Vol. 1, No. 3
JOURNAL OF CLINICAL MICROBIOLOGY Mar. 1975, p. 324-329 Copyright © 1975 American Society for Microbiology
Printed in U.S.A.
Assessment of Virus Infectivity by the Immunofluorescent and Immunoperoxidase Techniques NICHOLAS HAHON,* JANET SIMPSON, AND HERBERT L. ECKERT Appalachian Laboratory for Occupational Respiratory Diseases, U.S. Public Health Service,* and Departments of Pediatrics and Medicine, West Virginia University, School of Medicine, Morgantown, West Virginia 26505 Received for publication 20 November 1974
The immunofluorescent and immunoperoxidase cell-counting techniques were comparable in precision and reproducibility for the quantitative assessment of influenza virus inf'ectivity. The dose-response function was linear with each procedure, and comparable results were obtained for estimating neutralizing antibodies in antiviral serum.
Quantitative assays of' virus infectivity by enumeration of cells containing immunofluorescent viral antigens have been established for agents representative of' almost all major animal groups (5). These assays are dependent on a single cycle of infection and possess such outstanding attributes as rapidity (usually less than 24 h), specif'icity, reproducibility, precision, and high sensitivity. The versatility of the immunofluorescent cell-counting procedure has been demonstrated in fundamental and applied studies, namely, the kinetics of' virus-cell interactions (8, 12-14) and of' virus neutralization (9, 10), differentiation of' serologically similar viruses (4), and assessments of' interferon (11, 18) and virustatic compounds (25). Although the technique has been used widely in virology, circumstances that sometimes restrict its effectiveness are naturally occurring autof'luorescence, fading of stained cells during examination, and the limited resolution of' ultraviolet microscopy. In recent years, the immunoenzymatic technique, the coupling of' enzymes to immunoglobulins without significant alteration of their respective activities (2, 24), has been successfully employed to identif'y a wide variety of viral antigens in cell cultures (3, 6, 7, 16, 17, 19, 20, 22, 27-31). The technique embodies the advantages of' the immunofluorescence procedure without some of' its limitations. Comparable sensitivity in detecting viral antigens has been reported between the immunoenzymatic and immunofluorescence procedures (17. 32). The advantages of the former have been broadened by such considerations as the relative ease of' conjugate preparation, the reduction of nonspecific reactions, the stability of' the stain, and the enhanced visualization of' viral antigens under
light and electron microscopy. For assaying virus inf'ectivity, however, the immunoenzymatic technique has not been extensively exploited. This study was undertaken to evaluate quantitatively the assessment of virus infectivity by the immunofluorescent (IF) and immunoperoxidase (IP) cell-counting techniques. MATERIALS AND METHODS Virus. The B/Great Lakes/1739/54 influenza virus strain emploved in this study was obtained from the American Type Culture Collection, Rockville, Md. Stock pools of' virus, consisting of infective allantoic fluids, were prepared and stored in the manner described previouslV (12). Cell cultures. Clone 1-5c-4 derived from a variant line of' Chang's conjunctival cell (33) was propagated with Eagle minimum essential medium containing 10%c fetal calf serum. Cells were maintained with minimal essential medium plus 5% fetal calf' serum. For virus assay, cells were cultivated on circular cover slips (15 mm in diameter) inserted in flat-bottomed glass vials (19 by 65 mm). One milliliter of' cell suspension containing approximately 105 cells was introduced onto each cover slip which was then incubated at 35 C for 24 h or until a complete cell monolayer was formed. Virus antiserum. Initially, rabbits were injected intramuscularly in a 2.0-ml volume with an equal quantity of' virus suspension (1.0 x 108 cell-infecting units/ml) and Freund complete adjuvant. At weekly intervals thereafter, animals were twice injected intravenously with 1.5 ml of virus suspension; they were exsanguinated 10 days after the last injection. The pooled antiserum was stored at -20 C. IF conjugation and staining. The globulin fraction of' antiserum was precipitated with ammonium sulfate at 4 C (15) and conjugated with fluorescein isothiocyanate (Nutritional Biochemical Corp., Cleveland, Ohio) at the ratio of 0.02 mg of dye per mg of protein (23). Unbound dye was removed by passing
324
VOL. 1, 1975
VIRUS INFECTIVITY ASSESSMENT TECHNIQUES
325
conjugated globulin through a Sephadex G-50 column a 15-mm diameter cover slip was 4,773. To determine (Pharmacia Fine Chemicals Inc., Piscataway, N.J.). this number, the diameter of the microscopic field was To reduce nonspecific fluorescence, 5 ml of conjugated measured with a stage micrometer, and the calculated globulin was diluted with an equal volume of 0.01 M area was divided into the area (176.62 mm) of the phosphate-buffered saline (PBS), pH 7.1, and ad- cover slip. For each cover slip cell monolayer, 50 sorbed twice with 200 mg of acetone-dried mouse microscopic fields were examined. To calculate the liver powder (23). Conjugated globulin was stored in number of cell-infecting units of virus per milliliter, the average number of IF- or IP-positive cells per field 1-ml portions at -20 C. The direct method was used to demonstrate IF of was multiplied by the number of fields per cover slip. viral antigens in infected cells. Fixed cell cultures the reciprocal of the dilution of virus inoculum. and were washed with PBS and stained with conjugated the volume factor. antiserum for 20 min at room temperature (22 C). Cover slip cell monolayers were then rinsed in two RESULTS changes of PBS and mounted on slides with a Quantitative evaluations of the assays. semi-permanent medium (26). The sensitivity of IF and IP staining for detectIP conjugation and staining. Portions of the remaining globulin fraction of virus antiserum used ing intracellular viral antigens was determined for IF conjugation were labeled with horseradish by following their rate of appearance in infected peroxidase as described by Kurstak (19). Briefly, 12 cell monolayers during a 24-h incubation pemg of horseradish peroxidase (RZ 3.0, Type VI, Sigma riod. Infected cells were first detected 8 h after Chemical Co., St. Louis, Mo.) was dissolved in 0.5 ml infection by either staining procedure; however, of 0.1 M phosphate buffer solution, pH 6.8, and mixed the number of infected cells noted in IF-stained with 0.5 ml (approximately 5 mg) of globulin prepara- cell monolayers was three times higher (Fig. 1). tion. While the solution was gently stirred. 0.05 ml of a 1% aqueous solution of glutaraldehyde was added With each staining technique, maximal and dropwise. After 2 h at room temperature, the mixture equivalent numbers of infected cells appeared was dialyzed against two changes of PBS at 4 C at 16 h after infection (Figs. 2 and 3). Thereovernight and then clarified by centrifugation. Peroxi- after, the number was constant throughout the dase conjugates were used without further purifica- remainder of the incubation period. tion. Before use, the stock solution was diluted 1:10. The dose-response was linear between twofold For staining of cover slip cell cultures by the direct dilutions of virus over a range of 1.2 log1O units IP technique, fixed cell cultures were rinsed with and numbers of infected cells by using the IF- or PBS, covered with peroxidase-labeled antibody prep- IP-staining techniques (Fig. 4). The observed aration, and incubated at 37 C for 30 min. Cover slips were rinsed three times with PBS (10 min each) and once with distilled water (5 min). Freshly prepared horseradish peroxidase indicator, consisting of 10 mg O. 4.0 of 3,3'-diamino-benzidine tetrahydrochloride (Sigma Chemical Co.) in 10 ml of 0.01% H202 in PBS, was added onto cell cultures. When the staining was fully developed, usually within 5 to 10 min, cover slips were submerged in 2% formalin solution, dehydrated with ethyl alcohol, and mounted on slides with semipermanent medium. The Nadi reaction provides an alterna3.0-/X tive procedure for detecting horseradish peroxidase -
(19).
10 Virus assay. For this procedure, appropriate virus dilutions were prepared in PBS, pH 7.1. Inoculum in a 0.2-ml volume was introduced onto quadruplicate D10Ix cover slip cell monolayers, which were then incubated Z 1, at 35 C for 2 h. Residual inoculum was removed, and 1 o 1 ml of maintenance medium was added to each cover 2.0slip cell culture. After incubation at 35 C for 20 to 24 h, cell monolayers were rinsed twice with PBS fixed with cold (-60 C) acetone, and either prepared for staining or stored at -70 C. Microscopy and infected cell counting. IFstained cover slip cell monolayers were examined with a Zeiss fluorescence microscope equipped with an 0 4 8 12 16 20 24 FITC exciter filter and no. 50 and 65 barrier filters. HOURS The same microscope with an ordinary light source was used to examine IP-stained cell monolayers. With FIG. 1. Appearance of influenza virus-infected these optical systems at a magnification of x640. the 1-5c-4 cells during incubation (35 C) and stained by! number of microscopic fields contained in the area of the IF (-) or IP (x) techniques.
326
HAHON, SIMPSON, AND ECKERT
J. CLIN. MICROBIOL.
FIG. 2. IF-stained influenza virus-infected 1-5c-4 cells at 16 h after incubation. x 125.
4:e
FIG. 3. IP-stained influenza virus-infected 1-5c-4 cell at 16 h after infection. x200.
VOL. 1, 1975
0
36-
a.) 0
(A 0
28/ 24-
0
U,)
20:
16/ w 8.)
LA
12-
z
8/ w
Z
4/ 0
1
2
327
VIRUS INFECTIVITY ASSESSMENT TECHNIQUES
16
8
4
RELATIVE
VIRUS
CONCENTRATION
FIG. 4. Linear function between the number of infected cells stained by IF (0) or IP (0) techniques and relative concentration of virus.
proportionality between these elements established a valid system for virus assay based on counts of infected cells when stained by either technique. Twenty-four replicate determinations were made to estimate the precision of' the cell-counting assay by infecting cell monolayers with a standard quantity of virus inoculum. After the prescribed incubation period, twelve infected cover slip cultures were stained by the IF technique, and twelve were stained by the IP procedure. Results indicate that both assays show high precision. The number of cell-infected units of virus per ml of inoculum determined by IF staining ranged from 1.2 x 108 to 1.5 x 108 with a mean of 1.35 x 108, a standard deviation of 40.13 x 108, and a standard error of ±0.03 x 108. With IP staining, the number of cellinfected units of virus per ml ranged from 1.3 x 108 to 1.5 x 108 with a mean of 1.43 x 108, a standard deviation of ±0.08 x 108, and a standard error of ±0.02 x 108. There was
To determine the reproducibility of' the cellcounting assay, a standard virus pool was titrated in the usual manner at different time intervals and stained by the IF or IP techniques. Results attest to the high reproducibility of the assays (Table 1). At the 0.05 probability level, assessment values were not significantly different when either staining procedure was used. Serum-neutralizing antibody determinations. The applicability of the IF serum neutralization test that is based on infected cell counting has been reported for several viral agents (21). Replicate tests were performed to compare the IF and IP techniques for estimating serum-neutralizing antibodies against influenza virus. Appropriate dilutions (1:5,000, 1:10,000, 1:20,000, and 1:30,000) of antiviral or normal (control) serum were mixed with equal volumes of a constant quantity of virus (5.0 x 104 cell-infecting units/ml). Virus and serum dilutions were prepared in PBS. After test mixtures were incubated at 35 C for 1 h. they were assayed for unneutralized virus in the prescribed manner. The percentage reduction of infected cell counts for each antiviral serum dilution was computed from the control counts and then plotted against the logarithm of' the corresponding final dilutions of' antiviral serum. Results show that linear relationships were obtained over a critical range by both the IF and IP techniques (Fig. 5). Furthermore, 50% serum-neutralizing end points of 1:23,000 were demonstrated with both procedures.
DISCUSSION The findings of this study demonstrate that the IF and IP cell-counting techniques for quantitatively assessing virus infectivity were comparable. To obtain a valid comparison between the two methods, efforts were directed TABLE 1. Reproducibility of IF and IP cell-counting assays of influenza virusa Date of assav
6 Aug 1974 12 Aug 1974 23 Sept 1974 26 Sept 1974 1 Oct 1974 3 Oct 1974
no
significant difference between these determinations at the 0.05 probability level.
Cell-infecting units of virus (x 108 per ml)
IF
IP
1.3 1.1 1.0 1.6 1.4 1.5 1.31h
1.4 1.0 1.1 1.5 1.2 1.3 1.25b
a B/Great Lakes/1739/54 strain. b
Mean.
328
HAHON, SIMPSON, AND ECKERT CELL INFECTING UNITS OF RESIDUAL INFECTIVITY INFECTIVITY CELL INFECTING UNITS OF INITIAL
6.0
2%
F
10 15 20 30 40 50 60 70 8085 90 95
5
If
,
)
98%
0
0
-1-
50s
z
R
0
J. CLIN. MICROBIOL.
hinderance bv enzyme resulting in lower antibody avidity of' conjugates. or other technical f'actors as vet unrecognized. was not ascertained. That the IP technique was in close agreement with the IF procedure for estimating neutralizing antibodies in antiviral serum, augurs well f'or its applicability in serum neutralization tests. The potential and efficacy of the IP technique appears as great as that of' the IF technique in studies rele-vant to or dependent on quantitative measurements of' virus infectivity (4. 8-14, 25). ACKNOWLEDGMENTS The competent assistance of' James A. Booth and .John acknowledged. Stewart is
4.0O
gyratef'ullv
a:
LITERATURE CITED
3C 30
35
40
4.5
50
5.5
60
65
70
PROBITS
FiC,. 5. Determination of 50% neutralizing end ) and IP serum by the IF (@) techniques.
points of anti-influenza
towards maintaining a constancy ot' reagents and procedures. This was achieved by conjugating the same antiviral globulin with either fluorescein or horseradish peroxidase. by staining replicate infected cover slip monolavers by direct methods, and enumerating infected cells under the same microscope in a similar manner. The proportionality observed between relative virus concentration and numbers of' inf'ected cells with each staining technique established the validity of the virus assav procedures. Both techniques gave results that were comparable in precision and reproducibility. In comparing the sensitivity of' f'luoresceinand peroxidase-labeled antibodies f'or detecting virus antigens in cells inf'ected with SV4() virus. equivalent percentages of' infected cells were found with each technique (1). A slight diff'erence was observed in the sensitivity of' the IF and IP techniques for detecting intracellular influenza viral antigens during the early stages of' infection. At 8 h af'ter infection, counts of' infected cells in IF-stained cell monolayers were three times higher than those noted by the IP technique. In the latter stages of infection (16 to 24 h). equivalent counts of infected cells were obtained with each procedure. Whether this slight discrepancy in sensitivity is related to the purity of antibody. peroxidase reagent. steric
1. Avrameas, S. 197(0. Immunoenzyme techniquLes: enzymes as markers for the localization of' antigens and antibodies. Int. Rev. Cvtol. 27:349-385. 2. Avrameas. S.. and ,J. Uriel. 1966. Methode de marquage d'antig&nes et d'anticrops avec des enzymes et son application en immune diftf'usion. C. R. Acad. Sci. (Paris) 262:2543-2545. :3. Benjamin. D. R.. and C. G. Ray. 1974. Use of' immunoperoxidase f'or the rapid identif'ication of' human mvxolviruses and pararnyxoviruses in tissue ctultture. Appl. Microbiol. 28:47-51. 4. Carter. G. B. 1965. 'T'he rapid detection. titration and dit'f'erentiation of' variola and vaccinia viruses by a 5.
6.
7.
S. 9.
1(). 11. 12. 1:3.
14.
15.
fluorescent antibody coverslip cell monolaver system. Virology 25:659-661. Carter. G. B. 197:3. Immunofluorescence in the rapid identif'ication of viruses and the role ot f'luorescenit cell counting, p. 54-65. In F. J. Baker (ed.), Microbiology of the seventies. Butterworths. London. Dobardzic. R.. A. Boudreault. and V. Pavilanis. 197:3. L'identification du virus de l'inf'ltuenza a ['aide de l'immunoperoxidase. Can. -J. Microbiol. 19:146-149. Dubois-Dalcq. M.. and L. H. Barbosa. 197:3. Immunoperoxidase stain of' measles antigen in tissue culture. J. Virol. 12:9(09-918. Hahon. N. 1967. Multiple infection of' cell moInolayers by virus mixtures. Appl. Microbiol. 15:458--459. Hahon. N. 1969. The kinetics of' neutralization of' Venezuelan equine encephalomyelitis virus by antiserum and the reversibility of' the reaction. .J. Gen. \irol. 4:77 --88. Hahon. N. 197(0. Neutralization anti-IuG test for antisera to Venezuelan equine encephalomyelitis virus. *1. Gen. Virol. 6:285-291. Hahon. N. 1974. Depression of viral interteron induction in cell monolavers by coal dust. Br. .J. Ind. Med. 31:201-20(8. Hahon. N.. .J. A. Booth. and H. L. Eckert. 197:1. Cell attachment and penetration by intluenza virus. Infect. Immunity 7:341-351. Hahon. N.. and K. 0. Cooke. 1967. Primary virus-cell interactions in the immunofluorescence assay of Venezuelan equine encephalomyelitis virus. J. Virol. 1:317-:126. Hahon. N.. and H. L. Eckert. 1972. Cell surface antigen induced by inf'luenza virus. Inf'ect. Immunity 6:7:30-73:5. Hebert. G. A.. P. L. Pelham. and B. Pittman. 197:1.
VOL. 1. 1975
16.
17.
18.
19.
20.
21. 22.
2:3.
24.
VIRUS INFECTIVITY ASSESSMENT TECHNIQUES
Determination of the optimal ammonium sulfate concentration f'or the f'ractionation of rabbit, sheep. horse. and goat antisera. Appl. Microbiol. 25:26-:36. Herrmann. J. E.. and S. A. Morse. 197:3. Coupling of' peroxidase to poliovirus antibody: characteristics of' t he conjugates and their use in virus detection. Inf'ect. Immunitv 8:645-649. Herrmann. .J. E., S. A. Morse, and M. F. Collins. 1974. Comparison of techniques and immunoreagents used f'or indirect immunofluorescence and immunoperoxidase identit'ication of' enteroviruses. Infect. Immunity 10:220-226. Kozikowski. E. H.. and N. Hahon. 1969. Quantitative assay of interferon by the immunofluorescent cellcounting technique. J. Gen. Virol. 4:441-44:3. Kurstak. E. 1971. The immunoperoxidase technique: localization of viral antigens in cells. p. 433-44. In K. Maramorosch and H. Koprowski (ed.). Methods of' virology, vol. 5, Academic Press Inc.. New York. Kurstak. E.. S. Belloncik. P. A. Onji. S. Montplaisin. and B. Martineau. 1972. Localisation par limmunoperoxydase des antigens du virus cvtomegalique en culture cellulaire de fibroblasts humains. Arch. Gesamte Vi. rusforsch. 38:67-76. Kwapinski. J. B. G. 1972. Immune neutralization tests. p. 575-589. In Methodology of immunochemical and immunological research. Wiley-Interscience. New \ork. Levadite. .J. C.. P. AtanasiuL. A. Gamet. and .J. C. Guillon. 197:3. Problemes poses au diagnostic de la rage par les methodes d'immunofluorescence et d immunoperoxydase. Bull. Soc. Pathol. Exot. 66:12-20. Liu. C. 1964. Fluorescent-antibody technics. p. 177-19:1. In E. H. Lennette and N. J. Schmidt (ed.). Diagnostic procedures for viral and rickettsial diseases. :3rd ed. American Public Health Association Inc.. New Y'ork. Nakane. P. K.. and G. B. Pierce. 1966. Enzyme-labelled antibodies: preparation and application for the locali-
25.
26. 27.
28. 29.
30. .31.
:32.
3:3.
329
zation of antigens. J. Histochem. Cytochem. 14:929-9:31. Oxf'ord. .J. S.. and G. C. Schild. 1968. Immunofluorescent studies on the inhibition of influenza A and B viruses in mammalian cell cultures by amines and ammonium compounds. J. Gen. Virol. 2:377-384. Rodriguez. J.. and F. Deinhardt. 1960(. Preparation of' a semi-permanent mounting medium for fluorescent antibody studies. Virology 12:316-317. Shabo. A. L.. .J. C. Petriccana. and R. L. Kirschstein. 1972. Immunoperoxidase localization ot' herpes zoster and simian virus 40 in cell culture. Appl. Microbiol. 23:1(0(01-1009. Siverd. N. J., and N. Sharon. 1969. Immunohistochemical method f'or detection of' vaccinia virus. Proc. Soc. Exp. Biol. Med. 131:939-941. Sutmoller. P.. and K. M. Cowan. 1974. The detection of f'oot- and mouth disease virus antigens in inf'ected cell cultures by immuno-peroxidase techniquLes. .J. Gen. \'irol. 22:287-291. Tripathy. D. N. I. F. Hanson. and A. H. Killinger. 197:3. Immunoperoxidase technique for detecting fowlpox antigen. Avian Dis. 17:274-278. Ubertine. T., G. N. Wilkie. and F. Noronka. 1971. Ulse of' horseradish peroxidase-labeled antibody for light and electron microscope localization ot reovirus antigen. Appl. Microbiol. 21:5-34-5:38. Wicker. R. L. 1971. Comparison of immunotluorescent and immunoenz% matic techniques applied to the study ofv iral anti,ens. Ann. N. Y. Acad. Sci. 177:490-500. Wong. S. C.. and E. D. Kilbourne. 1961. Changing viral susceptibility ot a human cell line in continuous cultivation. I. Production of infective virus in a variant of the Chang conjunctival cell f'ollowning infection with swine or N-WS intluenza viruses. .J. Exp. Med. 113:95-110.
JOURNAL OF CLINICAL MICROBIOLOGY Mar. 1975, p. 324-329 Copyright © 1975 American Society for Microbiology
Printed in U.S.A.
Assessment of Virus Infectivity by the Immunofluorescent and Immunoperoxidase Techniques NICHOLAS HAHON,* JANET SIMPSON, AND HERBERT L. ECKERT Appalachian Laboratory for Occupational Respiratory Diseases, U.S. Public Health Service,* and Departments of Pediatrics and Medicine, West Virginia University, School of Medicine, Morgantown, West Virginia 26505 Received for publication 20 November 1974
The immunofluorescent and immunoperoxidase cell-counting techniques were comparable in precision and reproducibility for the quantitative assessment of influenza virus inf'ectivity. The dose-response function was linear with each procedure, and comparable results were obtained for estimating neutralizing antibodies in antiviral serum.
Quantitative assays of' virus infectivity by enumeration of cells containing immunofluorescent viral antigens have been established for agents representative of' almost all major animal groups (5). These assays are dependent on a single cycle of infection and possess such outstanding attributes as rapidity (usually less than 24 h), specif'icity, reproducibility, precision, and high sensitivity. The versatility of the immunofluorescent cell-counting procedure has been demonstrated in fundamental and applied studies, namely, the kinetics of' virus-cell interactions (8, 12-14) and of' virus neutralization (9, 10), differentiation of' serologically similar viruses (4), and assessments of' interferon (11, 18) and virustatic compounds (25). Although the technique has been used widely in virology, circumstances that sometimes restrict its effectiveness are naturally occurring autof'luorescence, fading of stained cells during examination, and the limited resolution of' ultraviolet microscopy. In recent years, the immunoenzymatic technique, the coupling of' enzymes to immunoglobulins without significant alteration of their respective activities (2, 24), has been successfully employed to identif'y a wide variety of viral antigens in cell cultures (3, 6, 7, 16, 17, 19, 20, 22, 27-31). The technique embodies the advantages of' the immunofluorescence procedure without some of' its limitations. Comparable sensitivity in detecting viral antigens has been reported between the immunoenzymatic and immunofluorescence procedures (17. 32). The advantages of the former have been broadened by such considerations as the relative ease of' conjugate preparation, the reduction of nonspecific reactions, the stability of' the stain, and the enhanced visualization of' viral antigens under
light and electron microscopy. For assaying virus inf'ectivity, however, the immunoenzymatic technique has not been extensively exploited. This study was undertaken to evaluate quantitatively the assessment of virus infectivity by the immunofluorescent (IF) and immunoperoxidase (IP) cell-counting techniques. MATERIALS AND METHODS Virus. The B/Great Lakes/1739/54 influenza virus strain emploved in this study was obtained from the American Type Culture Collection, Rockville, Md. Stock pools of' virus, consisting of infective allantoic fluids, were prepared and stored in the manner described previouslV (12). Cell cultures. Clone 1-5c-4 derived from a variant line of' Chang's conjunctival cell (33) was propagated with Eagle minimum essential medium containing 10%c fetal calf serum. Cells were maintained with minimal essential medium plus 5% fetal calf' serum. For virus assay, cells were cultivated on circular cover slips (15 mm in diameter) inserted in flat-bottomed glass vials (19 by 65 mm). One milliliter of' cell suspension containing approximately 105 cells was introduced onto each cover slip which was then incubated at 35 C for 24 h or until a complete cell monolayer was formed. Virus antiserum. Initially, rabbits were injected intramuscularly in a 2.0-ml volume with an equal quantity of' virus suspension (1.0 x 108 cell-infecting units/ml) and Freund complete adjuvant. At weekly intervals thereafter, animals were twice injected intravenously with 1.5 ml of virus suspension; they were exsanguinated 10 days after the last injection. The pooled antiserum was stored at -20 C. IF conjugation and staining. The globulin fraction of' antiserum was precipitated with ammonium sulfate at 4 C (15) and conjugated with fluorescein isothiocyanate (Nutritional Biochemical Corp., Cleveland, Ohio) at the ratio of 0.02 mg of dye per mg of protein (23). Unbound dye was removed by passing
324
VOL. 1, 1975
VIRUS INFECTIVITY ASSESSMENT TECHNIQUES
325
conjugated globulin through a Sephadex G-50 column a 15-mm diameter cover slip was 4,773. To determine (Pharmacia Fine Chemicals Inc., Piscataway, N.J.). this number, the diameter of the microscopic field was To reduce nonspecific fluorescence, 5 ml of conjugated measured with a stage micrometer, and the calculated globulin was diluted with an equal volume of 0.01 M area was divided into the area (176.62 mm) of the phosphate-buffered saline (PBS), pH 7.1, and ad- cover slip. For each cover slip cell monolayer, 50 sorbed twice with 200 mg of acetone-dried mouse microscopic fields were examined. To calculate the liver powder (23). Conjugated globulin was stored in number of cell-infecting units of virus per milliliter, the average number of IF- or IP-positive cells per field 1-ml portions at -20 C. The direct method was used to demonstrate IF of was multiplied by the number of fields per cover slip. viral antigens in infected cells. Fixed cell cultures the reciprocal of the dilution of virus inoculum. and were washed with PBS and stained with conjugated the volume factor. antiserum for 20 min at room temperature (22 C). Cover slip cell monolayers were then rinsed in two RESULTS changes of PBS and mounted on slides with a Quantitative evaluations of the assays. semi-permanent medium (26). The sensitivity of IF and IP staining for detectIP conjugation and staining. Portions of the remaining globulin fraction of virus antiserum used ing intracellular viral antigens was determined for IF conjugation were labeled with horseradish by following their rate of appearance in infected peroxidase as described by Kurstak (19). Briefly, 12 cell monolayers during a 24-h incubation pemg of horseradish peroxidase (RZ 3.0, Type VI, Sigma riod. Infected cells were first detected 8 h after Chemical Co., St. Louis, Mo.) was dissolved in 0.5 ml infection by either staining procedure; however, of 0.1 M phosphate buffer solution, pH 6.8, and mixed the number of infected cells noted in IF-stained with 0.5 ml (approximately 5 mg) of globulin prepara- cell monolayers was three times higher (Fig. 1). tion. While the solution was gently stirred. 0.05 ml of a 1% aqueous solution of glutaraldehyde was added With each staining technique, maximal and dropwise. After 2 h at room temperature, the mixture equivalent numbers of infected cells appeared was dialyzed against two changes of PBS at 4 C at 16 h after infection (Figs. 2 and 3). Thereovernight and then clarified by centrifugation. Peroxi- after, the number was constant throughout the dase conjugates were used without further purifica- remainder of the incubation period. tion. Before use, the stock solution was diluted 1:10. The dose-response was linear between twofold For staining of cover slip cell cultures by the direct dilutions of virus over a range of 1.2 log1O units IP technique, fixed cell cultures were rinsed with and numbers of infected cells by using the IF- or PBS, covered with peroxidase-labeled antibody prep- IP-staining techniques (Fig. 4). The observed aration, and incubated at 37 C for 30 min. Cover slips were rinsed three times with PBS (10 min each) and once with distilled water (5 min). Freshly prepared horseradish peroxidase indicator, consisting of 10 mg O. 4.0 of 3,3'-diamino-benzidine tetrahydrochloride (Sigma Chemical Co.) in 10 ml of 0.01% H202 in PBS, was added onto cell cultures. When the staining was fully developed, usually within 5 to 10 min, cover slips were submerged in 2% formalin solution, dehydrated with ethyl alcohol, and mounted on slides with semipermanent medium. The Nadi reaction provides an alterna3.0-/X tive procedure for detecting horseradish peroxidase -
(19).
10 Virus assay. For this procedure, appropriate virus dilutions were prepared in PBS, pH 7.1. Inoculum in a 0.2-ml volume was introduced onto quadruplicate D10Ix cover slip cell monolayers, which were then incubated Z 1, at 35 C for 2 h. Residual inoculum was removed, and 1 o 1 ml of maintenance medium was added to each cover 2.0slip cell culture. After incubation at 35 C for 20 to 24 h, cell monolayers were rinsed twice with PBS fixed with cold (-60 C) acetone, and either prepared for staining or stored at -70 C. Microscopy and infected cell counting. IFstained cover slip cell monolayers were examined with a Zeiss fluorescence microscope equipped with an 0 4 8 12 16 20 24 FITC exciter filter and no. 50 and 65 barrier filters. HOURS The same microscope with an ordinary light source was used to examine IP-stained cell monolayers. With FIG. 1. Appearance of influenza virus-infected these optical systems at a magnification of x640. the 1-5c-4 cells during incubation (35 C) and stained by! number of microscopic fields contained in the area of the IF (-) or IP (x) techniques.
326
HAHON, SIMPSON, AND ECKERT
J. CLIN. MICROBIOL.
FIG. 2. IF-stained influenza virus-infected 1-5c-4 cells at 16 h after incubation. x 125.
4:e
FIG. 3. IP-stained influenza virus-infected 1-5c-4 cell at 16 h after infection. x200.
VOL. 1, 1975
0
36-
a.) 0
(A 0
28/ 24-
0
U,)
20:
16/ w 8.)
LA
12-
z
8/ w
Z
4/ 0
1
2
327
VIRUS INFECTIVITY ASSESSMENT TECHNIQUES
16
8
4
RELATIVE
VIRUS
CONCENTRATION
FIG. 4. Linear function between the number of infected cells stained by IF (0) or IP (0) techniques and relative concentration of virus.
proportionality between these elements established a valid system for virus assay based on counts of infected cells when stained by either technique. Twenty-four replicate determinations were made to estimate the precision of' the cell-counting assay by infecting cell monolayers with a standard quantity of virus inoculum. After the prescribed incubation period, twelve infected cover slip cultures were stained by the IF technique, and twelve were stained by the IP procedure. Results indicate that both assays show high precision. The number of cell-infected units of virus per ml of inoculum determined by IF staining ranged from 1.2 x 108 to 1.5 x 108 with a mean of 1.35 x 108, a standard deviation of 40.13 x 108, and a standard error of ±0.03 x 108. With IP staining, the number of cellinfected units of virus per ml ranged from 1.3 x 108 to 1.5 x 108 with a mean of 1.43 x 108, a standard deviation of ±0.08 x 108, and a standard error of ±0.02 x 108. There was
To determine the reproducibility of' the cellcounting assay, a standard virus pool was titrated in the usual manner at different time intervals and stained by the IF or IP techniques. Results attest to the high reproducibility of the assays (Table 1). At the 0.05 probability level, assessment values were not significantly different when either staining procedure was used. Serum-neutralizing antibody determinations. The applicability of the IF serum neutralization test that is based on infected cell counting has been reported for several viral agents (21). Replicate tests were performed to compare the IF and IP techniques for estimating serum-neutralizing antibodies against influenza virus. Appropriate dilutions (1:5,000, 1:10,000, 1:20,000, and 1:30,000) of antiviral or normal (control) serum were mixed with equal volumes of a constant quantity of virus (5.0 x 104 cell-infecting units/ml). Virus and serum dilutions were prepared in PBS. After test mixtures were incubated at 35 C for 1 h. they were assayed for unneutralized virus in the prescribed manner. The percentage reduction of infected cell counts for each antiviral serum dilution was computed from the control counts and then plotted against the logarithm of' the corresponding final dilutions of' antiviral serum. Results show that linear relationships were obtained over a critical range by both the IF and IP techniques (Fig. 5). Furthermore, 50% serum-neutralizing end points of 1:23,000 were demonstrated with both procedures.
DISCUSSION The findings of this study demonstrate that the IF and IP cell-counting techniques for quantitatively assessing virus infectivity were comparable. To obtain a valid comparison between the two methods, efforts were directed TABLE 1. Reproducibility of IF and IP cell-counting assays of influenza virusa Date of assav
6 Aug 1974 12 Aug 1974 23 Sept 1974 26 Sept 1974 1 Oct 1974 3 Oct 1974
no
significant difference between these determinations at the 0.05 probability level.
Cell-infecting units of virus (x 108 per ml)
IF
IP
1.3 1.1 1.0 1.6 1.4 1.5 1.31h
1.4 1.0 1.1 1.5 1.2 1.3 1.25b
a B/Great Lakes/1739/54 strain. b
Mean.
328
HAHON, SIMPSON, AND ECKERT CELL INFECTING UNITS OF RESIDUAL INFECTIVITY INFECTIVITY CELL INFECTING UNITS OF INITIAL
6.0
2%
F
10 15 20 30 40 50 60 70 8085 90 95
5
If
,
)
98%
0
0
-1-
50s
z
R
0
J. CLIN. MICROBIOL.
hinderance bv enzyme resulting in lower antibody avidity of' conjugates. or other technical f'actors as vet unrecognized. was not ascertained. That the IP technique was in close agreement with the IF procedure for estimating neutralizing antibodies in antiviral serum, augurs well f'or its applicability in serum neutralization tests. The potential and efficacy of the IP technique appears as great as that of' the IF technique in studies rele-vant to or dependent on quantitative measurements of' virus infectivity (4. 8-14, 25). ACKNOWLEDGMENTS The competent assistance of' James A. Booth and .John acknowledged. Stewart is
4.0O
gyratef'ullv
a:
LITERATURE CITED
3C 30
35
40
4.5
50
5.5
60
65
70
PROBITS
FiC,. 5. Determination of 50% neutralizing end ) and IP serum by the IF (@) techniques.
points of anti-influenza
towards maintaining a constancy ot' reagents and procedures. This was achieved by conjugating the same antiviral globulin with either fluorescein or horseradish peroxidase. by staining replicate infected cover slip monolavers by direct methods, and enumerating infected cells under the same microscope in a similar manner. The proportionality observed between relative virus concentration and numbers of' inf'ected cells with each staining technique established the validity of the virus assav procedures. Both techniques gave results that were comparable in precision and reproducibility. In comparing the sensitivity of' f'luoresceinand peroxidase-labeled antibodies f'or detecting virus antigens in cells inf'ected with SV4() virus. equivalent percentages of' infected cells were found with each technique (1). A slight diff'erence was observed in the sensitivity of' the IF and IP techniques for detecting intracellular influenza viral antigens during the early stages of' infection. At 8 h af'ter infection, counts of' infected cells in IF-stained cell monolayers were three times higher than those noted by the IP technique. In the latter stages of infection (16 to 24 h). equivalent counts of infected cells were obtained with each procedure. Whether this slight discrepancy in sensitivity is related to the purity of antibody. peroxidase reagent. steric
1. Avrameas, S. 197(0. Immunoenzyme techniquLes: enzymes as markers for the localization of' antigens and antibodies. Int. Rev. Cvtol. 27:349-385. 2. Avrameas. S.. and ,J. Uriel. 1966. Methode de marquage d'antig&nes et d'anticrops avec des enzymes et son application en immune diftf'usion. C. R. Acad. Sci. (Paris) 262:2543-2545. :3. Benjamin. D. R.. and C. G. Ray. 1974. Use of' immunoperoxidase f'or the rapid identif'ication of' human mvxolviruses and pararnyxoviruses in tissue ctultture. Appl. Microbiol. 28:47-51. 4. Carter. G. B. 1965. 'T'he rapid detection. titration and dit'f'erentiation of' variola and vaccinia viruses by a 5.
6.
7.
S. 9.
1(). 11. 12. 1:3.
14.
15.
fluorescent antibody coverslip cell monolaver system. Virology 25:659-661. Carter. G. B. 197:3. Immunofluorescence in the rapid identif'ication of viruses and the role ot f'luorescenit cell counting, p. 54-65. In F. J. Baker (ed.), Microbiology of the seventies. Butterworths. London. Dobardzic. R.. A. Boudreault. and V. Pavilanis. 197:3. L'identification du virus de l'inf'ltuenza a ['aide de l'immunoperoxidase. Can. -J. Microbiol. 19:146-149. Dubois-Dalcq. M.. and L. H. Barbosa. 197:3. Immunoperoxidase stain of' measles antigen in tissue culture. J. Virol. 12:9(09-918. Hahon. N. 1967. Multiple infection of' cell moInolayers by virus mixtures. Appl. Microbiol. 15:458--459. Hahon. N. 1969. The kinetics of' neutralization of' Venezuelan equine encephalomyelitis virus by antiserum and the reversibility of' the reaction. .J. Gen. \irol. 4:77 --88. Hahon. N. 197(0. Neutralization anti-IuG test for antisera to Venezuelan equine encephalomyelitis virus. *1. Gen. Virol. 6:285-291. Hahon. N. 1974. Depression of viral interteron induction in cell monolavers by coal dust. Br. .J. Ind. Med. 31:201-20(8. Hahon. N.. .J. A. Booth. and H. L. Eckert. 197:1. Cell attachment and penetration by intluenza virus. Infect. Immunity 7:341-351. Hahon. N.. and K. 0. Cooke. 1967. Primary virus-cell interactions in the immunofluorescence assay of Venezuelan equine encephalomyelitis virus. J. Virol. 1:317-:126. Hahon. N.. and H. L. Eckert. 1972. Cell surface antigen induced by inf'luenza virus. Inf'ect. Immunity 6:7:30-73:5. Hebert. G. A.. P. L. Pelham. and B. Pittman. 197:1.
VOL. 1. 1975
16.
17.
18.
19.
20.
21. 22.
2:3.
24.
VIRUS INFECTIVITY ASSESSMENT TECHNIQUES
Determination of the optimal ammonium sulfate concentration f'or the f'ractionation of rabbit, sheep. horse. and goat antisera. Appl. Microbiol. 25:26-:36. Herrmann. J. E.. and S. A. Morse. 197:3. Coupling of' peroxidase to poliovirus antibody: characteristics of' t he conjugates and their use in virus detection. Inf'ect. Immunitv 8:645-649. Herrmann. .J. E., S. A. Morse, and M. F. Collins. 1974. Comparison of techniques and immunoreagents used f'or indirect immunofluorescence and immunoperoxidase identit'ication of' enteroviruses. Infect. Immunity 10:220-226. Kozikowski. E. H.. and N. Hahon. 1969. Quantitative assay of interferon by the immunofluorescent cellcounting technique. J. Gen. Virol. 4:441-44:3. Kurstak. E. 1971. The immunoperoxidase technique: localization of viral antigens in cells. p. 433-44. In K. Maramorosch and H. Koprowski (ed.). Methods of' virology, vol. 5, Academic Press Inc.. New York. Kurstak. E.. S. Belloncik. P. A. Onji. S. Montplaisin. and B. Martineau. 1972. Localisation par limmunoperoxydase des antigens du virus cvtomegalique en culture cellulaire de fibroblasts humains. Arch. Gesamte Vi. rusforsch. 38:67-76. Kwapinski. J. B. G. 1972. Immune neutralization tests. p. 575-589. In Methodology of immunochemical and immunological research. Wiley-Interscience. New \ork. Levadite. .J. C.. P. AtanasiuL. A. Gamet. and .J. C. Guillon. 197:3. Problemes poses au diagnostic de la rage par les methodes d'immunofluorescence et d immunoperoxydase. Bull. Soc. Pathol. Exot. 66:12-20. Liu. C. 1964. Fluorescent-antibody technics. p. 177-19:1. In E. H. Lennette and N. J. Schmidt (ed.). Diagnostic procedures for viral and rickettsial diseases. :3rd ed. American Public Health Association Inc.. New Y'ork. Nakane. P. K.. and G. B. Pierce. 1966. Enzyme-labelled antibodies: preparation and application for the locali-
25.
26. 27.
28. 29.
30. .31.
:32.
3:3.
329
zation of antigens. J. Histochem. Cytochem. 14:929-9:31. Oxf'ord. .J. S.. and G. C. Schild. 1968. Immunofluorescent studies on the inhibition of influenza A and B viruses in mammalian cell cultures by amines and ammonium compounds. J. Gen. Virol. 2:377-384. Rodriguez. J.. and F. Deinhardt. 1960(. Preparation of' a semi-permanent mounting medium for fluorescent antibody studies. Virology 12:316-317. Shabo. A. L.. .J. C. Petriccana. and R. L. Kirschstein. 1972. Immunoperoxidase localization ot' herpes zoster and simian virus 40 in cell culture. Appl. Microbiol. 23:1(0(01-1009. Siverd. N. J., and N. Sharon. 1969. Immunohistochemical method f'or detection of' vaccinia virus. Proc. Soc. Exp. Biol. Med. 131:939-941. Sutmoller. P.. and K. M. Cowan. 1974. The detection of f'oot- and mouth disease virus antigens in inf'ected cell cultures by immuno-peroxidase techniquLes. .J. Gen. \'irol. 22:287-291. Tripathy. D. N. I. F. Hanson. and A. H. Killinger. 197:3. Immunoperoxidase technique for detecting fowlpox antigen. Avian Dis. 17:274-278. Ubertine. T., G. N. Wilkie. and F. Noronka. 1971. Ulse of' horseradish peroxidase-labeled antibody for light and electron microscope localization ot reovirus antigen. Appl. Microbiol. 21:5-34-5:38. Wicker. R. L. 1971. Comparison of immunotluorescent and immunoenz% matic techniques applied to the study ofv iral anti,ens. Ann. N. Y. Acad. Sci. 177:490-500. Wong. S. C.. and E. D. Kilbourne. 1961. Changing viral susceptibility ot a human cell line in continuous cultivation. I. Production of infective virus in a variant of the Chang conjunctival cell f'ollowning infection with swine or N-WS intluenza viruses. .J. Exp. Med. 113:95-110.
