The immunogenicity of recombinant LC16mO vaccinia virus harboring HBsAg gene in mice

Biologicals

(1991)

19, ‘77-85

The tmm~ieity Vhs

tBiologica/

of Recombinant LCWmO Va#scinia Harboring HBsAg Glene in Mice

K. Watanabe,* M. *Chiba Science Institute, R *Tokyo Metropolitan §National

Mot-ha,* T. Sato,t K. YausQ. and A. KoJlms§ Serum Institute, Chiba 272, Japan and D Center, Nippon Zeon Co. Ltd., Kawasaki 210, Japan lnstitute for Neurosciences, Tokyo 183, Japan institute of Health, Tokyo 14 1, Japan

Abstract. It was found that hepatitis B surface antigen (HBsAg) expressed by recombinant vaccinia virus (RVV), rProHBm0143, harboring HBsAg gene was immunologically similar to plasma-derived HBsAg and immunogenicity of the rProHBm0143 was possibfe to evaluate by the skin scarification (SS) method using BALB/c mice. When we compared the immunogenicity of 10’ TCIDaO of the rProHBm0143 by the SS method with that of 0.125 ml of the plasma-derived hepatitis B vaccine (HB vaccine) given by intraperitoneal inoculation, ‘the anti-HBs antibody eliciting ability of its RVV was almost the same as that of the HB vaccine with maintenance of high antibody titers, and the antibody responses rose further by re-inoculation in association with HB vaccine, especially by using its RVV as a priming. Also, no virus was recovered from the liver, spleen or brain of the mice inoculated with rProHBm0143 by the SS method. Furthermore, in mice inoculated with rProHBm0143 and then inoculated with RVV harboring Japanese encephalitis virus (JEV) gene 24-weeks later, no effect was recognized on duration of anti-HBs antibody persistence while anti-JEV antibody is being produced. These results suggest that the rProHBm0143 is likely to become a practical live vaccine; a different immunization schedule to protect against hepatitis B virus and two or more kinds of RVV vaccines may be usable for the same animal or humans at intervals of some years. Introduction Recombinant vaccinia virus (RVV) as a live vaccine has a number of useful advantages and suggests the possibility of the development of polyvalent vaccines for several disease agents or vaccines against diseases which are at present di&ult to develop. Development of effective recombinant vaccinia vaccines depends on their ability to elicit protective immunological response against antigen genes inserted into a vaccinia virus gene as well as on the safety of RVVs. We have developed recombinant LC16mO (mO), rF’roHBmOI43 or LC16m8, rBroHBm839,‘*2 successfully expressing hepatitis B surface antigen (HBsAg), and have confirmed that these RVVs are safe for experimental animals and that laboratory attenuation markers inherited from the parental virus strains are genetically conserved with stable expression of the HBsAg gene. 3 The rProHBmO143 was Correspondence should he sent to: Michio Morita, !&6-l, Konodai, Ichikawa, Chiba 272, Japan. 1045-1056/91/020077+09

$03.00/O

highly immunogenic, showing high elicitation of antiHBs antibody in rabbits2 For an important advance in recombinant vaccinia vaccine development, it has been essential to evaluate the elicitation of anti-HBs antibody using suitable experimental animals and various inoculation routes. We now report a successful experiment for evaluation of the immunogenicity of rProHBm0143 by using mice and the practical use of rProHBm0143 for a new immunization schedule to protect against the hepatitis B virus (HBV). Materials

and methods

RVVs, vaccinia virus strain, HBsAg obtained by rProHBm0 143 (vaccinia HBsAg) and plasma-derived hepatitis B vaccine (HB vaccine) Two RWs were subjected in our experiments. The first RW is rProHBmOl43 expressing HBsAg,2 and the second RW is mOJ6 expressing three structural proteins of Japanese encephalitis virus (JEV), i.e. capsid and membrane proteins and envelope glycoprotein plus partial non-structural protein NSl .4 m0 @ 1991 The

International

Association

of Biological

Standardization

78

K. Watanabe et al.

vaccinia virus strain was used as the control virus in immunogenicity tests of RVV or as hemagglutinin antigen for the hemagglutination inhibition (HI) test. These viruses were titrated on monolayers of RK13 cells and expressed as TCIDsc/ml. Vaccinia HBsAg was purified by affmity chromatography using mouse anti-HBs antibodies,2 diluted to a final protein concentration of 40 pg per ml and adsorbed onto aluminum hydroxide (Wake Co., AT8268). A commercial HB vaccine containing approximately 40 pg protein per ml was used (Kitasato Institute, Lot 1). Experimental designs The following experiments were carried out to assess the immunogenicity and organ distribution of rProHBmO143 after its inoculation into 4-week-old BALB/c female mice. 1. Relative immunogenicity of vaccinia HBsAg to HB vaccine (experiment 7). To compare immunogenicity between the vaccinia HBsAg and the plasma-derived HBsAg, vaccinia HBsAg protein (40 pg protein per ml) adsorbed onto aluminum hydroxide (0.4 mg per ml) and commercial HB vaccines were intraperitoneally inoculated into BALB/c mice. Serial two-fold dilutions of the above-described samples were made with physiological saline solution, and 0.1 ml samples were inoculated intraperitoneally into BALB/c mice. At 5 weeks postinoculation (p.i.), each mouse was bled by heart puncture, and serum anti-HBs antibody titers were measured by the passive hemagglutinin (PHA) method. A parallel line assay was performed to analyse statistically the results of the mouse potency tests by the method of Saito et al5

2. Determination of suitable mouse strain or inoculation route for immunization tests of rProHBm0 143 (experiment 2). Preliminary experiments were carried out to determine a suitable mouse strain or inoculation route for immunization of rProHBm0143. The rProHBm0143 was inoculated into each of the outbred ddY and inbred BALB/c, C3H/He, DBAIB, A/J and C57BIJ6 mice by scarificing the skin of the tail-root, using a 24-gauge needle, where virus suspension of lo8 TCIDhc in 25 ~1 was placed on after cutting short hair [the skin scarification (SS) method] or by the intraperitoneal (IP) route. After a suitable mouse strain and inoculation route were determined by preliminary experiments, mouse immunogenicity

tests with rProHBm0143 were carried out by the SS method and with inbred BALB/c mice in comparison with HB vaccine. 3. Dose response of rProHBmO143 (experiment 3). To examine dose response of rProHBm0143,108~o to loo0 TCIDso of virus per mouse was inoculated into BALB/c mice by the SS method; anti-HBs and antivaccinia antibodies were assayed at 4 and 9 weeks p.i., and linearity of the regression line was tested by the method of Finney.’ 4. Long-term follow-up of antibodies and a priming or booster effect of rProHBmO143 (experiment 4). To test persistence of antibodies and strength of a priming or booster effect of rProHBmO143, BALB/c mice were inoculated with 108TCIDsc in 25 ~1 by the SS method. First, antibody response of mice inoculated with rProHBm0143 by the SS method was compared with that of mice inoculated with O-125 ml of HB vaccine by the IP route at 24 weeks p.i. Second, to test the priming or booster effect of the rProHBm0143 or HB vaccine, they were re-inoculated by the reverse procedure (HB vaccine and then rProHBm0143 or rProHBm0143 and then HB vaccine) to the primary inoculation at 24 weeks after the primary inoculation. As the control to the abovedescribed experiment, BALB/c mice were twice inoculated with O-125 ml of HB vaccine at intervals of 24 weeks. The antibody response followed 1 year after the primary inoculation. 5. Antibody response of another RVV after inoculation with rProHBmOl43 (experiment 5). To examine antibody response of another RVV harboring heterogeneous antigen gene against HBsAg gene within rProHBmO143, mOJ6 harboring JEV genes was inoculated into BALB/c mice already treated with rProHBm0143 by the SS method 24 weeks after primary inoculation and bled to measure antibody titers at 2 and 7 weeks later. 6. Virus distribution of rProHBm0143 in mice inoculated by the SS method (Experiment 6). To examine the distribution of rProHBm0143 to the liver, spleen and brain, BALB/c mice were inoculated with lo8 TCIDsc in 25 ~1 per mouse by the SS method or with the m0 strain of the same virus dose by the IP route. Mice of each group were sacrificed at 3,5 and 7 days p.i., and pieces of the liver, spleen and brain were collected to determine the virus contents of their 10% homogenate.

lmmunogenicity

of HBsAg-RW

Quantitation of antibody assays BALB/c mice were bled by cutting the tail vein with a razor, except for the mice used in experiment 1 for the quantification of succeeding antibodies. Serum anti-HBs antibodies were measured by the PHA method using EIZAI Kit (EIZAI Co.> and its antibody titers were expressed as the final dilution showing positive agglutination by the two-fold dilution method. Antibodies to JEV were carried out by the virus neutralizing method with pooled sera. The sera were heat-inactivated at 56°C for 30 min, diluted lo-fold with Hanks’ solution containing 2% calf serum, and equal parts of the serum dilutions and Beijing-l virus solutions of lo2 pfu were well mixed. The mixtures were incubated at 37°C for 15 h and inoculated with a dose of O-4 ml into four cultures of chick embryo cells in Petri dishes. The Beijing-l virus solutions of lo2 pfu were similarly inoculated to serve as controls. All the inoculated cultures were adsorbed at 37°C for 1.5 h in a CO2 incubator and were overlaid with the first overlaying agar medium and incubated for 2 days. The incubated cultures were overlaid with the second overlaying agar medium and further incubated for 1 day. The number of plaques formed on cultures was counted to obtain the reduction rates to the control groups. The neutralizing (NT> antibody titers at 50% plaque reduction rate was calculated as follows: Antibody

titer

Y = plaque reduction

=

,‘,,,“,“,

+ lo!atF

rates; x = serum dilution

(10”)

Anti-vaccinia virus antibodies were measured by the HI method. The titres were expressed as the reciproEa1 of serum dilution that can inhibit aggregation of chick erythrocytes with the m0 strain.7

Results A parallel line assay (experiment 1) Immunizing potency between vaccinia HBsAg preparation and HB vaccine were compared with sera obtained from BALB/c mice inoculated intraperitoneally. The data are presented in Fig. 1, and the variance analyses of the data are summarized in Table 1. No significant deviations from linearity and parallelism were found between the two doseresponse regression lines. Determination of suitable mouse strain or inoculation route for immunization test (experiment 2) Prior to studies for antibody responses and dose responses of rProHBm0143 in BALB/c mice, immunogenicity tests were carried out to examine whether antibody producing ability is different between among mouse strains or inoculation routes and determine a suitable mouse strain or inoculation route for these experiments. There were remarkable differences in anti-HBs antibody eliciting ability among mouse strains in both the SS method (Fig. 2) and the IP route (Fig. 31,

;1.,c I

I 5.0

I 2.5 F9

79

in mice

1

(b) .

I I.25

I

t

.

I x8

I xl6 Dilutm

am I x32

Figure 1. Dose-responseregression lines of vaccinia HBsAg and HB vaccine. Twelve BALB/c mice were divided into three dose-groups, 5,2~5 and 1.25 pg of vaccinia HBsAg adsorbed on aluminum hydroxide (a) and 15 BALB/c mice vireredivided into 3 dilution-groups, 8-(approximately 5 w protein), 16- and 32-fold dilutions of HB vaccine (b). Four mice of each group in the former and 5 mice of each group in the latter were inoculated with the above-described preparations in 0.5 ml by IP route. PHA antibody titer of each mice was measuredat 5 weeks p.i. and a parallel line assay was performed by the method of Saito et aZ.* 0, mean titer.

K. Watanabe et al.

80

Table

1. Variance analyses for the data in Fig. 1

Variation

SS

DF

Preparation Regression Parallelism Linearity

0.12234 11.12097 0*09087 0.01506

Dose Error

11.35004 5.32387

5 21

Total

16.67391

26

PHA 2 I

4 I

u 2;. -I 1.

?iik-

titer

6 I I

MS

1 0.12234 0.483 1 11.12097 43366 1 O-09087 O-358 2 O-00793 O-031

4.325 3.467

2.27001 O-25352

(2”)

Vaccinia

8 I

F(a = 0*05)

Fc

2 F

’

HI titer 4 1

,

(2”)

6 I

1 (6.25OkO.836)

I

1

(4.850+0.2963

I

tzE!

(1.75OItl.774)

(I~ooo+o)

t

2 0

SD)

I

5

Y

(mean+

8 I

(5.2505

(5,000+

0.447

(4.850

+ 0.296)

(4.875

+ 0,536)

)

I

0.974) I I

( I+000

20)

I

I

t

1

(5.250f0.471)

I t

I

-I

1

1

I

(5250+0*974)

I I

I

I

I J I

I

Figure 2. Antibody responseof different mousestrains te HBsAg or vaccinia virus; SS method. 10sTCIDm of rProHBmO143 in 25 ~1was inoculated into eight mice per mousestrains by the SSmethod, asdescribedin Materials anda Methods. PI-IA and anti-vaccinia HI antibody titers of eight serum samples per mouse strain were determined at 4 weeks p.i. Mean antibody titer and standard deviation are shown in parentheses.

lmmunogeniclty

PHA

titer

(2”

of HBsAg-WV

81

in mice

Voccima

1 2

4

HI titer

(2”)

6

8

(mean%SD)

(4*750?0.387)

I.3752

0+x37)

1.000+

(4.625

+ 0.622

)

0)

,279)

t-l 5

p

> s

, (2.375+2.141)

,

, I

(5.375kO.622)

1

I

I

(5.375_C2.363)

I

I

Figure 3. Antihodi response of difference mouse strains to HBsAg or vaccinia virus; IP route. 10 ‘ICI&, of rProHBmO143 in 100 ,~l was inoculated into eight mice per mousestrain by the IP route. PHA and anti-vaccinia HI antibody titers of each serum sample per mouse strain were determined as the SS method. Mean antibody titer and standard deviation are shown in parentheses.

whereas no remarkable difference was recognized in anti-vaccinia antibody producing ability among mouse strains in either inoculation route. The levels of antibody titers to HBsAg were higher in mice inoculated by the SS method than in mice inoculated by the IP route. In particular, BALB/c mice had much higher anti-HBs antibody titers than the other strains without varying levels of its antibody titers. Anti-HBs antibody was not detected in A/J mice inoculated by the SS method, in C57BW6 mice inoculated by the IP method and in C3H/He mice inoculated by both routes, while considerably high anti-vaccinia titers were produced in these mice.

Dose response of rProHBmO143 (experiment 3)

To examine dose response of rProHBmOl43, three single doses of 106, lo7 or 10” TCID50 in 25 ~1 per mouse were inoculated by the SS method. Each BALBlc mouse was bled at 4 and 9 weeks p-i. for titration of anti-HBs antibody by the PHA method. The results are presented in Fig. 4. The dose response regression lines at both 4 and 9 weeks p.i. appeared linear in the range from lo6 to 10s TClDm per mouse. Long-term follow-up of antibodies and a priming or booster effect of rProHBmO143 (experiment 4) To examine the persistence of antibodies and a

82

K. Watanabe et al.

(bi

I 6

I 7

I 0

6 log,,

7

8

TCID,,/mL

Figure 4. Dose response of BALB/c mice inoculated with rProHBmO143 at: (a) 4 weeks p.i.; (b) 9 weeks p.i. lOa, 10’ or lo6 TCID50 of rProHBm0143 in 25 ~1was inoculated into 20 BALBlc mice per virus dose, and 10 mice of each group were sacrificed at 4 or 9 weeks p.i. to determine PHA titer (@I, individual titer.

priming or booster effect of rProHBm0143 in association with HB vaccine, BALB/c mice were inoculated by the SS method or the IP route. PHA antibody titers of each mouse inoculated with rProHBm0143 by the SS method or with HB vaccine by the IP route rose to the high level of 2’ to 21° at 9 weeks p.i. The levels of anti-HBs antibody titers elicited by rProHBmO143 appeared to be similar to those produced by HB vaccine throughout 24 weeks follow-up observation period after the primary inoculation. When mice were re-inoculated with rProHBm0143 by the SS method or with HB vaccine by the IP route at 24 weeks after primary inoculation, the PHA antibody titers rose rapidly to approximately 2i5 to 21g, and the high antibody titers were maintained 28 weeks after booster inoculation, that is, by 1 year after primary inoculation. In particular, the mice inoculated with rProHBmO143 as the priming inoculation and HB vaccine as the booster inoculation maintained much higher titers than the mice with HB vaccine as the priming inoculation and rProHBm0143 as the booster inoculation or the mice inoculated twice with HB vaccine. In contrast, the anti-HBs antibody titer in the mice inoculated only once with rProHBmO143 decreased gradually. Finally, the magnitude of the antibody of the mice inoculated first with rProHBm0143 and later with HB vaccine compared with the mice inoculated with rProHBm0143 only had approximately 32-fold higher antibody titer when tested at the end of the experiment.

Anti-vaccinia HI titers rose to approximately 26 at 5 weeks p.i. in the mice receiving rProHBmOl43 as the priming inoculation, and thereafter their antibody titer decreased gradually and resulted in a titer of approximately 24 at 24 weeks p.i. Similarly, antivaccinia HI titer rose to 26 at 2 weeks after booster inoculation in mice inoculated with HB vaccine and subsequently inoculated with rProHBm0143 (Fig. 5). Antibody responses of mOJ6 in the mice inoculated with rProHBmO743 (experiment 5) mOJ6 was inoculated into rProHBm0143-infected BALB/c at 24 weeks after primary inoculation, and anti-JEV or anti-vaccinia antibody was measured 2 and 5 weeks later. NT antibody to JEV rose to 1Oo’ss 2 weeks later and 105.5 5 weeks later. HI antibody to vaccinia virus was boosted from approximately 23 to 25 at 2 weeks and to 28 at 5 weeks after the inoculation of the mOJ6. However, the results of anti-HBs antibody after the inoculation with the mOJ6 seemed to show that there was no effect on the duration of anti-HBs antibody persistence compared with the results for the mice inoculated only once with 10sTCID60 of rProHBmO143 in experiment 5 (Fig. 6). Virus distribution (experiment 6) To study the virus distribution in BALB/c mice inoculated with rProHBm0143 or m0 strain by the

lmmunogeniclty

Times after

of HBsAg-RW

primary

inoculation

in mice

(weeks)

Figure 6. Kinetics of antibody responseto rProHBmO143 or its clone in associationwith HB vaccine. BALBk mice were inoculated with 10’ TCID~ of rProHBm0143 (10 mice) or of m0 strain (5 mice) in 25 ,ol by the SS method, and five out of the 10 mice inoculated with the rProHBm0143 were re-inoculated with 0.125 ml of HB vaccine 24 weeks later. Ten mice were inoculated with 0.125 ml of HB vaccine by the IP route; five out of the 10 mice were re-inoculated with O-125ml of HB vaccine by the IP route, and the other five with 108TCIDm of rProHBmO143 in 25 pl by the BS method 24 weeks later. Bnmunological responseof all the mice wasfollowed 1 year after the primary inoculation. Anti-HBeAg and anti-vaccinia antibody titers in collected serum sampleswere measured as describedin Materials and Methods. Each point representa mean antibody titere. &, HB vaccine + rProHBm0143; --C-, rProHBmO143 -+ HB vaccine. -A-, HB vaccine+ HB vaccine. -A -, rF’roHBm0143 only; -Cl-, LC 16m0 strain only.

SS method in a direct comparison with the result of the IP route, I3ALBk mice were sacrificed at 3,5 and 7 days p.i. As shown in Fig. 7, no virus was isolated from the liver, spleen or brain of mice inoculated with m0 and rProHBmO143 by the SS method during 7 days observation period. On the other hand, virus was isolated from the liver and spleen at 3 days p.i. in the mice inoculated with rProHBmO143 and at 3 and 5 days in the mice inoculated with m0 strain by the IP route, but neither virus is detected in the brain. The mice inoculated with the m0 strain had much higher infectivity titers than those inoculated with the rProHBm0143. DIscussion From the results of the parallel was determined that vaccinia

line assay (Fig. l), it HBsAg has similar

immunological characteristics to plasma-derived HBsAg, and this fact will con&m that vaccinia HBsAg has similar physi~biological characteristics to plasma-derived HBsAg, as reported previously.2 Based on the foregoing results, mouse immunogenicity tests of rProHBmO143 were carried out to quantify the ability to elicit anti-HBs antibody in comparison with that of the HI3 vaccine. Furthermore, considering the practical use of rProHBm0143 to protect HBV infection, we thought that it was necessary to evaluate the priming or booster effect of rProHBm0143 by using it with HI3 vaccine. Prior to the immunological tests with rProHBm0143 in mice, we first of all planned experiments to determine suitable mouse strain or inoculation route for the tests. According to the present experiment with rProHBm0143, strain differences in mice were

84

K. Watanabe et al.

>6::

I

5

I

IO Tomes after

I I I

24 primary

lnoculotion

26

29

(weeks)

Figure! 6. Antibody responseto mOJ6 in rProHBm0143-infected BALB/c mice.mOJ6 was re-inoculated into 5 BALB/c mice inoculated already with 10sTCID50 of rProHBm0143 in 25 ~1 at 24 weeks after primary inoculation and bled to measure antibody titers 2 or 5 weeks later. Anti-JEV antibody titer in collected serum sampleswas measured as describedin Materials and Methods.

recognized in immunological response to vaccinia HBsAg expressed by rProHBmO143 when inoculated by the SS method or the IP route. There was, however, no notable strain difference in immunological response to vaccinia virus between the SS method and the IP route. BALBlc female mice inoculated by the SS method were comparatively high responders without varying levels of antibody activity among the strains tested. Thus, we determined that the SS method using BALB/c female mice were most suitable for the immunological test with RVV, rProHBm0143. It had been shown that murine humoral immune responses to HBsAg determinants are regulated by H-2 linked immune response genes’ and BALB/c female mice of H-2d haplotype were adequate for potency assay of HB vaccine from the results of anti-HBs antibody responses.g From the long-term follow-up studies of antibody responses in BALB/c female mice inoculated with rProHBm0143, the following encouraging results were observed. (1) The levels of anti-HBs antibody titers elicited by rProHBm0143 appeared to be similar to those produced by the HB vaccine throughout the 24 weeks follow-up observation period, in a direct comparison with their immunoge-

nicity between lOa TCIDao of the rProHBm0143 and O-125 ml of HB vaccine. (2) Higher anti-HBs antibody titers were obtained with the method of inoculating in association with rProHBm0143 and HB vaccine, first with the former and second with the later or the reverse inoculation procedure, than in single inoculation method; in addition high antibody titers were maintained 1 year after primary inoculation. (3) Furthermore, the secondary responses were higher in the mice inoculated with rProHBmO143 as a priming inoculation than in mice inoculated with the RW as a booster or in the mice inoculated twice with HB vaccine. (4) Dose responses were recognized by statistical analyses among virus doses of lo6 to 10’ TCIDso of rProHBm0143 as shown in the HB vaccine.“.” The results indicated that the rProHBm0143 was likely to become a practical live vaccine because of its high immunogenicity , and immunological data on the rProHBm0143 in association with HB vaccine raise the possibility of another different immunization schedule to protect HBV, considering also the result that no virus recovered from the liver, spleen or brain of the mice inoculated with the rProHBmO143 by the SS method. In the mice inoculated with rProHBm0143 and

lmmunogeniclty

of HBsAg-RW

in mice

85

References

Doys

after

inoculotlon

Figure 7. Virus distribution in mice inoculated with rPmHBmO143 or m0 strain by the SSmethod or IP route. Sixty BALB/c mice were divided into four groups consisting of 15 mice per group. Fifteen mice of each group were inoculated with 10s TCIDM of rProHBm0143 by the SS method or by IP route, or lOsTCID5o of m0 strain by the SS method or IP mute, respectively. Five mice of each group were sacrificedat 3,5 and 7 days p.i., and virus infectivity in the liver, spleen and brain were determined. Mean virus infectivity titer of 5 mice is shown. ,Ip;---,ss.o, LC 16m0, liver; A, LClGmO, spleen;Cl,LClGmO, brain; l , rProHBmO143, liver; A, rProHBm0143, spleen; n , rProHBm0143, brain.

subsequently inoculated with mOJ6 24 weeks later, no effect was recognized on the duration of anti-HBs antibody persistence while anti-JEV antibody rose by almost the same level, as recognized in mice inoculated once with mOJ6 (data not shown). It was, therefore, suggested that two or more kinds of RVVs harboring different antigen gene could be usable as

live vaccines for the same individual animal or humans at intervals of some years, as shown by other workers. i2,13

1. Morita M, Suzuki K, Yasuda A, Kojima A, Sugimoto M, Watanabe K, Kobayashi H, Kajiyama K, Hashixume S. Recombinant vaccinia virus LC16mO or LC16m8 that expresseshepatitis B surface antigen whiIe preserving the attenuation of the parental virus strain. Vaccine 1987; 5: 65-70. 2. Watanabe K, Kobayashi H, Ksjiyama K, Morita M, Yasuda A, Gotoh H, Saeki S, Sugimoto M, Saito H, Kojima A. Improved recombinant LC16mO or LC16m8 vaccinia virus successfully expressing hepatitis B surface antigen. Vaccine 1989; 7: 53-59. 3. Watanabe K, Morita M, Kojima A. Stability of recombinant vaccinia virus LC16mO or LC16m8: Preserved laboratory attenuation markers and conservedexpressionof inserted HBsAg gene.Vaccine 1989;7: 499-502. 4. Yasuda A, Kimura-Kuroda J, Ogimoto M, Miyamoto M, Sat0 T, Takamura C, Kurata T, Kojima A, Yasui K. Induction of protective immunity in animals vaccinated with recombinant vaccinia viruses that express pre M and E glycoproteins of Japanese encephalitis virus. J Virol199O; 64: 278s-2795. 5. Finney D-J. Statistical Method in Biological Assay, 2nd edn. London: Charles Griffin, 1971. 6. Saito H, Ishida, S, Yasuda J, Kurokawa M, Takahashi T. Application of the parallel line bioassay method to quantitative determination of HBs antigen in radioimmunossay. Japan J Med Sci Biol 1979; 32:

47-52. 7. Nakano JH. Poxviruses.

In: Lennette EH, Schmidt NJ, eds. Diagnostic Procedures for Viral, Rickettsial and Chlamydial Infection, 5th edn. Washington DC: American Public Health Association Press, 1979. 8. Mihch DR, Chisari FV. Genetic regulation of the immune response to hepatitis B surface antigen U-IBsAg). 1. H-2 restriction of the murine humoral immune response to the a and d determinants of HBsAg. J. Immunol1982; 129: 320-325. 9. Yuasa T, Shimojo H. Potency tests of hepatitis B vaccine by parallel line assaymethod in mice. Japan J Med Sci Biol 1985; 38: 9-18. 10. Takahashi T, Saito H, Kumkawa M, Tasuda J, Hashimoto T. Studies on the potency test of HB vaccine by the parallel line assay method in guinea pigs. Japan J Med Sci Biol 1982; 35: 53-62. 11. Marie-Christine M, Liliane R, Marc G, AIain G. Immunogenicity of hepatitis B vaccine in guinea-pigs, mice and man. J Biol Stand 1985; 13: 221-227. 12. Paoletti E, Lipinska BR, Samsonoff C, Mercer S, PanicaRi D. Construction of the live vaccines using genetically engineered poxviruses; biological activity of vaccinia virus recombinant expressing the hepatitis B virus surface antigen and herpes simplex virus glycoprotein D. Proc Nati Acad Sci USA 1984,81: 193-197. 13. Perkus ME, Picini A, Lipinska BR, Paoletti E. Recombinant vaccinia virus: Immunization against multiple pathogen. Science 1985; 229: 981-984.

Received for publication 16 July 1990; accepted 19 November 1990.