- Email: [email protected]
Optimization of in situ cellular ELISA performed on influenza A virus-infected monolayers for screening of antiviral agents
Journal of Virological Methods 77 (1999) 165 – 177
Optimization of in situ cellular ELISA performed on influenza A virus-infected monolayers for screening of antiviral agents Andrzej Myc a, Marie J. Anderson b, James R. Baker Jr. a,* a
Department of Internal Medicine, Di6ision of Allergy, Room 9220 MSRB III, Uni6ersity of Michigan, 1150 West Medical Center Dri6e, Ann Arbor, MI 48109 -0648, USA b Markey Molecular Medicine Center, Box 357720, Uni6ersity of Washington, Seattle, WA 98195 -7720, USA Received 27 July 1998; received in revised form 20 October 1998; accepted 20 October 1998
Abstract Viral susceptibility testing has been traditionally performed by the plaque reduction assay (PRA) which is laborious, time consuming, relatively expensive, and requires subjective input by the reader. An in situ cellular enzyme-linked immunosorbent assay (ELISA) has been developed with the potential to overcome many of the limitations of PRA and has been applied to a variety of viruses. This study establishes the specific conditions necessary for susceptibility testing of influenza A virus to antiviral agents such as amount of inoculum size, duration of incubation, fixative type, and cell number; factors which are critical to the performance of the in situ cellular ELISA. In situ cellular ELISA was found to correlate strongly with the plaque assay (PA) (R 2 = 0.997, PB 0.002). Both assays were applied to test the susceptibility of influenza A virus to a new antiviral emulsion agent and yielded comparable data. The optimized in situ cellular ELISA can serve as a reliable assay for the rapid screening of large numbers of antiviral agents. © 1999 Elsevier Science B.V. All rights reserved. Keywords: ELISA; Plaque reduction assay; Influenza A virus; Susceptibility; MDCK cells
1. Introduction Although the plaque reduction assay (PRA) has been accepted as the gold standard for measuring * Corresponding author. Tel.: +1-734-6472777; fax: +1734-9362990; e-mail: [email protected].
the susceptibility of viruses to antiviral agents, it is not convenient for screening a large numbers of antiviral agents. This is because PRA is laborious, time consuming, relatively expensive, and requires subjective input by the reader. Moreover, the virus detection range is very narrow (0–100 pfu), and this requires accurate predetermination of
0166-0934/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 0 9 3 4 ( 9 8 ) 0 0 1 5 0 - 5
166
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
infectious titer prior to undertaking susceptibility testing. All these drawbacks have prompted the development of rapid and less laborious assays to assess the activity of antiviral agents. These assays should be quantitative, overcome the limitations of the PRA, and yet be correlative with the traditional method. The MTT method has been used to assess the antiviral activity of compounds against the human immunodeficiency virus (Pauwels et al., 1988), herpes simplex virus type 1 (HSV-1) (Takeuchi et al., 1991), and influenza virus A (Fluv. A) (Hosoya et al., 1992; Watanabe et al., 1994) in cell system in vitro. However, this method is limited to these particular viruses because it measures the mitochondrial dehydrogenase activity to assess virus-induced cell damage. It cannot be utilized for viruses that do not grow well in culture or for cells with low enzymatic activity. An antiviral screening system using a lactate dehydrogenase (LDH) detection technique has been also reported (Watanabe et al., 1995). This system can be used to detect the infectivity of viruses that cannot be detected adequately by the MTT method. Although, both MTT and LDH assays are very easy to carry out, they are not specific to virus and detect the virus indirectly by assessing virus-induced cell damage. To study antiviral compounds, more specific assays have been developed and applied to offer better quantitation of virus. Berkowitz and Levin (1985) reported the enzyme-linked immunosorbent assay (ELISA) for evaluating drugs against varicella-zoster virus (Berkowitz and Levin, 1985). An ELISA has also been established for susceptibility testing of herpes simplex virus to acyclovir (Rabalais et al., 1987; Leahy et al., 1994, 1996). Hollingshead et al. utilized ELISA system for evaluating antiretroviral activity against Rauscher murine leukemia virus (Hollingshead et al., 1992). Influenza A virus, a common respiratory pathogen, was selected for this study as a model to test antiviral agents. To our knowledge there has been no report describing the conditions of in situ cellular ELISA performed on influenza A virus infected cells. We optimized the critical factors of the in situ cellular ELISA to make it comparable with PRA to quantitate influenza A virus. Advantages of the cellular ELISA over the PRA include:
(i) utilization of the 96-well format, which facilitates measurements in replicate, requires a small number of cells, and is adaptable to automation; (ii) less demanding which conserves costs, time, and labor; (iii) more quantitative; (iv) less subjective as it does not require the reader’s input. 2. Materials and methods
2.1. Virus Influenza virus A/AA/6/60, H2N2 (Herlocher, et al., 1996) was kindly provided by Hunein Maassab (School of Public Health, University of Michigan, Ann Arbor, MI). Influenza A virus was propagated in the allantoic cavities of fertilized pathogen-free hen eggs (SPAFAS, Norwich, CT) using standard methods described elsewhere (Barrett and Inglis, 1985). Virus stock was kept in aliquots (108 pfu/ml) of infectious allantoic fluids at − 80°C.
2.2. Cells Madin Darby canine kidney (MDCK) cells were purchased from the American Type Culture Collection (ATCC, Rockville, MD).
2.3. Media 2.3.1. Maintenance medium Cells were maintained in Eagle’s minimal essential medium with Earle’s salts, 2 mM L-glutamine, and 1.5 g/l sodium bicarbonate (Mediatech, Herndon, VA) containing 10% fetal bovine serum (FBS, Hyclone Laboratories, Logan, UT) supplemented with 0.1 mM non-essential amino acids, 1.0 mM sodium pyruvate, 100U penicillin/ml, streptomycin 100 mg/ml (Life Technologies, Gaithersburg, MD). 2.3.2. Infection medium Maintenance media (without fetal bovine serum) was formulated with 3.0 mg/ml of trypsin treated with tolylsulfonyl phenylalanyl chloromethyl ketone (TPCK-treated trypsin; Worthington Biochemical Corporation, Lakewood, NJ).
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
2.3.3. O6erlay medium Overlay medium consisted of equal amounts of 2X infection medium and 1.6% SeaKem ME agarose (FMC BioProducts, Rockland, MD). Staining agarose overlay medium consisted of agarose overlay medium plus 0.01% neutral red solution (Life Technologies, Gaithersburg, MD) without TPCK-treated trypsin. 2.4. Plaque and plaque reduction assays To directly compare the PA and cellular ELISA, MDCK cells were seeded at 2×105 cells/ well in 6-well Falcon plates (Becton Dickenson and Company, Franklin Lakes, NJ) and incubated at 37oC/5% CO2 for 3 days. Cells were washed free of complete media prior to use. Twofold dilutions of influenza A virus were made in infection medium. The cells were infected with 100 ml of the dilutions (for direct comparison with the cellular ELISA) and supplemented with an additional 400 ml of infection medium to prevent the drying of cell monolayers. The cells were incubated at 37oC/5% CO2 for 1 h to allow the virus to adsorb. The inoculum was removed and 1 ml of agarose overlay medium was added to each well. The plates were incubated at 37oC/5% CO2. Plaques appeared approximately 48 h post-infection and were stained with staining agarose overlay medium. The plates were returned to incubation at 37oC/5% CO2. Plaques were counted 6–12 h after staining. The plaque reduction assay was performed with a modification of the method described by Hayden et al. (1980). MDCK cells were seeded at 1× 105 cells/well in 12-well Falcon® plates and incubated at 37oC/5% CO2 for 3 days. Approximately 1×108 pfu of influenza A virus was incubated with BCTP as described below. Controls included influenza A virus plus infection medium and BCTP plus infection medium. The influenza A virus-BCTP treatments and controls were diluted in infection medium to contain 30 – 100 pfu/ 250 ml or the same concentration of BCTP. The three remaining wells served as cell controls and 250 ml infection medium/well was added. Confluent cell monolayers were inoculated in triplicate on three plates and incubated at 37oC/5% CO2 for
167
1 h. The inoculum/medium was aspirated and 1 ml of agarose overlay medium/well was added and plates were incubated at 37oC/5% CO2 until plaques appeared. Monolayers were stained with the agarose overlay medium and incubation was continued at 37oC/5% CO2. Plaques were counted 6–12 h after staining. The average plaque count from nine wells for each BCTP concentration was plotted against the average plaque count of untreated virus.
2.5. In situ cellular enzyme-linked immunosorbent assay To detect and quantitate viral proteins in MDCK cells infected with influenza A virus, the in situ cellular ELISA was developed based on the principles described by Berkowitz and Levin, 1985. The critical factors of the assay: the fixatives, time of virus adsorption, duration of incubation after infection, cell number, and number of virus particles were examined. MDCK cells in 100 ml complete medium were seeded on flat-bottom 96-well microtiter plates precoated with 0.5% gelatin and incubated overnight. On the next day, the culture medium was removed and cells were washed with maintenance medium. A total of 100 ml of viral inoculum was added to the wells and incubated for an indicated period of time. The viral inoculum was removed and replaced with 100 ml of infection medium + 2% FBS. Infected MDCK cells were incubated for an additional 6–48 h, as necessary, and medium was aspirated. The cells were washed once with PBS and fixed with one of the fixatives as described below. On the day of assay, fixed cells were washed with PBS and blocked with 1% dry milk in PBS for 30 min at 37°C. One hundred ml of ferret anti-influenza A virus polyclonal antibody (kindly provided by Dr Hunein Maassab) at 1× 10 − 3 dilution was added to the wells for 1 h at 37°C. The cells were washed four times with washing buffer (PBS and 0.05% Tween-20), and incubated with 100 ml at 1 × 10 − 3 dilution of goat anti-ferret peroxidase conjugated antibody (Kirkegaard and Perry Laboratories, Gaithersburg, MA) for 30 min at 37°C. Cells were washed four times and incubated with 100 ml of 1-Step™ Turbo TMB-ELISA substrate (Pierce
168
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Rockford, IL) until color was developed. The reaction was stopped with 1 N sulfuric acid and plates were read at a wavelength of 450 nm in ELISA reader.
2.6. Fixation procedure MDCK cells were washed once with PBS and were subjected to fixation with one of the following fixatives: acetone – ethanol (1:1) mixture, 0.5% glutaraldehyde, or 1% paraformaldehyde. One hundred microlitres of − 20oC acetone – ethanol mixture was added to each well on ice. Cells were fixed at −20oC for at least 15 min. One hundred microlitres of glutaraldehyde was added to each well and the plate was incubated for 30 min at room temperature (RT). Glutaraldehyde was aspirated and reactive aldehyde groups were blocked with PBSLE (PBS+ 100 mM lysine+100 mM ethanolamine) to avoid the covalent binding of proteins added at subsequent steps. The cells were washed with PBS, dried, and stored at 4oC. One hundred microlitres of paraformladehyde was added to each well on ice and incubated for 15 min then washed with PBS, dried, and stored at 4°C.
2.7. Influenza A treatment with BCTP or Triton X-100 In antiviral activity testing we evaluated BCTP, a new anti-microbial nanoemulsion disinfectant (Novavax, Rockville, MD), described elsewhere (Baker et al., 1997; Reuter et al., 1998) or a commonly available antiviral disinfectant, Triton X-100. Prior to assessing the efficacy of the nanoemulsion or Triton X-100 in our cellular ELISA, the direct cytotoxicity of BCTP and Triton X-100 on MDCK cells was determined by microscope inspection and MTT assay. Dilutions of a mixture of each virus and either the nanoemulsion or Triton X-100 applied to the cellular ELISA were always higher than the determined non-toxic concentration. Approximately 1× 108 pfu of influenza A virus were incubated either with BCTP or Triton X-100 at final concentrations of 1× 10 − 1, 1×10 − 2, 1× 10 − 3 or 5× 10 − 2, 5×10 − 3, 5×10 − 4, respectively for 30
min. at RT on a shaker. After incubation, serial dilutions of the BCTP/virus or Triton X-100/virus mixtures were made in infection medium and overlaid on MDCK cells to carry out either the cellular ELISA, PRA or both assays as described above.
3. Results
3.1. Optimization of cellular ELISA To optimize the in situ cellular ELISA, the following variables were examined: fixatives, virus adsorption time, post-infection incubation time, cell number, and amount of viral inoculum.
3.1.1. Fixati6es To test the effect of fixatives on the cellular ELISA performance three commonly used fixatives were compared: glutaraldehyde, paraformaldehyde, and the acetone–ethanol mixture. Although all three fixatives preserved the cell monolayer, as observed by microscope examination, different ELISA results were obtained (Fig. 1). Glutaraldehyde performed very poorly, producing little differentiation between infected and uninfected cells. Fixation with paraformaldehyde yielded greater resolution than glutaraldehyde and gave the smallest standard error. Using the acetone–ethanol mixture we obtained the best difference between uninfected and infected cells and the highest overall absorbance values. Based on these data, the acetone–ethanol mixture was used as the fixative in the succeeding experiments. 3.1.2. Effect of cell number on the performance of cellular ELISA We also investigated the upper limit of initial cell concentration and post-infection incubation time useful in the assay. Four different concentrations of cells in the range of 2.5× 103 –2.0×104 per well were used to seed the monolayer. Twenty four hours later the cells were infected with virus (2.5× 103pfu/well) and incubated for an additional 48 h. The highest cell concentration tested (2× 104 cells/well) yielded the highest absorbance readings and the best resolution. The three other
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
cell numbers tested: 2.5×103, 5.0 × 103, and 1.0×104 per well gave approximately 50% lower absorbance readings (Fig. 2). Cells at a concentration higher than 2.0×104 per well tended to overgrow and detach from the wells causing poor reproducibility of the assay (data not shown).
3.1.3. Virus adsorption and duration of post-infection incubation of MDCK cells Since cellular ELISA performance depends on efficient virus infection of plated cells, the effect of virus adsorption time on resolution and reproducibility of the assay was investigated. MDCK cells were incubated with influenza A virus for 15, 30, and 60 min. After each time point, the infection medium was replaced with virus free infection medium+2%FBS, and the cells were incubated for an additional 24 h prior to ELISA. There was not a significant difference in the
169
ELISA results in respect to the time of virus incubation on MDCK cells (data not shown). Virus reproductive cycles vary from a few to several hours, which may modify the cellular ELISA performance. Thus, it was important to estimate the optimal time of post-infection incubation for influenza A virus infected cells to obtain a highly reproducible ELISA. Cells were incubated from 6 to 48 h after infection followed by the cellular ELISA. Absorbance increase was logarithmic in the range over a 6–48 h incubation period (Fig. 3). The 6 h incubation period did not yield any increase in absorbance over baseline level. The highest absorbance was observed at the 48-h incubation period, but the reproducibility of the ELISA was poor due to cell lysis. Based on these data, the 24 h incubation time was chosen as the optimal, post-infection period for the in situ cellular ELISA protocol.
Fig. 1. Effect of fixatives on cellular ELISA performance. MDCK cells (104 cells/well) infected with influenza A virus (2.5 ×106 pfu/ml) were incubated for 24 h, fixed with the three different fixatives and subjected to ELISA. Each bar represents the mean of three replicates + /− 1 SE.
170
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Fig. 2. Influence of cell number on cellular ELISA performance. MDCK cells seeded at 2.5 × 103, 5.0 ×103, 1.0 ×104, and 2.0 ×104 cells per well were infected with influenza A virus (2.5× 103 pfu/well). Following a 48 h incubation, cells were fixed and subject to ELISA. Each bar represents the mean of three replicates + / −1 SE.
3.1.4. Amount of 6iral inoculum and linearity of cellular ELISA Two important features of the cellular ELISA are linearity and sensitivity. MDCK cells were infected with viral inoculum at the range 1.88× 1021 –1.37× 105 pfu/ml prepared by serial 3-fold dilution of the virus stock (1× 108 pfu/ml). The ELISA conducted 24 h after infection generated a linear absorbance up to 4.57×104 pfu/ml (R 2 = 0.983, PB0.0002) as shown in Fig. 4. Above this number of virus particles, the cellular ELISA plateaued. The resolution of cellular ELISA estimated in the linear range was 0.01 A450/2.4 ×102 pfu. 3.1.5. Direct comparison of the cellular ELISA and the PA To test the potential of substituting the cellular ELISA for the PA, we compared both assays by performing them in parallel. The same viral inoculum and infection time were used to infect MDCK cells in either a 96-well (cellular ELISA)
or 6-well (PA) plate format. Results are shown in Fig. 5a. Both assays yielded linear and highly correlated data at the tested range of viral inoculum (R 2 = 0.997, PB 0.002; Fig. 5b).
3.1.6. Testing of influenza A 6irus susceptibility to BCTP using PRA and cellular ELISA The antiviral activity of BCTP versus Triton X-100 was investigated prior to comparing the PRA and cellular ELISA. Influenza A virus was treated with either BCTP or Triton X-100 and cellular ELISA was applied to assess virus infectivity after treatment. Triton X-100 showed a very strong antiviral activity, and therefore had to be diluted 50 times more than BCTP. As shown in Fig. 6, cellular ELISA was able to detect and quantitate antiviral effect of both agents. Subsequently, the BCTP was used to investigate if the cellular ELISA can generate comparable data with the PRA for testing antiviral agents. Aliquots of viral stock were mixed with three dilutions of BCTP to a final concentration of
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
1× 10 − 1, 1 × 10 − 2, 1 × 10 − 3 or with infection media (untreated virus) and incubated for 30 min at RT. After incubation, the inoculum was diluted with the infection medium either to 1.25× 104 pfu/ml for the cellular ELISA or to 2.0× 102 pfu/ml for the PRA. As shown in Table 1, the PRA and cellular ELISA reflected the same degree of virucidal activity of BCTP at concentrations of 1× 10 − 1 and 1× 10 − 2. Incubation of virus with BCTP at the concentration of 1×10 − 3 had little or no effect on virus infection capacity as measured by both the cellular ELISA and PRA assays.
4. Discussion Although PRA has been accepted as the standard for measuring the susceptibility of viruses to
171
antiviral agents, it is laborious, time consuming, relatively expensive, and requires subjective input by the reader. All these drawbacks have prompted the development of rapid and less laborious assays to assess the activity of antiviral agents. The MTT method (Pauwels et al., 1988) and a lactate dehydrogenase (LDH) detection technique (Watanabe et al., 1995) have been used to assess the antiviral activity of compounds against various viruses. Although MTT and LDH assays are inexpensive and easy to carry out, unfortunately they are not specific to virus, since they detect the virus indirectly assessing virus-induced cell damage. Both assays take at least 4 days to assess screening results and are not able to discriminate between cell damage caused by virus itself and by some other unrelated factors. For the screening of antiviral agents, more specific assays have been developed: against varicella-zoster virus (Berkow-
Fig. 3. Effect of incubation time on detection of viral proteins in cellular ELISA. Influenza A virus (2.5 × 103 pfu/well) infected MDCK cells (2.0×104 cells/well) were incubated for different periods of time (6, 12, 24, 48 h), cells were fixed and subjected to ELISA. Each point represents the mean of three replicates +/ −1 SE.
172
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Fig. 4. Standard curve of cellular ELISA. MDCK cells (2.0 × 104cells/well) were infected with influenza A virus in the range of 1.2× 102 – 6.5 × 104 pfu/ml (3-fold dilution). After a 24 h incubation, cells were fixed and subject to ELISA. Each point represents the mean of replicates + /− 1 SE.
itz and Levin, 1985), Herpes simplex virus (Rabalais et al., 1987; Leahy et al., 1994, 1996), Rauscher murine leukemia virus (Hollingshead et al., 1992). Although influenza A virus is a common respiratory pathogen and has been used widely as a model system to test antiviral agents in vitro (Huang et al., 1991; Mammen et al., 1995; Karaivanova and Spiro, 1998) and in vivo (Waghorn and Goa, 1998; Mendel et al., 1998; Smith et al., 1998), there has been no report describing the use of ELISA to screen susceptibility of influenza A virus to antiviral agents. In this study, we have optimized an in situ cellular ELISA for susceptibility testing of influenza A virus to antiviral agents. In an attempt to conduct in situ cellular ELISA to screen antiviral agents, several factors were examined including fixatives, cell number, virus adsorption time, post-infection incubation time and the amount of viral inoculum to optimize the performance of this assay. In the second part of this study, we compared directly the optimized cellular ELISA with PA and also compared the cellular ELISA and PRA.
Although it was possible to carry out the cellular ELISA in the absence of fixation, a significant loss of the monolayer can occur due to extensive washes. A variety of commonly used fixatives were tested in an effort to maximize the persistence of an intact cell sheet (Bishop and Hwang, 1992; Smith et al., 1997) and preserve virus antigenicity (Berkowitz and Levin, 1985; Leahy et al., 1994). We have tested 0.5% glutaraldehyde, 1% paraformaldehyde, and acetone–ethanol (1:1) mixture as these three fixatives work in different ways. Glutaraldehyde is a popular choice of fixative for the cellular ELISA and it provides excellent morphological preservation (Petris, 1978). Paraformaldehyde and formaldehyde are the most commonly used fixatives for immunochemistry because the antigen structure is maintained (Samoszuk et al., 1987; Harrach and Robenek, 1990). However, these fixatives do not preserve the ultrastructure and integrity of the cells as well as glutarldehyde. The mixture of acetone and ethanol was the third fixative tested. Acetone, ethanol, and methanol are known to cause extraction of membrane lipids and facilitate intracellular
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
antigen exposure (Kiernan, 1990). Of these three fixatives glutaraldehyde performed the worst, producing little resolution between infected and uninfected cells. One reason for this result may be the loss of antigenicity. It has been reported that many cell surface molecules lose their antigenicity upon fixation with glutaraldehyde (Cunningham et al., 1987; Leu et al., 1988; McLean et al., 1989). In addition, fixation with glutaraldehyde results in extensive crosslinking which greatly retards the binding of labeled antibodies to the antigenic determinants (Kraehenbuhl et al., 1977; Ostrand-Rosenberg et al., 1979; Van Ewijk et al., 1980). Thus, it is possible that the combination of these two factors contributed to the poor results obtained upon fixation with glutaraldehyde. Fixation with paraformaldehyde yielded better resolution than glutaraldehyde and gave the smallest standard error which suggests
173
that in our assay preservation of virus antigen integrity was more important than the integrity of cell structure. Using the acetone–ethanol mixture provided the best resolution between uninfected and infected cells and the highest absorbance. Although we do not have direct evidence, it is possible that the extraction of membrane lipids led to both better exposure of viral antigens inside the cells and improved penetration of antibodies used in ELISA, and therefore increased the sensitivity and resolution of the assay (Fig. 1). Interestingly, background absorbance levels remained unchanged regardless of the fixative which was used. This indicates that fixatives themselves do not have any effect on the performance of ELISA. Based on these data, the acetone–ethanol mixture was used as the fixative in our optimized cellular ELISA protocol.
Fig. 5. Linear regression of PA and cellular ELISA (a) and linear regression analysis of relationship between pfu counted in PA and viral antigen expression as detected in cellular ELISA (b). The same volumes of influenza A inoculum were overlaid on six well plates (PA) and on 96-well flat bottom microtiter plates (cellular ELISA) and incubated for 1 h at 37oC. The virus inoculum was aspirated and replaced with agarose overlay medium (PA) or with infection media +2% FBS (cellular ELISA). Assays were performed as described in Section 2. Each point represents the mean of three (PA) or six (cellular ELISA) replicates +/ −1 SE.
174
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Fig. 6. Inhibition of influenza A virus infectivity by Triton X-100 (A) or BCTP (B) treatments as measured using cellular ELISA. Assay was performed as described in Section 2. Each point represents the mean of three replicates +/ − 1 SE.
It has been reported that the number of cells used to seed the monolayer is also important in this type of assay, as too high a cell number can result in elevated background absorbance readings and poor sensitivity (Leahy et al., 1994). Since the number of cells at the time of assay is a
function of initial cell concentration and the time of incubation, we investigated the upper limit of initial cell concentration and incubation time after infection. It was found that cell concentration of 2.0× 104 cells/well yielded the highest absorbance readings and the best resolution (Fig. 2). Cells
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
seeded at greater concentrations tended to overgrow and detach from the wells as soon as 24 h after infection. The virus adsorption times tested (15, 30, and 60 min) did not contribute significantly to the cellular ELISA results. Nevertheless, it must be emphasized that different influenza A virus strains have different hemagglutinin affinity for sialic acid and therefore different adsorption and subsequent infection capacities (Weis et al., 1988). Moreover, the virus tropism also depends on the number of sialic acid residues expressed on cell surface glycoproteins and may modify cell susceptibility to infection. In our experiments, only one strain of influenza A virus (A/AA/6/60) and one cell line (MDCK) was used, therefore the optimal adsorption time and subsequent infection time may be different for other influenza A virus strains and host cell types. The reproductive cycle of viruses consists of three phases, and all of them may contribute directly or indirectly to cellular ELISA results. The first interval is known as the eclipse phase. During this phase the virus genome is exposed to host and viral machinery necessary for its expression but progeny virus production has not yet been started. This phase is followed by an interval Table 1 Inhibition of influenza A virus infection by BCTP as measured using PRA and cellular ELISAa
No virus Untreated virus Virus treated with BCTP: 1×10−1 1×10−2 1×10−3
a
PRA
Cellular ELISA
0.00 (+/−0.00)c 50.88d (+/−0.25)
0.269b (+/−0.002) 1.856 (+/−0.050)
0.00 (+/−0.00) 0.00 (+/−0.00) 53.11 (+/−0.74)
0.258 (+/−0.012) 0.261 (+/−0.008) 1.392 (+/−0.103)
1.25×104 pfu/ml was used in cellular ELISA as it was within the linear portion of the curve (see Fig. 4). b Absorbance reading A(450nm). c Standard error. d Number of plaques.
175
in which progeny virus accumulates in the cell at an exponential rate. This interval is known as the maturation phase. After several hours infected cells cease all metabolic activity, lose their integrity and enter the lytic phase (Roizman and Palese, 1996). Therefore, the optimal phase of virus reproductive cycle to measure viral proteins in cellular ELISA is the late maturation phase before cells enter the lytic phase. Since virus reproductive cycles may vary from a few to several hours, it was important to determine the optimal time of post-infection incubation. Cells were incubated from 6 to 48 h after infection followed by cellular ELISA. At 6 h there was no increase in absorbance coinciding with the eclipse phase of virus reproduction cycle. The highest absorbance was observed at 48 h incubation period but the virus had entered the early lytic phase and reproducibility of the cellular ELISA was poor (Fig. 3). After 48 h incubation, we observed gross lysis of infected cells which had a detrimental effect on the assay generating, in some cases, background level absorbance (data not shown). Based on these data, we chose 24 h as the optimal, post-infection period for the in situ cellular ELISA protocol. Two important features of quantitative ELISA are linearity (Berkowitz and Levin, 1985), and sensitivity (Fulton et al., 1988). A linear relationship between virus inoculum and absorbance was observed at the range 1.88× 102 –4.57× 104 pfu/ ml (R 2 = 0.983, PB 0.0002). Above this number of virus particles, cellular ELISA readings plateaued (Fig. 4). After developing the cellular ELISA protocol, we compared directly the cellular ELISA and PA by running these two assays in parallel. Both assays yielded linear and highly correlated data however, we could directly compare only four out of 11 virus 3-fold dilutions. Dilutions higher than 1.88× 10 − 6, were the lower threshold of the cellular ELISA and at dilutions lower than 5.08× 10 − 5, the number of plaques/well became too numerous to count (Fig. 5). Leahy et al. (1994), reported that infection of cells in suspension resulted in greater absorbance values than inoculation of cell monolayers as a result of more rapid viral growth and improved the ability to measure dose-response. It is quite likely that infection of
176
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
cells in cell suspension could increase the sensitivity of our cellular ELISA as well, but we have not explored such possibility. Infection of the cells in suspension would introduce substantial variability to the experiment and thereby it would not make possible to compare both assays in parallel. The BCTP has been reported to have an extensive bactericidal (Hamouda et al. 1998a,b), sporicidal (Hayes et al. 1998), virucidal and fungistatic effects (Hamouda et al. 1998b). Since BCTP is a new class of anti-microbial agent, we compared its activity with Triton X-100. Antiviral activity of both agents was detected by cellular ELISA (Fig. 6). Subsequently, we applied the BCTP in both cellular ELISA and PRA assays to test the possibility of using cellular ELISA to screen antiviral agents. The cellular ELISA and PRA reflected a similar virucidal effect of BCTP on influenza A virus (Table 1). However, it is possible that some antiviral agents may generate discrepancies between these two assays as they are based on different principles. Nevertheless, the advantages of cellular ELISA, including the 96-well format and ability to automate, make this assay very convenient for the rapid screening of a large number of antiviral agents in a broad range of concentrations. Although in this report we described the cellular ELISA technique to detect H2N2 strain of influenza A virus, the same technique and antibody detection were used to assess influenza A infectivity of X-31 (H3N1) strain and it was found that the technique worked equally well (data not shown). We believe that by using other detecting antibodies directed to different influenza A and B strains, this technique would have more general application than reported in this paper.
Acknowledgements We thank Brian W. Donovan for assistance in plaque reduction assays. We also thank Michael M. Hayes and Dr Jon D. Reuter for their assistance in propagation of influenza A virus. Funding for the study was obtained from DARPA contract MDA 972-97-1-0007.
References Baker J.R., Jr., Wright D.C., Hamouda T., Hayes M.M., Brisker J., 1997. US Patent pending. Barrett, T., Inglis, S.C., 1985. Growth purification and titration of influenza viruses. In: Mahy, W.J. (Ed.), Virology: A Practical Approach. IRL Press, Washington, DC, pp. 119 – 151. Berkowitz, F.E., Levin, M.J., 1985. Use of enzyme-linked immunosorbent assay performed directly on fixed infected cell monolayers for evaluating drugs against varicellazoster virus. Antimicrob. Agents Chemother. 28, 207 – 210. Bishop, G.A., Hwang, J., 1992. Use of a cellular ELISA for the detection of cell surface antigens. BioTechniques 12, 326 – 330. Cunningham, C., Power, D.A., Innes, A., Lind, T., Catto, G.R.D., 1987. Maternal antibody responses during early pregnancy detected by a cellular enzyme linked immunospecific assay. Hum. Immunol. 19, 7 – 16. Fulton, R.E., Wong, J.P., Siddiqui, Y.M., Tso, M.-S., 1988. Sensitive fluorogenic enzyme immunoassay on nitrocellulose membranes for quantitation of virus. J. Virol. Methods 22, 149 – 164. Hamouda T., Wright D.C., Brisker J.M., Baker J.R. Jr., 1998a. Microbiocidal Effects of Liposome-like Microemulsions on Pathogenic Gram Negative Bacteria. American Society for Microbiology, 98th General Meeting, May 17 – 21, 1998, Atlanta, GA. Abstract c A-52. Hamouda T., Hayes M.M., Anderson M.J., Cao Z., Myc A., Reuter J.D., Donovan B., Johnosn K.J., Wright D.C., Brisker J., Baker J.R., Jr., 1998b. Broad spectrum antimicrobial activity of a novel surfactant nanoemulsion (personal communication). Harrach, B., Robenek, H., 1990. Polyclonal antibodies against formaldehyde-modified apolipoprotein A-I. An approach to circumventing fixation-induced loss of antigenicity in immunocytochemistry. Arteriosclerosis 10, 564 – 576. Hayden F.G., Cote K.M., Douglas G. Jr., 1980. Plaque inhibition assay for drug susceptibility testing of influenza viruses. Antimicrob. Agents Chemother. 17, 865 – 870. Hayes M.M., Cao Z., Chung Y., Wright D.C., Brisker J.M., Baker J.R. Jr., 1998. Biocidal effect of unique lipid surfactants on Bacillus spores. American Society for Microbiology, 98th General Meeting, May 17 – 21, 1998, Atlanta, GA. Abstract c A-49. Herlocher, M.L., Clavo, A.C., Maassab, H.F., 1996. Sequence comparison of A/AA/6/60 influenza viruses: mutations which may contribute to attenuation. Virus Res. 42, 11 – 25. Hollingshead, M.G., Westbrook, L., Ross, M.J., Bailey, J., Quallas, K.J., Allen, L.B., 1992. An ELISA system for evaluating antiretroviral activity against Rauscher murine leukemia virus. Antiviral Res. 18, 267 – 274. Hosoya, M., Matuyama, S., Baba, M., Suzuki, H., Shigeta, S., 1992. Effects of proteinase inhibitors on replication of various myxoviruses. Antimicrob. Agents Chemother. 36, 1432 – 1436.
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177 Huang, R., Dietsch, E., Lockhoff, O., Schuller, M., Reutter, W., 1991. Antiviral activity of some natural and synthetic sugar analogues. FEBS Lett. 291, 199–202. Kiernan, J.A., 1990. Fixation. In: Alan, J. (Ed.), Histological and Histochemical Methods: Theory and Practice. Pergamon Press, Oxford, UK; New York, pp. 25–27. Kraehenbuhl, J.P., Racine, L., Jamieson, J.D., 1977. Immunocytochemical localization of secretory proteins in bovine pancreatic exocrine cells. J. Cell Biol. 72, 406–423. Karaivanova, V.K., Spiro, R.G., 1998. Sulphation of N-linked oligosaccharides of vesicular stomatitis and influenza virus envelope glycoproteins: host cell specificity, subcellular localization and identification of substituted saccharides. Biochem. J. 329, 511 –518. Leahy, B.J., Christiansen, K.J., Shellam, G., 1994. Standardization of a microplate in situ ELISA (MISE-test) for the susceptibility testing of herpes simplex virus to acyclovir. J. Virol. Methods 48, 93–108. Leahy, B.J., Christiansen, K.J., Collins, P., 1996. A blinded comparison of two methods of viral susceptibility testing: plaque reduction assay versus microplate in situ ELISA. J. Virol. Methods 56, 85–90. Leu, R.W., Robinson, C.J., Wiggins, J.A., Shannon, B.J., Rummage, J.A., Horn, M.J., 1988. Photometric assays for FcR1-dependent cellular cytotoxicity mediated by monomeric IgGg2a in murine peritoneal macrophages. J. Immunol. Methods 113, 269–278. Mammen, M., Dahmann, G., Whitesides, G.M., 1995. Effective inhibitors of hemagglutination by influenza virus synthesized from polymers having active ester groups. Insight into mechanism of inhibition. J. Med. Chem. 38, 4179– 4190. McLean, L., Perrett, D., Winrow, V.R., Archer, J.R., 1989. Status of an unpaired thiol group on the HLA-B27 epitope. Clin. Exp. Immunol. 77, 417–421. Mendel, D.B., Tai, C.Y., Escarpe, P.A., et al., 1998. Oral administration of a prodrug of the influenza virus neuraminidase inhibitor GS 4071 protects mice and ferrets against influenza infection. Antimicrob. Agents Chemother. 42, 640 –646. Ostrand-Rosenberg, S., Edidin, M., Wetzel, B., 1979. Differential localization of H-2 and teratoma-defined antigens on mouse L cell fibroblasts. In: Becker, R.P., Johary, O. (Eds.), Scanning Electronic Microscopy, vol. 3. Scanning Electron Microscopy, AMF Ohare, IL, pp. 595–600. Pauwels, R., Balzarini, J., Baba, M., Snoeck, R., Schols, D., Herdewijn, P., Desmyter, J., De Clercq, E., 1988. Rapid and automated tetrazolum-based colorimetric assay for the detection of anti-HIV compounds. J. Virol. Methods 20, 309 – 321. Petris, D.E., 1978. Immunoelectron microscopy and immu-
.
177
nofluorescence in membrane biology. In: Korn, E. (Ed.), Methods in Membrane Biology, vol. 9. Plenum Press, New York, pp. 1 – 177. Rabalais, G.P., Levin, M.J., Berkowitz, F.E., 1987. Rapid herpes simplex virus susceptibility testing using an enzymelinked immunosorbent assay performed in situ on fixed virus-infected monolayers. Antimicrob. Agents Chemother. 31, 946 – 948. Reuter J.D., Myc A., Cao Z., Johnson K.J., Wright D.C., Brisker J., Baker J.R. Jr., 1998. Prevention of influenza A virus infection by non-ionic surfactant nanoemulsions in a murine model. Antimicrob. Agents Chemother. (submitted). Roizman, B., Palese, P., 1996. Multiplication of viruses: an overview. In: Fields, B.N., Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, vol. 1. Lippincott-Raven Publishers, Philadelphia, New York, pp. 101 – 111. Samoszuk, M., Sholly, S., Epstein, A.L., 1987. Screening for anti-idiotypic monoclonal antibodies on paraformaldehyde-fixed lymphoma cells. Hybridoma 6, 605 – 610. Smith, D.D., Cohick, C.B., Lindsley, H.B., 1997. Optimization of cellular ELISA for assay of surface antigens on human synoviocites. BioTechniques 22, 952 – 957. Smith, P.W., Sollis, S.L., Howes, P.D., et al., 1998. Dihydropyrancarboxamides related to zanamivir: a new series of inhibitors of influenza virus sialidases. 1. Discovery, synthesis biological activity, and structure-activity relationships of 4-guanidino- and 4-amino-4H-pyran-6-carboxamides. J. Med. Chem. 41, 787 – 797. Takeuchi, H., Baba, M., Shigeta, S., 1991. An application of tetrazolum (MTT) colorimetric assay for the screening of anti-herpes virus compounds. J. Virol. Methods 33, 61 – 71. Van Ewijk, W., Coffman, R.C., Weissman, I.L., 1980. Immunoelectron microscopy of cell surface antigens: a quantitative analysis of antibody binding after different fixation protocols. Histochem. J. 12, 349. Waghorn, S.L., Goa, K.L., 1998. Zanamvir. Drugs 55, 721 – 725. Watanabe, W., Konno, K., Ijichi, H., Inoune, T., Yokota, T., Shigeta, S., 1994. MTT colorimetric assay system for the screening of anti-orthomyxo- and anti-paramyxoviral agents. J. Virol. Methods 48, 257 – 265. Watanabe, W., Sudo, K., Asawa, S., Konno, K., Yokota, T., Shigeta, S., 1995. Use of lactate dehydrogenase to evaluate the anti-viral activity against influenza A virus. J. Virol. Methods 51, 185 – 192. Weis, W., Brown, J.H., Cusack, S., Paulson, J.C., Skehel, J.J., Wiley, D.C., 1988. Structure of influenza virus hemagglutinin complexed with its receptor, sialic acid. Nature 333, 426 – 431.
.
Optimization of in situ cellular ELISA performed on influenza A virus-infected monolayers for screening of antiviral agents Andrzej Myc a, Marie J. Anderson b, James R. Baker Jr. a,* a
Department of Internal Medicine, Di6ision of Allergy, Room 9220 MSRB III, Uni6ersity of Michigan, 1150 West Medical Center Dri6e, Ann Arbor, MI 48109 -0648, USA b Markey Molecular Medicine Center, Box 357720, Uni6ersity of Washington, Seattle, WA 98195 -7720, USA Received 27 July 1998; received in revised form 20 October 1998; accepted 20 October 1998
Abstract Viral susceptibility testing has been traditionally performed by the plaque reduction assay (PRA) which is laborious, time consuming, relatively expensive, and requires subjective input by the reader. An in situ cellular enzyme-linked immunosorbent assay (ELISA) has been developed with the potential to overcome many of the limitations of PRA and has been applied to a variety of viruses. This study establishes the specific conditions necessary for susceptibility testing of influenza A virus to antiviral agents such as amount of inoculum size, duration of incubation, fixative type, and cell number; factors which are critical to the performance of the in situ cellular ELISA. In situ cellular ELISA was found to correlate strongly with the plaque assay (PA) (R 2 = 0.997, PB 0.002). Both assays were applied to test the susceptibility of influenza A virus to a new antiviral emulsion agent and yielded comparable data. The optimized in situ cellular ELISA can serve as a reliable assay for the rapid screening of large numbers of antiviral agents. © 1999 Elsevier Science B.V. All rights reserved. Keywords: ELISA; Plaque reduction assay; Influenza A virus; Susceptibility; MDCK cells
1. Introduction Although the plaque reduction assay (PRA) has been accepted as the gold standard for measuring * Corresponding author. Tel.: +1-734-6472777; fax: +1734-9362990; e-mail: [email protected].
the susceptibility of viruses to antiviral agents, it is not convenient for screening a large numbers of antiviral agents. This is because PRA is laborious, time consuming, relatively expensive, and requires subjective input by the reader. Moreover, the virus detection range is very narrow (0–100 pfu), and this requires accurate predetermination of
0166-0934/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 0 9 3 4 ( 9 8 ) 0 0 1 5 0 - 5
166
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
infectious titer prior to undertaking susceptibility testing. All these drawbacks have prompted the development of rapid and less laborious assays to assess the activity of antiviral agents. These assays should be quantitative, overcome the limitations of the PRA, and yet be correlative with the traditional method. The MTT method has been used to assess the antiviral activity of compounds against the human immunodeficiency virus (Pauwels et al., 1988), herpes simplex virus type 1 (HSV-1) (Takeuchi et al., 1991), and influenza virus A (Fluv. A) (Hosoya et al., 1992; Watanabe et al., 1994) in cell system in vitro. However, this method is limited to these particular viruses because it measures the mitochondrial dehydrogenase activity to assess virus-induced cell damage. It cannot be utilized for viruses that do not grow well in culture or for cells with low enzymatic activity. An antiviral screening system using a lactate dehydrogenase (LDH) detection technique has been also reported (Watanabe et al., 1995). This system can be used to detect the infectivity of viruses that cannot be detected adequately by the MTT method. Although, both MTT and LDH assays are very easy to carry out, they are not specific to virus and detect the virus indirectly by assessing virus-induced cell damage. To study antiviral compounds, more specific assays have been developed and applied to offer better quantitation of virus. Berkowitz and Levin (1985) reported the enzyme-linked immunosorbent assay (ELISA) for evaluating drugs against varicella-zoster virus (Berkowitz and Levin, 1985). An ELISA has also been established for susceptibility testing of herpes simplex virus to acyclovir (Rabalais et al., 1987; Leahy et al., 1994, 1996). Hollingshead et al. utilized ELISA system for evaluating antiretroviral activity against Rauscher murine leukemia virus (Hollingshead et al., 1992). Influenza A virus, a common respiratory pathogen, was selected for this study as a model to test antiviral agents. To our knowledge there has been no report describing the conditions of in situ cellular ELISA performed on influenza A virus infected cells. We optimized the critical factors of the in situ cellular ELISA to make it comparable with PRA to quantitate influenza A virus. Advantages of the cellular ELISA over the PRA include:
(i) utilization of the 96-well format, which facilitates measurements in replicate, requires a small number of cells, and is adaptable to automation; (ii) less demanding which conserves costs, time, and labor; (iii) more quantitative; (iv) less subjective as it does not require the reader’s input. 2. Materials and methods
2.1. Virus Influenza virus A/AA/6/60, H2N2 (Herlocher, et al., 1996) was kindly provided by Hunein Maassab (School of Public Health, University of Michigan, Ann Arbor, MI). Influenza A virus was propagated in the allantoic cavities of fertilized pathogen-free hen eggs (SPAFAS, Norwich, CT) using standard methods described elsewhere (Barrett and Inglis, 1985). Virus stock was kept in aliquots (108 pfu/ml) of infectious allantoic fluids at − 80°C.
2.2. Cells Madin Darby canine kidney (MDCK) cells were purchased from the American Type Culture Collection (ATCC, Rockville, MD).
2.3. Media 2.3.1. Maintenance medium Cells were maintained in Eagle’s minimal essential medium with Earle’s salts, 2 mM L-glutamine, and 1.5 g/l sodium bicarbonate (Mediatech, Herndon, VA) containing 10% fetal bovine serum (FBS, Hyclone Laboratories, Logan, UT) supplemented with 0.1 mM non-essential amino acids, 1.0 mM sodium pyruvate, 100U penicillin/ml, streptomycin 100 mg/ml (Life Technologies, Gaithersburg, MD). 2.3.2. Infection medium Maintenance media (without fetal bovine serum) was formulated with 3.0 mg/ml of trypsin treated with tolylsulfonyl phenylalanyl chloromethyl ketone (TPCK-treated trypsin; Worthington Biochemical Corporation, Lakewood, NJ).
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
2.3.3. O6erlay medium Overlay medium consisted of equal amounts of 2X infection medium and 1.6% SeaKem ME agarose (FMC BioProducts, Rockland, MD). Staining agarose overlay medium consisted of agarose overlay medium plus 0.01% neutral red solution (Life Technologies, Gaithersburg, MD) without TPCK-treated trypsin. 2.4. Plaque and plaque reduction assays To directly compare the PA and cellular ELISA, MDCK cells were seeded at 2×105 cells/ well in 6-well Falcon plates (Becton Dickenson and Company, Franklin Lakes, NJ) and incubated at 37oC/5% CO2 for 3 days. Cells were washed free of complete media prior to use. Twofold dilutions of influenza A virus were made in infection medium. The cells were infected with 100 ml of the dilutions (for direct comparison with the cellular ELISA) and supplemented with an additional 400 ml of infection medium to prevent the drying of cell monolayers. The cells were incubated at 37oC/5% CO2 for 1 h to allow the virus to adsorb. The inoculum was removed and 1 ml of agarose overlay medium was added to each well. The plates were incubated at 37oC/5% CO2. Plaques appeared approximately 48 h post-infection and were stained with staining agarose overlay medium. The plates were returned to incubation at 37oC/5% CO2. Plaques were counted 6–12 h after staining. The plaque reduction assay was performed with a modification of the method described by Hayden et al. (1980). MDCK cells were seeded at 1× 105 cells/well in 12-well Falcon® plates and incubated at 37oC/5% CO2 for 3 days. Approximately 1×108 pfu of influenza A virus was incubated with BCTP as described below. Controls included influenza A virus plus infection medium and BCTP plus infection medium. The influenza A virus-BCTP treatments and controls were diluted in infection medium to contain 30 – 100 pfu/ 250 ml or the same concentration of BCTP. The three remaining wells served as cell controls and 250 ml infection medium/well was added. Confluent cell monolayers were inoculated in triplicate on three plates and incubated at 37oC/5% CO2 for
167
1 h. The inoculum/medium was aspirated and 1 ml of agarose overlay medium/well was added and plates were incubated at 37oC/5% CO2 until plaques appeared. Monolayers were stained with the agarose overlay medium and incubation was continued at 37oC/5% CO2. Plaques were counted 6–12 h after staining. The average plaque count from nine wells for each BCTP concentration was plotted against the average plaque count of untreated virus.
2.5. In situ cellular enzyme-linked immunosorbent assay To detect and quantitate viral proteins in MDCK cells infected with influenza A virus, the in situ cellular ELISA was developed based on the principles described by Berkowitz and Levin, 1985. The critical factors of the assay: the fixatives, time of virus adsorption, duration of incubation after infection, cell number, and number of virus particles were examined. MDCK cells in 100 ml complete medium were seeded on flat-bottom 96-well microtiter plates precoated with 0.5% gelatin and incubated overnight. On the next day, the culture medium was removed and cells were washed with maintenance medium. A total of 100 ml of viral inoculum was added to the wells and incubated for an indicated period of time. The viral inoculum was removed and replaced with 100 ml of infection medium + 2% FBS. Infected MDCK cells were incubated for an additional 6–48 h, as necessary, and medium was aspirated. The cells were washed once with PBS and fixed with one of the fixatives as described below. On the day of assay, fixed cells were washed with PBS and blocked with 1% dry milk in PBS for 30 min at 37°C. One hundred ml of ferret anti-influenza A virus polyclonal antibody (kindly provided by Dr Hunein Maassab) at 1× 10 − 3 dilution was added to the wells for 1 h at 37°C. The cells were washed four times with washing buffer (PBS and 0.05% Tween-20), and incubated with 100 ml at 1 × 10 − 3 dilution of goat anti-ferret peroxidase conjugated antibody (Kirkegaard and Perry Laboratories, Gaithersburg, MA) for 30 min at 37°C. Cells were washed four times and incubated with 100 ml of 1-Step™ Turbo TMB-ELISA substrate (Pierce
168
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Rockford, IL) until color was developed. The reaction was stopped with 1 N sulfuric acid and plates were read at a wavelength of 450 nm in ELISA reader.
2.6. Fixation procedure MDCK cells were washed once with PBS and were subjected to fixation with one of the following fixatives: acetone – ethanol (1:1) mixture, 0.5% glutaraldehyde, or 1% paraformaldehyde. One hundred microlitres of − 20oC acetone – ethanol mixture was added to each well on ice. Cells were fixed at −20oC for at least 15 min. One hundred microlitres of glutaraldehyde was added to each well and the plate was incubated for 30 min at room temperature (RT). Glutaraldehyde was aspirated and reactive aldehyde groups were blocked with PBSLE (PBS+ 100 mM lysine+100 mM ethanolamine) to avoid the covalent binding of proteins added at subsequent steps. The cells were washed with PBS, dried, and stored at 4oC. One hundred microlitres of paraformladehyde was added to each well on ice and incubated for 15 min then washed with PBS, dried, and stored at 4°C.
2.7. Influenza A treatment with BCTP or Triton X-100 In antiviral activity testing we evaluated BCTP, a new anti-microbial nanoemulsion disinfectant (Novavax, Rockville, MD), described elsewhere (Baker et al., 1997; Reuter et al., 1998) or a commonly available antiviral disinfectant, Triton X-100. Prior to assessing the efficacy of the nanoemulsion or Triton X-100 in our cellular ELISA, the direct cytotoxicity of BCTP and Triton X-100 on MDCK cells was determined by microscope inspection and MTT assay. Dilutions of a mixture of each virus and either the nanoemulsion or Triton X-100 applied to the cellular ELISA were always higher than the determined non-toxic concentration. Approximately 1× 108 pfu of influenza A virus were incubated either with BCTP or Triton X-100 at final concentrations of 1× 10 − 1, 1×10 − 2, 1× 10 − 3 or 5× 10 − 2, 5×10 − 3, 5×10 − 4, respectively for 30
min. at RT on a shaker. After incubation, serial dilutions of the BCTP/virus or Triton X-100/virus mixtures were made in infection medium and overlaid on MDCK cells to carry out either the cellular ELISA, PRA or both assays as described above.
3. Results
3.1. Optimization of cellular ELISA To optimize the in situ cellular ELISA, the following variables were examined: fixatives, virus adsorption time, post-infection incubation time, cell number, and amount of viral inoculum.
3.1.1. Fixati6es To test the effect of fixatives on the cellular ELISA performance three commonly used fixatives were compared: glutaraldehyde, paraformaldehyde, and the acetone–ethanol mixture. Although all three fixatives preserved the cell monolayer, as observed by microscope examination, different ELISA results were obtained (Fig. 1). Glutaraldehyde performed very poorly, producing little differentiation between infected and uninfected cells. Fixation with paraformaldehyde yielded greater resolution than glutaraldehyde and gave the smallest standard error. Using the acetone–ethanol mixture we obtained the best difference between uninfected and infected cells and the highest overall absorbance values. Based on these data, the acetone–ethanol mixture was used as the fixative in the succeeding experiments. 3.1.2. Effect of cell number on the performance of cellular ELISA We also investigated the upper limit of initial cell concentration and post-infection incubation time useful in the assay. Four different concentrations of cells in the range of 2.5× 103 –2.0×104 per well were used to seed the monolayer. Twenty four hours later the cells were infected with virus (2.5× 103pfu/well) and incubated for an additional 48 h. The highest cell concentration tested (2× 104 cells/well) yielded the highest absorbance readings and the best resolution. The three other
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
cell numbers tested: 2.5×103, 5.0 × 103, and 1.0×104 per well gave approximately 50% lower absorbance readings (Fig. 2). Cells at a concentration higher than 2.0×104 per well tended to overgrow and detach from the wells causing poor reproducibility of the assay (data not shown).
3.1.3. Virus adsorption and duration of post-infection incubation of MDCK cells Since cellular ELISA performance depends on efficient virus infection of plated cells, the effect of virus adsorption time on resolution and reproducibility of the assay was investigated. MDCK cells were incubated with influenza A virus for 15, 30, and 60 min. After each time point, the infection medium was replaced with virus free infection medium+2%FBS, and the cells were incubated for an additional 24 h prior to ELISA. There was not a significant difference in the
169
ELISA results in respect to the time of virus incubation on MDCK cells (data not shown). Virus reproductive cycles vary from a few to several hours, which may modify the cellular ELISA performance. Thus, it was important to estimate the optimal time of post-infection incubation for influenza A virus infected cells to obtain a highly reproducible ELISA. Cells were incubated from 6 to 48 h after infection followed by the cellular ELISA. Absorbance increase was logarithmic in the range over a 6–48 h incubation period (Fig. 3). The 6 h incubation period did not yield any increase in absorbance over baseline level. The highest absorbance was observed at the 48-h incubation period, but the reproducibility of the ELISA was poor due to cell lysis. Based on these data, the 24 h incubation time was chosen as the optimal, post-infection period for the in situ cellular ELISA protocol.
Fig. 1. Effect of fixatives on cellular ELISA performance. MDCK cells (104 cells/well) infected with influenza A virus (2.5 ×106 pfu/ml) were incubated for 24 h, fixed with the three different fixatives and subjected to ELISA. Each bar represents the mean of three replicates + /− 1 SE.
170
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Fig. 2. Influence of cell number on cellular ELISA performance. MDCK cells seeded at 2.5 × 103, 5.0 ×103, 1.0 ×104, and 2.0 ×104 cells per well were infected with influenza A virus (2.5× 103 pfu/well). Following a 48 h incubation, cells were fixed and subject to ELISA. Each bar represents the mean of three replicates + / −1 SE.
3.1.4. Amount of 6iral inoculum and linearity of cellular ELISA Two important features of the cellular ELISA are linearity and sensitivity. MDCK cells were infected with viral inoculum at the range 1.88× 1021 –1.37× 105 pfu/ml prepared by serial 3-fold dilution of the virus stock (1× 108 pfu/ml). The ELISA conducted 24 h after infection generated a linear absorbance up to 4.57×104 pfu/ml (R 2 = 0.983, PB0.0002) as shown in Fig. 4. Above this number of virus particles, the cellular ELISA plateaued. The resolution of cellular ELISA estimated in the linear range was 0.01 A450/2.4 ×102 pfu. 3.1.5. Direct comparison of the cellular ELISA and the PA To test the potential of substituting the cellular ELISA for the PA, we compared both assays by performing them in parallel. The same viral inoculum and infection time were used to infect MDCK cells in either a 96-well (cellular ELISA)
or 6-well (PA) plate format. Results are shown in Fig. 5a. Both assays yielded linear and highly correlated data at the tested range of viral inoculum (R 2 = 0.997, PB 0.002; Fig. 5b).
3.1.6. Testing of influenza A 6irus susceptibility to BCTP using PRA and cellular ELISA The antiviral activity of BCTP versus Triton X-100 was investigated prior to comparing the PRA and cellular ELISA. Influenza A virus was treated with either BCTP or Triton X-100 and cellular ELISA was applied to assess virus infectivity after treatment. Triton X-100 showed a very strong antiviral activity, and therefore had to be diluted 50 times more than BCTP. As shown in Fig. 6, cellular ELISA was able to detect and quantitate antiviral effect of both agents. Subsequently, the BCTP was used to investigate if the cellular ELISA can generate comparable data with the PRA for testing antiviral agents. Aliquots of viral stock were mixed with three dilutions of BCTP to a final concentration of
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
1× 10 − 1, 1 × 10 − 2, 1 × 10 − 3 or with infection media (untreated virus) and incubated for 30 min at RT. After incubation, the inoculum was diluted with the infection medium either to 1.25× 104 pfu/ml for the cellular ELISA or to 2.0× 102 pfu/ml for the PRA. As shown in Table 1, the PRA and cellular ELISA reflected the same degree of virucidal activity of BCTP at concentrations of 1× 10 − 1 and 1× 10 − 2. Incubation of virus with BCTP at the concentration of 1×10 − 3 had little or no effect on virus infection capacity as measured by both the cellular ELISA and PRA assays.
4. Discussion Although PRA has been accepted as the standard for measuring the susceptibility of viruses to
171
antiviral agents, it is laborious, time consuming, relatively expensive, and requires subjective input by the reader. All these drawbacks have prompted the development of rapid and less laborious assays to assess the activity of antiviral agents. The MTT method (Pauwels et al., 1988) and a lactate dehydrogenase (LDH) detection technique (Watanabe et al., 1995) have been used to assess the antiviral activity of compounds against various viruses. Although MTT and LDH assays are inexpensive and easy to carry out, unfortunately they are not specific to virus, since they detect the virus indirectly assessing virus-induced cell damage. Both assays take at least 4 days to assess screening results and are not able to discriminate between cell damage caused by virus itself and by some other unrelated factors. For the screening of antiviral agents, more specific assays have been developed: against varicella-zoster virus (Berkow-
Fig. 3. Effect of incubation time on detection of viral proteins in cellular ELISA. Influenza A virus (2.5 × 103 pfu/well) infected MDCK cells (2.0×104 cells/well) were incubated for different periods of time (6, 12, 24, 48 h), cells were fixed and subjected to ELISA. Each point represents the mean of three replicates +/ −1 SE.
172
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Fig. 4. Standard curve of cellular ELISA. MDCK cells (2.0 × 104cells/well) were infected with influenza A virus in the range of 1.2× 102 – 6.5 × 104 pfu/ml (3-fold dilution). After a 24 h incubation, cells were fixed and subject to ELISA. Each point represents the mean of replicates + /− 1 SE.
itz and Levin, 1985), Herpes simplex virus (Rabalais et al., 1987; Leahy et al., 1994, 1996), Rauscher murine leukemia virus (Hollingshead et al., 1992). Although influenza A virus is a common respiratory pathogen and has been used widely as a model system to test antiviral agents in vitro (Huang et al., 1991; Mammen et al., 1995; Karaivanova and Spiro, 1998) and in vivo (Waghorn and Goa, 1998; Mendel et al., 1998; Smith et al., 1998), there has been no report describing the use of ELISA to screen susceptibility of influenza A virus to antiviral agents. In this study, we have optimized an in situ cellular ELISA for susceptibility testing of influenza A virus to antiviral agents. In an attempt to conduct in situ cellular ELISA to screen antiviral agents, several factors were examined including fixatives, cell number, virus adsorption time, post-infection incubation time and the amount of viral inoculum to optimize the performance of this assay. In the second part of this study, we compared directly the optimized cellular ELISA with PA and also compared the cellular ELISA and PRA.
Although it was possible to carry out the cellular ELISA in the absence of fixation, a significant loss of the monolayer can occur due to extensive washes. A variety of commonly used fixatives were tested in an effort to maximize the persistence of an intact cell sheet (Bishop and Hwang, 1992; Smith et al., 1997) and preserve virus antigenicity (Berkowitz and Levin, 1985; Leahy et al., 1994). We have tested 0.5% glutaraldehyde, 1% paraformaldehyde, and acetone–ethanol (1:1) mixture as these three fixatives work in different ways. Glutaraldehyde is a popular choice of fixative for the cellular ELISA and it provides excellent morphological preservation (Petris, 1978). Paraformaldehyde and formaldehyde are the most commonly used fixatives for immunochemistry because the antigen structure is maintained (Samoszuk et al., 1987; Harrach and Robenek, 1990). However, these fixatives do not preserve the ultrastructure and integrity of the cells as well as glutarldehyde. The mixture of acetone and ethanol was the third fixative tested. Acetone, ethanol, and methanol are known to cause extraction of membrane lipids and facilitate intracellular
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
antigen exposure (Kiernan, 1990). Of these three fixatives glutaraldehyde performed the worst, producing little resolution between infected and uninfected cells. One reason for this result may be the loss of antigenicity. It has been reported that many cell surface molecules lose their antigenicity upon fixation with glutaraldehyde (Cunningham et al., 1987; Leu et al., 1988; McLean et al., 1989). In addition, fixation with glutaraldehyde results in extensive crosslinking which greatly retards the binding of labeled antibodies to the antigenic determinants (Kraehenbuhl et al., 1977; Ostrand-Rosenberg et al., 1979; Van Ewijk et al., 1980). Thus, it is possible that the combination of these two factors contributed to the poor results obtained upon fixation with glutaraldehyde. Fixation with paraformaldehyde yielded better resolution than glutaraldehyde and gave the smallest standard error which suggests
173
that in our assay preservation of virus antigen integrity was more important than the integrity of cell structure. Using the acetone–ethanol mixture provided the best resolution between uninfected and infected cells and the highest absorbance. Although we do not have direct evidence, it is possible that the extraction of membrane lipids led to both better exposure of viral antigens inside the cells and improved penetration of antibodies used in ELISA, and therefore increased the sensitivity and resolution of the assay (Fig. 1). Interestingly, background absorbance levels remained unchanged regardless of the fixative which was used. This indicates that fixatives themselves do not have any effect on the performance of ELISA. Based on these data, the acetone–ethanol mixture was used as the fixative in our optimized cellular ELISA protocol.
Fig. 5. Linear regression of PA and cellular ELISA (a) and linear regression analysis of relationship between pfu counted in PA and viral antigen expression as detected in cellular ELISA (b). The same volumes of influenza A inoculum were overlaid on six well plates (PA) and on 96-well flat bottom microtiter plates (cellular ELISA) and incubated for 1 h at 37oC. The virus inoculum was aspirated and replaced with agarose overlay medium (PA) or with infection media +2% FBS (cellular ELISA). Assays were performed as described in Section 2. Each point represents the mean of three (PA) or six (cellular ELISA) replicates +/ −1 SE.
174
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
Fig. 6. Inhibition of influenza A virus infectivity by Triton X-100 (A) or BCTP (B) treatments as measured using cellular ELISA. Assay was performed as described in Section 2. Each point represents the mean of three replicates +/ − 1 SE.
It has been reported that the number of cells used to seed the monolayer is also important in this type of assay, as too high a cell number can result in elevated background absorbance readings and poor sensitivity (Leahy et al., 1994). Since the number of cells at the time of assay is a
function of initial cell concentration and the time of incubation, we investigated the upper limit of initial cell concentration and incubation time after infection. It was found that cell concentration of 2.0× 104 cells/well yielded the highest absorbance readings and the best resolution (Fig. 2). Cells
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
seeded at greater concentrations tended to overgrow and detach from the wells as soon as 24 h after infection. The virus adsorption times tested (15, 30, and 60 min) did not contribute significantly to the cellular ELISA results. Nevertheless, it must be emphasized that different influenza A virus strains have different hemagglutinin affinity for sialic acid and therefore different adsorption and subsequent infection capacities (Weis et al., 1988). Moreover, the virus tropism also depends on the number of sialic acid residues expressed on cell surface glycoproteins and may modify cell susceptibility to infection. In our experiments, only one strain of influenza A virus (A/AA/6/60) and one cell line (MDCK) was used, therefore the optimal adsorption time and subsequent infection time may be different for other influenza A virus strains and host cell types. The reproductive cycle of viruses consists of three phases, and all of them may contribute directly or indirectly to cellular ELISA results. The first interval is known as the eclipse phase. During this phase the virus genome is exposed to host and viral machinery necessary for its expression but progeny virus production has not yet been started. This phase is followed by an interval Table 1 Inhibition of influenza A virus infection by BCTP as measured using PRA and cellular ELISAa
No virus Untreated virus Virus treated with BCTP: 1×10−1 1×10−2 1×10−3
a
PRA
Cellular ELISA
0.00 (+/−0.00)c 50.88d (+/−0.25)
0.269b (+/−0.002) 1.856 (+/−0.050)
0.00 (+/−0.00) 0.00 (+/−0.00) 53.11 (+/−0.74)
0.258 (+/−0.012) 0.261 (+/−0.008) 1.392 (+/−0.103)
1.25×104 pfu/ml was used in cellular ELISA as it was within the linear portion of the curve (see Fig. 4). b Absorbance reading A(450nm). c Standard error. d Number of plaques.
175
in which progeny virus accumulates in the cell at an exponential rate. This interval is known as the maturation phase. After several hours infected cells cease all metabolic activity, lose their integrity and enter the lytic phase (Roizman and Palese, 1996). Therefore, the optimal phase of virus reproductive cycle to measure viral proteins in cellular ELISA is the late maturation phase before cells enter the lytic phase. Since virus reproductive cycles may vary from a few to several hours, it was important to determine the optimal time of post-infection incubation. Cells were incubated from 6 to 48 h after infection followed by cellular ELISA. At 6 h there was no increase in absorbance coinciding with the eclipse phase of virus reproduction cycle. The highest absorbance was observed at 48 h incubation period but the virus had entered the early lytic phase and reproducibility of the cellular ELISA was poor (Fig. 3). After 48 h incubation, we observed gross lysis of infected cells which had a detrimental effect on the assay generating, in some cases, background level absorbance (data not shown). Based on these data, we chose 24 h as the optimal, post-infection period for the in situ cellular ELISA protocol. Two important features of quantitative ELISA are linearity (Berkowitz and Levin, 1985), and sensitivity (Fulton et al., 1988). A linear relationship between virus inoculum and absorbance was observed at the range 1.88× 102 –4.57× 104 pfu/ ml (R 2 = 0.983, PB 0.0002). Above this number of virus particles, cellular ELISA readings plateaued (Fig. 4). After developing the cellular ELISA protocol, we compared directly the cellular ELISA and PA by running these two assays in parallel. Both assays yielded linear and highly correlated data however, we could directly compare only four out of 11 virus 3-fold dilutions. Dilutions higher than 1.88× 10 − 6, were the lower threshold of the cellular ELISA and at dilutions lower than 5.08× 10 − 5, the number of plaques/well became too numerous to count (Fig. 5). Leahy et al. (1994), reported that infection of cells in suspension resulted in greater absorbance values than inoculation of cell monolayers as a result of more rapid viral growth and improved the ability to measure dose-response. It is quite likely that infection of
176
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177
cells in cell suspension could increase the sensitivity of our cellular ELISA as well, but we have not explored such possibility. Infection of the cells in suspension would introduce substantial variability to the experiment and thereby it would not make possible to compare both assays in parallel. The BCTP has been reported to have an extensive bactericidal (Hamouda et al. 1998a,b), sporicidal (Hayes et al. 1998), virucidal and fungistatic effects (Hamouda et al. 1998b). Since BCTP is a new class of anti-microbial agent, we compared its activity with Triton X-100. Antiviral activity of both agents was detected by cellular ELISA (Fig. 6). Subsequently, we applied the BCTP in both cellular ELISA and PRA assays to test the possibility of using cellular ELISA to screen antiviral agents. The cellular ELISA and PRA reflected a similar virucidal effect of BCTP on influenza A virus (Table 1). However, it is possible that some antiviral agents may generate discrepancies between these two assays as they are based on different principles. Nevertheless, the advantages of cellular ELISA, including the 96-well format and ability to automate, make this assay very convenient for the rapid screening of a large number of antiviral agents in a broad range of concentrations. Although in this report we described the cellular ELISA technique to detect H2N2 strain of influenza A virus, the same technique and antibody detection were used to assess influenza A infectivity of X-31 (H3N1) strain and it was found that the technique worked equally well (data not shown). We believe that by using other detecting antibodies directed to different influenza A and B strains, this technique would have more general application than reported in this paper.
Acknowledgements We thank Brian W. Donovan for assistance in plaque reduction assays. We also thank Michael M. Hayes and Dr Jon D. Reuter for their assistance in propagation of influenza A virus. Funding for the study was obtained from DARPA contract MDA 972-97-1-0007.
References Baker J.R., Jr., Wright D.C., Hamouda T., Hayes M.M., Brisker J., 1997. US Patent pending. Barrett, T., Inglis, S.C., 1985. Growth purification and titration of influenza viruses. In: Mahy, W.J. (Ed.), Virology: A Practical Approach. IRL Press, Washington, DC, pp. 119 – 151. Berkowitz, F.E., Levin, M.J., 1985. Use of enzyme-linked immunosorbent assay performed directly on fixed infected cell monolayers for evaluating drugs against varicellazoster virus. Antimicrob. Agents Chemother. 28, 207 – 210. Bishop, G.A., Hwang, J., 1992. Use of a cellular ELISA for the detection of cell surface antigens. BioTechniques 12, 326 – 330. Cunningham, C., Power, D.A., Innes, A., Lind, T., Catto, G.R.D., 1987. Maternal antibody responses during early pregnancy detected by a cellular enzyme linked immunospecific assay. Hum. Immunol. 19, 7 – 16. Fulton, R.E., Wong, J.P., Siddiqui, Y.M., Tso, M.-S., 1988. Sensitive fluorogenic enzyme immunoassay on nitrocellulose membranes for quantitation of virus. J. Virol. Methods 22, 149 – 164. Hamouda T., Wright D.C., Brisker J.M., Baker J.R. Jr., 1998a. Microbiocidal Effects of Liposome-like Microemulsions on Pathogenic Gram Negative Bacteria. American Society for Microbiology, 98th General Meeting, May 17 – 21, 1998, Atlanta, GA. Abstract c A-52. Hamouda T., Hayes M.M., Anderson M.J., Cao Z., Myc A., Reuter J.D., Donovan B., Johnosn K.J., Wright D.C., Brisker J., Baker J.R., Jr., 1998b. Broad spectrum antimicrobial activity of a novel surfactant nanoemulsion (personal communication). Harrach, B., Robenek, H., 1990. Polyclonal antibodies against formaldehyde-modified apolipoprotein A-I. An approach to circumventing fixation-induced loss of antigenicity in immunocytochemistry. Arteriosclerosis 10, 564 – 576. Hayden F.G., Cote K.M., Douglas G. Jr., 1980. Plaque inhibition assay for drug susceptibility testing of influenza viruses. Antimicrob. Agents Chemother. 17, 865 – 870. Hayes M.M., Cao Z., Chung Y., Wright D.C., Brisker J.M., Baker J.R. Jr., 1998. Biocidal effect of unique lipid surfactants on Bacillus spores. American Society for Microbiology, 98th General Meeting, May 17 – 21, 1998, Atlanta, GA. Abstract c A-49. Herlocher, M.L., Clavo, A.C., Maassab, H.F., 1996. Sequence comparison of A/AA/6/60 influenza viruses: mutations which may contribute to attenuation. Virus Res. 42, 11 – 25. Hollingshead, M.G., Westbrook, L., Ross, M.J., Bailey, J., Quallas, K.J., Allen, L.B., 1992. An ELISA system for evaluating antiretroviral activity against Rauscher murine leukemia virus. Antiviral Res. 18, 267 – 274. Hosoya, M., Matuyama, S., Baba, M., Suzuki, H., Shigeta, S., 1992. Effects of proteinase inhibitors on replication of various myxoviruses. Antimicrob. Agents Chemother. 36, 1432 – 1436.
A. Myc et al. / Journal of Virological Methods 77 (1999) 165–177 Huang, R., Dietsch, E., Lockhoff, O., Schuller, M., Reutter, W., 1991. Antiviral activity of some natural and synthetic sugar analogues. FEBS Lett. 291, 199–202. Kiernan, J.A., 1990. Fixation. In: Alan, J. (Ed.), Histological and Histochemical Methods: Theory and Practice. Pergamon Press, Oxford, UK; New York, pp. 25–27. Kraehenbuhl, J.P., Racine, L., Jamieson, J.D., 1977. Immunocytochemical localization of secretory proteins in bovine pancreatic exocrine cells. J. Cell Biol. 72, 406–423. Karaivanova, V.K., Spiro, R.G., 1998. Sulphation of N-linked oligosaccharides of vesicular stomatitis and influenza virus envelope glycoproteins: host cell specificity, subcellular localization and identification of substituted saccharides. Biochem. J. 329, 511 –518. Leahy, B.J., Christiansen, K.J., Shellam, G., 1994. Standardization of a microplate in situ ELISA (MISE-test) for the susceptibility testing of herpes simplex virus to acyclovir. J. Virol. Methods 48, 93–108. Leahy, B.J., Christiansen, K.J., Collins, P., 1996. A blinded comparison of two methods of viral susceptibility testing: plaque reduction assay versus microplate in situ ELISA. J. Virol. Methods 56, 85–90. Leu, R.W., Robinson, C.J., Wiggins, J.A., Shannon, B.J., Rummage, J.A., Horn, M.J., 1988. Photometric assays for FcR1-dependent cellular cytotoxicity mediated by monomeric IgGg2a in murine peritoneal macrophages. J. Immunol. Methods 113, 269–278. Mammen, M., Dahmann, G., Whitesides, G.M., 1995. Effective inhibitors of hemagglutination by influenza virus synthesized from polymers having active ester groups. Insight into mechanism of inhibition. J. Med. Chem. 38, 4179– 4190. McLean, L., Perrett, D., Winrow, V.R., Archer, J.R., 1989. Status of an unpaired thiol group on the HLA-B27 epitope. Clin. Exp. Immunol. 77, 417–421. Mendel, D.B., Tai, C.Y., Escarpe, P.A., et al., 1998. Oral administration of a prodrug of the influenza virus neuraminidase inhibitor GS 4071 protects mice and ferrets against influenza infection. Antimicrob. Agents Chemother. 42, 640 –646. Ostrand-Rosenberg, S., Edidin, M., Wetzel, B., 1979. Differential localization of H-2 and teratoma-defined antigens on mouse L cell fibroblasts. In: Becker, R.P., Johary, O. (Eds.), Scanning Electronic Microscopy, vol. 3. Scanning Electron Microscopy, AMF Ohare, IL, pp. 595–600. Pauwels, R., Balzarini, J., Baba, M., Snoeck, R., Schols, D., Herdewijn, P., Desmyter, J., De Clercq, E., 1988. Rapid and automated tetrazolum-based colorimetric assay for the detection of anti-HIV compounds. J. Virol. Methods 20, 309 – 321. Petris, D.E., 1978. Immunoelectron microscopy and immu-
.
177
nofluorescence in membrane biology. In: Korn, E. (Ed.), Methods in Membrane Biology, vol. 9. Plenum Press, New York, pp. 1 – 177. Rabalais, G.P., Levin, M.J., Berkowitz, F.E., 1987. Rapid herpes simplex virus susceptibility testing using an enzymelinked immunosorbent assay performed in situ on fixed virus-infected monolayers. Antimicrob. Agents Chemother. 31, 946 – 948. Reuter J.D., Myc A., Cao Z., Johnson K.J., Wright D.C., Brisker J., Baker J.R. Jr., 1998. Prevention of influenza A virus infection by non-ionic surfactant nanoemulsions in a murine model. Antimicrob. Agents Chemother. (submitted). Roizman, B., Palese, P., 1996. Multiplication of viruses: an overview. In: Fields, B.N., Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, vol. 1. Lippincott-Raven Publishers, Philadelphia, New York, pp. 101 – 111. Samoszuk, M., Sholly, S., Epstein, A.L., 1987. Screening for anti-idiotypic monoclonal antibodies on paraformaldehyde-fixed lymphoma cells. Hybridoma 6, 605 – 610. Smith, D.D., Cohick, C.B., Lindsley, H.B., 1997. Optimization of cellular ELISA for assay of surface antigens on human synoviocites. BioTechniques 22, 952 – 957. Smith, P.W., Sollis, S.L., Howes, P.D., et al., 1998. Dihydropyrancarboxamides related to zanamivir: a new series of inhibitors of influenza virus sialidases. 1. Discovery, synthesis biological activity, and structure-activity relationships of 4-guanidino- and 4-amino-4H-pyran-6-carboxamides. J. Med. Chem. 41, 787 – 797. Takeuchi, H., Baba, M., Shigeta, S., 1991. An application of tetrazolum (MTT) colorimetric assay for the screening of anti-herpes virus compounds. J. Virol. Methods 33, 61 – 71. Van Ewijk, W., Coffman, R.C., Weissman, I.L., 1980. Immunoelectron microscopy of cell surface antigens: a quantitative analysis of antibody binding after different fixation protocols. Histochem. J. 12, 349. Waghorn, S.L., Goa, K.L., 1998. Zanamvir. Drugs 55, 721 – 725. Watanabe, W., Konno, K., Ijichi, H., Inoune, T., Yokota, T., Shigeta, S., 1994. MTT colorimetric assay system for the screening of anti-orthomyxo- and anti-paramyxoviral agents. J. Virol. Methods 48, 257 – 265. Watanabe, W., Sudo, K., Asawa, S., Konno, K., Yokota, T., Shigeta, S., 1995. Use of lactate dehydrogenase to evaluate the anti-viral activity against influenza A virus. J. Virol. Methods 51, 185 – 192. Weis, W., Brown, J.H., Cusack, S., Paulson, J.C., Skehel, J.J., Wiley, D.C., 1988. Structure of influenza virus hemagglutinin complexed with its receptor, sialic acid. Nature 333, 426 – 431.
.