in the PN and MN tests was made from suckling mouse brains harvested when the animals became

Transcription

1 APPLIED MICROBIOLOGY, Nov. 1968, p Vol. 16, No. 11 Copyright 1968 American Society for Microbiology Printed in U.S.A. Yellow Fever Virus I. Development and Evaluation of a Plaque Neutralization Test SHELDON SPECTOR AND NICOLA M. TAURASO Laboratory of Virology and Rickettsiology, Division of Biologics Standards, National Institutes of Health, Bethesda, Maryland Received for publication 18 June 1968 Heretofore, the most reliable way of measuring yellow fever virus antibody was to use the mouse neutralization (MN) test employing either suckling or weanling mice. Certain disadvantages (e.g., expense both of animals and of maintaining a mouse colony, allergic reactions of many laboratory workers, and the relatively long time, 21 days, before end points are reached) are inherent in any program with mice or other laboratory animal species and have discouraged the use of the MN test by many laboratories. A previously reported plaque neutralization (PN) test with primary chick embryo cell cultures could not be consistently reproduced by later investigators. We have developed a convenient and reproducible PN test employing the MA-104 embryonic rhesus monkey kidney cell culture and a single agar-overlay procedure. When compared with MN tests with newborn (1 to 3 days old) and weanling (16 to 20 g, 24 to 28 days old) mice inoculated by the intracranial route, the PN test was the most sensitive for measuring neutralizing antibody; it was also less variable, less costly, and it achieved results in the shortest period of time. End points could be determined in 5 to 6 days for the PN test as compared to 21 days for the MN test. There are several ways to measure neutralizing activity to yellow fever (YF) virus. Neutralization and protection tests in monkeys have been used successfully in the past but are impractical because the use of large numbers of monkeys is expensive and inconvenient (1). The introduction of the susceptible white mouse provided a convenient and less expensive alternative and a means of testing large numbers of sera (15, 16). The basic technique, employing a constant serumvarying virus dilution, has been studied with different routes of inoculation, such as intracranial (ic), intraperitoneal (ip), and intraperitoneal with concomitant cerebral trauma (10, 17). A plaque neutralization (PN) test could provide a handy in vitro method of assaying YF virus neutralizing activity, but its success has varied in different laboratories. The use of chick embryo monolayers by Porterfield (7, 8) could not be reproduced by Henderson and Taylor (4, 5) or by Schulze and Schlesinger (12). We have developed a PN test with the MA- 104 embryonic rhesus monkey kidney cell line and a simple single agar-overlay procedure. In this report, we describe this PN test and compare it with the more conventional mouse neutralization (MN) test performed by inoculating mice by the ic or ip route. MATERIALS AND METHODS Virus. Infectious stock YF virus (17D strain) used in the PN and MN tests was made from suckling mouse brains harvested when the animals became sick and was prepared as a 20% suspension in phosphate-buffered saline (PBS; 2) with 0.5% bovine plasma albumin (BPA). The 17D vaccine strain was passaged three times in 10- to 12-g mice and four times in suckling mice in our laboratory. All pools were stored in sealed glass ampoules at -70 C until used. Animals. Newborn mice (1 to 3 days old), suckling mice prior to weaning (10 to 18 days old), and weanling mice (16 to 20 g, 24 to 28 days old; Swiss albino, NIH General Purpose strain) used for the MN tests were obtained from the Rodent and Rabbit Production Section, Laboratory Aids Branch, National Institutes of Health (NIH). In our studies, we used complete litters (containing from 6 to 16 newborn or suckling mice) and 10 weanling mice at each dilution point. Rhesus monkeys (Macaca mulatta) obtained from India were immunized with YF vaccine either on the day of their arrival at NIH or after 6 weeks of quarantine. Some monkeys were inoculated with commercial RIF (9) virus-contaminated YF vaccine and others were inoculated with an RIF virus-free vaccine (14) being developed in our laboratory. Antisera. Sera used in our neutralization tests were obtained from monkeys before and 28 days after 1770

2 VOL. 16, 1968 YELLOW FEVER VIRUS. I 1771 inoculation with YF vaccine. The sera were stored at -20 C until used. Cell culture. The MA-104 embryonic rhesus monkey kidney cell cultures (originally obtained from Microbiological Associates, Inc., Bethesda, Md.) were prepared in the routine manner established by the Tissue Culture Section of our laboratory. Monolayer cell cultures in 32-oz (0.946 liter) bottles were dispersed with a solution containing 0.25% trypsin and 0.2% ethylenediaminetetraacetic acid in Dulbecco's saline (2) without calcium and magnesium. Prescription bottles [2 oz (0.059 liter)] were seeded with 7 ml of cell suspension (80,000 cells/ml) in a growth medium consisting of minimum essential medium containing Hanks' solution base (11), 10% fetal bovine serum, and neomycin (50,g/ml). Bottle cultures [2 oz (0.059 liter)] were received in our laboratory 3 to 5 days after initial seeding of cells; growth medium was replaced with maintenance medium consisting of Earle's balanced salt solution containing 0.5% lactalbumin hydrolysate, 2% fetal bovine serum, penicillin G (100 units/ml), streptomycin (100,ug/ml), and nysstatin (50 units/ml). PN test. The PN test was performed in the following manner. All reagents were kept in an ice-water bath before and during mixing; unheated serum diluted 1:2 in PBS with 0.5% BPA was added to equal amounts of serial 10-fold dilutions of virus in the same diluent. After incubation at room temperature (23 to 25 C) for 1 hr, the mixtures were retumed to an ice-water bath, and 0.2 ml of each mixture was inoculated into each of two to four 2-oz (0.059 liter) bottle cultures of MA-104 cells. After adsorption for 1 hr at 36 C, the cultures were overlaid once with 5 ml of an agar medium (13). Plaques were read after incubation at 36 C for 5 to 6 days. End points were usually determined from dilutions containing 10 to 50 plaque-forming units (PFU). The difference in titer between serum samples taken before and 28 days after immunization represents the neutralizing capacity of the serum and is expressed as neutralization indices (NI) in this paper. MN tests. A sample of the same virus-serum mixtures used in the PN test was inoculated into mice. The standard inoculum was 0.03 ml inoculated ic and 0.2 ml inoculated ip. Deaths occurring over a period of 21 days were recorded; LD5o end points were determined by the Karber method (6) and were expressed as decimal exponent (Dex) values (3). RESULTS Characteristics of YF virus plaques. Plaques are best observed with indirect fluorescent lighting from an oblique angle. A plaque develops as a tiny (0.5 mm) white speck on day 5, progressing to 3 to 4 mm and developing a "foamy" appearance and an irregular border on day 7 to 8. As the plaque becomes larger, it can be seen as a clear area by direct lighting. In our experience, only two other viruses (West Nile and Murray Valley encephalitis) produce a similar "foamy" plaque in the MA-104 cell culture but they are TABLE 1. Plaque neutralization (PN) and intracranial weanling mouse neutralization (WMN) tests with sera from monkeys immuntized with yellow fever vaccinte Monkey no. a PN test Neutralization index WMN testb c : : : :5.7 5: : :3.4 a Monkey no. 1 to 12 were immunized with RIF (9) virus-contaminated YF primary seed (lot no. AB 237); no. 13 to 24 were immunized with the RIF virus-free candidate primary seed (14; National Drug lot no. 17D-51). b The mice weighed 16 to 20 g and were 24 to 28 days old. c Neutralization index, expressed as Dex values (3), represents the difference in titer between serum samples taken before and 28 days after immunization. larger than yellow fever virus plaques by several millimeters. Comparison of PN and intracranial weanling mouse neutralization (WMN) tests. Table 1 shows the results of PN and intracranial WMN tests performed with sera from 24 different monkeys (no. 1 to 24). Neutralization end points could not be determined from tests performed with postimmunization sera from monkey no. 1 to 8 because the virus preparation had a low titer (4.1 to 4.5 Dex PFU and 3.2 to 4.3 Dex mouse LD50/0.1 ml). (Titers have been corrected to 0.1 ml for comparison.) These tests could not be repeated because of a lack of serum. A higher infectious titer (6.4 to 6.7 Dex PFU and 5.7 to 6.8 Dex mouse LDto/0.1 ml) of a subsequently

3 1772 SPECrOR AND TAURASO APPL. MICROBIOL. TABLE 2. Plaque neutralization (PN) and intracranial newborn mouse neutralization (NBMN) tests with sera from monkeys immunized with yellow fever vaccine Monkey no." Neutralization index PN test NBMN testb c a Monkey no. 25 to 36 were immunized with RIF (9) virus-contaminated YF secondary seed (National Drug lot no. 5731); no. 37 to 47 were immunized with an RIF virus-free candidate secondary seed (14; National Drug lot no. 6676). b The mice were I to 3 days old., Neutralization index, expressed as Dex values (3), represents the difference in titer between serum samples taken before and 28 days after immunization. prepared virus pool enabled us to determine neutralization end points in tests performed with postimmunization sera from monkey no. 9 to 24. Because end points were obtained, these results are more meaningful for purposes of comparing the relative sensitivity of both neutralization tests. The PN test was consistently more sensitive than the intracranial WMN test for measuring neutralizing activity in the postimmunization sera. Comparison of PN and intracranial newborn mouse neutralization (NBMN) tests. Table 2 shows the results of the PN and intracranial NBMN tests performed with sera from 23 different monkeys (no. 25 to 47). The PN test was more sensitive than the intracranial NBMN tests for measuring neutralizing activity. Effect ofage and route of inoculation upon MN TABLE 3. Effect ofage and route ofinoculation upon the mouse neutralization test and comparison with the plaque neutralization (PN) test Neutralization index Monkey NBMN test" SMN test5 WMN test5 no. PN test icb ip ic ip ic ip c 1.3 > > > a NBMN, newborn (1 to 3 days old) mouse neutralization test; SMN, suckling (15 to 18 days old) mouse neutralization test; WMN, weanling (16 to 20 g, 24 to 28 days old) mouse neutralization test. b Intracranial, ic; intraperitoneal, ip. c Neutralization index, expressed as Dex values (3), represents the difference in titer between negative serum control and postimmunization serum samples. test and comparison with PN test. Results of PN, NBMN, suckling mouse neutralization (SMN), and WMN tests performed simultaneously with postimmunization sera from three different monkeys are shown in Table 3. Owing to a shortage of preinoculation sera, the negative serum control was performed with a pool of sera from the same three experimental monkeys. The NBMN test performed with mice inoculated by the ip route and the PN test were the most sensitive; from tests with sera from monkey no. 48 and 50 (Table 3), it appeared that the intraperitoneal NBMN test may have been more sensitive than the PN test if end points could have been obtained in the former. Suckling and weanling MN tests were unsuccessful with mice inoculated ip because the virus strain used was not highly lethal to mice inoculated by this route. Table 4 compares the virus infectivity end points obtained in the above-described experiment and demonstrates the effect of age and route of inoculation upon virus titer in the MN tests; a comparison is made with the MA-104 cell culture plaquing system. With nonimmune control serum, virus titers in ic inoculated newborn, suckling, and weanling mice were comparable as was the titer in the plaquing method, considering the fact that the inoculum was almost 10-fold greater in this latter system. However, differences in virus titers occurred with the immune sera; this indicated that the systems varied in their ability to measure neutralizing capacity of the sera. To resolve further the sensitivity of the various MN tests and the PN test, additional experiments

4 VoL. 16, 1968 were performed with a virus preparation having a higher infectious titer. Table 5 shows the results of this experiment with paired pre- and postimmunization serum samples from four different monkeys. The intraperitoneal NBMN test was consistently the most sensitive test for TABLE 4. Effect ofage and route ofinoculation upon virus titer in the mouse neutralization test and comparison with the MA-104 cell culture plaquing system Virus titer (LDio) Monkey Monkey no. PFJ4U Newborn Suckling Weanling miceb miceb mice6 icc ip ic ip ic ip Controld 5.8e < < < < < < < < < < <0.5 Per 0.2 ml of inoculum. bnewborn mice, 1 to 3 days old; suckling mice, 15 to 18 days old; weanling mice, 16 to 20 g, 24 to 28 days old. c Intracranial (ic) inoculation of 0.03 ml; intraperitoneal (ip) inoculation of 0.2 ml. d Pool of preinoculation sera from monkey no. 48, 49, and Dex value (3) per inoculum as calculated by the Karber method (6). Monkey TABLE 5. Bleed YELLOW FEVER VIRUS. I measuring neutralizing activity. We could not obtain neutralization end points in these tests with sera diluted 1:2, because virus infectivity titers with sera obtained before immunization were the lowest in ip inoculated newborn mice and because virus was neutralized and could not be detected in mixtures containing sera obtained after immunization. The PN test was the second most sensitive test for measuring neutralizing activity, followed by the intracranial NBMN test; the least sensitive test was the intracranial WMN test. Reproducibility of the PN test. To determine the reproducibility of the PN test, five separate tests were performed with samples of the same pre- and postinoculation sera obtained from a burro immunized with the Asibi strain of YF virus. The titers of virus in the presence of the preinoculation serum were 6.4, 6.3, 5.7, 5.8, and 5.9 Dex; in the presence of the postinoculation serum, the respective titers were 4.5, 4.4, 3.8, 4.5, and 4.5 Dex and the NI values were 1.9, 1.9, 1.9, 1.3, and 1.4 Dex, respectively. Additional data on reproducibility will be found in a subsequent paper. Effect of brain trauma upon virus titer. It was possible to increase the sensitivity of ip inoculated mice (Table 6) used to measure virus infectivity (lethality) by traumatizing the brains of these animals (e.g., ic inoculation of starch solution). Considering the labor and inconvenience Effect ofage and route ovf inoculation upon the mouse neutralization test and comparison with the plaque neutralization (PN) test Neutralization test no.a time6 PN NBMNC WM~Nd Titer" NI Titer NI Titer NI Titer NI 51 Pre > Post < Pre > Post < Pre Post < Pre > Post < a Monkeys were inoculated subcutaneously with 0.5 ml of a RIF virus-contaminated yellow fever vaccine (National Drug Co., lot no. 6135). 6 Time serum was obtained: pre = before inoculation; post = serum obtained 28 days after inoculation. c NBMN, newborn (1 to 3 days old) mouse neutralization; intracranial (columns 5 and 6) and intraperitoneal (columns 7 and 8) inoculations were used. d WMN, weanling (16 to 20 g, 24 to 28 days old) mouse neutralization; intracranial inoculation was used. e Titer expressed as Dex value (3); NI, neutralization index (difference in titer between serum samples taken before and 28 days after immunization) expressed as Dex values. 1773

5 1774 SPECTOR AND TAURASO APPL. MICROBIOL. TABLE 6. Effect ofbrain trauma upon sensitivity of intraperitoneally inoculated mice used for measuring infectivity of yellow fever virus Virus strain Traumaa Miceb Newborn Suckling Weanling 17D Without 4.6c <0. 5 <0.5 With <0.5 FN Without <0.5 With a Trauma consisted of inoculating 0.03 ml of a sterile 2% starch solution in physiological saline (0.85% NaCl) just prior to virus inoculation. b Newborn, 1 to 3 days old; suckling, 15 to 18 days old; weanling, 16 to 20 g, 24 to 28 days old. c Dex value (3) per inoculum as calculated by the Karber method (6). involved in the additional ic inoculation required to traumatize the brain, we did not feel that the slight increase in virus titer was enough to warrant the use of this procedure in the MN test. DISCUSSION A comparison between an in vitro PN test and the more conventional in vivo MN test for measuring YF virus antibodies can best be made with tests performed simultaneously on the same virusserum mixtures. Our studies provided an opportunity to evaluate the advantages and disadvantages of both methods. In our experience, the features of YF virus plaques in MA-104 cell cultures are characteristic enough to prevent confusion with plaques caused by other viruses in the same cell culture system. These unique features of the YF virus plaque enable one to observe an effect specific to YF virus. This is in contrast with the signs of illness produced by YF infection in mice, an effect which can be produced by a number of different arboviruses. Unlike Whitman (17), who found 18-day-old and younger mice more susceptible than older mice to ip virus inoculation, our 15- to 18-day-old suckling mice responded in the same way as our older weanling mice (16 to 20 g, 24 to 28 days old). This can be explained by possible differences in the susceptibility of mouse strains. We found the intraperitoneal NBMN test to be the most sensitive for measuring neutralizing activity. However, the low virus infectivity titers with the resulting lack of clear end points, as well as the general disadvantages of working with animals, emphasized to us the superiority of the PN test, which was the second most sensitive system for measuring YF virus neutralization. End points in the MA-104 cell culture plaquing system are reached in 5 to 6 days in contrast to the 21 days required for experiments with mice. Cost, a constant concern of most laboratories, is much lower for the PN test. Not only is the initial expense less for the MA-104 cell cultures than for either suckling or weanling mice, but it is also substantially less costly considering the money required to maintain a mouse colony and to care for the mice throughout the 21-day test period. Individual variation and host factors are minimal in the PN test. Mice, however, are susceptible to infection with viruses which might be indigenous to a conventional colony, such as lethal intestinal virus of infant mice and epizootic diarrhea of infant mice. In fact, diarrhea which affected our test animals on several occasions necessitated repetition of some of our experiments. With regard to potential health hazards to personnel, working with continuous cell cultures known to be free from biologically harmful agents is considerably safer than working with animals. Although the hazard of exposure to animals capable of transmitting an infectious disease to man is minimal in a modern laboratory, the problem of allergic reactions in personnel who work with laboratory animals is serious enough to warrant consideration as a potential health hazard. In our laboratory, two individuals developed serious allergic reactions (e.g., asthma and wheezing) from exposure to mice. The amount of virus-serum mixture required for the PN test (0.4 ml per two bottles per virus dilution) is slightly greater than the amount required for the intracranial MN test (0.3 ml) when 10 mice are used; the amount is considerably larger for the intraperitoneal MN test (2.0 ml). In our experience, a larger range of virus dilutions is required to achieve end points in titrations with mice and this larger range requires more serum than the above amounts which are calculated per dilution point. Previous investigators (10, 15, 17) have used the French neurotropic (FN) strain of YF virus in their studies. Although preparations of the FN strain have a higher virus titer in both our mouse and cell culture systems, we thought it logical to use the 17D strain as our challenge virus in the PN and MN tests since this was the virus strain used to inoculate our monkeys. Studies on the comparative serological and immunological properties of the 17D, FN, and Asibi strains of YF virus are currently being

6 VOL. 16, 1968 YELLOW FEVER VIRUS. I 1775 performed in our laboratory in order to determine whether virus strain is an important factor in YF virus neutralization tests. The recommended use of unheated serum (18) increased the sensitivity of both tests; a more detailed study of the effect of heat inactivation and of addition of complement is also under way. ACKNOWLEDGMENTS We thank Monroe Vincent (Microbiological Associates, Inc., Bethesda, Md.) for the MA-104 cell line and the Tissue Culture Section of our laboratory for the actual cell cultures. We also thank Michael J. Klutch, Roy W. Trimmer, Cero S. Hayes, and Thomas W. Conner for technical assistance. LITERATURE CITED 1. Beeuwkes, H., J. H. Bauer, and A. F. Mahaffy Yellow fever endemicity in West Africa with special reference to protection tests. Am. J. Trop. Med. 10: Dulbecco. R., and M. Vogt Plaque formation and isolation of pure lines with poliomyelitis viruses. J. Exptl. Med. 99: Haldane, J. B. S. "Dex" or order of magnitude? Nature 187: Henderson, J. R., and R. M. Taylor Arthropod-bome virus plaques in agar overlaid tube cultures. Proc. Soc. Exptl. Biol. Med. 101: Henderson, J. R., and R. M. Taylor Propagation of certain arthropod-bome viruses in avian and primate cell cultures. J. Immunol. 84: Lennette, E. H General principles underlying laboratory diagnosis of viral and rickettsial infections, p In E. H. Lennette and N. J. Schmidt (ed.), Diagnostic procedures for viral and rickettsial diseases, 3rd ed. Am. Public Health Assoc., Inc., New York. 7. Porterfield, J. S A plaque technique for the titration of yellow fever and related arthropod borne viruses. Nature 183: Porterfield, J. S A plaque technique for the titration of yellow fever virus and antisera. Trans. Roy. Soc. Trop. Med. Hyg. 53: Rubin, H A virus in chick embryos which induces resistance in vitro to infection with Rous sarcoma virus. Proc. Natl. Acad. Sci. U.S. 46: Sawyer, W. A The use of mice in tests of immunity against yellow fever. J. Exptl. Med. 54: Schmidt, N. J Tissue culture methods and procedures for diagnostic virology, p In E. H. Lennette and N. J. Schmidt (ed.), Diagnostic procedures for viral and rickettsial diseases, 3rd ed. Am. Public Health Assoc., Inc., New York. 12. Schulze, I. T., and R. W. Schlesinger Plaque assay of dengue and other group B arthropod-bome viruses under methyl cellulose overlay media. Virology 19: Tauraso, N. M., and A. Shelokov Protection against Junin virus by immunization with live Tacaribe virus. Proc. Soc. Exptl. Biol. Med. 119: Tauraso, N. M., S. L. Spector, W. G. Jahnes, and A. Shelokov Yellow fever vaccine. I. Development of a vaccine seed free from contaminating avian leukosis viruses. Proc. Soc. Exptl. Biol. Med. 127: Theiler, M Susceptibility of white mice to virus of yellow fever. Science 71: Theiler, M Studies on action of yellow fever virus in mice. Ann. Trop. Med. Parasitol. 24: Whitman, L A modified intraperitoneal protection test for yellow fever based on the greater susceptibility of immature white mice to extraneural injection of yellow fever virus. Am. J. Trop. Med. 23: World Health Organization Arthropodborne viruses. Report of a Study Group, World Health Organ. Tech. Rept. Ser. 219, Geneva, p. 68.