点击显示 收起
Department of Biology, Georgia State University
Measles Virus Section, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia
ABSTRACT
Proof of concept for a novel diagnostic assay for rubella virus (RUB) based on RUB replicons expressing reporter genes was demonstrated. RUB replicons have the structural protein coding region replaced with a reporter gene such as green fluorescent protein or chloramphenicol acetyltransferase. Previously, it was shown that a replicon construct with a specific in-frame deletion in the nonstructural protein coding region (NotI, approximately nucleotides 1500 to 2100 of the genome) failed to replicate and express the reporter gene unless rescued by a coinfecting wild-type helper RUB (W.-P. Tzeng et al., Virology 289:63-73, 2001). In the present study, it was found that rescue of reporter gene expression by NotI replicons occurred when coinfection was done with clinical specimens containing RUB, indicating that this system could be the basis for a diagnostic assay. The assay was sensitive, using laboratory RUB strains and as low a dose as one plaque-forming unit. The assay was specific in that it was positive for RUB strains of both genotypes and was negative for a panel of human viruses. It was also possible to genetically sequence the RUB present in positive clinical specimens detected in the assay for genotypic strain determination.
INTRODUCTION
Diagnosis of rubella virus (RUB) infection is essential after potential exposure of a woman in the first trimester of pregnancy (2), is necessary for confirmation of potential cases of congenital rubella syndrome in newborns (12), and is important in the surveillance component of rubella vaccination programs (4, 11, 20). Interestingly, diagnosis of RUB infection is also an important aspect of surveillance in measles vaccination programs since rubella is the most common nonmeasles rash illness screened during measles surveillance. Serodiagnostic assays for RUB infection are based on the detection of RUB-specific immunoglobulin M (IgM) in serum and saliva; IgM is also diagnostic of congenital RUB infection if present in fetal or newborn serum (1, 2). However, on the day of rash onset, only 50% of patients are IgM positive (1, 2). The percentage increases to 90% by 5 days after rash onset, and detection of RUB-specific IgG commences 3 days after rash onset; however, in light of the benign nature of the disease, it is often difficult to convince patients to return for collection of a second specimen (1, 2). In comparison, 90% of throat swab specimens taken on the day of rash onset are positive for the presence of RUB by virus isolation or reverse transcription-PCR (RT-PCR) assay (1). These techniques have also been used successfully to detect RUB in serum, saliva, nasopharyngeal washes, and urine (RT-PCR has been used to detect RUB genomes in chorionic villi or amniotic fluid in utero specimens [3, 14, 16, 17]). Thus, direct detection of virus is more sensitive than serodiagnosis in diagnosis of RUB infection on the days immediately after presentation of symptoms.
Both primary monkey kidney cell cultures or cultures of suitable continuous monkey kidney cell lines have been used to isolate RUB from clinical specimens. Detection of RUB in these cultures was classically done by challenge of the inoculated culture 7 to 10 days postinoculation with an enterovirus such as coxsackievirus A9, echovirus 11, or echovirus 40 and observance of the culture for development of cytopathic effect (9). RUB in clinical specimens induced interferon, which inhibited both replication and cytopathic effect induction by the challenge virus. Specificity was demonstrated by treating the clinical specimen with anti-RUB antibodies prior to inoculation of a parallel culture. Sometimes, passage of the culture fluid from the initially inoculated culture to a fresh culture, followed by incubation prior to challenge, was required, and thus this procedure generally took from 10 days to 2 weeks to perform. More recently, the presence of RUB in the inoculated culture has been demonstrated by immunofluorescence assay (IFA) with monoclonal antibodies to the RUB virions proteins (1) and by RT-PCR on RNA extracted either from the cells or the culture fluid (11, 13). Although more rapid than virus challenge, incubation of the inoculated culture for a period of a few days to a week to allow virus replication to ensue is necessary to generate a positive IFA or RT-PCR signal (1, 11, 13).
RUB has a plus-strand RNA genome that contains a 5' proximal open reading frame (ORF) that encodes nonstructural proteins involved in virus RNA replication and a 3'-proximal ORF that encodes the virion structural proteins (reviewed in reference 7). Using an infectious cDNA clone containing the RUB genome, we replaced the 3'-proximal ORF with a number of reporter genes such as green fluorescent protein (GFP) and chloramphenicol acetyltransferase (CAT) (18). These constructs, termed replicons, replicate when introduced into cells, express the reporter gene but do not spread from cell to cell unless helper wild-type (wt) virus is present to provide the virion structural proteins. In a previous study, we showed that replicons with a specific in-frame deletion of 600 nucleotides (nt) between two NotI sites in the nonstructural protein ORF, termed the NotI-NotI deletion or NotI, neither replicated nor expressed a reporter gene unless wt helper virus was present to rescue replication by the replicon. We later showed that the basis of this rescue was that capsid protein provided by the wt helper virus complemented the NotI replicon in trans; complementation exhibited by the capsid protein was not related to formation of virions or virus spread (19). We show here that NotI replicons can also be rescued by RUB present in clinical specimens, leading to proof of concept for a novel assay for RUB based on reporter gene expression by NotI replicons.
MATERIALS AND METHODS
Cells and viruses. Vero cells, obtained from the American Type Culture Collection (ATCC; Rockville, Md.) were maintained at 35°C under 5% CO2 in Dulbecco minimal essential medium (Invitrogen, Rockville, Md.) supplemented with 5% fetal bovine serum (FBS) and gentamicin (10 μg/ml) (Invitrogen) and were subcultured weekly by treatment with trypsin. The following RUB laboratory and vaccine strains as used in our lab have been described previously: W-Therien, F-Therien, RA27/3, and BRDII. The M33 strain of RUB was provided by Shirley Gillam, University of British Columbia. The Edmonton strain measles virus was provided by Paul Rota, Centers for Disease Control and Prevention (CDC), whereas the Sindbis virus (HR strain) and vesicular stomatitis virus (VSV; IND serotype) used in the study originated at the California Institute of Technology and the University of Pittsburgh, respectively. Herpes simplex virus type 1 (HSV-1) and HSV-2, human herpesvirus 6 variant A (HHV-6A), HHV-6B, and HHV-7 were provided by Phil Pellett (CDC), varicella-zoster virus (VZV) by Scott Schmid (CDC), and parvovirus B19 was provided by Dean Erdman (CDC).
Transfection and reporter gene expression. The replicon constructs RUBrep/GFP, RUBrep/GFP-NotI, RUBrep/CAT, and RUBrep/CAT-NotI constructs were described previously (18, 19). The plasmids were purified by CsCl isopycnic density centrifugation, followed by linearization with EcoRI (RUBrep/GFP-NotI) or SpeI (RUBrep/CAT and RUBrep/CAT-NotI) prior to in vitro transcription, as previously described (19). The transcription reaction mixtures were used directly for transfection without DNase treatment or phenol-chloroform extraction. Transfection was done with Lipofectamine 2000 (Invitrogen). To monitor GFP expression, the medium was removed from the transfected cells and replaced with a enough phosphate-buffered saline to keep the cells wet. The cells were then examined with a Zeiss Axioplan upright microscope with epifluorescence capability. Various magnifications were used, but the x10 objective was most useful for rapid screening for GFP fluorescence. If necessary, since the cells were still alive, the medium was replaced, and the cells were returned to the incubator for examination on subsequent days. To detect CAT expression, transfected cells were lysed, and the CAT activity was determined by using an assay described by Seed and Sheen (15).
Clinical specimens. For a preliminary study, four clinical specimens (throat swabs and urine samples) from patients with rubella-like illnesses and fluid from cultures inoculated with these specimens that were positive for RUB (12) were provided by Emily Abernathy (Georgia State University). For a subsequent study with previously untested specimens, 18 throat swabs were used that had been collected during a rubella outbreak in Westchester County, N.Y. (5), and provided by A. J. Huang, Westchester County Hospital. Vero cells were inoculated with these specimens (P0). At 6 to 7 days postinoculation, the P0 culture fluid was removed, and the cells were transfected with RUBrep/GFP-NotI transcripts and observed for GFP expression as described above. At 2 days posttransfection with RUBrep/GFP-NotI transcripts, total RNA was extracted by using Tri-Reagent (Molecular Research Center), and the presence of RUB genomic RNA was assayed with a genome-specific RT-PCR assay that amplified nt 8652 to 9124 within the E1 gene (lacking in RUBrep/GFP-NotI); these nucleotides represent the "molecular epidemiology window" previously shown to be optimal for genotypic strain determination (22). The assay used the oligonucleotide nucleotide primer 5'-TTTTTTTTTCTATACAGCAAC-3 [oligonucleotide(dT), followed by the complement of the 3' 12 nt of the RUB genome to prime RT with Superscript III reverse transcriptase (Invitrogen) and upstream primer 5'-GCTTCCCCACCGACACCGTG-3' (colinear with nt 8652 to 8671 of the RUB genome)] and downstream primer 5'-GAGTCCGCACTTGCGCGCCT-3' (complementary to nt 9105 to 9124 of the RUB genome) for PCR. LA Taq (Takara) was used for PCR under conditions supplied by the manufacturer. The sensitivity of this assay was determined to be 10 infected cells by using Vero cell culture infected with the Therien wt strain of RUB. The P0 culture fluid harvested prior to transfection was used infect fresh Vero cells (P1), and P1 culture fluid was subsequently passaged similarly (P2). Transfection with RUBrep/GFP-NotI transcripts and RT-PCR amplification of total cell RNA was done for P1 and P2 as described for P0. Amplicons produced from the P0 cultures that were positive for RUB were purified by using QIAquick gel extraction kit (Qiagen) and sequenced by using a 3100 Genetic Analyzer (Applied Biosystems/Hitachi). The sequencing reactions were done by using ABI Prism BigDye Terminator v3.1 cycle sequencing kit and the primers used for PCR. The sequence data was analyzed by using CLUSTAL W (version 1.8) multiple sequence alignment software.
RESULTS
As shown in Fig. 1B, cultures transfected with RUBrep/GFP transcripts express GFP that is detectable within 2 days posttransfection. In contrast, cultures transfected with RUBrep/GFP-NotI transcripts do not exhibit GFP expression (Fig. 1C). However, in cultures infected with wt helper RUB (F-Therien strain) 1 day before transfection with RUBrep/GFP-NotI transcripts, GFP expression is readily detectable (Fig. 1D and E). In the experiments shown in Fig. 1D to F, a stock of F-Therien virus with a titer of 106 PFU/ml was diluted 105-, 106-, and 107-fold prior to infection, resulting in additions of 10, 1, and <1 PFU to the plates shown in Fig. 1D, E, and F, respectively. GFP expression was detected in the plates shown in Fig. 1D and E, but not in 1F, indicating that as little as 1 PFU of wt helper virus is sufficient to rescue GFP expression by RUBrep/GFP-NotI. As shown in Fig. 2, similar sensitivity was exhibited when NotI replicons expressing the CAT gene (RUBrep/CAT-NotI) were used.
The RUBrep replicons are based on the Therien strain of RUB. We found that GFP expression by RUBrep/GFP-NotI could also be rescued by the M33 strain, a standard wt strain of RUB, and the RA27/3 vaccine strain, as well as by BRDII, a genotype II vaccine strain (Therien, M33, and RA27/3 are all genotype I) (Table 1). In contrast, a related plus-strand RNA virus member of the togavirus family, Sindbis virus, and a minus-strand RNA virus, VSV, did not rescue GFP expression by RUBrep/GFP-NotI. We also tested a number of human viruses associated with rash or rash-like illnesses or that could be expected to be present in clinical specimens taken to diagnose RUB infection, including measles virus, parvovirus B19, HSV-1 and -2, VZV, and HHV-6A, -6B, and -7, infection, with the result that none of these viruses rescued GFP expression by RUBrep/GFP-NotI. Thus, the effect is specific to RUB.
Finally, we tested clinical specimens starting with four specimens (one throat swab, two nasopharyngeal aspirates, and one urine sample) from which RUB had been isolated in a previous study (11). All four of these specimens rescued GFP expression by RUBrep/GFP-NotI (data not shown), demonstrating that the assay could be used to detect RUB in clinical specimens; the results with these specimens were reproducible in replicate assays. To expand the testing of clinical specimens, 18 throat swabs collected during a rubella outbreak in Westchester County, N.Y. (5), were used. Although viruses had been isolated from similar specimens collected during this outbreak (11), these 18 specimens had not previously been tested. As shown in Table 2, 8 of the 18 specimens were positive for RUB by RUBrep/GFP-NotI rescue in cells inoculated with the clinical specimens (P0). Examples of a negative specimen and positive specimens of different intensities are shown in Fig. 3; as can be seen, negative specimens produced no background and positive specimens could be readily distinguished. The results, both positive and negative, were confirmed by a RUB genome-specific RT-PCR assay (Fig. 4) with RNA extracted from the inoculated cells (which contained both virus and replicon RNAs). Medium collected from the cells inoculated with each of the clinical specimens before transfection with RUBrep/GFP-NotI was passaged twice to fresh Vero cells (P1 and P2); in each case the P0 result (+ or – for the presence of RUB) was repeated, and the results with RUBrep/GFP-NotI and RT-PCR were in complete agreement. The P0 RT-PCR amplicon from the eight positive specimens was sequenced and found to be identical to five viruses previously isolated from the Westchester County rubella outbreak (11) and different from any of the laboratory virus strains used.
DISCUSSION
We have provided here proof of concept for a novel diagnostic assay for RUB based on replicons expressing reporter genes. Transcripts from a replicon construct with a specific in-frame deletion in the nonstructural protein coding region of the genome (NotI) fail to replicate and express the reporter gene unless the virus capsid protein is provided by helper virus (19). With laboratory virus strains, this assay is sensitive to as low a dose as one PFU, and the assay works with viruses in clinical specimens as well. The assay is specific to RUB, including viruses from both genotypes, and is negative for viruses associated with rash illness or that might be expected to be present in clinical specimens taken to diagnose rubella infections (blood, nasopharyngeal aspirates, throat swabs, and urine). For conversion to a routine assay, development is necessary to determine optimal cell lines and reporter genes. We experimented with relative times of infection and transfection and found that the assay is successful when transfection is done either before or after infection (data not shown). Creation of a transgenic cell line that encodes an inducible NotI-reporter gene replicon (10) would be necessary for routine use of this assay. Interestingly, given that virus isolation is more sensitive than IgM detection for diagnosis of RUB infection at the time of rash onset (1), the novel replicon-based assay is also a potentially useful tool for the diagnosis of RUB infection, as would be other techniques that detect RUB. In this regard, the relative advantages of the replicon-based assay in comparison to other such techniques, such as IFA or RT-PCR, depends on the availability of equipment and reagents in the specific laboratory; in this regard the replicon-based assay requires cell culture and reporter gene detection capability. In terms of rapidity of the replicon-based assay, although cells inoculated with clinical specimens were incubated for 6 to 7 days before transfection with RUBrep/GFP-NotI transcripts to duplicate the virus isolation procedures used previously (11), we have found that cultures inoculated with clinical specimens can be transfected 2 days postinoculation with GFP expression detectable 1 day posttransfection, for a total assay time of 3 days (W.-P. Tzeng, preliminary observations).
In addition to virus detection, we showed that the novel replicon-based assay described here is also useful for virus genotyping, which remains a necessary part of the surveillance component of vaccination programs to determine virus distribution and transmission patterns (8, 11, 22, 23). The E1 gene has been used for all molecular epidemiology studies done on RUB thus far, although the region of the E1 gene used in these studies has not been standardized. Sensitive RT-PCR assays developed for direct RUB detection in clinical specimens produce amplicons of 125 to 300 nt (3, 14-17, 24), shorter than the 600- to 1,400-nt amplicons generated from virus isolates used for genotyping (8, 11, 22, 23), although recently a nested RT-PCR assay producing an 500-nt amplicon directly from clinical specimens that could potentially be used for both detection and genotyping was reported (6, 21). From cultures inoculated with clinical specimens shown to be positive for RUB by GFP expression after RUBrep/GFP-NotI transfection, we RT-PCR amplified an 500-nt window previously reported to be optimal for genotyping (22) that was RUB genome specific since it did not overlap with the replicon sequence. Sequencing of this amplicon confirmed the viruses present in positive clinical specimens to be identical to viruses previously isolated from companion clinical specimens from the same outbreak. It should also be pointed out that since detection of GFP expression can be detected in living cells (i.e., no fixation of the culture was required as is the case for IFA), virus from clinical specimens can potentially be directly isolated from culture fluid from positive cultures for further characterization and archiving.
ACKNOWLEDGMENTS
This study was supported by NIH grant AI21389 to T.K.F.
REFERENCES
Bellini, W. J., and J. P. Icenogle. 2003. Measles and rubella viruses, p. 1389-1403. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. ASM Press, Washington, D.C.
Best, J. M., and J. E. Banatvala. 2000. Rubella, p. 387-418. In A. J. Zuckerman, J. E. Banatvala, J. R. Pattison (ed.), Principles and practice of clinical virology, 4th ed. John Wiley, Chichester, United Kingdom.
Bosma, T. J., K. M. Corbett, M. B. Eckstein, S. O'Shea, P. Vijayalakshmi, J. E. Banatvala, K. Morton, and J. M. Best. 1995. Use of PCR for prenatal and postnatal diagnosis of congenital rubella. J. Clin. Microbiol. 33:2881-2887.
Castillo-Solorzano, C., P. Carrasco, G. Tambini, S. Reef, M. Brana, and C. A. De Quadros. 2003. New horizons in the control of rubella and prevention of congenital rubella syndrome in the Americas. J. Infect. Dis. 187(Suppl. 1):S146-S152.
Centers for Disease Control and Prevention. 1999. Rubella outbreak: Westchester County, New York, 1997-1998. Morb. Mortal. Wkly. Rep. 48:560-563.
Donadio, F. F., M. M. Siqueira, A. Vyse, L. Jin, and S. A. Oliveira. 2003. The genomic analysis of rubella virus detected from outbreak and sporadic cases in Rio de Janeiro state, Brazil. J. Clin. Virol. 27:205-209.
Frey, T. K. 1994. Molecular biology of rubella virus. Adv. Virus Res. 44:69-160.
Frey, T. K., E. S. Abernathy, T. J. Bosma, W. G. Starkey, K. M. Corbett, J. M. Best, S. Katow, and S. C. Weaver. 1998. Molecular analysis of rubella virus epidemiology across three continents, North America, Europe, and Asia, 1961-1997. J. Infect. Dis. 178:642-650.
Herrmann, K. L. 1979. Diagnostic procedures for viral, rickettsial and chlamydial infections, 5th ed., p. 725-766. Public Health Service, U.S. Department of Health and Human Services, Washington, D.C.
Olivo, P. D. 1996. Transgenic cell lines for detection of animal viruses. Clin. Microbiol. Rev. 9:321-334.
Reef, S. E., T. K. Frey, K. Theall, E. S. Abernathy, C. L. Burnett, J. Icenogle, M. M. McCauley, and M. Wharton. 2002. The changing epidemiology of rubella in the 1990s: on the verge of elimination and new challenges for control and prevention. JAMA 287:464-472.
Reef, S. E., S. Plotkin, J. F. Cordero, M. Katz, L. Cooper, B. Schwartz, L. Zimmerman-Swain, M. C. Danovaro-Holliday, and M. Wharton. 2000. Preparing for elimination of congenital rubella syndrome (CRS): summary of a workshop on CRS elimination in the United States. Clin. Infect. Dis. 31:85-95.
Revello, M. G., A. Sarasini, F. Baldanti, E. Percivalle, D. Zella, and G. Gerna. 1997. Use of reverse-transcription polymerase chain reaction for detection of rubella virus RNA in cell cultures inoculated with clinical samples. New Microbiol. 20:197-206.
Revello, M. G., F. Baldanti, A. Sarasini, M. Zavattoni, M. Torsellini, and G. Gerna. 1997. Prenatal diagnosis of rubella virus infection by direct detection and semiquantitation of viral RNA in clinical samples by reverse transcription-PCR. J. Clin. Microbiol. 35:708-713.
Seed, B., and J. Y. Sheen. 1988. A simple phase-extraction assay for chloramphenicol acyltransferase activity. Gene 67:271-277.
Tanemura, M., K. Suzumori, Y. Yagami, and S. Katow. 1996. Diagnosis of fetal rubella infection with reverse transcription and nested polymerase chain reaction: a study of 34 cases diagnosed in fetuses. Am. J. Obstet. Gynecol. 174:578-582.
Tang, J. W., E. Aarons, L. M. Hesketh, S. Strobel, G. Schalasta, E. Jauniaux, N. S. Brink, and G. Enders. 2003. Prenatal diagnosis of congenital rubella infection in the second trimester of pregnancy. Prenatal Diagn. 23:509-512.
Tzeng, W.-P., M.-H. Chen, C. A. Derdeyn, and T. K. Frey. 2001. Rubella virus DI RNAs and replicons: requirement for nonstructural proteins acting in cis for amplification by helper virus. Virology 289:63-73.
Tzeng, W.-P., and T. K. Frey. 2003. Complementation of a deletion in the rubella virus P150 nonstructural protein by the viral capsid protein. J. Virol. 77:9502-9510.
Venczel, L., J. Rota, V. Dietz, V. Morris-Glasgow, M. Siqueira, E. Quirogz, G. Rey, and C. A. De Quadros. 2003. The measles laboratory network in the region of the Americas. J. Infect. Dis. 187(Suppl. 1):S140-S145.
Vyse, A. J., and L. Jin. 2002. An RT-PCR assay using oral fluid samples to detect rubella virus genome for epidemiological surveillance. Mol. Cell Probes 16:93-97.
Zheng, D.-P., T. K. Frey, J. Icenogle, S. Katow, E. S. Abernathy, K.-J. Song, et al. 2003. Global distribution of rubella virus genotypes. Emerg. Infect. Dis. 9:1523-1530.
Zheng, D.-P., H. Zhu, M. G. Revello, G. Gerna, and T. K. Frey. 2003. Phylogenetic analysis of rubella virus isolated during a period of epidemic transmission in Italy, 1991-1997. J. Infect. Dis. 187:1587-1597.
Zheng, Q., Z. Zhu, W. Xu, and J. P. Icenogle. 2003. Sensitive RT-PCR to detect rubella virus RNA in clinical samples. Abstr. Am. Soc. Virol. Mtg., abstr. P34-1, p. 143.