON

FIG. 3. Ultrastructural details of HCMV particles. Occasionally, enveloped particles containing two nucleocapsids were observed in glycerol-tartrate virion preparations (A and B). Disrupted particles in such virion preparations showed capsids with sixfold (C and E) and fivefold (D) symmetry. (F), (G) and (H) show the capsid symmetry of (C), (D), and (E), respectively, more graphically. Individual capsomeres exhibited specific cleavage patterns (see arrows, E and H). Bar represents 100 nm in (A and B), 10 nm in (C-H).

observed in negatively stained preparations (Fig. 2), measurements made from these thin-sectioned preparations suggest that the diameter of the NIEP capsid is slightly smaller than that of the virion (i.e., 10-S%). The outer portion of both NIEP and virion capsids consisted of two concentric zones: an exterior stained region approximately 20 nm thick and an apposed interior nonstained region approximately 10 nm thick. The major difference in the appearance of thin-sectioned NIEPs and virions was in their core structure. That of NIEPs was lightly stained in the middle, and surrounded by a densely stained ring approximately 5-8 nm thick. The virion core, in contrast, was uniformly and densely stained. Occasionally, however, nonstained regions within this area were seen (e.g., 5% of particles).

NIEPs, virions, and dense bodies all were bounded by an envelope structure (see arrows, Figs. 2A, B, C) that was estimated to be 10 nm in thickness. Although well preserved in the negatively stained preparations, only vestiges of this structure (see arrows, Figs. 4A, C) survived the comparatively harsher thin-sectioning protocol.

Infectivity

and DNA content of CMV

particles. To determine the relative infectivity of these three types of virus particles, 3 ml of medium from an infected culture was subjected to rate-velocity centrifugation in a glycerol-tartrate gradient. The resulting gradient was then monitored at 280 nm as it was collected, and each fraction was assayed for infectivity by endpoint dilution. Figure 5 shows that 99% of the total infectivity in the gradient was present in the virion band. Neither NIEPs nor

HCMV

VIRION-LIKE

PARTICLE

125

FIG. 4. Thin sections of pelleted HCMV particles. NIEPS, virions, and dense bodies were recovered from gradients as shown in Fig. 1, then fixed, pelleted, thin sectioned, and stained, as described under Materials and Methods. Arrows in (A) and (C) indicate remnants of the particle envelope. Considerably more variability in the NIEP core structure exists than is suggested by (A). Bar represents 100 nm in all micrographs.

dense bodies exhibited significant infectivity. Next, the possibility was tested that NIEPs and dense bodies lack infectivity because they have no DNA, or contain a significantly altered DNA. This was done by radiolabeling a CMV-infected cell culture, using rH]thymidine; resolving the extracellular particles by centrifugation,

as above; and measuring the amount of 3Hradioactivity present in each fraction by scintillation spectrometry. A second culture, radiolabeled using 14C-amino acids, was processed in parallel for comparison. Figure 6A shows that only the virion band contained significant amounts of TCAprecipitable rH]thymidine. Neither NIEPs nor dense bodies had more than back-

126

IRMIERE

AND

GIBSON 0.2

DUlSe BOI

1 I z 4”

r,ions -0.1

:

20

Fraction Number FIG. 5. Infectivity of HCMV particles. Clarified infected cell medium was subjected to sedimentation in glycerol-tartrate gradients and subsequently collected in 0.24-ml aliquots. A lo-p1 aliquot of each fraction was used to determine infectivity by endpoint dilution. Shown here are results of the infectivity determinations plotted together with the optical absorbance (226 nm) pattern recorded as the gradient was collected. Percent infectivity was calculated by dividing the amount of infectious virus in each fraction by the total amount of infectious virus in the gradient, and multiplying X160.

ground levels of radioactivity (i.e., 0.5% amount in virion band). As expected, both NIEPs and virions were labelled with 14Camino acids, and their radioactivity profile reflected the optical absorbance pattern obtained in Fig. 6A. The relatively lower amount of dense-body radioactivity is due to a reduced amount of dense bodies in this particular gradient, as judged by light scattering. Proteins of AD169, NIEP.s, viritms, and dense booSea Since NIEPs and virions differ with regard to sedimentation properties, core structure and DNA content, and yet appear to have a very similar ultrastructural organization (Fig. 4), two experiments were done to compare their protein compositions. In the first, particles were biosynthetically radiolabeled with [%Imethionine or 32Pi, and their constituent proteins separated by electrophoresis in a polyacrylamide gel. Figure ‘7 shows an autoradiogram prepared from the resulting gel and demonstrates that NIEPs contain a 36,000 Da (36K) protein that is not pres-

ent in virions (compare channels “c” and “d”). The protein composition of these two types of virus particles was otherwise indistinguishable. The protein composition of dense bodies is comparatively simple. Densitometric measurements of such preparations indicate that at least 90% of their protein mass is represented by a protein that comigrates with the 69K matrix protein of virions and NIEPs. Another protein present in small amounts in NIEPs and virions (i.e., 80K) is also found in similar amounts (relative to the 69K matrix protein) in dense bodies (see Fig. 8). This autoradiogram also shows that NIEPs and virions each contain three predominant phosphorylated proteins (channels “a” and “b”), designated as the basic phosphoprotein (BPP, 149K) and two matrix proteins (MP, 74K and 69K). The NIEP phosphoprotein pattern contains an additional band (i.e., NIEP-specific 36K protein), and lacks the radioactive high-molecular-weight material (e.g., viral DNA) seen at the top of the adjacent virion chan-

HCMV

Otnrc Bodies

I

10

m 40 FractionLnber

VIRION-LIKE

A

127

PARTICLE

0.20

1

50

FIG. 6. DNA content of HCMV particles. Infected cells were radiolabeled with PH]thymidine or a mixture of “C-amino acids and the extracellular particles separated using glycerol-tartrate gradients. The gradients were scanned at 280 nm, fractionated, TCA precipitated, and analyzed by scintillation spectrometry-as described under Materials and Methods. Shown here is the optical absorbance profile of the gradient containing [3H]thymidine-labeled particles, together with the measured radioactivity in corresponding fractions (A). The pattern of i4C-amino acid radioactivity in the gradient analyzed in parallel is shown in (B).

nel. Both the 69K and 80K dense-body proteins are phosphorylated (unpublished observations; Roby and Gibson, manuscript in preparation). It should be noted that the basic phosphoprotein (BPP) comigrates with the major capsid protein (MCP, 153K) in our normal SDS-polyacrylamide gel system. However, when the ratio of methylenebisacrylamide to acrylamide is changed as described under Materials and Methods, the resolution of these two proteins shown in Fig. 7 was obtained. Further, when subjected to electrophoretic separation in SDS-containing, 10 or 14% polyacrylamide gels crosslinked with diallyltartardiamide

abed

e

FIG. 7. Fluorograms showing radiolabeled HCMV proteins. CMV-infected cells were radiolabeled with either [S2P]orthophosphate or [%I]methionine. Extracellular particles, recovered using glycerol-tartrate gradients, were solubilized, and the proteins separated by electrophoresis is a 10% “high bis” polyacrylamide gel. Shown here are fluorographic images of the resulting gel. Different exposure times were required for channels “a, b” and “e”. Shown here are NIEP and virion proteins labeled with m]orthophosphate (a, b) or pS]methionine (c, d), and dense body proteins labeled with [%I]methionine (e). Protein designations are as follows: HMWP (212K, high-molecular-weight protein), MCP (153K, major capsid protein), BPP (149K, basic phosphoprotein), MP (74K and 69K, matrix proteins), DGP, and DGP, (62K and 54K, discrete glycoproteins), 36K (36,000 Da NIEP-specific “assembly” protein), and mCP (34K, minor capsid protein) (see Gibson, 1988). Arrow at top of right-hand margin indicates top of resolving gel. Intense exposure in stacking gel and at top of resolving gel in the 82plabeled virion preparation is due to viral DNA, which is not present in NIEPs (see Fig. 6A).

128

IRMIERE

AND

GIBSON

tern seen in Fig. 7 was an accurate representation of the number of proteins present, and (ii) to obtain quantitative estimates of the proteins’ molar abundancies. For this purpose, NIEPs, virions, and dense bodies were recovered from the growth medium, concentrated, solubilized, and subjected to electrophoresis in a polyacrylamide gel-all as described above and under Materials and Methods. Following electrophoresis, the gel was stained with CBB, destained, and analyzed by densitometry. A photograph of the stained gel after it had been dried onto filter paper is shown in Fig. 8. As can be seen, the resulting pattern of CBB-stained proteins is qualitatively the same as that obtained by fluorography of radiolabeled proteins (Fig. 7). Estimates of the relative molar abundancies of NIEP and virion proteins, based on measurements from the stained gel shown in Fig. 8, are summarized in Table 1. These measurements indicate that the 36K protein is the most abundant species in NIEPs and show that there is a reduced amount of both matrix proteins in NIEPs (e.g., approximately lo-20% less 74K protein and 30-60% less 69K protein) compared with virions. TABLE FIG. 8. Proteins of HCMV NIEPs, virions, and dense bodies stained wtih CBB. NIEPs, virions, and dense bodies were recovered from infected cell medium by sedimentation in a glycerol-tartrate gradient. Concentrated particles were solubilized and subjected to electrophoresis in a lo%, “high his” polyacrylamide gel. Shown here is a photograph of the resulting gel following staining with Coomassie Brilliant Blue (CBB) stain, Protein designations shown in the lefthand margin are as described in the legend to Fig. 7. Numbers in the right-hand margin indicate the respective positions of myosin (l), beta-galactosidase (2), phosphorylase b (3), BSA (4), catalase (5). actin (6), aldolase (7), and carbonic anhydrase (8). Dots adjacent to the dense body channel, above the heavy band that migrates close to BSA (4), indicate the position of the 80K protein.

(DATD), the basic phosphoprotein was observed to migrate more slowly than the major capsid protein (unpublished observations). The second experiment was done (i) to determine whether the radiolabeling pat-

I

RELATIVE MOLAR ABUNDANCIES OF HCMV AND VIRION PROTEINS

NIEP

Relative molar abundancy” Protein”

Protein size (M x 10-3)

HMWP MCP BPP 115 Matrix Matrix 36K mCP

212K 153K 149K 115K 74K 69K 36K 34K

NIEP 0.21 1.06 0.81

0.25 0.78

Virion

0.23 1.00

0.78 0.24 1.00

2.00

1.17 1.61

NP”

0.75

0.56

“Abbreviations explained in legend to Fig. 7. b Calculations based on densitometric scans (AMO) of the CBB-stained gel shown in Fig. 8. Values determined by dividing the absorbance of the band (arbitrary units) by the molecular weight of the corresponding protein. All values normalized to that determined for the major capsid protein (MCP). ’ NIEP protein (NP) not present in virions.

HCMV

VIRION-LIKE

NIEP producticm by other strains. NIEPs were first noticed in our studies using the AD169 isolate of HCMV. In order to determine whether they were peculiar to that laboratory prototype strain, the following experiment was done. HFF cells were infected with laboratory prototype HCMVs (strains AD169, Davis, or Towne), wildtype HCMVs isolates (strains 751, CMT), strain Colburn CMV (simian-like isolate), or Old World monkey CMVs (strains CSG, RCMV), respectively. When most of the cells in the culture showed typical late CMV cytopathic effect (e.g., cell rounding and sloughing from culture surface; 5-10 days after infection), the culture media were collected and layered above glyceroltartrate gradients, as described under Materials and Methods. Following rate-velocity centrifugation, the gradients were inspected for the presence or absence of a NIEP band, and absorbance profiles were recorded for those in which the band was present. Neither strain Colburn nor the Old World monkey CMV isolates, nor HSV (types 1 and 2) gave rise to NIEPs. However, all of the HCMV isolates examined did. Measurements made from the absorbance profiles shown in Fig. 9 indicate that HCMV strains Davis, Towne, and 751 produce about 5-10s the amount of NIEPs as virions. In dramatic contrast, strain AD169 I

I

NltP

FIG. 9. Comparison of the relative amounts of NIEPs and virions produced by different strains of HCMV. Infected cell media were clarified, subjected to glycerol-tartrate gradient centrifugation, and the resulting gradients monitored at 280 nm-all as described in Materials and Methods. Shown here are absorbance profiles of the region of the gradients containing NIEPs and virions. All recordings were made at a full scale absorbance of 0.2.

129

PARTICLE

typically produces 50-75s the NIEPs as virions, and strain produced a comparatively large of NIEPs (i.e., 30% amount of this experiment.

amount of CMT also proportion virions) in

Cmparism of NIEP-sp&ic 36K protein from other HCMV strains. To determine whether the distinguishing 36K protein was present in NIEPs from other HCMV strains, NIEPs and virions were recovered, as described under Materials and Methods, from HFF cultures infected with either AD169, Towne, Davis, or ‘751 and biosynthetically labeled with 32Pi. The cells were separated into cytoplasmic and nuclear fractions using NP40. Following solubilization, the radiolabeled particles and the nuclear cell fractions were subjected to electrophoresis in polyacrylamide gels. Fluorograms prepared from the resulting gels are shown in Fig. 10 and reveal that (i) NIEPs from all HCMV strains tested contain a phosphorylated protein (35K36K) that is not present in corresponding virion preparations, (ii) the 36K protein of AD169 is slightly larger than the 751, Davis, and Towne counterpart (about 35K)-both in NIEPs and in the nuclear fraction, and (iii) neither CSG (Old World monkey isolate)-nor Colburn-infected cells produce NIEPs, and the virions from these strains do not contain a 36K phosphorylated protein. Two-dimensional (charge-size) separations of NIEP proteins in denaturing polyacrylamide gels have established that, although the AD169 NIEP-specific protein is slightly larger than its counterpart in other HCMV strains, it exhibits essentially the same net charge (basic) and charge heterogeneity as the others (data not shown). Thus, all HCMV strains examined have a protein with the following biochemical characteristics: present in NIEPs but not virions, located in the nuclear fraction of the infected cell, 35-36K in size, phosphorylated, basic, and heterogeneous in charge. DISCUSSION

This report describes the results of a series of experiments done to determine the biological, physical, and biochemical characteristics of a noninfectious, enveloped

130

IRMIERE

NIEP Viriins ‘al&&j

FIG. 10. Phosphorylated proteins present in CMVinfected cell nuclei, NIEPs, and virions. Infected cells, radiolabelled with [32P]orthophosphate, were separated into nuclear and cytoplasmic fractions; virions and NIEPs were prepared from the infected-cell media. Samples were solubilized and analyzed by electrophoresis in either a “high-bis” (“NUC.,” nuclear fractions) or normal (NIEP and virion preparations) 10% polyacrylamide gel. The position of the AD169 NIEP-specific phosphoprotein (i.e., 36K) is indicated in each gel. Other phosphoproteins (see Gibson, 1933) are shown for comparison in the right-hand two channels. Neither of these isolates (CSG or Colburn) produces NIEPs.

particle (NIEP) recovered from the growth medium of human foreskin fibroblast (HFF) cells infected with human strains of cytomegalovirus (HCMV). Although similar in appearance and protein composition to virions, NIEPs are distinguished from them principally by their lack of DNA, and, consequently, infectivity. The absence of DNA in these particles is also thought to account for their lighter density, altered “core” appearance and, as discussed further below, may relate to the presence of an additional protein species not found in virions. Earlier studies of thin-sectioned HCMV-infected cells have shown a virionlike enveloped particle with an atypical

AND

GIBSON

ring-shaped core structure (Ruebner et al, 1965; Kanich and Craighead, 1972). While we cannot be certain that these particles represent intracellular NIEPs, this possibility seems likely and is compatible (i) with the fact that NIEPs can be recovered from mechanically disrupted cells (our unpublished observations), and (ii) our evidence that NIEPs are not produced as artifacts of our culture conditions or isolation procedures, as summarized at the beginning of the Results section. Electron microscopy was initially used to evaluate the general structure and purity of the particles, and subsequently to obtain more detailed information about their architecture. Dimensions calculated for virions and some of their structural elements (e.g., capsomere, nucleocapsid, envelope) are similar to those previously described for herpesviruses in general (Wildy et al., 1975; Smith and Rasmussen, 1963; Watson, 1973; Vernon et al., 1974; Palmer et aL, 1975) and CMV in particular (Smith and Rasmussen, 1963; Wright et a& 1964). The NIEP capsid was found to be smaller in diameter than that of virions and to have a much less densely stained center. Both of these characteristics are ascribed to their lack of DNA. Central stain-excluding areas occasionally seen in the core region of thin-sectioned virions are consistent with the toroidal organization of encapsidated HSV DNA (Furlong et aL, 1972), and observations of a similar organization in CMV (Haguenau and Michelson-Fiske, 1975). Internal structures apparent in the core region of negatively stained NIEPs are also consistent with previously recognized features of CMV virus particles (Smith and Rasmussen, 1963; Wright et al, 1964; Benyesh-Melnick et al., 1966; Craighead et cd, 1972). The ultrastructural similarities between virions and NIEPs was reflected by their protein compositions. Both particles contain seven major protein species (Le., HMWP, BPP, MCP, 115K, 74K MP, 69K MP, mCP; see legend to Fig. 7), three of which (i.e., BPP, 74K MP, 69K MP) are predominant phosphoproteins. This description of HCMV virion proteins is compatible with results of previous studies (Sarov and Abady, 1975; Fiala et al, 1976;

HCMV

VIRION-LIKE

Kim et al, 1976; Gupta et a& 1977; SiqueiraLinhares et al, 1981; Mar et aL, 1981; Gibson, 1983). Two features of the NIEP protein composition distinguish it from that of virions. First, NIEPs contained significantly less (lo-30%) of the 74K and substantially less (30-60%) of the 69K matrix proteins than virions. This discrepancy could be explained if additional matrix protein, 69K in particular, were associated with the viral DNA or otherwise required for its incorporation into the particle. It could also be accounted for by volumetric considerations if the 35K NIEP protein occupies space in the particle that would otherwise contain matrix protein. And second, NIEPs contain a 35K protein that is not present in virions and which represents lo-15% of its protein mass, making it the most abundant NIEP protein species (Table 1). Although results presented here do not provide direct evidence concerning the ultrastructural location of this 35K NIEP protein within the particle, or its functional role, they do show that it has properties in common with the 37K B-capsid protein of strain Colburn CMV (simian-like isolate), and the 38-40K B-capsid protein of herpes simplex virus (HSV), both of which are capsid constituents. Each of these proteins is particle associated, about 35-40K in size; slightly basic and heterogeneous in charge, phosphorylated, located in the cell nucleus, and absent from the mature virion (Gibson and Roizman, 1972,1974; Heilman et al, 1981; Gibson, 1981). Accumulating evidence from preceding studies with pseudorabies (Ben-Porat et al, 1970; Ladin et al, 1970), HSV (Gibson and Roizman, 1972, 1974; Preston et aZ., 1983), equine herpes virus (Perdue et al, 1975, 1976), and strain Colburn CMV (Gibson, 1981), suggests that a protein of this general nature may play a common role in the assembly pathway of herpesgroup viruses. It has been proposed that this protein functions to mediate DNA packaging and/ or nucleocapsid envelopment, and in the normal flow of events is modified and/or removed from the particle (Gibson and Roizman, 1972; Gibson, 1981). To simplify discussing this herpesvirus group common species we refer to it by the term assembly protein rather than by its size which varies

PARTICLE

131

depending on both virus and strain. If, as results presented here indicate, the NIEPspecific protein is the HCMV assembly protein counterpart, then the observations presented here add to our understanding of its involvement during assembly as follows. First, the presence of the 35K protein in a particle lacking DNA (NIEP) suggests that its association with the capsid occurs prior to and independently of DNA packaging. Second, presence of the intact 35K protein in NIEPs demonstrates that its removal or degradation is not required for apparently normal capsid envelopment. And third, the existence of NIEPs establishes that the presence of DNA within the capsid is not an essential signal for its envelopment. The fact that a biochemical difference in the 35K protein of AD169 correlates with an overproduction of NIEPs, also suggests that NIEPs may prove useful in studying this protein’s function. As shown by Talbot and Almeida (1977), glycerol-tartrate gradients proved especially effective in separating virions from dense bodies. Structural and biochemical analyses of these comparatively pure preparations showed that dense bodies are simpler than NIEPs and virions in both structure and composition. As reported by others (Craighead et aL, 1972; Sarov and Abady, 1975; and Stinski, 1976), they appear as large (i.e., 250-600 nm) solid spheres, filled with homogenous material and bounded by an outer membrane. Results presented here established that they contain no DNA, in agreement with Sarov and Abady (1975), and that at least 90% of their protein mass is represented by the 69K matrix protein. Experiments with strain 751 HCMV indicates that both dense bodies and NIEPs contain a normal complement of virion glycoproteins (unpublished observations). The absence of DNA in NIEPs makes them of interest from both a basic research and clinical point of view. They represent attractive control particles for studies of (i) virion-associated enzyme activities, (ii) interactions between viral DNA and other capsid constituents and, (iii) early events in CMV infection, prior to viral DNA replication. Clinically, the development of a suitable vaccine has been suggested as a

132

IRMIERE

way to prevent CMV infections of neonates and immunocompromised patients (Lang, 1980; Osborn, 1981). As antigens, NIEPs would have advantages over both whole virus and subunit vaccines. Since NIEPs do not contain DNA, there is no associated risk of viral latency or oncogenicity, and since they have a full complement of virion proteins they should provoke an immune response to a greater variety of viral antigens than a subunit vaccine. ACKNOWLEDGMENTS We thank Dr. Jerry Wolinsky of the Department of Neurology for his generosity in preparing, examining, and photographing the thin-sectioned particle preparations; Dr. Thomas Pollard of the Department of Cell Biology and Anatomy for kindly sharing his electron microscope facilities with us; and Dr. Ueli Aebi of the same Department for advice on negative staining. We also thank Debra Livant aned Louise Flannery for typing the manuscript. These studies were aided by Research Grants AI 13718 and AI 16959 from the National Institute for Allergy and Infectious Diseases. Alice Irmiere is a predoctoral fellow in the Biochemistry, Cellular and Molecular Biology Training Program and was supported by Training Grant GM 07445 from the National Institutes of Health. REFERENCES BARZILAI, R., LAZARUS, L. H., and GOLDBLUM, N. (1972). Viscosity-density gradient for purification of footand-mouth disease virus. Archiv. Ges. Virusfmxh. 36, 141-146. BECKER, P., MELNICK, J. L., and MAYOR, H. D. (1965). A morphologic comparison between the developmental stages of herpes zoster and human cytomegalovirus. Exp. Mol. PathoL 4, 11-23. BEN-P• RAT, T., SHIMONO, H., and KAPLAN, A. S. (1970). Synthesis of proteins in cells infected with herpesvirus. IV. Analysis of the proteins in viral particles isolated from the cytoplasm and nucleus. I%rology 41, 256-264. BIJLENGA, R. K. L., AEBI, Il., and KELLENBERGER, E. (1976). Properties and structure of a gene 24-controlled T4 giant phage. J. MoL Bid 103,469-498. BONNER, W. M., and LASKEY, R. A. (1974). A film detection method for tritium-labeled proteins and nucleic acids in polyacrylamide gels. Eur. J. Biodem. 46,83-88. CRAIGHEAD, J. E., KANICH, R. E., and ALMEIDA, J. D. (1972). Nonviral microbodies with viral antigenicity produced in cytomegalovirus-infected cells. J. vird 10.766-775. DEMARCHI, J. M., BLANKENSHIP, M. L., BROWN, G. D., and KAPLAN, A. S. (1978). Size and complexity of human cytomegalovirus DNA. lrirology 89,643-646. FAIRBANKS, G., STECK, T. L., and WALLACH, D. F.

AND

GIBSON

(1971). Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Bie chemistry 10.2606-2617. FIALA, M., HONESS, R. W., HEINER, D. C., HEINE, JR., J. W., MURNANE, J., WALLACE, R., and GUZE, L. B. (1976). Cytomegalovirus Proteins I. Polypeptides of virions and dense bodies. J. Vim! 19, 243-254. FURLONG, D., SWIFT, H., and ROIZMAN, B. (1972). Arrangement of herpesvirus deoxyribonucleic acid in the core. J. lrird 10, 1071-1074. GEELEN, J. L. M. C., WALIG, C., WERTHEIM, P., and VAN DER NOORDAA, J. (1978). Human cytomegalovirus DNA I. Molecular weight and infectivity. J. ViroL 26, 813-816. GIBSON, W. (1981). Structural and nonstructural proteins of strain Colburn cytomegalovirus. virology 111, 516-537. GIBSON, W. (1983). Protein counterparts of human and simian cytomegaloviruses. virology 128, 391406. GIBSON, W., and ROIZMAN, B. (1972). Proteins specified by herpes simplex virus. VIII. Characterization and composition of multiple capsid forms of subtypes 1 and 2. J. ViroL 10, 1044-1052. GIBSON, W., and ROIZMAN, B. (1974). Proteins specified by herpes simplex virus. X. Staining and radiolabeling properties of B capsid and virion proteins in polyacrylamide gels. .J. ViroL 13, 155-165. GUPTA, P., ST. JEOR, S., and RAPP, F. (1977). Comparison of the polypeptides of several strains of human cytomegalovirus. .J. Gem. ViroL 34,447-454. HAGUENAU, F., and MICHELSON-FISKE, S. (1975). Cytomegalovirus: Nucleocapsid assembly and core structure. Intervirology 5, 293-299. HEILMAN, C., ZWEIG, M., and HAMPAR, B. (1981). Herpes simplex virus type 1 and 2 intracellular p40: Type-specific and cross-reactive antigenic determinants on peptides generated by partial proteolysis. .I ViroL 40, 508-515. IWASAKI, Y., FURUKAWA, T., PLOTKIN, S., and KoPROWSKI, H. (1973). Ultrastructural study on the sequence of human cytomegalovirus infection in human diploid cells. Arch Ges Vimsfwsch. 40,311324. KANICH, R. E., and CRAIGHEAD, J. E. (1972). Human eytomegalovirus infection of cultured fibroblasts II. Viral replicative sequence of a wild and an adapted strain. Lab. Invest. 27,273-282. KILPATRICK, B. A., and HUANG, E.-S. (1977). Human cytomegalovirus genome: partial denaturation map and organization of genome sequences. J. ViroL 24, 261-276. KIM, K. S., SAPIENZA, V. J., CARP, R. I., and MOON, H. M. (1976). Analysis of structural polypeptides of purified human cytomegalovirus. J. I%& 20,604611. LADIN, B. F., IHARA, S., HAMPL, H., and BEN-P• RAT, T. (1982). Pathway of assembly of herpesvirus capsids: an analysis using DNA+ temperature-sensitive

HCMV

VIRION-LIKE

PARTICLE

133

mutants of pseudorabies virus. Vi’irology 116, 544561. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Ladon) 227, 680-684. LANG, D. J. (1980). Cytomegalovirus immunization: status, prospects and problems. Rev. Infect Z&s. 2, 449-458. LASKEY, R. A., and MILLS, A. D. (1975). Quantitative film detection of ‘H and “C in polyacrylamide gels by fluorography. Eur. J. Biochem. 56,335-341.

Fraenkel-Conrat and R. R. Wagner, eds.), Vol. 3, pp. 229-403. Plenum, New York/London. RUEBNER, B. H., HIRANO, T., SLUSSER, R. J., and MEDEARIS, D. N., JR., (1965). Human cytomegalovirus infection. Electron microscopic and histochemical changes in cultures of human fibroblasts. Amer. J. Path& 46, 477-496. SAROV, I., and ABADY, I. (1975). The morphogenesis of human cytomegalovirus. Isolation and polypeptide characterization of cytomegalovirions and dense bodies. Virobgg 66, 464-473.

LASKEY, R. A., and MILLS, A. D. (1977). Enhanced autoradiographic detection of BP and ‘“I using intensifying screens and hypersensitized film. FEBS L&t 82, 314-316. LUSE, S. A., and SMITH, M. G. (1958). Electron microscopy of salivary gland viruses. J. Exp. Med 107, 623-632. MCGAVRAN, M. H., and SMITH, M. G. (1965). Ultrastructural, cytochemical, and microchemical observations on cytomegalovirus (salivary gland virus) infection of human cells in tissue culture. Ezp. Mol. Pathol. 4, l-10. MAR, E.-C., PATEL, P. C., and HUANG, E.-S. (1981). Human cytomegalovirus-associated DNA polymerase and protein kinase activities. J. Gen Viral 57.149-156.

SIQUEIRA-LINHARES, M. I., FAUCON-BIQUET, N., CHARDONNET, Y., and REV&LARD, J.-P. (1981). Polypeptides and major antigens of four new isolates of cytomegalovirus. Afed M&o&& Zmmunol 169,197208. SMITH, J. D., and DEHARVEN, E. (1974). Herpes simplex virus and human cytomegalovirus replication in WI38 cells. .Z. ViroL 14,945-956. SMITH, K. O., and RASMUSSEN, L. (1963). Morphology of cytomegalovirus (salivary gland virus). J. Batteriol. 85, 1319-25. STINSKI, M. F. (1976). Human cytomegalovirus: Glycoproteins associated with virions and dense bodies. J. Viral. 19, 594-609. TALBOT, P., and ALMEIDA, J. D. (1977). Human cytomegalovirus: Purification of enveloped virions and dense bodies. J. Gen. firoL 36, 345-349. VERNON, S. K., LAWRENCE, W. C., and COHEN, G. H. (1974). Morphological components of herpesvirus. I. Intercapsomeric fibrils and the geometry of the capsid. Zntewirolog~ 4.237-248. WATSON, D. H. (1973). Morphology. In “The Herpesvirus” (A. S. Kaplan, ed.), pp. 27-43. Academic Press, New York. WEBER, K., and OSBORN, M. (1969). The reliability of molecular weight determinations by dodecyl sulfate-polyacrylamide gel electrophoresis. J. Biol. Chm 244.4406-4412. WEINER, D., and GIBSON, W. (1981). Identification of a primate cytomegalovirus group-common protein antigen. virology 115, 182-191. WILDY, P., RUSSELL, W. C., and HORNE, R. W. (1960). The morphology of herpes virus. Virology 12,204222. WOLINSKY, J. S., BARINGER, J. R., MARGOLIS, G., and KILHAM, L. (1974). Ultrastructure of mumps virus replication in newborn hamster central nervous system. Lab. Invest. 31.403-412. WRIGHT, JR., H. T., GOODHEART, C. R., and LIELAUSIS, A. (1964). Human cytomegalovirus. Morphology by negative staining. virology 23, 419-424.

OSBORN, J. E. (1981). Cytomegalovirus:

immunology, 143,618-630.

and vaccine initiatives.

Pathogenicity, .Z. Z$ect Dis.

PALMER, E. L., MARTIN, M. L., and GARY, G. W., JR. (1975). The ultrastructure of disrupted herpesvirus nucleocapsids. Virology 65, 260-265. PATRIZI, G., MIDDLEKAMP, J. N., HERWEG, J. C., and THORNTON, H. K. (1965). Human cytomegalovirus: Electron microscopy of a primary viral isolate. J. Lab. Clin Med 65, 825-838. PERDUE, M. L., COHEN, J. C., KEMP, M. C., RANDALL, C. C., and O’CALLAGHAN, D. J. (1975). Characterization of three species of nucleocapsids of equine herpesvirus type-l (EHV-1). virob 64, 187-204. PERDUE, M., COHEN, J. C., RANDALL, C. C., and O’CALLAGHAN, D. J. (1976). Biochemical studies of the maturation of herpesvirus nucleocapsid species. Virology 74, 194-208. PRESTON, V. G., COATES, J. A. V., and RIXON, F. J. (1983). Identification and characterization of a herpes simplex virus gene product required for encapsidation of virus DNA. J. viral. 45, 1056-1064. B., and FURLONG, D. (1974). The replication of herpesviruses. In “Comprehensive Virology” (H.

ROIZMAN,