High
Frequency of Homologous Recombination in the Genome of Modified Vaccinia Virus
Ankara Strain (MVA)
ZHU
Li-Xin, XIE You-Hua, LI Guang-Di*, WANG
Yuan*
( Institute of Biochemistry and Cell
Biology, Shanghai Institutes for Biological Sciences,
the Chinese
Academy of Sciences, Shanghai 200031, China )
Abstract MVA
is a genetically modified vaccinia virus strain, which is replication defective
and extremely safe for research and clinical use. To further improve the safety
of MVA and achieve more efficient selection of recombinant constructs, a
transient selection marker system has been developed, which contains vaccinia
virus k1l gene flanked by two identical fragments. In this report, four
recombinant MVAs were constructed with this system, and homologous
recombination of recombinant MVAs was studied during the construction
procedure. The results showed that the recombination frequency of these rMVAs
was significantly high, though lower than that observed in other vaccinia virus
strains. The k1l free rMVAs could be conveniently obtained after 3 or 4
blind passages. Our data indicated that recombinant MVA with a transient
selection marker system was safe for use as live vaccine and gene therapy
vector for human. In addition, blind passages could enhance the efficiency in
isolation of k1l free recombinants.
Key
words recombination frequency;
homologous recombination; vaccinia virus; MVA
Vaccinia
virus(VV), the prototype orthopox-virus, is a large animal virus with
double-stranded linear DNA genome of about 190 kb, which potentially encodes
263 viral proteins[1]. VV was used as the live vaccine against variola
virus, the causative agent of smallpox. By the year of 1979, the smallpox
vaccination campaign resulted in the eradication of variola worldwide.
Homologous recombination of VV is the basis for the marker rescue technique
frequently used in the studies of many VV genes (e.g. reference [2,3]). More
importantly, the recombination between VV and plasmid harboring VV gene
fragments made vaccinia virus an efficient and versatile system for eucaryotic
gene expression study(reviewed by reference [4]) and vaccine
development(reviewed by reference [5]). People taking VV as a vector enjoys
many advantages of VV system, e.g. the ability of VV to grow high titer stocks
(1010 plaque forming units/ml); large capacity of VV genome for
foreign DNA; considerable latitude in choosing cells provided by the wide host
range of VV; proper processing, post-translational modification and
transportation of expressed eucaryotic proteins. Nevertheless, the occurrence
of rare adverse reactions to smallpox vaccination and the increased
susceptibility of immunodeficient individuals called for improved safety.
MVA
was generated from vaccinia Ankara strain CVA by hundreds of passages in CEF
cells. During the process, multiple deletions of more than 30 kb DNA including
at least two viral host range genes occurred[6], which turned this
VV vector extremely safe for research and clinical use[6,7]. In
addition, the comparatively smaller genome may enhance the capacity of MVA to
incorporate large foreign sequences.
Live
vaccines with MVA as the vector were able to efficiently induce humoral[8–10]
and cellular immune responses[11–13] against antigens expressed by
the recombinant MVAs. To ensure further safety and more efficient selection of
recombinant MVA, Staib et al[14] recently developed a
transient selection marker system which introduced a selection marker flanked
by two identical sequences into MVA genome. In this system, the k1l gene
that has been reported to play a critical role in VV host range selection[15]
was used as the selection marker. The exact mechanism of its function remained
unclear. Therefore, it is necessary to limit the use of k1l gene to
selection of rMVA only and discard this marker from the final recombinant MVA.
In this transient selection system, k1l gene will automatically
disappear through homologous recombination if selection pressure is released.
The homologous recombination between these identical sequences can be easily
monitored by PCR. In this study, four recombinant MVAs based on this system
were generated and the homologous recombination of recombinant MVAs was
studied. We report here the detailed analysis and calculation of the frequency
of the homologous recombination. Our data will improve the current
understanding on the safety of the recombinant MVA system using transient
selection method, on the stability of MVA genome, and guide future generation
of recombinant MVAs.
1 Materials and Methods
1.1 Plasmids,
viruses and cells
Vector
pIIIdHR-P7.5 which contains the selection marker k1l and wild type
MVA(wtMVA) used in this study (provided by Dr. Sutter G, National research
center for environment and health, Germany) were described previously[14].
Baby hamster kidney BHK-21 (ATCC CCL-10) and rabbit kidney RK-13 (ATCC CCL-37)
cells were grown in RPMI1640 supplemented with 10% FCS. Cells were maintained
in a humidified incubator at 37 ℃
and 5% CO2.
1.2
Generation of recombinant MVAs
Structures
of MVA and recombination events in generation of recombinant MVAs were depicted
in Fig.1. Recombinant viruses were constructed mainly according to the
procedure described previously[14]. Briefly, foreign gene fragments
encoding a polyprotein of different lengths (2.60 kb, 2.22 kb, 1.98 kb and 1.53
kb, which will be described in detail elsewhere) were inserted into the MCS of
pIIIdHR-P7.5. Respectively, the resulting plasmids pIIIdHR-P7.5A,
pIIIdHR-P7.5B, pIIIdHR-P7.5C and pIIIdHR-P7.5D were then used to transfect
BHK-21 cells that were pre-infected with wtMVA. Cells were harvested two days
post-transfection and passaged in RK-13 cells that did not support the
replication of wtMVA. Only recombinant MVAs derived from homologous
recombination between wtMVA and transfected plasmid containing the selection
marker k1l (MVA-k1lFS) were recovered. The second homologous
recombination aiming to delete the k1l gene from the MVA genome was
allowed to happen in BHK-21 cells, where growth pressure was released. To
facilitate the generation of MVA-FS, we introduced blind passages in the
procedure. After several blind passages in BHK-21 cells, each recombinant MVA
was plaque purified and MVA-FS was isolated after one plaque-picking passage.
During each passage, viral genomic DNA from several plaques was extracted for
PCR analysis and cell lysates tested by western blot for the expression of
foreign gene products.
Fig.1 Recombination in generation of
recombinant virus MVA-k1lFS and MVA-FS
Schematic maps of the MVA genome (HindIII
restriction map) is shown at the top of the figure. MVA-k1lFS is the
recombinant MVA containing selection marker k1l gene. Flank1 and Flank2
refer to the MVA-DNA sequences that are essential to target insertion of
foreign sequences (FS) to the site of deletion III within the MVA genome. rec2
indicates the position of a 283 bp repetitive MVA-DNA fragment homologous to
the right end of Flank1, which allows deletion of the k1l expression
cassette by homologous recombination. MVA-FS is the recombinant MVA in which
the k1l gene has been deleted through further recombination. The
annealing sites for the primers (MVA-III-5′ and MVA-III-3′) used in PCR
analysis are indicated on the genome of MVA-k1lFS.
1.3 Western blot analysis
BHK-21
cells were infected with recombinant MVA or wtMVA with an MOI of 10 IU per
cell. Infected cells were harvested 48 hours post infection, washed twice with
PBS and directly lysed with SDS-PAGE loading buffer. After electrophoresis, the
separated proteins were transferred onto nitrocellulose membrane (Schleicher
& Schuell). The membrane was then blocked in 5% fat-free milk powder
prepared in PBST (PBS with 0.5% Tween 20) before probing with rabbit anti-serum
raised against the C region of the polyprotein. After that, the membrane was
incubated with HRP-conjugated goad-anti-rabbit IgG(Dianova, Germany) and
developed with Lumi-light Western blotting substrate(Roche).
1.4 Monitoring
recombination by PCR
The
recovered MVA-k1lFS (MVA-k1lA, MVA-k1lB, MVA-k1lC
and MVA-k1lD) were blindly passaged in BHK-21 cells for 3 or 4 times. To
reduce intermolecular recombination, MOI lower than 0.01 IU per cell was used.
For each passage, cells were harvested in 1 ml of PBS three days
post-infection, thoroughly homogenized, diluted, and used to re-infect BHK-21
cells for the next passage.
To
determine the recombination frequency of rMVA, cells were thoroughly
homogenized by the end of the final blind passage. Cell lysates were diluted
and plated on BHK-21 cell monolayers at 10–5 MOI per cell. Two days
later, several plaques were picked and amplified. Recombinant MVA DNA were
prepared[16] and analyzed by PCR.
The
primer pair used in this study (MVA-III-5′ and MVA-III-3′, see Fig.1) was
described previously[14]. They anneal to the flanking regions of
deletion III in MVA genome. With wtMVA, k1l+ rMVA and k1l–
rMVA as template, PCR with this primer pair generates DNA fragments of 0.76 kb,
2.28 kb＋FS
(with k1l) and 0.98 kb＋FS
(without k1l), respectively. PCR reactions were performed with PCR
Master kit (Roche) with additional DMSO added to each tube (final
concentration: 4%). After denaturation at 92 ℃
for 2 min, 30 cycles of amplification were carried out as following: 92 ℃,
30 s; 55 ℃,
40 s; 68 ℃,
4 min. The procedure was ended with an additional elongation of 8 min at 68 ℃.
1.5 Calculation
of recombinant frequency
Assuming
that the recombination frequency during each passage was a constant, we could
calculate the recombination frequency as following.
Let
r denote the recombination frequency during each passage. After one
passage in BHK-21 cells, we have:
Ratio
of recombinants: r. Ratio of non-recombinants: 1–r.
During
the second passage in BHK-21 cells, recombination continued with those non-recombinants.
Therefore, after two passages, we have:
Ratio
of recombinants: r+(1–r)r = 2r–r2 =
1–(1–r)2.
Ratio
of non-recombinants: (1–r)2.
Similarly,
after three passages in BHK-21 cells, we have:
Ratio
of recombinants: 2r–r2+(1-r)2r =
2r–r2+r3–2r2+r
= r3–3r2＋3r
= 1–(1–r)3.
Ratio
of non-recombinants: (1–r)3.
In
general, after n blind passages in BHK-21 cells, we have:
Ratio
of recombinants: 1–(1–r)n.
Ratio
of non-recombinants: (1–r)n.
After
blind passages, the final recombination frequency (R) could be
determined by PCR analysis.
So
from R＝1–(1–r)n,
we can calculate the recombination frequency during each passage.
2 Results
2.1 Construction
of rMVA carrying k1l selection marker flanked by two identical sequences
Recombinant
MVA carrying two identical sequences (Fig.1) is a convenient system for study
of homologous recombination. For this purpose, four MVA-k1lFS were
constructed by homologous recombination between wtMVA and plasmids that
contained foreign sequence (FS) and k1l selection marker flanked by two
identical rec2 sequences (see materials and methods). Only recombinant MVA
carrying thek1l selection marker could replicate and propagate in RK-13
cells while k1l– MVA could not. Recombinant MVAs from
positive plaques were picked, diluted and used to re-infect RK-13 cells at a
low MOI per cell. After several passages in RK-13 cells, pure recombinant MVAs
carrying k1l selection marker flanked by two identical sequences of 283
bp were recovered.
Recovered
four MVA-k1lFS were subsequently characterized by PCR and western blot
analysis. PCR with primers specific for FS demonstrated that all rMVAs carried foreign
sequences. No wtMVA signal (0.76 kb) was observed in PCR with
MVA-III-5’/MVA-III-3′ primer pair (data not shown). The result indicated that
recovered viruses were recombinant and free of wild type MVA. Western blot
analysis with antibodies specific for FS product confirmed that all rMVAs were
able to express foreign inserts (Fig.2). These rMVAs are then subjected to the
recombination study.
Fig.2 Western blot analysis of the
expression products of all MVA-k1lFS
The
specific protein bands reacted with the anti-serum against the C region of
polyprotein were detected for each virus confirming the correct generation of
all MVA-k1lFS.
2.2 Analysis of homologous recombination by PCR
When
directly plating the above-mentioned rMVAs on BHK-21 cell monolayers which do
not exert any selection pressure, incubating for two days and analyzing
isolated viral plaques by PCR, we always observed two bands instead of only a k1l–
single band (Fig.3). The lack of a 0.76 kb band indicated that the isolates
were free of wtMVA. As to the PCR products from recombinant MVA templates, the
upper band was amplified from the k1l gene containing MVA genome, while
the lower band was from the k1l free recombinant MVA genome, which could
be generated during one passage of MVA-k1lFS on BHK-21 cells. As we and
others in our lab never observed pure k1l free MVA isolate from hundreds
of MVA-k1lFS isolates, we drew the conclusion that all original isolates
were k1l positive.
Fig.3 PCR analysis showing the
difficulty in isolation of k1l free recombinant MVA
Wild type free MVA-k1lD was
directly plated on BHK-21 cell monolayer. Six plaques were isolated and analyzed
by PCR method with primer pair MVA-III-5′/MVA-III-3′.
For all the samples of recombinant MVA, two bands were detected, the upper band
was amplified from the MVA genome containing k1l gene, while the lower
one from k1l free MVA genome. No pure k1l free virus represented
by a single lower band was detected. Analysis of other plaques gave similar
results. M, 1 kb ladder.
We
then blindly passaged the original k1l+ rMVAs for 3 or 4
times to facilitate the deletion of k1l gene. During each passage, viral
DNA was prepared and analyzed by PCR [Fig.4(A)]. All four rMVAs progressively
lost k1l signals during blind passages. After final blind passage, each
rMVA was plated on BHK-21 cell monolayers at an MOI of 10-5 IU per
cell. Plaques were isolated and analyzed by PCR to monitor the recombination
happened in the genome of recombinant MVAs. Fig.4(B) shows that we could obtain
k1l free recombinant MVA for each MVA-FS from only one plaque-picking
passage on BHK-21 cells. For each of the recombinant MVAs after the final blind
passage, more than 25% of the isolated plaques were the products resulted from
homologous recombination between the identical rec2 sequences, which were k1l
negative isolates. The results revealed that the k1l free MVA-FS could
be easily obtained after blind passages.
Fig.4 Analysis of the recombination
frequency after blind passages
(A) Four MVA-k1lFS were subjected
to 3 or 4 times of blind passages, when the k1l free recombinant MVAs
were accumulating. P1, P2, P3, P4 indicate the number of blind passages carried
on BHK-21 cells. (B) The recombinant MVA from final blind passages were then
subjected to plaque-picking passage on BHK-21 cell monolayer. S1―S8
indicate the isolated plaques (samples) analyzed by PCR. Pure k1l free
viruses represented by a single lower band were detected for each recombinant
virus. All PCR were performed with primer pair MVA-III-5’/MVA-III-3′. M, 1 kb
ladder.
2.3 Estimation of the homologous recombination frequency
From
Fig.4, recombinants were counted and the final recombination frequency was
calculated (Table 1).
The
recombination frequencies during one blind passage were also estimated as
described in Materials and Methods. An average of 0.136 with a standard
deviation of 0.056 was obtained. It indicated that MVA genomes carrying two
identical sequences underwent efficient homologous recombination when the
growth pressure was released.
3 Discussion
Homologous
recombination of VV was initially observed in cells co-infected with two
distinct VV strains. The studies of recombination in VV system with Southern blot
demonstrated a high frequency of homologous recombination[17].
Furthermore, the recombination frequency correlated linearly to the length of
homologous sequence[18].
In
this study, we constructed four MVA-FS based on the transient selection marker
system developed recently and carried out a detailed analysis on the events of
homologous recombination during the generation of these recombinant MVAs. We
observed that the recombination of MVA genome is significant, though not as
frequent as that of VV studied previously which reach as high as 40%[17].
After several blind passages on BHK-21 cell monolayer without selection
pressure, we calculated the recombination frequency per passage to be within 8%―20%.
This suggests that rMVA without the k1l selection marker can be easily
selected after three or four passages in BHK-21 cells. Therefore, our data
presents important evidence in supporting the conclusion that recombinant MVA
with the transient selection marker system is both safe and convenient, which
may turn out to be a valuable live vaccine and gene therapy vector for human.
The standard deviation of the calculated recombination frequencies was
comparably high, which was understandable, as the quantity of analyzed samples
was limited. For a precise measurement of the recombination frequency, more
samples will be required.
The
homologous recombination of vaccinia virus can be both intramolecular and
intermolecular. Southern blot analysis showed that intermolecular recombination
was much less frequent than intramolecular recombination[17],
possibly because two crossovers are needed for intermolecular recombination,
while intramolecular recombination only involves one crossover. To identify
rMVAs generated from intermolecular recombination would give a more accurate
estimation of the recombination frequency, which should be close to the present
result.
To
avoid multiple infection of one cell in analyzing the recombination frequency,
our approach was to homogenize the virus preparation as much as possible and to
use an extremely low MOI of MVA to infect cell monolayers. Comparing with
Southern blot analysis, PCR method was much more convenient and sensitive in
analyzing the structure of MVA genome, which had also been demonstrated in the
procedure of generating rMVAs[14]. However, the data from PCR
analysis on each plaque (Fig.3) was not quantitative. We noticed that the lower
band signal that represented the k1l– rMVA in the last blind
passages [Fig.4(A)] was stronger than that of the upper band for the k1l+
rMVA. This did not mean that there were more k1l– rMVAs than k1l+
rMVAs, since the recombination frequency was estimated within 8%―20%
per passage. It was likely that k1l– signal was much
exaggerated by PCR due to smaller fragment (1.3 kb smaller than k1l+)
being preferentially amplified.
The
capability to induce protective immune responses and safety of MVA made it a
promising vector in vaccine development. Recently, recombinant MVA-HIV was used
as the main component of an HIV vaccine in a study on macaques model[19].
Animals inoculated with such a vaccine were reported to produce high T-cell
immune responses that contained viruses after SHIV challenge. This breakthrough
in HIV vaccine development made rMVA more attractive in fighting viral diseases
and cancers. Therefore, the improvement on the technique to generate
recombinant MVAs, namely, transient selection marker system, will be more
widely used. We demonstrated that 3 or 4 blind passages would facilitate the
selection of k1l– rMVAs. Since blind passage translates to
much less work and more cost effective means than plaque-picking passage, our
data reported here would help to guide future generation of rMVAs based on the
transient selection marker system.
Acknowledgements The authors would like to thank Dr.
Sutter G and Dr. Staib C for their kind help in MVA manipulation.
References
1 Goebel
SJ, Johnson GP , Perkus ME, Davis SW, Winslow JP, Paoletti E. The complete DNA
sequence of vaccinia virus. Virology, 1990, 179(1): 247―266, 517―563
2 Kane
EM, Shuman S. Vaccinia virus morphogenesis is blocked by a
temperature-sensitive mutation in the I7 gene that encodes a virion component. J
Virol, 1993, 67(5): 2689―2698
3 Roper
RL, Wolffe EJ, Weisberg A, Moss B. The envelope protein encoded by the A33R gene
is required for formation of actin-containing microvilli and efficient
cell-to-cell spread of vaccinia virus. J Virol, 1998, 72(5): 4192―4204
4 Smith
GL. Vaccinia virus vectors for gene expression. Curr Opin Biotech, 1991,
2(5): 713―717
5 Ulaeto
D, Hruby DE. Uses of vaccinia virus in vaccine delivery. Curr Opin Biotech,
1994, 5(5): 501―504
6 Meyer
H, Sutter G, Mayr A. Mapping of deletions in the genome of the highly
attenuated vaccinia virus MVA and their influence on virulence. J Gen Virol,
1991, 72(5): 1031―1038
7 Drexler
I, Heller K, Wahren B, Erfle V, Sutter G. Highly attenuated modified vaccinia
virus Ankara replicates in baby hamster kidney cells, a potential host for
virus propagation, but not in various human transformed and primary cells. J
Gen Virol, 1998, 79(2): 347―352
8 Men
R, Wyatt L, Tokimatsu I, Arakaki S, Shameem G, Elkins R, Chanock R et al.
Immunization of rhesus monkeys with a recombinant of modified vaccinia virus
Ankara expressing a truncated envelope glycoprotein of dengue type 2 virus
induced resistance to dengue type 2 virus challenge. Vaccine, 2000, 18(27):
3113―3122
9 Ourmanov
I, Bilska M, Hirsch VM, Montefiori DC. Recombinant modified vaccinia virus
Ankara expressing the surface gp120 of simian immunodeficiency virus (SIV)
primes for a rapid neutralizing antibody response to SIV infection in macaques.
J Virol, 2000, 74(6): 2960―2965
10 Wyatt
LS, Whitehead SS, Venanzi KA, Murphy BR, Moss B. Priming and boosting immunity
to respiratory syncytial virus by recombinant replication-defective vaccinia
virus MVA. Vaccine, 1999, 18(5-6): 392―397
11 Belyakov
IM, Wyatt LS, Ahlers JD, Earl P, Pendleton. CD, Kelsall BL, Strober W et al.
Induction of a mucosal cytotoxic T-lymphocyte response by intrarectal
immunization with a replication-deficient recombinant vaccinia virus expressing
human immunodeficiency virus 89.6 envelope protein. J Virol, 1998, 72(10):
8264―8272
12 Drexler
I, Antunes E, Schmitz M, Wolfel T, Huber C, Erfle V, Rieber P et al. Modified
vaccinia virus Ankara for delivery of human tyrosinase as melanoma-associated
antigen: Induction of tyrosinase- and melanoma-specific human leukocyte antigen
A*0201-restricted cytotoxic T cells in vitro and in vivo. Cancer
Res, 1999, 59(19): 4955―4963
13 Seth
A, Ourmanov I, Schmitz JE, Kuroda MJ, Lifton MA, Nickerson CE, Wyatt L et al.
Immunization with a modified vaccinia virus expressing simian immunodeficiency
virus (SIV) Gag-Pol primes for an anamnestic Gag-specific cytotoxic
T-lymphocyte response and is associated with reduction of viremia after SIV
challenge. J Virol, 2000, 74(6): 2502―2509
14 Staib
C, Drexler I, Ohlmann M, Wintersperger S, Erfle V, Sutter G. Transient host
range selection for genetic engineering of modified vaccinia virus Ankara. BioTechniques,
2000, 28(6): 1137―1148
15 Wyatt
LS, Carroll MW, Czerny CP, Merchlinsky M, Sisler JR, Moss B. Marker rescue of
the range restriction defects of modified vaccinia virus Ankara. Virology,
1998, 251(2): 334―342
16 Drexler
I, Heller K, Ohlmann M, Erfle V, Sutter G. Recombinant vaccinia virus MVA for
generation and analysis of T cell responses against tumor associated antigens.
In: Walther W, Stein U eds. Methods in Molecular Medicine, Vol.35: Gene
Therapy: Methods and Protocols, Totowa, NJ: Humana Press, 1999, 57―73
17 Ball
LA. High-frequency homologous recombination in vaccinia virus DNA. J Virol,
1987, 61(6): 1788―1795
18 Fujitani
Y, Yamamoto K, Kobayashi I. Dependence of frequency of homologous recombination
on the homology length. Genetics, 1995, 140(2): 797―809
19 Amara
RR, Villinger F, Altman JD, Lydy S L, O’Neil SP, Staprans SI, Montefiori DC et
al. Control of a mucosal challenge and prevention of AIDS in rhesus
macaques by a multiprotein DNA/MVA vaccine. Science, 2001, 292(5514):
69―74
Received: April 13, 2001 Accepted: May 29, 2001
This work was supported by a grant from
the State 863 High Technology R&D Project of China (No.863-102-07-02-02)
and the Project CHN 98/112 (WTZ-Internationales Büro
des BMBF, Germany)
*Corresponding authors: Tel,
86-21-64374430; Fax, 86-21-64338357; e-mail, [email protected]