host range, growth property, and virulence of the smallpox vaccine: vaccinia virus tian tan strain

www.elsevier.com/locate/yviro

Virology 335 (20

Host range, growth property, and virulence of the smallpox vaccine:

Vaccinia virus Tian Tan strain

Qing Fanga,1, Lin Yanga,1, Weijun Zhua, Li Liua, Haibo Wanga, Wenbo Yua, Genfu Xiaoa,

Po Tiena, Linqi Zhanga,b,c, Zhiwei Chena,b,TaModern Virology Research Center and AIDS Center, National Key Laboratory of Virology, College of Life Sciences, Wuhan University,

Hubei 430072, PR ChinabAaron Diamond AIDS Research Center, The Rockefeller University, New York, NY 10016, USA

cAIDS Research Center, Chinese Academy of Medical Sciences, Peking Union Medical College, Beijing, PR China

Received 1 November 2004; returned to author for revision 16 December 2004; accepted 18 February 2005

Available online 19 March 2005

Abstract

Vaccinia Tian Tan (VTT) was used as a vaccine against smallpox in China for millions of people before 1980, yet the biological

characteristics of the virus remain unclear. We have characterized VTT with respect to its host cell range, growth properties in vitro, and

virulence in vivo. We found that 11 of the 12 mammalian cell lines studied are permissive to VTT infection whereas one, CHO-K1, is non-

permissive. Using electron microscopy and sequence analysis, we found that the restriction of VTT replication in CHO-K1 is at a step before

viral maturation probably due to the loss of the V025 gene. Moreover, VTT is significantly less virulent than vaccinia WR but remains

neurovirulent in mice and causes significant body weight loss after intranasal inoculation. Our data demonstrate the need for further

attenuation of VTT to serve either as a safer smallpox vaccine or as a live vaccine vector for other pathogens.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Smallpox vaccine; Vaccinia Tian Tan; Biological characterization

Introduction

Smallpox is a deadly human disease that is caused by the

infection of the variola virus, a member of the orthopoxvirus

genus (Moss, 1991). Because the mortality rate of smallpox

is over 50%, millions of lives were lost to this disease over

the last century before the introduction of an effective

vaccine. Vaccinia virus is another member of the orthopox-

virus genus (Moss, 1991). Since vaccinia virus has mutual

immunogenicity with variola virus and only causes mild

complications in humans, the use of the vaccinia virus as a

smallpox vaccine enabled aggressive immunization pro-

moted by the World Health Organization (WHO), which led

0042-6822/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.virol.2005.02.014

T Corresponding author. Modern Virology Research Center and AIDS

Center, National Key Laboratory of Virology, College of Life Sciences,

Wuhan University, Hubei 430072, PR China or Aaron Diamond AIDS

Research Center, 455 1st Ave., New York, NY 10016, USA.

E-mail address: [email protected] (Z. Chen).1 These authors contribute equally to this work.

to the declaration of complete variola eradication in the

world by 1979 (Breman and Arita, 1980; Fenner, 1982;

Mayr, 1999; Strassburg, 1982; or http://www.who.int/emc/

diseases/smallpox/Smallpoxeradication.html).

The history of smallpox in China can be traced back to

Han Dynasty (206–202 B.C.) (Ma, 1995). The smallpox

was not preventable until Chinese invented variolation in

North Song Dynasty (960–1127 A.D.), which was probably

the earliest practice of smallpox prevention (Ma, 1995). By

the time of Qing Dynasty (1644–1911 A.D.), variolation

was promoted in Kangxi reign to reduce the mortality in the

imperial family (Chang, 1996). Variolation was largely

supplanted by Edward Jenner’s cowpox vaccine discovered

in 1798, but it was not officially proscribed till 1950s

(Strassburg, 1982; Yutu et al., 1988). When the smallpox

vaccination program was interrupted, variolation was

renewed in some remote areas in China which caused

several outbreaks of smallpox in 1960s (Yutu et al., 1988).

During the smallpox eradication campaign in 1960–1970s,

05) 242–251

http:\\www.who.int

Q. Fang et al. / Virology 335 (2005) 242–251 243

the most extensively used smallpox vaccine in China was

the vaccinia Tian Tan (VTT) strain (Jin et al., 1997a,

1997b). VTT exhibits good immunogenicity, moderate

reactogenicity, and relatively mild complications in humans.

Due to these natural properties, VTT was successfully used

for millions of Chinese to prevent smallpox infection

between 1920 and 1980, which subsequently led to the

variola eradication in China before 1980 (http://www.who.

int/emc/diseases/smallpox/Smallpoxeradication.html).

Despite such extensive human use, however, many of the

biological characteristics of the vaccine strain VTT remain

unknown.

Recently, vaccinia virus attracted more attention for its

re-use as a vaccine against variola virus, due to the rise in

bio-terrorism threats around the world (Mahalingam et al.,

2004; Weltzin et al., 2003). This is a significant security

concern because about 40% of people in the United States,

for example, are younger than 30 years old and are not

vaccinated. Moreover, vaccinia virus can also be engineered

to deliver foreign antigens to serve as a live vaccine vector

for other pathogens (Moss, 1996; Paoletti, 1996). As an

AIDS vaccine candidate, an attenuated modified vaccinia

Ankara (MVA) expressing HIV-1 components has entered

clinical trials (McMichael and Hanke, 2002). Considering

the potential future use of VTT in humans, we attempt to

determine the biological properties of the virus.

Based on sequence analysis, VTT displays genetic

features distinct from other vaccinia viruses, including the

Western Reserve (WR) strain (Hou et al., 1985; Jin et al.,

1997a; Tsao et al., 1986). These data infer the plausibly

distinct biological properties of VTT. In recent years, VTT

was explored as a live vector for other pathogens such as

rabies virus and hepatitis B virus (Guo et al., 2001; Zhu et

al., 1996). These studies, however, are still preliminary. Up

to now, few data are available with respect to the baseline

biological properties of VTT, which is essential for assess-

ments of the safety and immunogenicity of a vaccinia

Table 1

Replication and spread of VTT in various cell types infected with 0.05 moi of th

Cell line ATCC code Species Organ

HeLa CCL-2 Human Cervix

MRC-5 CCL-171 Human Lung

293T CRL-11268 Human Kidney

WISH CCL-25 Human Amnion

Vero CCL-81 African green Monkey Kidney

COS-7 CRL-1651 African green Monkey Kidney

PK(15) CCL-33 Pig Kidney

MDCK CCL-34 Dog Kidney

RK13 CCL-37 Rabbit Kidney

C6 CCL-107 Rat Glial tum

CHO-K1 CCL-61 Chinese Hamster Ovary

BHK-21 CCL-10 Syrian Hamster Kidney

CEF Primary Chicken embryo Assorted

a Virus spread as visualized by immunostaining after 48 h. �, no stained cells; +

stained cells (Carroll and Moss, 1997).b Virus replication (fold increase in virus titer) determined by dividing the virus y

into permissive (P, N25-fold increase) and nonpermissive (NP, b1-fold increase) c

expression vector (Meyer et al., 1991; Moss, 1996; Paoletti,

1996). In this study, the smallpox vaccine VTT is studied in

relation to its host range, growth properties, morphogenesis,

and virulence. Our findings will serve as a baseline for

further modification of the virus, as either a safer smallpox

vaccine or a live vaccine vector for other pathogens.

Results

Host range of VTT strain

Twelve mammalian cell lines of various host origins and

primary chicken embryo fibroblasts (CEF) were tested in

this study (Table 1). Using immunochemical staining, we

found that each cell type tested can be infected by VTT.

This is indicated by the expression of specific viral antigen

stained with polyclonal anti-vaccinia serum in infected cells

(Fig. 1). Except for CHO-K1, all cell types show positive

foci, which consist of a cluster of stained cells. However,

the number of positive cells and the extent of cytopathic

effects is different in each cell type. We scored the observed

foci by counting the positive cells post-infection (pi)

according to a method previously described (Table 1)

(Carroll and Moss, 1997). CHO-K1 is deemed non-

permissive since only individually positive cells were

observed 24 h pi and the number of infected cells tended

to decrease 48 h pi (Fig. 1). In contrast, clear cytolytic

effects and large plaques were found 48 h pi in eight

permissive cell types including HeLa, MRC-5, Vero, COS-

7, MDCK, RK13, C6, BHK-21, and primary CEF cells.

Importantly, the growing size of each positive foci or

plaque overtime in these cell types also suggests that the

cell–cell contact is essential for viral transmission or spread.

In particular, COS-7 cells show the most rapid cell killing

(Fig. 1B). Large plaques were frequently seen in COS-7 by

24 h pi. Vero cells seem to give the highest yield of

e virus

Morphology Virus spreada Virus replicationb

Epithelial +++ 63.6 P

Fibroblast +++ 187.5 P





Epithelial ++ 100.8 P

Epithelial +++ 145 P


or Fibroblast +++ 156.7 P

Epithelial �/+ 0.12 NP



, foci of 1–4 stained cells; ++, foci of 5–25 stained cells; +++, foci of N25

ield at 72 h by the practical input titer. Cell lines were therefore categorized

ells (Carroll and Moss, 1997).

http:\\www.who.int

Fig. 1. Infection and cell-to-cell spread of VTT in various cell types including one primary chicken fibroblast (1A) and 12 cell lines (1B). The cells were

infected with 100 PFU of VTT per well in a 24-well-plate, and then fixed and immunostained with anti-vaccinia antibody at 24 and 48 h pi. All these photos

were taken at the 100� magnification.

Q. Fang et al. / Virology 335 (2005) 242–251244


progeny virions reflected by the fold of viral replication

reaching 562.5-fold by 72 h pi (Table 1). Such titer is more

than doubled than that of CEF (128.2-fold) which was used

for VTT manufacture before 1980 (Table 1). The data

suggest the potential use of Vero cell for manufacturing

VTT virus in the future. For 293T and WISH cells, typical

cytolytic plaques usually began to become predominant

after 72 h pi, suggesting relatively slower cell killing in

these cell types. In PK(15) cells, no severe cytolytic effects

were found although small clusters of infected cells were

identified, which may suggest a different cytopathic

mechanism in this cell type. Moreover, the much fewer

positive cells found in each foci at 48 h pi also indicated

that the viral spread via cell-to-cell contact is less efficient

in PK(15) cells.

Replication of VTT under low multiplicity of infection

To further understand the replication kinetics of vaccinia

VTT in various cell types tested, we infected the cells at a

low multiplicity of infection (moi). Using a 0.05 moi, we

determined the growth of viral replication for multiple

rounds (Carroll and Moss, 1997). We found that VTT

replicated with similar kinetics in four human cell lines

including HeLa, MRC-5, 293T, and WISH (Fig. 2A). Since

these cells were derived from human cervix epithelial, lung

Fig. 2. Replication of VTT in cell types tested. Cells were infected with 0.05 moi (

viral absorption, and 24, 48, and 72 h pi. Moreover, cells were infected with 5 moi

viral absorption, and 12 and 24 h pi. In these experiments, both culture medium

fibroblast, kidney epithelial, and amnion epithelial, respec-

tively, VTT can replicate efficiently in cell types with

diverse tissue origin. In these cells, the viral replication

quickly reached peak level around 24–48 h pi (Fig. 2A).

Interestingly, the cytolytic effects seem to occur quickly and

more severe in MRC-5 cells, a fibroblast lineage (Fig. 1B).

This finding is consistent with the data generated in nine

non-human cell types including Vero, COS-7, CHO-K1,

BHK-21, MDCK, RK13, PK(15), C6, and CEF. VTT tends

to replicate better and cause more severe cytopathic effects

in cells of fibroblast lineage such as COS-7 and C6

regardless of the host origin (Fig. 1B and Table 1). The

similar growth kinetics of VTT found in several cell types

makes it possible to use one of them to replace CEF cells for

vaccine production (Fig. 2B). Consistent with host range

study, TT does not seem to replicate in CHO-K1 cells with

the 0.05 moi for infection.

Replication of vaccinia VTT under high moi

A high moi usually results in synchronized infection in

viral target cells. Since a synchronized infection predom-

inantly reflects intracellular virus replication, we attempt to

predict the number of infectious viral particles produced

daily in each cell. In this experiment, the cells were infected

with a high moi of 5. After infection, the cells were

A and B). Viral replication was determined by measuring the viral titer after

(C and D). Viral replication was determined by measuring the viral titer after

and cells were harvested for testing.


processed to measure the yield of viral production within 24

h. Under these conditions, we found that significant yields

of VTT were obtained from the four human cell lines HeLa,

MRC-5, 293T, and WISH (Fig. 2C) as well as from many

non-human cells including Vero, BHK-21, MDCK, RK13,

COS-7, PK(15), C6, and CEF (Fig. 2D). As expected, we

did not see viral production in the non-permissive CHO-K1

cells (Fig. 2D). The yield of viral particles in each cell type

has been summarized in Table 2. Considering the data of

day 0 largely reflect the quantity of input virus, we have

subtracted the values of day 0 for the calculation of viral

yield for later time points. Generally, most cell types

produced more than one plaque forming unit (PFU) of

infectious VTT per cell except for CHO-K1 at 24 h pi (Table

2). Vero cells, on average, produced the highest amount of

virus with an approximated ratio of about 29 virions per cell

in 24 h pi (Table 2). Since Vero is currently used to produce

human rabies and Japanese B encephalitis vaccines in China

(Dong et al., 2002; Yan et al., 2002), our findings have

demonstrated the potential use of this cell line to produce

VTT-based smallpox vaccine if needed. Of interest, all cell

types tested, including the non-permissive CHO-K1 cell,

exhibited severe cytopathic effects 24 h after VTT infection

(Table 2). This finding indicated that VTT-infected cells

may undergo cell death rapidly. Moreover, this finding also

suggested that the blockade of viral replication in CHO-K1

cells may not be at the entry step because the observed cell

death might be a result of viral infection.

Virus morphogenesis in permissive and non-permissive cells

To understand the blockade of viral replication in CHO-

K1 cells, vaccinia VTT-infected cells were further subjected

to electron microscope (EM) analysis. By analyzing the

morphology of newly produced viral particles, we expected

Table 2

Cytopathic effects and yield of VTT in various cell types infected with 5

moi of the virus

Cell line CPEa PFU/cell, 24 h

12 h 24 h

HeLa ++++ ++++ 3.75

MRC-5 ++++ ++++ 17.18

293T ++++ ++++ 9.5

WISH ++++ ++++ 5.23

Vero ++++ ++++ 29.5

COS-7 ++++ ++++ 14.45

PK(15) ++ +++/++++ 12.54

MDCK ++++ ++++ 1.13

RK13 ++++ ++++ 5.31

C6 ++++ ++++ 7.1

CHO-K1 ++++ ++++ �0.11

BHK-21 ++ +++ 10.14

CEF ++ +++ 5.36

a CPE was categorized by the following criteria: no difference from

control, �; b25% CPE, +; 25–50% CPE, ++; N50–75% CPE, +++; N75–

100% CPE, or high level cell detachment, ++++ (Carroll and Moss, 1997).

to understand the life cycle of VTT in CHO-K1 cells. For

this reason, CHO-K1 and some permissive cells were

analyzed by EM 16 h pi. As shown in Fig. 3A, no

intracellular mature viruses (IMV) were found in infected

CHO-K1 cells although some immature virions were

frequently found in the cytoplasm of the infected cells. As

controls, the majority of the VTT particles showed the

characteristic brick-shape of IMV in the cytoplasm of

permissive cells including RK13, Vero, and CEF (Figs.

3B–D). Furthermore, viral particles at different stages of

viral maturation, like spherical immature virions, were

found in permissive cells as well (Figs. 3B–D). Therefore,

together with the finding of VTT-specific protein expression

in individual CHO-K1 cells (Fig. 1B), our data indicated

that the aborted viral replication in this cell line is likely due

to a blockade at the stage of viral assembly before

maturation. Considering that cowpox V025 gene determines

the virus growth in CHO cells (Ramsey-Ewing and Moss,

1995, 1998; Spehner et al., 1988), we have amplified the

gene homolog from our VTT. We found that our sequence is

99.9% identical to the V025 gene sequence of VTT

deposited in the GenBank (accession number AF095689).

Moreover, we found that VTT sequences contain many

similar mutations or deletions to WR strain when these

sequences were compared with the cowpox sequence

(AF482758) (data not shown). In particular, VTT and WR

share a nucleotide deletion in the V025 gene which resulted

in a reading frame shift and introduced a premature stop

codon at amino acid position 72. Therefore, like the

situation in WR, the loss of the functional V025 gene

probably contributes to the replication failure of VTT in

CHO-K1 cells.

Virulence of vaccinia VTT strain in BALB/c mice

Despite of the extensive use of vaccinia VTT in humans

in the last century, the virulence of the virus has not been

carefully determined. Here, we characterized the virulence

of the VTT by inoculating the virus into BALB/c mice. The

viral inoculums were based on the tissue infectious dose

(PFU) as determined in Vero cells. We first determine the

lethal dose (LD50 and LD90) of VTT by inoculating seven

groups of mice via the intracranial route. Each of the seven

groups, six mice per group, received one of 5-fold diluted

VTT. We found that the LD50 and LD90 of VTT were 3.1 �103 PFU and 1.4 � 104 PFU, respectively. The LD50 of

VTT, therefore, is about 300-fold less virulent in compar-

ison to that of vaccinia WR strain published previously,

which is about 10 PFU (Buller et al., 1985; Kutinova et al.,

1995). For a direct comparison, we also tested the LD50 and

LD90 of WR using the same method. The infectious dose of

WR was determined in Vero cells in parallel with VTT.

Most mice died after received WR except for some mice in

the group received 1 PFU of WR. We were able to obtain

the LD90 of WR which is 2.4 PFU. This LD90 value

indicates that WR is over 5000-fold more virulent than VTT

Fig. 4. Replication kinetics of VTT in mouse brain. BALB/c mice (3–4

weeks old) were inoculated intracranially with 10 Al containing 1.1 � 104

PFU of VTT. Mice were sacrificed daily, the intracranial contents were

homogenized in PBS, and the titer of infectious virus was determined by a

plaque assay on Vero cell monolayers. Each point on the curve represents

the average value of two mouse brains. The error bar indicates the

experimental variation.

Fig. 3. Electron microscopy analysis of four cell types infected with VTT. These cell types include CHO-K1 (A), RK13 (B), Vero (C), and CEF (D). The

analysis was performed at 16 h pi. The individual letters indicate different observed structures including: i, typical immature virion; m, brick-shaped mature

virion; n, nucleoid. The black bars shown at the left corner of each photo represent the equivalent physical sizes: 1 AM in A and B and 0.5 AM in C and D.


strain (LD90 = 1.4 � 104 PFU). To further determine the

replication kinetics of VTT in mouse brain, we inoculated 3

LD50 of VTT into ten mice via the same intracranial route.

Two animals were sacrifice daily for the titration of viral

replication in their brains. As shown in Fig. 4, VTT strain

replicated in the mouse brain with a peak viral yield of

1.71 � 104 PFU per brain on day 3 pi. Virus titer tended to

decline on day 4 pi, yet the remaining animals died on day

5 pi. These data demonstrated that VTT is likely a

neurovirulent virus.

The virulence of VTT was also tested by measuring the

body weight change of infected mice after they were

intranasally injected with viruses (Fig. 5). In contrast to

mock mice which received PBS, the infected mice showed

clear dose-dependent weight loss starting from the third day

after viral inoculation. There is a clear positive correlation

between the dose of virus inoculation and body weight loss

of the animals. A similar experiment was done by using the

WR strain. We inoculated two groups of BALB/c mice with

103 and 104 PFU of WR, respectively, via the intranasal

route. We found that 104 PFU of WR caused over 25%

weight loss in BALB/c mice, which was very similar to the

Fig. 5. Comparison of the virulence of VTT (A) and WR (B) in mice by the assessment of body weight loss of infected animals. Groups of five BALB/c mice

were infected intranasally with indicated doses of VTT or PBS, and WR or PBS on day 0 (arrow). The overtime body weight loss is represented by the mean

values of each group of animals post-viral infection. The error bar indicates the experimental variation.


mice that received 106 PFU of VTT. Similarly, 103 PFU of

WR resulted in comparable weight loss with 105 PFU of

VTT in mice. This dose-dependent pattern is comparable to

that of WR in mouse model published previously (Brandt

and Jacobs, 2001). These results demonstrated that WR is

about 100 more virulent than VTTwhen the virus was given

via the intranasal route. Taken together, our data suggested

that VTT is much less virulent than vaccinia WR strain

although it remains to be neurovirulent.

Discussion

In this study, we have characterized the biological

properties of the smallpox vaccine VTT strain by testing

its host cell range, growth properties, and virulence. We

found that VTT has a broad host cell range as the virus

infected all 13 different cell types of divergent host origins.

Various cell types, however, displayed different permissi-

bility and efficiency in supporting VTT replication. More-

over, we found that VTT could productively replicate in the

mouse brain and caused animal death after intracranial

infection. VTT also caused loss of mouse body weight after

intranasal infection. These findings indicated that VTT is a

neurovirulent virus and that it will need further attenuation

for future human use. Our data also established baseline

biological profiles for future VTT studies.

VTT has a broad host cell range. The biological

characteristics of multiple vaccinia strains have been

previously characterized such as strains of parental Ankara,

WR, Copenhagen, Wyeth, and Lister (Carroll and Moss,

1997; Kutinova et al., 1995; Meyer et al., 1991; Tartaglia et

al., 1992). One common property shared by these viruses is

their broad host cell range. Before the present study, the host

range of VTT was unclear despite the fact that the virus was

given to millions of Chinese as a smallpox vaccine before

1980, which probably made VTT the most extensively used

human smallpox vaccine in the world. Since VTT vacci-

nation was quite effective against smallpox, the side effects

related to the biological properties of VTT were probably

ignored and not reported. Here, we found that VTT has

broad cell tropism to enter all 13 cell types tested despite the

virus was passaged on calf skin or CEFs for many

generations during the production of smallpox vaccine

(The Committee for Biological Product Standardization in

China, 1979). Thus, like many vaccinia viruses, VTT

probably uses similar common receptor or mechanism to

enter the broad range of target cells.

Despite of the broad host cell range, the replication

kinetics of VTT varies among different cell types. VTT

productively replicates in 12 of 13 mammalian or avian cells

except in non-permissive CHO-K1 cells. The infected cells,

however, displayed different efficiency in supporting the

VTT replication. For example, Vero cell produced the

highest yield of VTT whereas CHO-K1 cells only had

abortive infection. The viral yield depends not only on the

capability of each cell type to produce the progeny virions

but also on the efficiency of viral spread among cells. The

reason that Vero cells made the highest yield of virions is

very likely due to the fact that each Vero cell could produce

more virions than other cell types and VTT could spread

among Vero cells quickly (Tables 1 and 2). Since Vero cells

are currently used for human rabies, hepatitis A, and

Poliovirus vaccines manufacture in China and other

countries (Dong et al., 2002; Sampath et al., 2005; Sheets,

2000; Sun et al., 2004), our data suggested that this cell type

may be used to replace CEF to produce smallpox vaccine

VTT for use against bio-terrorism. This replacement will

eliminate the use of special pathogen free eggs and hence

help minimize the chicken-derived pathogens and reduce

the vaccine production cost substantially (Sheets, 2000).

However, we should also be cautious on other issues related

to the use of Vero cells for vaccine production. For example,

the specific Vero cell line recommended by WHO should be

used to avoid the potential contamination with SV40 or

some unexpected cellular proteins which may cause auto-


immune reactions in humans (Monath et al., 2004; Sheets,

2000). Using EM analysis, we found that the abortive

infection of VTT in CHO-K1 cells is likely due to a

blockade at viral maturation stage. Since this finding is

consistent to the previous report on vaccinia Ankara and

WR, there might be a common mechanism shared by these

vaccinia strains with VTT (Carroll and Moss, 1997;

Ramsey-Ewing and Moss, 1995; Spehner et al., 2000).

There are several possible factors related to the abortive

viral replication in CHO-K1 cells. First, some host cell

proteins, which are essential for viral maturation, might be

missing in CHO-K1 cells. Second, due to the loss of V025

gene by our sequence analysis, the shutoff of some vaccinia

intermediate and late protein synthesis in CHO-K1 cells

might lead to the aborted viral maturation (Ramsey-Ewing

and Moss, 1995, 1998). Third, some viral genes D13L,

A14L and A17L might also play a role in the morpho-

genesis of vaccinia virus (Rodriguez et al., 1997; Smith et

al., 2002; Wolffe et al., 1996). The exact mechanism,

however, remains to be further studied for VTT.

VTT is a neurovirulent vaccinia stain. To make a

vaccinia-based human vaccine, the safety profile of the

virus is one of the most essential elements (Meyer et al.,

1991; Tartaglia et al., 1992). Based on previous reports, the

LD50 of other vaccinia strains, such as Copenhagen and

Wyeth, was about 104 PFU by intracranial route in mice

(Kutinova et al., 1995; Tartaglia et al., 1992). Similarly, we

found that the LD50 of VTT is 3.1 � 103 PFU. However, if

we compare with the neurovirulent WR strain which has a

LD90 of 2.4 PFU, VTT is about 5000-fold less virulent with

a LD90 of 1.4 � 104 PFU. Nevertheless, VTT was able to

replicate in mouse brain with growth kinetics somewhat

similar to that of vaccinia Wyeth (Lee et al., 1992).

Moreover, VTT infection resulted in over 25% body weight

loss in BALB/c mice after 106 PFU of VTTwas intranasally

inoculated. In comparison, only 104 PFU of WR (100-fold

less than VTT) was needed to cause similar body weight

loss in vivo. Taking together these findings, we conclude

that although weaker than WR strain VTT remains to be a

neurovirulent vaccinia strain. If VTT is produced as

vaccines for future human use, further attenuation of the

virus should be considered. This is a serious concern

because VTT may be lethal to humans especially to

immune-compromised patients who often fail to control

systemic vaccinia replication.

VTT may be engineered to become a safe live-attenuated

vaccine vector. Currently, several highly attenuated vaccinia

strains have been developed as live viral vectors to make

vaccines for other pathogens including human immunode-

ficiency virus (Moss, 1996). In this study, we have

established the biological profile of VTT which may serve

as a baseline for future modification of the virus. Previously

highly-attenuated modified vaccinia Ankara (MVA) strain

was successfully obtained by serial passages of its parental

virus in CEF cells (Mayr et al., 1975, 1978). Similar

strategies have also been used to attenuate VTT through

serial passages of the virus on chick embryos or on human

diploid cells (Yang et al., 1979; Zhao et al., 1981). Given no

biological characterization was reported on the passaged

VTT virus, we do not know to what extent those viruses

have been attenuated. After some host range and virulence

genes were identified from the genome of vaccinia virus, the

introduction of selected deletions in viral genome has

become another approach to attenuate vaccinia virus

(Mackett et al., 1985; Ramsey-Ewing and Moss, 1995;

Tartaglia et al., 1992). This approach became possible for

VTT after the entire genomic structure of the virus was

documented (Jin et al., 1997a). Some preliminary results

have indicated that removing thymidine kinase gene or C–K

genomic may reduce the virulence of VTT (Li et al., 1999;

Zhu et al., 1996). These works showed some promising

progress yet the detailed biological characteristics on the

attenuated virus remain unknown. Considering the extensive

and successful human use of VTT against smallpox in the

history, we believe that understanding the biological proper-

ties of VTT is one of the fundamental steps for under-

standing the basis of the virus and for developing a safe

human vaccine (Katz and Moss, 1997; Smith et al., 2002;

Spehner et al., 2000; Weltzin et al., 2003).

Materials and methods

Cells and virus

Twelve cell lines (Table 1) were grown under conditions

suggested by the American Type Culture Collection (ATCC,

Rockville, MD). The primary chick embryo fibroblasts

(CEFs) were prepared from 10-day-old chicken embryos.

The first generation of passaged CEF cells was used for

virus infections. VTT 761 strain was obtained from the

Institute of Virology, the Chinese Academy of Preventive

Medicine. Vaccinia WR strain was obtained from ATCC

(ATCC VR-1354, Rockville, MD). VTT and WR were

propagated in Vero cells. Viral stocks purified through

centrifugation over 36% sucrose cushion were used for in

vitro experiments while those used for in vivo testing were

further purified through sucrose density gradient centrifu-

gation (Joklik, 1962). Viral titers of VTT and WR were

determined on Vero monolayers by plaque assay using

crystal violet staining.

Immunostaining of virus plaques

Polyclonal anti-TT serum was produced in rabbits. This

antibody was used to detect virus-infected plaques or foci.

Two rabbits were immunized intradermally with 107 PFU of

VTT. Four weeks after, these rabbits were boosted twice

with 107 PFU of the purified TT stocks at a 3-week interval.

The animal was bled 4 weeks after the final inoculation. The

titer of anti-TT antibody in serum was determined in Vero

cells by a plaque reduction neutralization assay.


For the test of cell-to-cell spread, plaques or foci of VTT

were immunostained by the rabbit anti-VTT serum using a

method previously described (Carroll and Moss, 1997). The

target cells were grown in 350-mm culture dishes to 90–

100% confluence and then infected with TT stocks at a

multiplicity of infection (moi) of 0.01. After absorption for

90 min, cells were washed three times with culture medium

and then incubated at 37 8C for another 24 h or 48 h before

immunostaining. After the incubation, cells were fixed with

a cold 1:1 solution of methanol/acetone for 2 min and then

incubated for 1 h at room temperature with antiserum diluted

(1:80) with phosphate buffered saline (PBS) containing 2%

fetal bovine serum (FBS). Cells were washed and then

incubated for 1 h at room temperature with Protein A

conjugated to horseradish peroxidase (Zhongshan Biotech;

1: 1500 dilution). After wash, the color was developed by

incubation up to 30 min with substrate solution consisting of

10 Al of 30% H2O2 and 0.2 ml of an ethanol solution

saturated with dianishidine (Sigma) in 10 ml of PBS.

Virus replication in vitro

Cells were grown in Falcon 24-well culture dishes to form

monolayers. The number of cells was calculated before the

virus inoculation so as to determine the input amount of virus.

VTT was then applied onto the cell monolayers at an moi of

0.05 in 100 Al of medium containing 3% FBS. After 90-min

incubation at 37 8C, cells were washed three times with

medium containing 3% FBS and replenished with fresh

culture medium. After wash, cells and supernatant were

harvested at 0, 24, 48, and 72 h pi. For one step growth

curves, cells were infected with an moi of 5. The samples

were harvested at 0, 12, and 24 h pi. After freeze–thawing

thrice, the viral titers were determined in duplicates on Vero

cells.

Electron microscopic analysis

Confluent cell monolayers were infected with 5 moi of

VTT. The virus was allowed to adsorb for 90 min at 37 8C.The cells were then washed with medium three times and

incubated at 37 8C for an additional 16 h. Cells were then

trypsinized off the culture plates and washed with PBS

twice through centrifugation at 1200 rpm for 10 min. Then

the cell pellets were fixed in 2.5% glutaraldehyde and

processed for transmission electron microscopy using a

routine technique described previously (Wolffe et al., 1996).

Virulence in BALB/c mice

Groups of six 3- to 4-week-old BALB/c mice (equal

gender) were inoculated intracranially with one of 5-fold

diluted VTT or WR virus in 10 Al of sterile PBS. The LD50

was determined by the number of mice that succumbed

between 1 and 14 days pi by calculating the 50% end point

using the Reed-Muench method (Reed and Muench, 1938).

With respect to VTT replication in mouse brain, 3- to 4-week-

old mice were inoculated intracranially with 1.1 � 104 PFU

(3LD50 ) of purified virus in 10 Al of PBS. Two of tested mice

were sacrificed daily, and their brains were homogenized in

PBS before they were subjected to viral titration.

The virulence of VTT in mice was also studied by their

weight loss after viral inoculation using a method previously

described (Zhang et al., 2000). Groups of five 5-week-old

BALB/c mice were inoculated intranasally with 0, 104, 105,

and 106 PFU of VTT in 30 Al of PBS. For comparison, 5-

week-old BALB/c mice were inoculated with 103 or 104

PFU of WR under the same conditions, respectively. The

body weight change of each mouse was monitored daily

and the mean value was used in the graph of the weight

loss curve.

Acknowledgments

The authors would like to thank Xiaofeng Xie, Lei Ba,

and Christopher Yi for assistance. This study was supported

by the Special Fund of Wuhan University for Modern

Virology Research Center.

References

Brandt, T.A., Jacobs, B.L., 2001. Both carboxy- and amino-terminal

domains of the vaccinia virus interferon resistance gene, E3L, are

required for pathogenesis in a mouse model. J. Virol. 75 (2), 850–856.

Breman, J.G., Arita, I., 1980. The confirmation and maintenance of

smallpox eradication. N. Engl. J. Med. 303 (22), 1263–1273.

Buller, R.M., Smith, G.L., Cremer, K., Notkins, A.L., Moss, B., 1985.

Decreased virulence of recombinant vaccinia virus expression vectors is

associated with a thymidine kinase-negative phenotype. Nature 317

(6040), 813–815.

Carroll, M.W., Moss, B., 1997. Host range and cytopathogenicity of the

highly attenuated MVA strain of vaccinia virus: propagation and

generation of recombinant viruses in a nonhuman mammalian cell line.

Virology 238 (2), 198–211.

Chang, C., 1996. Time frame and reasons of Kangxi Emperor adopted

variolation. Zhonghua Yishi Zazhi 26 (1), 30–32.

Dong, G.M., Zheng, H.F., Liu, Z.S., Qu, X.S., Liu, J.H., Wu, X.H., 2002.

The clinic reaction and evaluation of Chinese purifying Vero cell rabies

vaccine. Chin. J. Biol. 5 (5), 312–314.

Fenner, F., 1982. A successful eradication campaign. Global eradication of

smallpox. Rev. Infect Dis. 4 (5), 916–930.

Guo, F., Lu, R.J., Lou, Y.M., Sun, Z.H., Ruan, L., 2001. Construction of

expression vectors of non-replication recombinant vaccinia virus

Tiantan strain with CK region deletion and studying of biological

properties of the recombinant viruses. Chin. J. Virol. 17 (1), 24–28.

Hou, Y.T., Yang, X.K., Hu, Y.W., 1985. Variation in the HindIII restriction

fragments of DNA from the Chinese Tian Tan strain of vaccinia virus.

J. Gen. Virol. 66, 1819–1823.

Jin, Q., Chen, L.H., Chen, S.X., Huang, J., Feng, Z.H., Yuan, J.S., Jin, D.Y.,

Bai, H.D., Hou, Y.D., 1997a. Characterization of the complete

genomic sequence of the vaccinia virus Tian Tan strain. Sci. China

27 (6), 562–567.

Jin, Q., Chen, N.H., Yao, E.M., Huang, J., Hou, Y.D., 1997b. Analysis of

open reading frames located in HindIII-C and -B fragments of vaccinia

virus strain TianTan. Chin. J. Virol. 13 (1), 1–5.


Joklik, W.K., 1962. The purification of four strains of poxvirus. Virology

18, 9–18.

Katz, E., Moss, B., 1997. Immunogenicity of recombinant vaccinia

viruses that display the HIV type 1 envelope glycoprotein on the

surface of infectious virions. AIDS Res. Hum. Retroviruses 13 (17),

1497–1500.

Kutinova, L., Ludvikova, V., Simonova, V., Otavova, M., Krystofova, J.,

Hainz, P., Press, M., Kunke, D., Vonka, V., 1995. Search for optimal

parent for recombinant vaccinia virus vaccines. Study of three vaccinia

virus vaccine strains and several virus lines derived from them. Vaccine

13 (5), 487–493.

Lee, M.S., Roos, J.M., McGuigan, L.C., Smith, K.A., Cormier, N., Cohen,

L.K., Roberts, B.E., Payne, L.G., 1992. Molecular attenuation of

vaccinia virus: mutant generation and animal characterization. J. Virol.

66 (5), 2617–2630.

Li, P., Hu, Q.L., Sun, Z.H., Wang, J.L., Zhu, J.H., Ruan, L., 1999.

Construction and characterization of non-replicating recombinant

vaccinia virus expressing rabies glycoprotein. Chin. J. Exp. Clin. Virol.

13, 170–174.

Ma, B., 1995. Variolation, pioneer of modern immunology. Zhonghua

YiShi ZaZhi 25 (3), 139–144.

Mackett, M., Yilma, T., Rose, J.K., Moss, B., 1985. Vaccinia virus

recombinants: expression of VSV genes and protective immunization of

mice and cattle. Science 227 (4685), 433–435.

Mahalingam, S., Damon, I.K., Lidbury, B.A., 2004. 25 years since the

eradication of smallpox: why poxvirus research is still relevant. Trends

Immunol. 25 (12), 636–639.

Mayr, A., 1999. Historical review of smallpox, the eradication of smallpox

and the attenuated smallpox MVA vaccine. Berl. Muench. Tieraerztl.

Wochenschr. 112 (9), 322–328 (Sep.).

Mayr, A., Hochstein-Mintzel, V., Stickl, H., 1975. Abstammung,eigen-

schaften und verwendung des attenuierten vaccinia-stammes MVA.

Infection 3, 6–14.

Mayr, A., Stickl, H., Muller, H.K., Danner, K., Singer, H., 1978. Der

pockenimptamm MVA: marker, genetische struktur, erfahrungen mit

der parenteralen schutzimpfung und verhalten im abwehrgesch wa

c̈hten organismus. Zentralbl. Bakteriol. Hyg. 167, 375–390.

McMichael, A., Hanke, T., 2002. The quest for an AIDS vaccine: is the

CD8+ T-cell approach feasible? Nat. Rev., Immunol. 2 (4), 283–291.

Meyer, H., Sutter, G., Mayr, A., 1991. Mapping of deletions in the genome

of the highly attenuated vaccinia virus MVA and their influence on

virulence. J. Gen. Virol. 72 (Pt. 5), 1031–1038.

Monath, T.P., Caldwell, J.R., Mundt, W., Fusco, J., Johnson, C.S., Buller,

M., Liu, J., Gardner, B., Downing, G., Blum, P.S., Kemp, T., Nichols,

R., Weltzin, R., 2004. ACAM2000 clonal Vero cell culture vaccinia

virus (New York City Board of Health strain)—a second-generation

smallpox vaccine for biological defense. Int. J. Infect. Dis. Suppl. 2,

S31–S44.

Moss, B., 1991. Poxviridae and their replication. In: Fields, B.N.,

Knipe, D.M. (Eds.), Fundamental Virology. Raven Press, New York,

pp. 953–985.

Moss, B., 1996. Genetically engineered poxviruses for recombinant gene

expression, vaccination, and safety. Proc. Natl. Acad. Sci. U. S. A. 93

(21), 11341–11348.

Paoletti, E., 1996. Applications of pox virus vectors to vaccination: an

update. Proc. Natl. Acad. Sci. U. S. A. 93 (21), 11349–11353.

Ramsey-Ewing, A., Moss, B., 1995. Restriction of vaccinia virus

replication in CHO cells occurs at the stage of viral intermediate

protein synthesis. Virology 206 (2), 984–993.

Ramsey-Ewing, A., Moss, B., 1998. Apoptosis induced by a postbinding

step of vaccinia virus entry into Chinese hamster ovary cells. Virology

242 (1), 138–149.

Reed, L.J., Muench, H., 1938. A simple method of estimating fifty percent

endpoints. Am. J. Hyg. 27, 493–497.

Rodriguez, J.R., Risco, C., Carrascosa, J.L., Esteban, M., Rodriguez, D.,

1997. Characterization of early stages in vaccinia virus membrane

biogenesis: implications of the 21-kilodalton protein and a newly

identified 15-kilodalton envelope protein. J. Virol. 71 (3), 1821–1833.

Sampath, G., Reddy, S.V., Rao, M.L., Rao, Y.U., Palaniappan, C., 2005. An

immunogenicity study of a newly introduced purified Vero cell rabies

vaccine (Abhayrab) manufactured in India. Vaccine 23 (7), 897–900.

Sheets, R., 2000. History and characterization of the Vero cell line. The

Vaccines and Related Biological Products Advisory Committee Meet-

ing. CBER/FDA, USA.

Smith, G.L., Vanderplasschen, A., Law, M., 2002. The formation and

function of extracellular enveloped vaccinia virus. J. Gen. Virol. 83

(Pt. 12), 2915–2931.

Spehner, D., Drillien, R., Proamer, F., Houssais-Pecheur, C., Zanta, M.A.,

Geist, M., Dott, K., Balloul, J.M., 2000. Enveloped virus is the major

virus form produced during productive infection with the modified

vaccinia virus Ankara strain. Virology 273 (1), 9–15.

Spehner, D., Gillard, S., Drillien, R., Kirn, A., 1988. A cowpox virus gene

required for multiplication in Chinese hamster ovary cells. J. Virol. 62

(4), 1297–1304.

Strassburg, M.A., 1982. The global eradication of smallpox. Am. J. Infect.

Control 10 (2), 53–59.

Sun, M.B., Jiang, Y.J., Li, W.D., Li, P.Z., Li, G.L., Jiang, S.D., Liao, G.Y.,

2004. A novel process for production of hepatitis A virus in Vero cells

grown on microcarriers in bioreactor. World J. Gastroenterol. 10 (17),

2571–2573.

Tartaglia, J., Perkus, M.E., Taylor, J., Norton, E.K., Audonnet, J.C., Cox,

W.I., Davis, S.W., van der Hoeven, J., Meignier, B., Riviere, M., 1992.

NYVAC: a highly attenuated strain of vaccinia virus. Virology 188 (1),

217–232.

The Committee for Biological Product Standardization in China, 1979.

Requirements for the Manufacture of Biological Products in China.

Chemical Industry Publishing House of China, Beijing, P.R. China.

Tsao, H., Ren, G.F., Chu, C.M., 1986. Gene coding for the late 11,000-

dalton polypeptide of the Tian Tan strain of vaccinia virus and its 5V-flanking region: nucleotide sequence. J. Virol. 57 (2), 693–696.

Weltzin, R., Liu, J., Pugachev, K.V., Myers, G.A., Coughlin, B., Blum, P.S.,

Nichols, R., Johnson, C., Cruz, J., Kennedy, J.S., Ennis, F.A., Monath,

T.P., 2003. Clonal vaccinia virus grown in cell culture as a new

smallpox vaccine. Nat. Med. 9 (9), 1125–1130.

Wolffe, E.J., Moore, D.M., Peters, P.J., Moss, B., 1996. Vaccinia virus

A17L open reading frame encodes an essential component of nascent

viral membranes that is required to initiate morphogenesis. J. Virol. 70

(5), 2797–2808.

Yan, S.H., Zhang, Sh.Y., Wu, Z.H., Zhang, L.Z., Tian, M.C., Zhao, Y.Y.,

2002. Observation of inoculation reaction and immune efficacy of

edulcorate-lyophilized Vero cell prepared Japanese B encephalitis (JBE)

vaccine in children. Chin. J. Vaccines Immun. 8 (1), 39–41.

Yang, H.F., Chen, C.Z., Pan, M.F., Zhu, Y.T., Lin, F.T., 1979. Screening and

study the attenuated strains form Vaccinia TT stocks. The Corpus for the

30th Anniversary of P.R. China, Chin. J. Virol., pp. 54–61.

Yutu, J.A., Jing, M., Guang, X.H., 1988. Outbreaks of smallpox due to

variolation in China, 1962–1965. Am. J. Epidemiol. 128 (1), 39–45

(Jul.).

Zhang, W.H., Wilcock, D., Smith, G.L., 2000. Vaccinia virus F12L protein

is required for actin tail formation, normal plaque size, and virulence.

J. Virol. 74 (24), 11654–11662.

Zhao, K., Wang, S.Z., Hao, S.L., Han, Y.R., Gong, Y.X., Zhang, Y.F., 1981.

Studies on attenuation of vaccinia virus. Chin. J. Microbiol. Immunol.

1, 421–424.

Zhu, J.H., Wang, J., Cai, B., Zhao, W., Zhu, Y., Chao, R., Chen, L., Xue,

H., Ying, B.L., Li, C.P., Hu, Q.L., Sha, J., Esposito, J.J., 1996.

Immunogenicity and relative attenuation of different vaccinia-rabies

virus recombinants. Arch. Virol. 141 (6), 1055–1065.

host range, growth property, and virulence of the smallpox vaccine: vaccinia virus tian tan strain

Documents