INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Human cytomegalovirus (HCMV) is a widespread infectious agent causing birth defects when primary infection is acquired during pregnancy and is associated with a wide range of clinical syndromes in immunocompromised patients, especially in transplant recipients and patients with AIDS. HCMV is a large betaherpesvirus with a complex structure consisting of a large double-stranded DNA genome exceeding 235 kb, with the potential to encode over 220 proteins, packaged in an icosahedral protein structure, the capsid (6, 8). A matrix of proteins, the tegument, covers the capsid, which is finally surrounded by a lipid envelope containing several viral glycoproteins (20). The maturation of the virus occurs in a multistep assembly process following replication of the HCMV genome in the nucleus of the infected cell (12). The nuclear matrix contains several domains, which are associated with different cellular processes, such as DNA replication, transcription, and pre-mRNA processing (30).

For human alpha- and betaherpesviruses, both viral DNA replication and viral RNA expression are confined to nuclear matrix structures, called nuclear dot domains (16, 27, 31). Cellular gene expression is located at a different site (16, 31). Sarisky and Hayward studied the constitution of HCMV viral replication centers and showed that HCMV replication centers could be induced in vitro by transfecting Vero cells with 11 HCMV genes essential for viral DNA replication (27). The formation of these HCMV replication compartments may be orchestrated by UL112 and UL113 gene products, as suggested by Penfold and Mocarski (21). Newly replicated viral DNA is formed in concatamers which are cleaved to unit length (17) and incorporated into the procapsid, thereby extruding the scaffolding protein, concomitant with a major conformational transition of the capsid shell (11). The nucleocapsid travels from the replication compartments to the nuclear boundary to become coated with a fuzzy layer of protein, the tegument (12). The inner nuclear membrane encloses the nucleocapsid, and after passing the perinuclear space, the capsid loses the inner nuclear membrane by exocytotic fusion with the membrane of the endoplasmic reticulum (10). Naked tegumented capsids are subsequently shed into the cytoplasm, where they may bind and penetrate by engulfing into Golgi and endosomal membranes, resulting in a double-enveloped viral particle (13, 23). Endosomal membranes may be derived from recycling cell surfaces, which contain the mature viral glycoprotein B protein (24), whereas Golgi membranes contain newly synthesized glycoprotein B. The engulfed virion traffics in Golgi or endosomal vesicles to the cellular membrane, where the enveloped virion is released from the host cell by exocytosis. In vivo HCMV spreads by close cell-to-cell contact, while production of cell-free virus seems to be restricted to some tissues (22). In vitro, the laboratory strain AD169 is abundantly released from infected cells into the supernatant. Besides structural proteins, the HCMV virion contains proteins such as pp71 (the product of the UL82 gene), the upper matrix protein, which is capable of transcriptional activation of the incoming viral genome shortly after penetration into the host cell and accelerates the infectious cycle (2, 15).

In studies of HCMV gene expression, nucleic acid sequence-based amplification (NASBA) technology was used to analyze HCMV RNA expression at early times postinfection (p.i.). When studying HCMV gene expression directly after the addition of strain AD169, it was noted that immediate-early (IE) and even late viral mRNA expression was already detectable at the moment of infection, even when UV-inactivated virus stocks were used. A more detailed analysis showed the persistent presence of IE and late mRNA even in gradient-purified HCMV. Besides viral RNA, RNA of cellular origin was also detectable in purified virions, whereas cellular DNA was absent. The presence and location of the viral and host-derived RNA in the virion appear to reflect nonspecific incorporation in the HCMV virion. In contrast to the function of tegument-associated proteins, our results suggest that virion-associated RNA may not serve a specific biologic function at early times p.i.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Virus and cell culture. Human fetal lung fibroblasts (HLF) were cultured in a 1:1 mixture of Ham's F-12 medium and Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum (HyClone, Logan, Utah). Virus stocks of HCMV strain AD169 were made by infecting HLF cells at a multiplicity of infection (MOI) of 0.1 PFU per cell and maintaining the cells until 3 days after they reached a 100% cytopathic effect (CPE) (18). The medium was collected and cleared of cellular fragments by centrifugation at 1,200 × g. For RNA studies, HLF cells were infected with cell-free HCMV virus at an MOI of 0.01 or 5 PFU per cell for 1 h, and medium with 10% fetal bovine serum was refreshed, followed by incubation for 3 days. Similarly, cells were inoculated with UV-inactivated virus and irradiated at 2 J/m² per s for 10 min with mixing every 2 min. Infected cells were washed twice with phosphate-buffered saline (PBS) and harvested by trypsinization, followed by centrifugation at 1,000 × g. The cell pellet was washed with PBS and dissolved in NASBA lysis buffer (4.7 M guanidinium thiocyanate, 46 mM Tris [pH 6.4], 20 mM EDTA, 1.2% [wt/vol] Triton X-100). Highly purified virions were obtained from ABI (Columbia, Md.). Purification of virions was performed with a density step gradient of 20 and 40% sucrose in PBS and collected at the 20 to 40% interphase. The ratio of virions to incomplete virions was 2:1 as determined by electron microscopy. De-enveloped capsids were prepared from the purified virus by incubation in detergent buffer (PBS with 1% Triton X-100) for 30 min at 4°C, and simultaneously released nucleic acids were degraded by adding 10 µl of RNase A (10 mg/ml; Qiagen, Hilden, Germany) and 30 U of DNase 1 (Pharmacia, Uppsala, Sweden). Highly concentrated viral nucleic acids were obtained from purified virions by sodium dodecyl sulfate (SDS)-mediated lysis and digestion with proteinase K, followed by phenol-chloroform extraction and ethanol precipitation; the procedure was performed by ABI.

Nucleic acid isolation. Total nucleic acids were isolated by silica-based extraction from samples representing 10⁴ infected cell equivalents, 10⁶ purified virion equivalents, 10⁶ and 10⁸ purified capsid equivalents, and 3 µl of viral DNA derived from 6.5 × 10⁷ virus particles per µl prior to isolation. Samples were dissolved in NASBA lysis buffer and isolated essentially as described by Boom et al. (5). The nucleic acids were bound to acid-activated silica for 10 min at room temperature. The silica was washed twice with wash buffer (5.3 M guanidinium thiocyanate, 50 mM Tris [pH 6.4]) and once in 70% ethanol and acetone, after which the silica was dried. Nucleic acids were eluted with 50 µl of elution buffer (1 mM Tris [pH 8.5]) at 56°C for 10 min.

Primers and probes. Primers and probes used in this study for NASBA and PCR are described in Table 1.

NASBA. Five microliters of nucleic acid eluate was used as input for NASBA analysis. The protocol was performed basically as described by Kievits et al. (14). The NASBA amplification reaction was performed with two primers designed for amplification of RNA specific for HCMV IE gene UL123 spanning an intron between exons 2 and 3 (29) (IE2.3 NASBA assay), for UL65 encoding the late tegument protein pp67 (3, 9), and for cellular genes encoding GAPDH (1) and U1A protein (28). NASBA reactions were carried out in a 20-µl reaction mixture containing 40 mM Tris (pH 8.5), 12 mM MgCl₂, 70 mM KCL, 15% (vol/vol) dimethyl sulfoxide; 5 mM dithiothreitol, each deoxynucleoside triphosphate at a concentration of 1 mM, 2 mM (each) ATP, CTP, and UTP, 1.5 mM GTP, 0.5 mM ITP, 2 µg of bovine serum albumin (Boehringer GmbH, Mannheim, Germany), 0.08 U of RNase H (Pharmacia), 32 U of T7 RNA polymerase (Pharmacia), 6.4 U of avian myeloblastosis virus reverse transcriptase (Seikagaku, Rockville, Md.), each primer at a concentration of 0.2 µM, and 5 µl of isolated nucleic acids. The NASBA reaction mixture without enzymes was incubated for 5 min at 65°C and was cooled to 41°C in 5 min, followed by addition of the enzymes. The reaction mixture was incubated for 90 min at 41°C. The quantitative NASBA assay was performed by competitive coamplification of calibrator RNA generated in vitro, shown to give amplification kinetics identical to those of wild-type RNA (A. E. Greijer, et al., submitted for publication). The sensitivity of the NASBA assays of pp67, GAPDH, and U1A was 10 to 100 molecules of in vitro-transcribed RNA, whereas the sensitivity of the IE2.3 NASBA assay was 1,000 molecules of RNA.

PCR. For DNA amplification by PCR, primers and probes (Table 1) were chosen from the UL65-coding region of the late tegument protein pp67 and GAPDH genes as described above. DNA amplification was essentially performed according to the method of Saiki et al. (26). For single PCR, 1 µl of silica eluate was amplified in a 20-µl reaction mixture containing 100 mM Tris-HCl (pH 8.3), 500 mM KCl, 15 mM MgCl₂, 0.01% (wt/vol) gelatin, and 5 nM each primer. The PCR was carried out in a thermal cycler (MJ Research, Watertown, Mass.) for 30 cycles (94°C for 30 s, 55°C for 30 s, and 72°C for 1 min). PCR products were analyzed on 1.5% agarose gels and stained with ethidium bromide.

Gel electrophoresis and blot analysis. The amplification products of the IE2.3 NASBA assay were analyzed with the CleanGel DNA Analysis kit (Pharmacia). The CleanGel was capillary blotted for 10 min to a Zetaprobe membrane (Bio-Rad, Hercules, Calif.), and nucleic acids were cross-linked by UV light. NASBA and PCR products were analyzed on 1 to 2% agarose gels and blotted by vacuum for 1 h. The blot was prehybridized with a hybridization mixture (5× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 7% SDS, 20 mM NaPi [pH 6.7; 0.1 M Na₂HPO₄ and 0.1 M NaH₂PO₄ in a 1:1.45] ratio, and 1× Denhardt's solution) for 30 min at 50°C. The probe was 5' end labeled with [-³²P]ATP and T4 polynucleotide kinase (Pharmacia) and was added to the blot. After overnight hybridization, the blot was washed with 3× SSC and 1% SDS and exposed to an X-ray film.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Viral RNA detection in the virion particle. HLF were infected with cell-free HCMV with an MOI of 5 PFU per cell. Surprisingly, we found that when using a highly sensitive NASBA assay for the detection of viral RNA transcripts, positive results were obtained for IE and late viral mRNA immediately after adding HCMV to the cells. The virus inoculum also proved to be positive for viral RNA. In addition, when UV-inactivated HCMV was used for inoculating fibroblasts, viral RNA encoded by UL123 and UL65 could be detected in the cell pellet harvested after 1 h. After refreshment of the medium, IE and pp67 RNA could no longer be detected by NASBA from 4 to 96 h p.i. The UV inactivation of HCMV adequately inhibited infection, since no CPE was observed in the cells 96 h p.i., and no viral proteins were detected by immunofluorescent staining, excluding new RNA synthesis as the likely source for the RNA detected at 1 h postinfection. These observations might be explained by the presence of RNA in the virus inoculum or by assuming that NASBA also uses DNA as a template for amplification in addition to RNA.

View larger version (55K): [in this window] [in a new window]   FIG. 1.   NASBA and PCR analyses of the nucleic acids detected in virions, capsids, viral DNA, and infected cells. Nucleic acids were isolated from gradient-purified virions (lanes 1 to 3), de-enveloped capsids (lanes 4 to 6), viral nucleic acids (lanes 7 to 9), and infected cells (lanes 10 to 12). (A) IE2.3 NASBA assay; (B) DNA amplification by PCR of pp67 (UL65); (C through E) NASBA assay of pp67 (UL65) (C); U1A (D); and GAPDH (E). , untreated; + DNase 1, DNase 1 treatment; + RNase A, RNase A treatment.

Presence of viral RNA in the capsid. Since virion capsid formation and envelopment occur at different sites in the cell, the localization of RNA in the virion was evaluated. The virion membrane was disrupted by adding 1% Triton X-100, and the permeabilized virions were incubated on ice in the presence of DNase 1 and RNase A. Purified de-enveloped capsids were checked for infectivity with negative results, indicating complete destruction of the viral envelope. By inoculating 10⁷ HLF cells with the capsid fraction (10⁸ particles), no CPE was seen in the cells even after 3 weeks of incubation. However, with IE2.3 NASBA IE RNA remained detectable in the capsid fraction (Fig. 1A, lanes 4 and 5). The amount of RNA is significantly smaller inside the capsid than in the intact virion (Fig. 1A, lane 1), since a 100-fold-higher input of capsid was used for IE2.3 NASBA analysis. This raised the question whether only an RNA-encoding IE gene was incorporated into the virion or whether RNA encoded by other viral genes could also be amplified from the purified virion and capsid preparation. Therefore, nucleic acid eluates from the virion, capsid, viral nucleic acids (ABI), and infected cells were used to amplify the intronless late gene UL65 encoding pp67 by using NASBA analysis before and after digestion with either DNase 1 or RNase A. In all cases, UL65 RNA and DNA yielded identically sized NASBA products (Fig. 1C), except for the infected cell-derived nucleic acids treated with RNase A. The remaining viral DNA is at too low a copy number to yield a UL65 product.

Presence of cellular messenger RNA in the virion. The presence of virus-encoded messenger RNA in the virion and capsid could be due to specific incorporation which might be of benefit in the infectious process. Alternatively, viral RNA could be incorporated at random in the virion during the assembly process. Consequently, if mRNA were enclosed nonspecifically into the virion, cellular RNA might also be present in the virion. Therefore, NASBA amplification was performed using primers specific for the U1A gene, a cellular gene with low rates of transcriptional activity. U1A mRNA was present in infected cells as seen in Fig. 1D, lanes 10 and 11, but was also detectable in purified HCMV virions and virion nucleic acids (lanes 1, 2, 7, and 8). RNase A treatment confirmed the specificity of the U1A reaction (lanes 2, 5, and 8). However, in purified de-enveloped capsids U1A mRNA could not be detected (lanes 4 and 5). In order to confirm the presence of cellular mRNA in virions but its absence in capsids, a highly abundant cellular mRNA encoded by the GAPDH gene was analyzed by NASBA. The results shown in Fig. 1E were similar to those for U1A (Fig. 1D). The virion showed detectable levels of GAPDH RNA (Fig. 1E, lanes 1), but the capsid did not contain RNA (lane 4). Again, RNase A treatment confirmed the specificity of the NASBA reaction, as no DNA-derived product was observed (Fig. 1E, lanes 3 and 6). Although the NASBA of intronless IE exon 4 (4), which has a higher sensitivity than IE2.3 NASBA (data not shown), and the pp67 NASBA were positive (Fig. 1C, lane 4), the GAPDH and U1A NASBA with identical sensitivity gave negative results. This was confirmed using 100-fold-larger input of nucleic acid derived from capsids (Fig. 1D and E, lanes 4 to 6). To ensure there was no cellular DNA contamination in the virion preparation, GAPDH PCR was performed. GAPDH DNA could not be detected in the virion and capsid fractions (data not shown).

Ratio of viral messengers in cell versus virion. Since the virion contained cellular RNA in addition to viral RNA, it seemed likely that RNA was packaged nonspecifically. In order to analyze whether RNA is randomly enclosed in the virion, the ratio of RNA encoded by the IE exon 4 gene and the UL65 gene whose product is pp67 was determined in virions and in infected cells by a quantitative NASBA analysis. In case of nonspecific incorporation, the ratio of RNA in virions and in extracts of infected cells would be similar. The results of quantitative NASBA analysis for the IE exon 4 gene and the UL65 gene and are shown in Table 2. The observed ratio of IE exon 4 versus pp67 RNA in the cells was 3.5 as compared to the ratio of 1 in virions (log difference of 0.54 with a standard deviation of ±0.36). The theoretical ratio for RNA enclosure without selection is 1, since the ratio would be the same in the virion as in the cytoplasm. The observed ratio of 3.5 indicates, rather, incorporation without selection, since specific incorporation is expected to give a more dramatic difference.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

As with most herpesviruses, HCMV virion assembly is a complex process and many aspects of DNA packaging, capsid assembly, perinuclear transport, and virion envelopment remain to be defined in detail. A new complexity was added by our finding of the presence of viral and cellular RNA in the virion. Using NASBA for detection, RNA was located in the virion and even associated with the capsid of HCMV, however in a lower quantity. Characterizing the RNA in the virion revealed the presence of spliced IE mRNA encoded by the UL123 gene and late mRNA encoded by the UL65 gene. In addition, mRNA is encoded by two cellular genes: U1A, expressing a low-abundant mRNA, and GAPDH, expressing an abundant mRNA, were located in purified virions. Upon detergent-mediated removal of the envelope, the capsid was shown to contain only viral mRNA, suggesting a different way that viral and cellular RNA are incorporated during virion formation. The presence of multiple viral and cellular mRNA molecules indicates random enclosure of RNA into the virion particle during envelopment rather than specific enrichment by a selective process. The random incorporation of RNA is further confirmed by the virtually identical ratio of IE1 to pp67 RNA in the virion and the infected cells. The different levels of cellular and viral RNA found in the capsid and the virion suggest that the mode of acquisition of RNA differs for the inner and outer capsid space. A possible explanation may be found in the route of assembly of the HCMV virion. The capsid is formed within the replication centers of HCMV at particular loci in the nucleus (21). At these centers, a procapsid scaffold is assembled, after which the DNA is packaged, leading to further maturation of the capsid (12). During formation of the capsid and incorporation of the viral genome into the capsid, nuclear fluid may be enclosed into the capsid as well. Since viral RNA synthesis is closely linked to viral DNA replication, especially at late times p.i., viral RNA may be present at the site of capsid assembly. Alternatively, RNA actively transcribed from the viral genome and still associated with the viral DNA may be cotransported into the procapsid during maturation. The latter explanation of the presence of viral RNA in the capsid is not likely, since the IE RNA (UL123) detected was found to be spliced between exons 2 and 3. The absence of cellular RNA in the capsid may be explained by the distant location of HCMV genome replication and packaging relative to the site of cellular RNA synthesis, which takes place outside the viral replication compartments (16, 31). RNA outside the capsid may be obtained in a different way. The capsid leaves the nucleus after obtaining a tegument layer at the nuclear boundary by passage through the inner nuclear membrane. Subsequently, the nucleocapsid loses the membrane by fusion with the membrane of the endoplasmatic reticulum, and the unenveloped capsid, coated with a fuzzy tegument layer, is released into the cytoplasm (12, 23). The coated capsid is finally engulfed by an endosome or inserted into Golgi membranes as described in the Introduction. The endosomal or Golgi membrane forms the final envelope of the virion (10, 13). In the process of enveloping tegument-coated nucleocapsids, cytoplasmic fluid may be enclosed. In the cytoplasm, mRNA from viral and cellular origin is present in high concentrations.

Consequently, viral and cellular RNA may be incorporated into the virion in a nonselective manner. In the case of random incorporation of RNA, the composition of the cytoplasm should be roughly equivalent to that in the virion, which is confirmed by the observed similar ratio of pp67 mRNA to IE1 mRNA in the infected cells versus that found in virions. Another indication for the random enclosure of RNA is the presence of late viral messengers, which are only expressed after the replication of HCMV DNA is initiated. Although it has been shown that viral tegument proteins, such as the higher matrix protein pp71 (the product of UL82), may be functionally relevant during the early stages of viral entry (2), the assumption that specific incorporation of late viral RNA into the virion would give an advantage to the virus upon cell entry is not very likely. Since the replication cycle of HCMV is 72 h long and active IE1 mRNA synthesis can already be detected at 0.5 hours p.i., the contribution of specific viral mRNA introduced with the virion is considered to be minimal. RNA encoded by UL65 and UL123 is abundantly present in the cytoplasm at the time of virion assembly at, respectively, 10⁵ and 10⁴ copies of RNA per cell (19), which may facilitate the nonspecific acquisition of pp67 and IE1 RNA by the virion during envelopment. Further characterization of the exact composition of RNA in the virion and capsid by gene array technology could shed more light on the diversity of RNA derived from viral and cellular origins (7). In recent experiments, the presence of immune evasion RNA encoded by the US3, US6, and US11 genes in the virion was detected as well (unpublished results). RNA found in the virion probably represents mature and functional RNA. In order to confirm the hypothetical functionality of the virion RNA, in vitro translation studies should be performed. However, the amount of virus needed to isolate enough virion-associated RNA for translation is extremely large, since the concentration of randomly acquired RNA is low as indicated by the quantification of viral RNA in virions. Since the virion assembly process is quite similar for most herpesviruses, the acquisition of mRNA might be a common phenomenon. Our results indicate that enclosure of RNA into virions during viral assembly may be without selection, therefore without a particular biological function.