Vol. 174, No. 12
Constitutive Mutations ofAgrobacterium tumefaciens
Transcriptional Activator virG
GREGORY J. PAZOUR,1t CHRISTOPHER N. TA,1 AND ANATH DASl12*
Department ofBiochemistry' and Plant Molecular Genetics Institute,2 University of
Minnesota, 1479 GortnerAvenue, St. Paul, Minnesota 55108
Received 4 December 1991/Accepted 21 April 1992
The virulence (vir) genes of Agrobacterium tumefaciens Ti plasmids are positively regulated by virG in
conjunction with virA and plant-derived inducing molecules. A procedure that utilizes both genetic selection
and a genetic screen was developed to isolate mutations in virG that led to elevated levels of vir gene expression
in the absence of virA and plant phenolic inducers. Mutants were isolated at a frequency of 1 in 107 to 108.
Substitution mutations at two positions in the virG coding region were found to result in the desired phenotype.
One mutant had an asparagine-to-aspartic acid substitution at residue 54, and the other contained an
isoleucine-to-leucine substitution at residue 106. In both cases, the mutant phenotype required the presence of
the active-site aspartic acid residue at position 52. Further analysis showed that no other substitution at residue
54 resulted in a constitutive phenotype. In contrast, several substitutions at residue 106 led to a constitutive
phenotype. The possible roles of the residues at positions 54 and 106 in VirG function are discussed.
The virulence (vir) genes of the plant pathogen Agrobac-
terium tumefaciens catalyze the transfer of a segment of Ti
plasmid-borne DNA to plant cells, leading to crown gall
tumor disease (reviewed in references 26 and 43). The vir
region encompasses about 35 kb ofDNA and is composed of
eight operons, virA to virH (7, 12, 17, 22, 31). Of these, virA
and virG, in conjunction with plant phenolics such as aceto-
syringone (AS), positively control vir gene expression (32,
33, 42). A chromosomal gene, chvE, is also required for
efficient induction of the vir genes (8). DNA sequence
analysis indicated that VirA and VirG, the polypeptide
products of virA and virG, respectively, are members of the
bacterial two-component regulatory system family (15, 41).
In two-component regulatory systems, one component, the
sensor, senses an environmental stimulus and transmits a
signal to the second component, the regulator, which then
controls a cellular function. The signal transduction process
is mediated through protein phosphorylation. In response to
a stimulus, the sensor protein is autophosphorylated at a
conserved histidine residue. This phosphate moiety is then
transferred to a conserved aspartic acid residue of the
regulator (reviewed in references 27 and 35). It is believed
that in the Agrobacterium vir system, VirA, a transmem-
brane protein, functions as the sensor and VirG, a cytosolic
protein, functions as the regulator. VirG is a sequence-
specific DNA-binding protein that binds at conserved virbox
sequences (11, 23) and is thought to function as a transcrip-
tional activator when chemically modified by phosphory-
To understand the signal transduction process, attempts
were made to isolate mutations in virA and virG by using a
genetic screen (25). All of the mutations identified in that
study mapped to the virA locus, although this approach
theoretically should have yielded mutations in virG as well.
With the assumption that mutations in virG are rare, we used
a combination of genetic selection and a genetic screen to
t Present address: Worcester Foundation for Experimental Biol-
ogy, Shrewsbury, MA 01545.
isolate virG mutants. In this report, we describe the effec-
tiveness of such a dual approach and an analysis of critical
residues of the regulator component of a two-component
Isolation and characterization of constitutive virG mutants.
Plasmid pGP358R (kanamycin resistant), which contains
virG and virB gene fusions with the ,B-lactamase (bla) and
1-galactosidase (3-gal) (lacZ) genes, was constructed as
follows. The kanamycin
pUC4K was first cloned as a PstI fragment into plasmid
pUC119 (36). An EcoRI site was created between the
ribosome-binding site sequence (30) and the translational
start codon of the 1-lactamase gene (1 residue upstream of
the ATG codon) by deoxyoligonucleotide-directed site-spe-
cific mutagenesis (14). The resultant plasmid was digested
with EcoRI and filled in with T4 DNA polymerase and
deoxynucleoside triphosphates. The 3.3-kb fragment con-
taining the promoterless bla gene was ligated to a 330-bp
SmaI fragment of plasmid pAD1221 (24). The SmaI fragment
contains the promoter-regulatory region and part of the first
open reading frame of theAgrobacterium virB locus. A clone
that contained virB in the same orientation
structural gene was isolated and designated pGP357. Plasmid
pGP357 was cloned as an EcoRI fragment (the SmaI frag-
ment containing virB sequences contains an EcoRI site
upstream of the virB promoter region) into the unique EcoRI
site of plasmid pGP220 to construct pGP358R. Plasmid
pGP220 contains virG and a virB-lacZ gene fusion on wide-
host-range vector pTJS75 (29). The virB-bla fusion was
constructed such that the bla gene lacks its native ribosome-
binding site sequence. The DNA sequence of the junction
region reads dgggatccccAAlTCTAlX; the lowercase letters
indicate virB sequences, and the underlined sequences indi-
cate the translational initiation codon for the bla gene. It was
necessary to construct this fusion because earlier experi-
ments showed that a virB-bla fusion that retained the bla
ribosome-binding site sequence can confer resistance to a
high level of carbenicillin, even in the absence of an inducer
(data not shown). We reasoned that the fusion in pGP358R
would be translated less efficiently because the bla gene
segment lacked its native ribosome-binding site sequence;
gene from plasmid
as the bla
JOURNAL OF BACTERIOLOGY, June 1992, p. 4169-4174
Copyright © 1992, American Society for Microbiology
therefore, increased transcription would be required to com-
pensate for the lower translation efficiency and, conse-
quently, for antibiotic resistance. A constitutive virG that is
independent of virA and plant signal molecules is expected to
induce transcription of the virB promoters in pGP358R,
leading to formation of carbenicillin-resistant (from virB-bla)
blue (from virB-lacZ) colonies on solid medium containing
To determine whether plasmid pGP358R can be used for
selection of mutants with increased vir gene expression, this
plasmid was introduced into Agrobacterium strains A136
(which lacks a Ti plasmid) and A348 (which contains oc-
topine Ti plasmid pTiA6). Under normal growth conditions,
both strains were sensitive to carbenicillin. To determine
whether vir gene-inducing conditions confer carbenicillin
resistance, 106 Agrobacterium strain A348/pGP358R bacte-
ria were plated on AB Mes (pH 5.5) solid medium (23) with
or without 100 ,uM AS. A filter paper disc impregnated with
300 ,ug of carbenicillin was placed in the center of the plate,
and cells were allowed to grow at 30°C. In the absence of
AS, the zone of inhibition was approximately 6 cm in
diameter, while that in the presence ofAS was only 0.2 to 0.6
cm (data not shown). These results indicated that this gene
fusion could be used to select mutants with elevated vir gene
Agrobacterium strain A136/pGP358R was mutagenized
with nitrous acid or nitrosoguanidine (19). For nitrosoguani-
dine mutagenesis, cells were grown at 30°C in AB Mes (pH
5.5) liquid medium to an A600 of -0.2. Nitrosoguanidine was
then added at a final concentration of 0.13 ,ug/ml, and cell
growth was continued for an additional 60 min. To stop the
mutagenesis, cells were collected by centrifugation and
washed with AB (pH 7.0) liquid medium. After resuspen-
sion, cells were plated on AB Mes (pH 5.5) solid medium
side (16 ,ug/ml) and carbenicillin (50 to 200 ,ug/ml). Nitrous
acid mutagenesis was performed essentially as described by
Miller (19), except that cells were not allowed to grow before
plating. In both mutagenesis procedures, 75 to 80% killing of
the cells was observed.
Carbenicillin-resistant cells that formed blue colonies on
ing plates in the absence of AS were picked, purified, and
resistant colonies appeared at a frequency of approximately
1 in 105 to 106 cells, of which only 1% turned blue in the
presence of 5-bromo-4-chloro-3-indolyl-3-D-galactopyrano-
side. By this combination of positive selection and screen-
ing, 43 Agrobacterium strains with elevated levels of virB
expression were isolated. Eight of these strains resulted
from nitrosoguanidine mutagenesis, and the remainder were
from nitrous acid mutagenesis. In liquid assays, the mutant
strains showed 200 to 1,000 U of ,B-gal activity. The virG
genes from five of these mutants were cloned into plasmid
pUC119, and the complete DNA sequence of the virG open
reading frame was determined by the dideoxy-chain termi-
nation method of Sanger et al. (28) by using a double-
stranded DNA template, a series of synthetic deoxyoligonu-
cleotide primers, and Sequenase (U.S. Biochemical Corp.,
Cleveland, Ohio). All mutants were found to contain a single
base change (an A-to-G change at position 432, numbered as
described by Winans et al. ). This transition creates a
new restriction endonuclease site for XbaI (TCTA-GA) and
leads to substitution of an asparagine residue at position 54
with an aspartic acid (N54D; amino acids were numbered as
described by Pazour and Das ; this mutation has also
3-gal activity in liquid medium. Carbenicillin-
FIG. 1. Effects of virG and its mutations on virB-lacZ expres-
sion. The N54D or 1106L mutation was introduced into the virG
gene of plasmid pGP109 (23) by deoxyoligonucleotide-directed
site-specific mutagenesis as described by Kunkel (14). Plasmid
pAD1092K (5) at its respective EcoRI site to construct pGP229 and
its derivatives. Plasmid pAD1092K contains a virB-lacZ gene fusion
on wide-host-range plasmid vector pTJS75 (29). pGP229 and its
derivatives were mobilized from Escherichia coli into Agrobacte-
rium strain A136 by triparental mating (6). Cells were grown
overnight in AB (pH 7) liquid medium, diluted 1:20 in AB Mes (pH
5.5) medium, and grown for various times as indicated. Procedures
for the ,B-gal activity assay were as described previously (23). WT,
wild-type virG; N54D, virG N54D; 1106L, virG I106L.
its mutant derivative was then fused
been isolated independently by S. Winans ). The rest of
the potential mutants were screened for this new restriction
site. Of the 43 mutants, 42 contained this site. DNA se-
quence analysis of the other mutant (obtained by nitrosogua-
nidine mutagenesis) indicated that it contains an A-to-C
change at position 585 which alters codon 106 from isoleu-
cine to leucine (I106L).
To confirm that the phenotypes observed were due to the
single base substitutions, these mutations were recreated by
(14). virG and its derivatives containing the N54D or 1106L
mutation were cloned into plasmid pAD1092K, which con-
tains a virB-lacZ reporter gene (5). After introduction of
these plasmids intoAgrobactenum strain A136, the resulting
strains were assayed for 13-gal activity in the absence of AS
(Fig. 1). The two strains with virG N54D or I106L showed a
significantly higher level of virB expression, even at time
zero (time of transfer of cells grown overnight in AB [pH 7]
medium into induction medium AB Mes [pH 5.5]), indicating
that the constitutive phenotype of the mutants is manifested
even at pH 7. In the virG mutant strains, virB expression
increased for about 8 h before reaching a plateau.
In a strain harboring wild-type virG, the level of virB
expression increased approximately fivefold during the 24-h
incubation period. The maximum level of virB expression in
the virG N54D mutant strain in the absence of the inducer
FIG. 2. Effect of pH on the phenotype of virG mutants. Proce-
dures and strains used were identical to those described in the
legend to Fig. 1, except that after overnight growth in AB (pH 7),
cells were diluted into AB Mes (pH 5.5 or 6.8), as indicated, and
grown for 24 h before assay. WT, wild type.
AS was found to be approximately fourfold higher than the
fully induced level (in the presence of virA and AS) in strains
containing wild-type virG, indicating that the vir genes are
not maximally expressed under standard induction condi-
tions. The level of virB expression observed with virG N54D
and I106L was greater than that observed with virG from the
supervirulent Ti plasmid pTiBo542 ofAgrobacterium strain
A281. The virG gene of pTiBo542 is believed to be respon-
sible for the supervirulent phenotype of Agrobacterium
strain A281. In a recent study, Chen et al. (3) reported that
virG from pTiBo542 causes
expression compared with virG from pTiA6. Whether the
large increase in virB expression we observed in strains
bearing the virG mutations leads to an increase in the amount
of T-strand DNA, and consequently to increased virulence,
remains to be seen.
Effect of pH on the virG constitutive phenotype. To study
the effect of pH on vir gene expression, strains containing
virB-lacZ and virG or virG mutations were grown in AB Mes
(pH 5.5 or 6.8) liquid medium and assayed for p-gal activity.
Both virG mutant strains exhibited an approximately four-
fold decrease in virB expression at pH 6.8 (Fig. 2). The basal
level of expression in the strain containing wild-type virG
was reduced twofold at the higher pH. These results indicate
that the transcriptional activator function of virG is modu-
lated by an extracellular acidic environment.
An acidic environment is essential for maximal expression
of the vir genes. It has been thought that virA is the primary
sensor of the environment, since deletion of the periplasmic
domain of virA largely relieves the requirement for an acidic
environment (18, 25). However, the phenotype produced by
virG N54D and 1106L is affected by the pH of the growth
medium, even in the absence of virA, indicating that there is
an additional step(s) that is pH dependent. Extracellular pH
is not expected to have a major effect on intracellular pH,
but it is known to alter the intracellular concentration of ions
a 1.7-fold increase in virB
TABLE 1. Effects of virA and AS on virG mutant phenotypea
aPlasmids pGP109, pGP109N54D, and pGP109I106L were cloned into
pGP119 as an EcoRI fragment to construct pGP159, pGP396, and pGP404,
respectively. After mobilization of the plasmids into Agrobacterium strain
A136, the resulting strains were assayed for 1-gal activity as described in the
legend to Fig. 1. Where indicated, AS was added at a final concentration of
such as K+ (9). The changes in ion concentration may
directly affect VirG activity. Alternatively, there may be an
additional unidentified component(s) that senses extracellu-
lar pH and affects virG expression and/or activity. Expres-
sion of virG is regulated by several factors, including pH
(40). The pH effect is exerted through transcriptional activa-
tion of a second promoter. It is unlikely that the pH effect
observed in this study is due to transcriptional activation
from the pH-inducible promoter, because these mutants
should autocatalytically (by transcriptional activation of the
virA, virG, AS-dependent promoter) increase their cellular
concentration much more significantly than expected from
the induction of the pH-sensitive promoter.
Effects ofVirA and AS on virG mutants. To study the effect
of VirA and AS on the virG mutants, plasmids were con-
structed which contained virA, the reporter virB-lacZ gene
fusion, and virG or its mutant derivatives. After introduction
of these plasmids into Agrobacterium strain A136, the re-
sulting strains were assayed for p-gal activity (Table 1). Both
mutations led to a large increase in virB-lacZ expression in
the absence of the inducer (AS). The N54D mutation led to
a 1,756-fold increase in virB-lacZ expression, while the
I106L mutation caused a 96-fold increase. In the presence of
AS, no significant change in virB expression was observed in
strains bearing the virG N54D mutation; however, an addi-
tional 18-fold increase in virB expression was observed in the
virG 1106L mutant strain.
Other substitutions at position 54 or 106. To determine the
effects of other substitution at either the N54 or the I106
position, additional substitutions were created by site-spe-
cific mutagenesis. Several substitutions at the N54 position
(N--E, F, I, L, P, R, S, W, or Y) abolished virG activity,
while others (N--+G, H, K, M, or T) attenuated virG activity
(Fig. 3). However, no substitution other than aspartic acid
led to a constitutive phenotype. In contrast, no substitution
at the 1106 position abolished virG activity and several
substitutions (I106-+F, L, N, P, or Y) led to a constitutive
Effect of virA on virG 1106 substitutions. We investigated
whether the phenotype of the virGI106 substitutions identi-
fied in studies described in the preceding section is indepen-
dent of virA. Plasmids containing virB-lacZ and virG muta-
tions were constructed and introduced into Agrobacterium
strain A136, and the strains were assayed for P-gal activity
(Fig. 4, open bars). All of the mutants exhibited a constitu-
tive phenotype in the absence of virA and AS, indicating that
the mutants function independently of VirA. The level of
virB expression in strains bearing the virG 1106F, I106N, or
I106Y mutation did not differ significantly in the presence or
absence of virA (Fig. 4). In contrast, virB expression in
VOL. 174, 1992
WT LP S R K M
V G T A E H
Q D Y W C F
FIG. 3. Effects of amino acid substitutions at positions 54 and
106 of virG on virB-lacZ expression. Substitutions were introduced
at the desired position by deoxyoligonucleotide-directed site-spe-
cific mutagenesis of pGP109 DNA as described previously (25). The
mutagenic primer contained a random assortment of bases at codon
position 54 or 106. The identities of the mutations were determined
by DNA sequence analysis (28). The mutant pGP109 derivatives
were fused to pGP119, introduced intoAgrobacterium strain A136,
and assayed for ,-gal activity as described in the legend to Fig. 1.
Amino acid substitutions are indicated by the single-letter code.
Open bars, no AS; dark bars, with AS; WT, wild type.
strains bearing wild-type virG, virG I106L, and virG I106P
was higher in the absence of virA. The reason for this
difference is not apparent and is under investigation.
Effect ofalteration ofthe active-site aspartic acid residue. In
two-component regulatory systems, an aspartic acid residue
(D52 for VirG) of the regulator is phosphorylated by the
corresponding sensor protein (35). The phosphorylated reg-
ulator is believed to be the active form that controls cellular
functions. Phosphorylation presumably causes a structural
change which activates the protein. The mutants isolated in
this study can function by (i) stabilizing the aspartyl phos-
phate, (ii) locking the protein in an active configuration to
mimic the effect of phosphorylation, or (iii) changing a site(s)
distal to the site of phosphorylation, e.g., a site necessary for
interaction with RNA polymerase. To distinguish between
these possibilities, we sought to determine whether the
N54D and 1106L mutations in virG act independently of the
aspartic acid residue at the active site. We constructed
double mutants containing virG D52E and N54D or I106L
and analyzed the effects of these mutations on virB expres-
sion. Substitution of the aspartic acid residue at position 52
with glutamic acid abolished VirG activity (Table 2). The
loss of activity in virG D52E may be due to (i) the inability of
this residue to be phosphorylated or (ii) structural alteration
N54D 1106L 1106F 1106P 1106N 1106Y
FIG. 4. Effect ofvirA on virB-lacZ expression. The effect of virA
on virB expression was monitored in different virG backgrounds.
virG and its mutant derivatives were cloned into plasmid pAD1092K
(virA) or pGP119 (virA+) and introduced intoAgrobacterium strain
A136, and the resulting strains were assayed for n-gal activity as
described in the legend to Fig. 1. WT, wild type.
associated with this mutation. Introduction of either the
N54D or the 1106L mutation into virG D52E did not restore
its transcriptional activator function, indicating that the
aspartic acid residue at position 52 is required for manifes-
tation of the constitutive phenotype of the mutants. These
results indicate that the third possibility for how these
mutants act is unlikely.
Activating mutations have been isolated in several two-
component system regulators, including cheY, degU, ginG,
and ompR (2, 4, 20, 38). Examples of mutations which act by
each of the first two possibilities mentioned above have been
found. cheY D13K, like phosphorylated CheY, causes a
tumble phenotype, but this protein is not readily phosphor-
ylated in vitro, making it likely that this mutation allows the
protein to function independently of phosphorylation. A
mutation in ompR, ompR3, acts by stabilizing the aspartyl
phosphate (1). It seems more likely that the mutants de-
scribed in the present study act by mimicking the phosphor-
ylated state of the protein rather than by stabilizing the
aspartyl phosphate, because these mutants are virA indepen-
dent. Therefore, any phosphorylation ofVirG would have to
be through cross talk (21). The latter mechanism would also
TABLE 2. Effect of mutation of the active-site aspartic acid
residue on VirG activitya
aThe desired mutations in virG were introduced by site-specific mutagen-
esis. Other procedures were as described in the legend to Fig. 1.
require that the aspartyl phosphate of wild-type VirG be
labile. In vitro, the half-life of VirG aspartyl phosphate is
rather long (>1 h; reference 10); this is within the range of
stability expected from studies of model compounds such as
acetyl phosphate (13). However,
whether the half-life of VirG aspartyl phosphate is as long in
The crystal structure of a VirG homolog, CheY, has been
determined (34). Since the N-terminal half of VirG is homol-
ogous to CheY, it can be assumed that VirG is structurally
similar to CheY. On this basis, both N54D and 1106L lie very
close to active-site aspartic acid D52. The conserved lysine
residue at position 109 in CheY (102 in VirG) appears to be
important for intramolecular changes that accompany acti-
vation. Lukat et al. (16) demonstrated the requirement of
this residue for CheY function and proposed that it either
active-site aspartic acid (D54 of CheY). Those researchers
also proposed that phosphorylation of the aspartic acid
would disrupt the original interaction but could create a new
one between the phosphate group and lysine 109. Analysis of
CheY structure at a higher resolution show that the e-amino
group of lysine 109 is bound to one of the carboxyl oxygens
of the active-site aspartic acid (37). This lysine is additionally
hydrogen bonded to aspartic acid 12 (Asp-9 in VirG) and to
an oxygen atom on a sulfate ion. This sulfate ion is also
bound to the side chain of asparagine 59 (Asn-54 of VirG, the
site of the N54D mutation). Phosphorylation of the active-
site aspartic acid residue would lead to local changes and
rearrangement in this bonding network, leading to a confor-
mational change. The N54D substitution in virG would
clearly disrupt many of these interactions. The other substi-
tutions at residue 106 are very close to the conserved lysine
residue (at position 102) and may affect its orientation. The
wide range of amino acid substitutions (asparagine, leucine,
phenylalanine, proline, and tyrosine) that confer a constitu-
tive phenotype argues against a specific interaction mediated
by these substitutions. Instead, it appears more likely that an
interaction(s) is disrupted by these substitutions.
it remains to be seen
a salt bridge
a hydrogen bond to the
We thank Yong-Hong Xie for excellent technical assistance,
Steve Winans for sharing unpublished results with us, Bonnie Allen
for typing, and our colleagues for critical reading of the manuscript.
This work was supported by a grant from the National Institutes
of Health (GM 37555), an American Cancer Society Faculty Re-
search Award (FRA 386), and a University of Minnesota Graduate
School Dissertation Fellowship (to G.P.).
1. Aiba, H., F. Nakasai, S. Mizushima, and T. Mizuno. 1989.
Evidence for the physiological importance of the phosphotrans-
fer between the two regulatory components, EnvZ and OmpR,
in osmoregulation in Escherichia coli. J. Biol. Chem. 264:14090-
2. Bourret, R. B., J. F. Hess, and M. I. Simon. 1990. Conserved
aspartic acid residues and phosphorylation in signal transduc-
tion by the chemotaxis protein CheY. Proc. Natl. Acad. Sci.
3. Chen, C.-Y., L. Wang, and S. C. Winans. 1991. Characterization
of the supervirulent virG gene of the Agrobacterium tume-
faciens plasmid pTiBoS42. Mol. Gen. Genet. 230:302-309.
4. Dahl, M. K., T. Msadek, F. Kunst, and G. Rapoport. 1991.
Mutational analysis of the Bacillus subtilis DegU regulator and
its phosphorylation by the DegS protein kinase. J. Bacteriol.
5. Das, A., and G. Pazour. 1989. Delineation of the regulatory
region sequences of Agrobacterium tumefaciens virB operon.
Nucleic Acids Res. 17:4541-4550.
6. Ditta, G., S. Stanfield, D. Corbin, and D. Helinski. 1980. Broad
host range DNA cloning system for gram-negative bacteria:
construction of a gene bank of Rhizobium meliloti. Proc. Natl.
Acad. Sci. USA 77:7347-7351.
7. Garfinkel, D. J., and E. W. Nester. 1980.Agrobacteriumtume-
faciens mutants affected in crown gall tumorigenesis and oc-
topine catabolism. J. Bacteriol. 144:732-743.
8. Huang, M.-L. W., G. A. Cangelosi, W. Halperin, and E. W.
Nester. 1990. A chromosomalAgrobacterium tumefaciens gene
required for effective plant signal transduction. J. Bacteriol.
9. Ingraham, J. 1887. Effect of temperature, pH, water activity,
and pressure on growth, p. 1543-1554. In F. C. Neidhardt,J. L.
Ingraham, K. B. Low, B. Magasanik, M. Schaechter,and H. E.
Umbarger (ed.),Escherichia coli and Salmonella typhimurium:
cellular and molecular biology, vol. 2. American Society for
Microbiology, Washington, D.C.
10. Jin, S., R. Prusti, T. Roitsch, R. Ankenbauer,and E. W. Nester.
1990. Phosphorylation of the VirG protein ofAgrobacterium
tumefaciens by the autophosphorylatedVirA protein:essential
role in the biological activity of VirG. J. Bacteriol. 172:4945-
11. Jin, S., T. Roitsch, P. Christie, and E. W. Nester. 1990. The
regulatory VirG protein specificallybinds to a cis-acting regu-
latory sequence involved in transcriptional activation ofAgro-
bacterium tumefaciens virulence genes. J. Bacteriol. 172:531-
12. Kanemoto, R. H., A. T. Powell, D. E. Akiyoshi, D. A. Regier,
R. A. Kerstetter,E. W. Nester,M. C. Hawes,and M. P. Gordon.
1989. Nucleotide sequence and analysis of the plant-inducible
locuspinF from Agrobactenium tumefaciens. J. Bacteriol. 171:
13. Koshland, D. E. 1952. Effect of catalysts on the hydrolysisof
acetyl phosphate. Nucleophilic displacement mechanisms in
enzymatic reactions. J. Am. Chem. Soc. 74:2286-2292.
14. Kunkel, T. 1985. Rapid and efficient site-specific mutagenesis
without phenotypic selection. Proc. Natl. Acad. Sci. USA
15. Leroux, B., M. Yanofsky, W. Winans, J. Ward, S. Zeigler,and
E. Nester. 1987. Characterization of the virA locus ofAgrobac-
terium tumefaciens: a transcriptional regulatory and host range
determinant. EMBO J. 6:849-856.
16. Lukat, G. S., B. H. Lee, J. M. Mottonen,A. M. Stock, and J. B.
Stock 1991. Roles of the highlyconserved aspartateand lysine
residues in the response regulator of bacterial chemotaxis. J.
Biol. Chem. 266:8348-8354.
17. Melchers,L. S.,M. J. Maroney,A. denDulk-Ras,D.Thompson,
H. van Vuuren, R. A. Schilperoort,and P. J. J. Hooykaas.1989.
Octopine and nopaline strains ofAgrobacterium tumefaciens
differ in virulence: molecular characterization of the virF locus.
Plant Mol. Biol. 14:249-259.
18. Melchers, L. S., T. J. G. Regensburg-Tuink, R. B. Bourret,
N. J. A. Sedee, R. A. Schilperoort,and P. J. J. Hooykaas.1989.
Membrane topology and functional analysis of the sensory
proteinvirA ofAgrobacterium tumefaciens.EMBO J. 8:1919-
19. Miller, J. H. 1972. Experiments in molecular genetics. Cold
SpringHarbor Laboratory,ColdSpring Harbor,N.Y.
20. Nara, F., S.-I. Matsuyama,T. Mizuno,and S. Mizushima. 1986.
Molecular analysis of mutant ompR genes exhibitingdifferent
phenotypeas toosmoregulationof theompFandompC genesof
Escherichia coli. Mol. Gen. Genet. 202:194-199.
21. Ninfa,A.J.,E. G. Ninfa,A.Lupas,A.Stock,B.Magasanik,and
J. Stock. 1988. Crosstalk between bacterial chemotaxis signal
transduction proteinsand the regulatorsoftranscriptionof the
Ntrregulon:evidence thatnitrogenassimilation and chemotaxis
are controlled bya commonphosphotransfermechanism. Proc.
Natl. Acad. Sci. USA 85:5492-5496.
22. Ooms, G.,P. M.KlapwiJk, J. A.Poulis,and R. A.Schilperoort.
1980. Characterization of Tn9O4 insertions inoctopineTiplas-
mid mutants ofAgrobacterium tumefaciens.J. Bacteriol. 144:
23. Pazour, G. J., and A. Das. 1990. VirG, anAgrobacterium
tumefaciens transcriptional activator, initiates translation at a
VOL. 174, 1992
UUG codon and is a sequence-specific DNA-binding protein. J.
24. Pazour, G. J., and A. Das. 1990. Characterization of the VirG
binding site ofAgrobacterium tumefaciens. Nucleic Acids Res.
25. Pazour, G. J., C. Ta, and A. Das. 1991. Mutants ofAgrobacte-
rium tumefaciens with elevated vir gene expression. Proc. Nat].
Acad. Sci. USA 88:6941-6945.
26. Ream, W. 1989. Agrobacterium tumefaciens and interkingdom
genetic exchange. Annu. Rev. Phytopathol. 27:583-618.
27. Ronson, C. W., B. T. Nixon, and F. M. Ausubel. 1987. Con-
served domains in bacterial regulatory proteins that respond to
environmental stimuli. Cell 49:579-581.
28. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-
ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.
29. Schmidhauser, T., and D. Helinski. 1985. Regions of broad-host-
range plasmid RK2 involved in replication and stable mainte-
nance in nine species of gram-negative bacteria. J. Bacteriol.
30. Shine, J., and L. Dalgarno. 1974. The 3'-terminal sequence of
Escherichia coli 16S ribosomal RNA: complementarity to non-
sense and ribosome binding sites. Proc. Natl. Acad. Sci. USA
31. Stachel, S., and E. Nester. 1986. The genetic and transcriptional
organization of the vir region of the A6 Ti plasmid ofAgrobac-
terium tumefaciens. EMBO J. 5:1445-1454.
32. Stachel, S., and P. Zambryski. 1986. virA and virG control the
plant-induced activation of the T-DNA transfer process of A.
tumefaciens. Cell 46:325-333.
33. Stachel, S. E., E. Messens, M. Van Montagu, and P. Zambryski.
signal molecules produced by
wounded plant cells that activate T-DNA transfer inAgrobac-
tenium tumefaciens. Nature (London) 318:624-629.
34. Stock, A., J. Mottonen, J. Stock, and C. Schutt. 1989. Three-
dimensional structure of CheY, the response regulator of bac-
terial chemotaxis. Nature (London) 337:745-749.
35. Stock, J., A. Ninfa, and A. Stock. 1989. Protein phosphorylation
and regulation of adaptive responses in bacteria. Microbiol.
36. Vieira, J., and J. Messing. 1987. Production of single-stranded
plasmid DNA. Methods Enzymol. 153:3-11.
37. Volz, K., and P. Matsumura. 1991. Crystal structure of Esche-
richia coli CheY refined at 1.7-A resolution. J. Biol. Chem.
38. Weglenski, P., A. J. Ninfa, S. Ueno-Nishio, and B. Magasanik.
1989. Mutations in the glnG gene of Escherichia coli that result
in increased activity of nitrogen regulator I. J. Bacteriol. 171:
39. Winans, S. Personal communication.
40. Winans, S. C. 1990. Transcriptional induction of an Agrobacte-
rium regulatory gene at tandem promoters by plant-released
phenolic compounds, phosphate starvation, and acidic growth
media. J. Bacteriol. 172:2433-2438.
41. Winans, S. C., P. R. Ebert, S. E. Stachel, M. P. Gordon, and
E. W. Nester. 1986. A gene essential for Agrobactenium viru-
lence is homologous to a family ofpositive regulatory loci. Proc.
Natl. Acad. Sci. USA 83:8278-8282.
42. Winans, S. C., R. A. Kerstetter, and E. W. Nester. 1988.
Transcriptional regulation of the virA and virG genes ofAgro-
bacterium tumefaciens. J. Bacteriol. 170:4047-4054.
43. Zambryski, P. 1988. Basic processes underlying Agrobacteri-
um-mediated DNA transfer to plant cells. Annu. Rev. Genet.