AVIAN DISEASES 47:549–558, 2003

Protection and Synergism by Recombinant Fowl Pox

Vaccines Expressing Multiple Genes from
Marek’s Disease Virus
Lucy F. Lee, R. L. Witter, S. M. Reddy,A P. Wu,B N. Yanagida,C and S. YoshidaD

USDA, Agricultural Research Service, Avian Disease and Oncology Laboratory,

3606 East Mt. Hope Road, East Lansing, MI 48823
Received 9 October 2002

SUMMARY. Recombinant fowl poxviruses (rFPVs) were constructed to express genes from
serotype 1 Marek’s disease virus (MDV) coding for glycoproteins B, E, I, H, and UL32 (gB1, gE,
gI, gH, and UL32). An additional rFPV was constructed to contain four MDV genes (gB1, gE,
gI, and UL32). These rFPVs were evaluated for their ability to protect maternal antibody–
positive chickens against challenge with highly virulent MDV isolates. The protection induced
by a single rFPV/gB1 (42%) confirmed our previous finding. The protection induced by rFPV/gI
(43%), rFPV/gB1UL32 (46%), rFPV/gB1gEgI (72%), and rFPV/gB1gEgIUL32 (70%)
contributed to additional knowledge on MDV genes involved in protective immunity. In
contrast, the rFPV containing gE, gH, or UL32 did not induce significant protection compared
with turkey herpesvirus (HVT). Levels of protection by rFPV/gB1 and rFPV/gI were comparable
with that of HVT. Only gB1 and gI conferred synergism in rFPV containing these two genes.
Protection by both rFPV/gB1gEgI (72%) and rFPV/gB1gEgIUL32(70%) against Marek’s disease
was significantly enhanced compared with a single gB1 or gI gene (40%). This protective
synergism between gB1 and gI in rFPVs may be the basis for better protection when bivalent
vaccines between serotypes 2 and 3 were used. When rFPV/gB1gIgEUL32 þ HVT were used as
vaccine against Md5 challenge, the protection was significantly enhanced (94%). This synergism
between rFPV/gB1gIgEUL32 and HVT indicates additional genes yet to be discovered in HVT
may be responsible for the enhancement.

RESUMEN. Protección y sinergismo de vacunas recombinantes de viruela aviar expresando

genes múltiples del virus de la enfermedad de Marek.
Se construyeron virus recombinates de viruela aviar expresando los genes del virus de la
enfermedad de Marek del serotipo 1 que codifican las glicoproteı́nas B, E, I, H y UL32 (gB1, gE,
gI, gH y gUL32). Se construyó un virus recombinante de viruela aviar adicional que contenı́a
cuatro genes del virus de la enfermedad de Marek (gB1, gE, gI y gUL32). Se evaluaron estos virus
recombinantes de acuerdo con su capacidad para proteger pollos con anticuerpos maternales
contra el desafı́o con aislamientos altamente virulentos del virus de la enfermedad de Marek. La
protección inducida por un virus recombinante de viruela aviar expresando la gB1 (42%)
confirmó nuestros hallazgos anteriores. La protección inducida por virus recombinantes
expresando la gI (43%), las glicoproteı́nas B1 y UL32 (46%), las glicoproteı́nas B1, E e I (72%) y
las glicoproteı́nas B1, E, I y UL32 (70%) contribuyeron al conocimiento adicional sobre los
genes del virus de la enfermedad de Marek involucrados en la inmunidad protectora. Por el
contrario, virus recombinantes expresando únicamente la glicoproteı́na E, H ó UL32 no
indujeron una protección significativa al ser comparados con el virus herpes de pavo. Los niveles
de protección obtenidos con el virus recombinante expresando la glicoproteı́na B1 y virus
recombinante expresando la proteı́na I fueron comparables con los niveles de protección
obtenidos con el virus herpes de pavo. Unicamente las glicoproteı́nas B1 e I proporcionaron

A
Present address: College of Veterinary Medicine, Texas A&M University, College Station, TX 77843-4467.
B
Present address: Biomune Company, 8906 Rosehill Road, Lenexa, KS 62215.
C
Present address: Nippon Zeon Company, Ltd., Tokyo 105, Japan.
D
Present address: Jichi Medical School, 3311-1 Yakushiji, Minamikawachi, Tochigi, 329-0498 Japan.

549
550 L. F. Lee et al.

sinergismo en virus recombinantes que contenı́an estos dos genes. La protección ofrecida por los
virus recombinantes expresando las glicoproteı́nas B1, I y E (72%) y las glicoproteı́nas B1, E, I y
UL32 (70%) contra la enfermedad de Marek se incrementó significativamente al ser comparada
con la protección ofrecida por los virus recombinantes que contenı́an únicamente el gen B1 o el
gen I (40%). Este sinergismo protectivo entre las glicoproteı́nas B1 e I en los virus recombinantes
de viruela aviar puede ser la base para una mejor protección cuando son empleadas las vacunas
bivalentes que contienen los serotipos 2 y 3. Se observó un incremento significativo en la
protección (94%) contra el desafı́o con la cepa Md5 al emplear el virus recombinante expresando
las glicoproteı́nas B1, I, E y UL32 en asociación con el virus herpes de pavo. El sinergismo entre
estos dos virus indica que genes adicionales en el virus herpes de pavo que aún no han sido
descubiertos pueden ser los responsables del incremento en la protección observada.

Key words: Marek’s disease, recombinant vaccine, fowl poxvirus, glycoprotein genes
Abbreviations: ab þ, ab
¼ antibody positive, antibody negative; CEF ¼ chicken embryo
fibroblast; CTL ¼ cytotoxic T lymphocytes; EHV ¼equine herpesvirus; g ¼ glycoprotein; HSV ¼
herpes simplex virus; HVT ¼ turkey herpesvirus; i.a. ¼ intra-abdominal; MAb ¼ monoclonal
antibody; MD ¼ Marek’s disease; MDV ¼ Marek’s disease virus; PBS ¼ phosphate-buffered
saline; PFU ¼ plaque-forming units; PI ¼ protective index; rFPV ¼ recombinant fowl poxvirus;
UL ¼ unique long; US ¼ unique short; ww ¼ wing web

Marek’s disease (MD) is a highly contagious,
lymphoproliferative disease of chickens and is caused
by a herpesvirus known as Marek’s disease virus
(MDV) (7). Three serotypes of MDV have been
described (2,5): serotype 1 includes pathogenic
isolates of chicken origin and their attenuated
derivatives, serotype 2 includes the naturally apatho-
genic isolates of chicken origin, and serotype 3
includes the naturally apathogenic isolates of turkey
origin known as turkey herpesvirus (HVT) (23).
Vaccines derived from all three serotypes offer
different levels of protection against the disease either
alone or in bivalent and trivalent combinations
(11,19,22) due to increasing virulence of MDV in
the field (20,24).
Recombinant fowl poxvirus (rFPV) containing
MDV glycoprotein (g) B1 gene was the first gene
shown to effectively protect chickens upon challenge
with a virulent virus (9,10). In herpes simplex virus
(HSV), gB is essential for virus infectivity and is
involved in virus penetration into host cells (3,4).
Antibodies to this glycoprotein neutralize the virus
(3,4). We and others have shown that gBs from all
three serotypes of MDV were highly conserved with
predicted amino acid identities of 83% and 82% for
serotype 2 gB (gB2) and serotype 3 gB (gB3), re-
spectively, when compared with serotype 1 gB (gB1)
(12,28). Analysis of the gB complex of MDV with
monoclonal antibody (MAb) 1AN86 showed that it is
composed of a precursor 100-kD molecular weight
protein and two cleavage products, 60-kD and 49-kD
glycoproteins (15). This MAb recognized only the gB
of serotypes 1 and 3 and not that of serotype 2 (15).
From previous studies, we found rFPV expressing
MDV genes serves as a good model to study the
biological function of a single gene effect. We were
interested in the function of other glycoprotein
genes including gE, gI, gH, gL, and UL32. All these
glycoproteins have significant homology to those of
HSV and are important genes involved in immune
responses.
The MDV UL32 homolog of HSV, located in the
unique long (UL) region of the MDV genome, was
identified and characterized as a glycoprotein with
a molecular weight of 82,000 kD and was
designated gp82 (25). We have shown that gp82
belongs to a class of O-linked membrane glyco-
proteins similar to gp300 of equine herpesvirus
(EHV) gene 28, which is homologous to HSV-1
UL32 (25). The gp300 of EHV is unusual in that it
is a glycoprotein modified by O-linked carbohy-
drate. Its function may be related to cell fusion
processes. For this reason, we were interested in the
function of MDV UL32 involved in cell-associated
infection processes.
Glycoproteins gE and gI, located in the unique
short region of the MDV genome (1), were reported
to be essential for virus growth, and deletion of
either gE or gI resulted in the production of virus
progeny that were unable to spread from cell to cell
(14). MDV gE and gI also form a disulfide-linked
heterodimer similar to all alpha-herpesviruses in-
vestigated to date (17).
Glycoproteins H and L of MDV located in the UL
region of the MDV genome have significant
homology to those of other herpesviruses. The mode
 Recombinant FPV vaccine against MD
of interaction of these two genes appeared to be
complex and variable among herpesviruses. Both of
these genes play an important role in virus entry to
host cells and cell-to-cell infection in herpesviruses.
Our recent data on gH and gL suggested that
coexpression of these two genes in the same cells is
required and necessary for subcellular translocation
and for cell surface expression. This observation
supports the previous finding that gL expressed in
FPV-infected chicken embryo fibroblasts (CEFs) was
consistently secreted into the culture medium (29).
We have shown that FPV expressing gB1 (rFPV/
gB1) induced high levels of protection against MD
challenge in maternal antibody–negative (ab
)
and -positive (abþ) chickens (9,10), and the
magnitude of the protection induced by rFPV/gB1
could not be differentiated from that induced by
HVT (9,10). In this report, we describe the use of
rFPVs expressing either a single glycoprotein gene
(gB1 gE, gI, gH, and UL32) of serotype 1 MDV or
containing multiple genes of the same group. These
recombinants, either alone or in combination with
HVT, were used to study their protective efficacy
against MD in chickens.
MATERIALS AND METHODS
Chickens. Chickens were F1 progeny of line 15I5
males and line 71 females. For most experiments, these
chickens were from breeder hens vaccinated with all
three MD viral serotypes as described (22) and were
considered positive for maternal antibodies (abþ). All
breeder chickens were maintained at the Avian Disease
and Oncology Laboratory and were free of antibodies
to avian leukosis virus, reticuloendotheliosis virus, and
various other poultry pathogens.
Viruses. Prototype vaccine viruses included the SB-
1 strain of serotype 2 (13) and the FC126 strain of
HVT (serotype 3) (23). Eight rFPVs (rFPV/gB1,
rFPV/gE, rFPV/gI, rFPV/gH, rFPV/UL32, rFPV/
gB1UL32, rFPV/gB1gEgI, and rFPV/gB1gEgIUL32)
were used as test vaccines. All vaccine stocks were
prepared in this laboratory according to published
procedure (23). The very virulent Md5 strain of MDV
was used as challenge virus (20). A large-plaque-
forming clone (8) from a cell-culture-propagated
vaccine isolate of FPV was used as the parent virus to
construct rFPVs expressing MDV genes.
Construction of rFPV. Cloning of MDV genes
and construction of rFPVs were essentially as reported
earlier (27). Recombinants were constructed that
expressed genes from serotype 1 MDV coding for
glycoproteins B (gB1), E (gE), H (gH), I (gI), and
UL32 (gp82) (12,17,25,26). Nomenclature and pre-
dicted function of these genes is based largely on
551
homology with herpes simplex virus. The early
poxvirus termination sequence, TTTTTNT, was
modified in MDV genes whenever present by site
directed mutagenesis (27). The insertion vector,
pNZ1729R, contained pigeon poxvirus sequences that
flank a cassette containing lacZ and a multiple cloning
site. The MDV genes of interest were cloned into the
multiple cloning site of pNZ1729R.
To effect transfer of the foreign gene from the
transfer vector to FPV, CEF cultures infected for 5 hr
at a multiplicity of infection of 0.2 with a vaccine
isolate of FPV were transfected with 10 lg of transfer
vector. The cassette of MDV genes and lacZ was
inserted into the fowl poxvirus vector via homologous
recombination at a nonessential site. With the
expression of lacZ as an indicator, cloned recombinant
viruses were identified and purified through several
CEF passages. Pure recombinant virus was usually
obtained after three to five serial passages of each clone.
Three blind passages were made to ascertain the purity
of each recombinant. Stocks of purified rFPV were
stored at
70 C, titered on CEF by plaque assay, and
used for vaccination of chickens.
Indirect immunofluorescence assays. Im-
munofluorescence staining was performed as described
(5). Briefly, MDV-infected cells were grown on micro-
titer plates or glass coverslips in 60-mm tissue culture
plates until clear cytopathic effects were visible about 72
hr after infection. The infected cells were washed with
phosphate-buffered saline (PBS) before the cells were
fixed with 60:40 alcohol and acetone mixture for 20
min. After another PBS wash, the cells were incubated
with primary antibodies for 1 hr. After three washes with
PBS (0.1% Tween 20), the secondary antibody
conjugated with fluorescein isothiocyanate was applied
for another hour, and the cells were examined under the
fluorescence microscope (Nikon).
Protection trials. Groups of 17 chickens were
vaccinated at 1 day of age with variable dosage from 101
to 106 plaque-forming units (PFU) of rFPV vaccines by
both the wing web (ww) and intra-abdominal (i.a.)
routes in trial 1. Monovalent vaccines of either SB-1 or
HVT were administered to chickens by the i.a. route at
a dose of 2000 PFU. For rFPV þ MDV vaccines, each
component was administered by a separate inoculation.
The abþchickens were used in all the experiments. After
vaccination, chickens were housed in modified Horsfall–
Bauer isolators.
At 6 days posthatch, groups of vaccinated and
unvaccinated chickens were challenged by i.a. inocula-
tion of 500 PFU of strain Md5. Mortality during the
course of the experiment was recorded, and chickens
were examined for gross MD lesions. At about 56 days
postchallenge (range 48–62 days), the remaining
chickens were killed and examined for gross MD lesions.
The gross lesions induced by Md5 were neural (80%)
and visceral tumors (20%). The visceral tumors were
552 L. F. Lee et al.

Fig. 1. Immunofluorescence for the detection of the expression of MDV genes in recombinant fowl poxvirus
(rFPV). (A) rFPV/gB1 with anti-gB1 monoclonal antibody. (B) rFPV/UL32 with anti-UL32 monoclonal antibody.
(C) rFPV/gE with anti-gE rabbit serum. (D) rFPV/gI with anti-gI rabbit serum. (E) rFPV/gB1gEgIUL32 with anti-
gB1 monoclonal antibody. (F) rFPV/gB1gEgIUL32 with anti-UL32 monoclonal antibody. (G) rFPV/gB1gEgIUL32
with anti-gE rabbit serum. (H) rFPV/gB1gEgIUL32 with anti-gI rabbit serum.

mainly from organs such as gonad, heart, kidney, and
spleen.
The percentage of MD based on the number of
chickens that died or were killed and had gross MD
lesions divided by the number of chickens at risk (total
less birds dying of other causes) 3 100 was determined
for each group. The protection percentage was calculated
for each vaccinated group as the percentage of MD in
unvaccinated, challenged controls less the percentage
of MD in the vaccinated, challenged group divided
by the percentage of MD in the unvaccinated,
challenged controls 3 100. The synergism percentage
was calculated for each bivalent-vaccinated group as
the protection percentage of the bivalent vaccine
minus the protection percentage of the better of the
two constituent monovalent vaccines divided by the
protection percentage of the best monovalent vaccine
3 100. Three trials were conducted; each trial
consisted of three replicates. Statistical analysis was
performed on data from all replicates. Differences in
protection percentage were analyzed by Fisher exact
test (16) and were considered to be significant when
P , 0.05.
RESULTS
Expression of the MDV genes in rFPV-
infected CEF cells. We have previously shown
that MDV gB complex protein consisted of gp100,
gp60, and gp49 and was abundantly expressed in
MDV-infected cells both in the cytoplasm and on the
cell surface by MAb 1AN86 (15). This antibody was
used to detect the expression of gB1 in rFPV by
immunofluorescence assay and found abundant
expression in the rFPV/gB1-infected CEF cells
(Fig. 1A). A second glycoprotein of MDV (gp82)
encoded by UL32 was inserted in FPV, and the rFPV/
UL32 expressed abundant gp82 in infected CEF cells
shown by MAb UL32 (Fig. 1B). Two additional
glycoproteins, gE and gI, were also cloned into rFPVs
and assayed by monospecific rabbit serum against
each protein. We found gE (Fig. 1C) and gI (Fig. 1D)
were highly expressed by these rFPVs in the
cytoplasm of the infected CEFs. rFPV/gH and
rFPV/gL were previously shown to express MDV
gH or gL proteins with monospecific rabbit serum
(26) and, therefore, are not included in this figure.
We also constructed a rFPV containing four
glycoprotein genes (rFPV/gB1gEgIUL32) and found
the expression of all four glycoproteins in infected
cells as detected by their respective antibodies. (Fig.
1E with gB1, 1F with UL32, 1G with gE, 1H with
gI).
Comparative studies on protective effi-
cacy of dose and route of inoculation of
rFPVs. A rFPV containing four MDV glycoprotein
genes (rFPVgB1gEgIUL32) was used to evaluate the
protective efficacy of vaccine dose and route of
inoculation. Table 1 summarizes the data on three
replicate experiments on vaccine dose ranging from
101 to 106 PFU and two inoculation routes, ww or
i.a. We found no statistical difference in protective
efficacy between ww and i.a. inoculation from
 Table 1. Evaluation of dose response and route of inoculation of rFPVgB1gEgIUL32.

VaccineA Replicate 1 Replicate 2 Replicate 3 Summary

Route Dose MD/total %MDB %PIC MD/total %MD %PI MD/total %MD %PI MD/total %MD %PI
6 b a a
i.a. 10 6/17 35 60 6/17 35 63 3/17 18 82 15/51 29 69a
i.a. 105 8/17 47 47b 8/17 47 50b 1/17 6 94a 17/51 33 65a
i.a. 104 7/17 41 53b 6/17 35 63a 6/17 35 65a 19/51 37 61a
i.a. 103 10/17 59 34c 7/17 41 56b 9/17 53 47b 26/51 51 46b
i.a. 102 12/17 71 19c 9/16 56 40b 13/17 76 24c 34/50 68 28c
i.a. 101 13/17 76 14c 12/17 71 24c 15/17 88 12c 40/51 78 17c
ww 106 6/14 43 51b 4/16 25 73a 8/17 47 53b 18/47 38 60a
ww 105 3/17 18 80a 5/16 31 67a 7/17 41 59b 15/50 30 68a
ww 104 6/17 35 60b 5/17 29 69a 11/17 65 35c 22/51 43 54b
ww 103 7/17 41 53b 9/17 53 44b 8/17 47 53b 24/51 47 50b
ww 102 11/17 65 28c 15/17 88 6d 16/17 94 6d 42/51 82 13c
ww 101 12/16 75 15c 14/17 82 13c 15/16 94 6d 41/49 84 11c
SB-1 þ HVT 6/17 35 60b 4/17 24 74a 1/17 6 94a 11/51 22 77a
None 15/17 88 12c 16/17 94 6d 17/17 100 0d 48/51 94 6d
A
rFPV/gB1gEgIUL32 vaccine was inoculated at 1 day of age with dosage between 101 and 106 PFU/chick via wing web (ww) or intra-abdominal (i.a.) route and
Recombinant FPV vaccine against MD

challenged with 500 PFU of Md5 at 5 days after vaccination.

B
%MD based on the number of chickens that died or were killed and had gross MD lesions divided by the number of chickens at risk 3 100.
C
%PI ¼ percentage of protective index. It is calculated from the formula as follows:
%MD in unvaccinated; challenged controls 2 %MD in vaccinated; challenged group 3 100
:
%MD in the unvaccinated; challenged controls
Values followed by the same lowercase superscript letter do not differ (P . 0.05); values with different lowercase superscript letters differ significantly
(P , 0.05).
553
554 L. F. Lee et al.
vaccine doses of 101, 103, 104, 105, and 106 PFU
except at 102 PFU between i.a. and ww (P , 0.05).
However, the protective efficacy of vaccine dose
varies from 11% to 69% depending on the dosage.
A dosage of 103 PFU/chick in three experiments
resulted in a protective index of 50%. Dosage
greater than 103 improved the protective efficacy
from 50% to 69% when the maximal dose of 106
was administered into day-old chicks. Dosage less
than 103 PFU/chick gave slight and variable
protection (11%–28%). On the basis of these data,
we routinely used the dose of 105 for day-old chicks,
and the route of inoculation was i.a.
Comparative efficacy of rFPV expressing
different MDV genes. Chickens with maternal
antibodies to three serotypes of MDV were immu-
nized with rFPVs with inserted MDV genes encoding
five different glycoproteins (gB1, gE, gI, gH, and
UL32). Conventional HVT vaccine was included for
comparative purposes. Table 2 summarizes three
replicate experiments on protective efficacy of MDV
genes in rFPVs against Md5 challenge. As shown,
rFPV/gB1 (43%) or gI (35%) or gE (12%) provided
significant protection when compared with no
vaccine control (0%). There was no significant
difference in protection between rFPV/gB1 and
rFPV/gI, and the level of protection was comparable
with that of standard HVT vaccine (37%). Looking
at gE, however, we noted that 12% is protective but
considerably less than gB1 gI, and HVT. The
remaining two vaccines, rFPV/gH and rFPV/UL32,
did not provide any significant protection.
Protective synergism between gB1 and gI
of rFPVs. rFPVs containing either a single MDV
gene or multiple genes were evaluated for synergism
by comparison of the protection percentage with
that of appropriate monovalent vaccine of HVT
(Table 3). As shown, the protective index was 42%
for rFPV/gB1, confirming the data shown in Table
2. The level of protection by rFPV/gB1 (42%) when
compared with HVT (54%) was not statistically
significant. As for rFPV/UL32, the protective index
was 18%, which is statistically significant when
compared with the no vaccine control (0%). When
rFPV/gB1UL32 was used as vaccine, the protective
index was 46%, which is not significantly different
from that of rFPV/gB1 (42%); therefore, the UL32
gene did not augment gB1 gene. When rFPV con-
taining three MDV genes (rFPV/gB1gEgI) was used
as vaccine, the protective index was 72%, an
increase of 30% in protection, and was significantly
different from that of rFPV/gB1 alone. Similarly,
when the rFPVs contained gB1, gE, gI, and UL32
genes (rFPV/gB1gEgIUL32), protection was also
significantly different from a single rFPV/gB1. In
both recombinant viruses, we observed a synergistic
interaction between two protective genes (gB1
and gI). No synergism was observed with gE or
UL32 genes because they gave minimal protection
against MDV when tested alone or in combination
(Table 3).
In the same experiment, synergism between rFPV/
gB1gEgIUL32 and HVTwas evaluated (Table 3). For
the rFPV/gB1 þ cell-associated HVT vaccine, the
mean protection of 83% was obtained as compared
with 54% for HVT vaccine. For the rFPV/
gB1gEgIUL32 þ HVT vaccine, a mean protection
of 94% was obtained, a 74% increase compared with
HVT alone. The level of protection was comparable
to conventional bivalent SB-1 þ HVT (90%).
DISCUSSION
The gB gene, as well as other glycoprotein genes
including gE, gI, gH, and UL32, has played
a significant role in induction of immunity against
other herpesvirus-induced diseases (3,4). Published
data showed that gB1 of MDV when cloned into
rFPV afforded the level of protection against MD
similar to that of HVT (9). In this study, we cloned
several glycoprotein genes into rFPV either singly or
in multiples. The degree of protection with rFPV/
gB1 confirmed the previous study in abþ chickens
where protection was induced by gB1 (9). The
protection induced by rFPV/gI and rFPV/gB1gE-
gIUL32 in the present study represents new
information on MDV genes involved in protective
immunity. The protective immunity induced by
rFPV/gI correlated well with cytotoxic T lympho-
cytes (CTL) responses against gI protein (6). Both
gB1 and gI gene products induced a significant
degree of protection similar to that of the standard
HVT vaccine. The rFPV containing multiple MDV
genes (gB1gEgIUL32) was more protective than
HVT.
In a number of herpesviruses, both gE and gI are
known to influence virus attachment and penetra-
tion and to play a significant role in immunity. In
our experiment, the protective index of rFPV/gE
(12%), however, was significantly different from
that of unvaccinated control (0%) but considerably
lower than that of gB1 (43%), gI (35%), or HVT
(37%). Similarly, the protective indexes of rFPV/
UL32 (18%) and FPV (10%) were considered
protective in one series of experiments (Table 3) but
not in other experiments (Table 2). This variation
 Table 2. Protective efficacy of MDV genes in rFPVs against Md5 challenge (three experiments).

Replicate 1 Replicate 2 Replicate 3 Summary

VaccineA MD/total %MDB %PIC MD/total %MD %PI MD/total %MD %PI MD/total %MD %PI
a a a
rFPV/gB1 10/16 63 37 10/17 59 41 8/16 50 50 28/49 57 43a
rFPV/gI 12/17 71 29a 11/17 65 35a 10/17 59 41a 33/51 65 35a
rFPV/gE 14/16 88 12b 15/17 88 12b 15/17 88 12b 44/51 88 12b
rFPV/gH 14/16 88 12b 16/17 94 6c 16/17 94 6c 46/50 92 8c
rFPV/UL32 17/17 100 0c 16/17 94 6c 15/16 94 6c 48/50 96 4c
rFPV 16/16 100 0c 13/14 93 7c 16/16 100 0c 45/46 98 2c
HVT 10/17 59 41a 11/16 69 31a 9/15 60 40a 30/48 63 37a
No vaccine 14/14 100 0c 16/16 100 0c 12/12 100 0c 42/42 100 0c
No vaccine, no challenge 0/5 0/5 0/5
A
rFPV containing single MDV genes and wild-type rFPV were inoculated at 1 day of age at the dose of 105 PFU/chick and challenged with 500 PFU of vvMDV
strain Md5 at 5 days after inoculation.
B
%MD based on the number of chickens that died or were killed and had gross MD lesions divided by the number of chickens at risk 3 100.
C
%PI ¼ percentage of protective index. It is calculated from the formula as follows:
Recombinant FPV vaccine against MD

%MD in unvaccinated; challenged controls 2 %MD in vaccinated; challenged group 3 100

:
%MD in the unvaccinated; challenged controls
Values followed by the same lowercase superscript letter do not differ (P . 0.05); values with different lowercase superscript letters differ significantly
(P , 0.05).
555
 556

Table 3. Protective efficacy of MDV genes in rFPVs against Md5 challenge (three experiments).

Replicate 1 Replicate 2 Replicate 3 Summary

A B C
Vaccine MD/total %MD %PI MD/total %MD %PI MD/total %MD %PI MD/total %MD %PI
rFPV/gB1 8/16 50 50c 7/16 44 53c 13/17 76 24d 28/49 57 42c
rFPV/UL32 14/17 82 18d 14/17 82 13d 12/16 75 25d 40/50 80 18d
rFPV/gB1UL32 9/17 53 47c 7/16 44 53c 11/17 65 35d 27/50 54 46c
rFPV/gB1gEgI 7/17 41 59c 2/17 12 87a 5/17 29 71b 14/51 27 72b
rFPV/gB1gEgI UL32 3/17 18 82b 4/17 24 74b 8/17 47 53c 15/51 29 70b
rFPV/gB1UL32 þ HVT 0/16 0 100a 1/16 6 94a 2/17 12 88a 3/49 6 94a
rFPV/gB1 þ HVT 3/15 20 80b 4/16 25 73b 1/17 6 94a 8/48 17 83a
HVT þ SB-1 2/17 12 88a 1/16 6 94a 2/17 12 88a 5/50 10 90a
HVT 7/17 41 59c 8/17 47 50c 8/17 47 53c 23/51 45 54c
FPV 15/17 88 12d 12/16 75 20d 17/17 100 0e 44/50 88 10d
No vaccine 15/15 100 0e 15/16 95 5e 15/15 100 0e 45/46 98 2e
L. F. Lee et al.

No vaccine, no challenge 0/5 0/5 0/5 0/15

A
rFPV containing single MDV genes and wild-type rFPV were inoculated at 1 day of age at the dose of 105 PFU/chick and challenged with 500 PFU of vvMDV
strain Md5 at 5 days after inoculation.
B
%MD based on the number of chickens that died or were killed and had gross MD lesions divided by the number of chickens at risk 3 100.
C
%PI ¼ percentage of protective index. It is calculated from the formula as follows:
%MD in unvaccinated; challenged controls 2 %MD in vaccinated; challenged group 3 100
:
%MD in the unvaccinated; challenged controls
Values followed by the same lowercase superscript letter do not differ (P . 0.05); values with different lowercase superscript letters differ significantly
(P , 0.05).
 Recombinant FPV vaccine against MD
between experiments is not uncommon for there are
many variable factors involved in animal experi-
mentation. One of the major variables may be the
presence of maternal antibody in chickens we used.
Those glycoprotein genes, which did not offer
additional protection against MD in abþ chickens,
may have already elicited a strong antibody re-
sponse. Maternal antibody to MDV has significant
influence on the outcome of the disease and
influences the level of vaccine protection (10,21).
It is possible that either rFPV/gE or rFPV/UL32
could have induced strong immune responses in ab

chickens because there was no pre-existing gE or
UL32 antibody to influence immune responses, as
would be the case of abþ chickens. We have
previously shown that gB from serotype 3 HVT
was marginally protective in abþ chickens but was
significantly protective in ab
chickens (10).
In the case of rFPV/gH, other factors in addition to
maternal antibody may be involved. Our recent data
on gH and gL indicated that coexpression of gH and
gL in the same cells is required and necessary for both
gH and gL subcellular translocation and cell surface
expression (26). Therefore, the interaction of gH and
gL is critical for the biological function of these two
glycoproteins and may be the reason for the inability
of gH protein to reach the cell surface to stimulate
a strong immune response.
Protective synergism has been described between
MD herpesviruses of serotypes 2 and 3 (18). The
synergism between rFPV/gB1 and HVT has also
been described (10). In the present study, we
described another type of synergism that is between
protective genes such as gB1 and gI in rFPV/
gB1gEgIUL32 virus. The level of protection exceeds
70% with gB1 and gI together in a single rFPV. This
protective synergism between gB1 and gI of serotype
1 in the rFPVs is of special interest. Could this
type of synergism between two protective genes be
the basis for the protective synergism observed
previously between SB-1 and HVT? Furthermore,
when HVT was administered in combination with
rFPV/gB1gEgIUL32 as a vaccine, the level of
protection was significantly increased to 94%. This
increase indicated that there are other protective
genes in HVT beyond the gB1 gE, gI, and UL32
and requires further study. It is also possible that
these four genes may function differently depending
on serotypes of virus. Our previous studies showed
that HVT gB gene was not protective in antibody-
positive chickens, unlike gB of serotype 1 MDV
(10). This great difference in protective efficacy of
gB gene between the serotypes was unexpected,
557
indicating some minor differences in gB sequences
may also be responsible for the level of protection.
Mean protection by the rFPV/gB1gEgIUL32 þ
HVT vaccine in all trials was 94% (range 88%–
100%). This level of protection was significantly
greater than that of HVT alone (54%). Cell-free
HVT vaccines, although less effective than cell-
associated products because of their greater suscep-
tibility to inhibition by maternal antibody (21), are
still used in certain area of the world where the
limited availability of liquid nitrogen precludes the
use of cell-associated vaccines. Our data indicate
that the rFPV/gB1gEgIUL32 þ HVT vaccine is
equivalent to bivalent SB-1 þ HVT. Because this
multiple rFPV is in a cell-free form, therefore, the
combination of cell-free HVT and rFPV/gB1gE-
gIUL32 may be as effective as bivalent SB-1 þ HVT
but is an entirely cell-free MD vaccine.
REFERENCES
1. Brunovskis, P., X. Chen, and L. F. Velicer. Analysis
of Marek’s disease virus glycoproteins D, I, and E. In: 4th
International Symposium on Marek’s Disease, 19th
World’s Poultry Congress, vol. 1. World’s Poultry Science
Association, Amsterdam, The Netherlands. pp. 118–122.
1992.
2. Bülow, V. v., and P. M. Biggs. Differentiation
between strains of Marek’s disease virus and turkey
herpesvirus by immunofluorescence assays. Avian Pathol.
4:133–146. 1975.
3. Glorioso, J., C. H. Schroder, G. Kumel, M.
Szczesiul, and M. Levine. Immunogenicity of herpes
simplex virus glycoproteins gC and gB and their role in
protective immunity. J. Virol. 50:805–812. 1984.
4. Highlander, S. L., D. J. Dorney, P. J. Gage, T. C.
Holland, W. Cai, S. Person, M. Levine, and J. C.
Glorioso. Identification of mar mutations in herpes
simplex virus type 1 glycoprotein B which alter antigenic
structure and function in virus penetration. J. Virol.
63:730–738. 1989.
5. Lee, L. F., X. Liu, and R. L. Witter. Monoclonal
antibodies with specificity for three different serotypes of
Marek’s disease viruses in chickens. J. Immunol.
130:1003–1006. 1983.
6. Markowski-Grimsrud, C. J., P. H. O’Connell, and
K. A. Schat. Syngeneic cell-mediated immune responses to
Marek’s disease virus (MDV)-encoded glycoproteins. In:
Current progress on Marek’s disease research. K. A. Schat,
R. M. Morgan, M. S. Parcells, and J. L. Spencer, eds.
American Association of Avian Pathologists, Kennett
Square, PA. pp. 127–130. 2001.
7. Nazerian, K. Marek’s disease: a herpesvirus-induced
malignant lymphoma of the chicken. In: Viral oncology.
G. Klein, ed. Raven Press, New York. pp. 665–682. 1980.
8. Nazerian, K., S. Dhawale, and W. S. Payne.
558 L. F. Lee et al.
Structural proteins of two different plaque-size phenotypes
of fowlpox virus. Avian Dis. 33:458–465. 1989.
9. Nazerian, K., L. F. Lee, N. Yanagida, and R.
Ogawa. Protection against Marek’s disease by a fowlpox
virus recombinant expressing the glycoprotein B of
Marek’s disease virus. J. Virol. 66:1409–1413. 1992.
10. Nazerian, K., R. L. Witter, L. F. Lee, and N.
Yanagida. Protection and synergism by recombinant fowl
pox vaccines expressing genes from Marek’s disease virus.
Avian Dis. 40:368–376. 1996.
11. Rispens, B. H., J. Van Vloten, N. Mastenbroek, H. J.
L. Maas, and K. A. Schat. Control of Marek’s disease in the
Netherlands. I. Isolation of an avirulent Marek’s disease
virus (strain CVI 988) and its use in laboratory vaccination
trials. Avian Dis. 16:108–125. 1972.
12. Ross, L. J. N., M. Sanderson, S. D. Scott, M. M.
Binns, T. Doel, and B. Milne. Nucleotide sequence and
characterization of the Marek’s disease virus homologue of
glycoprotein-B of herpes simplex virus. J. Gen. Virol.
70:1789–1804. 1989.
13. Schat, K. A., and B. W. Calnek. Characterization of
an apparently nononcogenic Marek’s disease virus. J. Natl.
Cancer Inst. 60:1075–1082. 1978.
14. Schumacher, D., B. K. Tischer, S. M. Reddy and
N. Osterrieder. Glycoproteins E and I of Marek’s disease
virus serotype 1 are essential for virus growth in cultured
cells. J. Virol. 75:11307–11318. 2001.
15. Silva, R. F., and L. F. Lee. Monoclonal antibody–
mediated immunoprecipitation of proteins from cells
infected with Marek’s disease virus and turkey herpesvirus.
Virology 136:307–320. 1984.
16. Steel, R. B. D., and J. H. Torrie. Principles and
procedures of statistics. McGraw-Hill Book Co., Inc.,
New York. 1960.
17. Tan, X., P. Brunovskis, and L. F. Velicer. Tran-
scriptional analysis of Marek’s disease virus glycoprotein
D, I, and E genes: gD expression is undetectable in cell
culture. J. Virol. 75:2067–2075. 2001.
18. Witter, R. L. Influence of serotype and virus strain
on synergism between Marek’s disease vaccine viruses.
Avian Pathol. 21:601–614. 1992.
19. Witter, R. L. Safety and comparative efficacy of the
CVI988/Rispens vaccine strain. In: 4th International
Symposium on Marek’s Disease, 19th World’s Poultry
Congress, vol. 1. World’s Poultry Science Association,
Amsterdam, The Netherlands. pp. 315–319. 1992.
20. Witter, R. L. Increasing virulence of Marek’s disease
virus. Avian Dis. 41:149–163. 1997.
21. Witter, R. L., and B. R. Burmester. Differential
effect of maternal antibodies on efficacy of cellular and
cell-free Marek’s disease vaccines. Avian Pathol. 8:145–
156. 1979.
22. Witter, R. L., and L. F. Lee. Polyvalent Marek’s
disease vaccines: safety, efficacy and protective synergism in
chickens with maternal antibodies. Avian Pathol. 13:75–92.
1984.
23. Witter, R. L., K. Nazerian, H. G. Purchase, and G.
H. Burgoyne. Isolation from turkeys of a cell-associated
herpesvirus antigenically related to Marek’s disease virus.
Am. J. Vet. Res. 31:525–538. 1970.
24. Witter, R. L., J. M. Sharma, and A. M. Fadly.
Pathogenicity of variant Marek’s disease virus isolants in
vaccinated and unvaccinated chickens. Avian Dis. 24:210–
232. 1980.
25. Wu, P., L. F. Lee, and W. M. Reed. Serological
characteristics of a membrane glycoprotein gp82 of
Marek’s disease virus. Avian Dis. 41:824–831. 1997.
26. Wu, P., W. M. Reed, and L. F. Lee. Glycoproteins
H and L of Marek’s disease virus form a hetero-oligomer
essential for translocation and cell surface expression.
Arch. Virol. 146:983–992. 2000.
27. Yanagida, N., R. Ogawa, Y. Li, K. Nazerian, and
L. F. Lee. Recombinant fowlpox viruses expressing the
glycoprotein B homolog and pp38 gene of Marek’s disease
virus. J. Virol. 66:1402–1408. 1992.
28. Yoshida, S., L. F. Lee, N. Yanagida, and K. Nazerian.
The glycoprotein B genes of Marek’s disease virus serotypes
2 and 3: identification and expression by recombinant
fowlpox viruses. Virology 200:484–493. 1994.
29. Yoshida, S., L. F. Lee, N. Yanagida, and K.
Nazerian. Identification and characterization of a Marek’s
disease virus gene homologous to glycoprotein L (gL) of
herpes simplex virus type 1. Virology 204:414–419. 1994.
ACKNOWLEDGMENTS
We thank Barry Coulson for excellent technical
assistance. The financial support of Nippon Zeon Co.
of Japan is greatly appreciated.