Assay of variola virus on Microtiter plates and its application to the neutralization test*

Bull. Org. nwnd. Sante 11973, 49, 251-256Bull. Wld Hlth Org. Assay of variola virus on Microtiter plates and its application to the neutralization test* TAKASHI KITAMURA & YUKIE TANAKA It was established that microfocus assay of the variola virus in HeLa cell cultures on Microtiter plates is a suitable methodfor titrating the virus in numerous specimens ofsmalt volume. The variola foci were formed within 48 hours ofinoculation, and the standard error for the number offoci per well did not exceed 15 %, with 96 flat-bottomed wells (6 mm diameter) containing 0.1 ml per well of the culture media on a plastic plate replacing 96 Petri dishes in the standard assay procedures. Microfocus assay was successfully applied to the quantification of neutralizing antibody against variola and vaccinia viruses, with a reproducibility equal to that of the standard focus reduction method on glass Petri dishes. Applications of this technique to the assay of field specimens and to the laboratory differentiation of variola viruses are discussed. Variola virus forms tiny " hyperplastic " foci in human cell monolayers and this phenomenon has been used in the in vitro quantification of the virus (7). When the foci were allowed to grow for 4-7 days they could be recognized macroscopically and could be counted when the cells were stained with neutral red (2). The neutralization test of anti-pox sera based on 50% reduction in focus formation is a well established procedure with excellent reproduci- bility and was employed for analyses of the antibody responses to revaccination with smallpox vaccines (3). This method, however, requires a rather large num- ber of cell cultures and laborious procedures, and simplified methods have been sought. The present paper describes a simplified and rapid procedure for microfocus counting of variola virus in the Micro- titer plate and its application in the neutralization test. MATERIALS AND METHODS Virus Variola major, strain Harvey (VHE), and variola minor, strain Butler (VAB), viruses were used as the homogenate of infected chorioallantoic membranes (CAM) as described earlier (2). * From the Division of Poxviruses, National Institute of Health, Shinagawa-ku, Tokyo 141, Japan. Cell culture HeLa cells propagated in the previously de- scribed (2) growth medium of 90% yeast extract- lactalbumin hydrolysate-Earle solution (YLE) and 10% bovine serum (BS) were suspended in the fresh medium (2.0 x 105 cells per ml) and 0.1 ml was dispensed into each of 96 wells of a flat-bottomed Microtiter plate (Cooke Engineering, No. 220- 29ART), which was then covered with a rigid lid (Cooke, No. 220-42). After 20-24 hours' incubation at 37°C, confluent monolayers were formed. Virus inoculation and microfocus count Growth medium was discarded by aspiration with a hypodermic needle and the cell sheets were washed once with 0.1 ml per well of PBS (pH 7.6). Serial dilution of the virus in HPS (a mixture of equal volumes ofHanks's solution and PBS) were prepared, 0.05-ml samples were inoculated with a fine-tipped 0.5-ml graduated pipette, and the plates were incu- bated at 37°C for 120 min, after which 0.05 ml per well of the maintenance solution (MS), consisting of 98% YLE and 2% BS, was added. MS was discarded by aspiration 40-48 hours after inoculation and the cell sheets were washed once with 0.1 ml per well of PBS and then stained with 0.05 ml per well of Wright's stain for 15 min at room temperature under heavy irradiation with germicidal lamps to secure the 3109 - 251 T. KITAMURA & Y. TANAKA virus-sterility of the microplates. When washed in the water, microfoci were observable as dark violet spots of approximately 0.07-0.12 mm in diameter on diffusely stained cell layers. Direct neutralization test Standard procedures for the neutralization test in ordinary cultures have been described (3). In the direct microneutralization test, virus-antibody mix- tures, incubated after the standard procedures, were inoculated and the serum dilution showing 50% reduction of the number of foci (ND50) was cal- culated. Neutralization test by use of the transfer plate A sterile transfer plate (Cooke, No. 220-43) (1) was placed on a sterile, dry rigid V-plate (Cooke, No. 220- 25A), which served as a holding sheath with a frame of stainless wire of 1.0 mm diameter to secure a distance between the transfer holes at the tips and the sheath bottom. Serial 2-fold dilutions of the serum were prepared in 3 rows of the transfer plate with 0.025 ml diluters (Cooke, No. 220-33), leaving each 0.025 ml of the dilution in the well. One drop of 0.025 ml of the challenge virus suspension containing about 1 200 FFU per ml was added with a micro- dropper (Cooke, No. 220-5) into each well and the plate in the sheath was shaken on a mechanical vibrator for a short time. The transfer plate-sheath complex covered with the rigid lid was wrapped air- tight in a plastic foil and incubated at 37°C for 60 min and 4°C overnight thereafter. Cell mono- layers in flat-bottomed wells of the microplates prepared on the previous day were washed with PBS and received 0.05 ml of the fresh MS per well. A transfer plate containing the virus-serum mixtures was put on a culture plate as described by Catalano et al. (1). After 40-48 hours, focus staining and ND60 determination were carried out as in the direct method. RESULTS Microfocus formation by the variola virus When the virus material was diluted in HPS and inoculated onto HeLa cell monolayers as described above, microfoci were discernible under the micro- scope as early as 18 hours p.i. and grew to the size of approximately 0.1 mm diameter by about 48 hours p.i. The foci stained deeply with Wright's solution (see Fig. 1) and under a dissecting microscope at x 10-16 magnification they could be counted in Fig. 1. Microfoci formed in HeLa cell monolayers on microplates. The cell sheets were washed with PBS and stained with Wright's solution 48 hours after the inoculation of VHE variola major virus. The diameter of the well was 6 mm. exactly the same way as macroscopic foci in a Petri dish. There was no difference between the foci induced by VHE and VAB viruses, as observed in the case of the standard focus method. Efforts were made to achieve optimal conditions for the formation of microfoci. The initial cell concentration for the preparation of cell monolayers was found to be important: the use of initial cell concentrations around 1.5-2.0 cells per ml resulted in the formation by the homogeneous monolayers of easily countable foci. With higher cell concentra- tions, cell clumps on overcrowded monolayers re- sembling the viral foci made the counting difficult. The use of Eagle's mediurn and calf or fetal calf serum as growth and maintenance media resulted in the formation of foci of rather irregular size, whereas with YLE and BS solid, homogeneous foci were produced. Focus counting was possible using either vital staining with neutral red or fixed staining with Giemsa, but the Wright staining was the most simple procedure and gave identical results. The parallel infectivity profiles in sucrose density gradient frac- tions of VHE virus suspension assayed by the stan- dard and microfocus titrations (Fig. 2) indicated that the foci in the present micro method were also formed by the variola virus. 252 MICROFOCUS ASSAY OF VARIOLA VIRUS Table 1. Standard error of the microplates focus counting on Average Standard Standard foci/well a deviation error (x) (3) (a/x in %) 85 14.6 17.2 68 7.8 1 1.5 50 4.4 8.8 41 5.1 12.2 34 4.4 12.9 26 4.8 18.5 1 9 2.7 14.2 1 5 3.3 22.0 a Average of each 12 wells. 1 5 10 15 2 Bottoa(50%) Fraction number Fig. 2. Comparative titration of variola virus fractions by microfocus and standard focus assay procedures. Partially purified variola virus (Harvey strain) suspen- sion was spun on 5 ml of linear density gradient of 40-50 % sucrose in PBS at 13 500 rev/min for 45 min. Fractions of 10 drops each were collected from the bottom and subjected to parallel tritrations of microfoci on microplates and standard foci on glass Petri dishes. Quantitative aspects of microfocus count As reported earlier (2), focus formation by the variola virus is quantitative only with inoculum doses permitting isolated foci on the cell layer, whereas higher doses of virus produce another type of cytopathic effect (CPE) consisting mainly of cell fusion. The upper limit of virus dose in the present microfocus count was checked by calculating the standard error of the focus count with each 12 wells inoculated with various doses of virus. As summa- rized in Table 1, focus counting was possible up to a virus dose of 85 FFU per well, and counting with a standard error around 10% was possible with a dose of 34-68 FFU per well. Thus, it may be concluded that, when the virus dose is kept within this range, a quantitative titration of the infectious virus should be possible. Another experiment was carried out to compare the standard error in the focus count by three inoculation methods, i.e., inoculation with cali- brated 0.5-ml pipettes or with a fixed volume dropper (0.05-ml, Cooke, No. 220-17) with or without a transfer plate. Differences among 3 operators with various degrees of experience of this technique were also compared. As shown in Table 2, inocula- tion with pipettes and droppers resulted in the same degree of standard error, regardless of the experience with the technique, whereas the use of a transfer plate resulted in a larger standard error; this finding Table 2. Comparison of standard errors among inocu- lation methods Inoculation Average Standard Standard method Operator foci/well a deviation error(x) (a) (a/x in %) S 31.5 2.64 8.4 0.05 ml with T 32.4 4.25 13.1 05. ml graduated K 34.8 4.47 12.8 pipette Average 32.6 11.4 S 35.6 3.95 11.1 T 30.4 2.06 6.8 0.05 ml dropper K 43.9 7.96 18.1 Average 36.6 12.0 S 41.8 1.30 31.1 0.05 ml T 45.4 6.76 14.9 dropper + transfer K 38.4 7.30 19.0 plate Average 41.9 21.7 a Average of each 12 wells. U/al(xio5) 253 T. KITAMURA & Y. TANAKA Per cent survival 100 0 46- 80 60 40Th IRS ampoules A B 201:820 100 400 1600 6400 25600 Serum dilution (usn) Fig. 3. Reproducibility of neutralization test on micro- plate. The contents of 4 ampoules of one lot of lyo- philized standard immune rabbit serum were reconsti- tuted separately and titrated separately. The ND5o titres obtained were 1 :1140 (A), 1 :1 230 (B), 1 :1 000 (C), and 1 : 820 (D). Per cent survival 160 640 2 100 400 160C Berm dilution (lin) !560 6400 Fig. 4. Neutralization curves of various antibody preparations. Sera from human adults previously vac- cinated against smallpox, immune rabbit sera, and vaccinia immune globulin fraction (VIG) from human serum pool were subjected to the neutralization test (direct procedure) described in the text. suggests that the use of the microtransfer technique in the quantitative neutralization test may give rise to some difficulties. The direct microneutralization test with standard anti- serum The direct microneutralization test was carried out with an antivaccinia hyperimmune rabbit serum serving as the national standard for neutralizing antibody (NA) assay. The contents of 4 lyophilized ampoules were reconstituted and assayed indepen- dently, 3 wells being inoculated with each reaction mixture. As shown in Fig. 3, almost identical neutral- ization curves and ND50 titres were obtained; this finding shows the high reproducibility of the assay. Another set of comparative NA assays carried out using transfer plates gave rather confusing results, with ND50 titres of 1: 1800, 1: 1400, 1: 460, and 1: 960, in contrast with those (1: 1140, 1: 1230, 1 : 1000, and 1: 820) obtained by the direct micro- neutralization test. The pattern of the neutralization curve was com- pared by the direct method in 4 sera from smallpox- vaccinated human adults (IHS), standard hyperim- mune rabbit serum (IRS), and a vaccinia immune human globulin preparation (VIG). As shown in Fig. 4, the ND50 may be determined with an identical degree of accuracy using any of the 3 kinds of antisera. Another set of IHS, IRS, and VIG were submitted to the comparative NA assay using the standard procedure and direct microneutralization procedure to compare the sensitivity ofNA detection. As summarized in Table 3, ND50 titres obtained with Table 3. Dual titration of NDso with the standard method and microplate method NDso titres obtained Origin Serum History Standard Microplate method method human Si recently revaccinated 34 49 S2 recently revaccinated 180 240 K 10 years after revaccination 75 94 VIG pool of revaccinated adult sera 1100 1 350 IRS-1 antivaccinia 450 490 rabbit IRS-2 antivaccinia 1 600 2 500 IRS-3 antivariola 133 160 254 MICROFOCUS ASSAY OF VARIOLA VIRUS the micro method were slightly higher than those obtained with the standard procedure. This may be accounted for by the less steep slope of the linear portion of the neutralization curves with the former than with the latter method. Another possibility is that the difference may be related to the rather larger fluctuations in ND50 titres with the micro method, although fluctuations may be within a narrow range, as seen in Fig. 3. DICUSSION Microtiter procedures have been applied by many authors to the quantification of virus infectivity as well as of neutralizing antibody. But in most cases, even with a flat-bottomed plate, the titration de- pended on the detection of viral CPE and calculation of TCID50 (8, 9, 6). The present authors established quantification of variola virus by direct microfocal lesion count with the same efficiency as with macro- focus or plaque assay. A plastic plate as small as 10 x 15 cm could replace 96 Petri dishes in the macrofocus method, with a standard error as small as 10-15 %. As a practical application of this method, a differential diagnosis of smallpox material was also established (4). A microneutralization test using the microtransfer procedure was successful with polio antibodies (6); here, the serum dilution endpoints determined were in close agreement with the results obtained by the conventional tube method. In the case of poxviruses, however, in vitro neutralization does not usually take place so readily as in the case of picorna or other small viruses, and as a result there is a breakthrough of CPE and fluctuating serum dilution endpoints. For the analysis of neutralizing antibody response, however, focal lesion reduction assay is possible, with excellent reproducibility (3). For this purpose, the microtransfer procedure seems to be rather inferior to the standard procedure, as observed above. The applicability of the microtransfer method in this field and its limitations need be defined by further studies. As a practical application of the direct microneu- tralization procedure for a neutralizing antibody (NA) assay of the field materials, 100 pairs of maternal and umbilical cord sera were collected during normal deliveries and examined for their variola NA levels (5). Comparison of ND50 titres in pairs gave a correlation coefficient of 0.8, which was interpreted as indicating identical levels of NA in maternal and fetal sera. The microneutralization procedure was found to have one restriction: with serum dilutions lower than 1: 10, the final serum content during incubation reaches more than 2.5% in MS, which seems to affect the growth conditions of the host cells, resulting in irregular focus formation and making focus counting difficult. ACKNOWLEDGEMENTS The present work was carried out as part of Special Studies on the Improvement of Smallpox Vaccines sponsored by the Ministry of Health and Welfare, Government of Japan, during the years 1970-1972. The authors are obliged to Dr I. Tagaya for his criticisms on the study. RItSUMt TITRAGE DU VIRUS VARIOLIQUE SUR PLAQUES MICROTITER ET APPLICATION A L'tPREUVE DE NEUTRALISATION On a mis au point une variante de la methode de num6ration des foyers d'hyperplasie dus au virus vario- lique en utilisant une culture de cellules HeLa en couche monocellulaire sur plaques Microtiter a fond plat. Chacun des 96 puits de la plaque regoit 0,1 ml d'une suspension de 2,0 x 10 cellules/ml qui forment une couche mono- cellulaire confluente apres 20-24 heures d'incubation. L'inoculation de virus rabique provoque l'apparition apres 48 heures de microfoyers d'environ 0,1 mm de diametre. L'ecart type de la numeration des foyers, par puits, a ete de l'ordre de 10 A 15% lorsque l'inoculum contenait 20-70 unit6s formatrices par puits. Le parallelisme entre les taux d'infectivite des suspensions de virus mesures 255 256 T. KITAMURA & Y. TANAKA par la m6thode standard ou par la microm6thode prouve que les foyers observ6s dans cette derniere sont dus i 1'activit6 du virus variolique. L'epreuve de neutralisation A I'aide de s6rums anti- virus variolique et anti-virus vaccinal bas6e sur la mesure de la r6duction des foyers d'hyperplasie a pu 8tre pra- tiqu6e par la microm6thode dont la reproductibilit6 s'est r6v61ke 6gale a celle de la m6thode standard. REFERENCES 1. CATALANO, L. W., Jr., ET AL. Appl. Microbiol., 18: 1094-1095 (1969). 2. KrrAmuRA, T. Virology, 36: 174-179 (1968). 3. KITAMURA, T. & SHmIo, N. Bull. Wld HNth Org., 46: 15-26 (1972). 4. KITAMURA, T. & TANAKA, Y. Bull. WId Hlth Org., 48: 495-497 (1973). 5. KITAMURA, T. ET AL. Japan J. Med. Sci. Biol., 26: 45-48 (1973). 6. MOR1TSUGU, Y. ET AL. Japan J. Med. Sci. Biol., 25: 43-46 (1972). 7. PIRSCH, J. B. & PURLSON, E. H. J. Immunol., 89: 632-637 (1962). 8. ROSENTHAL, L. J. & SHECHMEISTER, I. L. Appl. Micro- bioL, 21: 400-404 (1971). 9. SIDWELL, R. W. & HUFFMAN, R. W. Appl. Microbiol., 22: 797-801.