Download Archives of Microbiology

Arch Microbiol (1994) 161:310-315
Archives of
Hicrobiolegy
© SpringerVerlag 1994
Several phenotypic changes in the cell envelope
of Agrobacterium tumefaciens chvB mutants are prevented
by calcium limitation
Saskia Swart, Trudy J. J. Logman, Ben J. J. Lugtenberg, Gerrit Smit, Jan W. Kijne
Institute of Molecular Plant Sciences, Leiden University, P.O. Box 9505, 2333 AL Leiden, The Netherlands
Received: 24 May 1993/Accepted: 8 October 1993
Abstract. The chvB gene of Agrobacterium tumefaciens
encodes a 235 k D a proteinaceous intermediate involved
in the synthesis of/%1,2-glucan, chvB mutants show a
pleiotropic phenotype. Besides not to produce cyclic
/?-l,2-glucan, chvB mutants have been reported to be
avirulent, attachment-deficient, and nonmotile. In this
study we report additional differences from the parent
strain, p r o b a b l y all linked to changes in the cell envelope.
This pleiotropic phenotype - except for attachment and
virulence - could largely be prevented by growing chvB
cells with low levels of calcium. Although a role for
fi-l,2-glucan in o s m o a d a p t a t i o n has been proposed, the
m o d e of action of/3-1,2-glucan is not known. We speculate
that in A. tumefaciens fl-l,2-glucan stabilizes membranes,
which would be important especially in hypotonie media
containing calcium.
Key words: Agrobacterium - f l - 1 , 2 - G l u c a n - chvB Mut a n t s - Ca 2 +
chvB is one of the c h r o m o s o m a l genes of Agrobacterium
tumefaciens usually required for transformation of plants
(Douglas et al. 1982, 1985). The chvB gene is required for
the synthesis of a low molecular weight cyclic fi-l,2glucan. The gene encodes a 235 k D a cytoplasmic m e m brane protein, which is an intermediate in fl-l,2-glucan
synthesis (Zorreguieta and Ugalde 1986). Another virulence gene, chvA, is required for transport of fi-l,2-glucan
to the periplasmic space (Cangelosi et al. 1989). However,
chvB mutants are not only impaired in fi-l,2-glucan
production. They also exhibit reduced motility and
resistance to certain phages, due to absence of flagella
(Bradley et al. 1984), and are attachment-deficient and
Abbreviations: Cb, carbenicillin; Km, kanamycin; TCA, trichloroacetic acid; Kay fraction of the stationary gel volume available for
diffusion; LPS, lipopolysaccharide; SDS-PAGE, Sodium dodecyl
sulphate polyacrylamide gel electrophoresis
Correspondence to: S. Swart, Mogen International, Einsteinweg 97,
2333 CB Leiden, The Netherlands Tel. (0) 71 25 82 82;
Fax (0) 71 22 1471
avirulent (Douglas et al. 1982, 1985). A large part of the
chvB gene is not required for glucan synthesis. Also
attachment ability and virulence are not impaired by
deletion of this part of the chvB gene, suggesting that the
latter properties are coupled. Since wild-type A. tumefaciens cells accumulate fl-l,2-glucan in the periplasmic
space when grown under hypo-osmotic conditions, synthesis of cyclic fi- 1,2-glucan appears to be osmoregulated
(Miller et al. 1986). Cangelosi et al. (1990) reported for
chvB mutants a growth rate lower than that of wild-type
cells and an altered periplasmic and cytoplasmic protein
content when mutant cells were cultured in low-osmoticstrength media, but not for cells cultured in high-osmoticstrength media. These observations support a role for
periplasmic glucan in osmoadaptation. Inadequate osm o a d a p t a t i o n m a y result in deficiencies of the cell
envelope. However, the mechanism of action of fi-l,2glucan is not known.
This study shows that the pleiotropic phenotype of
chvB mutants of A. tumefaciens can largely be prevented
by growing chvB cells with low levels of calcium. We
speculate that fl-l,2-glucan is involved in stabilization of
membranes, especially in hypo-osmotic media containing
calcium.
Materials and methods
Bacterial strains and culture conditions
Agrobacterium tumefaciens strains and mutants were generous gifts
from E. W. Nester (University of Washington, Seattle, Wash., USA),
and have been described by Douglas et al. (1985). A. tumefaciens
strain A348 has a wild-type C58 chromosome and harbours plasmid
pTiA6NC (Garfinkel et al. 1981). A348 has a wild-type phenotype.
chvA mutants ME42 and ME66 and chvB mutants MEll7 and
ME 118 have a Tn3Hoho 1 insertion, including a Cb resistance gene,
in the chromosomal virulence A (chvA) or B (chvB) gene,
respectively, chvB mutants A1011 and A1045 have a Tn5 insertion,
including a Km resistance gene, in the chvB gene. Bacteria were
grown at 28 °C in 100-ml Erlenmeyer flasks containing 50ml
medium on a rotary shaker (180 rpm). The composition of TY
medium was 0.5% tryptone-0.3% yeast extract. According to the
manufacturer (Difco Laboratories, Detroit, Mich., USA), the Ca 2÷
concentration of this medium is 0.15 mM and the Mg 2+ concentra-
311
tion is 0.13 mM. TYC medium is TY medium supplemented with
7 mM CaClz (Beringer 1974). Swarm plates contain in addition
0.3% (w/v) agar.
Plant culture conditions
Seeds of pea (Pisum sativum cv. Finale, purchased from Cebeco,
Rotterdam, The Netherlands) were surface-sterilized and cultivated
as previously described (Smit et al. 1986). Kalanchob"daigremontiana
plants were cultivated as described by Ooms et al. (1980).
Measurement of medium osmolarity
The osmolarity of media was measured with an osmomat 030
cryoscopic osmometer (Salm and Kipp, Breukelen, The Netherlands).
Results
Growth
Attachment assay
The attachment assay used was previously described by Smit et al.
(1986). Briefly, bacteria were grown to early stationary phase,
centrifuged and resuspended in 25 mM potassium phosphate buffer,
pH 7.5, to 3 x 108 bacteria per nal. After incubation of lateral roots
from 7-day-old axenically grown pea plants for 2 h in this suspension, the roots were washed in phosphate buffer and screened for
attached bacteria by phase-contrast microscopy. In this study
attachment to root hairs was ranked into three classes: class 1, no
attached bacteria; class 2, few bacteria directly attached to the root
hair; class 3, many attached bacteria forming two or more layers
or a caplike aggregate on top of the root hair. The variability of
the test was about 10% and depended largely on the condition of
the roots.
ldrulence assays
In the greenhouse, leaves of K. daigremontiana plants were wounded
with a toothpick. Cells of early stationary phase cultures of A.
tumefaciens strain A348 or strain MEll7, grown in TYC or TY
medium, were harvested and resuspended in phosphate buffer to
an A6z0 value of 1.0. Ten gl of this suspension were added per
wound site. Tumour formation was quantified three weeks after
inoculation.
Purification of flagella
Flagella were isolated from the bacterial cell surface and purified
by sucrose gradient ultracentrifugation according to the method of
Carsiotis et al. (1984).
Isolation of cell envelope fractions and lipopolysaccharide
Bacteria were grown to early stationary phase. Cell membrane
preparations were made according to De Maagd et al. (1989).
Proteins were separated on 11% polyacrylamide gels as described
by Lugtenberg et al. (1975). For analysis of LPS, cell envelope
preparations in sample buffer, each containing an equal amount of
protein, were treated first with proteinase K (0.2 mg/ml) for 60 min
at 60 °C and then applied to a 15% SDS-polyacrylamide gel.
Proteins and LPS were visualized by silver-staining (Blum et al.
1987).
Purification of fl-l,2-glucan
To test the production of fl-1,2-glucan, cells from 11 early stationary
phase cultures were harvested by centrifugation at 10,000 x g for
10 min. Pellets were extracted with 1% TCA for 30 min at room
temperature. TCA extracts were neutralized, concentrated, and
applied to a Sephadex G50 column (63.5 x0.8 cm), which was
subsequently eluted with 7% propanol. One ml fractions were tested
for the presence of carbohydrates by using the phenol-sulphuric
acid method (Hanson and Philips 1981). The saccharide composition of these carbohydrate fractions was determined by gas-liquid
chromatography after sample hydrolysis (Albersheim et al. 1967)
(liquid phase: 3% ECNSS/M on GasChromQ [Applied Science,
Oud-Beijerland, The Netherlands]).
Agrobacterium tumefaciens strain A348, g r o w n in the
s t a n d a r d g r o w t h m e d i u m TYC, h a d a doubling time of
a b o u t 2.1 h, reaching a final A620-value of 1.2 to 1.3. As
shown in Table 1, omission of the calcium supplement
(TY medium) yielded for this strain the same g r o w t h
characteristics as seen in T Y C medium. The osmolarity
of TY and T Y C media is 0.043 and 0.065 osM, respectively. Because chvB m u t a n t s have been reported to be
impaired in g r o w t h in low osmolarity m e d i u m (Cangelosi
et al. 1990), we tested whether chvB m u t a n t s grew slower
in TY c o m p a r e d to T Y C media. However, chvB m u t a n t
strain M E l l 7 appeared to grow slower in T Y C with a
doubling time of 3.4 h, reaching a final A6zo-value of 0.8
to 0.9, whereas the doubling time and the g r o w t h
m a x i m u m a p p e a r e d to be restored to nearly the level of
strain A348 by omitting the calcium supplement from
the m e d i u m (Table 1). Similar results were obtained with
chvB m u t a n t s A1011, A1045, and M E l l 8 , and chvA
m u t a n t s M E 4 2 and M E 6 6 (data n o t shown).
Strain M E l l 7 h a r b o u r s a T n 3 H o H o l t r a n s p o s o n in
chvB carrying a Cb resistance gene. Despite the presence
of this gene, M E l l 7 does n o t grow in T Y C m e d i u m
supplemented with 100 gg/ml Cb (Table 1). In contrast,
in TY m e d i u m M E l l 7 grows n o r m a l l y in the presence
of Cb, indicative of expression of the Cb resistance gene.
Several other chvB strains were tested for this phenotype.
M E l l 8 also h a r b o u r e d a C b resistance gene, and appeared to be Cb-resistant only when g r o w n in TY medium.
Similar results were obtained with chvA m u t a n t s M E 4 2
and M E 6 6 (data not shown). The other chvB strains
studied, A1011 and A1045, are T n 5 - m u t a n t s thus harb o u r i n g a K m resistance gene. P h e n o t y p i c expression of
K m resistance appeared not to depend on the calcium
concentration of TY m e d i u m (data not shown). Cells of
strain A348 are not Cb resistant and did grow neither in
TY nor in T Y C media supplemented with 100 gg/ml Cb.
W h e n the Cb concentration was lowered to 10 gg/ml Cb,
cells of strain A348 showed an impaired g r o w t h in b o t h
TY and T Y C media, being only slightly m o r e impaired
in T Y C medium.
Production of fl- l,2-glucan
W h e n strain A348 was tested for p r o d u c t i o n of fl-l,2glucan using gel permeation c h r o m a t o g r a p h y , a c a r b o h y drate-positive peak was observed at a Kay of 0.6, corresp o n d i n g with the Kav previously found for cyclic/~-l,2glucan (Cangelosi et al. 1989; Miller et al. 1986). The
a m o u n t of cell-associated soluble c a r b o h y d r a t e with a
Kay of 0.6 p r o d u c e d by A348 in 1 1 T Y C m e d i u m (5 g
312
Table 1. Characteristics of Agrobacterium
tumefaciens strain A348 and its chvB mutant
strain ME117 grown in TY or TYC a media
Strain
Growth b
medium
Final
A62ovalue
Doubling
time, h
Cb r~ Motility Attachment
Virulence
A348
TYC
TY
TYC
TY
1.2 1.3
1.2-1.3
0.8-0.9
1.1 - 1.2
2.1
2.1
3.4
2.5
.
+
+
+
MEll7
(chvB)
+
+
.
.
+
+
-.
-
-
a TY is TYC medium (0.5% tryptone 0.3% yeast extract-7 mM CaC12)without the calcium
supplement
u Determined by the method of Smit et al. (1986). More than 80% class 1 was designated
"no attachment"
Growth in TY or TYC media supplemented with 100 gg/ml Cb
fresh weight of cells) was 31 _+ 3 mg. However, from
bacteria grown in TY medium only 11 _+ 2 mg of carbohydrate with a Kay of 0.6 could be isolated. The saccharide
composition of both carbohydrate fractions, as determined by gas-liquid chromatography, appeared to be identical, with glucose representing the major saccharide detected. Next production of fi-l,2-glucan by strain ME117
grown in T Y C or TY medium was tested. As judged from
the a m o u n t of carbohydrate-positive material detected,
ME117 does not produce fi-l,2-glucan when grown in
either TY or T Y C media. In view of the possibility that
fi-1,2-glucan was excreted, the growth media were chekked for the presence of fi-l,2-glucan. Extracellular cyclic
/?-1,2 glucan was found neither with the parent strain nor
with chvB cultures.
chvB mutants have been reported to show a reduced
n u m b e r of flagella (Bradley et al. 1984). Electron microscopic observations showed that the number of flagella
on the surface of ME117 increased significantly when the
bacteria were grown in TY medium (data not shown).
We c o m p a r e d flagella preparations of A348 and ME117,
grown either in TY or TYC, by S D S - P A G E analysis
(Fig. 1 A). S D S - P A G E of purified flagella of A. tumefaciens strain A348 showed two polypeptides of 30 k D a
and 32 kDa, respectively (Fig. 1A, lanes 1 and 2). At
present it is unknown whether the band with higher
mobility represents an artefact, e.g., a proteolytic product
of flagellin, or whether the two bands correspond to two
isotypes of flagellin, the m a j o r protein of a flagellum. F o r
Motility
Parent strain A348 is motile during exponential growth,
in both TY and T Y C media. Only in the late stationary
phase the cells become nonmotile. In contrast, cells of
strain ME117 are nonmotile when grown in T Y C medium, as judged from phase-contrast microscopy (data not
shown). However, when grown in TY medium M E l l 7
cells are motile, although not to the same extent as are
cells of strain A348. Motility was also assessed by using
swarm plates. On TY swarm plates cells of strain ME117
formed swarms, although smaller than did cells of the
parent strain. On T Y C plates m u t a n t cells formed only
a small colony, confined exclusively to the site of inoculation, whereas A348 cells formed swarms as large as those
seen on TY swarm plates (Fig. 1B). The same observations were made for the other chvB mutants tested,
A1011, A1045 and ME118. Also the chvA mutants ME42
and ME66 showed an increase in mobility when grown
in TY c o m p a r e d to T Y C (data not shown). However, for
these mutants the difference was smaller, chvA mutants
grown in T Y C were more motile than chvB mutants
grown in TYC, whereas the motility of both strains grown
in TY was similar, chvA mutants have been reported to
have a leaky phenotype. The mutants still export 2% of
the wild-type a m o u n t of fi-l,2-glucan (O'Connell and
H a n d e l s m a n 1989). Moreover, chvA mutants occasionally
form small tumours, and when harbouring a pSym9
plasmid, can form small root nodules (Van Veen et al.
1987). Therefore, we decided not to further use chvA
mutants in this study.
A
B
kDa
97-66-45-36-29--
18--
12345
Fig. 1A. Silver-stained SDS-PAGE (15%) profiles of crude flagella
preparations of Agrobacterium tumefaciens strain A348 (lanes 1 and
2) and chvB mutant strain MEl17 (lanes 3, 4 and 5) grown in TY
medium (lanes 1 and 3) or TY supplemented with 7 mM CaC12
(TYC) (lanes 2, 4 and 5). The amount applied corresponds with
flagella of 7 x 108 bacteria, with the exception of lane 5 (1.4 x 101°
bacteria). Positions of molecular weight markers are indicated at
the left; B 0.3 % agar swarm-plates containing TYC (uppear plate)
or TY (lower plate) medium spotted with A348 (1 and 3) and
ME117 (2 and 4)
313
the chvB mutant the yield of the two polypeptides
corresponding with the flagellin bands was strongly
reduced in the presence of calcium (Fig. 1 A, lanes 3, 4
and 5).
Use of a concentration series of calcium in TY medium
revealed that at an amount of between 2.0 and 2.5 m M
calcium added, the mutant cells became nonmotile. Since
the calcium concentration in TY is 0.15 mM, the critical
calcium concentration that directly or indirectly prevents
motility apparently is between 2.15 and 2.65 raM.
Next to calcium the addition of 7 m M MgC12 to TY
medium reduced growth and motility of ME117, but the
effect was less pronounced. For instance, the final A620
of M E l l 7 cells grown in TY supplemented with 7 m M
MgC12 was 1.1 and the diameter of the colony formed
on corresponding swarm plates was larger than on TYC
swarm plates, although not as large as on TY swarm
plates (data not shown). Addition of 10.5 m M NaC1 or
KC1 did not result in reduced growth and motility of
chvB mutants. These results indicate that not just salt,
but rather a higher concentration of divalent cations and
especially Ca 2 ÷ has an inhibitory effect on growth and
motility of chvB mutants.
Cell envelope proteins and L P S
Since some of the observed changes in chvB mutants,
such as lack of flagella and attachment-deficiency, were
A
B
associated with the cell envelope, we examined the effect
of reduced Ca2+-levels on profiles of cell envelope
proteins and the LPS of strains A348 and M E l l 7 .
At reduced Ca2+-levels, the cell envelope protein
profiles of the bacteria are altered. However, this alteration was not specific for M E l l 7 bacteria, since also
A348 bacteria show a different cell envelope protein
profile when grown in TY and TYC. Nevertheless,
comparison of the profiles of A348 and M E l l 7 yielded
several differences. These include an increase in the
amount of a 20 kDa protein band in the profile of ME117
grown in TYC (Fig. 2A).
To compare the LPS profiles and the amount of LPS
of strains A348 and M E l l 7 grown in TYC or TY, cell
envelope isolates containing equal amounts of protein
were treated with proteinase K and separated in a
SDS-polyacrylamide gel. The relative amount of LPS
isolated from A348 grown in TYC was smaller than the
amount of LPS from A348 grown in TY (Fig. 2B).
However, whereas the amount of LPS from strain ME117
grown in TY was only slightly reduced compared to that
from A348, the amount of LPS from M E l l 7 grown in
TYC was severely reduced (Fig. 2B). Since the protein
content of cell envelopes of ME117 cells grown in TYC
is probably not increased, M E l l 7 cells grown in TYC
apparently contain a decreased amount of LPS. Furthermore, the profile of M E l l 7 LPS appears to differ
from that of the parent strain in that the banding pattern
of the region with higher motility, i.e. the core region, is
different.
Attachment and virulence
kDa
The attach.ment ability of A. tumefaciens cells grown in
TYC and TY media was tested. Wild-type cells grown in
TYC attach to pea root hairs. However, omission of
calcium from the medium resulted in loss of attachment
ability (Smit et al. 1987) (Table 1). Cells of strain M E l l 7
did not attach to pea root hair tips when grown in either
TYC or TY media (Table 1).
Virulence was tested on KalanchoY leaves. Wild-type
cells grown in TYC or TY media form tumours. Since
attachment presumably is a prerequisite for tumour
formation, the plant apparently supplies the calcium
needed for the attachment of wild-type cells grown in TY
medium. Cells of chvB strain M E l l 7 grown in TYC or
TY media were avirulent (Table 1).
66--
45-36-29--
18-14--
Discussion
1
2
3
4
1
2
3
4
Fig. 2A, B. Silver-stained ceil envelope profiles of A. tumefaciens
cells separated by SDS-PAGE. A An 11% gel of ceil envelope
proteins of chvB mutant strain ME117 (lanes 1 and 2) and parent
strain A348 (lanes 3 and 4) grown in TY medium (lanes 1 and 3)
or TY supplemented with 7 mM CaC12 (TYC) (lanes 2 and 4).
Positions of molecular weight markers are indicated at the left. An
arrow head indicates the position of the 20-kDa protein. B LPS
profiles. Cell envelope isolations containing an equal amount of
protein were treated with proteinase K and applied to a 15% gel.
Lanes are numbered as in A
fi-l,2-Glucan is a cyclic oligosaccharide found in Rhizobiaceae and reported to be involved in adaptation to
hypo-osmotic conditions. Miller et al. (1986) found less
fi-l,2-glucan produced by wild-type Agrobacterium tumefaciens cells grown in medium of high osmolarity than
by cells grown in medium of low osmolarity. Furthermore, fi-l,2-glucan negative mutants are impaired in
growth in medium of low osmolarity (Cangelosi et al.
1990). However, the mechanism of action of cyclic fl-l,2glucan is not known. In a hypotonic environment cells
314
have to deal with an influx of water which causes swelling
of the cells. Gram-negative bacteria have a relatively rigid
cell wall, but when the influx of water proceeds and the
cell wall cannot extend further, rupture due to a process
of brittle fracture will occur (Corner and Marquis 1969).
It has been reported that sugars protect membrane
structures from damage, for instance during freeze-thaw
cycles (Crowe et al. 1984a; Rudolph and Crowe 1985).
Furthermore, several reports point out interactions between carbohydrates and phospholipid monolayers. Carbohydrates would coordinate to the polar headgroups of
the lipidlayer by hydrogen bonds and in doing so alter
their mobility and/or packing density thereby increasing
the film area (Crowe et al. 1984b; Johnston et al. 1984).
We speculate that cyclic fi-l,2-glucan is involved in
stabilizing the lipid bilayers of the cytoplasmic and outer
membranes of Rhizobiaceae grown in hypotonic media.
Calcium is known to adsorb onto membrane surfaces
causing disruption of the membrane and by that the cell,
especially in aqueous solutions (BfikAs and Disalvo 1991;
Lis et al. 1981). A disruption of cell envelope membranes
would result in impaired assembly of membrane structures such as LPS and flagella. This offers a possible
explanation for the toxicity of calcium for fl-l,2-glucannegative mutants, as described in this study. Apparently
fi-l,2-glucan-negative mutants are very sensitive to high
calcium concentrations. The pleiotropic phenotype of
chvB mutants, i.e. attachment inability, avirulence, lack
of flagella, altered cell envelope profile and altered LPS
profile, mainly corresponds to changes in the cell envelope. A chvB mutant seems to show impaired assembly
of outer membrane components when grown in TYC
medium. If periplasmic fl-l,2-glucan can stabilize membranes and may compete for Ca 2 +-binding sites at polar
headgroups, a fi-l,2-glucan-eontaining cell would have
less problems with disruption by calcium than would
have a fi-l,2-glucan-negative mutant. We found two
indications for a correlation between fi-l,2-glucan and
calcium: (i) A348 bacteria produce two to three times
more fi-l,2-glucan when grown in TYC medium than do
cells grown in TY medium, despite the fact that TYC
medium has a higher osmolarity than TY medium; this
suggests that the osmolarity of the medium is one but
not the only factor influencing the production of fl-l,2glucan. (ii) Omission of calcium from the medium
restored the growth of chvB mutant M E l l 7 to nearly
wild-type levels and allowed growth of M E l l 7 in
100 gg/ml Cb. Moreover, these bacteria became motile
in TY and the production of flagella was restored. Also,
the cell envelope profiles and the amount of LPS were
more like those of the parent strain when the mutant
bacteria were grown in TY (although also the parent
strain cells showed a difference in these characteristics
when grown in TY versus TYC medium).
Attachment and virulence were not restored when
calcium was omitted from TYC medium. This suggests
that (i) fi-l,2-glucan could be directly involved in attachment; although this seems unlikely, however, since exogenously added fi-l,2-glucan does not affect attachment
(O'Connell and Handelsman 1989; S. Swart, unpublished
results), (ii) fl-l,2-glucan is needed for the synthesis or
functioning of an attachment factor, or (iii) calcium is
needed for virulence and/or attachment. The calciumbinding protein rhicadhesin, which is involved in the first
step of attachment of Rhizobiaceae to plant cells (Smit
et al. 1989) has been reported to be anchored to the
bacterial cell surface in a calcium-dependent manner
(Smit et al. 1991). As a hypothesis, we propose that above
a critical level (between 2.15 and 2.65 raM) calcium is
toxic to fi-l,2-glucan deficient mutants but that, on the
other hand, calcium is required for anchoring, stabilization and/or activation of rhicadhesin. Absence of fi-l,2glucan apparently results in deregulation of this process.
Ongoing research is focused on the role of rhicadhesin
in virulence ofA. tumefaciens bacteria and on the possible
connection between fi-1,2-glucan, calcium and rhicadhesin.
Acknowledgements. This work was supported by the Foundation
for Fundamental Biological Research (BION), which is subsidized
by the Netherlands Organization of ScientificResearch (NWO). We
thank Otto Geiger for stimulating discussions.
References
Albersheim P, Nevins D, English PD, Karr A (1967) A method for
the analysis of sugars in plant cell wall polysaccharides by
gas-liquid chromatography. Carbohydr Res 5:340-345
B/tkgts LS, Disalvo EA (1991) Effect on the cryoprotective action
of trehalose. Cryobiology 28:347-353
Beringer J (1974) R factor transfer in Rhizobium leguminosarum.
J Gen Microbiol 84:188-198
Blum H, Beier H, Gross HJ (1987) Improved silver staining of plant
proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 8:93-99
Bradley DE, Douglas CJ, Peschon J (1984) Flagella-specific bacteriophages of Agrobacterium tumefaciens: demonstration of
virulence of nonmotile mutants. Can J Microbiol 30:676-681
Cangelosi GA, Martinetti G, Leigh JA, Lee CC, Theines C, Nester
EW (1989) Role for Agrobacterium tumefaciens chvA protein
in export of fl-l,2-glucan. J Bacteriol 171:1609 1615
Cangelosi GA, Martinetti G, Nester EW (1990) Osmosensitivity
phenotypes of Agrobacterium tumefaciens mutants that lack
periplasmic fl-l,2-glucan. J Bacteriol 172:2172 2174
Carsiotis M, Weinstein DL, Karch H, Holder IA, O'Brien AD (1984)
Flagella of Salmonella typhimurium are a virulence factor in
infected C57BL/6J mice. Infect Immun 46:814 818
Corner TR, Marquis RE (1969) Why do bacterial protoplasts burst
in hypotonic solutions? Biochim Biophys Acta 183:544-558
Crowe LM, Mourdian R, Crowe JH, Jackson SA, Womeisly C
(1984a) Effect of carbohydrates on membrane stability at low
water activities. Biochim Biophys Acta 769:141-150
Crowe JH, Whittam MA, Chapmen D, Crowe LM (1984b) Interactions of phospholipid monolayers with carbohydrates. Biochim
Biophys Acta 769:151-159
De Maagd RA, Rijk R de, Mulders IHM, Lutgenberg BJJ (1989)
Immunological characterization of Rhizobium leguminosarum
outer membrane antigens using polyclonal and monoclonal
antibodies. J Bacteriol 171:1136 1142
Douglas CJ, Halperin W, Nester EW (1982) Agrobacterium tumefaciens mutants affected in attachment to plant cells. J Bacteriol
152:1265-1275
Douglas CJ, Staneloni RJ, Rubin RA, Nester EW (1985) Identification and genetic analysis of an Agrobacterium tumefaciens
chromosomal virulence region. J Bacteriol 161:850-860
Garfinkel D J, Simpson RB, Ream LW, White FF, Gordon MP,
Nester EW (1981) Genetic analysis of crown gall: fine structure
315
map of the T-DNA by site-directed mutagenesis. Cell 27:
143-153
Hanson RS, Philips JA (1981) Chemical composition. In: Gerhardt
P, Murray RGE, Costilow RN, Nester EW, Wood WA, Krieg
NR, Philips GB (eds) Manual of methods for general bacteriology. American Society for Microbiology, Washington DC,
pp 328-364
Johnston DS, Coppard E, Parera PV, Chapman D (1984) Langmuir
film balance study of the interaction between carbohydrates
and phospholipid monotayers. Biochemistry 23:6912 6919
Lis LJ, Parsegian VA, Rand RP (1981) Binding of divalent cations
to dipalmitoylphosphatidylcholine bilayers and its effect on
bilayer interaction. Biochemistry 20:1761-1770
Lugtenberg BJJ, Meyers J, Peters R, Hoek P van der, Alphen L
van (1975) Electrophoretic resolution of the major outer membrane protein of Escherichia coli K12 into four bands. FEBS
Lett 58:254-258
Miller KJ, Kennedy EP, Reinhold VN (1986) Osmotic adaptation
by gram-negative bacteria: possible role for periplasmic oligosaccharides. Science 231:48 51
O'Connetl KP, Handelsman J (1989) ChvA locus may be involved
in export of neutral cyclic /~-l,2-1inked glucan from Agrobacterium tumefaciens. Mol Plant-Microbe Int 2:11 16
Ooms G, Hooykaas PJJ, Poulis JA, Schitperoort RA (1980)
Characterization of Tn904 insertions in octopine Ti plasmid
mutants of Agrobacterium tumefaciens. J Bacteriol 144:82 91
Rudolph A, Crowe JH (1985) Membrane stabilization during
freezing: the role of two natural cryoprotectants, trehalose and
proline. Cryobiology 22:367 377
Smit G, Kijne JW, Lugtenberg BJJ (1986) Correlation between
extracellular fibrils and attachment of Rhizobium leguminosarum to pea root hair tips. J Bacteriol 168:821-827
Smit G, Kijne JW, Lugtenberg BJJ (1987) Involvement of both
cellulose fibrils and a Ca2+-dependent adhesin in the attachment of Rhizobium leguminosarum to pea root hair tips. J
Bacteriol 169:4294-4301
Smit G, Logman TJJ, Boerrigter METI, Kijne JW, Lugtenberg BJJ
(1989) Purification and partial characterization of the Rhizobium leguminosarum biovar viciae Ca2+-dependent adhesin
which mediates the first step in attachment of cells of the family
Rhizobiaceae to plant root hair tips. J Bacteriol 171:4054-4062
Smit G, Tubbing DMJ, Kijne JW, Lugtenberg BJJ (1991) Role of
Ca 2+ in the activity of rhicadhesin from Rhizobium Ieguminosarum biovar viciae, which mediates the first step in attachment
of Rhizobiaceae cells to plant root hair tips. Arch Microbiol
155:278 283
Van Veen RJH, Den Dulk-Ras H, Schilperoort RA, Hooykaas PJJ
(1987) Chromosomal nodulation genes: Sym-plasmid containing Agrobacterium strains need chromosomal virulence genes
(chvA and chvB) for nodulation. Plant Mol Biol 8:105-108
Zorreguieta A, Ugalde RA (1986) Formation in Rhizobium and
Agrobaaerium spp. of a 235 kDa protein intermediate in /~D(1,2)-glucan synthesis. J Bacteriol 167:947 951