Research Paper |
|
|||
|
Acta Biochim Biophys Sin |
||||
|
doi:10.1111/j.1745-7270.2005.00101.x |
Vaccination against Very Virulent
Infectious Bursal Disease Virus Using Recombinant T4 Bacteriophage Displaying
Viral Protein VP2
Yong-Chang CAO*, Quan-Cheng
SHI, Jing-Yun MA, Qing-Mei XIE, and Ying-Zuo BI
Genetic
Engineering Laboratory, Animal Science College, South China Agricultural
University, Guangzhou 510642, China
Received: June 20,
2005
Accepted: July 30,
2005
This work was
supported by the grants from the National High Technology Research and
Development Program of China (No. 2002AA245051) and the National Natural
Science Foundation of China (No. 39870550)
*Corresponding
author: Tel, 86-20-85280283; Fax, 86-20-85280740; E-mail, [email protected]
Abstract In order to develop a desirable
inexpensive, effective and safe vaccine against the very virulent infectious
bursal disease virus (vvIBDV), we tried to take advantage of the emerging T4
bacteriophage surface protein display system. The major immunogen protein VP2
from the vvIBDV strain HK46 was fused to the nonessential T4 phage surface
capsid protein, a small outer capsid (SOC) protein, resulting in the 49 kDa
SOC-VP2 fusion protein, which was verified by sodium dodecylsulfate
polyacrylamide gel electrophoresis and Western blot. Immunoelectromicroscopy
showed that the recombinant VP2 protein was successfully displayed on the
surface of the T4 phage. The recombinant VP2 protein is antigenic and showed
reactivities to various monoclonal antibodies (mAbs) against IBDV, whereas the
wild-type phage T4 could not react to any mAb. In addition, the recombinant VP2
protein is immunogenic and elicited specific antibodies in immunized specific
pathogen free (SPF) chickens. More significantly, immunization of SPF chickens
with the recombinant T4-VP2 phage protected them from infection by the vvIBDV
strain HK46. When challenged with the vvIBDV strain HK46 at a dose of 100 of
50% lethal dose (LD50) per chicken 4 weeks after the booster
was given, the group vaccinated with the T4-VP2 recombinant phage showed no
clinical signs of disease or death, whereas the unvaccinated group and the
group vaccinated with the wild-type T4 phage exhibited 100% clinical signs of
disease and bursal damages, and 30%–40%
mortality. Collectively, the data herein showed that the T4-displayed VP2
protein might be an inexpensive, effective and safe vaccine candidate against
vvIBDV.
Key words infectious bursal disease
virus; VP2; T4 bacteriophage; vaccination
Infectious bursal disease virus (IBDV) is a member of the Birnaviridae
family, the genome of which consists of two segments (A and B) of
double-stranded RNA [1]. Segment B (approximately 2.8 kb) encodes VP1, a 90
kDa protein with RNA polymerase and capping enzyme activities [2]. Segment A
(approximately 3.2 kb) encodes a polyprotein that can be cleaved by
autoproteolysis to form mature viral proteins VP2, VP3 and VP4 [3]. IBDV
particles have non-enveloped, icoasahedral capsids with a diameter of
approximately 60 nm. The structure of the virus is based on the T=13 lattice
and the capsid subunits are predominantly trimer clustered [4]. VP2 and VP3
form the outer and inner capsid of the virus, respectively [4,5]. The antigenic
site responsible for the induction of neutralizing antibodies is highly
conformation-dependent and located on VP2 [6].
IBDV multiplies rapidly in developing B lymphocytes in the bursa of
Fabricius, and causes infectious bursal disease (IBD) in chickens, as well as
other diseases, by immunosuppression [7,8]. Control of IBD in young chickens
had been primarily achieved by vaccination with a live attenuated strain of
IBDV at the age of 0 to 5 weeks or by transferring high levels of maternal
antibody induced by the administration of live and killed IBD vaccines to
breeder hens [8]. However, since the late 1980s, the emergence of very virulent
IBDV (vvIBDV) has rendered IBD even more difficult to control [9–12]. vvIBDV
showed antigenicity similar to the classical strains [10], and can break
through high levels of maternal antibody and cause 60%–100% mortality in specific
pathogen free (SPF) chickens. More challengingly, vaccines against vvIBDV will
induce pathogenic effects [13]. Recently, in vivo studies showed that
vvIBDV adapted to chicken embryo cell cultures using site-directed mutagenesis
and a reverse genetics approach was partially attenuated [14]. It is apparent
that the risk of reversion of the partially attenuated vvIBDV to wild-type
vvIBDV limits its application as a potential live vaccine.
A subunit vaccine overcomes the risk of reversion. As the main host
protective antigen harboring most of the neutralization sites, VP2 has been
used for the development of subunit vaccines in the following schemes.
(1) Immunization of chickens with the expression vector that in
vivo expresses VP2 continuously, for example, avian herpesvirus vector
[15], fowlpox virus vector [16], fowl adenovirus vector [17], Marek’s disease
virus vector [18] and Semliki Forest virus vector [19]. However, this method
is costly and complicated.
(2) Expression of VP2 as the recombinant protein in systems
including Escherichia coli [20], yeast [21] and insect cells [22].
However, the recombinant VP2 expressed in E. coli failed to provide full
protection upon challenge [20]. Although the VP2 expressed in yeast and insect
cells proved to be effective, the cost is also very high.
(3) Production of virus-like particles of IBDV. Virus-like particles
resemble the native viral capsids structurally and immunologically and elicit
excellent antibody responses [23], but the cost of production is very high.
The T4 bacteriophage surface protein display system enables the
fusion of foreign proteins with the C terminal of small outer capsid (SOC)
proteins and the N terminal of highly antigenic outer capsid (HOC) proteins, so
that foreign proteins can be displayed on the surface of the T4 phage [24]. In
addition to advantages possessed by other phage display systems, this system
features large capacity, high copy numbers of interesting antigens, and
assemblage of the T4 phage inside the host cell without recourse to a
secretory pathway. The system has been applied to display several peptides,
including a 271-amino acid heavy and light chain fused IgG anti-egg white
lysozyme antibody [24], the CD4 receptor of a 183-amino acid HIV-1 [25] and a
36-amino acid PorA peptide from Neisseria meningitides [26]. However,
it has been a challenge to display any protein or polypeptide with more than
400 amino acid residues.
HK46 is a strain of vvIBDV isolated from southern China [10]. We
have reported the full-length cDNA sequence of the vvIBDV HK46 strain [10,27].
In this study, VP2 of the vvIBDV HK46 strain consisting of 441
amino acids was successfully expressed using the T4 phage surface protein
display system. The recombinant VP2 maintains its immunogenicity and
antigenicity, and renders full protection of SPF chickens against the challenge
of the vvIBDV HK 46 strain.
Materials and Methods
Plasmids, bacterial strains, T4 phage and IBDV
The integrative plasmid pR and phage T4-Z1 were kindly provided by
Dr. Zhaojun REN (Department of Biological Chemistry, School of Medicine,
University of Maryland, Baltimore, USA) [24] and maintained at the Animal Genetic
Engineering Laboratory (South China Agricultural University, Guangzhou, China).
Phage T4-Z1 is a lysozyme-dependent SOC gene-deleted mutant of the T4
phage. The integrative plasmid pR contains the SOC gene, part of the
lysozyme gene upstream of SOC, and part of the denV gene
downstream of SOC. The partial lysozyme gene and denV were used
for homologous recombination with T4-Z1. Foreign genes were inserted into the 3‘
terminus of the SOC gene at the EcoRI site. The HK46 strain was
propagated in 5-week old SPF chickens.
Amplification of VP2 cDNA
Segment A of the HK46 strain has been cloned into the plasmid
pAlter-FA [27]. Primers VP2-S/VP2-A were employed to amplify the cDNA of the VP2
gene by polymerase chain reaction (PCR) from the plasmid pAlter-FA. The primer
VP2-S (5‘-TGAAGAATTCTATGACGAACCTGCAA-3‘) was synthesized
according to nucleotide 131–145 of segment A and contained an EcoRI restriction site in
italic. The primer VP2-A (5‘-ATTTGAATTCCTATAGTGCCCGAATTATGTCCTT-3‘)
was synthesized according to nucleotide 1463–1480 of segment A and
contained an EcoRI restriction site in italic and a TAG termination
codon in bold. The length of the amplified VP2 cDNA was expected to be
1350 bp.
Construction and amplification of the recombinant plasmid
According to Ren et al. [24], the VP2 cDNA was digested
by EcoRI and ligated to a CIAP-treated pR plasmid by T4 DNA ligase,
resulting in the recombinant calculate the B/B value. The bursa index accounted
for the ratio of the B/B value of the testing group and that of the control
group.
Results
Integration of HK46 VP2 cDNA into T4-Z1 phage
The VP2 cDNA was approximately 1350 bp in length. Primers
VP2-S and VP2-A, corresponding to the 5‘ end and the 3‘ end of
the VP2 cDNA, respectively, were employed to amplify the VP2 cDNA
from a plasmid containing segment A of the vvIBDV strain HK46. In order to
facilitate the subsequent cloning into the integrative plasmid, an EcoRI
restriction site was incorporated into each primer. The amplified VP2
cDNA was cloned into the integrative plasmid pR to form pR-VP2, which was
amplified in bacteria.
To express VP2 in the T4-VP2 phage, correct orientation
within the pR-VP2 was needed. The orientation of the VP2 cDNA within the
pR-VP2 was verified by either PCR or direct sequencing (data not shown). It is
noted that for the verification by PCR, one primer SOC-S corresponding to the
5‘ terminus sequence of the SOC gene of T4 phage was employed to
amplify the SOC-VP2 fusion gene together with the other primer VP2-A
corresponding to the 3‘ terminus sequence of VP2 cDNA. If the
correct PCR product (1600 bp in our case) was obtained, it strongly suggested
or even verified that the integrated VP2 had the correct orientation.
The integration of the VP2 cDNA in pR-VP2 into the T4-Z1
phage through homologous recombination was done following the procedure
described in “ Materials and Methods”. The resultant T4-VP2 recombinant phage
was selected and verified by either PCR or direct sequencing (data not shown).
VP2 expression in T4-VP2 recombinant phage
T4-VP2 recombinant phages and T4-Z1 mutant-type phages were enriched
by dialysis and subjected to SDS-PAGE analysis. In our experiments, the titer
of the phages reached over 1010 phages/ml. In the T4-VP2 phage, VP2
is expressed as part of the SOC-VP2 fusion protein with a molecular weight of
approximately 49 kDa. On the electrophoretogram, an extra band of approximately
49 kDa was visible in the T4-VP2 sample. Following SDS-PAGE, the proteins were
transferred to a nitrocellulose membrane, which was subjected to Western blot
with IBDV monoclonal antibody R63 against VP2. The band of approximately 49 kDa
was specifically recognized by the R63 antibody (Fig. 1). It
demonstrated that VP2 could be expressed in the T4 phage display system, even
though the size of VP2 was larger than those peptides or proteins previously
expressed.
VP2 expression on the surface of T4-VP2 recombinant phage
Reactivities of various mAbs and chicken sera to IBDV and phages are
shown in Table 1. The D78 strain of IBDV reacted to anti-sera and mAbs
B29, B69, R63, 9-6, 17-82 and 39A, but not mAb 3-1. The HK46 strain of vvIBDV
and recombinant phage T4-VP2, in which VP2 from the HK46 strain was expressed
on the head of the phage, showed the same reactivities to anti-sera and various
mAbs. Both of them showed reactivities to mAbs B29, R63, 3-1, 9-6, 17-82 and
39A, whereas the wild-type phage T4 could not react to any antibody against
IBDV. The ELISA data demonstrated that the VP2 expressed on the surface
of the T4-VP2 phages was antigenic.
Immunoelectronmicroscopy was employed to directly show whether the
VP2 protein was expressed on the surface of the T4-VP2 recombinant phages.
Under the transmission electron microscope, both the wild-type and T4-VP2 recombinant
phages presented typical morphology, the heads and tails were distinctly
visible in white (Fig. 2). When the T4-VP2 recombinant phages were
stained with mAb R63, only the heads were positively stained by the antibody
R63, as indicated by the increased electron density, which showed that VP2 was
expressed on the surface of the T4-VP2 recombinant phages.
T4-VP2 recombinant phage as vaccine against vvIBDV
We tested whether the surface-expressed VP2 could be used as an effective
vaccine against the HK46 strain of vvIBDV. White leghorn chickens were divided
into three groups (10 chickens/group), immunized and boosted with T4-VP2
recombinant phage (Group 1), T4 wild-type phage (Group 2) or saline (Group 3).
In another negative control group, chickens were not immunized or challenged.
The specific antibody of IBDV was detected with a commercial ELISA
kit and quantified using A630. A630 for the
negative control serum is represented as N, A630
value for positive control sera as P, A630
value for the sera from any group as S, and the IBDV antibody level as S/P:
S/P=(S–N)/(P–N)
The IBDV antibody levels for the three groups over six consecutive
weeks are summarized in Table 2.
As can be seen from Table 2, only Group 1 immunized with the
T4-VP2 recombinant phage vaccine showed IBDV-specific antibodies. At the end of
these experiments, when the chickens were 8 weeks old, the T4-VP2 vaccine
elicited the specific antibodies of IBDV in about 90% of the positive control
in the sera. As expected, the chickens in Group 2, immunized with the T4
wild-type phage, showed no IBDV-specific antibodies, and the same result was
seen in Group 3.
To investigate whether the specific antibodies of IBDV induced by
T4-VP2 were effective to protect the immunized chickens from infection by the
HK46 strain of vvIBDV, all chickens from the three groups were challenged with
the HK46 strain at a dose of 100 of LD50 per bird 4
weeks after the booster was given. The health of all the challenged chickens
was recorded 7 d post-infection and the results are summarized in Table 3.
No chicken from the group vaccinated with T4-VP2 showed clinical signs of IBD.
In contrast, all chickens from the Group 2, immunized with wild-type T4 phage,
and Group 3, given saline, showed clinical signs of IBD. Four chickens in Group
2 and three in Group 3 died within the recording period. The apparently healthy
chickens from the group immunized with the T4-VP2 showed variable signs of
temporary bursa damages that were soon fully repaired. At day 7 post-infection
with the HK46 strain, the bursa indexes of chickens immunized with T4-VP2, the
wild-type T4 phage and saline were 0.93, 0.72 and 0.78 respectively (Table
4).
Discussion
We have demonstrated here that the VP2 protein expressed on the
surfaces of T4 phages maintained enough requisite conformational epitopes to
elicit high humoral responses in immunized chickens and, more importantly, to
effectively protect the immunized chickens from infection by vvIBDV.
The development of an inexpensive, effective and safe vaccine
against vvIBDV has been an ongoing effort. When vvIBDV first appeared, the
effective vaccines were partially attenuated vvIBDV. However, they carry the
intrinsic hazard of being pathogenic due to incomplete inactivation of the
virulent virus, or due to a reversal to virulence of the attenuated viruses. To
overcome the shortcomings of whole virus-based vaccines, a variety of subunit
vaccines based on VP2 have been reported [14–23].
VP2 is known to be the main host protective antigen. The antigenic
epitopes located within VP2 are highly conformation-dependent and able to
elicit serologically-specific neutralizing antibodies. In addition, the mature
VP2 protein in the vvIBDV particle contains more than one glycoside by
post-translational glycosylation. It has been assumed that proper glycosylation
is required for maintaining the conformational epitopes of VP2. It is natural
that most of the efforts on developing VP2-based vaccines have been focused on
eukaryotic expression systems, for example, formation of virus-like particles
in insect cells [23,30], expression of recombinant VP2 proteins in yeast [21]
or insect cells [22], and expression of VP2 in vivo using different
expression vectors [15–19]. However, none of these eukaryotic expression systems are
practically applicable in developing countries because of their complexity and
the high costs of production and administration. Nonetheless, it was reported
that expression of VP2 together with VP3 and VP4 in E. coli produced
viral particles. Even though the viral particles from E. coli only
provided partial protection against infection by vvIBDV, it strongly suggested
that certain conformational epitopes on VP2 proteins were properly maintained
in the viral particles.
The T4 phage surface protein display system is able to display
heterologous proteins as fusion proteins with its own SOC and HOC proteins on
the surface of the phage particles. In consideration of the spatial limits for
the display of foreign proteins on the surfaces of T4 particles, the sizes of
the diversified peptides or proteins displayed are relatively small [24–26]. It is
possible to suspect that a large foreign protein may destroy the configuration
of the endogenous surface proteins, resulting in no production of T4 viral
particles.
Our results have shown that the VP2 protein of 441 amino acids from
the HK46 strain was successfully expressed on the surface of the T4 phages as
the VP2-SOC fusion protein. Although the largest size of the foreign protein
that could be displayed on the surface of T4 phages is not known, the
expression of VP2 clearly demonstrated that the T4 display system has great
elasticity in the accommodation of large size proteins. Our work opens the
door for displaying large proteins on the T4 phages.
As discussed earlier, VP2 proteins undergo glycosylation during
maturation. VP2 from the HK46 strain has three potential glycosylation sites.
Even though VP2 proteins displayed on the T4 phages lack any glycosylation,
they not only elicit a high level of specific antibodies for IBDV, but also
provide effective protection against the challenge of the HK46 strain. We do
not have data to compare the configuration of VP2 displayed on T4 phages with
its configuration on native particles, but we have every reason to believe
that the displayed VP2 proteins maintain their conformational epitopes. It is
plausible that the maintenance of conformational epitopes may be due to the
spatial constraints imposed on VP2 by SOC.
The VP2 displayed on the T4 phages as vaccines conferred full
protection to the immunized birds. All the birds immunized with the T4-VP2 were
free of clinical signs of disease or death. It is noted that an anatomical investigation
revealed temporary bursal damages in the immunized birds. The bursal damages
were repaired spontaneously shortly after infection. This system will be a
useful tool in studying the molecular mechanisms that explain how vvIBDV
causes bursal damages, clinical signs of disease or death.
In summary, we have developed an effective and safe VP2-based
subunit vaccine against vvIBDV. The employment of the T4 display system
provides various advantages, including inexpensive production, convenient
storage and transportation, and easy administration. This inexpensive,
effective and safe vaccine is particularly applicable to developing countries.
Acknowledgements
We thank Dr. Zhao-Jun REN for kindly
providing plasmid pR and phage T4-Z1, and Dr. George D. LIU for kindly
revising the manuscript.
References
1 Muller H, Scholtissek C, Becht H.
Genome of infectious bursal disease virus consists of two segments of
double-stranded RNA. J Virol 1979, 31: 584–589
2 Spies U, Muller H, Becht H.
Properties of RNA polymerase activity associated with infectious bursal
disease virus and characterization of its reaction products. Virus Res 1987, 8:
127–140
3 Muller H, Becht H. Biosynthesis of
virus-specific proteins in cells infected with infectious bursal disease virus
and their significance as structural elements for infectious virus and
incomplete particles. J Virol 1982, 44: 384–392
4 Hudson PJ, McKern NM, Power BE,
Azad AA. Genomic structure of the large RNA segment of infectious bursal
disease virus. Nucleic Acids Res 1986, 14: 5001–5012
5 Bottcher B, Kiselev NA,
Stel’Mashchuk VY, Perevozchikova NA, Borisov AV, Crowther RA. Three-dimensional
structure of infectious bursal disease virus determined by electron
cryomicroscopy. J Virol 1997, 71: 325–330
6 Caston JR, Martinez-Torrecuadrada
JL, Maraver A, Lombardo E, Rodriguez JF, Casal JI. C terminus of
infectious bursal disease virus major capsid protein VP2 is involved in
definition of the T number for capsid assembly. J Virol 2001, 75: 10815–10828
7 Becht H, Muller H, Muller HK.
Comparative studies on structural and antigenic properties of two serotypes of
infectious bursal disease virus. J Gen Virol 1988, 69: 631–640
8 Becht H. Infectious bursal disease
virus. Curr Top Microbiol Immunol 1980, 90: 107–121
9 van den Berg TP. Acute infectious
bursal disease in poultry: A review. Avian Pathol 2000, 29: 175–194
10 Cao YC, Yeung WS, Law M, Bi YZ, Leung FC,
Lim BL. Molecular characterization of seven Chinese isolates of infectious
bursal disease virus: Classical, very virulent, and variant strains. Avian Dis
1998, 42: 340–351
11 Chettle N, Stuart JC, Wyeth PJ. Outbreak
of virulent infectious bursal disease in East Anglia. Vet Rec 1989, 125: 271–272
12 van den Berg TP, Gonze M, Meulemans G.
Acute infectious bursal disease in poultry: Isolation and characterization of a
highly virulent strain. Avian Pathol 1991, 20: 133–143
13 Mazariegos LA, Lukert PD, Brown J.
Pathogenicity and immunosuppressive properties of infectious bursal disease
“intermediate” strains. Avian Dis 1990, 34: 203–208
14 van Loon AA, de Haas N, Zeyda I, Mundt E.
Alteration of amino acids in VP2 of very virulent infectious bursal disease
virus results in tissue culture adaptation and attenuation in chickens. J Gen
Virol 2002, 83: 121–129
15 Tsukamoto K, Saito S, Saeki S, Sato T,
Tanimura N, Isobe T, Mase M et al. Complete, long-lasting protection
against lethal infectious bursal disease virus challenge by a single
vaccination with an avian herpesvirus vector expressing VP2 antigens. J Virol
2002, 76: 5637–5645
16 Shaw I, Davison TF. Protection from
IBDV-induced bursal damage by a recombinant fowlpox vaccine, fpIBD1, is
dependent on the titre of challenge virus and chicken genotype. Vaccine 2000,
18: 3230–3241
17 Francois A, Eterradossi N, Delmas B,
Payet V, Langlois P. Construction of avian adenovirus CELO recombinants in
cosmids. J Virol 2001, 75: 5288–5301
18 Tsukamoto K, Kojima C, Komori Y, Tanimura
N, Mase M, Yamaguchi S. Protection of chickens against very virulent infectious
bursal disease virus (IBDV) and Marek’s disease virus (MDV) with a recombinant
MDV expressing IBDV VP2. Virology 1999, 257: 352–362
19 Phenix KV, Wark K, Luke CJ, Skinner MA,
Smyth JA, Mawhinney KA, Todd D. Recombinant Semliki Forest virus vector
exhibits potential for avian virus vaccine development. Vaccine 2001, 19: 3116–3123
20 Rogel A, Benvenisti L, Sela I, Edelbaum
O, Tanne E, Shachar Y, Zanberg Y et al. Vaccination with E. coli
recombinant empty viral particles of infectious bursal disease virus (IBDV)
confer protection. Virus Genes 2003, 27: 169–175
21 Pitcovski J, Gutter B, Gallili G, Goldway
M, Perelman B, Gross G, Krispel S et al. Development and large-scale use
of recombinant VP2 vaccine for the prevention of infectious bursal disease of
chickens. Vaccine 2003, 21: 4736–4743
22 Dybing JK, Jackwood DJ. Antigenic and
immunogenic properties of baculovirus-expressed infectious bursal disease viral
proteins. Avian Dis 1998, 42: 80–91
23 Kibenge FS, Qian B, Nagy E, Cleghorn JR,
Wadowska D. Formation of virus-like particles when the polyprotein gene
(segment A) of infectious bursal disease virus is expressed in insect cells.
Can J Vet Res 1999, 63: 49–55
24 Ren ZJ, Black LW. Phage T4 SOC and HOC
display of biologically active, full-length proteins on the viral capsid. Gene
1998, 215: 439–444
25 Ren ZJ, Lewis GK, Wingfield PT, Locke EG,
Steven AC, Black LW. Phage display of intact domains at high copy number: A
system based on SOC, the small outer capsid protein of bacteriophage T4.
Protein Sci 1996, 5: 1833–1843
26 Jiang J, Abu-Shilbayeh L, Rao VB. Display
of a PorA peptide from Neisseria meningitides on the bacteriophage T4
capsid surface. Infect Immun 1997, 65: 4770–4777
27 Lim BL, Cao YC, Yu T, Mo CW. Adaptation
of very virulent infectious bursal disease virus to chicken embryonic
fibroblasts by site-directed mutagenesis of residues 279 and 284 of viral coat
protein VP2. J Virol 1999, 73: 2854–2862
28 Snyder DB, Lana DP, Cho BR, Marquardt WW.
Group and strain-specific neutralization sites of infectious bursal disease
virus defined with monoclonal antibodies. Avian Dis 1988, 32: 527–534
29 Fahey KJ, McWaters P, Brown
MA, Erny K, Murphy VJ, Hewish DR. Virus-neutralizing and passively protective
monoclonal antibodies to infectious bursal disease virus of chickens. Avian
Dis 1991, 35: 365–373
30 Martinez-Torrecuadrada JL, Saubi N,
Pages-Mante A, Caston JR, Espuna E, Casal JI. Structure-dependent efficacy of
infectious bursal disease virus (IBDV) recombinant vaccines. Vaccine 2003, 21:
1952–1960