Download Surface polysaccharide mutants of Rhizobium sp. (Acacia) strain

Microbiology (1995), 141, 573-581
Printed in Great Britain
Surface polysaccharide mutants of Rhizobium
sp. (Acacia) strain GRHZ: major requirement of
lipopolysaccharide for successful invasion of
Acacia nodules and host range determination
Isabel Maria Ldpez-Lara,’ Guy Orgambide,2 Frank B. Dazzo,’
Jose Olivares’ and Nicolds Toro’
Author for correspondence : Nicolas Toro. Tel :
1
Department of
Microbiology, Estaci6n
Experimentaldel Zaidln,
Consejo Superior de
lnvestigaciones Cientificas,
Prof. Albareda 1, 18008
Granada, Spain
2
Department of
Microbiology, Michigan
State University, East
Lansing, MI 48824, USA
+34 58 121011 . Fax : + 34 58 129600.
Two transposon Tn5-induced mutants of wild-type broad-host-range Rhizobium
sp. GRHZ were isolated and found to harbour different alterations in surface
polysaccharides. These mutants, designated GRHZ-14 and GRHZ-50, induced a
few, empty nodules on Acacia and lost the ability to nodulate most host
herbaceous legumes. Whereas mutant GRHZ-14 produces an acidic
exopolysaccharide (EPS) similar to the wild-type, the acidic EPS of mutant
GRHZ-50 lacks galactose and the pyruvyl and 3-hydroxybutyryl substituents
attached to this sugar moiety. In addition, both mutants GRHZ-50 and GRHZ-14
were altered in smooth lipopolysaccharides (LPS). DNA sequence analyses of
the corresponding Tn5 insertions revealed that strain GRHZ-50 was mutated in
a DNA locus homologous to gal€, and in vitro enzyme assays indicated that the
UDPglucose 4-epimerase (GalE) activity was missing in this mutant strain. DNA
hybridization studies showed that the GRHZ-50 mutant DNA has homologous
sequences within the different biovars of Rhizobium leguminosarum.
However, no DNA homology to GRHZ-14 altered DNA was found in those
rhizobial strains, indicating that it represents a new chromosomal Ips locus in
Rhizobium sp. (Acacia) involved in symbiotic development.
Keywords : Rhixobium sp. GRH2, legume tree, polysaccharide, Tn5
INTRODUCTION
Acacia is the largest and most diverse genus of legume
trees, containing approximately 1500 species, and serves
as a model legume for studies of symbiotic nitrogen
fixation under conditions of desiccation stress. We are
investigating the role(s) of Rhixobizrm surface glycoconjugates in the development of nitrogen-fixing rootnodule symbiosis with Acacia. For these studies, we are
using wild-type Rhixobium sp. GRH2 which was originally
isolated from root nodules of Acacia yanopLylla in Chile
(Herrera e t al., 1985). A unique aspect of this rhizobial
strain is its wide host range, which includes legume trees
(e.g. Acacia, Prosopis) and a diversity of herbaceous
legumes such as Trzfolizrm, Lotzrs, Phaseolus, Vicia and
Siratro (Herrera e t al., 1985; L6pez-Lara e t al., 1993).
Although many reports suggest an involvement of
Rhixobium acidic heteropolysaccharides (exopolysaccha..........................................................................................................................................................
Abbreviation : EPS, exopolysaccharide.
0001-9303 0 1995 SGM
rides ; EPS) and lipopolysaccharides (LPS) in symbiotic
infection of herbaceous legume roots (Abe e t al., 1984;
Canter Cremers e t al., 1990; Carlson, 1982; Dazzo e t al.,
1992; Herrera e t al., 1985; Lamb e t al., 1982; Maier &
Brill, 1978; Muller e t al., 1988; Noel e t al., 1986; Reuber
e t al., 1991 ; Zhan e t al., 1992), little is known about their
importance in root infection of legume trees. We have
recently shown that the excreted acidic heteropolysaccharide produced by wild-type GRH2 and R. legzrminosarum bv. trifolii ANU843 are very similar in structure,
and the analysis of an Exo- mutant derivative of GRH2
indicates that this surface glycoconjugate is important for
invasion of host cells in Acacia nodules (L6pez-Lara e t a/.,
1993). In this study, we isolated and characterized two
different surface polysaccharide mutants of GRH2 in
order to examine the role of smooth LPS in symbiotic
development. This study of new GRH2 mutants identifies : (i) a major requirement of smooth LPS for successful
invasion of Acacia nodules and in determination of host
range; (ii) a dual role of exoB encoding UDPglucose 4epimerase involved in production of both smooth LPS
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
573
I. M. L O P E Z - L A R A a n d O T H E R S
and acidic EPS and in development of root nodule
symbiosis; a n d (iii) a n e w chromosomal Ips locus in
Rhipbizlm involved in symbiotic development.
METHODS
Bacterial strains and genetic manipulations. Rhixobitlm sp.
GRH2 was originally isolated from root nodules of A .
yanopbylla (Herrera e t al., 1985). Other bacterial strains used in
this work are listed in Table 1. Bacteria were routinely grown in
minimal medium (MM; Robertsen e t al., 1981), T Y (Beringer,
1974) or T Y supplemented with mannitol (YMT). Transposon
Tn5: :mob random mutagenesis was done by mating Escherichia
coli S17-1 (pSUP5011) with RhiTobium sp. GRH2. Kanamycin
(Km)-resistant transconjugants were selected on MM agar
(L6pez-Lara e t al., 1993). Antibiotics were used at the following
concentrations : kanamycin sulfate (Km), 180 pg ml-' ; tetracycline (Tc), 10 pg ml-'. Plasmid isolation, random primer
DNA labelling, DNA hybridization, and Southern blotting
were performed according to standard procedures (Maniatis e t
al., 1982).
Cloning of the GRHZ-14 and GRHZ-50 Tn5 insertions. The
total DNA of mutants GRH2-14 and GRH2-50 was EcoRI
digested and cloned in plasmid pSUP102. After ligation, D N A
was transformed in E. coli DH5a (Bethesda Research Laboratories). Km-resistant colonies were selected and analysed for
plasmid content. Plasmid clones with the T n 5 insertion were
designated pIS14 and pIS50.
DNA sequencing. The left junctions of Tn5insertions 14 and 50
were cloned as Sall fragments in pUC18 from plasmids pIS14
and pIS50, respectively. The sequence data were obtained by the
chain-termination method (Sanger e t al., 1977) using the ISSOL
specific primer 5'-AAAGGTTCCGTTCAGGACGC-3'.
Sequence analysis. Sequence analysis and homology search
were done with the GCG software package (Devereux e t al.,
1984). The amino acid sequences deduced from the nucleotide
sequences were compared with the actual version in the
Swissprot and PIR Protein Data Banks using the program FASTA
(Pearson & Lipman, 1988).
LPS analyses by SDS-PAGE. Cell cultures (1.5 ml) were pelleted,
suspended in SDS sample buffer (Laemmli, 1970), denatured at
100 "C for 3 min, and treated with proteinase K. After
discontinuous SDS-PAGE, gels were stained by periodic
acid/silver for carbohydrates (Hitchcock & Brown, 1983).
Isolation and fractionation of the EPS components. The
native EPS was isolated and fractionated as previously described
(Lopez-Lara e t al., 1993). Alternatively, EPS was obtained from
bacteria grown on solid medium. Cells were suspended in
10 mM sodium phosphate buffer (5-7 ml per plate), and pelleted
by centrifugation. EPS was isolated by precipitation of the
culture supernatant with 2 vols cold ethanol. The precipitated
EPS was redissolved in water and then dialysed against water
for 48 h at 4 "C.
Compositional and spectroscopic analyses of the EPS. Sugar
composition, glycosidic linkage determination and 'H-NMR
analyses of the native EPS were carried out as previously
described (Lopez-Lara e t al., 1993).
Nodulation assays. Plants were grown on nitrogen-free medium (Rigaud & Puppo, 1975) under microbiologically controlled conditions and inoculated as described by Olivares e t al.
(1980). Nodulation of tree legumes and herbaceous legumes was
followed for 6 months and 1 month, respectively.
Table 1. Strains used or constructed in this work
Strain
Rhixobium sp.
GRH2
Source/reference
Broad-host-range Rhixobium sp. isolated
from A . ganopbylla root nodules
GRH2 derivative obtained by Tn5
mutagenesis, exo-57, EPSGRH2 derivative obtained by T n i
mutagenesis, exo- 14 with altered LPS
GRH2 derivative obtained by Tn5
mutagenesis, exo-50, lacks the
UDPglucose 4-epimerase
Herrera e t al. (1985)
R. meliloti
Rm2011
Wild-type, SmR
C a d e t al. (1979)
R. leguminosarum
VF39
RS1050
Wild-type, bv. viciae, SmR
Wild-type, bv. triflii
8002
Wild-type, bv. phaseoli
Priefer (1989)
F. Rodriguez-Quiiiones,
Facultad Farmacia,
Universidad de Sevilla,
Spain
Lamb e t al. (1982)
B. japonicum
122DES
Wild-type, Hup'
Emerich et al. 1980)
A . tumefaciens
A348
A136 containing pTiA6NC
Garfinkel & Nester
(1980)
GRH2-57
GRH2-14
GRH2-50
574
Relevant characteristics
Lopez-Lara et al. (1993)
This work
This work
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
Rhixobitcm sp. LPS mutants and symbiotic behaviour
Enzyme assay for UDPglucose 4-epimerase. Bacteria were
grown on MM medium containing glucose or galactose as sole
carbon source. Cells were disrupted by sonication and debris
removed by centrifugation. Cell-free extracts were used to
measure the UDPglucose 4-epimerase activity according to
Fukasawa e t al. (1980).
Plant microscopy. Root nodules were processed as previously
described (Lopez-Lara et al., 1993). Two micrometre sections
were stained with alkaline toluidine blue solution and examined
by light microscopy. Ultrathin sections were stained with uranyl
acetate and lead citrate, and examined by transmission electron
microscopy using a Philips CM-10 electron microscope. Computer-assistedimage analysis of nodule histology was performed
using Bioquand System IV software as described previously
(Dazzo & Petersen, 1989).
RESULTS
Isolation and characterization of Rhizobium sp. GRH2
surface polysaccharide mutants
RhixobiZlm sp. GRH2 was mutagenized with Tn5: :mob,
and Kmr transconjugants were selected on MM agar with
mannitol as carbon source. By screening 10000 random
T n 5 insertion mutants, colonies with a non-mucoid,
supermucoid or a rough morphology were selected
visually (Lopez-Lara e t al., 1993). Two of the mutant
strains with a rough colony morphology were designated
GRH2-14 and GRH2-50. A Southern blot of EcoRIdigested total DNA of GRH2-14 and GRH2-50 mutant
strains, probed with Tn5, showed a single insertion
located in 2.4 and 3-2 kb EcoRI fragments, respectively
(data not shown). In addition, Southern blotting an
electrophoretic gel of Eckhart lysates from these mutants
using T n 5 probe indicated that the T n 5 insertions were
located on the chromosome, not on any of the five large
plasmids (Toro e t al., 1984). T o determine whether the
T n 5 insertions were responsible for the altered colony
morphology exhibited by the mutants, the insertions were
cloned as EcoRI fragments (Fig. la) in the pSUP102
vector and transferred back to the wild-type strain GRH2.
Transcon jugants showed the same colony morphology to
that of GRH2-14 and GRH2-50 mutant strains indicating
that the T n 5 insertion was responsible for the rough
colony phenotype. These results were further confirmed
by hybridization of EcoRI-digested total DNA from
RS
R
k1
I
1
2
3
4
5
Fig. 2. LPS profiles on an SDS-polyacrylamide gel. Periodic
acid/siIver-stainedSDS-polyacrylamide gel (12 %) of proteinaseK-treated SDS extracts of Rhizobiurn sp. GRH2 (lane l), EPSGRHZ-57 (lane 2), GRH2-50 (lane 3), GRH2-50 complemented
with plasmid p501 (lane 4) and GRH2-14 (lane 5).
GRH2 and mutants using the mutant cloned restriction
fragments as D N A probes.
The rough colony phenotype exhibited by GRH2-14 and
GRH2-50 strains suggested that they were Ips mutants.
Proteinase-K-digested cell extracts of wild-type and
mutant strains were separated by SDS-PAGE and stained
by the periodic acid/silver procedure. As shown in Fig. 2,
the relative staining intensity of the slow-migrating
carbohydrate band of ‘smooth’ LPS present in the wildtype was considerably reduced in the mutants. Apparently, the GRH2-14 and GRH2-50 mutants have an
alteration in the smooth LPS, due likely to a loss in 0antigen content. Reintroduction of the corresponding
T n 5 insertions into the wild-type GRH2 also reproduced
the altered LPS profile.
Complementation of Rhizobium sp. GRH2-14 and
GRH2-50 mutants
(b) R S
t l
H
I
RSS
I.
I
RS
I
1
B
I
R
.
I
p501
1 kb
Fig. 1, (a) Restriction fragments containing the Tn5 insertion of
GRH2-14 and GRH2-50 mutant strains. (b) Restriction map of
GRH2-50 complementing plasmid p501. R, EcoRI; B, BarnHI; HI
Hindlll; 5, Sall.
We attempted to restore the wild-type phenotype of
GRH2-14 and GRH2-50 mutants by carrying out a
conjugative mating with a wild-type GRH2 cosmid
(pLAFR1) clone bank (Lopez-Lara e t al., 1993). Mucoid
colonies, occurring at a frequency of 0.1 %, were obtained
only in the case of GRH2-50 and no complementation was
achieved for the GRH2-14 mutant strain. Cosmids
isolated from the mucoid colonies of GRH2-50 deriva-
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
575
I. M. L O P E Z - L A R A a n d O T H E R S
tives were identical and designated p501. A restriction
map of cosmid p501 is shown in Fig. l(b). Upon
reintroduction of cosmid p501 into mutant GRH2-50, the
wild-type mucoid phenotype was restored, which indicated that this phenotype was encoded by plasmid p501.
Hybridization of p501 with the EcoRI Tn5-mutant
fragment isolated from GRH2-50 further confirmed that
the DNA inserted in p501 carries the wild-type DNA that
was mutated by the T n 5 insertion. In addition, plasmid
p501 also restored the wild-type LPS profile (Fig. 2, lane
4) of mutant GRH2-50 but not of mutant strain GRH214.
GRHZ
GRH2-50
Identification of DNA homologous to GRH2-14 and
GRH2-50 mutant loci in Rhizobium
Genomic DNA from various Rbixobiztm species was
examined for homology to the mutant loci of GRH2-14
and GRH2-50 by Southern blot hybridization. DNA
fragments homologous to the GRH2-50 DNA probe
(Fig. la) were found in R. legztminosarzlm bv. viciae, bv.
trzfolii and bv. pkaseoli but no homology was detected with
Rhixobatlm meliloti, Bradyrbixobitlm japonicztm or Agrobacteriztm tztmefaciens DNA. On the contrary, no DNA
fragments homologous to the GRH2-14 DNA probe
(Fig. la) were found in the four fast-growing rhizobia
analysed nor in B. japonicztm or A. tztmefaciens (data not
shown). In addition, no homology was detected with the
R. legztminosarmz strain VF39 Ips genes containing plasmid
pCos4 (Priefer, 1989).
Sequencing of Tn5 insertions
In order to better understand the phenotypes of GRH2-14
and GRH2-50 mutations, we determined the DNA
sequence of the corresponding left Tn5-DNA junctions.
The partial amino acid sequences deduced were compared
with sequences in the Swissprot and PIR Data Banks. N o
homologous sequences were found to the predicted
product encoded by the mutant locus of GRH2-14. O n
the other hand, a significant homology (up to 43%) to
GalE protein (UDPglucose 4-epimerase) from different
organisms was found for the putative protein encoded by
the GRH2-50 mutant locus. This result suggested that
strain GRH2-50 was an exoB mutant.
Analysis of the UDPglucose 4-epimerase activity in
wild-type GRHZ and derivative strains
To further confirm that strain GRH2-50 was an exoB
mutant, cell-free extracts from wild-type GRH2 and the
mutant GRH2-50 were assayed for the presence of
UDPglucose 4-epimerase activity. Data indicated that this
enzyme activity was present in the wild-type [99 nmol
substrate used min-' (mg protein)-'] but absent in the
GRH2-50 mutant strain. Similar values were obtained for
cells grown with glucose or galactose as the sole carbon
source indicating a constitutive expression. In the cell-free
extract prepared from GRH2-50 complemented with
cosmid p501,the enzymic activity was recovered [98 nmol
substrate used min-' (mg protein)-'].
576
5
4
3
p.p.m.
2
1
Fig. 3. 'H-NMR spectra of the native acidic EPS from GRH2,
GRH2-50 and GRHZ-50 complemented with plasmid p501. 3-HB,
3-Hydroxybutyryl substituent; Ac, acetate; Pyr, pyruvyl
substituent.
Structural evaluation of the acidic
heteropolysaccharides from GRH2-14, GRH2-50 and
GRH2-50(~501)
Gel filtration analyses indicated that while GRH2-14
produced a similar amount of EPS to the wild-type
GRH2, the yield of EPS produced by mutant GRH2-50
was fivefold lower. Based on sugar composition and 'HNMR spectral features, the high molecular mass acidic
EPS made by the insertion-mutant GRH2-14 was indistinguishable from the wild-type GRH2 acidic EPS,
which consists of an octasaccharide-repeat-unit-based
polymer with glucose, glucuronic acid and galactose
glycosyl components in a 5 :2 : 1 ratio, and acetyl, pyruvyl
and 3-hydroxybutyryl substitutions (Lopez-Lara e t al.,
1993).
In contrast, 'H-NMR analysis indicated that GRH2-50
EPS was altered in both glycosidic and non-carbohydrate
substitutions composition (Fig. 3). The spectrum of this
EPS was totally devoid of 3-hydroxybutyryl resonances
and displayed a single pyruvyl resonance whose relative
integration indicated that GRH2-50 EPS bears a single
pyruvate residue per repeat unit, as compared to two
residues in GRH2 EPS. Compositional analysis showed
that GRH2-50 polysaccharide contains glucose and glucuronic acid in an approximate 3: 1 ratio similar to the
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
Rhipbitlm sp. LPS mutants and symbiotic behaviour
Fig. 4. Longitudinal sections of an effective root nodule of A. cyanophy//a infected with wild-type Rhizobium sp. (Acacia)
strain GRH2 examined by light microscopy (a) and transmission electron microscopy (b). Note the numerous infected host
cells interspersedwithin the large central zone in (a), and a portion of an infected host cell with numerous symbiosomes
containing pleomorphic bacteroids surrounded by peribacteroid membranes in (b). Scale bars: (a) 100 pm; (b) 1 pm. CW,
Cell wall; B, bacteroid; PM, peribacteroid membrane.
wild-type EPS, but no trace of a galactosyl component
was detected.
1H-NMR spectroscopy and compositional analyses indicated that the introduction of plasmid p501 into the
GRH2-50 background resulted in restoration of the
production of the wild-type EPS. The 'H-NMR spectrum
of GRH2-50 (p501) EPS was superimposable with that of
the GRH2 EPS for both glycosidic and non-carbohydrate
substitutions resonances (Fig. 3). Assuming a stoichiometry of two pyruvyl residues per GRH2 EPS repeat
unit (Lopez-Lara e t a/., 1993), the pyruvate/acetate/3hydroxybutyrate ratio was 2-00: 1-62:0.84 for the EPS of
the complemented mutant, as compared to 2-00: 1.56 :0.78
in the wild-type EPS. In addition, both the EPS sugar
ratios were similar (glucose/glucuronic acid/galactose
approximately 5 :2 : l), showing the restoration of the
galactose moiety as a component of the EPS of the
strain.
The structures of GRH2-50 and GRH2-50 (p501) EPS
were further examined by GC/MS analysis of the
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
577
I. M. LOPEZ-LARA a n d O T H E R S
Fig. 5. Longitudinal sections of an ineffective root nodule of A. cyanophylla induced by Ips mutant GRH2-14, examined
by light microscopy (a and b) and transmission electron microscopy (c and d). Scale bars: (a) 100 pm; (b) 20 pm; (c) 10 pm;
(d) 5 pm. NP, Nodule parenchyma; CT,central tissue; V, vacuole; GM, electron-dense granular material.
glycosidic linkage composition. The analysis of GRH2-50
(p501) EPS showed the predominance of 4-linked glucose,
and the occurrence of 4-linked glucuronic acid, 4,6-linked
glucose, 4,6-linked galactose and 3,4,6-linked glucose.
This pattern of partially methylated alditol acetates was
similar to the one obtained from the GRH2 wild-type
EPS (Lopez-Lara e t a/., 1993). In the GRH2-50 EPS, 4linked glucose was also predominant ;4-linked glucuronic
acid and 4,6-linked glucose were also present but no other
derivative was found. This result reinforced the above
NMR and GC/MS data indicating that the GRH2-50
57%
mutation results in an altered EPS lacking the galactose
component, as well as one pyruvyl moiety and 3hydroxybutyryl substitutions, whereas the introduction
of the p501 plasmid into the GRH2-50 background
restores the production of the wild-type EPS.
Symbiotic phenotype of GRH2-14 and GRHZ-50
mutant strains
Wild-type strain GRH2 is able to elicit reddish, elongated
root nodules within 20 d of inoculation on A . cyanupbylla,
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
Rhizobim sp. LPS mutants and symbiotic behaviour
Fig. 6. Longitudinal sections of an ineffective root nodule of A. cyanophylla induced by Ips mutant GRH2-50, examined
by light microscopy (a and b) and transmission electron microscopy (c and d). Scale bars: (a) 100 pm; (b) 20 pm; (c) 2 pm;
(d) 2 pm. NP, Nodule parenchyma; CT, central tissue; V, vacuole; GM, electron-dense granular material.
Acacia melanoxjlon and Prosopis chilensis plants. Inoculation of these tree-legumes with GRH2-14 and GRH2-50
mutant strains resulted in alteration of their root system
leading to shorter and thicker roots without the formation
of reddish, elongated nodules. Occasionally, inoculated
roots of tree legumes developed a few small white nodules
after 1 month of incubation with either of these two
mutant strains. Reintroduction of the corresponding T n 5
insertions into the wild-type GRH2 also reproduced the
altered symbiotic phenotype. Combined light and transmission electron microscopy of median longitudinal
sections through 1-month-old A. cyanoph~lla nodules
indicated that many host cells in the central tissue were
invaded by the wild-type strain GRH2, which developed
into pleiomorphic endosymbiotic bacteroids within symbiosomes (Fig. 4a, b). In contrast, Acacia nodules were
not invaded by either GRH2-14 or GRH2-50 mutant
strains (Figs 5a-c and 6a-c). Other distinguishing mor-
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
I. M. LOPEZ-LARA a n d O T H E R S
phological features of median sections through ' empty '
nodules induced by GRH2-14 and GRH2-50 mutant
strains were a smaller proportional area of the central
tissue (7 % and 15 %, respectively, compared to 59 YOfor
nodules induced by wild-type GRH2) and a more
expansive nodule parenchyma between the peripheral
nodule cortex and the central tissue. Nodule cells in this
layer surrounding the central tissue stained intensely with
toluidine blue (Figs 4a, 5a and 6a) and contained large
vacuoles in which was deposited electron-dense granular
material that was not bacteria (Figs 5d and 6d). Neither
GRH2-14 nor GRH2-50 mutant strains were able to
induce nodules on the herbaceous legumes which are
nodulated by the GRH2 wild-type strain, such as Lottls
corniculattls, Vicia kirstlta, Triflitlm stlbterraneum, Tr;folium
repens and Triflitlm incarnattlm. However, like wild-type
GRH2, mutants GRH2-14 and GRH2-50 elicited nodules
on Phaseoltls vdgaris.
DISCUSSION
Broad-host-range. rhizobia are very useful for investigating the molecular mechanisms of symbiotic interactions between bacteria and plants. For this study, we
have used wild-type Rhixobitlm sp. strain GRH2, which
was isolated from root nodules of A. cyanoplylla and has a
broad host range which includes legume trees and
herbaceous legumes (Herrera et al., 1985). Here we report
the characterization of two mutant strains of GRH2
which are defective in production of smooth LPS. One of
them, strain GRH2-50, is an exoB mutant which lacks
UDPglucose 4-epimerase activity. A consequence of this
mutation is the inability to produce UDPgalactose and
thus it is defective in production of glycoconjugates
containing galactosyl components. In a previous study
(L6pez-Lara e t al., 1993), we found that the acidic EPS of
wild-type GRH2 is closely related to the EPS of wild-type
R. legtlminosartlm bv. trifalii ANU843 (Hollingsworth e t
al., 1988). The simultaneous lack of galactose, one pyruvyl
residue, and the 3-hydroxybutyryl substituents in the EPS
of the GRH2-50 exoB mutant is best explained by
proposing that the EPS of wild-type GRH2, like wildtype ANU843, hai a galactose residue at the terminus of
the branch in each oligosaccharide repeat unit which is
substituted with one pyruvate residue and is also the site
of esterification of 3-hydroxybutyrate substitutions. The
inability of the GRH2-50 exoB mutant to produce smooth
LPS implies that UDPgalactose is also required for
synthesis of this cell surface polysaccharide. Preliminary
studies of GRH2 LPS suggest that the sugar composition
is different from that of strain ANU843, and that the LPS
of mutant GRH2-50 lacks galactose and contains galacturonic acid instead (S. Hollingsworth, personal communication).
Attempts to complement the mutant phenotypes exhibited by GRH2-14 d GRH2-50 with a wild-type
cosmid bank were only successful with the latter mutant
strain. Southern blot hybridization and complementation
studies showed that both GRH2-50 and GRH2-14 mutant
strains carry a single T n 5 insertion. Reintroduction into
580
the wild-type strain GRH2 of the corresponding T n 5
insertions reproduced the colony morphology, LPS and
symbiotic mutant phenotypes. Thus, the lack of complementation of strain GRH2-14 suggests that: (i) the
wild-type DNA was not well represented in the genomic
bank used, (ii) the mutation could be dominant, or (iii)
several copies of the wild-type DNA region could be
lethal. Whereas the GRH2-50 mutant locus is well
conserved among R. legtlminosartlm strains, GRH2-14
mutant DNA did not show homology in these rhizobial
strains nor with pCos4. In addition, sequence analysis of
the left junction of the T n 5 insertion 14 did not reveal
homologous sequences in the Swissprot and PIR Protein
Data Banks. Therefore, the GRH2-14 mutant locus may
represent a newly described Ips gene for Acacia rhizobia.
In a previous study we reported an EPS mutant derivative
of GRH2 (mutant GRH2-57) which was partially defective in nodule invasion in Acacia. It still nodulated this
host but exhibited a fivefold reduction in nodule occupancy (Lopez-Lara e t al., 1993). The LPS mutant strains
described here (GRH2-14 and GRH2-50) have even more
drastic defects in their symbiotic phenotypes on this tree
legume host. These mutant strains are only able to induce
a few small white nodules in Acacia, and these nodules are
devoid of bacteria. Therefore, whereas a mutation which
blocks acidic EPS biosynthesis results in a partial loss in
ability to fully invade Acacia nodules, a mutation which
blocks biosynthesis of smooth LPS completely prevents
bacterial invasion of nodule cells on this preferred tree
legume host. This demonstrates the critical importance of
LPS and acidic EPS in enabling rhizobia to overcome the
host cell barriers in order to achieve nodule invasion in
the nitrogen-fixing Rhixobitlm-Acacia symbiosis.
Nodulation tests with a diversity of legumes have revealed
that these polysaccharide mutants of Acacia rhizobia are
also altered in host range. In a previous study (Lopez-Lara
e t al., 1993), the EPS mutant GRH2-57 lost the ability to
nodulate clovers and vetch, but retained the ability to
nodulate Phaseoltls and Lottls. Therefore, the importance
of EPS in nodulation and host range on herbaceous
legume hosts is correlated with the type (determinant or
indeterminant) of nodule ontogeny. In contrast, the LPS
mutant strains GRH2-14 and GRH2-50 were able to
nodulate Phaseoltls but not Lottls. These results suggest
that the symbiotic defect exhibited by these two LPS
mutants on herbaceous legume hosts was not dependent
on the nodule ontogeny. Thus, our results highlight the
contributing role(s) of the surface polysaccharides (LPS
and acidic EPS) in determining the host range of Acacia
rhizobia.
ACKNOWLEDGEMENTS
This work was supported by Comision Asesora de Investigacion
Cientifica y Ticnica grants BI090-0747 and BI093-0677 (to
J.O. and N.T.), National Institute of Health Grant no.
GM-34331-06 and the Michigan State University Biotechnology Research Center (to F. B. D.). I. L-L. was supported by a
M. E. C. fellowship. We thank H. S. Pankratz and J. Palma for
technical assistance.
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
Rhipbitlm sp. LPS mutants a n d symbiotic behaviour
REFERENCES
Abe, M., Sherwood, 1. E., Hollingsworth, R. 1. & Dazzo, F. B.
(1984). Stimulation of clover root hair infection by lectin-binding
oligosaccharides from the capsular and extracellular polysaccharides
of Rhixobium trfolii. J Bacteriol160, 517-520.
Beringer, J. E. (1974). R factor transfer in Rhixobium leguminosarum.
J Gen Microbiol84, 188-1 98.
Canter Cremers, H. C. J., Batley, M., Redmond, J. W., Eydems, L.,
Breedveld, M. W., Zevenhuizen, L. P. T. M., Pees, E., Wijffelman,
C. A. & Lugtenberg, B. J. 1. (1990). Rhixobizlm leguminosarum exoB
mutants are deficient in the synthesis of UDP-glucose 4'-epimerase.
J Biol Chem 265,21122-21 127.
Carlson, R. W. (1982). Nitrogen fixation. Rhixobium. In Surface
Chemisty, vol. 2, pp. 199-234. Edited by W. J. Broughton. Oxford :
Oxford University Press.
Cass6, F., Boucher, C., Julliot, 1. S., Michel, M. & Denari6, J. (1979).
Identification and characterization of large plasmids in Rhixobium
meliloti using agarose gel electrophoresis. J Bacteriol113, 229-242.
Dazzo, F. B. & Petersen, M. (1989). Applications of computerassisted image analysis for microscopic studies of the Rhixobiumlegume symbiosis. Symbiosis 7 , 193-210.
Dazzo, F. B., Truchet, G. L., Hollingsworth, R. I., Hrabak, E. M.,
Pankratz, H. S., Philip-Hollingsworth, S., Salzwedel, J. L., Chapman, K., Appenzeller, L., Squartini, A., Gerhold, D. & Orgambide,
G. (1992). Rhixobium lipopolysaccharide modulates infection thread
development in white clover root hairs. J Bacterioll73, 5371-5384.
Devereux, 1.. Haeberli, P. & Smithies, 0. (1984). A comprehensive
set of sequence analysis programs for the VAX. Nucleic Acids Res
12, 387-395.
Emerich, D. W., Ruiz-ArgUeso, T., Russell, 5. A. & Evans, H. J.
(1980). Investigation of the H, oxidation system in Rhixobium
japonicum 122DES nodule bacteroids. Plant Physiol66, 1061-1 066.
Fukasawa, T., Obonai, K., Segawa, T. & Nogi, Y. (1980). The
enzymes of the galactose cluster in Saccharomyces cerevisiae. J Biof
Chem 255, 2705-2707.
Garfinkel, D. 1. & Nester, E. W. (1980). Agrobacterium tumefaciens
mutants affected in crown gall tumorigenesis and octopine catabolism. J Bacteriol 144, 732-743.
Herrera, M. A., Bedmar, E. J. & Olivares, J. (1985). Host specificity
of Rhiqobium strains isolated from nitrogen-fixing trees and
nitrogenase activities of strain GRH2 in symbiosis with Prosopis
chilensis. Plant Science 42, 177-1 82.
Hitchcock, P. 1. & Brown, T. M. (1983). Morphological hetero-
geneity among Salmonella lipopolysaccharide chemotypes in silverstained polyacrylamide gels. ] Bacteriol154, 269-277.
Hollingsworth, R. I., Dazzo, F. B., Hallenga, K. & Musselman, B.
(1988). The complete structure of the Trifoliin A lectin-binding
capsular polysaccharide of Rhixobium triflii 843. Carbohydr Res 172,
97-1 12.
Laemmli, U. K. (1970). Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680-685.
Lamb, J. W., Hombrecher, G. &Johnston, A. W. B. (1982). Plasmid-
determined nodulation and nitrogen fixation abilities in Rhixobium
phaseoli. Mol & Gen Genet 186, 444452.
L6pez-Lara, M. I., Orgambide, G., Dazzo, F. B., Olivares, J. & Toro,
N. (1993). Characterization and symbiotic importance of acidic
extracellular polysaccharides of Rhixobium sp. strain GRH2 isolated
from Acacia nodules. J Bacteriol 175, 2826-2832.
Maier, R. J. & Brill, W. 1. (1978). Involvement of Rhixobium
japonicum 0 antigen in soybean nodulation. J Bacteriol 133,
1295-1299.
Maniatis, T., Fritsch, E. F. & Sambrook, 1. (1982). Molecular Cloning:
a Laboratoy Manzlal. Cold Spring Harbor, NY : Cold Spring Harbor
Laboratory.
MUller, P., Hynes, M., Kapp, D., Niehaus, K. & PUhler, A. (1988).
Two classes of Rhizobium meliloti infection mutants differ in
exopolysaccharide production and in coinoculation properties with
nodulation mutants. Mol & Gen Genet 211, 17-26.
Noel, K. D., Vandenbosch, K. A. & Kulpaca, B. (1986). Mutations in
Rhixobium phaseoli that lead to arrested development of infection
threads. J Bacterioll68, 1392-1401.
Olivares, J., Casadesus, 1. & Bedmar, E. J. (1980). Methods for
testing degree of infectivity of Rhixobium meliloti strains. Appl
Environ Microbiol39, 967-970.
Pearson, W. R. & Lipman, D. J. (1988). Improved tools for
biological sequences comparison. Proc Natl Acad Sci U S A 85,
2444-2448.
Priefer, U. B. (1989). Genes involved in lipopolysaccharide production and symbiosis are clustered on the chromosome of
Rhixobium leguminosarum biovar viciae VF39. J Bacteriol 171,
6161-6 168.
Reuber, T. L., Reed, 1. W., Glazebrook, J., Glucksmann, A. M.,
Ahman, D., Marra, A. & Walker, G. C. (1991). Rhixobium meliloti
exopolysaccharides : genetic analysis and symbiotic importance.
Biochem Soc Trans 19, 636-641.
Rigaud, J. & Puppo, A. (1975). Indole-3-acetic acid catabolism by
soybean bacteroids. J Gen Microbiol88, 223-228.
Robertsen, B. K., Aiman, P., Darvill, A. G., McNeil, M. & Albersheim, P. (1981). The structure of acidic extracellular polysaccharides
secreted by Rhixobium leguminosarum and Rhixobium trifalii. Plant
Physi0l67, 389-340.
Sanger, F., Nicklen, 5. 81Coulson, A. R. (1977). DNA sequencing
with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74,
5463-5467.
Toro, N., Herrera, M. A. & Olivares, J. (1984). Location of nifgenes
on large plasmids in Rhixobium strains isolated from legume tree
root nodules. F E M S Microbiol Lett 24, 113-1 15.
Zhan, Y., Hollingsworth, R. 1. & Priefer, U. (1992). Characterization
of structural defects in the lipopolysaccharides of symbiotically
impaired Rhixobium leguminosarum bv. viciae VF-39 mutants. Carbobydr Res 231, 261-271.
Received 21 June 1994; revised 11 October 1994; accepted 26 October
1994.
Downloaded from www.microbiologyresearch.org by
IP: 78.47.27.170
On: Fri, 14 Oct 2016 11:48:26
58 1