:7.
. , 73:. ; I _.
2 r. « ...

:
3,4... _
,3: i.

2.1.51.2 '
2:11.: .

>....

23%....» 3
2.25 .1

s
.33... .71.. 3.5..
... y 1.. 21:25.5..5xt

)
3311.55

(3 {9.7.5
:33... 32.11,
\za.,l\.?..a‘ Fit: 1
rwsfxnsxgnsbi,
.Iiiu. 3.. ..IZ ).

1.3::
3:4 l .

 

 

 

 

STATE U

mllfllllijfililllmlllilu ”

020 1642

This is to certify that the

dissertation entitled

CARBOHYDRATE BINDING ACTIVITIES OF BRADYRHIZOBIUM
JAPONICUM: LOCALIZATION AND EXPRESSION OF THE
LECTIN BJ38

 

1‘ ~ 1‘ .l't'l’ i-V , iil‘ i
presented by

JOHN TZINKUAN LOH

has been accepted towards fulfillment
of the requirements for

PH.D degree in BIOCHEMISTRY

 

 

 

 

1 . («J ‘—\
Major professor T

Date 19 AUGUST 1994

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

LURARY
Michigan State
‘ University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID F INES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

—‘

l 1 II J!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fit I!

MSU Is An Affirmative Action/Equal Opportunity Institution
Warm-9.1

 

 

 

 

 

 

 

CARBOHYDRATE BINDING ACTIVITES OF BRADYRHIZOBI UM
JAPONICUM: LOCALIZATION AND EXPRESSION OF THE LECTIN BJ38

By

John TzinKuan Loh

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1994

ABSTRACT

CARBOHYDRATE BINDING ACTIVITIES OF BRAD YRHIZOBI UM JAPONI C UM:
LOCALIZATION AND EXPRESSION OF THE LECTIN B138

By

John TzinKuan Loh

B138 is a lactose/galactose-specific binding protein that is proposed to mediate
the attachment of Bradyrhizobium japonicum to soybean roots. Using a polyclonal
antiserum generated against B138, the lectin has been unipolar localized on the
bacterium cell surface. More importantly, B138 localization is coincident with the site
of B. japonicum attachment to soybean cells. This result indicates that the
distribution of B138 is consistent with its potential role in the polar adhesion of the
bacterium to soybean roots.

The expression of B138, assayed at the polypeptide level, was found to be
elevated when B. japonicum cells were cultured in the presence of saccharides. This
induction effect was of the order lactose > galactose > mannose ~ no hapten,
which is analogous to the relative affinities of these saccharides for B138. A similar

effect was also observed at the mRNA level, using Northern blot analysis.

Treatment of B. japonicum cells with genistein also resulted in elevated levels
of B138 at both the mRNA and polypeptide levels. In parallel, expression of
nodD 1, the transcriptional regulator of the nod genes, was also found to be elevated
by flavonoids in a fashion similar to that of B138 induction. These results raise the
possibility that B138 may be a member of the nod gene family under the control of
nodD 1.

Finally, the levels of nodD, transcription could also be elevated by saccharide
treatment. This observation suggests that alternative inducers, besides flavoniods,

may control the expression of the nod genes in B. japonicum.

To my parents
and
To Pearl

for their love and faith in me

iv

ACKNOWLEDGEMENTS

I would like to thank my mentor Dr. John Wang for his patience in guiding
me through the time spent in his laboratory. I thank him for teaching me about
science and life, but most of all, for instilling in me the understanding that both
commitment and desire are essential elements to success. I would also like to extend
my gratitude to Dr. Melvin Schindler for his encouragement, support and insight into
both science and sports. I thank him for providing a conducive environment for both
scientific and social growth.

I would also like to express my appreciation to members of my guidance
committee, Dr. Richard Anderson, Dr. Robert Hausinger, Dr. Lee Kroos, and Dr.
Natasha Raikhel for their advice and criticisms.

I owe a special note of thanks to Dr. Siu-Cheong Ho for his friendship and
help that he has provided me throughout the years. His constant encouragement made
the rough times much easier to endure.

Finally, I wish to thank my friends and colleagues in the Wang and Schindler
Laboratories who shared both the ups and downs and who made my time in the lab
very special: Anandita, Patty Voss, Dr. Jamie Laing, Dr. Neera Agrwal, Sung-Yuan
Wang, Mark Kadrofske, Yeou Guang Tsay, Eric Arnoys, Agnes Ho, and Sharon
Grabski.

TABLE OF CONTENTS

LIST OF TABLES .................................
LIST OF FIGURES ................................
LIST OF ABBREVIATIONS ..........................
CHAPTER 1: LITERATURE REVIEW ................
INTRODUCTION .............................
i. R. leguminosarum bv. tnfolii-clover system ..........
ii. R. Ieguminosarum bv. viciae-pea system ............

iii. B. japonicum-soybean system ...................

Saccharide-specific binding activities of B.

japonicum ..........................

LITERATURE CITED ..........................

CHAPTER II: CARBOHYDRATE BINDING ACTIVITIES
OF BRADYRHIZOBIUM J APONICUM.
Unipolar Localization of the Lectin B138 on

the Surface of Bmdyflzizobium japom'cm .......
FOOTNOTES ................................
ABSTRACT .................................

INTRODUCTION .............................

vi

Page

...... x

...... Xi

...... 2
..... 6
..... 7
..... 8
10

...... 17

..... 24
...... 25
...... 26

...... 27

 

 

METHODS AND MATERIALS .......................... 29
Cell Cultures ................................. 29
Isolation of B138 ............................... 30
Antibody Reagents .............................. 31
Purification of CPS and LPS ........................ 32
Fluorescence Studies ............................ 33
Electron Microscopy ............................ 34

RESULTS ....................................... 36
Antibodies Directed Against B138 .................... 36
Immunofluorescence of B138 on B. japonicum Cell Surface ..... 39
Ultrastructural Confirmation of Polar Localization of B138 ..... 39
Identification of B138 at the Point of Contact .............. 44

DISCUSSION ..................................... 50

ACKNOWLEDGEMENTS ............................. 53

REFERENCES .................................... 54

CHAPTER III: CARBOHYDRATE BINDING ACTIVITIES OF
BRADYRHIZOBI UM JAPONICUM.
Effect of Lactose and Flavones on the Expression - :
of the Lectin, B138 ........................ 57 ;.

 

ABSTRACT ...................................... 58 l ~
INTRODUCTION .................................. 59
METHODS AND MATERIALS .......................... 61

Bacterial Cultures and Isolation of B138 ................. 61

Use of [1251] Anti-B138 to Quantitate Cell Surface

Expression of B138 ............................. 62

B. japonicum Binding to SB-l Cells ................... 63

Fluorescence Microscopy .......................... 64

Glutamine Synthetase Assay ........................ 64

vii

 

 

 

 

RESULTS ....................................... 66
Lactose Induction of B1 38 Expression in
B. japonicum ................................. 66
Specificity of the Lactose Induction Effect ............... 74

Effect of Lactose Induction on Cell Surface Expression of B1 38 . . 74
Effect of Lactose Induction on the Adhesive properties

of B. japonicum ............................... 82
Induction of B1 38 Expression by Flavones and Derivatives ..... 89
DISCUSSION ..................................... 90
ACKNOWLEDGEMENTS ............................. 95
REFERENCES .................................... 96

CHAPTER IV: CARBOHYDRATE BINDING ACTIVITIES OF
BRADYRHIZOBIUM JAPONICUM.
Analysis of the Effects of Saccharides and Flavones

on the Expression of BJ 38 at the mRN A level ...... 100
FOOTNOTE ..................................... 101
ABSTRACT ..................................... 102
INTRODUCTION ................................. 103
METHODS AND MATERIALS ......................... 105
Cell Cultures ................................ 105
Determination of the Partial Amino Acid Sequence of B138 . . . . 106
Polymerase Chain Reaction (PCR) Amplification and
Nucleotide Sequence of PCR Products ................. 106
RNA Isolation and Northern Blot Analysis .............. 107
Southern Blot Analyses .......................... 109
120le, glnA, galectin—3 oligonucleotide probes ............ 109
RESULTS ...................................... 111

viii

 

 

 

A Molecular Probe for the B138 gene(s) ................ 111

Southern and Northern Blot Analysis Using the B138 probe . . . . 111

Effect of Flavonoids on BJ38 and nodD, expression ......... 119

Effect of Saccharides on B1 38 and nodD, expression ........ 124
DISCUSSION .................................... 126
REFERENCES ................................... 131
CHAPTER V: CONCLUDING STATEMENT ................ 136

ix

 

 

LIST OF TABLES
Page
CHAPTER I
TABLE I. Classification of Rhizobia and their Symbiotic Hosts ..... 3
CHAPTER IV
TABLE I. Comparison of A. tumefaciens vir gene
induction and B. japonicum nod gene
induction ............................... 129

 

CHAPTER I

Figure 1.

CHAPTERII

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

LIST OF FIGURES

Models illustrating the role of B138 in the
binding of B. japonicum to:

(a) Iac—Sepharose beads; (b) SB—l cultured
soybean cells; (c) Soybean roots; and

((1) other B. japonicum cells, leading to

autoagglutination ......................

Immunoblot analyses of B138, CPS and LPS

isolated from B. japonicum ................

Immunofluorescent labeling of B. japonicum
cell surface by anti-B138, anti-Bij , and

preimmune sera .......................

Immunogold labeling of B. japonicum cell
surface by anti—B138, anti-Brj, and

preimmune sera .......................

Immunofluorescent labeling of B. japonicum
cells that were attached to a) other star—forming
B. japonz'cum; b) SB—l cells; c) Lac-

Sepharose beads .......................

FITC-SBA labeling of B. japonicum cells that
were a) in suspension; b) in star-formation;

0) attached to SB-l cells ..................

xi

Page

..... 12

..... 38

..... 41

..... 43

..... 46

..... 48

 

CHAPTERIII

Figure l. Dose-response curve of the effect of Lac on the
production of B138 .......................... 68

Figure 2. The effect of the growth phase of B. japonicum

cells on the induction of B138 expression ............ 70
Figure 3. The effect of time of exposure of B. japonicum

to Lac on the induction of B138 expression ........... 73
Figure 4. Saccharide specificity in the induction of B138 ......... 76
Figure 5. Quantitation of the effect of saccharides on the

cell surface expression of B138 .................. 79
Figure 6. The effect of saccharides on the expression of

B138 at the cell surface of B. japonicum as
detected by immunofluorescence staining
with anti-B138 ............................. 81

Figure 7. The effect of inclusion of Lac in the final liquid
culture of B. japonicum on the subsequent binding of
bacteria to 83-1 soybean cells as observed by
light microscopy ........................... 84

Figure 8. The effect of inclusion of Lac in the final
liquid culture of B. japonicum on the subsequent
binding of bacteria to SB-l soybean cells as

quantitated by a radioimmunoassay ................ 86
Figure 9. Comparison of the effect of flavones and

saccharides on the induction of B138

expression in B. japonicum cells .................. 88

CHAPTERIV

Figure 1. A) The partial amino acid sequence of peptides
derived from B138
B) The deduced amino acid sequences of PCRl
fragment derived from the nucleotide
sequence ............................ 113

xii

Figure 2. Genomic Southern blot of B. japonicum DNA that
had been digested with endonucleases that do
not contain cleavage sites within the PCRl probe ...... 115

Figure 3. A) Northern blots of RNA that had been
hybridized with 32P-labeled PCRl
B) Northern blots of B. japonicum RNA that had
been hybridized with: a) PCRl; b) nodD 1;
c) glnA; d) galectin-3 probes .................. 118

Figure 4. The effect of flavonoids on the mRNA levels of
B138 and nodD, ........................... 121

Figure 5. The effect of saccharides on the mRNA levels of
B138 and nodD, ........................... 123

xiii

ATP
Api
B138

BSA
CBP
cDNA
CPS
DOC
DNA
FITC
Gal
GalN Ac
Gen
Glc

Lac
LPS
Lut

mRN A
Nat
PAGE
PBS
RNA
SBA

TCA
YEG

LIST OF ABBREVIATIONS

adenosine triphosphate

apigenin

carbohydrate binding protein 38 of Bradyrhizobium
japonicum

Bradyrhizobiwn japonicum

bovine serum albumin
carbohydrate binding protein
complementary deoxribonucleic acid
capsular polysaccharide
deoxycholate

deoxyribonucleic acid

fluorescein isothiocyanate
galactose

N -acetyl-D-galactosamine
genistein

glucose

kilodaltons

lactose

lipopolysaccharide

luteolin

mannose

messenger RNA

naringenin

polyacrylamide gel electrophoresis
phosphate buffer saline
ribonucleic acid

soybean agglutinin

tris-buffer saline

trichloroacetic acid
yeast-cxtract-gluconate

xiv

CHAPTER I

LITERATURE REVIEW

INTRODUCTION

Rhizobia, belonging to the genera Rhizobium, Bradyrhizobium and
Azorhizobium, are able to infect the roots of leguminous plants leading to the
establishment of a root nodule symbiosis with the legume (1). Within the nodule, the
bacteria differentiate into bacteriods that fix nitrogen which the plant uses as a
nitrogen source. In return, the plant supplies the bacteria with carbon as energy
sources. The infection process is characterized by a high degree of host-specificity
(Table I). For example, Rhizobium leguminosarum bv. viciae (R. leguminosarum)
nodulates only pea and vetch; Rhizobium leguminosarum bv. trifolii (R. mfolit)
nodulates clover; and Bradyrhizobium japonicum (B. japonicum) nodulates soybean.
This host specificity is observed throughout the nodulation process, including stages
prior to the initiation of infection. Several steps or levels of recognition appear to be
involved in this host-specific nodulation process, and these include the interaction
between components derived from both the bacteria and the plant symbiont (2-4).

At the first level, there are diffusable signals secreted from the plant to the
bacteria. These signals include phenolics and flavonoid compounds, the latter of
which are products of the phenylpropanoid pathway and almost ubiquitous in the plant

kingdom (5). Rhizobia are chemotactic toward these flavonoid compounds (6-9).

 

 

I Bradyrhizobiwn

Azorhizobium

 

R. fredii

111.141.0122-

R. Ieguminosarum

R. Ioti

R. meliloti

biovar phaseoli
biovar trifolii

biovar viciae

B. japonicum

B. Iupini

A. caulinodans

   

 

_1_‘ll'_ L .01. :01.

Wild soybean

Bean (Phaseolus)

Clover (Trifolium)

Pea (Pisum); Vetch (V icia)
Trefoil (Lotus)

Alfalfa (Medieago)
Soybean (Glycine)

Lupine (Lupinus)

Sesbania (stem-nodulating)

4

The secreted flavonoids are plant-specific (10-12) and are able to stimulate the
coordinate transcription of an important set of bacterial genes (nod genes) that are
necessary for nodulation. This process of induction by flavonoids is mediated by
NodD, a transcriptional activator protein encoded for by the nodD gene (13-15). The
expression of nodD varies among rhizobial species; it is induced by flavonoids in B.
japonicum, autoregulated in R. leguminosarum, and constitutively expressed in
Rhizobium meliloti (R. meliloti) (13-14, 16). In the presence of flavonoids, the
NodD protein binds to the conserved nod box DNA sequence preceeding the nod
genes and activates the transcription of these genes (17-20). This interaction of the
NodD protein with flavonoids is specific in that the NodD proteins from different
species of bacteria are able to recognize only specific flavonoids (ll, 21). Thus,
genistein, a specific inducer of nod genes in B. japonicum only interacts with the
NodD protein of B. japonicum (12, 16) while luteolin, the equivalent inducer in R.
meliloti, only activates the NodD protein in that species (10, 13). In this connection,
it has also been demonstrated that the transfer of a nodD gene from Rhizobium sp.
NGR234 into other Rhizobium strains leads to extended host range for the nodD gene
recipient (22).

Flavonoids can also serve as repressors of nod gene expression in non-
symbiotic strains. For example, diadzein and genistein, which are strong inducers in
B. japonicum, are potent inhibitors of gene expression in R. leguminosarum and R.
meliloti (23, 24). The molecular recognition process involving flavonoids, therefore,

constitutes an important first level determinant of host-Rhizobium specificity (24).

5

The second level of recognition involves diffusible signals from the bacterium
to the plant that are termed Nod factors. The Nod factors are able to initiate many of
the early nodulation responses in the plant symbiont. For example, alfalfa roots
exposed to the Nod factors from R. meliloti demonstrate membrane depolarization
(25) and morphogenic changes associated with the infection process; these changes
include root-hair deformation, curling and the formation of pseudo-nodules (26, 27).
The Nod factors are synthesized by proteins that are the product of the nod genes.
These nod genes fall into two general groups; the common nod genes (nodABC) that
are essential for the nodulation of any host, and the host specific nod genes which are
rhizobium-strain specific (2, 4, 28). This latter group of nod genes determine host
specificity by species-specific chemical modification of the B-1,4 linked N-acetyl—D-
glucosamine sugar backbone of the Nod factor. Thus the addition of a sulfate group
to the sugar backbone by the nodHPQ gene products is an important determinant of
host specificity of the R. meliloti factor (27-29), while the O-acetyl group and nature
of the fatty acyl substituent affect the biological activity of the R. leguminosarum
factor (30). Mutations to these host-specific nod genes can lead to structural changes
of these signals resulting in a change in the host range of the bacterium. For
instance, a nodH mutant in R. meliloti produces a non-sulfated Nod factor and is
unable to nodulate alfalfa; instead, it has gained the ability to nodulate vetch, a R.
leguminosarum symbiont (31, 32).

The third level of recognition occurs at the attachment of rhizobia to the root

surface. Rhizobia are found to bind predominantly to the root hairs and it is in this

6

region that the infection process begins (33). Inoculation of rhizobia with seedling
roots leads to an initial clumping and binding of bacterial cells at the tips of the roots.
Rhicadhesin (Mr 14,000), a Ca“ binding protein is proposed to mediate this initital
attachment (34). Rhicadhesin is a nod gene independent bacterial surface protein
found in most rhizobial strains (35). A second binding phase follows in which the
bacteria begin to attach to the root tips in an end-to-end (polar) fashion. This second
phase results in firm attachment of rhizobia to the legume roots (36) and is proposed
to control the host-specific interaction of rhizobia with its symbiont (37,38). Much of
our understanding of how this host-specific attachment occurs has been guided by the
lectin-recognition hypothesis proposed by Kriipe (39). In this hypothesis, legume
lectins are proposed to control the specificity of the rhizobia-legume attachment by
specifically interacting with saccharide components on the bacterial cell surface. This
hypothesis is discussed in the context of the following rhizobial-legume systems.
(i) R. le minosarum bv. tri olii-clover stem

The clover-R. trifolli symbiosis provides the most convincing evidence of the
role of the plant lectin in controlling the host specific attachment of rhizobia to the
legume root. Trifoliin A, a clover lectin (Mr 50,000) that binds to 2-deoxyglucose
has been identified and proposed to function in this role (40). Trifoliin A is
multivalent in binding to and can agglutinate R. mfolii. Infective strains of R. trifolii
contain polysaccharides that are antigenically cross-reactive with the clover root hair
cell surface (41). This cross-reactive antigen (CRA) binds to trifoliin A as well as to

root cells in a fashion matching the distribution of the lectin. The clover root lectin

7

thus functions as a cross—bridge between R. tnfolii and the root hair. Consistent with
this idea, the sugar 24eoxyglucose can specifically inhibit binding of R. trifolii or its
capsular polysaccharide to clover root hairs (42). A strong correlation also exists
between lectin binding and infectivity of R. trifolii (40,43). When the expression of
lectin receptors on the bacterial cell surface as a function of culture age were
examined, the number of lectin receptors on R. tnfolii directly correlated with the
number of cells bound to the clover root hairs (44). Hence, for the R. tnfolii-clover
symbiosis, all the presently available evidence lend support to the lectin-recognition
hypothesis.
(ii) R, leguminosarum bv. viciae-pea system

The attachment of R. leguminosarum to pea roots appears to be mediated by
rhicadhesin, as evidenced by the ability of purified rhicadhesin to inhibit rhizobial
attachment to pea roots (35). A pea lectin is also present in the roots of the pea
legume. However, unlike the R. tnfolii—clover symbiosis, the plant lectin does not
appear to play a direct role in the attachment process. This belief arises from the
observation that rhizobial attachment cannot be inhibited by saccharides, including
affinity haptens (mannose (Man) and glucose (Glc)) of the pea lectin. Moreover, on
surveying the distribution of the pea lectin, the lectin was found on the external
surface of the plasma membrane rather than the outer surface of the plant cell wall
where it could mediate attachment (45).

The pea lectin, though not directly involved in the attachment process,

nevertheless can enhance nodulation as the secreted lectin can contribute to the

8

accumulation of rhizobia at the root hair tips (46). The most striking and persuasive
evidence for a role of the pea lectin in root infection is the finding that transfer of the
psl gene coding for the pea lectin to white clover roots resulted in nodulation of the
transgenic roots by R. leguminosarum (47). N ormally, white clover roots release
appropriate flavonoids for the induction of R. leguminosarum nod genes and signal
molecules produced by the bacteria are recognized by the clover root hairs (21, 48).
However, root hair curling is usually abnormal and infection threads are not found.
Transformation of white clover roots with the pea lectin gene, however, conferred
upon these roots the ability to be nodulated by R. leguminosarum. Thus it appears
that the pea lectin might indeed play a key role in the normal pea-R. leguminosarum

symbiosis.

(iii) B. M. nicum-soybean system

A plant lectin has also been demonstrated to be present in soybean roots. This
lectin, isolated on the basis of its carbohydrate-binding activity, exhibits molecular
and saccharide binding properties similar to seed soybean agglutinin (SBA) (49-51).
All available evidence, however, fails to demonstrate that the plant lectin actually
plays an indispensable role in the recognition process. This viewpoint stems from the
following observations (52). First, the binding of SBA to soybean nodulating bacteria
is not selective in that SBA does bind to certain rhizobial strains that do not infect
soybean roots. Second, some Rhizobium strains that nodulate soybean roots do not

bind detectable amounts of SBA. Third, soybean roots that lack SBA are still able to

9

form root nodules. Fourth, exogenously added SBA, at a concentration sufficient to
saturate the Rhizobium cell surface, resulted in only 35% inhibition of binding to
soybean roots (53). Fifth, there is a glaring discrepancy in the saccharide specificity
of SBA, and the specificity of saccharide inhibition of B. japonicum binding to
soybean roots. If indeed SBA mediates attachment of B. japonicum attachment to
soybean roots, then the addition of SBA haptens galactose (Gal) and N-acetyl-D-
galactosamine (GalN Ac) would inhibit bacterial attachment. However, studies by
Vesper et al. (54) and Ho et al. (55) have shown that bacterial binding was only
inhibited by Gal but not GalN Ac. The specificity of saccharide inhibition of B.
japonicum attachment is not consistent with SBA’s proposed role in mediating the
bacterial attachment to its symbiont. This saccharide specificity in itself may reflect
recognition components, other than the plant lectin, which exhibit Gal-specific
characteristics.

Finally, the distribution of SBA binding polysaccharides is incompatible with
its role in mediating B. japonicum attachment. Studies by Tsien et al. (56) have
revealed two distinct poles in B. japonicum: the nucleoid portion containing the
cytoplasm and the bacterial chromosome, and the reserved polymer end which
contains hydroxybutyrate and glycogen granules. The SBA-binding polysaccharides
are localized in the nucleoid portion (57). As the nucleoid end of the bacterium is the
portion of the bacterium away from the point of attachment, it is highly unlikely that
the plant lectin can play a direct role in bacterial attachment.

Similar to the pea lectin of R. leguminosarum-pea interaction, SBA may still

 

 

10

have a role to play in the nodulation process. SBA can interact with the bacterial
surface and initiate metabolic responses (58-60). A mutant of B. japonicum that
exhibits a delay in nodulation can be reversed by pretreating these cells with SBA or
soybean root exudate (60). This effect is specific in that it could not be repeated
using root exudates or lectins from other legume species. In addition, SBA could also
enhance the nodulation efficiency in wild-type bacteria, facilitating nodule initiation at
a low inoculum of bacterium (60). These observations clearly demonstrate that
though the plant lectin performs no direct role in the attachment of bacterium to the

soybean root, it does participate in facilitating the nodulation process.

Saccharide-specific binding activities of B. iaponicum

In the course of studies to characterize the rhizobium-soybean interaction, it
was discovered that B. japonicum exhibited four carbohydrate-binding properties: (i)
heterotypic binding to soybean root; (b) heterotypic adhesion to cultured soybean SB-l
cells, originally derived from soybean roots; (0) homotypic autoagglutination; and ((1)
adsorption to sepharose beads derivatized with lactose (Lac-Seph) (55). All of these
binding properties are inhibited by Gal, but not by GalNAc. In light of the similar
saccharide specificities of carbohydrate inhibition, it was proposed that these four
assays may be mediated by the same mechanism or components. This hypothesis is
diagrammed schematically in Fig. 1. The key feature of these models is that in each

of the cases of heterotypic binding, the lectin resides on the bacterium, while the

 

 

 

11

Figure 1. Models illustrating the role of B138 in the binding of B. japonicum to: (a)
Lac-Sepharose beads; (b) SB-l cultured soybean cells; (c) soybean roots; and ((1)
other B. japonicum cells, leading to autoagglutination.

 

 

12

 

 

 

 

 

 

13

carbohydrate ligand is presented by the other partner. In the case of homotypic
agglutination, both the lectin and carbohydrate ligand are found on the bacterial
surface. Consistent with this notion, a carbohydrate-binding protein, B138 (M,
38,000) was purified from B. japonicum cells. BJ38 bound specifically to Gal and
Gal-containing sugars, but not to the epimeric sugars mannose (Man) or glucose (Glc)
(61). In addition, a modification of Gal at the C-2 position resulted in a drastic
lowering of affinity as demonstrated by an 18-fold reduction of binding by GalNAc.
The carbohydrate-binding specificity of B1 38 was, therefore, in good agreement with
the specificity of saccharide inhibition of B. japonicum binding in the four saccharide—
specific assays. B138 is thus a likely candidate responsible for mediating the
saccharide-specific binding of B. japonicum.

This hypothesis was further tested with mutants that were defective in
attachment. If indeed B138 was responsible for mediating the carbohydrate-specific
binding in all four assays, two observations would follow. First, mutants isolated on
the basis of an impairment in one of the binding activities should show a concomitant
loss of the other binding activities. Second, the mutants would produce little or
defective B138 compared to the parental strain. Chemical mutagenesis of wild-type
bacteria with N -methyl-N’-nitro-N -nitrosoguanidine, followed by selection based on
reduced binding to soybean SB—l cells, resulted in the isolation of two mutants, N4
and N6 (55). Compared to the wild-type, these mutants showed no difference in
morphology, growth patterns and staining with both SBA, and antibodies generated

against the lipopolysaccharide and capsular polysaccharide fractions of B. japonicum

 

 

 

 

14

(anti—Brj). However, when the mutants were analyzed for the ability to participate in
all four assays, absent or reduced binding was observed. When extracts from the N4
and N6 mutants were fractionated over a Lac-Seph column, no B138 could be
identified. These results further support the hypothesis that the saccharide binding
properties are mediated by a common mechanism/ components and that B] 38 may be
directly involved in the attachment process.

The importance of B138 and bacterial binding in the infection process was
investigated by comparing the nodulation efficiency of the mutants with the wild-type
bacteria. When soybean seedlings were inoculated with three doses each of wild-
type, N4 and N6 bacteria, the plants treated with the mutant cells showed a reduction
in nodule formation, relative to the parental strain (62). This result suggests that the
carbohydrate-specific binding process, and possibly BJ38, may be an important
component in the symbiosis between soybean plants and B. japonicum. Consistent
with this, purified B1 38 was found to bind predominantly to the young emergent root
hairs of soybean roots and to a much lesser extent, to the root cap, mature root hairs,
epicotyl or hypocotyl region. This positional specificity was identical to that observed
in B. japonicum cell binding to the soybean root. In previous studies, it has been
documented and postulated that rhizobia tend to colonize at the root hair zone as the
rhizosphere surrounding it were rich in nutrients suitable for bacterial viability and
growth. Nonetheless, this reasoning could not account for the preferential binding of
B1 38 to these regions, as the lectin’s binding is not dependent on nutrients and factors

of the environment. Thus, it seems more likely that both B. japonicum and B138 bind

 

 

 

 

 

 

 

 

15

bind selectively to the emergent root hairs as these regions express the carbohydrate
structures recognized by B1 38. In studies by Bauer and coworkers (63, 64), it was
also demonstrated that the emergent root hair zones were regions of the soybean root
that were most susceptible to root-hair deformation, a key step in the infection process
leading to the formation of nitrogen fixing nodules. Thus three key correlations are
observed which support the hypothesis that B1 38 mediates B. japonicum binding to
soybean roots and that the attachment is a significant step leading to successful

nodulation. First, B. japonicum cells preferentially bind to the young, emergent

 

soybean root hairs; second, BJ38 binds to sites that are coincident with bacterial
binding; third, these sites have been shown to be most susceptible to infection and
nodule initiation.

In summary, though the lectin-recognition hypothesis appears to be upheld in
the R. tnfolz’i—clover system, this hypothesis has fared less well in the pea and soybean
systems. The documentation of a carbohydrate specific lectin isolated from B.

japonicum presents a new and alternative View to the lectin-recognition hypothesis; in

 

that the bacteria provides the protein and the plant the carbohydrate.
In this dissertation, I describe the characterization of BJ38 in terms of its

distribution in B. japonicum. This determination is important as the B. japonicum

 

cells bind to soybean roots in a polar fashion. Thus any hypothesis implicating a role
for B138 in bacterial binding will require the lectin to be exposed at the cell surface
where the attachment occurs. In addition, I also describe the characterization of B138

in terms of factors that regulate its expression. The infection process occurs via a

   

16

series of steps that involve the activation of nod genes. As the attachment process,
possibly involving BJ38, is a requisite step leading to the formation of nitrogen fixing
nodules, the possibility exists that B138 may be a member of the nod gene family.
This possibility is studied in terms of the ability of specific inducers of the nod genes

in B. japonicum to similarly induce BJ38.

 

 

 

10.

17

LITERATURE CITED

Brewin, N. J. 1991. Development of the legume root nodule. Ann. Rev. Cell
Biol. 7: 191-226.

Fisher, R. F., and S. R. Long. 1992. Rhizobium-plant signal exchange. Nature
(Land). 357: 655-660.

Spaink, H. P. 1992. Rhizobia] lipo-oligosaccharides: answers and questions.
Plant Mol. Biol. 20: 977-986.

Dénarié 1., F. Débellé, and C. Rosenberg. 1992. Signaling and host range
variation in nodulation. Ann. Rev. Microbial. 46:497-531.

Lamb, C.J., M. A. Lawton, M. Dron, and R. A. Dixon. 1989. Signals

and transduction mechanisms for activation of plant defenses against microbial
attack. Cell 56: 215-224.

Caeteno-Anollés, G., D. K. Grist-Estes, and W. D. Bauer. 1988. Chemotaxis
of Rhizobium melilori to the plant flavone luteolin requires functional
nodulation genes. J. Bacterial. 170:3164-3169.

Dharmatilake, A. 1., and W. D. Bauer. 1992. Chemotaxis of Rhizobium
melilati toward nodulation gene-inducing compounds from alfalfa roots. Appl.
Environ. Microbial. 58: 1153-1158.

Armitage, J. P., A. Gallagher, and A. W. B. Johnston. 1988. Comparison of
the chemotactic behavior of Rhizobium leguminosarum with and without the
nodulation plasmid. Mol. Microbiol. 2:743-748.

Kappe, R., M. Parniske, and D. Werner. 1991. Chemotaxis and nod gene
activity of Bradyrhizabium japanicum in response to hydroxycinnamic acids
and isoflavonoids. Appl. Environ. Microbial. 57:316-319. ;

Peters, N. K., J. W. Frost, and S. R. Long. 1986. A plant flavone luteolin

induces expression of Rhizobium melilati nod genes. Science 233:917-1008.

 

 

 

 

 

 

11.

12.

13.

14.

15.

16.

17.

18.

18

Zaat, S. A. 1., C. A. Wijffelrnan, I. H. M. Mulders, A. A. N. van Brussel,
and B. J. J. Lugtenberg. 1988. Root exudates of various host plants of R.
leguminasarum contain different sets of inducers of Rhizobium nodulation
genes. Plant Physiol. 86:1298-1303.

Kosslak, R. M., R. Bookland, 1. Barkei, H. E. Paaren, and E. R. Appelbaum.
1987. Induction of Bradyrhizabium japanicum common nod genes by
isoflavones isolated from Glycine max. Proc. Natl. Acad. Sci. USA 84:7428-
7432.

Mulligan, J. T., and S. R. Long. 1985. Induction of Rhizobium melilati
nadC expression by plant exdudate requires nadD. Proc. Natl. Acad. Sci.
USA 82:6609-6613.

Rossen, L., C. A. Shearman, A. W. B. Johnston, and J. A. Downie. 1985.
The nodD gene of Rhizobium leguminasarum is autoregulatory and in the
presence of plant exudate induces the nadA, B, C genes. EMBO J. 4:3369-
3373.

Shearman, C. A., L. Rossen, A. W. B. Johnston, and J. A. Downie. 1986.
The Rhizobium leguminasarum nod gene nadF encodes a polypeptide similar to
acyl-carrier protein and is regulated by nadD plus a factor in pea root exudate.
EMBO J. 5:647-652.

Banfalvi, Z., A. Nieuwkoop, M. Schell, L. Besl, and G. Stacey. 1988.
Regulation of nod gene expression in Bradyrhizabium japanicum. Mal. Gen.
Genet. 214:420-424.

Fisher, R. F., T. T. Egelhoff, J. T. Mulligan, and S. R. Long. 1988.
Specific binding proteins from Rhizobium melilati cell free extracts containing
N odD to DNA sequences upstream of inducible nodulation genes. Genes Dev.
2:282-293.

Hong, G. G., J. E. Burn, and A. W. B. Johnston. 1987. Evidence that DNA
involved in the expression of nodulation (nod) genes in Rhizobium binds to the

product of the regulatory gene nadD. Nucleic Acids Res. 15:9677-9690.

 

 

 

19

19. Winsor, B. A. T. 1987. A nod at differentiation: the nadD gene product and
initiation of Rhizobium nodulation. Trends Genet. 5: 193-201.

20. Kondorosi, E., J. Gyuris, J. Schmidt, M. John, E. Duda, B. Hoffman, J.
Schell, and A. Kondorosi. 1989. Positive and negative control of nod gene
expression in Rhizabium melilati is required for Optimal nodulation. EWO J.
8:1331-1334.

21. Spaink, H. P., C. A. Wijffelman, E. Pees, R. J. H. Okker, and B. J. J.
Lugtenberg. 1987. Rhizobium nodulation gene nadD as a determinant of host
specificity. Nature (Land). 328:337-340.

22. Horvath, B., C. W. B. Bachem, J. Schell, and A. Kondorosi. 1987. Host-
specific regulation of nodulation genes in Rhizobium is mediated by a plant
signal, interacting with the nadD gene product. EMBO J. 6:841-848.

23. Firmin, 1. L., K. E. Wilson, W. L. Rossen, and A. W. B. Johnston. 1986.
F lavonoid activation of nodulation genes in Rhizobium reversed by other
compounds present in plants. Nature (Land). 324:90-92.

24. Gyorgypal, 2., E. Kondorosi, and A. Kondorosi. 1991. Diverse signal
sensitivity of NodD protein homologs from narrow and broad-host range
rhizobia. Mol. Plant Microbe Interact. 4:356-364.

25. Ehrhardt, D. W., E. M. Atkinson, and S. R. Long. 1992. Depolarization of
alfalfa root hair membrane potential by Rhizobium melilati Nod factors.

Science 256:998-1000.

26. Lerouge, P., P. Roche, C. Faucher, F. Maillet, G. Truchet, J. C. Promé, and
J. Dénarié. 1990. Symbiotic-host specificity of Rhizobium melilati is
determined by a sulphated and acylated glucosamine oligosaccharide signal.
Nature (Land). 344:781-784.

27. Truchet, G., P. Roche, P. Lerouge, J. Vasse, S. Camut, F. de Billy, J. C.

Promé, and J. Dénarié. 1991. Sulfated lipooligosaccharide signals of

Rhizobium melilati elicit root nodule organogenesis in alfalfa. Nature (Land).
351:670-673.

 

 

 

 

 

 

 

28.

29.

30.

31.

32.

33.

34.

35..

20

Dénarié, 1., and J. Cullirnore. 1993. Lipooligosaccharide nodulation factors -
a minireview: new class of signaling molecules mediating recognition and
morphogenesis. Cell 74:951-954.

Roche, P., F. Débellé, F. Maillet, P. Lerouge, C. Faucher, G. Truchet, J.
Dénarié, and J . C. Promé. 1991. Molecular basis of symbiotic host
specificity in Rhizobium melilati: nadH and nadPQ genes encode the sulfation
of lipo-oligosaccharide signals. Cell 67: 1131—1143.

Spaink, H. P., D. M. Sheeley, A. A. N. van Brussel, J. Glustika, W. S.

York, T. Tak, O. Geiger, E. P. Kennedy, V. N. Reinhold, and B. J. J.

Lugtenberg. 1991. A novel highly unsaturated fatty acid moiety of lipo-
oligosaccharide signals determines host specificity of Rhizobium. Nature
(Land). 354:125-129.

Horvath, B., E. Kondorosi, M. John, J. Schmidt, 1. Torok, Z. Gyorgypal, I.
Barabas, U. Wieneke, J. Schell, and A. Kondorosi. 1986. Organization,
structure and symbiotic function of Rhizobium melilati nodulation genes
determining host specificity. for alfalfa. Cell 46:335-343.

Faucher, C., F. Maillet, J. Vasse, C. Rosenberg, A. A. N., van Brussel, G.
Truchet, and J . Dénarié. 1988. Rhizobium melilati host range nadH gene
determines production of an alfalfa-specific extracellular signal. J. Bacterial.
170:5489-5499.

Bhuvaneswari, T. V., A. A. Bhagwat, and W. D. Bauer. 1981. Transient
susceptibility of root cells in nodulation by rhizobia. Plant Physiol. 68:1144-
1149.

Smit, G., S. Swartz, B. J. J. Lugtenberg, and 1. W. Kijne. 1992. Molecular
mechanisms of attachment of rhizobium bacteria to plant roots. Mal.
Microbial. 6:2897-2903.

Smit, G., G. J. J. Logman, M. E. T. I. Boerrigter, 1. W. Kijne, and B. J. J.
Lugtenberg. 1987. Purification and partial characterization of the Rhizobium

leguminasarum biovar viciae Ca2+-dependent adhesin, which mediates the first

 

 

 

 

 

 

36.

37.

38.

39.

40.

41.

42.

43.

 

21

step in attachment of cells of the family Rhizobiaceae to plant root hair tips.
J. Bacteriol. 171:4054-4062.

Smit, G., and G. Stacey. 1990. In Microbial Cell Surface Hydrophobicity.
R. Doyle and M. Rosenberg, editors. ASM Publications, Washington, D. C.
pp. 179-210.

Dazzo, F. B. , and G. Truchet. 1983. Interaction of lectins and their
saccharide receptors in the Rhizobium-legume symbiosis. J. Membrane Biol.
73: 1-16.

Halverson, C. 1., and G. Stacey. 1986. Signal exchange in plant-microbe
interactions. Microbial. Rev. 50:193—225. I

Krfipe, M. 1956. Blutgrappenspezifische pflanzliche Eiweisskorper
(Phytagglutine). Ferdinand Enke, Stuggart. pp. 131.

Dazzo, F. B., W. E. Yanke, and W. J. Brill. 1978. Trifoliin A: A
Rhizobium recognition protein from white clover. Biachim. Biophys. Acta.
539:276-286.

Dazzo, F. B., and D. H. Hubbell. 1975. Cross-reactive antigens and lectin
as determinants of symbiotic specificity in the Rhizobium-clover association.
Appl. Environ. Microbiol. 30:1017-1033.

Dazzo, F. B., and W. J. Brill. 1977. Receptor site on clover and alfalfa
roots for Rhizobium. Appl. Environ. Microbial. 33:132-136.

Dazzo, F. B. 1981. Bacterial attachment as related to cellular recognition in
the Rhizobium-legume symbiosis. J. Supramal. Struct. Cell Biochem. 16:29—
41.

Hrabak, E. M., M. R. Urbano, and F. B. Dazzo. 1981. Growth-phase
dependent immunodeterminants of Rhizobium tnfolii lippolysaccharides which

bind Trifoliin A, a white clover lectin. J. Bacteriol. 148:697-711.

 

 

 

 

 

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

 

22

Diaz, C. L., P. C. van Spronsen, R. Bakhuizen, G. J. J. Logman, B. J. J.
Lugtenberg, and J. W. Kijne. 1986. Correlation between infection by
Rhizobium leguminasarum and lectin on the surface of Pisum sativum L. roots.
Planta 168:350-359.

Diaz, C. L., M. Hosselet, G. J. J. Logman, E. van Driessche, B. J. J
Lugtenberg, and J . W. Kijne 1990. Distribution of glucose/mannose-specific
isolectins in pea (Pisum sativum L.) seedlings. Planta 181:451-461.

Diaz, C. L., L. S. Melchers, P. J. J. Hooykaas, B. 1. J. Lugtenberg, and J.
W. Kijne. 1989. Root lectin as a determinant of host-plant specificity in the
Rhizobium-legume symbiosis. Nature (Land. ) 338:579-581.

Spaink, H. P., R. J. H. Okkar, C. A. Wijffelman, E. Pees, and B. J. J.
Lugtenberg. 1987. Promoters in the nodulation region of the Rhizobium
leguminosarum sym plasmid pRLlJ 1. Plant Mol. Biol. 9:27-39.

Gade, W., M. A. Jack, J. B. Dahl, E. L. Schmidt, and F. Wold. 1981. The
isolation and characterization of root lectin from soybean (Glycine max (L),
cultivar Chippewa). J. Biol. Chem. 256:12905-12910.

Gade, W., and E. L. Schmidt. 1983. Evidence for the existence of an
intracellular root-lectin in soybean. Planta 158: 108-110.

Malek-Hedayat, S. , S. A. Meiners, T. N. Metcalf, M. Schindler, and J. L.
Wang. 1987. Endogenous lectin from cultured soybean cells. Chemical
characterization of the lectin from SB-l cells. J. Biol. Chem. 262:7825-7830.
Pueppke, S. G. 1983. Soybean lectin: does it have an essential role in the
Rhizobium-soybean symbiosis? In Chemical Taxonomy, Molecular Biology,
and Function of Plant Iectins. I. 1. Goldstein and M. E. Etzler, editors.
Alan R. Liss, Inc., New York, NY. pp. 225-236.

Pueppke, S. G. 1984. Adsorption of slow-and-fast-growing rhizobia to
soybean and cowpea roots. Plant Physiol. 75:924-928.

VeSper, S. 1. , and W. D. Bauer. 1985. Characterization of Rhizobium
attachment to soybean roots. Symbiosis 12139-162.

 

 

 

 

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

23
Ho, S.-C., J. L. Wang, and M. Schindler. 1990. Carbohydrate binding

activities of Bradyrhizabi um japanicum. I. Saccharide-specific inhibition of
homotypic and heterotypic adhesion. J. Cell Biol. 111:1631-1638.

Tsien, H. C., and E. L. Schmidt. 1977. Polarity in the exponential-phase
Rhizobium japanicum cell. Can. J. Microbial. 23:1274-1284.

Tsien, H. C. 1982. Ultrasturcture of the free-living cell. In Nitrogen
Fixation. Vol.2. W. J. Broughton, editor. Clarendon Press, Oxford. pp.
182-198.

Halverson, L. J. , and G. Stacey. 1984. Host recognition in the Rhizobium-
soybean symbiosis. Detection of a protein factor in soybean root exudate
which is involved in the nodulation process. Plant Physiol. 74:84-89.
Halverson, L. 1., and G. Stacey. 1985. Host recognition in Rhizobium-
soybean symbiosis. Evidence for the involvement of lectin in nodulation.
Plant Physiol. 77:621-625.

Halverson, L. 1., and G. Stacey. 1986. Effect of lectin on nodulation of
wild-type Bradyrhizabium japanicum and nodulation defective mutant. Appl.
Environ. Microbial. 51:753-760.

Ho, S.-C., M. Schindler, and J. L. Wang. 1990. Carbohydrate binding
activities of Bradyrhizobium japanicum. II. Isolation and characterization of a
galactose-specific lectin. J. Cell Biol. 111:1639-1643.

Ho, S.-C., J . L. Wang, M. Schindler, and J. T. Loh. 1994. Carbohydrate
binding activities of Bradyrhizabium japanicum. III. Lectin expression,
bacterial binding, and nodulation efficiency. Plant J. 52873-884.
Bhuvaneswari, T. V., B. G. Turgeon, and W. D. Bauer. 1980. Early events
in the infection of soybean (Glycine max.L. Merr) by Rhizobium japanicum.

1. Location of infectible root cells. Plant Physiol. 66: 1027-1031.

Calvert, H. E., M. K. Pence, M. Pierce, N. S. A. Malik, and W. D. Bauer.
1984. Anatomical analysis of the deveIOpment and distribution of Rhizobium
infections in soybean roots. Can. J. Bot. 62:2375-2384.

 

 

 

 

CHAPTER II

CARBOHYDRATE BINDING ACTIVITIES
OF BRAD YRHIZOBI UM JAPONICUM
Unipolar Localization of the Lectin BJ38

on the Surface of B. japanicum

24

 

25

FOOTNOTES

1This work was previously published in the Proceedings of National Academy of
Science (1993) 90:3033-3037 and is reproduced with permission from the publisher.

2The electron microscopy work was performed by Dr. Siu-Cheong Ho.

 

 

 

 

 

26
ABSTRACT

A polyclonal antiserum generated against the Bradyrhizabium japanicum lectin
B1 38 was characterized to be specifically directed against the protein. Treatment of
B. japanicum cells with this antiserum and subsequent visualization with transmission
electron and fluorescence microscopy revealed B1 38 at only one pole of the bacteri-
um. B] 38 appeared to be organized in a tuft-like mass, separated from the bacterial
outer membrane. BJ 38 localization was coincident with the attachment site for: a)
homotypic agglutination to other B. japanicum cells; b) adhesion to the cultured
soybean cell line, SB-l, and c) adsorption to Sepharose beads covalently derivatized
with lactose. In contrast, the plant lectin soybean agglutinin labeled the bacteria at
the pole distant from the bacterial attachment site. The results documented indicate
that the topological distribution of B1 38 is consistent with a potential role for this

bacterial lectin in the polar binding of B. japanicum to other cells and surfaces.

 

 

_.._._._. __-.-_-4 v P

 

 

 

 

 

27
INTRODUCTION

Many eukaryotic and microbial cells can recognize and interact specifically
with other cells, either cells of the same type (homotypic) or cells of a different type
(heterotypic). One example is the binding of the bacterium Rhizobium to the root
cells of leguminous plants, leading to the formation of the symbiotic association and
nitrogen fixing nodules (1,2). Several lines of evidence indicate that one key event in
the establishment of such a symbiosis is the initial binding between the bacterium and
the host cell (3).

In previous studies (4), we have documented that Bradyrhizabium japanicum,

 

which normally infects the roots of soybean plants, exhibits four saccharide-specific
binding activities: (a) adsorption to Sepharose beads derivatized with lactose (Lac—
Sepharose); (b) homotypic autoagglutination; (c) heterotypic binding to cultured
soybean (SB-1) cells; and (d) heterotypic adhesion to soybean roots. In all four of
these assays, galactose inhibited the binding but a C-2 derivative of the monosaccha-
ride, N-acetyl-D—galactosamine, failed to yield the same effect. When mutants of B.
japanicum, designated as N4 and N6, were isolated on the basis of a defect in one
binding activity (SB-1 cell binding), it was found that they showed a concomitant loss
in the other three binding capacities (4). These observations suggested that all four of
the carbohydrate-specific binding processes may be mediated by the same compo-
nent(s) and mechanism(s). Consistent with this notion, we have purified a carbohy-

drate-binding protein, designated BJ38 (M, ~ 38,000), which has an affinity for

 

 

 

 

28

galactose and Lac about 13- and 240-fold higher, respectively, than that for N -acetyl-
D-galactosamine (5). This carbohydrate binding specificity correlates well with that
of the B. japanicum binding activities. It was proposed, therefore, that B138 may
mediate the carbohydrate binding activities of B. japanicum.

In all the four carbohydrate-specific activities tested, B. japanicum bound in a
polar fashion (4). Thus, any hypothesis implicating a role of BJ38 in these binding
assays would require that the lectin must not only be anisotropically localized at the
pole of the bacterium, but also must be exposed at the surface of the bacterium where
attachment occurs. In the present communication, we report on the characterization
of a specific anti-B1 38 antibody that has allowed us to identify the topological

distribution of B1 38 on the bacterial cell surface.

 

 

 

29
MATERIALS AND METHODS

Cell Cultures

B. japanicum (R110d) was originally obtained from Dr. Barry Chelrn of
Michigan State University. This bacterial strain was maintained on agar plates
containing yeast extract-sodium gluconate medium (YEG) and 50 mM lactose. YEG
contained (per liter): 1.28 g K2PO4, 0.2 g MgZSO4.7HZO, 7.35 mg CaC12.2 H20, 28
mg sequestrene, 5.0 g gluconic acid, and 1.0 g yeast extract, and was supplemented
with trace elements: 2.5 mg NazEDTA, 4.39 mg ZnSO4.7 H20, 0.77 mg
MnSO4.HzO, 0.15 mg CuSO4.5 H20, 2.49 mg NazMoO4.2 H20, 0.23 mg CoC12.6
H20, 0.46 mg Na2B4O7.10 H20, 0.38 mg Na3VO4, and 0.1 mg NaSeO3, pH 6.0. The
bacteria (3-day-old) were transferred from the agar plate to 50 ml YEG in a 125 ml
Erlenmeyer flask and cultured for one day. The bacterial suspension was then
inoculated into 2 liters of YEG and cultured for 2 days on a gyratory shaker (120
rpm) at 30°C. Aliquots of this culture (300 ml) were then inoculated into six
Fembach flasks, each containing 1.5 liters of YEG. The bacteria were further
cultured for about 30 h until a value of 1.7-2.0 was obtained for the absorbance at
620 nm (A620). The bacterial cells were then used for the isolation of BJ38, capsular
polysaccharide (CPS), lipopolysaccharide (LPS), and for irnmunolocalization studies.

The SB-l cell line, derived from soybean roots (Glycine max (L) Merr. cv.
Mandarin) was kindly provided by Dr. G. Lark (Department of Biology, University

of Utah, Salt Lake City). Cultures were grown in 50 ml of 1B5C medium at 27°C in

 

 

 

30

a gyratory shaker in the dark as described before (6). Continuous cultures were
maintained by subdividing cells every 4 days, transferring 15 ml of culture to 50 ml

of fresh 1B5C.

Isolation of B1 38

The bacterial culture (A620 = 1.7) was centrifuged at 8,000 rpm for 15 min in
a Sorvall GS-3 rotor. The pelleted bacteria were frozen at -20° C overnight, thawed
and then subjected to a passage through the French press at 1,600 psi. The bacterial
cell extract was then centrifuged at 15,000 rpm in a Sorvall SS-34 rotor for 30 min
and the supernatant was subjected to affinity chromatography on Sepharose covalently
derivatized with lactose (Lac-Sepharose) at 4° C. The column (1.5 x 7 cm) was
prepared by coupling lactose to activated Sepharose 4B beads using the procedure of
Matsumoto et al. (7) and equilibrating with PBS (10 mM sodium phosphate, 0.14 M
NaCl, 4 mM KCl, pH 7.4). After loading the extract, the column was washed with
300 ml of PBS and material bound to the column was then eluted with 50 ml of
0.1 M Lac in PBS. Fractions containing the lectin were identified by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 10% acrylamide gels
(8). Polypeptides were revealed by silver staining (9) or by immunoblotting.

For immunoblotting, proteins were transferred to the Immobilon—p (Millipore,
Naperville, IL) membrane by electrophoresis (400 mA, 2 h, 25° C). The membrane
was blocked overnight in TBS (20 mM Tris, 500 mM NaCl, pH 7.5) containing 5%

(wt/vol) bovine serum albumin (BSA). Membranes were then incubated with 1:20

 

 

 

 

31

dilutions of the primary antiserum for 8 h and then washed three times with 10 min
incubations in TTBS (TBS containing 0.05 % Tween 20). The membranes were
incubated with a 1:1000 dilution of goat anti-rabbit IgG coupled to alkaline phospha-
tase in 5% BSA/PBS for 1 h at 25° C, followed by washing in TTBS (3 times, 10 min
each). The immunoreactive material was then developed with nitro blue tetrazolium

and 5-bromo-4-chloro-3-indolyl phosphate as substrates.

Antibody Reagents

Fractions containing BJ 38 lectin isolated from 100 liters of bacterial culture
were concentrated by acetone precipitation (10). The B1 38 fractions were mixed with
ice-cold acetone at a ratio of 4:1 (v/v) and incubated in a dry-ice/methanol bath for 30
min before centrifugation at 10,000 rpm using a SS-34 rotor. The pellet was
resuspended in SDS sample buffer, and analyzed by silver staining after SDS-PAGE
(10% acrylamide). Gel slices containing BJ38 lectin were excised. This sample,
containing about 10 ug of B1 38, was homogenized with Titer-Max adjuvant (Cthx
Corporation, N orcross, GA) and injected into a female New Zealand White rabbit.
Booster injections of 6 pg of B138 in Titer-Max adjuvant were administered every
three weeks. Blood from the rabbit was collected one week after each boost. The
antiserum was collected and stored at -200 C in 1 ml aliquots.

To obtain monospecific antibodies directed against B1 38 from the antiserum,
B1 38 fractions eluted from the Lac-Sepharose column were subjected to SDS—PAGE

and immunoblotted with anti-B1 38 antiserum. A strip of the blot was cut out and the

 

 

32

position of migration of B138 was determined by developing the irnmunoblot. Strips
corresponding to B138 were then cut out from the remainder of the blot and antibody
specific to B138 was eluted by 0.1 M glycine, pH 2.3 for 30 sec with vortexing,
according to the method of Smith and Fisher (11). The eluant was neutralized with 2
M Tris, pH 7.5. This treatment was repeated 5 times, and the eluant was pooled
together and dialyzed against PBS overnight. The dialysate was lyophilized and the

monospecific, affinity purified anti-B138 antibody was resuspended in water.

Purification of CPS and LPS

The CPS fraction of B. japanicum was isolated by the method described by
Mort et al. (12). The bacterial culture (A620 = 1.7) was centrifuged at 8,000 rpm for
15 min in a Sorvall GS-3 rotor. CPS was then removed from B. japanicum by a 2
min treatment in a Waring blender. These cells were centrifuged at 15,000 rpm for
30 min in a Sorvall SS-34 rotor. The supernatant was dialyzed 24 h against water
and the dialysate lyophilized.

The LPS fraction was purified from B. japanicum by phenol-water extraction
(13). The pelleted bacteria were first washed with 0.5 M NaCl three times to remove
exopolysaccharide. Equal volumes of water and phenol was added to the cells and the
mixture incubated at 65°C for 15 min. The aqueous layer was collected and dialyzed
24 h against distilled water. The dialysate was passed over a Dowex AG 1X8
(acetate form) (Sigma, St. Louis, MO) column. The flow-through fraction containing

LPS was lyophilized.

33

Flu r en t

a) B. igganicum cells in suspension culture: B. japanicum cells (A620 = 1.7)
were fixed onto coverslips according to the method of Aplin and Hughes (14). The
cells were blocked for 2 h in 5% (wt/vol) BSA in YEG. One set of the sample was
incubated with fluorescein (FITC) labeled soybean agglutinin (SBA; Sigma) for 30
min, followed by washing (2 times, 10 min each) in PBS. The remainder of the
sample was incubated for 4 h with 1:20 dilutions (in 5% BSA/YEG) of preimmune
serum and anti-B138 of this study and anti-Brj raised against heat-killed B. japanicum
cells (6). After washing in YEG (3 times, 10 min each), FITC-conjugated goat-anti
rabbit antibody (1:50 dilution, Boehringer Mannheim Biochemicals, Indianapolis, IN)
in 5 % BSA/YEG was added and incubated for 1 h. The samples were washed with
PBS to remove the unbound antibody. Stained cells were then visualized with a
fluorescence microscope (Carl Zeiss D-7082 Oberkochen; excitation filter: 546 nm;
chromatic beam splitter: 580 nm; barrier filter: 590 nm).

b) Autoagglutination of B. £2. onicum that exhibited star formation: B.
japanicum cells were grown on a YEG agar plate for 5 days. These cells were then
fixed onto coverslips according to the method of Aplin and Hughes (14). The cells
on the coverslips were blocked in 5% BSA/YEG for 2 h. The antibody or SBA

labeling procedure was performed as in (a).

 

c) B. z‘aQanicum bindm' g to SB-l cell_s or Lac-Sepharose Eds: B. japanicum
cells (A620 = 1.7) were incubated with either SB-l cells (2-day-old) or Lac-Sepharose

beads in a petri dish for 3 h. The mixture was then transferred to 5 ml culture tubes

 

34

to allow the bound fraction to settle to the bottom of the tube. Unbound B.
japanicum cells were removed and the bound cells were fixed in 2%
paraformaldehyde in YEG for 30 min. The cell cultures were subjected to three
washes in YEG (10 min each) and blocked in YEG containing 5% (wt/vol) BSA for 2
h. The B. japanicum cells that were bound to SB-l cells were treated with either
FITC-SBA or anti-B138 antiserum, while the B. japanicum cells bound to Lac-
Sepharose beads were treated with only anti-BJ 38 antiserum. The SBA and antibody

labeling procedures were then performed as above in (a).

Electron Microscopy

Both B. japanicum cells from suspension culture (similar to the immunofluo—
rescence studies) and actively growing cells in solid agar YEG medium (a 3-day-old
culture in which a majority of the cells participated in star formation) were used. For
B. japanicum grown on agar plates, the cells were first suspended in 2 ml of YEG
medium. Cell clumps were dispersed by resuspension with a Pasteur pipette. The
star-forming clusters or single cells from suspension culture were fixed for 15 min on
ice in 0.25 % glutaraldehyde (EM Grade, Electron Microscopy Sciences, Fort
Washington, PA) in YEG. The cells were then washed with PBS by centrifugation
(3,000 rpm, HL-4 rotor, 15 min) and resuspension. The cell pellets were suspended
in BSA-PBS to inactivate the remaining aldehyde groups of the fixative. After 1 h on
ice, the samples were centrifuged (3,000 rpm, HL-4 rotor, 15 min) and the pelleted

cells were suspended in anti-B1 38, anti-Brj or normal rabbit sera (1 :20 dilution in

 

 

 

 

 

35

BSA/PBS, 4 h on ice). The samples were then washed three times with 4 ml PBS
and labeled with goat-anti-rabbit IgG coupled with colloidal gold (1:5 dilution,
particle size of 20 nm, Polysciences, Warrington, PA). After incubation for l h at
room temperature, the cells were washed three times and resuspended in 4 ml PBS,
and stored at 4° C overnight. The bacteria were then centrifuged at 3,000 rpm for 15
min in a HL-4 rotor and the pelleted cells were resuspended in 0.2 ml of PBS before
being applied to the grids.

The formvar coated grids were first treated with 1 N HCl (5 min) to provide a
negatively charged surface. After washing two times with water, the grids were then
coated with 1 % polylysine (M, 100,000, Sigma) in water for 5 min. Excess amounts
of polylysine were removed by rinsing in water three times. A drop of the bacterial
sample was then applied on each grid and the cells were allowed to settle on the grid
for 15 min. Excess sample was blotted dry. The grids were washed three times with
water to remove residual salt. The samples were then either directly observed under
an electron microscope or stained with 1 % phosphotungstic acid, pH 7 .2 before

observation.

 

I”

 

 

 

36
RESULTS

Antibodies Directed Against B138

B1 38 was purified by affinity chromatography on Lac-Sepharose and subjected
to SDS-PAGE. The gel slices corresponding to the M, 38,000 region were used as
the immunogen for generating a rabbit antiserum against the lectin. The resulting
antiserum, designated rabbit anti-BJ38, yielded a single band (M, 38,000) on
immunoblotting the B138 sample (Fig. 1, lane a). This band was not observed with
control preimmune serum (Fig. 1, lane h). However, the same BJ38 sample yielded
multiple bands when immunoblotted with anti—Brj antibody (Fig. 1, lane e). This
latter antibody had been generated against heat-killed B. japanicum bacteria and had
been shown to react with LPS and other bacterial components (6). The " ladder "
pattern obtained when the B1 38 sample was immunoblotted with anti-Brj (Fig. 1, lane
e) was similar to those obtained when LPS or CPS isolated from B. japanicum were
immunoblotted with the same antibody (Fig. 1, lanes f and g). These results suggest
that although the BJ 38 sample was purified in terms of polypeptides, there was
nonetheless a small amount of contamination by LPS and CPS.

For the purpose of the intended use of the anti-BJ38 antiserum as a reagent for
immunolocalization, it was important to establish that anti-B1 38 reacted only with
B] 38 and not with any contaminating polysaccharides. The results indicate that anti-
BJ38 specifically reacted with BJ38, but did not react with either LPS or CPS

fractions derived from B. japanicum (Fig. 1, lanes b and c). As an additional

 

37

Figure 1. Immunoblot analyses of BJ38, CPS and LPS isolated from B.

japanicum. Lanes a. d.e.h. BJ38 sample (0.1 ng) isolated by 0.1 M Lac elution of

B. japanicum cell extracts that had been fractionated on a Lac—Sepharose column;
Lanes b,f,i, LPS (0.4 pg); Lanes c,g,j.

resed on 10% acrylamide gels, transferr
with anti—B138 (lanes a-c), monospecific
preimmune serum (lanes h-j); all sera w
tase conjugated goat anti-rabbit immuno
secondary antibody. Immunoreactive

ed to immobilon-P membrane and probed d
anti—BJ38 (lane d), anti—Brj (lanes e-g), an

globulin antibody (1 : 1000) was use? as the

materials were then developed with mtro blue
tetrazolium and 5-bromo-4-chloro—3-indolyl phosphate as substrates. The arrow on
the left indicates the position of migration of a polypeptide of M, 38,000, [Clam/e to
known molecular weight markers.

 

CPS (04 ug). The samples were electropho-

ere used at 1:20 dilution. Alkaline-phospha'

 

38

c d e f g h i J

-flr‘-

 

39
precaution, we affmity-purified the anti-B1 38 bound to the M, 38,000 band on

immunoblots. This affinity-purified monospecific antibody preparation yielded
identical results as the original anti-BJ38 antiserum in immunoblotting (e.g. Fig. 1,

lane d) and in immunofluorescence experiments.

Immtmgfluorescence of B1 38 on B. IE2. onicum Cell Surface

The availability of anti-BJ38 antibody allowed for the localization of BJ38 on
the bacterial surface. When B. japanicum cells were labeled with anti-B138 antiser-
um, followed by FITC-conjugated goat anti-rabbit IgG, irnmunofluorescent staining of
B138 was observed uniquely at one pole of the cell (Fig. 2a). The remainder of the
cell surface was not labeled. A similar pattern of labeling was also observed with the
monospecific affinity-purified anti-B1 38 antibody. Under our optimal conditions for
labeling, about 70% of the bacterial population showed anti-B1 38 staining.

In contrast, when anti-Brj antibody was used to label the B. japanicum cells,
immunofluorescence staining was observed evenly distributed throughout the entire
cell surface (Fig. 2b). The preimmune antibody did not show any fluorescent labeling

(Fig. 2c). These results indicate that BJ38 is located at one tip of the bacterium.

r ur onfirm inofPolar ' in fBJ
To confirm the polar localization of BI 38 on B. japanicum at the ultrastructur-
al level, B. japanicum cells were first incubated with anti-BJ38, followed by colloidal

gold-labeled anti-rabbit antibody. B138 was found at one pole of the bacterial cell

40

Figure 2. Immunofluorescent labeling of B. japanicum cell surface by anti-BJ38,
anti-Brj and preimmune sera. Cells were fixed on the activated coverslips, bIOCked
in 5% BSA/YEG for 2 h and then incubated in (a) anti-BJ38, (b) anti-Brit and (c)
preimmune serum for 4 h (all at 1:20 dilution). After washing in YEG, the samples
were incubated with FITC-conjugated goat anti-rabbit IgG antibody (1:50) for 1 h-
The samples were washed in PBS and observed under a fluorescence microscope.
PH, phase contrast; FL, fluorescence. Bar represents 3 yum.

 

41

0)
uy anti-l
slips, 11
1', all
these? b)
I) for 13:
result1

e)

 

 

 

 

 

 

 

 

 

 

 

 

 

42

Figure 3. Immunogold labeling of B. japanicum cell surface by anti-BJ38, antl-
Brj and preimmune sera. a, cells from suspension culture; b-d, cells from star
aggregates grown on a YEG agar plate. Suspension cells or star aggregates were b)
fixed with glutaldehyde, blocked in 5 % BSA/PBS for 1 h and then incubated m (a-
anti-BJ38; (c) anti-Brj; and (d) preimmune serum for 4 h (all at 1:20 dilution)- After
washing in PBS, the samples were incubated with goat—anti—rabbit IgG antibody
coupled with colloidal gold (1:5 dilution, particle size of 20 mm) for 1 h. The

samples were washed in PBS and observed under an electron microscope. Bar
represents 0.5 nm.

__,.-_..—

 

 

 

43

 

44

(Fig. 3a), as was seen with the irnmufluorescence data. Moreover, B1 38 appeared to
be in a tuft-like mass, localized at a distance away from the outer membrane of the
bacteria. A similar result was obtained by negative staining with phosphotungstic acid
after immunogold labeling. The positions of the gold particles did not appear to be
associated with any filamentous or pilus structure.

As expected, when anti-Brj antibody was used as the primary antibody,
immunogold labeling was observed throughout the cell surface (Fig. 3c). Preimmune

serum showed no labeling of the cell surface (Fig. 3d).

Identification of B1 38 at the Point of Contact

When B. japanicum was cultured on YEG agar plates, the bacteria tend to
form aggregates (stars) by attaching to each other at one of their tips. Previous study
had shown that this star formation can be inhibited by galactose and Lac (4). It was
proposed that BJ38 might be involved in the star formation of B. japanicum. When
these star-forming bacteria were used for BI 38 immunolocalization studies, irnmuno—
fluorescence was observed at the focal point of cell-cell contact (Fig. 4a). This result
suggests that BJ 38 is present at the contact point where the bacteria interact with each
other. This data is supported by the results of electron microscopic studies using
immunogold labeling techniques (Fig. 3b).

When B. japanicum was allowed to attach to SB-l cells or Lac—Sepharose
beads, immunofluorescence analysis also indicates that BJ38 is localized at the point

of contact (Fig. 4b,c). In contrast, there was no labeling of the other end of the

 

45

lfi'gure 4' lunofluorescent labeling of B. japanicum cells that were attached to

a) other star-forming B. japanicum; b) SB-l cells; and c) Lac-Sepharose bead. _
The B- Japonicum were blocked in 5% BSA/YEG for 2 h and then labeled w1th ant1-
BJ38 antiserum (1320 dilution) for 4 h. After washing in YEG, the samples were
mCUbated f°r 1 h With FITC‘CODJUgated goat-anti—rabbit IgG antibody (1:50). The
samples were washed with PBS and then viewed under a fluorescence mlcroscope.
PH indicates phase contrast; FL, fluorescence. Bar represents 3 am.

 

 

0)

attachedi
2 bed.
with 112*
as were
3). m , 1:)
amps.

c)

 

47

Fi u _ - . .
ings “1:2“ 1: EC SBA labeling 0f B. Japonrcum cells that were a) in suspension; b)
. , 10“} c) attached to SB-l cells. FITC-SBA (5 jig/ml) was incubated

 

48

 

II
C

 

49

bacterial cell. It should be noted that not all of the attached B. japanicum cells were
labeled, possibly due to the low sensitivity of the immunofluorescence technique or
hindered accessibility of the antibody. In any case, the results obtained with anti-
B1 38 were distinctly different from the corresponding data using SBA.

When FITC-SBA was used to label the B. japanicum cells, a cap-like structure
was observed covering a third of the bacterial cell (Fig. 5a). This was strikingly
different from the point fluorescence observed with anti-BJ38 (Fig. 2a). When FITC-
SBA was used to label the star-forming B. japanicum cells, FITC-SBA labeling was
shown to be at the pole of the bacterium away from the attachment point (Fig. 5b).
No fluorescent labeling was observed at the center of the stars, where B. japanicum
cells attached to one another. Similarly, SBA labeled B. japanicum cells which were
attached to SB-l cells at the end of the bacterium, away from the point of contact
(Fig. 5c). These data indicate that the receptor site for SBA binding is distinctly

different from the site where BJ 38 is localized.

 

 

 

 

50
DISCUSSION

The experiments reported in this paper were performed to determine the
localization of BJ38 on the bacteria, using an antiserum directed specifically against
the lectin. The results provide several key items of information. First, B1 38 is found
at the cell surface. Second, BJ38 is localized at the pole of the bacterial cell which is
involved in: a) homotypic adhesion to other star forming B. japanicum; b) heterotyp-
ic binding to SB-l cells; and c) attachment to Lac-Sepharose beads. Since these
binding processes occur in a polar fashion (4), the distribution of BJ38 has satisfied
three topological requirements necessary for it to mediate bacterial adhesion. These
are that B138 must be localized in a polar fashion, that it appears on the bacterial cell
surface, and that the lectin must be found in the region of cell-cell contact.

The pole of BJ38 localization is distinct from that labeled by SBA. Consistent
with our previous report (15), SBA labeled the pole of the bacteria away from which
bacterial attachment occurred. In addition, SBA also labeled approximately a third of
the cell surface. The latter result indicates that SBA-binding polysaccharides (12, 16,
17) were covering a third of the bacterial cell surface, forming a cap like structure.
This is consistent with the observations of Tsien and coworkers (18, 19). They
showed, by ruthenium red staining in transmission electron microscopy, that the
capsular polysaccharides in B. japanicum were unevenly distributed on one half of the
bacterium. They further showed that the internal structure of the bacterium can be

divided into two distinct poles: the nucleoid portion containing chromosome and

 

51

cytoplasmic components and the reserve polymer half which contains glycogen and
poly-beta-hydroxybutyrate granules. Ruthenium red staining indicates that SBA—
binding polysaccharides are found in the nucleoid portion of the bacteria (19). In
contrast, the granular portion is the half with which the bacterium participates in star
formation. In the present study, we have shown that B138 and the SBA receptor
reside at opposite ends of B. japanicum. The granular portion is thus the end at
which B1 38 would be located.

Tsien and Schmidt (19) observed tuft-like structures in B. japanicum and
termed them extracellular polar bodies. They contain fibrillar material of both
polysaccharide and protein. It was suggested that these extracellular polar bodies
mediate star formation. Indeed, our own ultrastructural localization of B138 in a tuft-
like mass at a distance away from the bacterial outer membrane is consistent with this
hypothesis and with the notion that the lectin plays a key role in the autoagglutination
event.

On the other hand, our electron microscopic studies also indicate that B1 38
does not appear to be associated with any well—defmed proteinaceous filaments. Such
filamentous pili have been found in R. lupini (20), R. melilati, R. leguminasarum, and
B. japanicum (21) and have been prOposed to mediate attachment (22). It has been
reported that the isolated pili of B. japanicum consists of protein subunits with M,
21,000 and M, 18,000 (23). The subunit molecular weight of B138 makes it unlikely
that it is a part of the isolated pilus structure. Moreover, our previous gel filtration

data, in which the purified B138 protein chromatographed to a position corresponding

 

 

 

 

 

 

52

to M, 38,000 (13), suggest that the lectin most likely does not form polymers leading
to a filamentous structure.

Taken all together, the results detailed in this paper provide further evidence to
support the proposed role of B138 in mediating the carbohydrate binding properties of
B. japanicum. Not only is B138 saccharide specificity similar to that of B.
japanicum, but its topological distribution is consistent with the polar binding
activities of the bacterium. The importance of this latter observation is further
demonstrated in studies with the N4 and N6 mutants (24). These mutants showed no
detectable amounts of B1 38 on the bacterial cell surface and a decreased ability to
nodulate soybean roots as compared to the wild—type B. japanicum. These results

indicate that B] 38 may play an important role in the symbiotic infection process.

 

 

 

1

 

 

53
ACKNOWLEDGEMENTS

We thank Dr. Natasha Raikhel for the use of the Zeiss fluorescence microscope.
We also thank Mrs. Linda Lang for her help in the preparation of the manuscript. This
work was supported by Grant GM45200 from the National Institutes of Health and by

the Michigan Agricultural Experimental Station.

 

 

 

 

10.

54
REFERENCES

Bauer, W. D. 1981. Infection of legumes by Rhizobia. Annu. Rev. Plant
Physiol. 32:407-449.

Halverson, L. 1., and G. Stacey. 1986. Signal exchange in plant-microbe
interactions. Microbial. Rev. 50:193-225.

Ho, S.-C., and J. W. Kijne. 1991. Lectins in rhizobium-legume symbiosis.
In Lectin Reviews. Vol. 1. D. C. Kilpatrick, E. Van Driessche, and T. C.
Bog-Hansen, editors. pp. 171-181.

Ho, S.—C., J. L. Wang, and M. Schindler. 1990. Carbohydrate binding
activities of Bradyrhizabium japanicum. I. Saccharide-specific inhibition of
homotypic and heterotypic adhesion. J. Cell Biol. 111:1631-1638.

Ho, S.-C., M. Schindler, and 1. L. Wang. 1990. Carbohydrate binding
activities of B. japanicum. H. Isolation and characterization of a galactose-
specific lectin. J. Cell Biol. 111:1639-1643.

Ho, S.-C., W. Ye, M. Schindler, and J. L. Wang. 1988. Quantitative assay
for binding of Bradyrhizabium japanicum to cultured soybean cells. J.
Bacterial. 170:3882-3890.

Matsumoto, I., Y. Mizuno, and N. Seno. 1979. Activation of Sepharose with
Epichlorohydrin and subsequent immobilization of ligand for affinity adsor-
bent. J. Biochem. 85:1091-1098.

Laemmli, U. K. 1970. Cleavage of structural protein during the assembly of
the head of bacteriophage T4. Nature (Land). 227:680-685.

Blum, H., H. Beier, and H. Gross. 1987. Improved silver staining of plant
proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 8:93-99
Hager, D. A., and R. R. Burgess. 1980. Elution of proteins from sodium
dodecyl sulfate-polyacrylamide gels. Anal. Biochem. 109:76-86.

 

 

ll.

12.

13.

14.

15.

16.

17.

18.

19.

20.

55
Smith, D. E. , and P. A. Fisher. 1984. Identification, developmental regula-

tion, and response to heat shock of two antigenically related forms of a major
nuclear envelope protein in Drasaphila embryos: Application of an improved
method for affinity purification of antibodies using polypeptides immobilized
on nitrocellulose blots. J. Cell Biol. 99:20-28.

Mort, A. 1., and W. D. Bauer. 1980. Composition of the capsular and
extracellular polysaccharides of B. japanicum. Plant Physiol. 66:158-163.
Carlson, R. W., R. E. Saunders, C. A. Napoli, and P. Albersheim. 1978.
Host-symbiont interactions. III. Isolation and partial characterization of
lipopolysaccharides from Rhizobium. Plant Physiol. 62:912-917.

Aplin, J. D., and R. C. Hughes. 1981. Protein derivatized glass coverslips
for the study of cell-to—substratum adhesion. Anal. Biochem. 113:144-148.
Ho, S.-C. 1991. Distinct entities between soybean agglutinin receptor and
soybean root hair binding site on Bradyrhizabium japanicum cell surface. In
Plant Biotechnology and Development. P. M. Gresshoff, editor. CRC Press,
Inc. pp. 25-32.

Bal, A. K., S. Shantharam, and S. Ratnam. 1978. Ultrastructure of
Bradyrhizabium japanicum in relation to its attachment to root hairs. J. of
Bacterial. 133: 1393-1400.

Bhuvaneswari, T. V., S. G. Pueppke, and W. D. Bauer. 1977. Role of
lectins in plant-microorganism interactions. Plant Physiol. 60:486-491.
Tsien, H. C. 1982. Ultrastructure of the free—living cell. In Nitrogen
Fixation. Vol. 2. W. J. Broughton, editor. Clarendon Press, Oxford. pp.
182-198.

Tsien, H. C., and E. L. Schmidt. 1977. Polarity in the exponential-phase
Rhizobium japanicum cell. Can. J. Microbial. 23:1274-1284.

Heumann, W. 1968. Conjugation in star forming Rhizobium lupini. Mol.
Biol. Biochem. 5:47-81.

 

 

 

 

 

 

 

21.

22.

23.

24.

56
Dazzo, F., J. W. Kijne, K. Haahtela, and T. K. Korhonen. 1986. Fimbrae,

lectins and agglutinins of nitrogen fixing bacteria. In Microbial Lectins and
Agglutinins. D. Mirelman, editor. John Wiley and Sons, New York, NY.
pp. 237-254.

Isaacson, R. E. 1985. Pilus adhesions. In Bacterial adhesion. D. C. Savage
and M. Fletcher, M., editors. Plenum Publishing Corp., New York. pp.
307-336.

Vesper, S., and W. D. Bauer. 1986. Role of pili (firnbrae) in attachment of
B. japanicum to soybean roots. App. Environ. Microbial. 52:134-141.

Ho, S.-C., 1. L. Wang, M. Schindler, and J. T. Loh. 1994. Carbohydrate
binding activities of Bradyrhizabium japanicum. III. Lectin expression,

bacterial binding, and nodulation efficiency. Plant J. 5:873-884.

 

 

 

 

CHAPTER III

CARBOHYDRATE BINDING ACTIVITIES
OF BRADYRHIZOBIUM JAPONICUM

Effect of Lactose and Flavones on the

Expression of the Lectin, B138

57

 

 

 

 

 

58
ABSTRACT

BJ38 is a galactose/lactosespecific lectin (M, ~ 38,000) found at one pole of
Bradyrhizobium japanicum. It has been implicated in mediating the adhesion of the
bacteria to soybean roots, leading to the establishment of a nitrogen-fixing symbiosis.
When the ligand lactose is added to cultures of the bacteria for at least one hour prior
to harvesting the cells for B1 38 isolation, the yield of the protein was found to be
elevated in a dose-dependent fashion. Half maximal stimulation was observed at ~ 50
uM; the effect was saturated at ~ 1 mM, where a 10-fold higher yield of BJ38 was
obtained. Saccharides with a lower affinity for BI 38 than lactose yielded a corre-
spondingly smaller induction effect when compared at a concentration of 1 mM. The
higher level of B1 38 induced by lactose is also manifested by an elevated amount of
B1 38 detectable at the cell surface and by a higher number of B. japanicum cells
adsorbed onto soybean cells. Surprisingly, the induction of B1 38 expression seen with
lactose was also observed with certain, but not all, flavonoids that induce the nod
genes of the bacteria; genistein mimicked the induction observed with lactose,

whereas luteolin failed to stimulate B138 production.

 

 

 

 

 

 

59
INTRODUCTION

In previous studies (1), we documented that Bradyrhizabium japanicum
exhibits four saccharide-specific binding activities: (a) adsorption to Sepharose beads
derivatized with lactose (Lac); (b) autoagglutination (star formation); (c) binding to
cultured soybean (SB-1) cells; and (d) adhesion to soybean roots. In all four of these
assays, galactose (Gal) inhibited the binding, but C-2 derivatives of the monosaccha-
ride (e. g., N -acetyl-D-galactosamine, (GalNAc)) failed to yield the same effect.
Mutants of B. japanicum, isolated on the basis of a defect in one binding activity (SB-
1 cell binding), showed a concomitant loss in the other three binding capacities.
These observations suggested that all of these carbohydrate-specific processes may be
mediated by the same component and mechanism.

On the basis of these observations, we began to search for Gal/Lac-specific
carbohydrate-binding proteins from B. japanicum. We reported the purification of
one such lectin, designated B1 38 (M, ~ 38,000), that bound Lac and Gal but showed
much lower affinity toward GalNAc (2). A polyclonal antiserum generated against
B1 38 was used in transmission electron microscopy and in confocal fluorescence
microscopy to localize the lectin at the cell surface of the bacterium (3). More
importantly, the lectin was found at only one pole of the cell, coincident with the
attachment site for: (a) autoagglutination of other B. japanicum cells; (b) adhesion to

soybean cells; and (c) adsorption to Lac-Sepharose beads. These results indicate that

 

 

 

 

 

6O .
the localization of B138 is consistent with a suggested role for this bacterial lectin in
the polar binding of B. japanicum to other cells and surfaces.

In the course of these studies, we observed that the yield of BJ38 purified
from B. japanicum was consistently higher when isolated from cultures containing
Lac than from corresponding cultures without the disaccharide. Therefore, a system-
atic study was undertaken to document the requirements, specificity , and the conse-
quences of this Lac effect. We report in the present communication the data derived
from such a study, as well as the surprising observation that the Lac induction effect
can also be mimicked by flavonoid compounds that are a part of the signaling

components between the Rhizobium bacteria and their leguminous hosts.

 

 

61
MATERIALS AND METHODS

Bacterial Cultures and Isolation of BJ38

B. japani cum cells (R110d) were obtained from the laboratory of the late
Dr. Barry Chelm (Michigan State University). The bacterial strain was maintained on
agar plates containing yeast extract gluconate (Y EG) as described before (3). Liquid
cultures of B. japanicum were initiated by inoculating B. japanicum (3-day-old cells)
from. YEG agar plates into 50 ml YEG and culturing the bacteria on a gyratory shaker
(100 rpm, 30°C) for two days. This culture was then inoculated into 2 liters of YEG.
After two days, aliquots of B. japanicum (300 ml) were inoculated into 1.5 liters of
YEG and the bacteria cultured for different lengths of time (see below).

Routinely, saccharides and flavonoids were added to this final liquid culture
step to determine their effect on B1 38 expression. Lac was obtained from Kodak
(Rochester, New York); mannose, GalNAc, genistein, apigenin, naringenin from
Sigma (St. Louis, MO); and glucose from MCB (Cincinnati, OH). Dr. Franz De
Bruijn (Michigan State University) kindly provided the luteolin.

For the isolation of BJ38, B. japanicum cells were harvested by centrifugation
(11,000 g, 15 min) in a GS3 rotor and the bacteria were lysed by passage through a
French Press (20,000 psi). The lysate was centrifuged at 27,000 g for 30 min and the
supernatant fraction was precipitated by ammonium sulfate (65 % saturation). The
precipitate was centrifuged and resuspended in 3 ml phosphate-buffered saline (PBS)

and dialyzed against PBS overnight at 4°C. The bacterial lectin was then isolated by

 

 

 

 

62

affinity chromatography on a Lac-Sepharose column (2). Lac-eluted fractions
containing BJ38 were identified by SDS-PAGE (4) and silver staining (5). Samples
were first concentrated by precipitation with 0.015 % deoxycholate - 7.2% trichloro-
acetic acid (6), washed with acetone, resuspended in sample buffer of the SDS-PAGE
system, and subjected to electrophoresis. The gels were silver-stained and the
intensities of the bands corresponding to B1 38 were quantitated on the basis of
comparison with the staining intensities of known amounts of an arbitrary standard,
carbonic anhydrase. Protein determination of cell extracts was performed by the

method of Bradford et al. (7).

Use of lmll Anti-BJ38 to Quantitate Cell Surface Expression of BJ38

The generation and characterization of the anti-B138 antiserum used in this
study have been reported previously (3). The immunoglobulin fraction of this
antiserum was isolated by adsorption onto protein A-Sepharose (Sigma, St. Louis,
MO) and was then labeled with 125I by the chloramine T method (8). Naml (1 mCi)
was added to 100 pl of antibody and chloramine T (1 mg/ml; 25 #1) was added to
start the reaction. The reaction was stopped with the addition of 25 pl of sodium
metabisulfite (1 mg/ml) in PBS. Free 12SI was removed by passage through an
AG1X8 ion exchange column (Bio-Rad Laboratories, Richmond, CA). Radioactivity
was determined in a 'y-spectrometer counter (LKB Instruments, Inc. , Rockville, MD).

For quantitation of cell surface B138, aliquots of B. japanicum cells (5 x 10°)

were first incubated in PBS containing 5 % bovine semm albumin (BSA) for 1 hour at

 

 

63
4°C. 12‘SI-labeled anti-BJ38 (specific activity = 11.7 Ci/ g) was added to the B.

japanicum suspension. After 4 hours, individual samples were filtered through a
membrane filter (pore size 0.22 am, Millipore, Naperville, IL) that had been pre-
soaked in 5% BSA/PBS. The membrane was washed three times in 0.1% BSA/PBS

and the bound radioactivity determined.

B. iaganicum Binding to SB-l Cells

 

SB-l cells, originally derived from soybean roots (Glycine max (L.) Merr. cv.
Mandarin) was kindly provided by G. Lark (Department of Biology, University of
Utah, Salt Lake City, UT). Continuous cultures were maintained in 1B5C media by
transferring 15 ml of culture (4-day-old) to 50 ml fresh 1B5C and culturing the
suspension cultures at 27°C on a gyratory shaker in the dark.

For quantitation of bacterial binding to SB—l cells, B. japanicum cells were
washed three times in YEG to remove the saccharides that had been added to the
YEG media. A binding assay described by Ho et al. (9) was then used. SB—l cells
(2-day-old, 0.5 ml of cell suspension (20% v/v)) were incubated with 109 B.
japanicum in a 35 mm culture dish. After incubation for 0.5, 2 and 4 hours, the cells
were transferred to 12 x 75 mm culture tubes and washed three times with PBS by
centrifugation (460 g, 2 min) and resuspension.

Bacterial cells bound to the soybean host were quantitated by radioimmunoas-
say, using anti-Brj antibody (antibody generated against whole B. japanicum cell) (9).

The cells were incubated with anti-Brj antibody (75 rig/ml containing 10° cpm [1251]

 

 

 

 

 

 

64

anti-Brj (specific activity 2.32 Ci/g)) at 4°C. The samples were washed three times
by centrifugation (460 g, 2 min) in 0.1% BSA/PBS and the bound radioactivity was
quantitated to compare the relative amount of B. japanicum binding to 83-1 cells. In
parallel, the conclusions derived from the quantitative assay were confirmed qualita-

tively by microscopic examination.

Fluorescence Microscopy

B. japanicum cells were attached to coverslips as described (3). The cells
were blocked in 5 % BSA/PBS: for 2 hours, followed by incubation in either
preimmune or anti-BJ38 antiserum (1:20 in 5% BSA/PBS). Following washing in
PBS three times, fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit
antibody (1:50 dilution; Boehringer Mannheim, Indianapolis, IN) in 5% BSA/PBS
was added. After 1 hour incubation, the samples were washed three times with PBS.
Stained cells were visualized with a fluorescence microscope (Zeiss D-7802
Obserkochen; excitation filter, 546 nm; chromatic beam splitter, 580 nm; barrier

filter, 590 nm).

Glutamine Smthetase Assay

Glutamine synthetase activity was determined using the y-glutamyl transferase
(y-GT) assay described by Bender et al. (10). Cell extract (pl amounts) was added to
the y-GT assay mixture consisting of 135 mM irnidazole, 18 mM hydroxylamine,

0.27 mM MnClz, 25 mM potassium arsenate, 0.36 mM sodium ADP, pH 7.55. The

 

 

 

 

 

65

samples were incubated at 37°C for 5 min. The reaction was initiated by addition of
L—glutamine (20 mM final concentration). Addition of "stop-mix" (0.2 M
FeCl3.6HZO, 0.12 M trichloroacetic acid, and 0.25 M HCl) terminated the reaction.
Precipitate was removed from the samples by centrifugation and the absorbance at 540
nm was determined. Using the value that 1 umol of glutamyl hydroxamate gives
0.532 units of absorbance at 540 nm as determined by Bender et al. (10), the specific
activity of glutamine synthetase was calculated and expressed as nmol y-

glutamylhydroxamate produced per minute per mg cell extract.

 

 

 

66
RESULTS

Lama Induction pf L138 Expression in B. (m nicum

In the protocol for the isolation of BJ38, liquid cultures of B. japanicum were
initiated by inoculating the bacteria maintained on agar plates in YEG medium. This
initial liquid culture (50 ml) was expanded twice (for details, see Material and
Methods). The final liquid culture (~ 2 liters) reached late logarithmic phase in ~ 27 h,
at which time the cells were harvested for lectin purification. Addition of Lac to the
final liquid culture resulted in a lO-fold increase in the yield of B138, when compared
to the corresponding yield from cultures without Lac (Fig. l). The effect of Lac was
dose-dependent, with saturation at ~ 1 mM. Half maximal effect was observed at ~ 50
uM.

Three sets of experiments were performed to determine the length of time
required to observe the Lac effect. First, Lac was added to a final concentration of 1
mM at various times during the final liquid culture and all the cultures were harvested
at the same time (i.e. at the end of the 30-h period of the final liquid culture). In this
scheme, the cell density and growth phase of the bacteria at the time of harvest were
kept constant, while the initial time (cell density and growth phase of bacteria) of
exposure to Lac, as well as the length of exposure to Lac, were allowed to vary. The
results showed that Lac added at early time points during the final liquid culture

yielded the greatest effect (Fig. 2A). Thus, addition of Lac at 2 h and 8 h of the final

67

Figure 1. Dose-response curve of the effect of Lac on the production

of BJ38. B. japanicum cells were cultured for 27 h in YEG media that had been
supplemented with different concentrations of Lac. The bacterial cells were lysed by
French pressing and the cell extracts were fractionated over a Lac-Sepharose column.
B138 was eluted with 0.1 M Lac. The amount of lectin was quantitated by comparing
the intensities of the silver-stained polypeptide bands (inset) with the staining intensi-
ties of known amounts of carbonic anhydrase, and is presented in the figure as ng of
B138 isolated per mg of cell extract.

 

68

 

 

 

 

 

12
1
ion _, “- -'-‘.-
1111111 3101 0 0.1 10 100 1000 5000 20000
wait 31
nosed o 8..
-- . . 0
WE )1 ° 6.
151 1,
gure. 1 0’ 1
~ g
‘ 8 41
.,
m l
a 2
c 1
0 .
.0001 .001 .01 .1 1 10 100 1000 10000100000

[Lac] (uM)

 

 

 

 

 

69

Figure 2. The effect of the growth phase of B. japanicum cells on the Lac
induction of BJ38 expression. Lac was added to a concentration of 1 mM at various
times during the final liquid culture (300 ml of a 2 liter B. japanicum culture
passaged into 1.5 liter YEG). In A, all cultures were harvested at the same time (~ 30
h after initiation of the final liquid culture). In B, the cultures were harvested at
different times (in each case, 4 h after the addition of Lac). The cell density of the
sample at the time of harvest is presented as absorbance at 620 nm (A620 0). The
amount of B138 isolated is expressed as ng B138/mg cell extract (A). The inset
shows the intensities of the silver-stained bands of the isolated BJ38 normalized to
total protein added. These experiments were repeated three times, with essentially the
same results.

70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 limolao 93556 no: 822:: W
1 ..t 8 6 4 2 0 ._ lilac 95230 no: 8232 _
. .17 . . . L . . i 5 M 0.0 A“ “I 00 _
n . 1r 1P1 .1? 3 _
L

O

3

H 2

u 2

Him

«a

«a

3 1 0

 

their

thn
1111111

saith
math
and
201) 1

 

 

71

liquid culture resulted in an approximately 8 times higher yield of BJ38 than if Lac
was added immediately before the harvest. In such an experimental scheme, we
could not distinguish whether the observed differences were due to a difference in the
length of exposure of the B. japanicum cells to Lac or to the time (cell density and/or
growth phase) of the initial exposure.

Therefore, the second experimental scheme tested the effect of Lac addition
and harvest at various times throughout the final liquid culture period of 30 hours,
keeping the length of exposure constant at 4 h. The results showed that exposure to
Lac early during the final liquid culture period yielded the highest amount of B138,
with a monotonic decrease in the yield with later times of exposure (Fig. 2B). It
should be noted that the yield of BJ38 is expressed as ng of B138 isolated per mg of
cell extract so that the value is normalized to the total cell number» used in the sample.
Thus, the results suggest that the cell density and/or growth phase of the bacteria was
critical at the time of Lac addition. In this experiment, however, the density (and the
growth phase) of the B. japanicum cells at the time of harvest for BJ38 isolation was
different for each sample. Therefore, a third experimental scheme was devised to
eliminate this variable. Lac was added at the initiation of the final liquid culture
period and the cells were harvested after different lengths of exposure to the saccha-
ride. The results showed that the control samples (B. japonicum cells without any
Lac) yielded "baseline" levels of B138, while Lac-treated cells yielded increasing
amounts of B1 38 with longer exposures to the saccharide (Fig. 3). Exposures as short
as three-quarters of an hour were sufficient for appreciable enhancement of the yield

of B138. The high yield of BJ38 in the sample that received 21 h of exposure to Lac

 

ll

 

 

72

 

ount of BJ38 isolated is expressed as ng BJ38/mg cell
c intensities of the silver-stained bands of the isolated
otein content. This experiment was repeated three tlmes

extract. The inset shows th
BJ38 normalized to total pr
with similar results.

 

 

no BJ38/mg cell extract

73

 

10

 

1-N0hapten ~o~'-~N

I we 9 .

  

  
   

   

O 0.75

7 7

tlme (II)

 

 

21

74

was important because this showed that the high cell density and the stationary phase

of the bacteria at the time of harvest did not impair the Lac effect.

Specificity of the Lactose Induction Effect

The specificity of the Lac effect was tested in terms of two key questions: (a)
can other saccharides mimic the effect of Lac? and (b) does Lac affect the expression
of other B. japanicum proteins, besides B138? Glutamine synthetase was chosen as an
irrelevant protein control, to be compared to B1 38. Thus, aliquots of the same B.
japanicum culture were inoculated into the final liquid YEG step that contained either
no hapten (control) or various saccharides (1 mM). For each of these final liquid
cultures, the amount of B138 isolated and the specific activity of glutamine synthetase
were quantitated (Fig. 4). Neither Lac, nor any of the other saccharides, altered the
level of activity of glutamine synthetase. On the other hand, the stimulatory effect of
Lao on the yield of B138 was much more pronounced than the other saccharides. The
8-fold induction due to Lac should be compared to the 4—fold induction seen with Gal
and the 2.5-fold effect due to GalNAc. This relative order of the ability to induce
B1 38 expression is similar to the relative binding affinity of BJ38 for the three
saccharides (2). Finally, Man, which does not bind to B138 (2), had little or no

effect on B138 expression.

Effect of Lactose Induction on Cell Surface Expression of BJ38

 

Thus far, the expression of BJ38 has been assayed at the level of amount of

the lectin that can be purified by Lac-Sepharose affinity chromatography. The

 

 

75

Figure 4. Saccharide specificity in the induction of B138. Final liquid

cultures of B. japanicum were incubated for 10 h in the presence of 1 mM concentra-
tions of Lac, Gal, GalNAc, and Man, or without any saccharide. The amount of
BJ38 isolated (inset) was quantitated and is expressed as ng BJ38/mg cell extract. In
parallel, the specific activity of glutamine synthetase in the cell extracts was also
determined (see Materials and Methods). This experiment was reproduced three
times.

 

76

. 9:83.: 83:58:

E330=§23u no: 83253

6 4 2
D b P

LLO

 

 

'll' 11’

No ‘
Lac Gal GalNAc Man hapten

and- an“?

 

 

Man

No hapten

 

V/

 
     

  

GalNAc

r//////////////2a
r////////////e.

 

8

..o . a... . a . 10
82:8 :3 9533 a... I

 

net.
1 l
at '1
rec

saccharlde

 

qut
ma:
cell

for

of 1
Th
wh

ass

det

stat
cul
10'
ant

inc

77

question is, therefore, raised whether the induction of B1 38 expression by Lac is also
manifested at the level of increased amounts of B138 at the cell surface. B138 at the
cell surface was detected in two assays, using a polyclonal rabbit antiserum specific
for the lectin (3). First, rabbit anti-B138 was labeled with 1251 and the amount of
antibody required to saturate the B1 38 found on the cell surface of B. japanicum was
determined by quantitating the radioactivity bound as a function of antibody added
(Fig. 5, inset). Using a saturating dose of 125I-labe1ed anti-BJ38 (~ 44 ng), the level of
B1 38 expressed on the cell surface was compared for B. japanicum cells derived from
final liquid cultures containing various saccharides (1 mM) or containing no hapten
(control). The presence of Lac in the final liquid culture greatly increased the amount
of BJ38 at the cell surface, accessible to binding by 125I-labeled anti-BJ38 (Fig. 5).
The other saccharides had little or minimal effects. This includes GalNAc and Gal,
which yielded 2.5— to 4-fold induction effects when the expression of B138 was
assayed at the level of isolation of bulk amounts of the lectin.

Second, immunofluorescence staining with rabbit anti-BJ38 was used to
determine the fraction of B. japanicum expressing BJ38 at the cell surface. Approxi-
mately 40 % of the bacteria derived from final liquid culture containing Lac (1 mM)
stained with anti-B138 (Fig. 6a). In contrast, bacteria derived from final liquid
cultures that contained no hapten or a control saccharide (1 mM glucose) yielded
10% immunofluorescent cells (Fig. 6b and 60). Thus, assaying at the level of
antibody-accessible B138, the induction effect of Lao is manifested mainly as an

increase in the percent of cells exhibiting cell surface exposed lectin.

 

 

78

Figure 5. Quantitation of the effect of saccharides on the cell surface

expression of BJ38. B. japanicum cells were cultured in the presence of Lac, Gal,
GalNAc, Man, and Glc (all at 1 mM), or without any saccharide. Bacteria from each
treatment (5 x 10° cells) were then blocked with 5% BSA/PBS for l h and then
incubated for 4 h with 44 ng of [”51]anti-B138. The bacteria were then filtered
through a millipore filter (pore size 0.22 am) and the membrane washed three times
in PBS. The amount of radioactivity retained on the membrane was then determined.
The data represent means :1; standard deviations of triplicate determinations. Inset:
Determination of the amount of anti-B1 38 needed to saturate the B138 found on the
cell surface of B. japanicum. B. japanicum (5 x 10°) were incubated with various
amounts of [1251]anti-BJ38 for 4 h. B. japanicum was then filtered through a millipore
filter and the amount of bound radioactivity determined.

 

 

79

 

 

    
     
 

 

n I A

  

DIV-#0)

 

 

 

cpm bound 1: 10'4

 

6‘ '.
‘1'
. c - t r . . I . , .
11 v- 0 1o 20 30 4o 50
at x anti-8.138(ng)
ll '24..
l :1
it 3
l“ s
i ”2‘
11
O"!
Lac Gal GalNAc Man Glc No hapten
saccharide

 

 

 

80

Figure 6. The effect of saccharides on the expression of BJ38 at the cell
surface of B. japanicum as detected by immunofluorescence staining with anti-
BJ38. Bacteria cells were cultured in the presence of 1 mM concentrations of (a)
Lac; (b) Glc; and (e) no hapten for 24 h. The cells were then stained with anti-BJ38
as described in the Materials and Methods. PH, phase contrast; FL, fluorescence.

 

81

 

inn ‘
ms of 11* 1
111 in? 1

l

 

 

all

111]

1g:
b0

 

82

Effect of Lactose Induction on the Adhesive Properties of B. iaponicum

Since Lac induction of BJ38 expression is manifested not only at the level of
total BJ38 isolatable but also in terms of the fraction of cells exhibiting BJ38 detect-
able at the cell surface, the question is raised whether the capacity of B. japanicum to
bind to soybean cells is also increased upon Lac induction. Bacteria derived from
Lac-containing final liquid cultures indeed showed a high rate of adhesion to cultured
soybean SB-l cells (Fig. 7). Within half an hour following co—culture of SB-l cells
with B. japanicum, bacteria derived from Lac—induced cultures showed many cells
attached to the soybean cells, compared to the corresponding co—culture using
uninduced bacteria.

We had previously developed a radioimmunoassay, using an antiserum raised
against whole B. japanicum cells (anti-Brj), to quantitate the number of bacteria
bound on soybean cells (9). The differences in the number of B. japanicum attached
to SB-l cells, qualitatively represented in the photomicrographs of Figure 7, are
corroborated by the quantitative data (Fig. 8). There was about a 4—fold increase in
the number of attached cells, comparing Lac—induced versus control samples, half an

hour after co-culture. This difference decreased to about 2-fold upon longer term

culture.

 

83

Figure 7. The effect of inclusion of Lac in the final liquid culture of

B. japanicum on the subsequent binding of the bacteria to SB-l soybean

cells as observed by light microscopy. B. japanicum (109 cells) derived from final
liquid cultures with (YEG + Lac) and without (YEG) saccharide were incubated with
0.5 ml cell suspension (20% v/v) of 83-1 cells as described in Materials and Meth-
ods. Light micrographs were taken at (a) 0.5, (b) 2, and (c) 4 h after B. japanicum
was added to the SB-l cells.

 

 

YEG+LAC YEG

 

 

 

85

Figure 8. The effect of inclusion of Lac in the final liquid culture of B.
japanicum on the subsequent binding of the bacteria to SB-l soybean ‘
cells as quantitated by a radioimmunoassay. The experiment was carried out as in
Figure 7 and the number of bacteria binding to the SB—l cells was quantitated usmg

[lzsflanti-Brj as detailed in Materials and Methods. The data represent means i
standard deviations of triplicate determinations.

86

 

 

 

 

 

”.2. x 252. Ego

 

 

87

Figure 9. Comparison of the effect of flavones and saccharides on the
induction of BJ38 expression in B. japanicum cells. Final liquid cultures of B..
japanicum were incubated for 20 h in the presence of flavones (2 11M) or saccharides
(1 mM) or with no addition. The amount of B1 38 isolated (inset) was quantitated and
is expressed as ng BJ38/mg cell extract. In parallel, the specific activity of glutamine

synthetase in the cell extracts was also determined. This experiment was repeated
three times with similar results.

 

 

its;
the:
F 0151‘;

 

 

.ng BJ38/mg cell extract

88

 

 

 

 

wet-~54 1.

Gen Lut Api Nar Lac Man N0

 

 

Genistein Luteolin Aplgenin Narlngenln

lnducer

hapten

 

 

Lac

 

 

Man No hapten

E

or x toenxe ueo Btu/mulnown)
esoteutulls eululetms

‘-

.0V

89
Induction pf Q38 Expression by Flavones and Derivatives

Using the same protocol for Lac induction, we have also tested the effect of
several flavonoid compounds implicated in the induction of the nod genes that are
required for early host responses in the Rhizobium-host plant signalling ( 11-15).

Thus, the flavonoids (2 uM) were included in the final liquid culture for 20 h prior to
harvesting the bacterial cells for BJ38 isolation. The isoflavone genistein yielded a
strong induction effect on BJ38 expression, comparable to that seen with the "positive
control," Lac (Fig. 9). In contrast, the flavone luteolin failed to yield any effect; the
amount of BJ38 isolated corresponded to those of "negative controls," Man and no
hapten inducer. Finally, the flavone apigenin and the flavonone naringenin resulted in

intermediate levels of induction of BJ38 expression.

90
DISCUSSION

The key findings of this study include: (a) Carbohydrate ligands of the lectin
BJ38 can regulate the level of the protein; (b) This is manifested both in terms of the
amount of protein that can be isolated on the basis of its carbohydrate-binding activity
and in terms of cell surface exposed protein accessible to binding by a specific
antibody; (c) Among the saccharides tested, Lac was the best inducer, with a half

maximal effect observed at ~ 50 11M and saturation at ~ 1 mM. When tested at 1 mM,

 

the order of potency in stimulating the expression of B1 38 was Lac > Gal >
GalN Ac.
There does not appear to be a direct correspondence between the amount of

B1 38 isolated by affinity chromatography and the amount of B1 38 detected at the cell

 

surface. Lac induced an 8- to 10-fold increase in the yield of B1 38, but the increase
in the percent of cells with B1 38 detectable at the cell surface is only about 4-fold.
Moreover, the increase in B1 38 assayed by these methods translates into a 2-fold
increase in the number of cells adhering to soybean cells. These results raise the
question whether the increase in B1 38 is due to an elevated expression of the lectin in
cells that already contain the protein or may represent a recruitment of a new
subpopulation of B. japanicum cells that do not exhibit B1 38 in the absence of the
carbohydrate ligand. It is not known how this subp0pulation of bacteria might be

different from those B. japanicum cells expressing the protein.

 

 

 

 

91

At the molecular level, it is not known whether the Lac induction effect occurs
at the transcriptional or post-transcriptional levels. For example, it is possible that
ligand binding merely protects or sequesters the lectin from proteolytic degradation,
thus resulting in high levels of the lectin. It seems more likely, however, that the
induction effect is a result of transcriptional regulation of the B1 38 gene.
Transcriptional regulation has been observed in the control of several saccharide
binding proteins which mediate the transport of maltose, arabinose, xylose and lactose
into bacteria (16-19). In these systems, the genes encoding the binding proteins are
located in operons that are regulated by proteins which either activate or repress
transcription of the operon. The transcription of these genes are regulated by the
haptens of their respective gene products. In the case of B138, Lac is not the only
saccharide inducer. Moreover, the potency of various saccharides in stimulating B1 38
expression follows the same order of affinities as between the lectin and its carbohy-
drate ligands. For example, we had previously quantitated that the binding of the
disaccharide Lac was 13-fold higher than that of Gal, which in turn was about 18—fold
higher than that of GalNAc (2). Consistent with this relative order of affinities, Lac
induced an 8- to 10-fold increase in B138, while the effects of Gal and GalNAc were
4- and 2.5-fold, respectively. This raises the distinct possibility that it is BJ38 itself
that is binding to the promoter region and regulating the transcription of its own gene.
For example, binding of carbohydrate ligands may preclude B1 38 from repressing the

transcription of its own gene, possibly due to a ligand-induced conformational change.

 

 

 

92

To our surprise, we found that several flavonoid compounds known to regulate
the nod genes of Rhizobium (ll-15) also induced B138 expression. In accord with
previous reports (11,20), genistein, which was most potent in inducing nodABC
expression, was also the most potent inducer of BJ38. Similarly, the flavone luteolin
had negligible effects with respect to both nodABC and B1 38 induction and apigenin
yielded induction levels intermediate between genistein and luteolin in both assays.
Naringenin also induced B1 38, at a level below that of genistein but comparable to
apigenin; this flavanone had no effect on nodABC induction in B. japanicum cells
(11,20).

These results raise the possibility that the B1 38 gene, demonstrated to be
saccharide-sensitive, is also controlled by N odD, which mediates the activation of
other nod genes (21,22). This scheme suggests that BJ38 may belong to the family of
nod genes necessary for successful infection of the host plant. Such a possibility is
intriguing in light of the fact that sequence analyses of known nod genes suggest that
the nodABC, nadH, nadL, nadM and nodZ gene products are capable of binding
carbohydrate (22-25). These carbohydrate binding proteins function as enzymes
proposed to be involved in the synthesis of low molecular weight lipooligosaccharide
compounds such as the nod factors essential for successful nodulation. In our present
case, however, B1 38 would be a carbohydrate binding product of a nod gene whose
function is to mediate the attachment of B. japanicum to soybean roots (1). This
present scheme, however, does not preclude the possibility that flavonoids may

stimulate B1 38 transcription through bacterial response pathways other than nadD.

 

 

 

93

For instance, studies by Parniske et al. (26) have also demonstrated isoflavonoid-
induced resistance of B. japanicum to the phytoalexin glyceollin. This resistance has
been demonstrated in nadD,D2YABC deletion mutants, suggesting alternative sites of
genistein recognition.

At the functional level, the reasons for such a dual control of B1 38 expression
can only be speculated at this point. Lac and Gal could function to enhance B.
japanicum attachment to soybean roots. Lac and Gal are major constituents of the
plant cell wall and may serve as a plant signal to enhance bacterial attachment to the
soybean root through elevated levels of B138. Our present observation that an
increase in B. japanicum binding to soybean roots is paralleled by an increase in the
levels of BJ38 provide correlative evidence that this may indeed be so. A similar role
for saccharide mediated induction has been proposed for Agrabacterium tumefaciens.
Ankenbauer and N ester have reported that ChvE, a periplasmic sugar binding protein
mediates the induction of the vir genes in A. tumefaciens (27, 28), leading to the
infection of the host plant. As most of the sugars able to bind to ChvE and induce
the vir genes are constituents of plant cell wall polysaccharides, it was postulated that
these compounds served to signal the bacterium of the presence of the host plant and,
therefore, activate the bacterium for infection. In a similar fashion, soybean wall
constituents, specifically Lac and Gal, may serve as signals to B. japanicum and
activate that bacterial strain’s attachment to soybean roots.

Flavonoids could also function to enhance B. japanicum attachment to soybean

roots. In previous studies (29), we had documented that both BJ38 and B. japanicum

 

94

bind preferentially to the emergent root hair zone of the soybean root. This preferen-
tial binding has been attributed to recognition of specific carbohydrate structures on
the soybean root by BJ38. In studies on the spatial distribution of flavones capable of
inducing nod gene expression along the plant root surface, the emergent root hair
zone has also been shown to be the zone of maximum induction of bacterial nod genes
by flavonoids (30—33). Given these observations and the results documented in the
present study, induction of BJ38 expression by flavonoids would lead to an enhanced
binding of B. japanicum to soybean roots and, in particular, to the emergent root hair

Z0116.

 

 

 

 

95
ACKNOWLEGEMENTS

We thank Mrs. Linda Lang for her help in the preparation of the manuscript.
This work was supported by Grant GM45200 from the National Institutes of Health and

by the Michigan Agricultural Experiment Station.

10.

96
REFERENCES

Ho, S.-C, 1. L. Wang, and M. Schindler. 1990. Carbohydrate binding
activities of Bradyrhizabium japanicum. I. Saccharide-specific inhibition of
homotypic and heterotypic adhesion. J. Cell Biol. 111:1631-1638.

Ho, S.-C, M. Schindler, and 1. L. Wang. 1990. Carbohydrate binding
activities of Bradyrhizabium japanicum. II. Isolation and characterization of a
galactose-specific lectin. J. Cell Biol. 111:1639-1643.

Loh, J. T, S.-C Ho, A. W. deFeijter, J. L Wang, and M. Schindler. 1993.
Carbohydrate binding activities of Bradyrhizabium japanicum: unipolar
localization of the lectin B138 on the bacterial cell surface. Proc. Natl. Acad.
Sci. USA 90:3033-3037.

Laemmli, U. K. 1970. Cleavage of structural proteins during assembly of the
head of bacteriophage T4. Nature (Lond). 227:680—685.

Blum, H., H. Beier, and H. Gross. 1987. Improved silver staining of plant
proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 8:93-99.
Peterson, G. L. 1977. A simplification of the protein assay method of Lowry
et al. which is more generally applicable. Anal. Biochem. 83:346-356.
Bradford, M. 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principles of protein-dye binding.
Anal. Biochem. 72:248-254.

Langone, 1.1. 1980. 125I-labeled protein A: reactivity with IgG and use as a
tracer in radioimmunoassay. Methods Enzymol. 70:356-361.

Ho, S.-C., W. Ye, M. Schindler, and J. L. Wang. 1988. Quantitative assay for
binding of Bradyrhizabium japanicum to cultured soybean cells. J. Bacterial.
170:3882-3890.

Bender, R. A., K. A. Janssen, A. D. Resnick, M. Blumenberg, F. Poor, and B.
Magasanik. 1977. Biochemical parameters of glutamine synthetase from
Klebsiella aerogenes. J. Bacterial. 129:1001-1009.

20.

97
Kosslak, R. M, R. Bookland, J. Barkei, H. E. Paaron, and E. R. Appelbaum.

1987. Induction of Bradyrhizabiumjaponicum common nod genes by isoflavones
isolated from Glycine max. Proc. Natl. Acad. Sci. USA 84:7428-7432.
Banfalvi, Z., A. Nieuwkoop, M. Schell, L. Besl, and G. Stacey. 1988.
Regulation of nod gene expression in Bradyrhizobium japanicum. Mol Gen Genet
214:420-424.

Peters, K. N, J. W. Frost, and S. R. Long. 1986. A plant flavone luteolin
induces expression of Rhizobium melilati nod genes. Science 233:977—979.
Firmin, J. L, K. E. Wilson, L. Rossen, and A. W. B. Johnston. 1986.
Flavonoid activation of nodulation genes in Rhizobium reversed by other
compounds present in plants. Nature (Lond). 324290—92.

Zaat, S. A. J., C. A. Wijffelman, H. P. Spaink, A. A. N. van Brussel, R. J. H.
Okker, and B. J. J. Lugtenberg. 1987 . Root exudates of various host plants of
Rhizobium leguminasarum contains different set of inducers of Rhizobium
nodulation genes. J. Bacteriol. 1692198—204.

Ahlem, C., W. Huisman, G. Neslund, and A. S. Dahrns. 1982. Purification and
properties of a periplasmic D-xylose-binding protein from Escherichia cali K-12.
J. Biol. Chem. 257:2926-2931

Williams, S. G., J. A. Greenwood, and C. W. Jones. 1992. Molecular analysis
of the lac operon encoding the binding-protein-dependent-lactose transport system
and B-galactosidase in Agrabacterium radiabacter. Mal. Microbial. 6: 1755-1768.
Kolodrubetz, D., and R. Schleif. 1981. Regulation of the L—arabinose transport
opersons in E. Coli. J. Mol. Biol. 151:215-227.

Débarbouillé, M., H. A. Shuman, T. J. Silhavy, and M. Schwartz. 1978.
Dominant constitutive mutations in malT, the positive regulator gene of the
maltose regulon in Escherichia cali. J. Mol Biol 124:359-371.

Kosslak, R. M., R. S. Joshi, B. Bowen, H. E. Paaren, and E. Applebaum. 1990.
Strain-specific inhibition of nod gene induction in Bradyrhizabium japanicum by

flavonoid compounds. Appl. Environ. Microbial. 56:1333-1341.

 

 

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

98

Long, S. R. 1989. Rhizobium-legume nodulation: life together in the
underground. Cell 56:203—214.

Dénarié, J., F. Debéllé, and C. Rosenberg. 1992. Signaling and host range
variation in nodulation. Annu. Rev. Microbiol. 46:497-531.

Spaink, H. P. 1992. Rhizobial lip—oligosaccharides: answers and questions.
Plant Mal. Biol. 20:977-986.

Fisher, R. F., and S. R. Long. 1992. Rhizobium-plant signal exchange. Nature
(Lond). 357:655-660.

Stacey, G. S., S. Luka, J. Sanjuan, Z. Banfalvi, A. J. Nieuwkoop, J. Y. Chun,
L. S. Forsberg, and R. Carlson. 1994. nodZ, a unique host—specific nodulation
gene, is involved in the fucosylation of the lipooligosaccharide nodulation signal
of Bradyrhizobium japanicum. J. Bacterial. 1762620—633.

Parniske, M., B. Ahlborn, and D. Werner. 1991. Isoflavonoid-inducible
resistance to the phytoalexin glyceollin in soybean rhizobia. J. Bacterial.
173:3432-3439.

Huang, M. W., G. A. Cangelosi, W. Halperin, and E. W. Nester. 1990. A
chromosomal Agrobacterium tumefaciens gene required for effective plant signal
transduction. J. Bacteriol. 172:1814-1822.

Ankenbauer, R. G., and E. W. Nester. 1990. Sugar—mediated induction of
Agrobacterium tumefaciens virulence genes: specificity and activities of
monosaccharides. J. Bacterial. 172:6442-6446.

Ho, S.—C., J. L. Wang, M. Schindler, and J. T. Loh. 1994. Carbohydrate
binding activities of Bradyrhizabiumjaponicum. III. Lectin expression, bacterial
binding and nodulation efficiency. Plant J., in press.

Redmond, J. W., M. Barley, M. A. Djordjevic, R. W. Innes, P. L. Kuempel,
and B. G. Rolfe. 1986. Flavones induce expression of nodulation genes in
Rhizobium. Nature (London) 323:632—635.

 

 

 

31.

32.

33.

99
Peters, N. K., and S. R. Long. 1988. Alfalfa root exudates and compounds

which promote or inhibit induction of Rhizobium melilati nodulation genes. Plant
Physiol. 88:396-400.

Djordjevic, M. A., 1. W. Redmond, M. Batley, and B. 1. Rolfe. 1987. Clovers
secrete specific phenolic compounds which either stimulate or repress nod gene
expression in Rhizobium trifolii. EMBO J. 6:1173-1179.

Gulash, M., P. Ames, R. C. Larosiliere, and K. Bergman. 1984. Rhizobia are
attracted to localized sites on legume roots. Appl. Environ. Microbiol. 48:149—
152.

 

CHAPTER IV

CARBOHYDRATE BINDING ACTIVITIES
OF BRADYRHIZOBI UM JAPONICUM
Analysis of the effects of saccharides and flavones

on the expression of BJ38 at the mRNA level.

100

 

 

 

 

 

 

 

101
FOOTNOTE

1The generation of the PCRl probe and Southern blot analyses were performed by Dr.

Siu-Cheong Ho.

102
ABSTRACT

B138 is a galactose/lactose-specific binding protein implicated in mediating the
adhesion of Bradyrhizabium japanicum to soybean roots. In previous studies, we had
demonstrated that B138 expression, assayed at the polypeptide level, was increased
when B. japanicum cells were cultured in the presence of either lactose or the
flavone genistein. In the present study, we have compared the effect of these
inducers on the expression of BJ38 mRN A. Treatment of B. japanicum cells with
saccharides and flavones resulted in enhanced B138 mRN A expression. Lactose
yielded the greatest effect on BJ38 expression, followed by galactose; mannose had
little effect on BJ38 mRNA levels. Of the flavones tested, induction of B138 mRN A
was greatest with genistein, followed by apigenin. Luteolin showed little increase in
BJ38 mRN A levels. Interestingly, a similar order of efficacy for induction by
saccharides and flavones was also observed for the expression of the nadD, gene in B.

japanicum.

103
INTRODUCTION

The rhizobium-legume interaction leads to the invasion of the roots of the host
plant and the formation of a nitrogen-fixing symbiosis. A key step leading to
successful infection and nodulation involves the attachment of bacteria to the host
root. In the Bradyrhizabium japanicum-soybean interaction, the attachment of
bacteria to the host-legume is galactose (Gab-specific (1-3), suggesting the
involvement of a carbohydrate-binding protein. Consistent with this observation, we
have described the isolation of a carbohydrate binding protein, BJ38, from B.
japanicum. B138 exhibits a saccharide specificity similar to that of bacterial adhesion
to soybean roots (4). In addition, the lectin has also been immunolocalized at the
attachment site of bacterial attachment to soybean cells (5). Purified B138 binds to
soybean roots at sites coincident with B. japanicum attachment: namely, at the
emergent root hair zones (6). These regions had previously been demonstrated by
Bauer and coworkers (7, 8) to be the most susceptible to initiation of infection by B.
japanicum. More recently, we have reported that the expression of B138 can be
induced by both lactose (Lac) and the isoflavonoid genistein (9). As genistein is a
potent inducer of the nod genes of B. japanicum, this latter result suggests the
possibility that BJ38 may be a member of the nod gene family.

The nod genes comprise a set of key bacterial elements in the infection process
and are transcribed in response to specific flavonoid compounds secreted from the

host-plant (10-12). This requires the presence of the nadD gene product (13-15),

104

which, in association with the appropriate flavonoids, binds to the nod box promoter
sequence preceeding the nod genes and activates the transcription of these genes (16,
17). In B. japanicum, two copies of nadD, nadD, and nodDz, have been identified
(18). Of these two nadD genes, only nadD, is both inducible by isoflavonoids and
necessary for nod gene induction. In this present study, we document results which
characterize the effect of saccharides and flavonoid compounds on the expression of
B138 mRNA . Simultaneously, we also monitored the effects of these compounds on

nadD, mRNA expression.

105
MATERIALS AND METHODS

Cell Culturg

B. japanicum cells (R110d) were maintained on YEG (yeast-extract-gluconate)
agar plates for 3 days as described previously (9). The bacterial cells were then
inoculated into 50 ml YEG and cultured in a gyratory shaker for 2 days. This culture
was then transfered to 2 liter of YEG and grown for 2 more days. Aliquots of 300
ml of the 2 liter bacterial cultures were then transferred to 1.5 liter YEG to obtain the
final liquid culture step. On the initiation of the final liquid culture, saccharides and
flavones were added to a final concentration of 1 mM, and 2 pM, respectively and the
bacteria cells were cultured for 10 h prior to analysis. Lac was obtained from Kodak
(Rochester, New York); mannose, genistein, apigenin, naringenin from Sigma (St.
Louis, MO). Luteolin was kindly provided by Dr. Franz De Bruijn (Michigan State
University).

The R. melilati 1021 and R. leguminosarum bv. trifolii (R. trifolii) ANU843
strains were kindly provided by Dr. Rawle Hollingsworth (Michigan State University)
and maintained on agar plates containing Minimum Bergensens Media III (MBM)
(19). Liquid cultures of these strains were initiated by inoculation of the bacterial
cells (2 day old, MBM agar plates) into 50 ml MBM. The bacterial cells were grown
for 12 h at 30°C before being transfered to 1 liter of MBM. This final 1 liter culture

was then allowed to grow for 10 h.

 

 

 

106

Determination of the Partial Amino Acid Sequence of BJ38

Purified B138 (12 pg) was digested with Staphylococcus aureus V-8 protease
(Miles Laboratories, Naperville, IL) as described by Cleveland et al. (20). The
digestion was carried out with 0.6 pg of enzyme for 0.5 h. The peptides generated
were subjected to SDS-PAGE electrophoresis (21), transfered to a problot membrane
(Applied Biosystems, Foster City, CA) and the digestion products revealed by
Coomassie blue staining. Amino acid sequencing was performed by the Edman
degradation method (22) on a Model 477A Gas Phase Sequencer (Macromolecular

Structure Facility, Michigan State University).

Polymerase Chgin Reaction (PCR) Amplification and Nucleotide Sequence of

PCR products

 

Using the partial amino acid sequences obtained for B1 38, degenerate
oligonucleotide primers were synthesized based on the codon usage in B. japanicum.
PCR amplification was peformed, with B. japanicum genomic DNA as a template, on
an Automated DNA Thermal Cycler 9600 (Perkin Elmer Cetus, N orwalk, CT) using
the protocol described by Saiki (23). The procedure consists of a denaturing step
(94°C, 2 min) followed by 35 amplification cycles of denaturation (94°C, 1 min),
annealing (55°C, 1 min) and extension (72°C, 2 min). For BJ38 amplification, 5
additional cycles [denaturation (94°C, 1 min), annealing (37°C, 5 min) and extension
(72°C, 2 min)] were included prior to the 35 amplification cycles. Following

amplification, further extension at 72°C (5 min) was allowed before cooling to 4°C.

 

 

T——_ A. I

107
PCR products were then isolated using a QIAGEN DNA isolation kit (QIAGEN Inc.,

Chatsworth, CA) and eluted with 20 pl of 10 mM Tris, 0.1 mM EDTA, pH 8.0.
DNA sequencing was performed by the dideoxychain termination method of
Sanger et al. (24) using the pBluescript 11 SK (+) phagemid (Stratagene, San Diego,

CA). For subcloning of PCR derived DNA fragments into the pBluescript vector, the

 

phagemid was first digested with EcoRV (Boerhinger Mannheim Biochemicals,
Indianapolis, IN). Addition of T’s to the blunt ends of the vector was then carried
out using Taq DNA Polymerase and dTTP (25, 26). PCR fragments were then
annealed to the phagemid utilizing the property that PCR products synthesized by Taq 1
DNA polymerase result in the addition of single non-templated A’s to the 3’ ends of

the duplex DNA strands. Sequencing was carried out using the Taquence kit Version

2.0 (US Biochemical Corp., Cleveland, OH) and 35S-dATP (Dupont, NEN; 10

mCi/ml) according to the manufacturers instructions.

RNA Isolation and Northern Blot Analysis

For the extraction of mRNA, the bacterial cells were harvested by
centrifugation in a GS3 rotor (11,000 g, 15 min). The pelleted cells were
resuspended in 20 mM sodium acetate (pH 5.5), 1 mM EDTA, 10 mM (3—
mercaptoethanol, 10 mM sodium dodecyl sulfate (SDS). The suspension was
immediately placed in a 65°C water bath and an equal volume of hot phenol (65°C)
that had been equilibrated with 20 mM sodium acetate (pH 5.5) was added. After 5

min, the suspension was subjected to centrifugation (30,000 g, 20 min) in an SS-34

 

108

rotor at 4°C to remove DNA and protein debris. The supernatant was extracted three
times with phenol-chloroform ( 1:1), once with chloroform and precipitated by
addition of 2.5 volumes of ethanol and 0.1 volume of 3M sodium acetate (pH 5.5).
The RNA was washed with 80% ethanol, dried and resuspended in water.

Total RNA (20 pg) was subjected to gel electrophoresis in 1% agarose-
formamide denaturing gels in 20 mM MOPS, 5 mM sodium acetate (pH 7.0), 10 mM
EDTA, and transferred to nitrocellulose membranes (Schleicher and Schuell, Keene,
NH) by capillary transfer (27). The filters were blocked in hybridization buffer [50%
formamide, 5X SSPE (1X SSPE is 150 mM NaCl, 10 mM sodium phosphate, 1 mM
EDTA), 0.1% SDS, 100 pg/ml denatured salmon sperm DNA, 5X Denhardt’s
solution] for 1 h. The composition of 5X Denhardt’s solution was 0.1% bovine
serum albumin, 0.1% polyvinlypyrolidone, and 0.1% Ficoll in H20. DNA probes
were labeled with [a-32P]dCTP (Dupont NEN, 10 mCi/ml) using random
oligodeoxynucleotide primer labeling (28), and were used at 10° cpm/ml. The filters
were incubated for 24 h at 42°C. The blots were washed three times with 2X SSPE,
0.1% SDS (15 min each), and once in 0.1X SSPE, 0.1% SDS (1 h). The washed
filters were exposed to Kodak X-Omat AR film with an intensifying screen at -80°C.
Relative intensities of the bands were determined by scanning densitometric analysis

with a Visage 110 densitometer (Bioirnage Prod., Ann Arbor, MI).

 

ge:

(dl

 

 

 

109

Southern Blot Analyses

 

Genomic DNA was isolated by the standard phenol-chloroform method (29).
The DNA (3 pg) was then digested with 20 units of EcoRI and BamHI (Boerhinger
Mannheim Biochemicals, Indianapolis, IN) using reaction conditions suggested by the
manufacturer. The digested DNA was electrophoresed on 0.8% agarose gels, and
transferred to nytran membranes (Schleicher and Schuell, Keene, NH). The filters
were then blocked in hybridization buffer and hybridized with 32P-labeled probes as
described for the Northern blot analyses. The filters were then washed twice in 2X
SSC (1X SSC = 0.15 M NaCl and 0.5 M sodium citrate) and 0.1% SDS at 42°C for
15 min each, and once in 0.1X SSC and 0.1% SDS for 15 min at 42°C before being

exposed to Kodak X-Omat AR film.

n_0d_Dy, glnA, galectin-3 Oligonucleotide Probes

The nadD, probe was obtained by PCR amplification of the B. japanicum

 

genomic DNA using the sequences -(dATCTAAATCTTCTCGTTGCGCTC)- and -
(dCGAGCAATATCCGACGCATCCAGA)- complementary to the 5’ and 3’ ends of
the published nadD, sequence (18) respectively. A PCR product of 870 bp was
obtained whose size was as expected from the sequence. In addition, a 270 bp stretch
of the PCR product was sequenced and this segment demonstrated 100% identitity to
the reported nadD, sequence (18). The cDNA of galectin—3 from mouse 3T3 cells
was as described in Jia et al. (30).

The glutamine synthetase probe (glnA) was kindly provided by Dr. Greg

 

 

l 10
Martin of Purdue University, in the form of the plasmid pB153 (31). The plasmid

was digested with EcoRI and BglII to liberate a 0.8 kb fragment from within the
coding sequence, as expected from restriction map analyses. This 0.8 kb glnA

fragment was chosen as the glutamine synthetase probe in Northern blot analyses.

111
RESULTS

A Molecular Probe for the BJ38 ge_rm(§)

Coomassie blue staining of purified B138, that had been digested with V-8
protease and subjected to SDS-PAGE, identified four peptides: V1 (38 kd), V2 (31
kd), V3 (21 kd) and V4 ( 14 kd). Amino acid sequence analyses of V2 and V4
yielded partial amino acid sequences as indicated in Fig. 1A. N 0 sequence was
obtained from V1 and V3. Two degenerate primers (a sense oligonucleotide primer
derived from V2 and an anti-sense primer derived from V4) were synthesized on the
basis of their partial amino acid sequences and codon usage. Using genomic DNA
isolated from B. japanicum as a template for PCR amplification, 3 DNA fragment
(designated PCRl) of about 0.45 kb was obtained. The PCR fragment was sequenced
and the deduced amino acid sequence of the PCR fragment matched exactly the
experimentally determined amino acid sequence, both at the ends corresponding to the
PCR primers, as well as internal regions not specified by the primers (Fig. 1B).

PCRl was thus used as an authentic probe for the B1 38 gene.

Southern and Northern Blot Analysis using the BJ38yobe

 

In these experiments, chromosomal DNA was first digested with the
restriction endonucleases EcoRI and BamHI. These enzymes were chosen because
they do not possess cleavage sites within the PCRl probe. Southern blot analysis of

the digested genomic DNA samples probed with PCRl revealed, in each case, two

 

 

 

112

Figure l. A) The partial amino acid sequences of peptides derived

from BJ38. BJ38 was digested with V8 protease, subjected to SDS-PAGE,
transferred to a Problot membrane and Coomassie blue stained. Peptides were
subjected to amino acid sequence analysis and the sequences for V2 and V4 are
shown at the right. The arrows highlight the amino acid sequences used to synthesize
the 5’- and 3’- primers (direction indicated by the arrow) used for PCR amplification.

B) The deduced amino acid sequences of PCRl fragment derived from the
nucleotide sequence. The underlined amino acid residues represent direct matches
between the amino acid sequence deduced from the nucleotide sequence of the PCR
product and the experimentally determined amino acid sequence from the protein.

113

A. Experimentally determined peptide sequence:

VI- .3 )-
VZ— --- -Thr Asa — Als — Asp Gly — Thr — Asp Asa Lea Ala llc — Ala Gin -- Asn llc

 

V3— «—

#—
V4— 1"“ -Vsl Val Phc Lea Vsl Thr Asp Gly Vsl Gly Asp Lg llc Val Scr Gly Als Ser

B. Dcduced amino acid sequences from DNA sequence analyses:
Asp Asa Lea Ala lle Ar; Ala glp Ar: Q [Is Thr Les Ar: lle Asp

..........(l40 amino acids In so open reading frame of the nucleotide sequence).____...._..

TIIr Pro Gls Gin Val Vsl fie Leu Vsl Thr Asp Gly Vsl Qty Qp m lie

 

 

 

 

 

114

Figure 2. Genomic Southern blot of B. japanicum DNA that had been
digested with endonucleases that do not contain cleavage sites Within

the PCRl probe. 3 pg of B. japanicum DNA was digested with 20 units of EcoRL
and BamHI and the digested DNA subjected to 0.8% agarose gel electrophoresis.
DNA was transferred to N ytran membranes and the filters probed with 32P-labeled
PCRl. Lanes a and c , EcoRI digestion; lanes b and d, BamHI digestion. Lanes a
and b, ethidium bromide staining; lanes c and d, hybridization with 32P-labeled PCRl.

 

115

 

 

 

 

 

 

 

 

 

 

116

 

bands (Fig. 2). These results suggest the possibility that there are either two genes
for BJ38, or the presence of another gene, besides the BJ38 gene, that can hybridize
with the PCRl probe.

When Northern blots of B. japanicum RNA were probed with PCRl, the B138
probe hybridized to 2 bands of 1.5 kb and 2.8 kb (Fig. 3A, lane a). As a control,
PCRl was also used to probe RNA isolated from R. melilati and R. trifolii, two
rhizobial strains that do not bind to soybean roots. In both of these strains, no
hybridization of PCRl was observed (Fig. 3A, lanes b, c). When the nadD, probe
was used to hybridize B. japanicum RNA, a single mRNA species of 1.6 kb was
observed (Fig. 3B, lane b). The glutamine synthetase probe revealed a single band of
~ 1.5 kb as expected for the glnA transcript (Fig. 3B, lane c) (31, 32). As a negative
control, the cDNA of galectin—3 from mouse 3T3 cells was used to probe the RNA
blots (30). As expected, this probe showed no hybridization to B. japanicum RNA
(Fig 3B, lane (1).

The fact that the PCRl probe hybridizes to two bands in both Southern and
Northern blots suggests that there may indeed be two genes, each with its own distinct
transcript. As of this writing, the sequence of the B138 gene (either one of the
possible two) has not been completely determined. Nor is it clear whether the two
mRNA species observed on the Northern blots are derived from one or distinct genes.
In light of this problem, we have decided to describe the effects of saccharide and
flavones on BJ38 expression in terms of both the two mRN A transcripts, hereafter,

the 1.5 kb and 2.8 kb species will be designated as transcript I and transcript II, respectively.

 

 

 

117

Figure 3. A) Northern blots of RNA that had been hybridized with
32P-labeled PCRl. RNA was isolated as described in the methods and materials,
subjected to 1% agarose gel electrophoresis and transfered to nitrocellulose
membranes. The blots were then hybridized with the 32P-labeled PCRl, washed, and
analyzed by audioradiography. Lanes: a) B. japanicum; b) R. melilati; and c) R.
mfalii RNA.

B) Northern blots of B. japanicum RNA that had been hybridized

with: a) PCR]; b) nodD,; c) glnA; d) galectin-3 probes.

118

15» . .. rs.

abs abcd

119

Em of fipvonoids on B138 and n_a__t_l_l_)l expression

Our previous observation that B1 38 expression, assayed at the polypeptide
level, was induced by the flavonoid genistein, a known nod gene inducer in B.
japanicum, raised the possibility that BJ38 may itself be a nod gene. Thus the effects
of flavonoids on the accumulation of mRN A for BJ38 and nadD, were tested. In
these experiments, RNA was isolated from liquid cultures of B. japanicum that had
been exposed to 2 pM concentrations of flavones for 10 h. This concentration of
flavonoids was the same as what we had reported for genistein induction of B1 38 (9).
Similarly, a 10 h exposure to flavonoids had previously been shown to be sufficient
for nod gene induction (12). Finally, the transcript for the constitutively expressed
glutamine synthetase (glnA) gene (33, 34) was used as an internal control in Northern
blot analyses, as we had previously shown that the level of glutamine synthetase
enzymatic activity was unaffected by treatment with flavonoids (9). Consistent with
this previous report, the glnA transcript levels were not altered by the addition of
flavonoids.

When RNA from B. japanicum was probed with PCRl, the B1 38 probe
hybridized to two bands of 1.5 (transcript I) and 2.8 kb (transcript II) in both the
untreated and flavonoid treated samples. The order of flavonoid induction for both
transcript I and II was similar to that reported at the protein level, with genistein
yielding the greatest effect (Fig. 4A). Luteolin yielded little or no effect, while
apigenin yielded intermediate effects on BJ38 expression. In parallel, an examination

of nadD, expression revealed that the nadD, gene was similarly induced by

120

Figure 4. The effect of flavonoids on the mRNA levels of B138 and
nadD, . Final liquid cultures were incubated for 10 h in the presence of 2 pM
concentrations of genistein, apigenin, luteolin or without any flavonord treatment.

The amount of each transcript is represented in intensity units. A) B138 transcripts;
B) nadD, transcript.

 

 

121

 

 

 

21M

3311116111.

 

 

 

 

 

transcript

 

 

122

Figure 5. The effect of saccharides on the mRN A levels of BJ38

and nadD, . Final liquid cultures of B. japanicum were incubated for 10 h in the
presence of 1 mM concentrations of Lao, Gal, Man and without any saccharide. A)
B1 38 transcripts; B) nadD, transcript.

123

 

 

 

 

 

 

 

 

 

 

 

 

flavonoi
lnteolir

consiste

Effect 1

culture
was e11
(9). l
accum
a cone
exposr
increa
previc

japan

 

124

flavonoids. Genistein yielded the greatest effect, followed by apigenin (Fig. 4B).
Luteolin failed to have an effect on nadD, mRN A levels. These latter results are

consistent with previously reported effects on nadD, expression ( 12, 15).

Effect of saccharides on the expression of Bdfi am] nad_D,

In our previous study, we have also shown that when B. japanicum cells were
cultured in the presence of saccharides, the amount of BJ 38 isolated from these cells
was elevated in an order similar to the relative affinities of the saccharides for BI 38
(9). Using the same protocol for flavone induction, the effects of saccharides on the
accumulation of mRN A for BI 38 and nadD, were tested. Saccharides were added to
a concentration of 1 mM and the cells cultured for 10 h. This 10 h length of
exposure to Lac was selected as it had previously been shown to result in a 4-5- fold
increase in the amount of BJ38 isolated from B. japanicum (9). Consistent with our
previous report, the glnA mRNA levels were unaffected by treatment of B.
japanicum with saccharide.

Northern blot analyses of saccharide treated samples also revealed mRN A
transcripts of 1.5 kb and 2.8 kb similar to that obtained with the flavone induction
(Fig. 5). The intensities of both of these bands were strongest in the Lac-treated
cells, followed by Gal and mannose (Man). This relative order of induction of both
transcripts was similar to that reported for saccharide induction of B1 38 at the protein
level (9).

When the nadD, probe was used to probe the RNA samples isolated from both

 

 

 

 

untreated at
was observr
asaccharid
observed 11
which had

nadDl trar

125
untreated and saccharide treated B. japanicum cells, a single nadD, species of ~ 1.6 kb

was observed. Interestingly nadD, transcript levels were also found to be elevated in
a saccharide dependent fashion. The order of saccharide induction was similar to that
observed with BJ38 expression, with lactose yielding the greatest effect. Mannose,
which had no little or no effect on BJ38 expression, had only a minimal effect on

nadDl transcription.

 

11
level cor
and lac
flavonoi
level. '
were e1
that re;
demon
(C) no

order .

probe
genon
nan
dedur
sequt
regic
In g1

restr

126
DISCUSSION

In previous studies (9), we have shown that BJ38 expression at the polypeptide
level could be induced by treatment of B. japonicum cells with either the genistein
and Lac. We now document the results of a study to characterize the effect of both
flavonoid compounds and saccharides on BJ38 and nadD, expression at the molecular
level. The key findings of this study are (a) The levels of BJ38 mRNA transcripts
were elevated by both saccharides and flavonoids in an order of efficacy similar to
that reported at the polypeptide level (9); (b) nadD, induction by flavonoids
demonstrated a strong correlation to that observed in the flavonoid regulation of BJ38;
(c) nadD, mRNA levels were also elevated in a saccharide dependent fashion. This
order of induction was: Lac > Gal > Man.

A key factor in the interpretation of these results is the validity of the BJ38
probe PCRl. This probe was obtained by PCR amplification of B. japonicum
genomic DNA using degenerate primers derived from the partial amino acid
sequences of BJ38 that had been digested with V—8 protease. PCRl yielded a
deduced amino acid sequence that matched the experimentally derived B] 38 amino
sequence at both ends corresponding to the PCR primers, as well as in internal
regions bounded by the primers. PCRl was thus used as an authentic probe for BJ38.
In genomic Southern blot analyses of B. japanicum DNA that had been digested with

restriction enzymes that do not possess cleavage sites within the PCRl probe, two

 

 

 

 

 

bands were
2.8 kb (tra
At the pre
the presen
it is also
or two di
both tran
BJ38 car
that B13
monitor
marker
for the
nadD,
BJ38.
culturr

activa

indie.
Com
with
with

soyl

 

127

bands were revealed. Consistent with this, two transcripts of 1.5 kb (transcript I) and
2.8 kb (transcript II) were detected in Northern blot analyses of B. japonicum RNA.
At the present time, we do not know whether these results reflect two BJ38 genes or
the presence of another gene, besides BJ38, that can hybridize to PCR]. In addition,
it is also not known whether both transcript I and transcript II are derived from one
or two different genes. The present results nevertheless demonstrate inducibility for
both transcripts and provoke several intriguing ideas. Firstly, the observation that
B138 can be induced by flavonoid compounds, such as genistein, raises the possibility
that BJ38 is a member of the nod gene family. This possibility was tested by
monitoring, in parallel, the levels of the nodDl transcript. nadD, was chosen as a
marker for nod gene expression as it is inducible by flavonoids (15), as well as codes
for the transcriptional regulator of other nod genes (18). Flavonoid induction of
nodD, demonstrated a strong correlation to that observed for flavonoid regulation of
BJ38. The possibility exists, therefore, that the addition of genistein to B. japanicum
cultures results in an increased expression of the nadD, gene product, which, in turn,
activates the transcription of the BJ38 gene.

Second, the observation that ’10le mRNA levels were induced by saccharides
indicate alternative inducers of this gene, besides the flavonoid compounds.
Consistent with this, Banfalvi et al. (15) have reported that the induction of nod genes
with soybean exudate result in induction levels of nod genes greater than that obtained
with the flavonoid inducers alone. As saccharides are a major component of the

soybean root exudate, the possibility exists that saccharides may provide an additional

 

 

 

 

 

 

 

componen
soybean rt
derivative
the nadD,
flavonoid
on the re
connectir
japanicu
tumefaci
by both
transduc
binding
transdul
bound 1
transcri
of nod
likelih.
leading
such a
and D
Protei

comn

 

128

component leading to the maximal induction of nodD, expression observed with
soybean root exudate. Recently, Smit et al. (35) have also reported that glycoside
derivatives of flavonoid compounds are actively involved in the specific induction of
the nadD, gene. The possibility of a dual control of the nadD, gene by both
flavonoids and saccharides in B. japanicum is especially intriguing in light of studies
on the regulation of the vir genes in Agrobacterium tumfaciens (36, 37). In this
connection, there is a striking analogy between the control of the nod genes of B.
japonicum and the vir genes in A. tumefaciens (Table I). For instance, the A.
tumefaciens vir genes necessary for the infection of the host plant are also activated
by both saccharides and phenolics. This activation involves a two—component sensory
transduction pathway composed of a sensor-kinase protein (VirA) and a DNA
binding protein (VirG). Saccharide induction of the vir genes through this
transduction system is mediated by the carbohydrate binding protein ChvE. When
bound by its ligands, ChvE activates the sensory transduction pathway resulting in the
transcription of the vir genes. Our present observation that the saccharide specificity
of nadD, induction is similar to the saccharide binding specificity of BJ38 suggests the
likelihood that BJ38 may serve as an analogous saccharide receptor in B. japonicum
leading to the activation of the nadD, gene. If indeed nadD, regulation did occur via
such a sensory transduction pathway, a possible candidate for both the sensor kinase
and DNA binding protein components in B. japanicum would be the NodV and NodW
proteins respectively. These proteins are essential for the full expression of the

common nadD, and nodYABCSUIJ operons and have been proposed to form an

 

 

 

Table I.
attend!

 

If

#

lnducer

 

 

 

 

129

Table I. Comparison of A. tumetaciens vir gene induction and B. iagonicum nod
gene induction.

 

 

vir gene induction nod gene induction
Inducer a. saccharides: a. saccharides:
galacturonic acid > Lac > Gal > Man

glucuronic acid >
xylose > glucose

 

 

 

b. phenolics: b. flavonoids:
eg. acetosyringone eg. genistein
carbohydrate—binding ChvE BJ38 ?
receptor
sensory protein Vir A NodV
DNA binding protein VirG NodW
NodD

 

 

 

 

 

 

 

 

altemativ
analyses
11

BJ38 are
correlati
nadD, 6
family i
for indt
also cu:
the B13
inducil
This s
blots a

derive

 

130

alternative pathway for nod gene induction by flavonoids (38). In addition, sequence
analyses of NodV have also shown it to be homologous to the VirA protein (39).

In conclusion, our present results have demonstrated that mRNA levels of
BJ38 are increased by both genistein and Lac. In this connection, there is a direct
correlation between the efficacy of induction by different flavonoids on BJ38 and
nadD, expression. This result suggests the possibility that BJ38 may belong to the
family of nod genes required for nodulation. As the nadD, gene product is essential
for induction of nod genes, the effects of these compounds on BJ38 expression are
also currently being tested in nodD, mutants. Studies are also in progress to sequence
the BJ38 gene, and to determine if the conserved nod box sequence preceeding the
inducible nod genes in Rhizobium are found in the promoter sequences of these genes.
This sequence analysis will also determine whether both bands detected in Southern
blots are indeed BJ38 genes, and whether both transcripts I and II are transcripts

derived from one or two distinct genes.

 

 

 

1 3 1
REFERENCES

Vesper, S. J., and W. D. Bauer. 1985. Characterization of Rhizobium
attachment to soybean roots. Symbiosis 1:139-162.

Ho, S.-C., J. L. Wang, and M. Schindler. 1990. Carbohydrate binding
activities of Bradyrhizobium japonicum. I. Saccharide—specific inhibition of
homotypic and heterotypic adhesion. J. Cell Biol. 111:1631—1638.

Smith, G. B., and A. G. Wollum II. 1992. Physicochemical and D-galactose-
mediated interactions in attachment of Bradyrhizobium japanicum to roots of
Glycine max. Can. J. Microbiol. 39:245-251.

Ho, S.-C., M. Schindler, and J. L. Wang. 1990. Carbohydrate binding
activities of Brodyrhizobium japanicum. H. Isolation and characterization of a
galactose—specific lectin. J. Cell Biol. 111:1639-1643.

Loh, J. T., S.-C. Ho, A. W. de Feijter, J. L. Wang, and M. Schindler.
1993. Carbohydrate binding activities of Bradyrhizobium japonicum:
Unipolar localization of the lectin BJ38 on the bacterial cell surface. Proc.
Natl. Acad. Sci. USA 90:3033-3037.

Ho, S. -C., J. L. Wang, M. Schindler, and J. T. Loh. 1994. Carbohydrate
binding activities of Bradyrhizobium japonicum. III . Lectin expression,
bacterial binding, and nodulation efficiency. Plant J. 5:873-884.
Bhuvaneswari, T. V., B. G. Turgeon, and W. D. Bauer. 1980. Early events
in the infection of soybean (Glycine max. L. Merr) by Mizobium japanicum.

1. Location of infectible root cells. Plant Physiol. 66:1027—1031.

Calvert, H. E., M. K. Pence, M. Pierce, N. S. A. Malik, and W. D. Bauer.
1984. Anatomical analysis of the development and distribution of Rhizobium
infections in soybean roots. Can. J. Bot. 62:2375—2384.

 

 

 

 

10.

ll.

12.

l3.

14.

15.

16.

10.

ll.

12.

13.

14.

15.

16.

132

Loh, J. T., S.-C. Ho, J. L. Wang, and M. Schindler. 1994. Carbohydrate
binding activities of Bradyrhizobium japonicum. IV. Effect of lactose and
flavones on the expression of the lectin, BJ38. Glycoconjugate J., in press.
Peters, N. K., J. W. Frost, and S. R. Long. 1986. A plant flavone luteolin
induces expression of Rhizobium meliloti nod genes. Science 233:917-1008.
Zaat, S. A. J., C. A. Wijffelrnan, I. H. M. Mulders, A. A. N. van Brussel,
and B. J. J. Lugtenberg. 1988. Root exudates of various host plants of R.
leguminosarum contain different sets of inducers of Rhizobium nodulation
genes. Plant Physiol. 86:1298-1303.

Kosslak, R. M., R. Bookland, J. Barkei, H. E. Paaren, and E. R. Appelbaum.
1987. Induction of Bradyrhizobium japonicum common nod genes by
isoflavones isolated from Glycine max. Proc. Natl. Acad. Sci. USA 84:7428-
7432.

Mulligan, J. T., and S. R. Long. 1985. Induction of Rhizobium melilati
nodC expression by plant exudate requires nadD. Proc. Natl. Acad. Sci. USA
82:6609—6613.

Rossen, L., C. A. Shearman, A. W. B. Johnston, and J. A. Downie. 1985.
The nodD gene of Rhizobium leguminosarum is autoregulatory and in the
presence of plant exudate induces the nodA, B, C genes. EMBO J. 4:3369-
3373.

Banfalvi, Z., A. Nieuwkoop, M. Schell, L. Besl, and G. Stacey. 1988.
Regulation of nod gene expression in Bradyrhizobium japonicum. Mol. Gen.
Genet. 214:420—424.

Fisher, R. F., T. T. Egelhoff, J. T. Mulligan, and S. R. Long. 1988.
Specific binding proteins from Rhizobium melilati cell free extracts containing
N odD to DNA sequences upstream of inducible nodulation genes. Genes Dev.
2:282-293.

 

 

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

133
Hong, G. G., J. E. Burn, and A. W. B. Johnston. 1987. Evidence that DNA

involved in the expression of nodulation (nod) genes in Rhizobium binds to the
product of the regulatory gene nodD. Nucleic Acids Res. 15:9677-9690.
Gottfert, M., D. Holzhauser, D. Bani, and H. Hennecke. 1992. Structural
and functional analysis of two different nadD genes in Bradyrhizobium
japanicum USDAl 10. Mal. Plant-Microbe Interact. 5:257-265.

Dazzo, |F.B. 1982. In Experimental Microbial Ecology. R. Burns, and T.
Slater, editors. Blackwell Scientific Publications, Oxford, U. K. pp. 431-
446.

Cleveland, D. W., S. G. Fisher, M. W. Kirschner and U. K. Laemmli.

1977. Peptide mapping by limited proteolysis in sodium dodecyl sulfate and
analysis by gel electrophoresis. J. Biol. Chem. 252: 1102-1106.20.
Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature (Land. ). 227:680-685.

Edman, P., and G. Begg. 1967. A protein sequenator. Eur. J. Biochem.
1:80-91

Saiki, R. K. 1990. Amplification of genomic DNA. In PCR protocols: A
guide to methods and applications. M. A. Innis, D. H. Gelfand, and J. J.
Sninsky, editors. Academic Press, Inc., San Diego, CA. pp.l3-20.

Sanger, F., S. Nicklen and A. R. Coulson. 1977. DNA sequencing with
chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467.
Kovalic, D., J. H. Kwak, and B. Weisblum. 1991. General method for
direct cloning of DNA fragments generated by PCR. Nucl. Acids. Res.
19:4560.

Marchuk, D. M. Drumm, A. Saulino, and F. S. Collins. 1991. Construction
of T-vectors, a rapid and general system for direct cloning of unmodified PCR
products. Nucl. Acids Res. 19:1154.

 

 

27.

28.

29.

30.

31.

32.

33.

34.

35

 

27.

28.

29.

30.

31.

32.

33.

34.

35.

134

Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. In Molecular cloning: a
laboratory manual. N. Ford, M. Ferguson, and C. Nolan, editors. Cold
Spring Harbor press, Cold Spring Harbor, NY.

Feinberg, A. P. and B. Vogelstein. 1983. A technique for radiolabeling
DNA restriction endonuclease fragments to high affinity. Anal. Biochem.
132:6

Zyskind, J. W. and S. Bernstein. 1992. Isolation and purification of bacterial
and Drosophila chromosomal DNA. In Recombinant DNA Laboratory
Manual. Academic Press, Inc., San Diego, CA. pp.12-13.

Jia, S., R. P. Mee, G. Morford, N. Agrwal, P. Voss, I. C. Moutstatsos, and
J. L. Wang. 1987. Carbohydrate binding protein 35:molecular cloning and
expression of a recombinant polypeptide with lectin activity in Escherichia
coli. Gene 60:197-204.

Carlson, T. A., M. L. Guerinot and B. K. Chelm. 1985. Characterization of
the gene encoding glutamine synthetase I (glnA) from Bradyrhizobium
japanicum. J. Bacteriol. 162:698—703.31.

Turner, N. E., S. J. Robinson, and R. Harelkom. 1983. Different promoters
for the Anabaena glutamine synthetase gene during growth using molecular or
fixed nitrogen. Nature (Land) 306:337-341.

Adams, T. H. , and B. K. Chelm. 1988. Effects of oxygen levels on the
transcription of nif and gln genes in Bradyrhizobium japanicum. J. Gen.
Microbiol. 134:611-618.

Carlson, T. A., G. B. Martin, and B. K. Chelm. 1987 . Differential
transcription of the two glutamine synthetase genes of Bradyrhizobium
japanicum. J. Bacteriol. 169:5961-5866.

Smit, G., V. Puvanesarajah, R. W. Carlson, W. M Barbour and G. Stacey.
1992. Bradyrhizobium japanicum nadD, can be specifically induced by
soybean flavonoids that do not induce the nodYABCSUIJ operon. J. Biol.
Chem. 267 2310-318.

 

 

 

 

 

36.

37.

38.

39.

 

 

 

 

36.

37.

38.

39.

135
Cangelosi, G. A., R. G. Ankenbauer, and E. W. Nester. 1990. Sugars

induce the Agrabacterium virulence genes through a periplasmic binding
protein and a transmembrane signal protein. Proc. Natl. Acad. Sci. USA
87:6708-6712.

Ankenbauer, R. G. , and E. W. Nester. 1990. Sugar-mediated induction of
Agrobacterium tumefaciens virulence genes: structural specificity and
activities of monosaccharides. J. Bacteriol. 172:6442-6446.

Sanjuan, J ., P. Grob, M. Gottfert, H. Hennecke and G. Stacey. 1994.
NodW is essential for full expression of the common nodulation genes in
Bradyrhizobium japanicum. Mol. Plant-Microbe Interact. 72364-369.
Gottfert, M. , P. Grob, and H. Hennecke. 1990. Proposed regulatory
pathway encoded by the nodV and nodW genes, determinants of host-
specificity in Bradyrhizobium japanicum. Proc. Natl. Acad. Sci. USA
87:2680-2684.

 

 

the

Bra

(G:

de:

Sp:

 

 

 

 

136

CHAPTER V

CONCLUDING STATEMENT

The attachment of rhizobium to the roots of the host legume is a key step in
the infection process leading to the establishment of a nitrogen fixing symbiosis.
When I began my thesis research, it had already been documented that, in the
Bradyrhizobium japanicum-soybean interaction, this attachment is galactose
(Gal)/ lactose (Lac)-inhibitable. In addition, a carbohydrate binding protein,
designated, BI 38, had been isolated from B. japanicum. BJ38 exhibited a pattern of
specificity in saccharide binding similar, if not identical, to that of B. japanicum
adhesion to soybean cells.

In this dissertation, I have demonstrated, using a polyclonal antiserum
generated against BJ38, that the lectin is localized at one pole of the bacterium cell
surface. More importantly, BJ38 localization is coincident with the site of B.
japanicum attachment to soybean cells. Thus, the cell surface display of B] 38 is
consistent with its role in mediating the polar adhesion of the bacterium to the
soybean roots.

In the course of studies to obtain large amounts of BJ 38 for antibody
generation, the level of BJ 38 expression, assayed at the polypeptide level, was found

to be elevated when the bacterial cells were first cultured in the presence of

 

 

 

 

137

saccharides. This induction effect followed the order Lac > Gal > mannose ~ no
hapten, which is similar to the relative affinities of these saccharides for BJ38. The
effect of Lac and Gal on the expression of BJ38 was also observed at the mRNA
level, as determined by Northern blot analysis.

Surprisingly, when the effects of flavonoids on BJ38 expression was tested,
BJ38 expression was also found to be elevated, at both the transcriptional and
polypeptide levels, by genistein, a known nod gene inducer in B. japanicum. This
observation raised the possibility that BJ38 may be a member of the nod gene family
required for successful nodulation. In this connection, we have also observed that the
transcript levels of nodD,, the transcriptional regulator of nod genes, were elevated by
flavonoid compounds in a fashion similar to that of BJ38 induction. The possibility
exists, therefore, that genistein activation of the nadD, gene results in increased levels
of NodD, which then activates the transcription of the BJ38 gene. Further
investigation is needed to determine if the nod box sequences preceeding the nod
genes are found in the regulatory/promoter sequences of BJ38.

Finally, we have also shown that the levels of nadD, transcription could also
be elevated by saccharide treatment of B. japanicum. This observation suggests that
alternative inducers, besides flavonoids, may control the expression of the nod genes

in B. japanicum.

RIES

n

l

l

y

|
02

lllllllllll

. 4...... .2

ll

31|fli|5|2i301

V
I
N
U
E
T
g
T
s
N
n
G
I
H
C
I
n

. .11.... ...............1 . . . . . :3;

:1...

 

 

._1...... 1....
.....11.... . 2.....1:
. . . Z...