Journal of Virological Methods, 39 (1992) 3946 0 1992 Elsevier Science Publishers B.V. / All rights reserved / 0166-0934/92/%05.00
39
VIRMET 01367
Enhanced detection of respiratory viruses using the shell vial technique and monoclonal antibodies Spencer Department
H.S. Lee, James E. Boutilier, Margaret and Kevin R. Forward
A. MacDonald
of Microbrology, Victoria General Hospital and Dalhousie University, Halifax. Nova Scotia (Canada)
(Accepted 21 February
1992)
Summary The shell vial technique using A549 and MDCK cells, coupled with the use of Bartels respiratory viral monoclonal antibodies, was evaluated initially for the detection of 28 previously isolated respiratory viruses. All viruses were recovered and correctly identified. The shell vial-monoclonal antibody technique was then evaluated for virus isolation from 338 respiratory specimens and compared with the conventional tube method. Both methods gave rise to a total of 83 virus isolates. Of these isolates, 68 (20.1%) were isolated and identified by the shell vial-monoclonal method; 60 (17.8%) were culture-positive by the conventional tube method; forty-five (13.3%) were positive by both methods. The shell vial-monoclonal antibody method yielded 12 isolates of influenza A, two isolates of parainfluenza type 3 and one each of parainfluenza types 1 and 3, which were missed by the conventional tube method, indicating the superior sensitivity and specificity of the shell vialmonoclonal antibody method (Chi-square analysis, P= 0.001) for the detection of these viruses. Of the 50 RSV isolates, 29 were detected by both methods and there were 21 discrepancies between the two methods. The shell vialmonoclonal antibody method also improved the turn-around time for the respiratory virus groups. Respiratory
virus; Monoclonal
antibody panel; Shell vial method
Correspondence to: S.H.S. Lee, Department of Microbiology, Victoria General Hospital, 1278 Tower Road, Halifax, Nova Scotia, Canada B3H 2Y9.
Introduction In recent years, the increasing availability of specific antibodies against respiratory viruses has enabled clinical laboratories to develop and/or implement rapid and cost-effective approaches for the diagnosis of respiratory viral infections. A number of investigators have reported the use of these antibodies to detect viral antigens directly in respiratory specimens, using enzyme immunoassay (Ahluwalia et al., 1987; Hughes et al., 1988; Swierkosz et al., 1989) or the immunofluorescence technique (Bell et al., 1983; Ray and Minnich, 1987; Ahluwalia et al., 1987) as a rapid alternative to conventional virus isolation. Significant progress has largely been made in the direct detection of respiratory syncytial virus (RSV) antigens (Ahluwalia et al., 1987; Waner et al., 1990). However, virus isolation is still desirable in many circumstances, especially when virus typing, serological and antiviral sensitivity testing with the virus isolate may be deemed necessary. Thus, some investigators (Halstead et al., 1990; Smith et al., 1991) have focussed on enhancing the virus isolation procedure by the use of the shell vial method, which was shown to be effective for the rapid detection of herpes viruses (Darougar et al., 1981; Gleaves et al., 1984; Gleaves et al., 1985). Widespread use of this method has been limited by the commercial unavailability of a pool of specific antibodies which would allow the detection of influenza types A and B, parainfluenza types 1, 2, and 3 RSV and adenovirus. An apparent solution to this limitation was recently provided by the availability of the Bartels Indirect Fluorescent Antibody Viral Respiratory Panel (Baxter Healthcare Co., Bellevue, WA) which is intended for the qualitative detection of these respiratory viral pathogens. This led us to undertake an evaluation of the panel using the shell vial method in a clinical laboratory setting. The evaluation was carried out in comparison with a routine virus isolation protocol using conventional tube cultures.
Materials and Methods Clinical specimens From November 1989 to March 1990, 338 respiratory specimens from 259 patients were cultured in parallel by the shell vial method and the conventional tube method. The specimens included 255 nasopharyngeal aspirates, 33 throat swabs, 12 tracheal secretions, five sputa, four pleural fluids, three bronchial washings, three biopsies of lung’, esophagus and lymph node and two lung aspirates. Specimens were collected from children less than 17 yr old at the Izaak Walton Killam Hospital for Children. In addition 21 throat washings from adults were collected by sentinel physicians monitoring influenza activity. All specimens were collected in viral transport medium, consisting of 0.5% gelatin in Hank’s balanced salt solution (HBSS) at pH 7.2.
41
Tissue culture HEp-2 cells (ATCC CCL 23) and Madin-Darby canine kidney (MDCK) cells (ATCC CCL 34) were grown at 37°C in minimum essential medium with Earle’s salts (MEME) (Gibco Laboratories, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Flow Laboratories, Mississauga, Ontario, Canada), 100 U of penicillin per ml and 50 pg of streptomycin per ml. Human A549 cells, obtained from Dr. F.T. Jay (Department of Medical Microbiology, University of Manitoba, Winnepeg, Manitoba) were grown under similar conditions, except using RPM1 1640 medium (Gibco Laboratories, Grand Island, NY). Primary rhesus monkey kidney (PRMK) cells in cell culture tubes were obtained weekly from Connaught Laboratories (Willowdale, Ontario, Canada). For virus culture, cell monolayers were prepared in conventional cell culture tubes and shell vials. When the monolayers were approximately 70% confluent (usually 24 h following cell seeding), the growth medium was replaced with maintenance medium. This consisted of MEME or RPM1 1640 supplemented with 2% FBS, 10 pg per ml of ampicillin, 6.25 ,ug per ml of netilmycin and 5 pg per ml of amphotericin B. Reisolation
of viruses
For initial assessment of specificity of the Bartels viral respiratory monoclonal antibody panel reagents, coded samples of 28 previously isolated viruses, including two influenza A viruses, eight parainfluenza type 1, one parainfluenza type 2, ten parainfluenza type 3, five adenovirus and two RSV, were used. These viruses had been stored frozen at - 70°C in their original cell culture maintenance medium. Each sample was inoculated in 0.2 ml volumes to a shell vial culture of MDCK and A549. Virus culture (i) Conventional tube method. Each specimen was vortexed and 0.5 ml portions were inoculated, respectively, into one tube of HEp-2 and two tubes of PRMK cells, each containing 1 ml of replenished maintenance medium. An inoculum adsorption step was eliminated in our laboratory previously as it did not produce additional isolates but increased toxicity and cell deterioration. After overnight incubation at 35°C in a stationary position the cultures were examined microscopically for toxicity and cytopathic effect. If the cell monolayer displayed no visible effect, the inoculum was replaced with fresh maintenance medium. Cultures showing toxicity were subcultured. Cultures were further incubated and examined daily (except Sundays) for cytopathic effect for 5-7 days. To detect the presence of myxovirus and paramyxovirus, one PRMK cell tube was hemadsorbed at day 7, using guinea pig erythrocytes (Woodlyn Laboratories, Guelph, Ontario). If hemadsorption was positive, the
42
PRMK cell tube was then frozen and thawed twice and the disrupted cell debris was centrifuged at 1500 rpm and the pellet examined by electron microscopy, using a Jeol 1200EX microscope, for the presence of viruses with characteristic myxovirus or paramyxovirus morphology. Influenza and parainfluenza viruses were typed by a direct fluorescent-antibody (DFA) test of the infected cells in the duplicate PRMK culture using the ImagenTM Influenza Virus A and B Test (API Laboratory Product Ltd., St. Laurent, Quebec, Canada) and monoclonal parainfluenza antibodies (Whittaker Bioproducts Inc., Walkersville, MD), respectively. RSV was identified by the formation of typical syncytial cells in HEp-2 culture and confirmed by a DFA test with Ortho RSV Identification Reagent (Ortho Diagnostic Systems Inc., Raritan, NJ). Adenovirus was identified on the basis of cytopathic effect in HEp-2 and PRMK cells and characteristic virus morphology by electron microscopy. Although reports were usually initiated after 7 days of culture, all negative cultures were frozen and thawed twice and subcultured to fresh cells and incubated for a further 7 days before the culture was confirmed negative. For specimen inoculation, maintenance medium of two (ii) Shell vial method. MDCK and two A549 shell vial cultures was aspirated from the monolayers and each vial was inoculated with 0.2 ml of the vortexed specimen. The shell vials were centrifuged using a Beckman Model J6-B centrifuge for 40 min at 750 x g in a Beckman Model JS 4-2 swinging bucket rotor at 3&33”C. After centrifugation, 1 ml of maintenance medium was added to each vial. The cultures were incubated at 35°C in a humidified CO2 incubator. Each culture was examined daily and cultures showing toxicity were frozen and thawed twice and subcultured into fresh SV cultures. On the third day the culture medium was removed from one A549 and one MDCK vial and the cells rinsed twice with 1 ml of fluorescent antibody testing (FAT) buffer consisting of Ca2+ and Mg2 + -free phosphate-buffered saline. The cells were detached by gentle scraping into 1 ml of FAT buffer and centrifuged at 200 x g for 10 min. The cell pellet was resuspended in 100 ~1 of FAT buffer and smears were prepared on lo-well acetone-methanol washed epoxy-coated slides (Cell Line Ass., Newfield, NJ). Uninoculated cell cultures served as negative controls and cultures infected with known viruses served as positive controls. The slides were air-dried and fixed in acetone for 10 min at room temperature (RT). Fixed slides were stained, usually on the same day, using Bartel’s monoclonal antibody panel. If the screening was found to be negative, the staining was repeated on the remaining shell vial on the 7th day of culture. No subculturing was carried out using this method. Immunojluorescence
staining using the Bartels reagents
The screening slides were stained with the antiviral monoclonal antibody pool, consisting of monoclonal antibodies directed against adenovirus, influenza A and B, parainfluenza types 1, 2 and 3 and RSV. In brief, each
43
fixed smear was overlaid with 10 ~1 of the monoclonal antibody pool and incubated in a humid chamber at 35-37°C for 30 min, washed for 5-10 min in FAT buffer and air-dried at RT. The smears were then overlaid with 10 ~1 of anti-mouse fluorescein-conjugated antibody containing Evan’s blue dye. Negative control smears received 10 ,~l of non-immune antibody as per manufacturer’s directions. The slides were reincubated at 35-37°C for 30 min, washed in FAT buffer and air-dried. The slides were then mounted with buffered glycerol medium and examined under a fluorescent microscope at 400 x g magnification. If either smear was found to be positive a slide with seven additional smears was prepared from the original pellet and stained with the individual antiviral monoclonal antibodies. To insure reagent reactivity, one slide positive for each of the seven respiratory viruses included with the kit was stained prior to the routine use of the kit, as well as at fixed intervals.
Results A preliminary study was conducted to ascertain the efficiency and the specificity of the Bartel’s monoclonal antibody panel. Twenty-eight known respiratory virus isolates were included in the evaluation. These isolates were coded and recultured using the shell vial method. After three days incubation all smears were screened positive with the Bartels monoclonal antibody panel and all were correctly identified in a blinded fashion with the individual monoclonal antibodies. Table 1 shows the number and the types of respiratory viruses isolated from the 338 specimens by the shell vial-monoclonal antibody and the conventional tube methods. A total of 83 viruses were isolated. Sixty-eight (20.1%) specimens yielded respiratory viruses by the shell vial-monoclonal antibody method and 60 (17.8%) were culture positive by the conventional tube method. Forty-five (13.3%) were positive by both methods. There were differences in sensitivity between the two methods. The shell vial-monoclonal antibody method yielded twelve isolates of influenza A, two isolates of parainfluenza TABLE 1 Culture results of conventional
tube and shell vial-monoclonal
antibody
Virus
No. of results
Conventional tube method Shell vial-monoclonal antibody method
Positive Positive
Positive Negative
Adenovirus Parainfluenza Parainfluenza Parainfluenza RSV Influenza A All
2 1 3 1 29 9 45
1 0 0 0 14 0 15
type 1 type 2 type 3
methods
Negative Positive 0 1 : 7 12 23
Total Positive 3 2 4 3 50 21 83
44 TABLE 2 Average times for virus culture by shell vial-monoclonal Virus
Adenovirus Parainfluenza type 1 Parainfluenza type 2 Parainfluenza type 3 RSV Influenza A Influenza Bb d Values in parentheses ’ Mock specimens.
antibody
and conventional
tube methods
Days Shell vial-monoclonal antibody method
Conventional tube method
5 4 4 4 3 4 4
10 ( 3) 7( 1) 7 ( 3) 7 ( 1) 5 (43) 7 ( 9) 7 ( 2)
( 2)” ( 2) ( 4) ( 3) (36) (21) ( 2)
are the number of virus isolates.
type 3 and one each of parainfluenza types 1 and 2, which were missed by the conventional tube method. One adenovirus isolate was recovered only by the conventional tube method but after 14 days of incubation. In all there were fifty RSV isolates. Twenty-eight were detected by both methods and there were 21 discrepancies between the two methods. Of the 14 RSV isolates detected by the conventional tube method and negative by the shell vial-monoclonal antibody method, five were positive at two days, two at seven days and seven were positive only after subculture. The seven RSV isolates detected by the shell vial-monoclonal antibody method and negative by the conventional tube method were all positive at three days incubation. The average time required for isolation and identification of each of the respiratory viruses as well as the total number of isolates by the shell vialmonoclonal antibody and the conventional tube methods are shown in Table 2. The turn-around time was substantially improved by the shell vial-monoclonal antibody method.
Discussion In the present study we explored the sensitivity and specificity of the monoclonal antibody panel from Bartels to screen for and to identify viruses in shell vial cultures of A549 and MDCK cells. and to compare this method with a routine conventional tube method. The results of the study show that the shell vial-monoclonal antibody method was more sensitive than the conventional tube method (Chi-square analysis, P < 0.001) for the isolation of influenza A and parainfluenza viruses, even though a relatively smaller inoculum was used. It would appear that this increased sensitivity of the shell vial-monoclonal antibody method is not attributable to the differing susceptibility of MDCK, as compared with PRMK cells, since previous investigators have shown that they are of equal (Frank et
45
al., 1979) or lesser (Waris et al., 1990) susceptibility to influenza viruses. The increased sensitivity of the shell vial method for parainfluenzaviruses is surprising in light of the relative superiority of PRMK over MDCK cells (Frank et al., 1979). It is more likely that the increased positive rate is attributable to the centrifuged inoculation used by the shell vial-monoclonal antibody method. Other investigators have reported the enhancement of cytomegalovirus (Gleaves et al., 1984) herpes simplex virus (Darougar et al., 1981; Gleaves et al., 1985), influenza A (Waris et al., 1990) and RSV (Halstead et al., 1990) detection achieved by initial centrifugation of the inoculum. It is of interest to note that the influenza A isolates were from specimens sent in by the sentinel physicians monitoring influenza activity. Their recovery coincided with influenza activity in the community, as evidenced by increased school absenteeism, a number of sero-conversions in non-study patients and the isolation of influenza A viruses from specimens not processed as part of this study. Two of the isolates were passaged once in 8-day-old hen’s embryonated eggs by the allantoic route and subsequently identified by hemagglutinationinhibition test as Shanghai-like H3N2 strains of influenza A. There was no influenza B virus isolated in the study. However, when two mock throat washings, containing influenza B, were processed in a blinded fashion, both the shell vial-monoclonal antibody method and the conventional tube method detected the virus in three days and seven days, respectively. The shell vialmonoclonal antibody method missed only one adenovirus, which was detected by the conventional tube method only after 14 days of culture. This may be attributed to the relatively low concentration of virus in the specimen. Fifty RSV isolates were recovered. Both the shell vial-monoclonal antibody and conventional tube methods detected 29 isolates. Between them, there were 21 discrepancies, including 14 culture-positive by the conventional tube method and seven by the shell vial-monoclonal antibody method. The reason for the discrepancies was not apparent. Of the seven positive, by the shell vialmonoclonal antibody method, all became positive in 3 days and none of these seven was positive by the conventional method, even following a blind subculture. Of the 14 specimens cultured positive for RSV by the conventional tube method and missed by the shell vial-monoclonal antibody method, five were positive at two days. However, two of these were refrigerated ovLernight before being cultured by the shell vial-monoclonal antibody method. Two were positive at 7 days and seven became positive only after the 7-day subculture suggesting a very low inoculum especially in the latter specimens. The average time in days for the isolation and identification of the respiratory viral agents from clinical specimens by the shell vial-monoclonal antibody method was significantly reduced when compared with that by the conventional tube method (Chi-square analysis, P = 0.001). This, coupled with the ease and convenience of the shell vial-monoclonal antibody technology, as well as the sensitivity and specificity of Bartels monoclonal antibody panel, should facilitate its widespread use. In spite of the few RSV discrepancies we have instituted the shell vial-monoclonal antibody method into our routine
46
diagnostic setting, emphasizing prompt inoculation of specimens, The costs incurred by this change were minimal and can be further reduced by only using MDCK cells during the influenza season.
Acknowledgments We are indebted to the staff of the microbiology laboratory at the IWK Hospital for Children for their assistance in the study. We gratefully acknowledge the excellent technical assistance of Deborah Anthony, Carolyn Baker and Sherry Baker. We also acknowledge Baxter/Bartels Diagnostic Division for supplying the monoclonal antibody kits under evaluation. References Ahluwalia, G., Embree, J., McNicol, P., Law, B. and Hammond, G. (1987) Comparison of nasopharyngeal aspirate and nasopharyngeal swab specimens for respiratory syncytial virus diagnosis by cell culture, indirect immunofluorescence assay and enzyme-linked immunosorbent assay. J. Clin. Microbial. 25, 763-767. Bell, D.W., Walsh, E.E., Hruska, J.F., Schnabel, K.C. and Hall, C.B. (1983) Rapid detection of respiratory syncytial virus with a monoclonal antibody. J. Clin. Microbial. 17, 1099-l 101. Darougar, S., Gibson, J.M. and Thacker, U. (1981) Effect of centrifugation on herpes simplex virus isolation. J. Med. Virol. 8, 231-235. Frank, A.L., Couch, R.B., Griffis, CA. and Baxter, B.D. (1979) Comparison of different tissue cultures for isolation and quantitation of influenza and parainfluenza viruses. J. Chn. Microbial. 10, 32-36. Gleaves, CA., Smith, T.F., Shuster, E.A. and Pearson, G.R. (1984) Rapid detection of cytomegalovirus in MRC-5 cells inoculated with urine specimens by using low-speed centrifugation and monoclonal antibody to an early antigen. J. Clin. Microbial. 19, 917-919. Gleaves, C.A., Wilson, D.J., Wold, A.D. and Smith, T.F. (1985) Detection and serotyping of herpes simplex virus in MRC-5 cells by use of centrifugation and monoclonal antibodies 16 h postinoculation. J. Clin. Microbial. 21, 29-32. Halstead, D.C., Todd, S. and Fritch, G. (1990) Evaluation of five methods for respiratory syncytial virus detection. J. Clin. Microbial. 28, 1021-1025. Hughes, J.H., Mann, D.R. and Hamparian, V.V. (1988) Detection of respiratory syncytial virus in clinical specimens by viral culture, direct and indirect immunofluorescence, and enzyme immunoassay. J. Clin. Microbial. 26, 588-591. Ray, C.G. and Minnich, L.L. (1987) Efftciency of immunofluorescence for rapid detection of common respiratory viruses. J. Clin. Microbial. 25, 355-357. Smith, M.C., Creutz, C. and Huang, Y.T. (1991) Detection of respiratory syncytial virus in nasopharyngeal secretions by shell vial technique. J. Clin. Microbial. 29, 463465. Swierkosz, E.M., Flanders, R., Melvin, L., Miller, J.D. and Kline, M.W. (1989) Evaluation of the Abbott TESTPACK RSV enzyme immunoassay for detection of respiratory syncytial virus in nasopharyngeal swab specimens. J. Chn. Microbial. 27, 1151-l 154. Waner, J.L., Whitehurst, N.J., Todd, S.J., Shalaby, H. and Wall, (1990) Comparison of Directigen RSV with viral isolation and direct immunofluorescence for the identification of respiratory syncytial virus. J. Clin. Microbial. 28, 480-483. Waris, M., Ziegler, T., Kivivirta, M. and Ruuskanen, 0. (1990) Rapid detection of respiratory syncytial virus and influenza A virus in cell cultures by immunoperoxidase staining with monoclonal antibodies. J. Clin. Microbial. 28, 1159-l 162.
39
VIRMET 01367
Enhanced detection of respiratory viruses using the shell vial technique and monoclonal antibodies Spencer Department
H.S. Lee, James E. Boutilier, Margaret and Kevin R. Forward
A. MacDonald
of Microbrology, Victoria General Hospital and Dalhousie University, Halifax. Nova Scotia (Canada)
(Accepted 21 February
1992)
Summary The shell vial technique using A549 and MDCK cells, coupled with the use of Bartels respiratory viral monoclonal antibodies, was evaluated initially for the detection of 28 previously isolated respiratory viruses. All viruses were recovered and correctly identified. The shell vial-monoclonal antibody technique was then evaluated for virus isolation from 338 respiratory specimens and compared with the conventional tube method. Both methods gave rise to a total of 83 virus isolates. Of these isolates, 68 (20.1%) were isolated and identified by the shell vial-monoclonal method; 60 (17.8%) were culture-positive by the conventional tube method; forty-five (13.3%) were positive by both methods. The shell vial-monoclonal antibody method yielded 12 isolates of influenza A, two isolates of parainfluenza type 3 and one each of parainfluenza types 1 and 3, which were missed by the conventional tube method, indicating the superior sensitivity and specificity of the shell vialmonoclonal antibody method (Chi-square analysis, P= 0.001) for the detection of these viruses. Of the 50 RSV isolates, 29 were detected by both methods and there were 21 discrepancies between the two methods. The shell vialmonoclonal antibody method also improved the turn-around time for the respiratory virus groups. Respiratory
virus; Monoclonal
antibody panel; Shell vial method
Correspondence to: S.H.S. Lee, Department of Microbiology, Victoria General Hospital, 1278 Tower Road, Halifax, Nova Scotia, Canada B3H 2Y9.
Introduction In recent years, the increasing availability of specific antibodies against respiratory viruses has enabled clinical laboratories to develop and/or implement rapid and cost-effective approaches for the diagnosis of respiratory viral infections. A number of investigators have reported the use of these antibodies to detect viral antigens directly in respiratory specimens, using enzyme immunoassay (Ahluwalia et al., 1987; Hughes et al., 1988; Swierkosz et al., 1989) or the immunofluorescence technique (Bell et al., 1983; Ray and Minnich, 1987; Ahluwalia et al., 1987) as a rapid alternative to conventional virus isolation. Significant progress has largely been made in the direct detection of respiratory syncytial virus (RSV) antigens (Ahluwalia et al., 1987; Waner et al., 1990). However, virus isolation is still desirable in many circumstances, especially when virus typing, serological and antiviral sensitivity testing with the virus isolate may be deemed necessary. Thus, some investigators (Halstead et al., 1990; Smith et al., 1991) have focussed on enhancing the virus isolation procedure by the use of the shell vial method, which was shown to be effective for the rapid detection of herpes viruses (Darougar et al., 1981; Gleaves et al., 1984; Gleaves et al., 1985). Widespread use of this method has been limited by the commercial unavailability of a pool of specific antibodies which would allow the detection of influenza types A and B, parainfluenza types 1, 2, and 3 RSV and adenovirus. An apparent solution to this limitation was recently provided by the availability of the Bartels Indirect Fluorescent Antibody Viral Respiratory Panel (Baxter Healthcare Co., Bellevue, WA) which is intended for the qualitative detection of these respiratory viral pathogens. This led us to undertake an evaluation of the panel using the shell vial method in a clinical laboratory setting. The evaluation was carried out in comparison with a routine virus isolation protocol using conventional tube cultures.
Materials and Methods Clinical specimens From November 1989 to March 1990, 338 respiratory specimens from 259 patients were cultured in parallel by the shell vial method and the conventional tube method. The specimens included 255 nasopharyngeal aspirates, 33 throat swabs, 12 tracheal secretions, five sputa, four pleural fluids, three bronchial washings, three biopsies of lung’, esophagus and lymph node and two lung aspirates. Specimens were collected from children less than 17 yr old at the Izaak Walton Killam Hospital for Children. In addition 21 throat washings from adults were collected by sentinel physicians monitoring influenza activity. All specimens were collected in viral transport medium, consisting of 0.5% gelatin in Hank’s balanced salt solution (HBSS) at pH 7.2.
41
Tissue culture HEp-2 cells (ATCC CCL 23) and Madin-Darby canine kidney (MDCK) cells (ATCC CCL 34) were grown at 37°C in minimum essential medium with Earle’s salts (MEME) (Gibco Laboratories, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Flow Laboratories, Mississauga, Ontario, Canada), 100 U of penicillin per ml and 50 pg of streptomycin per ml. Human A549 cells, obtained from Dr. F.T. Jay (Department of Medical Microbiology, University of Manitoba, Winnepeg, Manitoba) were grown under similar conditions, except using RPM1 1640 medium (Gibco Laboratories, Grand Island, NY). Primary rhesus monkey kidney (PRMK) cells in cell culture tubes were obtained weekly from Connaught Laboratories (Willowdale, Ontario, Canada). For virus culture, cell monolayers were prepared in conventional cell culture tubes and shell vials. When the monolayers were approximately 70% confluent (usually 24 h following cell seeding), the growth medium was replaced with maintenance medium. This consisted of MEME or RPM1 1640 supplemented with 2% FBS, 10 pg per ml of ampicillin, 6.25 ,ug per ml of netilmycin and 5 pg per ml of amphotericin B. Reisolation
of viruses
For initial assessment of specificity of the Bartels viral respiratory monoclonal antibody panel reagents, coded samples of 28 previously isolated viruses, including two influenza A viruses, eight parainfluenza type 1, one parainfluenza type 2, ten parainfluenza type 3, five adenovirus and two RSV, were used. These viruses had been stored frozen at - 70°C in their original cell culture maintenance medium. Each sample was inoculated in 0.2 ml volumes to a shell vial culture of MDCK and A549. Virus culture (i) Conventional tube method. Each specimen was vortexed and 0.5 ml portions were inoculated, respectively, into one tube of HEp-2 and two tubes of PRMK cells, each containing 1 ml of replenished maintenance medium. An inoculum adsorption step was eliminated in our laboratory previously as it did not produce additional isolates but increased toxicity and cell deterioration. After overnight incubation at 35°C in a stationary position the cultures were examined microscopically for toxicity and cytopathic effect. If the cell monolayer displayed no visible effect, the inoculum was replaced with fresh maintenance medium. Cultures showing toxicity were subcultured. Cultures were further incubated and examined daily (except Sundays) for cytopathic effect for 5-7 days. To detect the presence of myxovirus and paramyxovirus, one PRMK cell tube was hemadsorbed at day 7, using guinea pig erythrocytes (Woodlyn Laboratories, Guelph, Ontario). If hemadsorption was positive, the
42
PRMK cell tube was then frozen and thawed twice and the disrupted cell debris was centrifuged at 1500 rpm and the pellet examined by electron microscopy, using a Jeol 1200EX microscope, for the presence of viruses with characteristic myxovirus or paramyxovirus morphology. Influenza and parainfluenza viruses were typed by a direct fluorescent-antibody (DFA) test of the infected cells in the duplicate PRMK culture using the ImagenTM Influenza Virus A and B Test (API Laboratory Product Ltd., St. Laurent, Quebec, Canada) and monoclonal parainfluenza antibodies (Whittaker Bioproducts Inc., Walkersville, MD), respectively. RSV was identified by the formation of typical syncytial cells in HEp-2 culture and confirmed by a DFA test with Ortho RSV Identification Reagent (Ortho Diagnostic Systems Inc., Raritan, NJ). Adenovirus was identified on the basis of cytopathic effect in HEp-2 and PRMK cells and characteristic virus morphology by electron microscopy. Although reports were usually initiated after 7 days of culture, all negative cultures were frozen and thawed twice and subcultured to fresh cells and incubated for a further 7 days before the culture was confirmed negative. For specimen inoculation, maintenance medium of two (ii) Shell vial method. MDCK and two A549 shell vial cultures was aspirated from the monolayers and each vial was inoculated with 0.2 ml of the vortexed specimen. The shell vials were centrifuged using a Beckman Model J6-B centrifuge for 40 min at 750 x g in a Beckman Model JS 4-2 swinging bucket rotor at 3&33”C. After centrifugation, 1 ml of maintenance medium was added to each vial. The cultures were incubated at 35°C in a humidified CO2 incubator. Each culture was examined daily and cultures showing toxicity were frozen and thawed twice and subcultured into fresh SV cultures. On the third day the culture medium was removed from one A549 and one MDCK vial and the cells rinsed twice with 1 ml of fluorescent antibody testing (FAT) buffer consisting of Ca2+ and Mg2 + -free phosphate-buffered saline. The cells were detached by gentle scraping into 1 ml of FAT buffer and centrifuged at 200 x g for 10 min. The cell pellet was resuspended in 100 ~1 of FAT buffer and smears were prepared on lo-well acetone-methanol washed epoxy-coated slides (Cell Line Ass., Newfield, NJ). Uninoculated cell cultures served as negative controls and cultures infected with known viruses served as positive controls. The slides were air-dried and fixed in acetone for 10 min at room temperature (RT). Fixed slides were stained, usually on the same day, using Bartel’s monoclonal antibody panel. If the screening was found to be negative, the staining was repeated on the remaining shell vial on the 7th day of culture. No subculturing was carried out using this method. Immunojluorescence
staining using the Bartels reagents
The screening slides were stained with the antiviral monoclonal antibody pool, consisting of monoclonal antibodies directed against adenovirus, influenza A and B, parainfluenza types 1, 2 and 3 and RSV. In brief, each
43
fixed smear was overlaid with 10 ~1 of the monoclonal antibody pool and incubated in a humid chamber at 35-37°C for 30 min, washed for 5-10 min in FAT buffer and air-dried at RT. The smears were then overlaid with 10 ~1 of anti-mouse fluorescein-conjugated antibody containing Evan’s blue dye. Negative control smears received 10 ,~l of non-immune antibody as per manufacturer’s directions. The slides were reincubated at 35-37°C for 30 min, washed in FAT buffer and air-dried. The slides were then mounted with buffered glycerol medium and examined under a fluorescent microscope at 400 x g magnification. If either smear was found to be positive a slide with seven additional smears was prepared from the original pellet and stained with the individual antiviral monoclonal antibodies. To insure reagent reactivity, one slide positive for each of the seven respiratory viruses included with the kit was stained prior to the routine use of the kit, as well as at fixed intervals.
Results A preliminary study was conducted to ascertain the efficiency and the specificity of the Bartel’s monoclonal antibody panel. Twenty-eight known respiratory virus isolates were included in the evaluation. These isolates were coded and recultured using the shell vial method. After three days incubation all smears were screened positive with the Bartels monoclonal antibody panel and all were correctly identified in a blinded fashion with the individual monoclonal antibodies. Table 1 shows the number and the types of respiratory viruses isolated from the 338 specimens by the shell vial-monoclonal antibody and the conventional tube methods. A total of 83 viruses were isolated. Sixty-eight (20.1%) specimens yielded respiratory viruses by the shell vial-monoclonal antibody method and 60 (17.8%) were culture positive by the conventional tube method. Forty-five (13.3%) were positive by both methods. There were differences in sensitivity between the two methods. The shell vial-monoclonal antibody method yielded twelve isolates of influenza A, two isolates of parainfluenza TABLE 1 Culture results of conventional
tube and shell vial-monoclonal
antibody
Virus
No. of results
Conventional tube method Shell vial-monoclonal antibody method
Positive Positive
Positive Negative
Adenovirus Parainfluenza Parainfluenza Parainfluenza RSV Influenza A All
2 1 3 1 29 9 45
1 0 0 0 14 0 15
type 1 type 2 type 3
methods
Negative Positive 0 1 : 7 12 23
Total Positive 3 2 4 3 50 21 83
44 TABLE 2 Average times for virus culture by shell vial-monoclonal Virus
Adenovirus Parainfluenza type 1 Parainfluenza type 2 Parainfluenza type 3 RSV Influenza A Influenza Bb d Values in parentheses ’ Mock specimens.
antibody
and conventional
tube methods
Days Shell vial-monoclonal antibody method
Conventional tube method
5 4 4 4 3 4 4
10 ( 3) 7( 1) 7 ( 3) 7 ( 1) 5 (43) 7 ( 9) 7 ( 2)
( 2)” ( 2) ( 4) ( 3) (36) (21) ( 2)
are the number of virus isolates.
type 3 and one each of parainfluenza types 1 and 2, which were missed by the conventional tube method. One adenovirus isolate was recovered only by the conventional tube method but after 14 days of incubation. In all there were fifty RSV isolates. Twenty-eight were detected by both methods and there were 21 discrepancies between the two methods. Of the 14 RSV isolates detected by the conventional tube method and negative by the shell vial-monoclonal antibody method, five were positive at two days, two at seven days and seven were positive only after subculture. The seven RSV isolates detected by the shell vial-monoclonal antibody method and negative by the conventional tube method were all positive at three days incubation. The average time required for isolation and identification of each of the respiratory viruses as well as the total number of isolates by the shell vialmonoclonal antibody and the conventional tube methods are shown in Table 2. The turn-around time was substantially improved by the shell vial-monoclonal antibody method.
Discussion In the present study we explored the sensitivity and specificity of the monoclonal antibody panel from Bartels to screen for and to identify viruses in shell vial cultures of A549 and MDCK cells. and to compare this method with a routine conventional tube method. The results of the study show that the shell vial-monoclonal antibody method was more sensitive than the conventional tube method (Chi-square analysis, P < 0.001) for the isolation of influenza A and parainfluenza viruses, even though a relatively smaller inoculum was used. It would appear that this increased sensitivity of the shell vial-monoclonal antibody method is not attributable to the differing susceptibility of MDCK, as compared with PRMK cells, since previous investigators have shown that they are of equal (Frank et
45
al., 1979) or lesser (Waris et al., 1990) susceptibility to influenza viruses. The increased sensitivity of the shell vial method for parainfluenzaviruses is surprising in light of the relative superiority of PRMK over MDCK cells (Frank et al., 1979). It is more likely that the increased positive rate is attributable to the centrifuged inoculation used by the shell vial-monoclonal antibody method. Other investigators have reported the enhancement of cytomegalovirus (Gleaves et al., 1984) herpes simplex virus (Darougar et al., 1981; Gleaves et al., 1985), influenza A (Waris et al., 1990) and RSV (Halstead et al., 1990) detection achieved by initial centrifugation of the inoculum. It is of interest to note that the influenza A isolates were from specimens sent in by the sentinel physicians monitoring influenza activity. Their recovery coincided with influenza activity in the community, as evidenced by increased school absenteeism, a number of sero-conversions in non-study patients and the isolation of influenza A viruses from specimens not processed as part of this study. Two of the isolates were passaged once in 8-day-old hen’s embryonated eggs by the allantoic route and subsequently identified by hemagglutinationinhibition test as Shanghai-like H3N2 strains of influenza A. There was no influenza B virus isolated in the study. However, when two mock throat washings, containing influenza B, were processed in a blinded fashion, both the shell vial-monoclonal antibody method and the conventional tube method detected the virus in three days and seven days, respectively. The shell vialmonoclonal antibody method missed only one adenovirus, which was detected by the conventional tube method only after 14 days of culture. This may be attributed to the relatively low concentration of virus in the specimen. Fifty RSV isolates were recovered. Both the shell vial-monoclonal antibody and conventional tube methods detected 29 isolates. Between them, there were 21 discrepancies, including 14 culture-positive by the conventional tube method and seven by the shell vial-monoclonal antibody method. The reason for the discrepancies was not apparent. Of the seven positive, by the shell vialmonoclonal antibody method, all became positive in 3 days and none of these seven was positive by the conventional method, even following a blind subculture. Of the 14 specimens cultured positive for RSV by the conventional tube method and missed by the shell vial-monoclonal antibody method, five were positive at two days. However, two of these were refrigerated ovLernight before being cultured by the shell vial-monoclonal antibody method. Two were positive at 7 days and seven became positive only after the 7-day subculture suggesting a very low inoculum especially in the latter specimens. The average time in days for the isolation and identification of the respiratory viral agents from clinical specimens by the shell vial-monoclonal antibody method was significantly reduced when compared with that by the conventional tube method (Chi-square analysis, P = 0.001). This, coupled with the ease and convenience of the shell vial-monoclonal antibody technology, as well as the sensitivity and specificity of Bartels monoclonal antibody panel, should facilitate its widespread use. In spite of the few RSV discrepancies we have instituted the shell vial-monoclonal antibody method into our routine
46
diagnostic setting, emphasizing prompt inoculation of specimens, The costs incurred by this change were minimal and can be further reduced by only using MDCK cells during the influenza season.
Acknowledgments We are indebted to the staff of the microbiology laboratory at the IWK Hospital for Children for their assistance in the study. We gratefully acknowledge the excellent technical assistance of Deborah Anthony, Carolyn Baker and Sherry Baker. We also acknowledge Baxter/Bartels Diagnostic Division for supplying the monoclonal antibody kits under evaluation. References Ahluwalia, G., Embree, J., McNicol, P., Law, B. and Hammond, G. (1987) Comparison of nasopharyngeal aspirate and nasopharyngeal swab specimens for respiratory syncytial virus diagnosis by cell culture, indirect immunofluorescence assay and enzyme-linked immunosorbent assay. J. Clin. Microbial. 25, 763-767. Bell, D.W., Walsh, E.E., Hruska, J.F., Schnabel, K.C. and Hall, C.B. (1983) Rapid detection of respiratory syncytial virus with a monoclonal antibody. J. Clin. Microbial. 17, 1099-l 101. Darougar, S., Gibson, J.M. and Thacker, U. (1981) Effect of centrifugation on herpes simplex virus isolation. J. Med. Virol. 8, 231-235. Frank, A.L., Couch, R.B., Griffis, CA. and Baxter, B.D. (1979) Comparison of different tissue cultures for isolation and quantitation of influenza and parainfluenza viruses. J. Chn. Microbial. 10, 32-36. Gleaves, CA., Smith, T.F., Shuster, E.A. and Pearson, G.R. (1984) Rapid detection of cytomegalovirus in MRC-5 cells inoculated with urine specimens by using low-speed centrifugation and monoclonal antibody to an early antigen. J. Clin. Microbial. 19, 917-919. Gleaves, C.A., Wilson, D.J., Wold, A.D. and Smith, T.F. (1985) Detection and serotyping of herpes simplex virus in MRC-5 cells by use of centrifugation and monoclonal antibodies 16 h postinoculation. J. Clin. Microbial. 21, 29-32. Halstead, D.C., Todd, S. and Fritch, G. (1990) Evaluation of five methods for respiratory syncytial virus detection. J. Clin. Microbial. 28, 1021-1025. Hughes, J.H., Mann, D.R. and Hamparian, V.V. (1988) Detection of respiratory syncytial virus in clinical specimens by viral culture, direct and indirect immunofluorescence, and enzyme immunoassay. J. Clin. Microbial. 26, 588-591. Ray, C.G. and Minnich, L.L. (1987) Efftciency of immunofluorescence for rapid detection of common respiratory viruses. J. Clin. Microbial. 25, 355-357. Smith, M.C., Creutz, C. and Huang, Y.T. (1991) Detection of respiratory syncytial virus in nasopharyngeal secretions by shell vial technique. J. Clin. Microbial. 29, 463465. Swierkosz, E.M., Flanders, R., Melvin, L., Miller, J.D. and Kline, M.W. (1989) Evaluation of the Abbott TESTPACK RSV enzyme immunoassay for detection of respiratory syncytial virus in nasopharyngeal swab specimens. J. Chn. Microbial. 27, 1151-l 154. Waner, J.L., Whitehurst, N.J., Todd, S.J., Shalaby, H. and Wall, (1990) Comparison of Directigen RSV with viral isolation and direct immunofluorescence for the identification of respiratory syncytial virus. J. Clin. Microbial. 28, 480-483. Waris, M., Ziegler, T., Kivivirta, M. and Ruuskanen, 0. (1990) Rapid detection of respiratory syncytial virus and influenza A virus in cell cultures by immunoperoxidase staining with monoclonal antibodies. J. Clin. Microbial. 28, 1159-l 162.
