|
|
|
|||||||
|
|||||||||
(Received for publication, August 4,
1994; and in revised form, October 12, 1994) From the
T-DNA processing during agroinfection of plants is initiated by
site- and strand-specific incision at the T-DNA border sequences of the
Ti plasmid. Two proteins are required for this reaction: VirD2 (49.6
kDa), catalyzing a site-specific cleaving-joining reaction on
single-stranded DNA in vitro (Pansegrau, W., Schoumacher, F.,
Hohn, B., and Lanka, E.(1993) Proc. Natl. Acad. Sci. U. S. A. 90, 11538-11542), and VirD1 (16.1 kDa), an accessory protein
required for VirD2-mediated specific cleavage of double-stranded DNA.
Following efficient overproduction, VirD1 was isolated in active form
from inclusion bodies and purified to near homogeneity. The protein was
applied together with purified VirD2 protein for specific cleavage of
double-stranded T-DNA border sequences in vitro. The reaction
proceeds on negative superhelical DNA and requires Mg Agrobacterium tumefaciens is a prokaryotic plant
pathogen that transforms plant cells by a unique interkingdom DNA
transfer system. A particular DNA stretch of its 200-kb ( Functions required for
T-DNA transfer are encoded by the virulence (Vir) region ( Data have accumulated that
suggest a close relationship between T-DNA transfer from Agrobacterium spp. to plants and IncP plasmid-mediated
bacterial conjugation. Sequence relations have been found between four
components: the nick regions of T-borders and the IncP transfer origin
(Pansegrau and Lanka, 1991; Waters et al., 1991; Waters and
Guiney, 1993) and gene clusters of the VirD operon/relaxase operon
(Lessl and Lanka, 1994; Pansegrau and Lanka, 1991; Pansegrau et
al., 1994a, 1994b; Ziegelin et al., 1991), VirC/leader
operon (Pansegrau et al., 1994a), and VirB/Tra2 region of
pTi/IncP plasmids (Lessl et al., 1992), respectively. The gene
organization of the corresponding operons is highly conserved, and a
high degree of sequence identity between the primary structures of gene
products indicates evolutionary relationship of the DNA transfer
systems (Pansegrau et al., 1994a). Although there is little
sequence similarity detectable, the pTi virE2 gene product
seems to be functionally analogous to the TraC1 primase of IncP plasmid
RP4 concerning DNA binding properties. Both proteins have been shown to
be transported together with the DNA to recipient cells and therefore
are supposed to coat and protect the ssDNA during transfer (Gietl et al., 1987; Christie et al., 1988; Citovsky et
al., 1989, 1992, 1994; Rees and Wilkins, 1990). However, a primase
activity has not been detected in the VirE2 protein. ( The pTi VirB region is related to three other
macromolecular transport systems, the pilus gene clusters of IncW ( The mechanism of T-DNA transfer
shows striking similarity to bacterial conjugation mediated by IncP
plasmids; in both systems, a DNA single strand is transferred with the
5` end leading (Al-Doori et al., 1982; Grinter, 1981;
Zambryski, 1992). During transfer, the 5` terminus is thought to remain
covalently associated with the relaxase (VirD2/TraI)
(Dürrenberger et al., 1989; Howard et
al., 1989; Pansegrau et al., 1990b, 1993a, 1993b; Young
and Nester, 1988). Functional similarity between VirD2 and TraI has
also been detected in vitro; VirD2 and TraI catalyze
site-specific cleaving-joining reactions on single-stranded
oligonucleotides reaching an equilibrium when 35-40% of the input
DNA exists in the cleaved form (Pansegrau et al., 1993a,
1993b). The reactions are Mg In the IncP system, the corresponding
reaction on dsDNA in addition to the TraI relaxase requires the
accessory protein TraJ (Pansegrau et al., 1990a). TraJ is a
specific oriT-binding protein (Ziegelin et al., 1989)
proposed to form a nucleoprotein complex with negatively superhelical oriT DNA that can be recognized by TraI (Pansegrau et
al., 1990a). VirD1 could play an analogous role in the Ti system
since virD1 and virD2 are the only determinants
essential for T-border-specific cleavage in vivo. Any DNA
sequence located between T-borders can be transferred efficiently;
therefore, plant oncogenes and genes for opine catabolism may be
replaced by any other gene of interest. The resulting so-called
disarmed vectors are extremely helpful as widely used tools for genetic
manipulation or engineering of plant cells by transformation. However,
T-DNA transfer is limited in its host range, allowing efficient
transformation of most dicotyledonous, but of only a few
monocotyledonous, plants. Since monocots are of great economic
importance, it is desirable to find efficient methods allowing their
genetic transformation. A possible way could consist in generation of
the T-complex in vitro and subsequent transfer into plant
cells by common methods such as microinjection or particle gun
transformation. Since VirD2 and VirE2 possess nuclear localization
signals (Citovsky et al., 1992, 1994; Rossi et al.,
1993; Tinland et al., 1992), higher transformation rates than
with conventional methods are expected. A prerequisite for these
experiments is the in vitro reconstitution of the initiation
complex active in T-border-specific DNA cleavage. Here we report the
overproduction and purification of VirD1 protein of the A.
tumefaciens plasmid pTiC58. Efficient overproduction of VirD1 in Escherichia coli required exchange of the natural ribosome
binding site against that of bacteriophage T7 gene 10 and
replacement of a cluster of rare Arg codons within the virD1 structural gene. The purified VirD1 protein together with VirD2
relaxase (Pansegrau et al., 1993a) was applied to cleave
superhelical plasmid DNA containing T-border sequences in
vitro. The in vitro cleavage reaction was shown to mimic
the virD1/virD2-dependent T-border cleavage reaction observed in vivo in terms of site and strand specificity, resulting in
covalent attachment of the VirD2 relaxase to the 5` terminus of the
cleaved DNA strand. Implications for the mechanism of T-DNA processing
and functional analogies to DNA processing during bacterial conjugation
are discussed.
Figure 1:
Construction of virD1 overexpression plasmid pPS20. Upper panel,
nucleotide sequence flanking the initiation codons in pTiC58/pVir97.89
(Alt-Mörbe and Schröder, 1986)
and a sequence stretch 40 bp upstream of the stop codon. The
Shine-Dalgarno sequence (S/D) is indicated by a horizontalline; potential initiation codons are shown as blackboxes with whiteletters. MaeIII restriction enzyme recognition site is marked by a bracket above the sequence. Lowerpanel, to
construct the virD1 overexpression plasmids, the 464-bp MaeIII-SacI fragment of pVir97.89 carrying the virD1 reading frame except for the first 10 codons was
inserted in the NdeI and SacI sites of the polylinker
of the T7 promoter
The virD1 gene of pTiC58 including
the original translational initiation site was placed under control of
the LacI-regulated tac promoter in the expression vector
pMS119HE. The resulting plasmid was called pPS9 (Table 1). To
allow effective translational initiation in E. coli fusions of
start 1 or start 2 with the S/D sequence of phage T7 gene 10 were constructed that resulted in plasmids pPS15 (start 1) and
pPS13 (start 2). The rare Arg codons in these two plasmids were
replaced by alternative Arg codons using site-directed mutagenesis.
Resulting plasmids were named pPS22 (start 1) and pPS20 (start 2) ( Fig. 1and Table 1). Gene products specified by all the
constructs were identified and quantified by immunoblotting (Fig. 2). As expected, cells carrying plasmid pPS9 with the
original 5` end of virD1 expressed low amounts of a 16-kDa
polypeptide (Fig. 2, lane a). In cells containing
plasmid pPS13 (start 2, wild type Arg codons) the yield of this
polypeptide was approximately 10 times higher (Fig. 2, lane
b). Truncated forms of VirD1 were still produced but were absent
in extracts of SCS1(pPS20) containing the alternative Arg codons in virD1 (Fig. 2, lanes b and d; Table 1). In cells carrying constructs with start 1 (pPS22 and
pPS15) translation initiated at both, start 1 and 2; consequently two
VirD1 polypeptide versions were produced. Four products were observed
if the rare Arg codons were present (Fig. 2, lanes c and e; Table 1). The N termini of these four
polypeptides were determined by microsequencing. The amino acid
sequence of the two upper bands corresponded to peptides derived from
start 1, that of the two lower bands to translational initiation at
start 2. These results verify that both AUG codons are used for
initiation of translation. It also suggests that truncation occurs at
the C terminus of VirD1 resulting from premature termination of
translation caused by the cluster of rare Arg codons. The N-terminal
sequences of the product of SCS1(pPS20) (S/D T7 gene 10/start
2 and alternative Arg codons) and the lower band produced by
SCS1(pPS22) (S/D T7 gene 10/start 1 and alternative Arg
codons) are identical and correspond to start 2. We concluded that
start 2 functions as the predominantly used translational initiation
site because the size of the polypeptide specified by pPS9 and that of
the smaller one specified by pPS22 are identical. In both constructs
the original sequence around start 2 was preserved. In SCS1(pPS20) virD1 translation initiates at start 2 and high levels of
VirD1 are produced without contaminating truncated forms; hence, this
clone was used for overproduction and subsequent purification of the
protein. Following chemical induction of gene expression by IPTG
addition to the culture medium, approximately 8% of SDS-soluble cell
protein consisted of VirD1 (Fig. 3, lane b).
Figure 2:
Solid phase immunoassay of overproduced
VirD1 proteins. Extracts of SCS1 cells induced for 4 h by addition of 1
mM IPTG to the culture carrying various plasmids were loaded
on a 17.5% polyacrylamide gel containing 0.1% SDS. Following
electrophoresis proteins were transferred to a nitrocellulose membrane
by electroblotting and detected as described under
``Experimental Procedures.'' Lanea, pPS9 (300 A
Figure 3:
Purification of VirD1 protein. Samples
were electrophoresed on a 17.5% polyacrylamide gel containing 0.1% SDS.
The gel was stained with Serva Blue R. Lane a, SCS1(pPS20), no
IPTG was added; lane b, SCS1(pPS20), cells were induced for 4
h with 1 mM IPTG; lanes c-e, fractions
I-III (35, 14, and 9 µg of protein, respectively); lane
f, molecular mass standards: bovine serum albumin (BSA,
68 kDa), ovalbumin (OVA, 46 kDa), chymotrypsinogen A (CHYA, 25.7 kDa), lysozyme (LYS, 14.3 kDa), and
aprotinin (APR, 6.5 kDa).
Figure 4:
Specific cleavage of T-border plasmids
catalyzed by VirD1 and VirD2. A, various plasmids were
incubated with VirD proteins under conditions described under
``Experimental Procedures'' and electrophoresed in a
0.7% agarose gel (89 mM Tris borate, pH 8.0, 2 mM EDTA) at 1.5 V/cm. Lanes a-e, pPS100 (right border,
0.7 µg of DNA); a, proteins omitted; b, 500 ng of
VirD1; c, 60 ng of VirD2; d, 500 ng of VirD1 and 60
ng of VirD2, MgCl
Previously, a nonspecific DNA
relaxation activity was reported for VirD1 containing extracts (Ghai
and Das, 1989). Under our conditions and also using the conditions
described by Ghai and Das(1989), this activity was never observed with
purified VirD1 protein that is active in T-border cleavage (Fig. 4, lane b). E. coli and calf thymus
topoisomerase I served as positive controls; upon incubation with the
topoisomerases, all of our DNA substrates gave a spectrum of
topoisomers independent of whether vector DNA or T-border DNA was used
(data not shown). Under our conditions, superhelicity of the DNA was an
essential prerequisite for specific T-border cleavage by VirD1/VirD2,
and form III and form IV DNA did not act as a substrate (data not
shown). This finding is analogous to that of the cleavage reaction at
RP4 oriT catalyzed by TraI and TraJ protein, which also
required superhelical DNA substrates (Pansegrau et al.,
1990a). While the nick regions of RP4 oriT and pTi T-borders
share close sequence similarity (Pansegrau and Lanka, 1991; Waters et al., 1991) and VirD2 cleaves RP4 nick region
oligonucleotides (Pansegrau et al., 1993a). VirD1 and VirD2
together failed to cleave oriT form I plasmid DNA (data not
shown). TraJ instead of VirD1 in the reaction mixture did not stimulate
cleavage of RP4 oriT DNA by VirD2 (data not shown). These
results indicate the higher degree of specificity of the dsDNA cleavage
reaction as compared to ssDNA cleavage and suggest that VirD1
contributes to the stringency of T-border recognition on dsDNA
substrates.
Figure 5:
Analysis of cleavage products. A,
pPS100 form II DNA specifically cleaved by VirD1/VirD2 in vitro was isolated from 0.7% agarose gels. 1.5 µg of DNA were
linearized with either XmnI or AatII, denatured in
0.1 M NaOH and electrophoresed for 6 h (3.5 V/cm) on a
alkaline 0.7% agarose gel (McDonnell et al., 1977). Lane M contains a 1-kb ladder (Life Technologies, Inc.). B, the
physical structure of pPS100 and the corresponding cleavage products
including the position of the cleavage site (nic) are drawn
schematically.
Figure 6:
Nucleotide sequencing of the 5` terminus
of specifically cleaved T-border plasmid DNA. I, pPS100 form
II DNA (0.5 µg) specifically cleaved by VirD1/VirD2 was used as
substrate for primer extension. Reactions were initiated at a 24-mer
primer [d(GTGCGGCGACGATAGTCATGCCCC)] hybridizing 77 bp
downstream from the nic site. II, as a control
template pPS100 form I DNA (2 µg) was used for sequencing
reactions.
A key event in the initiation of DNA processing for T-DNA
transfer resulting in the production of T-complexes is the site- and
strand-specific cleavage at the Ti plasmid border sequences. Based on
genetic studies, it was demonstrated that this reaction is mediated by
products of the Ti plasmid's VirD operon. In particular, the virD1 and virD2 gene products have been shown to be
the only Ti-encoded polypeptides required for T-border cleavage (De Vos
and Zambryski, 1989; Filichkin and Gelvin, 1993; Jayaswal et
al., 1987; Porter et al., 1987; Stachel et al.,
1987; Yanofsky et al., 1986). In a previous study, it was also
shown that VirD2 alone exerts a site-specific DNA cleaving-joining
reaction on single-stranded DNA, indicating that this protein bears the
catalytic activity required for DNA scission (Pansegrau et
al., 1993a). In vitro experiments employing VirD2 in the
presence of dsDNA, either relaxed or supercoiled, failed to detect any
border-specific cleavage, suggesting that an essential accessory
component for dsDNA cleavage is missing (Pansegrau et al.,
1993a). Others reported that even the combination of the purified
proteins VirD1 and VirD2 was inactive in specific cleavage of dsDNA
(Jasper et al., 1994). In the present study we demonstrate
that the proteins VirD1 and VirD2 together are sufficient to catalyze
the T-border-specific cleavage on dsDNA in vitro. The reaction
takes place on negative superhelical DNA in the presence of
Mg What could the role of
VirD1 in the dsDNA cleavage reaction be? Obviously, VirD2 recognizes
specific nucleotide sequences in ssDNA; therefore, the VirD2 target
surface in dsDNA probably is buried. Binding of VirD1 to the T-border
could locally distort the DNA double helix structure, exposing the
T-border nick region as a single strand for cleavage by VirD2. Negative
supercoiling of the DNA, which is an essential requirement for the
reaction to occur, could lower the energy required for local strand
separation. Alternatively, VirD1 could assemble in a complex with VirD2
in the absence of DNA, resulting in an altered recognition specificity,
directed against dsDNA. Thus far, there is no clear evidence to
decide between the two hypotheses. Complex formation of VirD1 with
linear double-stranded DNA under our conditions was not detectable by
the fragment retardation assay (not shown). However, negative
supercoiling of the DNA might also be a prerequisite for binding of
VirD1. On the other hand, detection of complex formation between VirD1
and VirD2 by chemical cross-linking in the absence of DNA and analysis
of the products by gel electrophoresis did not yield specific
VirD1-VirD2 protein-protein complexes (not shown). Recently, a
topoisomerase activity of type I was described for VirD1-containing
extracts (Ghai and Das, 1989). This activity was attributed to VirD1
and proposed to be required for relaxing the DNA in order to prepare it
for cleavage by VirD2 (Ghai and Das, 1989). In our hands, the purified
VirD1 protein never showed any topoisomerase activity. Assuming that
the denaturation/renaturation procedures that were applied in our VirD1
purification protocol restored the protein's activities only
partially (indeed, the VirD1 preparation is fully active as an
accessory component in dsDNA border cleavage), any hypothetical
topoisomerase I activity of VirD1 would be anti-productive for
T-complex formation since the cleavage reaction requires negative
superhelical DNA. It is therefore extremely unlikely that the
topoisomerase activity observed by Ghai and Das(1989) originates from
VirD1. It is rather VirD2 that possesses a topoisomerase I-like
activity (Pansegrau et al., 1993a). The reaction that is
catalyzed by VirD2 on ssDNA is a cleaving-joining reaction that reaches
an equilibrium when approximately 35-40% of the substrate exist
in the cleaved form (Pansegrau et al., 1993a). The observation
that only 35-40% of the input form I plasmid DNA can be converted
to form II by the combination VirD1/VirD2, even when the proteins are
added in great molar excess, parallels this result and suggests that
the reaction on dsDNA is an equilibrium reaction too. However,
spontaneous relaxation of supercoiled plasmid DNA containing T-borders
in the presence of VirD1 and VirD2 has not been observed, indicating
that the cleaved plasmid DNA species might be topologically constrained
by strong protein-DNA interactions. Indeed, this has been shown for the
analogous TraJ/TraI system encoded by the conjugative plasmid RP4 (videinfra). Application of a TraI mutant
that is impaired in the interaction with the oriT nick region
(TraI S74A) resulted in spontaneous release of plasmid topoisomers
demonstrating continuous cleaving-joining (Pansegrau et al.,
1994b). As in the VirD1/VirD2 system, cleaved reaction products could
be captured only upon addition of protein-denaturing agents like SDS or
proteases. The reactions described for VirD1/VirD2 strongly resemble
those catalyzed by relaxosomes of the IncP The
operons encoding the components required for initiation of transfer DNA
replication in the Ti and for relaxosome formation in the IncP system
are organized in a strikingly similar manner, and the amino acid
sequences of several gene products are similar too, indicating that
both systems have evolved from a common ancestor (Pansegrau et
al., 1994a; Ziegelin et al., 1991). Formation of stable
relaxosomes at T-border sequences has not yet been observed (not
shown). However, this is conceivable since relaxosome assembly in the
IncP systems requires a third polypeptide component: the chaperone-like
TraH protein, which is required for stabilization of the nucleoprotein
complex assembled at oriT (Pansegrau et al., 1990a).
A possible candidate for a TraH-analogue in the Ti system is the
product of the virD3 gene that is, like TraH, non-essential
for DNA transfer (Kado and Lin, 1993; Vogel and Das, 1992b). The virD3 gene is located downstream of virD2 at a
position that corresponds to that of traH in the IncP system
(Pansegrau et al., 1994a).
Volume 270,
Number 3,
Issue of January 20, 1995 pp. 1269-1276
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
PURIFIED PROTEINS VirD1 AND VirD2 CATALYZE SITE- AND
STRAND-SPECIFIC CLEAVAGE OF SUPERHELICAL T-BORDER DNA IN VITRO(*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
ions. Relaxed DNA is not cleaved. The 5` terminus of the broken
DNA strand is covalently associated with protein, most probably VirD2,
and the cleavage site is located at the same position that is found in vivo, indicating that the in vitro reaction mimics
the one that takes place in induced agrobacteria. Relaxation of plasmid
DNA occurs only upon addition of protein denaturants, suggesting that
the DNA in the VirD1/VirD2 complex is topologically constrained by
strong protein-DNA interactions. The characteristics of the
VirD1/VirD2-mediated cleavage reaction strongly resemble those observed
with relaxosomes of IncP plasmids involved in initiation of transfer
DNA replication during bacterial conjugation.
)tumor-inducing (Ti) plasmid, the T-DNA, is transferred and
stably integrated into the plant genome (Winans, 1992; Zambryski,
1992). Expression of the transferred DNA (20-30 kb) in the
plant cell leads to autonomous production of the plant hormones auxin
and cytokinin, resulting in tumor formation. Furthermore, the T-DNA
directs the synthesis of specific amino acid derivatives (opines) that
are exported by the transformed plant cell allowing A. tumefaciens to utilize these opines as the sole carbon and nitrogen source for
growth (Zambryski et al., 1989).
30 kb)
of the Ti plasmid. Expression of the DNA transfer machinery specified
by this region is stringently regulated. vir gene expression
is induced by chemical signal molecules (phenolics and sugars) released
from wounded plant cells (Winans, 1992). Recognition of these signals
by the sensor protein VirA leads to phosphorylation of VirG protein,
upon which VirG activates expression of the other vir operons (virB, virC, virD, virE, and virF) (Jin et al., 1990; Stachel and Zambryski,
1986). DNA processing involves (i) recognition and complexing of 25-bp
direct repeats flanking the T-DNA, the T-border sequences, by virD gene products (Filichkin and Gelvin, 1993; Jayaswal et
al., 1987; Yanofsky et al., 1986); (ii) single-stranded
incisions within the T-borders leading to covalent attachment of the
VirD2 protein to the 5` terminus of the T-DNA
(Dürrenberger et al., 1989; Howard et
al., 1989; Pansegrau et al., 1993a; Young and Nester,
1988); (iii) DNA transport requiring displacement of the T-DNA and its
protection by coating with the ssDNA-binding protein VirE2 (Gietl et al., 1987; Christie et al., 1988; Citovsky et
al., 1989). The resulting nucleoprotein complex, the T-complex, is
exported to the plant cell, probably through a transmembrane pore
consisting of gene products specified by the virB operon
(Kado, 1994; Kuldau et al., 1990; Lessl and Lanka, 1994;
Shirasu et al., 1990; Thompson et al., 1988; Ward et al., 1988). Inside the plant cell, targeting of the
T-complex into the nucleus is directed by nuclear localization signals
present in VirD2 (Citovsky et al., 1994; Rossi et
al., 1993; Tinland et al., 1992) and VirE2 (Citovsky et al., 1992, 1994) using the nuclear import machinery of the
plant. Insertion of the T-DNA in one of the host's chromosomes
might involve catalysis by the 5`-terminal attached VirD2 protein
(Gheysen et al., 1991; Mayerhofer et al., 1991;
Pansegrau et al., 1993a).
)
)and IncN plasmids (Pohlman et al., 1993) and the
Ptl operon of Bordetella pertussis mediating export of the
pertussis toxin (Weiss et al., 1993). However, the closest
relationship exists between the Ti system and IncP plasmids because
both the DNA processing and the DNA transport functions are similar in
organization and primary structure. The relationship to IncN and IncW
plasmids appears to be limited mainly to the DNA transport system.
These findings suggest that in the course of the evolution of the
different DNA transfer systems modules of different origin have been
combined and adapted to form functional conjugation machineries or, in
one case, a toxin transport system. ion-dependent and occur
at the same position, relative to the consensus nick region. Cleavage
results in formation of a covalent VirD2/TraI-oligonucleotide adduct in
which the protein is attached to the 5`-terminal nucleotide at the nick
site. Peptide mapping of such VirD2-oligonucleotide adducts identified
tyrosine 29 as the residue that forms a phosphodiester with the
nucleotide at the 5` end (Pansegrau et al., 1993a). A
mutational analysis of the virD2 gene had also shown that
tyrosine 29 is the only tyrosine residue within VirD2 that cannot be
replaced without loss of activity (Vogel and Das, 1992a). In agreement
with previous studies that assigned the attachment site of the IncP
relaxase TraI to tyrosine 22 (Pansegrau et al., 1993b),
tyrosine 29 of VirD2 is located within the conserved relaxase motif I
(Ilyina and Koonin, 1992; Koonin and Ilyina, 1993; Pansegrau et
al., 1994a, 1994b).
Bacterial Strains, Plasmids, and Medium
E.
coli SCS1 (endA1, hsdR17(r
-,
m-), supE44, thi-1, recA1, gyrA96, relA1), a
DH1 derivative (Hanahan, 1983), was used as host for plasmids. Plasmids
pMS119HE (Balzer et al., 1992) and pBR329 have been described
previously (Covarrubias and Bolivar, 1982). Plasmids pT7-7 (Sano and
Cantor, 1990) and pVir97.89 (Alt-Mörbe and
Schröder, 1986), were kindly provided by Stanley
Tabor (Harvard Medical School) and Juliane
Alt-Mörbe (Institut für
Biologie III, Universität Freiburg, Freiburg,
Germany), respectively. Substrates for T-border cleavage were
constructed by insertion of synthetic complementary oligonucleotides in
pBR329: right border of pTiC58,
d(GATCCTCTTAGGTTTACCCGCCAATATATCCTGTCAAACACTG) and
d(TCGACAGTGTTTGACAGGATATATTGGCGGGTAAACCTAAGAG); left border of pTiC58,
d(GCAATTTGTTTACACCACAATATATCCTG
CCACCAGCCGACGT) and
d(CGGCTGGTGGCAGGATATATTGTGGTGTAAACAAATTGCTGCA) (the specific nick
positions are indicated by an arrow); overdrive of pTiA6:
d(CTAGCATCAAACAAACATACACAGCGACTTATTCACACGG) and
d(GATCCCGTGTGAATAAGTCGCTGTGTATGTTTGTTTGATG). Substrates were
constructed as follows: pPS100, pBR329
(BamHI-SalI, right T-border of pTiC58, 43 bp);
pPS101, pPS100
(PstI-AatII, left T-border of
pTiC58, 43 bp); pPS110, pPS100
(NheI-BamHI,
overdrive of pTiA6, 40 bp); pPS111, pPS101
(NheI-BamHI, overdrive of pTiA6, 40 bp). Cells were
grown in YT medium (Miller, 1972) supplemented with 25 mM MOPS
(sodium salt; pH 8.0), 0.1% glucose, 0.1% casamino acids, and 25
µg/ml thiamine hydrochloride. When appropriate ampicillin, sodium
salt, was added (100 µg/ml).
DNA Methodology
Standard molecular cloning methods
were performed as described (Sambrook et al., 1989).
Substrates for T-DNA cleavage and plasmid DNA sequencing were purified
by ethidium bromide-CsCl gradients. Following treatment with alkali
(0.2 M NaOH), DNA sequencing was performed using T7
Sequenase (Pharmacia) and specific synthetic
oligonucleotides 18-25 nucleotides in length as primers. Primer
extension was performed as described (Pansegrau et al.,
1990b).
Site-directed Mutagenesis
Rare Arg codons in virD1 were changed by site-specific oligonucleotide-directed
mutagenesis as described by Sayers et al.(1988).
Oligonucleotide primer applied for mutagenesis was:
d(GTATCACGCCGGCGTATCGAC). The resulting derivatives were examined by
nucleotide sequencing.Proteins
VirD2 protein was purified as described
(Pansegrau et al., 1993a). Rabbit antiserum to the VirD1
protein was raised against a TraL/VirD1 fusion protein purified by
preparative gel electrophoresis. The fusion protein derived from a gene
fusion constructed in the expression plasmid pGZ161-2 (Ziegelin,
1989). E. coli topoisomerase I was generously provided by
Eberhard Scherzinger (Max-Planck-Institut für
Molekulare Genetik, Berlin). Calf thymus topoisomerase I was purchased
from MBI Fermentas.Solid Phase Immunoassay
Proteins were visualized
by immunoblotting as described previously (Towbin et al.,
1979). For detection anti-TraL/VirD1 serum (1:500, 2 h) and
dichlorotriazinyl aminofluorescein-conjugated goat anti-rabbit IgGs
(Jackson ImmunoResearch; 1:100) were used. Fluorescence signals were
quantified with the FluorImager 575 (Molecular Dynamics).In Vitro Reconstitution of T-border Cleavage
Under
standard conditions, mixtures of VirD proteins and form I plasmid DNA
(0.7 µg) were incubated in a total volume of 20 µl (20 mM Tris-HCl, pH 8.8, 50 mM NaCl, 5 mM MgCl, 0.1 mM EDTA) for 45 min at 37 °C.
Reactions were stopped by addition of EDTA, SDS, and proteinase K to
concentrations of 25 mM, 1% (w/v), and 100 µg/ml,
respectively. Following incubation at 37 °C for 10 min, samples
were analyzed electrophoretically. The reaction products were
quantified by scanning ethidium bromide-stained agarose gels with the
FluorImager 575 (Molecular Dynamics). Since form I and form II DNA give
different specific fluorescence intensities, the yield of specifically
cleaved plasmid DNA was calculated from the decrease of the
fluorescence signal of the form I relative to the untreated substrate
DNA.
Overproduction of VirD1
Previous attempts in
expression vector cloning of virD1 resulted in rather low
yields of protein and additional peptides of unexpected size were
observed (De Vos and Zambryski, 1989; Jasper et al., 1994). A
careful inspection of the nucleotide sequence revealed features that
could affect virD1 expression. The initiation codon of virD1 (pTiC58) was proposed to be an AUG located at nucleotide
positions 22-24 (here referred to as start 2, Fig. 1)
(Yanofsky et al., 1986). An additional potential AUG start
codon (start 1) in the same reading frame is located 12 bp upstream (Fig. 1). Potential initiation codons at this position are
conserved in at least two additional Ti-type plasmids: pTiA6NC (UUG)
(Jayaswal et al., 1987; GenBank accession number M17989) and
pRiA4b (AUG) (Hirayama et al., 1988; GenBank accession number
X12867). Furthermore, a cluster of three rare CGG Arg codons was
noticed 40 bp upstream of the stop codon. Consecutive rare codons are
known to stimulate frameshifting or premature termination (Rosenberg et al., 1993). We constructed a set of virD1-overproducing plasmids to study the effect of these two
features on gene expression, to determine the translation initiation
site, and finally to obtain amounts of VirD1 protein suitable for
biochemical analysis.
10/gene 10 S/D expression
plasmid pT7-7. The 5` end of the gene was restored using synthetic
oligodeoxyribonucleotides (printed in lowercaseletters) to link the SacI cohesive end of the virD1 fragment to the NdeI-end of the vector. The
oligonucleotides carried codons from the MaeIII site to start
1 or start 2, respectively. Nucleotides of rare CGG Arg codons 40 bp
upstream of the stop codon (nucleotides 1048-1056 of Atuvird,
GenBank accession number M33673) indicated by arrows were
changed by site-directed mutagenesis. To place the manipulated virD1 gene under the control of the chemically inducible tac promoter a XbaI-SacI fragment (533 bp),
carrying the manipulated gene, together with the T7 gene 10 S/D sequence, was inserted in the multi-cloning site of pMS119HE.
Resulting overexpression plasmids (Table 1) are: pPS20, carrying
alternative Arg codons; pPS13, carrying wild-type Arg codons; pPS22,
carrying alternative Arg codons and start 1; pPS15, carrying wild-type
Arg codons and start 1. A plasmid encoding the gene under its original
translational control was obtained inserting a 530-bp HindIII-SacI fragment from pVir97.89 in the
multicloning site of pMS119HE, resulting in pPS9. The amino acid
sequence of VirD1 is shown below the nucleotide sequence. The
underlined part was confirmed by N-terminal microsequencing.
Polypeptides resulting from initiation of translation at start 1
yielded the N-terminal sequence
M-E-E-A-M-S-Q-G-S-R.
units of cells); lane b, pPS13 (75 A
); lane c,
pPS15 (25 A
); lane d, pPS20 (51 A
); lane e, pPS22 (17 A
); lane f, pMS119HE (75 A
); lane g, molecular mass standards
bovine serum albumin (BSA, 68 kDa) and lysozyme (LYS,
14.3 kDa) labeled with fluorescein
isothiocyanate.
Purification of VirD1
The protein was purified by
a three-step procedure to near homogeneity (Fig. 3, lanes
c-e; Table 2). Following induction of SCS1(pPS20) with
IPTG, VirD1 was found in inclusion bodies. After extensive washing with
3 M urea of the insoluble pellet formed after lysis and
centrifugation, VirD1 was solubilized in buffer containing 8 M urea, 10 mM DTT. The protein was renatured by
(NH
)SO precipitation and stepwise
removal of urea. Aggregated molecules that were not properly refolded
and contaminating proteins were separated by chromatography on
DEAE-Sephacel and phosphocellulose P11. The recovery of this procedure
was 8% of approximately 96% pure VirD1. The existence of VirD1 dimers
was shown by chemical cross-linking with glutaraldehyde as described
elsewhere (Balzer et al., 1992; data not shown).
VirD1 and VirD2 Specifically Cleave T-border Plasmid Form
I DNA
In vivo data suggest that T-border cleavage
during the initiation of T-DNA transfer is catalyzed by VirD1 and VirD2
(Alt-Mörbe and Schröder, 1986;
Yanofsky et al., 1986). Purified VirD2 protein was shown to
bear site-specific cleaving and joining activity on ssDNA, but failed
to cleave form I, form II, or form III DNA (Pansegrau et al.,
1993a). Therefore, we incubated combinations of purified VirD proteins
with plasmid DNA containing T-border sequences. Following digestion
with proteinase K, the reaction products were analyzed on 0.7% agarose
gels (Fig. 4; Table 3). Neither VirD1 nor VirD2 alone
mediated conversion of T-border plasmid DNA from form I to form II (Fig. 4, lanes b and c; Table 3). VirD2
alone failed to cleave form I T-border plasmid DNA even when added in
200-fold molar excess (data not shown). If both proteins were present,
35-40% form I DNA of a plasmid containing a unique right border
sequence of pTiC58 was converted to form II (Fig. 4, lane
e; Table 3). In analogy to the cleavage of ssDNA by VirD2,
the reaction requires Mg ions (Fig. 4, lane d; Table 3). Substrate DNA containing one left and
one right T-border was converted to 70-75% into form II (Fig. 4, lanes f and g). Since vector DNA
under these conditions did not yield significant amounts of form II,
cleavage is T-border-specific (Fig. 4, lanes h and i; Table 3). The presence of the overdrive sequence in
plasmids pPS110 or pPS111 had no significant influence on the yield of
cleaved products (not shown).
omitted; e, 500 ng of VirD1 and
60 ng of VirD2; lanes f and g: pPS101 (right and left
border, 0.7 µg of DNA); f, proteins omitted; g,
500 ng of VirD1 and 60 ng of VirD2; lanes h and i:
pBR329 (0.7 µg of DNA); h, proteins omitted; i,
500 ng of VirD1 and 60 ng of VirD2. B, schematic diagram of
design of T-border plasmids. Synthetic oligonucleotides symbolized by boxes were inserted in resistance genes of plasmid pBR329
(shown as arrows). Base pair coordinates of restriction sites
in pBR329(4151) are BamHI (606), SalI (882), PstI (2755), and AatII (3432) (Covarrubias and
Bolivar, 1982). The corresponding size of the substrates is given in
base pairs.
The in Vitro Cleavage Reaction by VirD1 and VirD2 Is
Strand- and Site-specific
To verify the specificity of the
cleavage reaction, products (form II DNA) were linearized with a
restriction enzyme and analyzed on denaturing alkaline agarose gels (Fig. 5). Three distinct bands (A/A`, B/B`, C/C`; Fig. 5, lanes AatII and XmnI, respectively) were observed,
one corresponding to the full-length linear form of the plasmid (A/A`).
The sizes of the two additional fragments with higher mobility indicate
cleavage in the T-border sequence. The sum of the fragment sizes
matches that of one full-length linear plasmid, demonstrating that
scission does occur exclusively at the T-border sequence and nowhere
else on the plasmid. Radioactive labeling of the 5` termini only
resulted in phosphorylation of fragments A and C (data not shown). The
fragment C starting with the 5`-terminal nucleotide of the nick site
was blocked, suggesting covalent association with protein. This protein
most likely is VirD2, since it was shown to be attached to the T-strand in vivo and in vitro following cleavage of T-border
sequences (Pansegrau et al., 1993a; reviewed by Winans, 1992).
The in Vitro Reaction Catalyzed by VirD1 and VirD2 Mimics
T-border Cleavage in Vivo
Following induction of vir genes, scission of a single strand in the right border sequence
initiates T-DNA transfer to plant cells. The T-strand displacement is
thought to proceed from this nick with the VirD2 protein covalently
bound to the 5` end of the strand (reviewed by Zambryski(1992)). The
purified VirD1 and VirD2 proteins catalyze a border-specific cleavage
reaction in vitro, resulting in a product covalently modified
at its 5` terminus. The position of the nick produced in vitro was determined by primer extension to demonstrate further identity
with the reaction taking place in vivo (Fig. 6). These
data show that the cleavage site coincides with that determined in
vitro by VirD2-mediated oligonucleotide cleavage (Pansegrau et
al., 1993a) and deduced from cleaved intermediates obtained from E. coli cells (Dürrenberger et
al., 1989).
ions. dsDNA cleavage results in covalent
attachment of a protein (most likely VirD2) to the DNA 5` terminus as
it has been shown for T-DNA single strands that were isolated from
bacterial cells (Dürrenberger et al.,
1989; Howard et al., 1989; Young and Nester, 1988). The
position of the cleaved phosphodiester bond is the same that is found in vivo (Dürrenberger et al.,
1989), indicating that the in vitro reaction described here is
mimicking that occurring in agrobacteria.
plasmid RP4. The RP4
relaxase (TraI) is analogous to VirD2, specifically recognizing
single-stranded substrates containing the nick region of the IncP
transfer origin (oriT) (Pansegrau et al.,
1993b). The corresponding reaction with dsDNA requires the accessory
protein TraJ that binds specifically to a target sequence adjacent to
the nick region (Pansegrau et al., 1990a; Ziegelin et
al., 1989). Formation of the TraJ
oriT nucleoprotein
complex is the first step in a cascade that culminates in the formation
of stable relaxosomes (Pansegrau et al., 1990a). It is
tempting to speculate that TraJ and VirD1 are functional analogs,
although there is only little similarity in their amino acid sequences
(Llosa et al., 1994; Pansegrau et al., 1994a). The
low degree of amino acid sequence similarity might also reflect
different recognition specificities, concerning the target DNA as well
as the respective relaxases with which the proteins interact.
)
-D-galactopyranoside; bp, base pair(s);
DTT, dithiothreitol.))
We are grateful to Heinz Schuster for generous
support. We greatly appreciate the expert technical assistance of
Marianne Schlicht. We thank Werner Schröder for
synthesizing oligonucleotides and sequencing VirD1 N termini and Karl
Rak for preparing the VirD1 antiserum.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Lacroix, A. Loyter, and V. Citovsky Association of the Agrobacterium T-DNA-protein complex with plant nucleosomes PNAS, October 7, 2008; 105(40): 15429 - 15434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Rommens, O. Bougri, H. Yan, J. M. Humara, J. Owen, K. Swords, and J. Ye Plant-Derived Transfer DNAs Plant Physiology, November 1, 2005; 139(3): 1338 - 1349. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Jakubowski, E. Cascales, V. Krishnamoorthy, and P. J. Christie Agrobacterium tumefaciens VirB9, an Outer-Membrane-Associated Component of a Type IV Secretion System, Regulates Substrate Selection and T-Pilus Biogenesis J. Bacteriol., May 15, 2005; 187(10): 3486 - 3495. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Abu-Arish, D. Frenkiel-Krispin, T. Fricke, T. Tzfira, V. Citovsky, S. G. Wolf, and M. Elbaum Three-dimensional Reconstruction of Agrobacterium VirE2 Protein with Single-stranded DNA J. Biol. Chem., June 11, 2004; 279(24): 25359 - 25363. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Caryl, M. C. A. Smith, and C. D. Thomas Reconstitution of a Staphylococcal Plasmid-Protein Relaxation Complex In Vitro J. Bacteriol., June 1, 2004; 186(11): 3374 - 3383. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Tzfira, L. R. Frankman, M. Vaidya, and V. Citovsky Site-Specific Integration of Agrobacterium tumefaciens T-DNA via Double-Stranded Intermediates Plant Physiology, November 1, 2003; 133(3): 1011 - 1023. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Bako, M. Umeda, A. F. Tiburcio, J. Schell, and C. Koncz The VirD2 pilot protein of Agrobacterium-transferred DNA interacts with the TATA box-binding protein and a nuclear protein kinase in plants PNAS, August 19, 2003; 100(17): 10108 - 10113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Gelvin Agrobacterium-Mediated Plant Transformation: the Biology behind the "Gene-Jockeying" Tool Microbiol. Mol. Biol. Rev., March 1, 2003; 67(1): 16 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Y. Szpirer, M. Faelen, and M. Couturier Mobilization Function of the pBHR1 Plasmid, a Derivative of the Broad-Host-Range Plasmid pBBR1 J. Bacteriol., March 15, 2001; 183(6): 2101 - 2110. [Abstract] [Full Text]
|
|
|
|
|
|
J. Zhu, P. M. Oger, B. Schrammeijer, P. J. J. Hooykaas, S. K. Farrand, and S. C. Winans The Bases of Crown Gall Tumorigenesis J. Bacteriol., July 15, 2000; 182(14): 3885 - 3895. [Full Text]
|
|
|
|
|
|
B. Relić, L. Rossi, Y. Nagamine, and B. Hohn Interaction of the DNA modifying proteins VirD1 and VirD2 of Agrobacterium tumefaciens: Analysis by subcellular localization in mammalian cells PNAS, August 4, 1998; 95(16): 9105 - 9110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Hansen, R. D. Shillito, and M.-D. Chilton T-strand integration in maize protoplasts after codelivery of a T-DNA substrate and virulence genes PNAS, October 14, 1997; 94(21): 11726 - 11730. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Hansen and M.-D. Chilton "Agrolistic" transformation of plant cells: Integration of T-strands generated in planta PNAS, December 10, 1996; 93(25): 14978 - 14983. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |