Download Applied and Environmental Microbiology

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, JUlY 1976, P. 166-171
Copyright © 1976 American Society for Microbiology
Vol. 32, No. 1
Printed in U.S.A.
Adsorption of Bacteria to Roots as Related to Host Specificity
in the Rhizobium-Clover Symbiosis'
FRANK B. DAZZO,*2 CAROLYN A. NAPOLI, AND DAVID H. HUBBELL
Departments of Microbiology and Soil Science, University of Florida, Gainesville, Florida 32611
Received for publication 10 December 1975
Recently, it has been established that there
are cell surface polysaccharides unique to Rhizobium trifolii that are cross-reactive with antigens on the outer surface of clover roots (5).
Analytical studies of the purified polysaccharide cross-reactive antigen from R. trifolii
strain 403 (infective on clover) indicated that it
is a high-molecular-weight (>4.6 x 106), amorphous, 8-linked, acidic heteropolysaccharide
containing 2-deoxyglucose, galactose, glucose,
and glucuronic acid. A multivalent lectin with
2-deoxyglucose sugar specificity capable of
binding to this unique antigen was extracted
from Trifolium repens (white clover) seeds.
This lectin agglutinated cells of numerous infective R. trifolii strains (including infective
revertants) but not several noninfective R. trifolii strains or strains of R. japonicum, R. meliloti, R. leguminosarum, R. phaseoli, and several R. species (cowpea miscellany). Based on
these data, a model was proposed to explain
host specificity in the R. trifolii-Trifolium symbiosis based on a preferential adsorption of infective bacterial cells on the surface of clover
roots by a cross-bridging of their common surface antigens with the multivalent clover lectin. This model differs from other hypotheses
(1, 7) concerning the possible role of lectins in
the specific adsorption of rhizobia in that it
places equal emphasis on the lectin and the
unique, complementary surface polysaccharides of the infective bacteria and the appropriate legume root cells.
The purpose of this study was to investigate
the adsorption ofRhizobium to roots as a possible expression of host specificity in the establishment of the Rhizobium-clover symbiosis.
MATERIALS AND METHODS
Rhizobium cultures. The formation of infection
threads in root hairs of the appropriate legume host
was used to verify the infectiveness of R. trifolii and
R. meliloti strains. The slide culture method of Fahraeus (6) was used for this determination. The R.
trifolii strains were obtained from sources previously described (5). R. meliloti strains F28 and F51
were obtained from J. C. Burton, The Nitragin Co.,
Milwaukee, Wis. R. meliloti strain 142 and R. ja-
ponicum strain 61A76 were obtained from W. J.
Brill, University of Wisconsin, Madison, Wis. Unless otherwise indicated, all strains of Rhizobium
were cultured on a chemically defined medium (4).
The effect of 2-deoxyglucose on the growth rate ofR.
trifolii strain 403 was determined during exponential growth in 500-ml side-arm flasks containing 50
ml of this medium (4). The flasks were shaken at 220
' Florida Agricultural Experiment Station Journal Se- rpm and 30°C, and growth was monitored by a KlettSummerson colorimeter (640 nm). Filter-sterilized
ries no. 8030.
2
Present address: Department of Bacteriology and Cen- 2-deoxyglucose was added to two of the four flasks at
ter for Studies of Nitrogen Fixation, University of Wiscon- mid-exponential phase (30 mM final concentration),
sin, Madison, Wis. 53706.
and then cell growth was monitored for 6 h. Floccu166
Downloaded from http://aem.asm.org/ on October 17, 2015 by guest
Quantitative microscope techniques were utilized to examine the adsorption of
rhizobial cells to clover root hairs. Adsorption of cells of noninfective strains of
Rhizobium trifolii or infective R. meliloti strains to clover root hairs was four to
five times less than that of the infective R. trifolii strains. Attachment of the
rod-shaped bacteria to clover root cells occurred in a polar, end-on fashion.
Viable or heat-killed R. trifolii cells precoated with a clover lectin having 2deoxyglucose specificity had increased adsorption to clover roots. Adsorption of
bacteria to roots was not increased if the clover lectin was inactivated by heat or
2-deoxyglucose treatment prior to incubation with R. trifolii. Adsorption of R.
trifolii to clover root hairs was inhibited by 2-deoxyglucose (30 mM) but not by 2deoxygalactose or a-D-glucose. Adsorption of R. meliloti cells to alfalfa root hairs
was not affected by 2-deoxyglucose at that concentration. These results suggest
that expression of host specificity in the Rhizobium-clover symbiosis involves a
preferential adsorption of infective cells to clover root hairs through a 2-deoxyglucose-sensitive receptor site.
VOL. 32, 1976
ADSORPTION OF R. TRIFOLII TO ROOTS
167
c
Downloaded from http://aem.asm.org/ on October 17, 2015 by guest
lation of R. trifolii strain 403 cells (13) did not occur tative adsorption of the bacteria to root hairs. The
until the culture was in the stationary phase of adsorption of R. meliloti strain F28 to alfalfa root
growth.
hairs in the presence or absence of 2-deoxyglucose
Assay for adsorption of bacteria to roots. Fah- (30 mM) was also examined. This concentration of 2raeus slide assemblies (6) of sterile T. repens L.
deoxyglucose inhibited the agglutination ofR. trifovariety Louisiana Nolen (white clover) or Medicago lii cells by either the clover lectin or the crosssativa var. Florida 66 (alfalfa) seedlings without reactive anti-white clover root antibody (5).
agar were inoculated with approximately 1.9 x 108
rhizobial cells. Cell numbers were determined by
RESULTS AND DISCUSSION
direct microscope counting in a Petroff-Hausser
Adsorption of rhizobia to roots. The model
chamber. Slide assemblies were incubated in a
growth chamber (Warren Sherer, Marshall, Mich.; utilized to explain host specificity in the R.
model CEL 255-6) programmed at 22°C isothermal, trifolii-Trifolium symbiosis (5) requires that
12-h photoperiod, and a light intensity of 2,160 lux. there be a preferential adsorption of infective
After 12 h of incubation (10 h of darkness followed by cells to clover root hairs when compared with
2 h of light), the cover slips were removed, and the noninfective R. trifolii cells or other Rhizobium
roots were washed on the slides by a gentle stream of species incapable of infecting this host. This
saline (0.15 M NaCl). New cover slips were placed criterion was fulfilled when adsorption of bacteover the roots, and the assemblies were filled with
to clover root hairs was evaluated by direct
saline and then examined by phase-contrast micros- ria
microscopy.
The mean number of R. trifoiji cells
copy. Four optical median planes of two primary
roots per bacterial strain were evaluated. The opti- adsorbed to root hairs of approximately 200 ,tm
cal median plane is parallel to the microscope slide in length after 12 h of incubation are presented
and transects through the center of the root cylin- in Table 1. When pooled together, the mean
der. Root hairs along the optical median plane are number (+ standard deviation) of adsorbed inequally distant from the cover slip and the micro- fective and noninfective cells were 25.8 + 5.9
scope slide, and therefore any possible effects of the
and 5.6 + 1.9, respectively. The populations
glass surfaces on the adsorption of the bacteria to were compared statistically using the t tests
the roots would be minimized in this system. The
number of bacterial cells of each strain which were computed after the square root transformation
adsorbed to root hairs approximately 200 ,um in of the means (19). For every strain combination
length was determined, and the data were evaluated examined, the mean number of infective cells
statistically. Approximately 15 to 20 root hairs of adsorbed per root hair examined was signifithis length could be found in four optical median cantly greater (at the 99.5% level) than the
planes. In addition, some Fahraeus slide cultures corresponding noninfective cells. Of particular
were incubated for 4 days for photomicrography.
importance was the finding that the infective
Inoculation of T. repens seedlings with R. trifolii revertant strain BA-L had a significantly
cells coated with clover lectin. A saline suspension
of infective R. trifolii strain 403 containing 3.1 x 109
TABLE 1. Adsorption ofRhizobium cells to T. repens
cells per ml was divided into two aliquots of 5 ml
root hairsa
each. One aliquot received 20 ml of sterile saline,
Infective
Adsorbed
and the other received 20 ml of clover seed extract
Strain
on T. re- cells (means
Species
(5) containing the clover lectin (an end point agglu+ SD)b
pens
tinating titer of 16) previously filtered through a
+
R. trifolii
2S-2
21.00 + 1.00
0.20-,um sterile membrane filter (TCM-20, Gelman
R. trifolii
2L
2.00 + 1.73
Company, Ann Arbor, Mich.). The cells were incu+
R. trifolii
0435
22.50 ± 2.81
bated at 30°C for 30 min, washed three times with
6.67 ± 1.63
R. trifolii
0435-2
sterile saline, resuspended in saline, and used to
+
22.75 + 2.22
R. trifolii
T37
inoculate 2-day-old white clover seedling roots in
4.80 + 1.64
R. trifolii
Bio-9
slide assemblies. Seedling roots were inoculated
+
R. trifolii
WU-290-I
25.50 ± 4.12
with approximately 2 x 107 cells (0.25 ml) of either
R. trifolii
WU-290-N
8.00 + 2.45
+
R. trifolii
25.67 + 0.58
403
lectin-coated or uncoated R. trifolii cells. Lectin was
R.
A
5.33
Bart
+ 2.31
trifolii
detected on the surface of the coated cells directly by
+
BA-L
37.33 + 9.48
R. trifolii
cell agglutination after continued incubation and
R. trifolii
403c
3.50 + 1.05
indirectly by rabbit anti-lectin extract antiserum
F51
2.90 + 1.10
R. meliloti
using indirect immunofluorescence (5). Sterile sa2.44 + 1.01
R. meliloti
F28
as
a
uninoculated
plants
line (0.25 ml) was added to
2.56 + 1.30
R. meliloti
142
control. Washed slide assemblies were examined for
aFrom 15 to 20 root hairs were examined per treatment,
bacterial root adsorption after 12 h (10 h of darkness
each approximately 200 jAm in length.
followed by 2 h of light).
b Values for infective R. trifolii cells were significantly
Inhibition of bacterial adsorption to clover roots greater
than corresponding noninfective mutants at the
by 2-deoxyglucose. Fahraeus slide assemblies of R. 99.5% level. Values
for R. meliloti strains were significantly
trifolii strain 403 and white clover were prepared in less than pooled means of infective R. trifolii cells (25.80) at
the presence of 30 mM 2-deoxyglucose, 2-deoxyga- the 99.5% level. SD, Standard deviation.
Heat killed (80°C for 10 min).
lactose, or a-D-glucose and then assayed for quanti-
168
DAZZO, NAPOLI, AND HUBBELL
APPL. ENVIRON. MICROBIOL.
I
FIG. 1. Adsorption of R. trifolii 403 to a T. repens root hair. Bar
=
10 ,um.
FIG. 2. Adsorption of R. trifolii 403 to epidermal root cells of T. repens. Bar = 10 ,um.
Downloaded from http://aem.asm.org/ on October 17, 2015 by guest
greater frequency of adsorbed cells per root hair were not removed from roots by extensive shakthan the corresponding noninfective R. trifolii ing by hand or a Vortex mixer, and the data
strain Bart A. R. meliloti strains could adsorb were not analyzed statistically.
firmly to clover root hairs, but their mean numThe mean number (-+standard deviation) of
ber of adsorbed cells per root hair was signifi- four additional replicates of R. trifolii strain
cantly less thaxi the pooled means of infective 403 adsorbed on white clover root hairs were
R. trifolii cells at the 99.5% level. Thus, it can 24.25 ± 1.60 25.10 - 2.18, 24.90 ± 1.79, and
be concluded that the phenotypic trait of infec- 24.69 ± 1.74 and, therefore, this root adsorption
tivity for R. trifolii is directly correlated with assay yields very reproducible data.
the extent of their cell adsorption to root hairs
Polar attachment. Adsorption of the rodof the clover host. Heat-killed R. trifolii strain shaped bacteria occurred mostly through per403 cells could adsorb firmly to clover root pendicular attachment of one of their poles to
hairs, but quantitatively they were fewer in the root hair surface (Fig. 1). The bacteria also
numbers than the corresponding viable cells. attached to other clover root epidermal cells in
Studies by others (15) have indicated no corre- a polar fashion (Fig. 2). Polar attachment could
lation between the extent of root adsorption be observed for at least 20 days. When viewed
with the infective capabilities of the microorga- along the optical median plane of the epidermal
nisms. Indirect measurements involving cell root cells, the bacteria were attached perpencounts after their removal from roots by shak- dicular to their longitudinal axis and appeared
ing were utilized. Unfortunately, this latter rod shaped. When focused on top of the root
approach did not distinguish root hairs from cells, the bacteria appeared as spheres, since
other adsorptive root surfaces, nondispersable they were viewed parallel to their longitudinal
flocs from single cells, and the variability of axis. Polar orientation has been reported as the
root surface areas among individual plants. position of attachment of single cells of R. trifoFurthermore, many adsorbed Rhizobium cells lii (11, 12, 13, 17), R. meliloti (3), R.japonicum
VOL. 32, 1976
ADSORPTION OF R. TRIFOLII TO ROOTS
cum strain 61A76 cells were treated with the
clover lectin extract and then inoculated on alfalfa or white clover seedling roots, respectively. Additionally, the massive adsorption
was not observed if the clover lectin was inactivated by heat (80'C for 10 min) or 2-deoxyglucose (30 mM). Furthermore, pretreatment of R.
meliloti strains F51, F28, or 142 with an alfalfa
seed extract prepared in an identical fashion (5)
did not increase substantially the adsorption of
these cells to white clover roots (4.42 + 1.14
cells adsorbed per root hair of 200 Am in
length). These results indicate that the clover
root does not serve as an inert nonspecific support for flocculated growth.
Inhibition of bacterial adsorption to roots
by 2-deoxyglucose. The outer surface of gramnegative bacteria contains a mosaic of possible
antigenic determinants. Infective strains of R.
trifolii possess a surface acidic polysaccharide
containing a unique antigenic determinant not
detected immunochemically on the outer surfaces of other species of Rhizobium (5). This
unique antigen has sufficient structural similarity with an antigen on the surface of clover
roots to be considered cross-reactive. The 2deoxyglucose is the only sugar component of
the antigen that is capable of blocking its binding to the multivalent clover lectin (5). We
hypothesized that this sugar may interfere with
the bacterial adsorption to clover roots by occupying all the lectin binding sites. Lack of inhibition by this sugar would therefore indicate
independence of bacterial cell attachment and
lectin binding. The results indicated that free 2deoxyglucose (30 mM) but not 2-deoxygalactose
or a--glucose significantly inhibited (at 99%
level) the quantitative adsorption of R. trifolii
strain 403 cells to white clover root hairs. A 10fold reduction in the numbers of adsorbed cells
per root hair was observed (Table 2). The generation time of R. trifolii strain 403 during exponential growth did not change in the presence
of 30 mM 2-deoxyglucose (4.5 h), and therefore
neither stimulation nor inhibition of cell division was observed. Unlike the R. trifolii-clover
system, the quantitative adsorption of R. meliloti cells to alfalfa root hairs was not affected by
2-deoxyglucose (Table 2). Therefore, 2-deoxyglucose specifically inhibits the adsorption of R.
trifolii cells to clover root hairs. Subsequent
events, e.g., infection thread and nodule formation, might be nonspecifically inhibited by 2deoxyglucose, since this sugar can interfere
with the metabolism of Saccharomyces cerevisiae (8).
The data presented here suggest that expression of host specificity in the Rhizobium-clover
symbiosis involves a preferential adsorption of
Downloaded from http://aem.asm.org/ on October 17, 2015 by guest
(16), and Rhizobium sp. (Aeschynomene nodulating strain [14]) to their appropriate legume
host root surfaces.
The molecular basis for the preferred polar
attachment of Rhizobium cells to their appropriate legume host roots is unknown. This polar orientation has been observed also with bacteria attached to epithelial surfaces in the gastrointestinal tracts of mice (18) and Flexibacter
and Hyphomicrobium cells at solid-water and
oil-water interfaces (10). Alignment of cells of
Hyphomicrobium and Flexibacter perpendicular to oil-water and air-water interfaces suggested to Marshall and Cruickshank (10) that
attachment at one pole represented a rejection
of a hydrophobic portion of the cells from the
aqueous environment. Rosettes in aqueous suspensions were interpreted by them as micelles
with the hydrophobic ends of the cells attracted
to each other. Polar orientation could not be
explained by localization of surface ionogenic
groups (10), and both infective and noninfective
R. trifolii strains had equal electrophoretic mobilities (9). Recent studies have indicated that
certain R. japonicum strains bind soybean lectin mostly at one of their cell poles (2). Savage
(personal communication) has suggested that
end-on attachment would result in more microbial colonization per unit of surface area and
more microbial surface exposed to the environment for the organism to extract nutrients.
Studies with lectin-coated cells. The effect
of coating infective R. trifolii cells with the
clover lectin on their ability to adsorb to white
clover roots was examined. Cells precoated
with clover lectin greatly increased their ability to firmly adsorb to the white clover root
surface within 12 h. Photomicrographs of
washed, sterile seedling roots (control) and
washed roots incubated with viable R. trifolii
strain 403 cells with and without clover lectin
coating are shown in Fig. 3a, b, and c, respectively. Approximately 25 uncoated infective R.
trifolii cells were adsorbed per root hair (200
,um in length); adsorption with lectin-coated
viable or heat-killed cells was so great that a
quantitative estimate was impossible. Higher
magnification revealed that the massive coating on the roots consisted of bacteria. Since the
ability of heat-killed R. trifolii cells to adsorb to
clover roots was tremendously increased by
prior incubation with the clover lectin preparation, the observed effect was not an artifact due
to stimulated growth of the bacterial cells.
The massive bacterial adsorption observed
with clover lectin-coated R. trifolii cells to clover roots was specific for the R. trifolii-clover
system, since it was not observed under conditions where R. trifolii strain 403 or R. japoni-
169
170
APPL. ENVIRON. MICROBIOL.
DAZZO, NAPOLI, AND HUBBELL
41.
.,
losI.
t
.
,
TABLE 2. Inhibition of adsorption of Rhizobium
cells to legume root hairsa
Strain
Strain
R. trifolii strain 403T. repens
R. meliloti strain F28M. sativa
Treatment"
Treatmenth
None
2-Deoxyglucose
2-Deoxygalactose
cells
Adsorbed
(means + SD)
1.80
1.06
2.34
2.75
a-n-Glucose
24.70 +
2.60d +
23.55 ±
24.00 +
None
2-Deoxyglucose
35.50 ± 4.11
34.25 + 4.79
a From 15 to 20 root hairs were examined per treatment,
each approximately 200 gm in length.
b Final sugar concentration was 30 mM.
e SD, Standard deviation.
d
Significantly less than untreated control at 99% level.
infective cells to clover roots through a 2-deoxyglucose-sensitive receptor site. This receptor
site is unique to the R. trifolii-clover system.
Currently, we are trying to determine if the
specific 2-deoxyglucose-sensitive lectin is on the
clover root hair surface at sites where it could
cross-bridge the unique cross-reactive surface
antigens of both symbionts.
ACKNOWLEDGMENTS
This research was supported by grant no. DEB 75-14043
and BM574-01134 from the National Science Foundation
and by a grant-in-aid for research from Sigma Xi, the
Scientific Research Society of North America.
Downloaded from http://aem.asm.org/ on October 17, 2015 by guest
FIG. 3. (a) Sterile T. repens seedling root; (b) T. repens seedling root inoculated with clover lectin-coated
R. trifolii strain 403 cells; (c) T. repens seedling root inoculated with R. trifolii strain 403 cells. Bar (in each)
= 100 ,um. Higher magnification of (b) revealed that this massive coating consisted of bacterial cells. The
results of (b) could be repeated with heat-killed, clover lectin-coated R. trifolii cells, but not if R. japonicum
cells, alfalfa roots, or inactivated clover lectin were substituted.
ADSORPTION OF R. TRIFOLII TO ROOTS
VOL. 32, 1976
LITERATURE CITED
1. Bohlool, B. B., and E. L. Schmidt. 1974. Lectins: possible basis for specificity in the Rhizobium-legume root
nodule symbiosis. Science 185:269-271.
2. Bohlool, B. B., and E. L. Schmidt. 1976. Immunofluorescent polar tips of Rhizobium japonicum: possible
3.
4.
5.
7.
8.
9.
10.
11.
13.
14.
15.
16.
17.
18.
19.
Bushby. 1975. The orientation of certain root-nodule
bacteria at interfaces, including legume root-hair
surfaces. J. Gen. Microbiol. 91:198-200.
Menzel, G., H. Uhig, and G. Weischsel. 1972. Settling
of rhizobia and other soil bacteria on the roots of some
legumes and non-legumes. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 2 127:348-358.
Napoli, C., F. Dazzo, and D. Hubbell. 1975. Production
of cellulose microfibrils by Rhizobium. Appl. Microbiol. 30:123-132.
Napoli, C., F. Dazzo, and D. Hubbell. 1975. Ultrastructure of infection and "common antigen" relationships
in Aeschynomene, p. 1-3. In J. M. Vincent (ed.),
Proceedings of the 5th Australian Legume Conference, March 18-21, 1975, Brisbane, Australia.
Paper no. 15. University of Sydney, Sydney, Australia.
Peters, R. J., and M. Alexander. 1966. Effect of legume
exudates on the root nodule bacteria. Soil Science
102:380-387.
Reporter, M., D. Raveed, and G. Norris. 1975. Binding
of Rhizobium japonicum to cultured soybean root
cells: morphological evidence. Plant Sci. Lett. 5:7376.
Sahlman, K., and G. Fahraeus. 1963. An electron microscopic study of root hair infection by Rhizobium. J.
Gen. Microbiol. 33:425-427.
Savage, D. C., and R. V. H. Blumseshine. 1974. Surface-surface associations in microbial communities
populating epithelial habitats in the murine gastrointestinal ecosystem: scanning electron microscopy.
Infect. Immun. 10:240-250.
Steel, R. G. D., and J. H. Torrie. 1960. Principles and
procedures of statistics, p. 157. McGraw-Hill Book
Co., New York.
Downloaded from http://aem.asm.org/ on October 17, 2015 by guest
6.
site of attachment or lectin binding. J. Bacteriol.
125:1188-1194.
Dart, P. J., and F. V. Mercer. -1964. The legume rhizosphere. Arch. Mikrobiol. 47:344-378.
Dazzo, F. B., and D. H. Hubbell. 1975. Antigenic differences between infective and noninfective strains of
Rhizobium trifolii. Appl. Microbiol. 30:172-177.
Dazzo, F. B., and D. H. Hubbell. 1975. Cross-reactive
antigens and lectin as determinants of symbiotic
specificity in the Rhizobium-clover association. Appl.
Microbiol. 30:1017-1033.
Fahraeus, G. 1957. The infection of clover root hairs by
nodule bacteria studied by a simple glass slide technique. J. Gen. Microbiol. 16:374-381.
Hamblin, J., and S. P. Kent. 1973. Possible role of
phytohaemagglutinin in Phaseolus vulgaris L. Nature (London) New Biol. 245:28-30.
Lratky, A., P. Beily, and S. Bauer. 1975. Mechanism of
2-deoxy-D-glucose inhibition of cell wall polysaccharide and glycoprotein biosynthesis. Eur. J. Biochem.
54:459-467.
Marshall, K. C. 1967. Electrophoretic properties of fast
and slow growing species of Rhizobium. Aust. J. Biol.
Sci. 20:429-438.
Marshall, K. C., and R. H. Cruickshank. 1973. Cell
surface hydrophobicity and the orientation of certain
bacteria at interfaces. Arch. Mikrobiol. 91:29-40.
Marshall, K. C., R. H. Cruickshank, and H. V. A.
12.
171