MSU

 

 

RETURNING MATERIALS:

 

Place in book drop to
remove this checkout from

 

 

LIBRARIES .
W your Y‘QCOY‘d. FINES W1“
be charged if book is
returned after the date
stamped below.
«11“ 41"} “'. ”7'" r77 ' “ fit. 1 ', n,‘ 7-:
£ , j," If; K 9 ,_.5 «5:86.. bud lb 63 as
“ft/“MW 53"" {‘a*‘
E‘x‘ iii .4 7:13;: E 3i f 1...," “r L V
r { I! 'I. If n '\ 'TI ”‘1 f9; ““5"”
;r: I i‘. A 75"

 

 

 

STUDIES ON THE PLASMID-CODING OF NODULATION
AND NITROGEN FIXATION GENES IN TWO
STRAINS OF RHIZOBIUM TRIFOLII

 

By

Alan H. Christensen

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1983

ABSTRACT

STUDIES ON THE PLASMID-CODING OF NODULATION
AND NITROGEN FIXATION GENES IN TWO
STRAINS OF RHIZOBIUM TRIFOLII

 

By

Alan H. Christensen

Large plasmids coding for genes essential for the Rhizobium
trifolii-clover symbiosis have been identified. Transfer of pJBSJI, a
3;_leguminosarum plasmid coding for pea nodulation (52g) and the nitro-
genase (git) structural genes, into 3; trifolii strain T37 generated
transconjugants containing a variety of plasmid profiles formed from
the recombination of pJBSJI and pRtT37a. The symbiotic properties
exhibited on both hosts and the plasmid profiles were stably maintained
even after reisolation from root nodules. Approximately 30% of the
transconjugants, which contained a "hybrid" plasmid corresponding in
size to pJBSJI, formed effective (nitrogen-fixing) nodules on peas but
were unable to nodulate clover. This suggests that the 3;_trifolii nod
genes had been deleted. Hybridization analysis indicated that the g;
trifolii Elf genes, which are located on pRtT37a, had also been
deleted. Another 40% of the transconjugants harbored "hybrid" plasmids
of various molecular weights. These strains contained the 3;_trifolii
.gij_genes and formed effective nodules on clover. However, these

strains lacked the §;_leguminosarum nif genes and formed ineffective

Alan H. Christensen

nodules on peas. This suggests that the 3; trifolii nif genes were

 

unable to complement the deleted 3; leguminosarum nif genes in pea

 

nodules. The remaining 30% of the transconjugants contained both sets
of symbiotic genes and formed effective nodules on both peas and
clover. Results from the analysis of the strain T37 (pJBSJI)
transconjugants indicated that genes essential for clover nodulation
and nitrogen fixation are encoded on pRtT37a. The linkage of Eli and
£2g_genes on a plasmid in §;_trifolii strain 0403 was demonstrated.
Analysis of plasmid DNA from a spontaneous non-nodulating mutant of
strain 0403 indicated that about 70 Mdal of DNA had been deleted from
pRt0403a, the smallest plasmid in this strain. This non-nodulating
mutant was able to attach to clover root hairs, but could not induce
the formation of infection threads, suggesting that genes essential for
infection thread formation had been lost. Southern hybridization
analysis indicated that the E; trifolii nif genes which are encoded on
pRt0403a, had been deleted. Thus, the 31: genes and genes essential
for nodulation are linked on a 70 Mdal region of pRt0403a DNA.

To Mom and Dad

ii

ACKNOWLEDGEMENTS

I wish to thank my thesis advisor, Dr. Karel R. Schubert, for his
help and guidance throughout my graduate career. Thanks are also due
to my committee members, Drs. Paul Kindel, Hsing-Jien Kung, Harold
Sadoff, and Clarence Suelter. I also wish to thank Dr. Arnold Revzin
for his generous gifts of §29_RI and the members of Dr. Kung's
laboratory for their helpful suggestions. Financial support from the
National Institute of Health, the Monsanto Agricultural Products
Company, the Jessie Smith Noyes Foundation, and the Biochemistry
Department at Michigan State University is gratefully acknowledged.
Finally, special thanks are extended to my very dear friend, Deborah
Ann Polayes, whose understanding and support during my graduate career

is gratefully appreciated.

iii

List of Tables . . . . .
List of Figures. . . . .
List of Abbreviations. .
Introduction . . . . . .

Literature Review. . . .
Biological Nitrogen

Chapter I.

TABLE

OF CONTENTS

Fixation.
Lectin Recognition Hypothesis . .
The Role of Bacterial Polysaccharides
Plant Cell Responses. 1. Root Hair Curling.
Plant Cell Responses. II. Infection Thread F rm
Bacteroid Formation and Nodule Development. . .
Organization and Regulation of the nif Genes. .
Plasmid DNA in Rhizobium. . . . . . . . . . . . .
Genetic Functions Encoded on Plasmids in Rhizobium.

O

at1

oooéooooo

Preliminary Studies on the Presence of Plasmids and

Plasmid-Coding of Nodulation and Nitrogen Fixation Genes in

Rhizobium

Introduction. . . . . .

Materials and Methods .
Materials. . . . .
Bacterial Strains.

Media. 0 O O I O O

Currier-Nester Plasmid

CsCl'EtBr Density Equilibrium Centrifugation

Isolation

Procedure
Agkaline Lysis Procedure for Plasmid Isolation
[ HJThymidine Labeling of Bacterial DNA. . . .
Sucrose Gradient Centrifugation. . . . . . .
Agarose Gel ElectrOphoresis. . . . . . . . .

Electron Microscopy of Plasmid DNA . . . . .
Acetylene Reduction Assay on Rhizobium in Culture..
Growth of R. “cowpea” Strain SZHI at Elevated Temperature 40
Acridine Orange Treatment. . . . . . . . . . . . . . . .
Nodulation Test. . . . . . . . . . . . . . . . . . . . .
Acetylene Reduction Assay on Nodulated Plants. . . . . .

iv

Page

vii

.viii

xi

H

NHHHH
#O‘wNO-‘CONU'IU'I

4O
41
42

Results and Discussion. . . . . . . . . . . . . . . . . .
Isolation of Plasmid DNA . . . . . . . . . . . . . .
Sucrose Gradient Centrifugation. . . . . . . . . . .
Agrose Gel Electrophoresis of Plasmid DNA. .....
CsCl'EtBr Density Equilibrium Centrifugation . .
Electron Microscopy of Plasmid DNA . . . . . . . .
Plasmid Curing Experiments . . . . . . . . . . .

Chapter II. The Identification of a Rhizobium trifolii Strain
Plasmid Coding for Nitrogen Fixation and Nodulation Genes
Its Interaction with pJBSJI, a Rhizobium leguminosarum
Plasmid

IntrOducti on. O C O O O O O O O O O O O O O O O O O O O 0

Materials and Methods . . . . . . . . . . . . . . . . .
Materials. . . . . . . . . . . . . . . . . . . . .
Bacterial Strains, Plasmids, and Phage ......
Media. . . . . . . . . . . . . . . . . . . . .
Plasmid Transfer . . . . . . . . . . . . . . .
Ekhardt Agarose Gel Electrophoresis Technique.
Nodulation Tests . . . . . . . . . . . . . .
Acetylene Reduction Assay. . . . . . . . .
Isolation of Bacteria from Root Nodules. .
Fahraeus Slide Technique . . . . . . . . .
Isolation of Total DNA . . . . . . . . . .
Determination of DNA Concentration . . . .

Agarose Gel Electrophoresis. . . . . . . .
Growth of Bacteriophage AzTn g . . . . . .
Isolation of DNA from Phage Particles. . .
Cleared Lysate Procedure for Plasmid Isolatio
Isolation of DNA Fragments from Agarose Gels
Preparation of nif Probe DNA . . . . . . . .
§§olation of pRtT37a and pJBSJI. . . . . .

P-labeling of DNA by Nick Translation. .
Southern Hybridization . . . . . . . . . .

o o o o o 3 o o o o o o o o 0

Results . . . . . . . . . . . . . . . . . . . . . . . . .
Transfer Frequency of pJBSJI . . . . . . . . . . . .
Characterization of R;_trifolii Strain 0403 (pJBSJI)

Transconjugants. . . . . . . . . . . . . . . . .
Plasmid Analysis of R;_trifolii Strain T37 (pJBSJI)

Transconjugants. O 0 I I O O I O O O O I O O O O O O

Symbiotic Properties of R;_trifolii Strain T37 (pJBSJI)

Transconj ugants. O O O O O O O O O O O O 0 I O O
Fahraeus Slide Analysis of Infection Process .
Tn 5 Hybridization Analysis. . . . . . . . . .

nif‘Hybridization Analysis . . . . . . . . .
BDBSJI and pRtT37a Hybridization Studies . .

DiSCUSSion. O O 0 O O O O O O 0 O O O O O O O O O O O O O

118
126

Page
Chapter III Characterization of a Spontaneous Non-nodulating
Mutant of Rhizobium trifolii Strain 0403: Linkage of
Nodulation and Nitrogen Fixation Genes

Int rOdUCtiono O O O O O O O O O O O O O O O O O O O O O O O O 134

Materials and Methods . . . . . . . . ..... . ...... 135
Materials. . . . . . . . . . . . . . . . . . . . . . . . 135
Bacterial Strains and Plasmids . . . . . ....... . 135
Media. . . . . . . . . . . . . ...... . . . . . . . 135

mthOds. O O O O O O I O O O O O O O O O O O O O O O O O 136

Results . . . . . . . . . . ................ . 137
Plasmid Analysis . . . . . . . . . . . . . . . . . . . . 137
Clover Nodulation Test . . . . . . . . . ....... . 139
Fahraeus Slide Analysis of Infection Process ...... 139

.gif Hybridization Analysis . . . . . . . . . . . . . . . 144
DiSCUSSion. O O O O O O O O O O O O O O O O O O O O O O O O O 152
Summary. . . . . . . .............. . . . . . . . . . 158

References 0 O O O O O O O O O O O O O O O O O O O O O O O I O O I 16]-

vi

4.

5.
6.
7.
8.

9.

10.

ll.

LIST OF TABLES

Page

Classification of Rhizobium and their hosts . . . . . . . . . 6

.5; pneumoniae nif gene products . . . . . . . . . . . . . . . 18

In vitro nitrogen fixation by R. u"cowpea strains grown in
the presence of acridine orange . . . . . . . . . . . . . . . 64

Nodulation of white clover by R; trifolii strain T37 grown
in the presence of acridine orange. . . . . . . . . . . . . . 67

Bacterial strains . . . . . . . . . . . . . . . . . . . . . . 73
P] asmids. O O O O O O O O O O O O O O O O O 0 O O O O O O O O 74
Transfer of pJBSJI from R; leguminosarum strain T83K3 . . . . 91

Diagramatic representation of plasmids of strain T37 (pJBSJI)
transconjugants after agarose gel electrophoresis . . . . . .

Nodulation and nitrogen fixation abilities of the four classes
of transconjugant strains on white clover and peas. . . . . . 100

Frequency of transfer of kanamycin resistance from R. trifolii
strain T37 (pJBSJI) transconjugants to R. leguminosarum strain
726 O O O O O O O O O O I I O O O O O 0 O O O O O O O O O O O 11]-

Comparison of the symbiotic properties of isolates of L.

trifolii strain 0403 containing pRt0403a or pRt0403aA . . . . 141

vii

4.

5.

10.
ll.

l2.

l3.

14.
15.

LIST OF FIGURES

Physical map and transcriptional organization of the njj_gene
Cluster of &. PLeumoniaeo O O 0 O O O O O O O 0 O O O O O O O

 

A model of £1: regulation in 5; pneumoniae. . . . . . . . . .

Agarose gel electrophoresis of plasmid DNA in A; tumefaciens
and various Rhizobium strains . . . . . . . . . . . . . . . .

 

A§kaline sucrose gradient sedimentation profile of
[ H]thymidine-labeled DNA from cells lysed using the alkaline
IySi s prOCEdure O O O O O O O O O O O O O O O 0 O O O I O O O

Ngutral sucrose gradient sedimentation profile of
[ H]thymidine-labeled cell lysate of R; trifolii strain T37 .

Agarose gel electrophoresis of plasmid DNA. . . . . . . . . .

CsCl'EtBr density equilibrium gradient of plasmid DNA
isolated by the alkaline lysis procedure. . . . . . . . . . .

Electron micrograph of an open circular (0C) molecule of RP4.

Electron micrograph of an open circular plasmid molecule from
R; leguminosarum strain 128C53. . . . . . . . . . . . . . . .

 

Growth of R; "cowpea" strain 32H1 at elevated temperature . .

Acetylene reduction assays of free-living Rhizobium grown on
cs7 medi u“. 0 O O O O O O O O O O O O O O O O O O I O O O I 0

Treatment of strains of Rhizobium with various concentrations
Of acridi ne orange. 0 O O O O O O O O O O O I O O O O O O O O

Agarose gel electrophoresis of plasmid DNA isolated from
“R; trifolii strain T37 isolates treated with acridine orange.

Partial restriction maps of pSA30 and pRmRZ . . . . . . . . .

Molecular weight of plasmid DNA versus log relative mobility
of plasmids in agarose gels . . . . . . . . . . . . . . . . .

viii

Page

17
22

28

46

48
SO

52
55

57
59

6O

63

68
85

90

16.

17.

18.

19.

20.

21.

22.
23.

24.

25.

26.

27.

28.

29.

30.

Page

Ekhardt agarose gel electrophoresis of plasmid DNA of R;

trifolii strain 0403 (pJBSJI) transconjugants . . . . . . . . 93

Autoradiogram of 32P-labeled pRmRZ to a Southern filter of
plasmid DNA from R; trifolii strain 0403 (pJBSJI)
transconjugants . . . . . . . . . . . . . . . . . ...... 95

Ekhardt agarose gel electrOphoresis of plasmid DNA of R;

trifolii strain T37 (pJBSJI) transconjugants. . . . . . . . . 96

Ekhardt agarose gel electrophoresis of bacteria isolated from
pea "OdUIES O O O O O O O O O O O O O O O O O O O O O O O O I 101

Fahraeus slide analysis of root hairs of clover plants
inoculated with strains of Rhizobium. . . . . . . . . . . . . 104

Infection thread in a root hair of clover plant inoculated
With .8; -trif01‘ii Strain T37 O O O O O O O O O O O O O O O O O 106

Partial restriction map of Tn §.. . . . . . . . . . . . . . . 108

Autoradiogram of 32P—labeled AzzTn.§ DNA hybridized to
restriction endonuclease-digested DNA . . . . . . . . . . . . 109

Ekhardt agarose gel electrophoresis of isolates of bacterial
crosses offiL trifolii strain T37 (pJBSJI) transconjugants
with R; leguminosarum strain 726. . . . . . . . . . . . . . . 112

Autoradiogram of 32P-labeled AzzTn.§ DNA hybridized to a
Southern filter of an Ekhardt agarose gel . . . . . . . . . . 113

Hybridization analysis of plasmids of Class IV transconjugant
Strains 6135 and 6137 O O O O O O O I O O O O O O O O O O O O 115

Autoradiogram of 32P-labeled‘RLmeliloti nif DNA hybridized
to Southern filter of an Ekhardt agarose gel similar to
Figure 13 O O O O O O O O O O O O O O O O O I O O O O O O O O 116

Autoradiogram of 32P-labeled‘fi; penumoniae nif DNA hybridized
to Southern blots of §29_RI-digested DNA from Rhizobium
strai "s O O O O O O O O O O O O O O O I O O O O O O O O O O O 117

Autoradiogram of 32P-labeled pJBSJI DNA hybridized to a
Southern filter of Eco RI-digested total DNA from
transconjugant straTEE. . . . . . . . . . . . . . . . . . . . 120

Autoradiogram of 32P-labeled pJBSJI DNA hybridized to a
Souther filter of Egg RI-digested total DNA from donor,
recipient and transconjugants . . . . . . . . . . . . . . . . 122

ix

31.

32.

33.

34.

35.

36.
37.

38.

39.

40.

4].

Autoradiogram of 32P-1abe1ed pJBSJI DNA hybridized to a
Southern filter of §2g_RI-digested DNA from donor and

Page

rECipient strains 0 O O O O O O O O O I O O O O O O O O O O O 123

Autoradiogram of 32P-labeled pRtT37a DNA hybridized to a
Southern filter of §22_RI-digested DNA from donor, recipient
and transconjugant strains. . . . . . . . . . . . . . . . .
A model for the generation of the "hybrid" plasmids of R;
Ekhardt agarose gel electrophoresis offiL trifolii strain
0403 and Strai n 2016. O O O O O O O O O O O O O O O O O O 0

Molecular weight of plasmid DNA versus log relative mobility
of plasmids in agarose gels . . . . . . . . . . . . . . . .

Attachment of_R_L trifolii to clover root hair tips. . . . .

Root hair curling of clover plants inoculated with R;

trifOIii strains. 0 I O O O O O O O O O O O O O O O O O O 0

Infection thread in a curled root hair of a clover plant
inoculated‘withfiL trifolii strain 0403 . . . . . . . . . .

Hybridization of a heterologous an probe to plasmid DNA
from L. trifolii strain 0403 and— strain 20l6 separated on
an Ekhardt agarose gel. . . . . . . . . . . . . . . . . . .

Autoradiogram of a heterologous an probe hybridized to a
Southern blot of Eco RI- -digested total DNA of L. trifolii .

A speculative model for IS-mediated deletion of plasmid DNA
containing Elf and Egg genes in Rhizobium . . . . . . . . .

trifolii strain T37 (pJBSJI) transconjugants. . . . . . . . .

124

133

138

. 140

143

146

148

150

151

157

ampr
BSA
CCC
DNase
dNTP
DTT
EDTA
EtBr
EtOH
FeMo-co
Fix
‘gln
IS

kanr

LIST OF ABBREVIATIONS

ampicillin resistance
bovine serum albumin
covalently closed circular
deoxyribonuclease
deoxynucleotide triphosphate
dithiothreitol
(ethylenedinitrilo)-tetraacetic acid, disodium
ethidium bromide

ethyl alcohol
iron-molybdenum cofactor
nitrogen fixation phenotype
glutamine regulatory gene
insertion sequence
kanamycin resistance
kilobase

megadalton

micro Curie

micro Einstein

nitrogen fixation genes
nodulation phenotype
nodulation gene

nitrogen regulatory gene

xi

0C
231
PEG
pfu
Egg
POPOP
PPO
PVP
rifr
RNase
rpm
$101
505
str
éym
tetr
TCA
Tn
Tris
5:2
UV

w.t.

open circular, relaxed
origin of replication
polyethylene glycol
plaque-forming units
phenylalanine auxotroph
l,4-bis[2-(4-methyl-5-phenyloxazoyl)J-benzene
2,5-diphenyloxazole
polyvinylpyrrolidone
rifampicin resistance
ribonuclease
revolutions per minute
standard deviation
sodium dodecyl sulfate
streptomycin resistance
symbiotic gene
tetracycline resistance
trichloroacetic acid
transposon
Tris(hydroxymethyl)aminoethane
tryptophan auxotroph
ultraviolet

volume

weight

wild-type

xii

INTRODUCTION

Scientists throughout the world are confronted with the task of
increasing agricultural productivity to feed an expanding world popula-
tion. Past increases in food production can be attributed primarily to
the greater use of mechanization, improved crap species, and chemicals,
such as herbicides, pesticides, and nitrogenous fertilizer. In fact,
the majority of the increases in productivity of cereal crops over the
last twenty years have resulted from increased applications of nitro-
genous fertilizer (77,182).

The production of nitrogenous fertilizer, however, is an energy
intensive process. The synthesis of ainnonia from dinitrogen is carried
out under high pressure and temperature by the Haber-Bosch process.
This industrial process requires a vast input of energy, usually from
natural gas. Currently 6% of the seven trillion cubic feet of natural
gas consumed by industrial processes in the United States is used for
the production of nitrogenous fertilizer (l). The natural gas used for
feedstock and fuel for ammonia synthesis is estimated to account for
83% of the production cost by l985 (l). For this reason, the cost of
nitrogenous fertilizer is closely correlated with the cost and avail-
ability of natural gas. The supply of natural gas is not inexhaust-
ible, however, and is subject to possible interruptions due to the
current partial dependence upon foreign sources. These disadvantages

of industrially produced nitrogenous fertilizer (i.e., high cost,

unstable energy supply) have aroused great interest in the search for
alternative technologies.

Some prokaryotic organisms can convert dinitrogen to ammonia for
growth. This process, termed biological nitrogen fixation, occurs in a
diverse group of prokaryotes. Some bacteria, such as Azotobacter, fix
dinitrogen under aerobic conditions, while others (e.g., Klebsiella and
Clostridium) fix dinitrogen anaerobically. Another genus of bacteria,
Rhizobium, normally fixes dinitrogen only in a species-specific symbio-
tic association with leguminous plants, such as clover, soybeans, and
peas. The symbiosis is beneficial to the plant in that all of the
nitrogen required for plant growth can be supplied by the Rhizobium.

In return, the host plant provides the bacteria with all of their
nutrient requirements. Thus, the energy-intensive production and
application of nitrogenous fertilizer for plant growth is obviated.

The extension of biological nitrogen fixation to important food
crops, such as the cereals, corn, wheat, and rice, is therefore an
obvious, albeit, long-range goal. A prerequisite for the extension of
symbiotic nitrogen fixation to other craps would be a better under-
standing of the genetic and biochemical events involved in the estab-
lishment and maintenance of the Rhizobium - legume symbiosis.

The establishment of the symbiosis involves the specific recogni-
tion between the species of Rhizobium and the host legume root (6,43,
44). This recognition is proposed to be mediated by the interaction
between host plant lectins and specific residues in the Rhizobium cell
wall polysaccharide (16,20,48,74,207). After attachment to the root
hair, the Rhizobium invade the root hair via a structure called the

infection thread,(43,44).

Tumor-like growths, termed nodules, develop on the legume root at
some of the infection sites (43,44). Within the nodule, bacteria are
released from the infection threads into membrane vesicles in the nod-
ule cell cytoplasm. The bacteria then differentiate into bacteroids,
the form in which biological nitrogen fixation occurs.

At the time this research was initiated, several reports indicated
that genes essential for nodulation (god) and nitrogen fixation (git)
might be located on plasmid DNA in strains of Rhizobium. Culturing
Rhizobium under certain growth conditions or in certain soils resulted
in a loss of nodulation ability (59,175,187). Treatment of strains of
Rhizobium with plasmid "curing" agents, such as intercalating dyes or
505, also resulted in the loss of nodulation ability (59,85,216).
Dunican and Tierney (60) observed conjugal transfer of 21: genes from
3;_trifolii to a 311‘ strain of E; aerogenes, and suggested that this
was plasmid mediated. Although all of these studies suggested that
plasmids might be involved in the formation of the symbiosis, there was
no direct evidence for the presence of plasmid DNA in Rhizobium.

To address this problem, experiments were designed to determine
whether plasmids were present in Rhizobium and, if so, whether the 21:
genes and/or 59g genes were plasmid-encoded.

In Chapter I, preliminary experiments are described for the detec-
tion, isolation, and characterization of plasmid DNA in strains of
Rhizobium. Plasmid "curing“ experiments were carried out, and isolates
from the treated cultures were examined for the loss of symbiotic prop-
erties and of a specific plasmid.

In Chapter II, the identification of a plasmid (pRtT37a) in L

trifolii strain 137 which encodes the Eli structural genes and genes

essential for clover nodulation is reported. This was accomplished by
the characterization of plasmids resulting from the interaction between
pJBSJI, a §;_leguminosarum plasmid, and pRtT37a, and by direct hybridi-
zation of a heterologous gjf_probe to a Southern filter of pRtT37a
DNA.

In Chapter III, the characterization of a non-nodulating mutant of
B;_trifolii strain 0403 which resulted from the spontaneous deletion of
plasmid DNA is described. The-n1: and Egg genes were localized to a

specific plasmid (pRt0403a) in R; trifolii strain 0403.

LITERATURE REVIEW

Biological Nitrogen Fixation. The process whereby dinitrogen is
enzymatically converted to ammonia is termed biological nitrogen fixa-
tion. This process occurs only in prokaryotic organisms, both free-
living and symbiotic nitrogen-fixing bacteria. The genetics and bio-
chemistry of nitrogen fixation have been extensively studied in the
free-living organisms (25,26,131). However, the genetic and biochemi-
cal events involved in the establishment and maintenance of an effec-
tive (nitrogen-fixing) symbiosis between Rhizobium and leguminous
plants are only now being elucidated.

Rhizobia are aerobic, gram-negative, rod-shaped (0.5 to 0.9 x l.2
to 3.0 pm) soil bacteria which proliferate in the rhizosphere of poten-
tial host plants (2,43,44,173). Rhizobia are capable of forming a
species-specific symbiosis with legumes, some of which are agronomical-
ly important. The specificity of the symbiotic association is the
basis for the classification of Rhizobium. Seven species of Rhizobium
and their host plants are listed in Table l.

The establishment of the Rhizobium-legume symbiosis involves a
complex series of steps (see 6,43 and 44 for reviews). Initially, the
host plant must recognize the correct species of Rhizobium among the
many soil bacteria. This recognition and attachment step has been sug-
gested to be mediated by plant lectins, proteins which bind specific

carbohydrate moieties (16,20,48,74,114,207). Rhizobium induces curling

.z o— on m we

we?“ cowumcmcmm m mmumo_uc_ mum; spzoem 20pm m; o op e we me_u cornmemcmm m mwumupu=_ mum; :uzoem ummun

.Amsflv Hemaazum soc; camemucoo mew: ounce

 

 

Amp_mumeemmv mz>wumm maqosu_:eo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Amscvaapv .qam m:c_m:4 30pm Pcwmap 4m. swan;
Apscmmqv mammoax: m_;ume<
Anon commwav cameo mzcmnmu
Acmmn m:=£~.oumwume m:m_>
Acmmn me__v mapmcap mzpommmsm .11
Ammazoov mumpzowzmca m:m_> zopm emmazouz .m amazcu
Acmmnxomv xme m=_uxpo zapm E=u_:ommw.qm somehow
Aacpae_ev a>_uam omed_uaz Same Puo_w_de.qm ae_ee_<
Acm>o_ov .qam E=P_o$_ep amok __—ow_ca.qm em>opu
€me :muccmv $.53; 3:585. .53 Zommugm 1am seam
Amma ummzmv maumeouo macxapwu
A__u:mpv mweo=PP=u mamA
Azoum> coeaouv o>_umm mwo_>
%cmma emoenv one; mmum>
Ame emceemv Ea>wuom sump;
Rama u_o_$v mmcm>em Ezmpa ummm Eaeamo:_ssum— um. guum> use mm;
pew—a umo; nouns mm_uwgm gaoem “mo:
ucmueoaa_ xp—uupeocoem< guzoew

 

 

magma; e_a;p new e=_no~_;¢ co eczema.e_mma_o .3 appeh

and deformation of root hairs of its host plant, and also the develop-
ment of a tubular structure, termed the infection thread (6,27,43,44,
212). Rhizobia are carried with the infection thread as it elongates
toward the root cortex. Plant cells in the root cortex are stimulated
to divide and enlarge, thus producing the nodule. When the infection
thread reaches the root cortex, the bacteria are released from the
infection thread into membrane vesicles (44). The bacteria then dif-
ferentiate into bacteroids, the symbiotic form which carries out bio-
logical nitrogen fixation (114). These steps will be discussed in more
depth in the next sections.

Lectin Recognition Hypothesis. One aspect of the Rhizobium -
legume symbiosis is the specificity whereby a particular legume is nod-
ulated only by a certain species of Rhizobium. The legume must, in
some fashion, recognize the correct microsymbiont (Rhizobium species)
in the soil microflora, and permit its entry to the exclusion of other,
possibly harmful, microorganisms.

Recently, plant lectins (phytolectins, phytohemaglutinins) have
been proposed to play an important role in the recognition and binding
between Rhizobium and the host root (16,20,48,74,207). Lectins are a
group of proteins, found in both plants and animals, which have the
capability of binding carbohydrates and carbohydrate-containing mole-
cules in a highly specific fashion (for review, see 115). In most spe-
cies of plants, the highest concentrations of lectins are found in the
seed, although lectins may also be present to a lesser extent in the
roots, stems and leaves (194).

The physiological role lectins play in the plant is not entirely

clear at present. One propbsed role for plant lectins in legumes is

the mediation of the selective interaction between legume roots and the
Rhizobium cell wall during the initial stages of the infection (20,48,
74). Hamblin and Kent (74) reported that phytohemagglutinin (PHA) from
seeds of Phaseolus could agglutinate 3; phaseoli cells. Production of
PHA by mature roots of Phaseolus was also observed. The binding of R;
phaseoli cells to sites on the roots suitable for infection was sug-
gested to be mediated by PHA.

A specific interaction between FITC-conjugated soybean lectin and
all but three of twenty-five strains of R; japonicum has also been
reported (20). The lectin did not bind to 23 heterologous strains of
Rhizobium which do not nodulate soybeans. Subsequent studies in
another laboratory have achieved similar results (16).

Nolpert and Albersheim (207) and Kamberger (103) have reported a
specific interaction between legume lectins and isolated cell surface
polysaccharides from several Rhizobium strains. Lectins isolated from
seeds of a legume apparently bind to the polysaccharide from the spe-
cific strain of Rhizobium capable of nodulating the host legume. No
interaction was observed between Rhizobium polysaccharides and lectins
from non-host plants.

Dazzo and Hubbell (48) have proposed a simple model for the basis
of host specificity. The clover lectin, trifoliin, is proposed to form
.3 cross-bridge between R;_trifolii and clover root hairs. The binding
of the lectin occurs to cross-reactive antigenic sites on the bacteria
and the clover root hair (48). These antigenic sites appear to be
localized to the tips of clover root hairs (46), and have a transient

appearance on the cell surface of R;_trifolii (49,96).

Some results obtained cast doubt upon the role of lectins in the
infection process. Some strains offiL japonicum formed effective nod-
ules on soybeans, but were unable to bind soybean lectin (128). This
may reflect a transient appearance of the lectin receptor on the cell
surface of R; japonicum as has been reported for R; trifolii (49, 96).

Wong (208) has observed that binding of the legume lectin to a
Rhizobium strain does not necessarily mean that the strain will nodu-
late the legune. Concanavalin A, the jack bean lectin, bound to all
Rhizobium strains tested, although only one strain could nodulate jack
beans (208). Lack of binding of host legume lectin to some Rhizobium
strains which nodulate lentils, peas, and broad beans was also observed
(208). This may again be due to the transient nature of the lectin
receptor (16,49,96).

A number of soybean lines which lack lectin in the seeds and in
the roots have been found (151,188). These soybean lines can still be
effectively nodulated. Whether there is a different non-cross reactive
lectin in the roots of these soybeans which would mediate binding of‘RL
japonicum is not known.

The inconsistent results and gaps in the data make defining a
clear role for lectins in the recognition step of the Rhizobium -
legume symbiosis difficult. Hopefully, further research on this prob-
lem will result in a better understanding of how the recognition
between the host legume and Rhizobium is mediated.

Role of Bacterial Polysaccharides. Cell surface polysaccharides
of Rhizobium are probably important in the establishment of the symbio-
sis, in conjunction with the proposed role for legume lectins (l5,16,

20,48,49,120,133,180,207). A number of different cell surface

10

polysaccharides have been described for Rhizobium: exopolysaccharides

(EPS), capsular polysaccharides (CPS), lipopolysaccharides (LPS), and

various glucans, including cellulose (6).

Structural analyses of exopolysaccharide (EPS) isolated from L

tri folii have shown a repeating unit of eight glycosyl residues (6,99,

 

162) . However, 2-deoxyglucose, the specific sugar hapten for the

c] over lectin, was not detected. This suggests that either the EPS was

not the polysaccharide which mediates the specific recognition between

R- trifolii and clover root hairs, or that the model (48) proposing a

 

Cr‘os s-bridging role for clover lectin between the plant root hair and a
bacterial polysaccharide is incorrect.
The EPS on which structural analysis was performed was isolated

fronl cultures in the phase of growth which Dazzo g; 91 reported (49)

1acked the ability to bind clover lectin. However, possible tests for

bi 01 ogical activity of the purified EPS such as lectin binding, induc-
ti on of root hair curling, etc., were not carried out.
Several polysaccharides from L 11% have been reported to

posSess biological activity. Dazzo _eLa_l_. (49) suggested that the

t"‘ansient clover lectin receptor in L trifolii was a capsular polysac-
char‘ide. Another report by the same workers found that the biological-
]3' active polysaccharide of L trifolii exhibits the characteristics of
1 ‘DOpolysaccharides (47). Thus, the exact nature of the biologically
1:“hctional polysaccharide of L trifolii is still questionable.

The exopolysaccharide (EPS) of L japonicum has been reported to
react in a biologically Specific manner with soybean lectin (16,196).

Kamberger (103) observed an interaction between soybean lectin and L

iaponicum EPS but not between the lecti n and the lipopolysaccharide of

11

‘R;_japonicum. Others, however, have observed interactions between
Rhizobium LPS and the host plant lectin (6,47,103,145,207). The purity
of the LPS used in these studies was questionable, as was the specific-
ity of binding since sugar hapten controls were not employed. Highly
purified LPS has been obtained from several strains of Rhizobium but
has not been tested for biological activity (6). The sugar composition
of the LPS from these Rhizobium strains varied as much between strains
as between species. How the great variability in the composition of
LPS would mediate the specificity of the legume-Rhizobium symbiosis is
not clear at present.

Plant Cell Responses. 1. Root Hair Curling. The first observable
plant response during the establishment of the symbiosis is root hair
curling or deformation (6,27,43,44). This response can be induced with
heterologous strains of Rhizobium, but the response is not as great as
observed with homologous strains (44,109).

The compound(s) responsible for the induction of root hair curling
has not yet been elucidated. Sterile filtrates of liquid cultures of
Rhizobium induce the typical root hair deformations, but not to the
same extent as live cells (97,184,211,212). Indole acetic acid (IAA)
has been implicated in the curling phenomenon since IAA can be synthe-
sized by Rhizobium (43,44). However, root hair deformations caused by
IAA are distinctly different than those observed with live Rhizobium
cells (6,43,44).

Hubbell (97) has reported that the curling inducer(s) for clover
root hair curling is present in a crude extracellular polysaccharide
preparation. The curling agent(s) which was dialysable and heat labile

was not characterized further. Solheim and Raa (184) observed two

 

 

12

fractions of culture filtrate which caused root hair curling. One
fraction was sensitive to nucleases and the other was believed to be a
polysaccharide or protein. Similar results have also been reported by
Yao and Vincent (212).

The function of root hair curling in the infection process has not
been well established. While infections generally occur on curled root
hairs, there is no evidence that curling is a prerequisite for infec-
tion. Indeed, several reports have indicated that infection of legume
root hairs by Rhizobium cells can occur on straight root hairs (45,143,
163). The curling of the root hair tip may result in the entrapment of
the Rhizobium cells. This might present a greater opportunity for
infection to occur (134,163).

Bauer recently preposed a speculative mechanism for root hair
curling (6). The root hair cell wall appears to be composed of two
layers (43,44). The outer layer is flexible while the inner layer
which does not extend over the growing tip region is rigid and fibril-
lar (43). Attachment of Rhizobium cells is proposed to result in
localized inhibition of synthesis of the inner rigid layer of the root
hair cell wall. As the root hair continues to elongate, the tip curls
around the Rhizobium attachment site. This eventually results in
envelopment of the attached Rhizobium cells between the root hair cell
walls.

Elgnt Cell Responses: II. Infection Thread Formation.

The bacteria invade the root hairs and are enclosed in a structure
known as the infection thread. This is a tubular structure which elon-
gates towards the base of the root hair and through preexisting plant

cell walls as it advances into the root cortex (6,43,44). The

 

13

infection thread may branch several times before reaching the root
cortex (43,44).

The infection thread appears to result from an invagination of the
root hair cell plasma membrane and cell wall (133,163). The wall of
the infection thread has been shown in electron microscopic studies to
be contiguous with the root hair cell wall (84,134,169). Histochemical
studies have shown that the infection thread is composed of the same
components as the root hair cell wall (44).

The event(s) which triggers the initiation of the infection thread
is unknown, although mechanical rupture or enzymatic degradation of the
root hair cell wall, or the entrapment of Rhizobium cells by the curled
root hair have been suggested (43,44). In the rare cases where infec-
tions occur in non-curled root hairs, the initiation of infection
threads may occur by attached Rhizobium cells at the point of contact
of two non-curled root hairs (6).

The Rhizobium cells enter the infection thread and swim freely and
multiply within it. As the infection thread branches and elongates
towards the root cortex, Rhizobium cells are carried (or swim) along.
When the infection thread reaches the root cortex, the Rhizobium cells
are released into the cytoplasm of root cortex cells in a membrane
vesicle derived from the infection thread, termed the peribacteroid
envelope (44).

Bacteroid Formation and Nodule Development. As the branching
'infection thread elongates into the root, proliferation of root corti-
<:al tissue cells occurs (44,114). The branches of the infection thread
penetrate into this meristematic region and the bacteria are released

‘into the cortical cells (43,44). The bacteria multiply within the

 

14

peribacteroid envelope and begin to differentiate into the symbiotic
form called bacteroids. This differentiation process results in little
morphological changes in R; japonicum and R;_"cowpea" bacteroid cells,
whereas, 3; leguminosarum and_R_L trifolii cells form branched and lobed
structures up to forty times larger than the free-living bacterial
cells (114).

New gene products are also expressed in the bacteroids during and
following the differentiation process. The 21: gene products, dinitro-
genase and dinitrogenase reductase, are expressed in the bacteroid
state. Dinitrogenase, a soluble iron-molybdenum protein, is a tetramer
(0282) of two different subunit types (62,131). The enzyme con-
tains a small iron-molybdenum cofactor (FeMo-co) which has been sug-
gested to be the active site for the reduction of dinitrogen to ammonia
(153,180). Dinitrogenase reductase is a non-heme iron protein composed
of two identical subunits (62,131). Dinitrogenase reductase transfers
electrons for reduction of substrate from a flavodoxin and/or ferridox-
in to dinitrogenase during nitrogen fixation (131).

Changes in the cytochrome composition also occur during differen-
tiation of bacteria to bacteroids. Cytochromes a, a3, and o are
expressed in bacteria but not in bacteroids (8). These cytochromes are
replaced in the bacteroids by several new cytochromes, including cyto-
chrome c (552), P-450, and P-420. This change in the composition of
electron transport proteins may reflect the adaptation of the bacteroid
t1) the low oxygen concentration which exists in the root nodule (8).

The activity of the heme biosynthetic enzymes increases dramati-

cally during nodule development (132). The heme produced is the

 

15

prosthetic group of leghemoglobin, an oxygen binding protein present at
high concentrations in legume root nodules.

The genetic and biochemical role of the host legume in the symbio-
sis is only now being elucidated. Classical genetic studies have indi-
cated that a number of plant genes are involved in nodule development
and the formation of an effective (nitrogen—fixing) symbiosis (32,92,
140). Apart from leghemoglobin, the identity and function of these
gene products is unknown.

Leghemoglobin is truly a unique protein. The legume host synthe-
sizes the globin apoprotein (28,53) while Rhizobium bacteroids synthe-
size the heme moiety (42,132). Leghemoglobin appears to be located in
the plant cell cytoplasm (199,200) and not inside the peribacteroid
envelope with the Rhizobium bacteroids, as previously reported (9,160).

The function of the leghemoglobin appears to be the regulation of
oxygen tension in the root nodule (3,206). The high concentration of
leghemoglobin in nodules (1 mM in soybean nodules) reduces the concen-
tration of free oxygen to less than 10 nM (3). The dinitrogenase
enzyme complex, which is sensitive to oxygen inactivation, thus remains
active. A high flux of oxygen to the respiring bacteroids is also
maintained by the high concentration of leghemoglobin.

Legocki and Verma (112,113) have detected at least 18 to 20 poly-
peptides in addition to leghemoglobin which are specific to soybean
root nodules. These nodule-specific proteins, termed nodulins, are
synthesized by the host plant. Nodulins account for about 10% of the
[355] methionine-labeled protein synthesized in the host cell cyto-
iilasm. The proteins are not detected in uninfected roots, and appear

‘to be plant gene products necessary for the development and maintenance

 

16

of the root nodule. Most of the nodulins have molecular weights
between 12,000 and 20,000 (113), while one nodulin of unknown function
has a molecular weight of 35,000 (112).

Nodules which develop from inoculation of soybeans with ineffec-
tive (Fix') strains of R; japonicum generally contain lower levels of
the nodulins (112). Differential expression of nodulins was observed
in nodules of soybeans inoculated with different mutant strains of_R_L
japonicum. At present, the mechanism by which B; japonicum (and mutant
strains) influence the expression of the host nodulin genes is not
known.

Organization and Regulation of the nif Genes. Genetic analysis of
the 21: genes in strains of Rhizobium has been difficult since these
strains normally fix nitrogen only during the symbiosis with the legume
plant. Most of the genetic analyses of the nif genes have been per-

formed on the free-living, nitrogen-fixing bacterium, 5. pneumoniae.

 

The nif operon in 5; pneumoniae consists of at least 17 contiguous

 

genes located on a 24 kb segment of the chromosome (26,160). The Elf
genes are arranged into 7 or 8 operons which are transcribed in the
same direction (Figure l). Many of the Eli gene products have been
identified and functions have been assigned for some. These data are
sunmarized in Table 2.

The 31: H gene codes for the iron-containing protein, dinitro-
genase reductase (161). The gene product of £1: H is processed by the
products of 11E M and 1111 S to yield the active protein (150,159,161).
The a- and s-subunits of dinitrogenase, the FeMo proteins, are coded
‘for by Eli D and 21: K, respectively (161). An iron-molybdenum cofac-

‘tor'(FeMo-co) is synthesized and processed by the gene products of git

17

came
mzP

.5: fl... old 38.5 2; AG: Am mm 3233. so: 8538 2a,,
.ama Pmomeza any m>onm mzoeea men an umumuPucP mew :oPuaPeumcmeu Po :oPuumePu as» new mcoemgo cm>mm
.mmPcosawea .x Po emumzPu memo PP: use Po :oPumNPcmmeo PecoPuaPeumemeu use nae Pnquxzm .P meszm

 

 

 

4|

 

 

11 «[1 | I l 1.1

18

Table 2. ‘5; pneumoniae nif gene products

 

 

 

 

 

 

Gene Molecular Weight Function

J l20,000 electron transport

H 35,000 dinitrogenase reductase subunit
0 56,000 a-subunit of dinitrogenase

K 60,000 B-subunit of dinitrogenase

Y 19 24,000 (-)

E 40 46,000 FeMo-co processing or insertion
N 50,000 FeMo-co processing or insertion
x l8,000 (-)

u 22 28,000 (-)

S 42 45,000 processing of 91: H gene product
V 38 42,000 FeMo-co processing

M 27,000 processing of 51: H gene product
F l0 22,000 electron transport

L 45 55,000 negative regulation

(transcriptional)
A 57 66,000 positive regulation
(transcriptional)

B (-) FeMo-co processing or insertion
Q (-) FeMo-co processing or insertion

IJata were summarized from Roberts et al. (161) and Puhler and Klipp

(150).

gene product is not known.

(-) indicates that the molEEuTEr weight or function of this

19

E, N, V, B, and Q and is inserted into dinitrogenase to form the active
site of dinitrogen reduction (26.125,150,161,170). The reduction of
dinitrogen to ammonium requires the presence of electron donors. 12.
3132, the electrons are transported to the dinitrogenase complex by the
‘gif F and g1: J gene products (161,170). The Elf F protein has been
identified as a flavodoxin (86,135,159), while the nif J protein is an
iron-sulfur protein (19). At least three other Elf genes, Eli U, X,
and Y, have been identified, but functions for these genes have not yet
been established (4,34,150).

The regulation of the Eli operon in Klebsiella is complex. Oxygen

 

inactivates dinitrogenase as well as repressing expression of the 31:
genes (31,61,171). Ammonium (161) and temperatures above 37°C (83,214)
also repress nitrogen fixation. Two genes, git L and nif_A, appear to
exert regulatory control over the entire nit operon. The.nif L protein
is a negative regulatory factor and mediates oxygen repression and to a
lesser extent ammonium repression (30,87,127). The 31: A gene product
is a positive regulatory factor and is required for expression of all
of the £1: operons except 21: LA. The 31: A protein is thermolabile
(214), so 51f gene expression is repressed at temperatures above 37°C.
Other proteins involved in the general nitrogen metabolism in the
cell are also involved in the regulation of nit gene expression. The
‘glg 0 (Qt: C) and 913 F (gt; A) gene products, which are required for
the expression of the glutamine synthetase operon (gln_ALG), are also
required for positive activation of the g1: operon (50.64). The acti-
vation of _r_1_i_f expression by glg G (n33; C) and gl_n_ F (fit; A) gene pro-
ducts is mediated at the 31: LA operon, as is the repression of 31f

expression by 9111 L (_n_t_r_- B) and 913 6 (fit: C) (50,58,64,142).

20

A model of _n_i_f_ regulation in _K_. pneumoniae is shown in Figure 2.
In response to nitrogen starvation, the gln G (n_t£ C) gene product in
concert with the gl_n F (M A) protein activates transcription of the
_n_i_f_ LA operon (50,64). The products of the HE A and gl_n F (it: A)
genes are required for the expression of all other njj genes (58,142).
In addition, the _n_ij A gene product can substitute for the gl_n G (nt_r
C) protein in the activation of the _g_l_n_ ALG Operon, other nitrogen
assimilatory genes, and its own (n31 LA) operon (58,142). The substi-
tution is not reciprocal, however, since the 1(_._ pneumoniae gln G (n_t£
C) protein does not activate the I5; mieumoniae nif H promoter and prob-
ably not promoters for the other fli_f_ genes as well (142).

Repression of m expression in the presence of high (30 mM)
amoniun is mediated at the _rfif LA operon by the _gl_n L (_r_1_t_r; B) and SM
G (1t_r C) proteins. The _ni_f L protein represses synthesis of the other
_n_i_f proteins in the presence of oxygen and amnonium (30,87,127).

The regulation of nitrogen fixation in Rhizobium may occur by a
somewhat different mechanism. Rhizobium strains are not repressed for
dinitrogenase synthesis by "fixed” nitrogen compounds, such as ammonium
(17). This might be expected since nitrogen-fixing bacteroids excrete
amnoniun into the plant cytosol.

Gene fusions constructed with the 3; meliloti n_i_f H promoter and
the 119 Z gene from g; £9_l_i_ show that the g_i_f_ H promoter can be acti-
vated by the .K_. pneumoniae n_i_i: A protein (189). Sequencing data indi-
cates about 50% homology between the _rflj H promoters for R_._ meliloti
and 5; pneumoniae, with several regions of exact homology (189). This
suggests that 5. meliloti may contain a regulatory protein analogous to

the 5. pneumoniae _n_fi A protein. The .913 0 gene product may directly

21

.mPcoEEm Po uncommon ogu :klmmzoem chezu weeps—game

m>Pummm= oum men a mam.ncm 4 mam. .EchoEEo go cmmxxo Po accommea ms“ cP mmcmm PP: emzuo ms» Po Lommhumme

PmcoPqueumemcu m>Puommc a mP 4 Hflm .mmcmm “Hm.cmzuo «nu ma :oPqugomcmgp mmum>Puoa chuoeg < PP: ogu

.coPuPuum cP .Louo>Puom co molmwwpoea o mamrwm» LoP wuauPumnzm coo :Pmuoea <.Hflm.m;P .mmcmm xuoum_:mme
:_

:mmoeuP: emguo new .coemao <4 PP: any .Pu4< V :oewno ammuwgucam chEmuaPm on» mpm>Puum chmuoea m mam.
new a mam..;uzogm Po mcoPuPccou chuPEPPncmmo

mmw: mmwxao .Emthnaume :mmogch Pmmwcmm mgu Po mgoum 3mm;
PmcoPuanumcmgu m>Pquoa men m mmw.nco w :P .mmPcoaamcm,.x :P :oPuismm; PP: Po Ponce < .N we: Pm

 

 

22

zo_mmm_mn_mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

,r 4er .oSulH 1:
Pi 1m P L «:2 16.4.2
_ _
_ .
i P _ P .
> mocm - P u u . mace
3%me P 3: 3: Pie a“
1 I -
P P,, P P P
1w... “_Su.@:1\\
r “EBQEW 1L

 

 

29.25 _H 04

23

regulate MI expression in some Rhizobium. The _E_._c_oll'gfl G (33:3 C)
gene product, acting in conjunction with the _l_-Z_. all g_l_n F (_n_t_r_ A)
activated the 3;_meliloti nif H promoter in 31: Hglag Z gene fusions
(142). In contrast, the K;pneumoniae gln G (gt5_C) gene product can-

not activate K; pneumoniae nif H transcription (142). The hypothesis

 

that n1: expression can be regulated by the glfl_G (gt: C) gene product
is supported by sequence analysis of five promoters under the general
nitrogen regulatory control of gln G and gln F gene products (143).
These promoters, which included K; pneumoniae nif L, g; gglj_glg_A, 3;
meliloti Elf H and Salmonella typhimurium arg Tr and ghg_A were found
to have a 7 base pair consensus sequence (TTTTGCA) in the -l5 region.
The K; pneumoniae nif H promoter, which is not under glg_Gggln F con-
trol has only partial homology (CCCTGCA) in this region.

Ruvkun and Ausubel (166) have shown that the K;pneumoniae nif
genes are homologous to DNA from l9 other diverse nitrogen-fixing bac-
terial strains, and that the genes did not hybridize to DNA from l0
different non-nitrogen-fixing bacteria. The homologous region of 31:
DNA is localized to a l.6 kb region of the 31: H and nif_D genes.

The interspecies homology of 21: DNA has been useful in the isola-
tion of 31: DNA from gene libraries of 3; meliloti and g; japonicum DNA
(82,166). The availability of cloned 21: DNA from Rhizobium has ena-
bled preliminary studies of the genetic organization of the git operon
in Rhizobium to be carried out. In §;_meliloti and g;_leguminosarum,
the 51: H, D, and K genes are located adjacent to each other as in K;
pneumoniae (5,40,143). However, the organization of the 31: H, D, and
K genes in 3;_japonicum apparently is different. Transcript mapping
indicates that the £1: 0 gene is adjacent to the g1: K gene and that

24

the genes are transcribed together. The Elf H gene is located a short
distance upstream from gij_DK, however, and is probably in another
transcription unit (147).

The region of DNA adjacent to and containing the 3; meliloti nif
HDK genes has been analyzed using transposon Tn.§ insertion mutagenesis
(39,40,168). The sites of insertion of Tn.§ have been physically
mapped, and complementation analysis between genomic nifzzTn‘g inser-
tions and nifzzTn.§ insertions on mobilizable cloning vectors has been
carried out. Fix‘ and Fix+ phenotypes were observed on alfalfa
plants inoculated with the Tn.§ insertion mutants. The Fix“ mutants
were clustered in a l4 to l5 kb region of DNA, of which l.9 kb was not
essential for nitrogen fixation (39). This compares to a 24 kb segment
of DNA which encodes the g1: genes in K; pneumoniae (Figure l). The
presence of three transcriptional units in the 14 to l5 kb region of
DNA containing the gfij HDK genes were deduced from complementation
analysis data (166). The 31: HDK genes constitute one transcription
unit about 6.3 kb long. The other two transcription units are located
upstream from the 21: HDK transcription start site. One of these
transcriptional units is apparently transcribed in the opposite direc-
tion as compared to the nif HDK operon (39). These results indicate
that the organization of the 3; meliloti £1: operon is different from
that of K; pneumoniae where only one transcription unit, 21: J is
upstream from £1: HDK (see Figure l).

Plasmid DNA in Rhizobium. Historically, many investigators have
reported that the nodulation (Nod) and nitrogen fixation (Fix) pheno-
types are unstable (2,59,175). Prolonged storage of Rhizobium on cer-

tain media (117) or in sterilized soil (139,154) have resulted in up to

25

35% of the recovered colonies possessing a Fix“ phenotype (192).
Preliminary studies did not uncover a genetic basis for the instability
of the symbiotic phenotypes (59).

More recently, Higashi (86) was able to demonstrate transfer of
clover nodulation from R; trifolii to R; phaseol'. The clover nodula-
tion phenotype could be partially eliminated by growth of the R;
trifolii strain and the R; trifoliiAR; phaseoli transconjugant in the
presence of acridine orange. Several other reports have since been
published showing the loss of nodulation ability from strains of
Rhizobium as a result of growth in the presence of intercalating dyes
(59,216). Intercalating dyes are known to eliminate some plasmids by
interfering with their replication (38,88,91), suggesting that the
genetic information coding for nodulation and nitrogen fixation may be
located on plasmid DNA in Rhizobium.

Plasmids are extra chromosomal genetic elements present in a wide
variety of gram—negative and gram-positive bacteria (38,136). Plasmids
replicate autonomously (38) and exist in the cells as covalently closed
circular (CCC) DNA molecules (183). Plasmids have a considerable
variation in size and also in copy number, the number of plasmid mole-
cules of one type stably maintained in a bacterial cell (38). A wide
variety of functions are encoded on plasmid DNA, including: bacterio-
cin production, antibiotic resistance and promotion of conjugation (see
38,81,136 for Reviews).

Using techniques developed for the detection and isolation of
plasmid DNA in other bacteria, a number of effective (Nod+ Fix+)
and some ineffective (Nod+ Fix‘) Rhizobium strains were analyzed

for plasmid DNA. Although plasmids were not detected in all of the

26

strains examined, plasmids of molecular weight ranging from 5.5 x

l06 to 64 x l06 were observed in R;_"cowpea" (190), R;_trifolii
(106,195,217), and R;_japonicum (104). No correlation could be made
between nodulation and nitrogen fixation abilities and the presence or
absence of plasmid DNA in these effective and ineffective strains of
Rhizobium.

Agrobacterium, a genus of bacteria closely related to Rhizobium

 

(69,80) causes tumorous growths, or crown galls, on dicotyledonous
plants (57). The genes coding for tumor formation in Agrobacterium
are encoded on large plasmids having molecular weights greater than l00
x l06 (57,198,203). The similarity between the formation of crown
gall tunors by Agobacterium and the induction of root nodules by
Rhizobium led to the suggestion that large plasmids in Rhizobium may
encode genes essential for the Rhizobium-legume symbiosis. Methods
developed for the isolation of large plasmids from A; tumefaciens (41,
111,213) have been applied to Rhizobium. A wide range of large plas-
mids have been observed in R; leguminosarum (23,137,146), 3;_meliloti
(35), and R; japonicum (72,123). Using these procedures, the isolation
of plasmids greater than 200 Mdal was difficult and recovery was poor
(11,23). However, new techniques (5,164,177), especially the "in gel“
lysis technique of Ekhardt (63), allow the detection and isolation of
very large plasmids. Indeed, plasmids much greater than 300 Mdal in
size, termed "megaplasmids", have been detected in virtually all
strains of R; meliloti using these techniques (5,164).

The number and size of the plasmids in Rhizobium varies greatly
(11). The diversity of plasmids in Rhizobium can be seen in the

Ekhardt agarose gel shown in Figure 3. As many as seven plasmids may

27

be present in a strain (Figure 3, lane c). Plasmids of molecular
weight ranging from about 30 x l06 to greater than 325 x l06 can be
detected in Rhizobium.

Genetic Functions Encoded on Plasmids in Rhizobum. In Rhizobium,
plasmid DNA can account for 25 to 30% of the total DNA due to the num-
ber and large size of the plasmids (ll). However, few genetic func-
tions have been determined to be specified on these plasmids.

Bacteriocins, compounds which inhibit the growth of closely
related species of bacteria, are known to be plasmid encoded in entero-
bacteria (76). Production of bacteriocins by Rhizobium has been des-
cribed by many investigators (165,176,178) but was not known to be
plasmid encoded.

Hirsch (89) identified three transmissible plasmids in R;

leguminosarum which code for a bacteriocin. The bacteriocin was class-

 

ified as medium in size since it diffused about 5 mm in agar plates but
could not diffuse through a dialysis membrane. A second locus on these
plasmids repressed the synthesis of a small bacteriocin when the
plasmids were transferred into R;_leguminosarum strains which normally
produce the small bacteriocin.

Pea nodulation ability was found to cotransfer at high frequency
(l0'3 to l0‘2) with the ability to produce the medium bacterio-
cin (22,100). Transfer of pRllJI, a conjugative plasmid in R;
leguminosarum which codes for the production of medium bacteriocin pro-
duction, into Nod' strains of R; leguminosarum resulted in transcon—
jugants with a Nod+ Fix+ phenotype on pea plants (100). Transfer
of mutant (Fix') derivatives of pRllJI (29) into a Nod‘ strain of

.3; leguminosarum containing a deletion in one of its plasmids resulted

28

 

Figure 3. Ekhardt agarose gel electrophoresis of plasmid DNA in A:
tumefaciens and three strains of Rhizobium. Lanes a and e, A.
tumefaCiens strain 058(RP4); b, _R_. meliloti strain Su27; c, E
leguminosarum strain T69; d, B_. trif—o‘fl strain 0403. Samples were
prepared and Ekhardt agarose gel electrophoresis performed by the
author as described in Chapter II Methods.

 

 

29

in transconjugants with Nod+ Fix' phenotypes (22). This indicates
that both Egg and 211 genes were contained on the region of deleted
plasmid DNA. Similar results have been obtained in other species of
Rhizobium. Transfer of a.R; trifolii plasmid from Nod+ strains of R;
trifolii to Nod' strains conferred clover nodulation ability on the
recipient (95,179,215). Deletion of plasmid DNA in R; meliloti (5,164)
and the elimination of a 111 Mdal R; leguminosarum plasmid (146)
resulted in a Nod' phenotype on the respective host plant.

The 51: structural genes have been located on large plasmids in
many Rhizobium strains (5,79,94,123,138,164). A heterologous Elf probe
from K; pneumoniae hybridized to a Southern filter of restriction endo-
nuclease-digested total plasmid DNA from R;_leguminosarum (138).
Results from similar Southern hybridization filter analyses have indi-
cated that one of the plasmids in strains of R; trifolii and R;
phaseoli encode the £1: structural genes (147).

The specific plasmid which encodes the_gif genes has since been
identified in some Rhizobium strains. Hybridization of a heterologous
.gif probe to a Southern filter of plasmids separated on agarose gels
has identified the gjjfcontaining plasmid in R;_leguminosarum (94,148),
.3; phaseoli (94), and R; japonicum (79,123).

Strains of'3;_meliloti harbor very large plasmids ("megaplasmids")
coding for genes required for nodulation of alfalfa (7,164). Hybridi-

zation with cloned R;_meliloti nif DNA has shown that the Eli’structur-

 

al genes are also located on this plasmid (5,164). Using a series of
overlapping cosmid clones of R; meliloti DNA, Long gt al. (116) have

localized a nodulation gene with 30 kb of the £1: structural genes.

30

Hombrecher gt al. (77) have reported linkage of £1: and Egg genes on
plasmids in R;_leguminosarum and R;_phaseoli.
Genetic determinants for a hydrogen uptake enzyme (Hup) have been

shown to be plasmid-encoded in R; leguminosarum (24). The uptake

 

hydrogenase permits recycling of hydrogen gas, a by-product of the
dinitrogenase-catalyzed reduction of dinitrogen (174). Nodulation

ability (NodT) was transferred from a Hup+ R; leguminosarum strain

 

to a Nod‘ Hup' strain using a kanamycin-resistant derivative of a
plasmid (pRl3JI) which is known to mobilize other plasmids (22).
Approximately 70% of the kanamycin-resistant transconjugants were
Nod+ Hup+, suggesting 39g and hug genes are linked on a plasmid in
this 3; leguminosarum strain.

The expression of plasmid DNA in the bacteroid state has also been
analyzed (191,202,203). Krol gt al. (107) have obtained hybridization
of R;_leguminosarum plasmid DNA to RNA from bacteroids. Little or no
hybridization was observed to RNA from vegetative bacteria. The RNA
hybridized to only one of the two plasmids present in the R;
leguminosarum strain (108). Regions of DNA of a different 5;
leguminosarum plasmid which is heavily transcribed in bacteroids have
been mapped (148). One of the regions contains the gjj_structural
genes. The functions of the genes in the other transcribed areas are
at present unknown, but presumably are involved in the maintenance of
nitrogen-fixing nodules.

Other genetic functions which may be involved in the establishment
and maintenance of the symbiosis have been determined to be plasmid-
encoded. A plasmid in a.R; leguminosarum strain appears to code for

genes required for synthesis of exopolysaccharide (146). Isolates

31

which had lost a 111 Mdal plasmid as a result of "heat-curing", had a
rough colony morphology, indicative of an inability to synthesize
exopolysaccharide. These isolates also had altered phage sensitivities
and were unable to nodulate pea plants.

The ability of strains of R;_meliloti to produce polygalacturonase
has been linked to the presence of a 59.6 Mdal plasmid (141). Curing
of the plasmid by growth of R;_meliloti strains in the presence of
acridine resulted in isolates with low levels of polygalacturonase
activity. The role of this enzyme in the degradation of the root hair

cell wall during the infection process is still in dispute (6,43,44).

CHAPTER I

Preliminary Studies on the Presence of Plasmid DNA
and Plasmid-Coding of Genes for Nodulation and
Nitrogen Fixation in Rhizobium.

INTRODUCTION

At the time this research was initiated, plasmid coding of gym
genes had been implicated in strains of Rhizobium. Treatment of
strains of Rhizobium with plasmid "curing" agents, such as acridine
dyes (85,216), or culturing Rhizobium under certain growth conditions
resulted in the loss of nodulation ability (59,175,187). Conjugal
transfer of clover nodulation and nitrogen fixation (Elf) genes was
observed, and was suggested to be plasmid-mediated (60,85). However,
no direct correlation of the absence or transfer of a specific plasmid
with the loss or transfer of symbiotic properties had been reported at
that time.

Two widely used plasmid isolation techniques, the SDS-salt precip-
itation technique (73) and the cleared-lysate technique (37) are suita-
ble for the isolation of plasmids of molecular weight <l00 Mdal. How-
ever, these techniques were not suitable for the detection of plasmids
in most strains of Rhizobium. Subsequently, techniques for the isola-
tion of large (>l00 Mdal) plasmids have been developed (35,41,111).

In this Chapter, I report and discuss: l) two methods for the
isolation and detection of plasmids in strains of Rhizobium; 2)
attempts to eliminate or "cure" plasmids from strains of Rhizobium and
the examination of isolates from the treated cultures for the loss

of symbiotic properties.

32

MATERIALS AND METHODS

Materials. Yeast extract and Bacto agar were obtained from Difco
Laboratories, Detroit, MI. Acridine orange, ethidium bromide, cyto-
chrome c (Type III), pronase (Type XIV) and Dowex 50 were purchased
from Sigma Chemical Company, St. Louis, MO. Sodium N-laurylsarcosinate
(sarkosyl) was obtained from ICN Pharmaceuticals, Plainview, NY.

Sodium dodecyl sulfate (SDS) was purchased from either Pierce Chemical
Company, Rockford, IL, or Bio-Rad Laboratories, Richmond, CA. Sucrose
was obtained from Schwartz/Mann, Orangeburg, NY. Agarose was purchased
from Bio-Rad Laboratories. 3H-Thymidine (2 Ci/mmole) was obtained
from New England Nuclear, Boston, MA. Platinumzpaladium (80:20; 0.008
inch diameter) was obtained from Ted Pella, Inc., Tustin, CA. Scintil-
lation fluid contained 66.7% toluene, 33.3% triton, 0.5% PPD, 0.0l%
POPOP (v:v:w:w). Photography of ethidium bromide-stained gels was car-
ried out using Polaroid Type 57 or Type 667 film (Polaroid Corporation,
Cambridge, MA).

Bacterial strains. The strains of Rhizobium used were: .3;
leguminosarum strain l28C53, obtained from J. Burton, The Nitragin
Company, Milwaukee, NI; R. trifolii strain T37, obtained from F.B.
Dazzo, Michigan State University, East Lansing, MI; R;_"cowpea" strain
C8756, obtained from J. Tjepkema, University of Maine, 0rono, ME; ‘3;
"cowpea" strain 32H1, obtained from J.M. Vincent, University of

Australia, Sydney, Australia; and Agrobacterium tumefaciens strain

33

34

C58(RP4) obtained from T.C. Currier, Kansas State University,
Manhattan, KS.

All strains of Rhizobium were symbiotically effective on their

normal host plant. R. trifolii strain T37 (118) as well as R;_"cowpea

 

strains 08756 and 32H1 (144) have been reported to fix nitrogen in cul-
ture. A. tumefaciens strain C58(RP4) contains the 36 Mdal broad-host-
range plasmid RP4 (52), a l30 Mdal tumor-inducing (Ti) plasmid (93),
and a 325 Mdal cryptic plasmid (35).

.flEQiE- Bacteria were routinely grown on yeast extract-mannitol
(YEM) medium containing (per liter redistilled water): l0.0 g mannitol,
2.0 g yeast extract, 0.2 g KH2P04, 0.3 g KZHP04, 0.2 g
MgS04'7H20, and 0.l 9 NaCl. YEM medium was solidified with l.5%

Bacto agar. Bacteria grown for plasmid isolations were cultured in
YEZ medium, which was similar to YEM but contained no mannitol.
Bacterial DNA was labeled by culturing bacteria in Wright's medium
(195) supplemented with 0.l mCi/ml 3H-thymidine (2 Ci/mmole).

Wright's medium contained: 0.2 g NaCl, 0.5 g KZHP04, 0.2 g
MgSO4°7H20, 0.1 g CaSO4°7H20, 0.l g CaC03, and l.0 g

yeast extract per liter redistilled water. Bacteria to be assayed for
dinitrogenase activity (acetylene reduction) in culture were grown on
CS7 mediun (144). CS7 medium was prepared by mixing 0.l vol une of
Solution A and 0.0l volume of Solution 8 with 0.8 volumes of redis-
tilled water. Agar (l g/l) was added and the medium was sterilized by
autoclaving. The autoclaved medium was cooled to 50°C and 0.l volume
of filter sterilized Solution C was added. The CS7 medium was dis-
pensed into 3-dram vials. Solution A contained: 0.3 g KH2P04, 0.l

g CaClz°2H20, 0.07 9 KCl, 0.035 g MgS04°7H20, l.0 g

35

.mygeinositol, and 2.55 g succinic acid in l00 ml of redistilled water.
Solution B contained: 98 mg MnSO4'H20, 50 mg H3803, l0 mg ZnSO4'7H20,
l0 mg KI, 2 mg CuSO4'5H20, l.0 mg Na2M004'2H20, l.0 mg CoClz'6H20, and
20l mg disodium EDTA per l00 ml of redistilled water. Solution C con-
tained 9.38 g arabinose, 73l mg glutamine, 37.5 mg FeSO4'7H20, l3 mg
thiamine'HCl, l3 mg nicotinic acid and 1.2 mg pyridoxine HCl per l00
ml of redistilled water. The pH of the medium was adjusted to pH 5.9.
Currier-Nester Plasmid Isolation Procedure. This procedure is
suitable for the isolation of large (5130 Mdal) plasmids (41). Cells
were grown in YEZ medium to late log phase (l50-l75 Klett units; l to
2 x l09 cells/ml). Bacteria were pelleted by centrifugation at
l0,000 x g for 20 min at 4°C, washed once with l0 mM TriS°HCl (pH
8.0), l M EDTA (TE buffer) and resuspended in 0.2 vol unes TE buffer.
Pronase (l0 mg/ml predigested at 37°C for l.5 h) and $05 (l0%) were
added to a final concentration of 0.5 mg/ml and 1%, respectively. The
cells were incubated for 45 min at 37°C. The viscosity of the lysate
was reduced by shearing the DNA for l min with a vortexer (Scientific
Products Model 38220) or with a Serval Omni Mixer (for volumes > l00
ml). The sheared lysate was adjusted to pH l2.4 with 3 M NaOH and
incubated at room temperature for l0 min with occasional stirring. The
lysate was neutralized with 2 M Tris'HCl (pH 7.0) to pH 8.5-9.0 and
stirred gently for 3 min. Sodium chloride (30%; w/v) was added to the
neutralized lysate to a final concentration of 3% (w/v) NaCl and the
lysate was extracted with an equal volume of redistilled phenol equili-
brated with 3% (w/v) NaCl. The phenol and aqueous phases were sepa-

rated by centrifugation at 6000 x g for l0 min at 4°C. Plasmid DNA was

36

contained in the aqueous phase while the denatured chromosomal DNA
banded at the interface. The aqueous phase was recovered using an
inverted 5 ml pipet and was extracted with an equal volume of
chloroformzisoamyl alcohol (24:l). The aqueous phase was recovered as
before, and 0.l volume of 3 M sodium acetate and 2 volumes of 95%
ethanol were added to precipitate the DNA. After a minimum of 4 h at
-20°C, the DNA was pelleted by centrifugation at l2,000 x g for 20 min
at 4°C. The DNA pellet was dried and was resuspended in 200-500 ul TE
buffer. Samples were analyzed for plasmid DNA by agarose gel electro-
phoresis or CsCl'EtBr density equilibrium centrifugation.

Alkaline Lysis Procedure for Plasmid Isolation. This method (35)
utilizes a non-enzymatic and more gentle lysis than the Currier-Nester
procedure. Bacteria were grown in YEZ medium to a density of about
50 Klett (5 x l08 cells/ml). Cells were pelleted by centrifugation
at l2,000 x g for l0 min at 4°C, washed first with l M NaCl, l0 mM
EDTA, and then washed with TE buffer. The cells were resuspended in TE
buffer at a concentration of 0.2 9 cells/ml. The cell suspension (0.5
ml) was added to 9.5 ml alkaline lysis buffer (1% (w/v) Sarkosyl in TE
buffer, pH l2.45). A 2.5 cm stir bar was added and the mixture was
stirred for 90 sec at l00 rpm. The mixture was incubated at 34°C for
20 min, and neutralized with 0.6 ml 2 M Tris-HCl (pH 7.0), and was
stirred for 2 min at l00 rpm. Sodium chloride (0.l volume of a 30%
(w/v) solution) was added and the solution was stirred another l0 sec.
The lysate was incubated at room temperature for 30 min. An equal
volume of redistilled phenol equilibrated with 3% (w/v) NaCl was added.
The mixture was stirred rapidly (about 300 rpm) for l0 sec, and then

the stirring rate was decreased to l00 rpm for 2 min. The two phases

37

were separated by centrifugation at 5000 x g for l0 min at 4°C. The
clear upper aqueous phase was transferred to a new tube using an
inverted 5 ml pipet. The plasmid solution was brought to 0.3 M sodium
acetate with 0.l volume of 3 M sodium acetate and 2 volumes of 95%
ethanol were added. DNA was precipitated at -20°C and pelleted by
centrifugation at l2,000 x g for 20 min at 4°C. The supernatant fluid
was decanted and the DNA pellet was allowed to dry. The pellet was
resuspended in l00 pl TE buffer. Samples were analyzed for plasmid DNA
by agarose gel electrophoresis or further purified by CsCl°EtBr den-
sity equilibrium centrifugation. The procedure can be scaled up 5-fold
for the isolation of greater amounts of plasmid DNA.

LEHJThymidine Labelinggof Bacterial DNA. The inocula for
[3H]thymidine labeling studies were grown in YEM medium to late log
phase (1 to 2 x 109 cells/ml; 150-175 Klett units). An aliquot (0.1
ml) was used to inoculate l0 ml of Nright's medium containing l0 uCi/ml
[3H]thymidine (2 Ci/mmole). The cultures were grown for 24 h at 30°C
with shaking. The culture labeled with [3H]thymidine was mixed with
a 40 ml of unlabeled culture. Cells were pelleted by centrifugation at
12,000 x g for l0 min at 4°C and washed once with l M NaCl, l mM EDTA
and once with TE buffer. The cells were resuspended at a concentration
of 0.2 g cell/ml and lysed using the alkaline lysis procedure. Analy-
sis for plasmid DNA was carried out on sucrose gradients.

Sucrose Gradient Centrifugation. Neutral and alkaline linear
5-20% sucrose gradients were used to analyze 3H-labeled bacterial DNA
for plasmids (111,213). DNA samples to be analyzed on neutral sucrose
gradients were isolated by the alkaline lysis procedure. Samples were

layered on linear 5-20% neutral sucrose gradients (l0 ml) prepared in

38

100 mM NaCl, 10 mM Tri5°HCl (pH 8.0), l mM EDTA (STE buffer). The
gradient was centrifuged at 37,000 rpm for 60 min at 4°C in a Beckman
SN41 rotor. Fractions (100 pl) were collected from the bottom directly
into scintillation vials. Redistilled water (0.5 ml) and 5 ml of scin-
tillation fluid (see Materials) were added and radioactivity was deter-
mined by liquid scintillation spectrometry. Samples to be analyzed on
alkaline sucrose gradients were lysed as described in the alkaline
lysis procedure, but were not neutralized. Instead, lysates were
sheared for 30 sec (two-15 sec pulses) using a vortexer (at full speed)
and 1.0 ml of the lysate was layered on an alkaline sucrose gradient.
Linear 5-20% alkaline sucrose gradients (10 ml) were prepared in l M
NaCl, 0.3 M NaOH, 10 mM Tris'HCl (pH 8.0), 50 mM EDTA. The gradient
was centrifuged at 37,000 x g for 20 min at 4°C in a Beckman SN4l
rotor. Fractions were collected and radioactivity was determined as
described for neutral sucrose gradients.

Agarose Gel Electrophoresis. Plasmid DNA was separated on verti-
cal 0.7% agarose gels (14 mm x 12 mm x 3 mm). Tris-borate - EDTA (TBE)
electrophoresis buffer consisted of 89 mM Tris (Trizmabase), 89 mM
H3803 and 8.9 mM EDTA (129). Electrophoresis was carried out at
100 volts for 3 h. The gel was stained for 15 min with ethidium bro-
mide (0.5 pg/ml). DNA was visualized with UV light and the gel was
photographed.

CsCl'EtBr Density Equilibrium Centrifugation. Supercoiled (CCC)
plasmid DNA was separated from nicked plasmid and linear chromosomal
DNA by CsCl°EtBr density equilibrium centrifugation (152). Cesium
chloride (9 g) was dissolved in 9 ml TE buffer. The resuspended plas-
mid DNA in 1 m1 of TE buffer and 0.3 ml of ethidium bromide (l0 mg/ml)

39

were added to the CsCl solution and the solutions were gently mixed to
avoid shearing the DNA. The gradients were centrifuged at 35,000 rpm
for 36 h at 4°C in a Beckman Ti 50.1 rotor. The plasmid DNA was visu-
alized with UV light and removed from the gradient. Cesium chloride
was removed from the plasmid DNA solution by dialysis against one l TE
buffer (4 changes). Ethidium bromide was removed from the DNA by
Dowex-50 chromatography in TE buffer (152) or by repeated extraction
with isopropyl alcohol. Plasmid DNA was concentrated by ethanol pre-
cipitation at -20°C. The concentration of plasmid DNA was determined
by the absorbance at 260 nm.

Electron Microscopy of Plasmid DNA. The microspreading modifica-
tion (98) of the protein monolayer technique of Kleinschmidt (105) was
used to prepare DNA samples for electron microscopy. Hyperphase solu-
tions (10 01) containing 1 09 DNA, 1 ug cytochrome c, and 0.5 M ammoni-
um acetate were spread down a glass rod onto approximately 2 ml of
hypophase solution (0.25 M ammonium acetate). The protein monolayer
and plasmid DNA were picked up with Parlodion- coated, ZOO-mesh copper
grids. The DNA was stained (205) for 30 seconds with a 1:100 dilution
of a solution of 5 mM uranyl acetate, 5 mM HCl in 95% ethanol, and
rotary shadowed with platinumzpaladium (80:20). Electron microscopy
was carried out on a Siemens Elmiskop 1A electron microsc0pe. Magnifi-
cation was calibrated using a carbon replica of a diffraction grating
(54,864 lines/inch). Contour length measurements of DNA molecules were
determined from about 3X print enlargements using a Keufel and Esser
opisometer. The print magnification was determined from the ratio of
the distance between two objects on the print to the corresponding dis-

tance on the negative. Molecular weights were calculated from contour

40

length measurements assuming a mass to length ratio of 2.07 Mdal/um
(38).

Acetylene Reduction Assay on Rhizobium in Culture. Dinitrogenase
activity was assayed by measuring the production of ethylene from
acetylene (78). Strains to be tested for dinitrogenase activity in
culture were grown in YEM medium to late log phase. Three loops of the
culture were stabbed into 2.5 ml of CS7 medium solidified with 1% agar
in a three-dram vial. Three replicates of each strain were inoculated.
The vials were incubated at 30°C for 7 days. The screw caps were
removed aseptically and the vials were sealed with sterile serum caps.
Air (1 ml) was removed and 1 ml of acetylene (generated from calcium
carbide and water) was added. The vials were incubated another 2 days
at 30°C, at which time 0.2 ml samples of gas were removed. The gas
samples were analyzed for ethylene on a gas chromatograph (Varian Model
3700) equipped with a Porapak N column (2 mm x 2 m).

Growth of R. "cowpea" strain 32H1 at Elevated Temperature. .3;
"cowpea" strain 32H1 was grown in two - 100 ml cultures of YEM medium
at 30°C with shaking. When the cell density reached 30 Klett units
(about 2 x 108 cells/ml), one culture was shifted to 37°C and both
cultures were incubated for another 12 days. Suitable lO-fold serial
dilutions of the culture incubated at 37°C were plated on YEM agar
plates to follow cell viability. At the end of 12 days, suitable
lO-fold serial dilutions of heat-treated (37°C growth) culture were
plated on YEM agar. Fifty colonies were randomly chosen and tested for
the ability to fix nitrogen in culture.

Acridine Orange Treatment. Bacteria were grown in YEM medium for

3 days at 30°C. An aliquot of the culture was diluted to give a cell

41

density of about 5 x 106 cells/ml and 1 ml of the diluted culture was
transferred to flasks containing 50 ml of YEM medium supplemented with
0, 5, 10, or 20 ug/ml acridine orange. The cultures were incubated for
24 h at 30°C with shaking and then lO-fold serial dilutions were plated
on YEM agar. Plates were incubated at 30°C and single colonies were
randomly chosen. The isolates were tested for the ability to reduce
acetylene in culture and to nodulate the appropriate host plant.

Nodulation Test. The symbiotic properties of R. trifolii strain
T37 isolates were examined by inoculating white clover (Trifolium
532335 var. Ladino). Clover seeds were surface-sterilized by washing
with 70% ethanol for 30 sec followed by a 5 min treatment with 0.2%
HgClz acidified with 5 ml 12 N HCl per liter. The seeds were rinsed
eight times with sterile redistilled water and germinated in the dark
at 20°C on water agar plates. Two seedlings were transferred
aseptically to each 18 x 150 mm culture tube containing 9 ml of Jensen
medium agar (201). Alternatively, the surface-sterilized clover seeds
were germinated directly on the Jensen agar slant. Jensen medium con-
tained (per liter redistilled water): 1.0 g CaHP04, 0.2 g K2HP04,
0.2 g MgSO4°7H20, 0.2 9 NaCl and 0.1 g FeCl3. Bacteria to be I
used as an inoculum were grown to late log phase (1 to 2 x 109
cells/ml; 150-175 Klett units) in YEM medium. An aliquot (0.1 ml) of
the culture was pipeted directly onto the seedlings. The plants were
placed in a growth chamber at 20°C with a 16 h photoperiod. After 5
weeks, plants were scored for root nodules and were tested for acety-
lene reduction activity.

The symbiotic properties of isolates of R; "cowpea" strain 32H1

were examined on cowpea (Vigna unguiculata L). Cowpea seeds were

42

surface-sterilized as described for clover seeds and were germinated
directly in pots (10 cm) containing Perlite. Plants were inoculated
with 5 ml cultures of isolates of R; "cowpea" strain 32Hl treated with
acridine orange. The plants were grown in a greenhouse and were main-
tained on nitrogen-free nutrient solution (66). The photoperiod was
extended to 16 h with supplemental fluorescent lighting (200 uE/mz).
After 4 weeks, plants were examined for root nodules and were tested
for acetylene reduction activity.

Acetylene Reduction Assay on Nodulated Plants. Tubes containing
clover plants were sealed with serum caps and 0.1 volume (2 ml) of air
was removed and replaced with an equal volume of acetylene. After 15
min at room temperature, 0.2 ml samples were removed and analyzed for
ethylene on a gas chromatograph.

Cowpea plants were removed from the pots and the shoot was
excised. The roots were placed in a 250-ml Erlenmeyer flask. The
flask was sealed with a serum cap and 0.1 volume (25 ml) of air was
removed and replaced with an equal volume of acetylene. Gas samples
(0.5 ml) were taken after l5 min and analyzed for ethylene on a gas

chromatograph.

RESULTS AND DISCUSSION

Isolation of Plasmid DNA. Two procedures were used for the isola-
tion of plasmid DNA from Rhizobium and Agrobacterium: the Currier-
Nester procedure and the alkaline lysis procedure. The Currier-Nester
procedure employed a SDS-pronase lysis, followed by a brief shearing
step to reduce the viscosity and facilitate the denaturation of the
DNA. The alkaline lysis procedure used included several modifications
which increased the yield of plasmid DNA and also allowed the isolation
of larger plasmids. The bacteria were lysed non-enzymatically at high
pH, thus accomplishing the cell lysis and denaturation of the DNA in
one step. The lysis of cells with pronase (which may be contaminated
with nucleases) and the DNA shearing step found in the Currier-Nester
procedure were not utilized in the alkaline lysis procedure. The rate
of stirring during the cell lysis and subsequent steps was also care-
fully controlled at about 100 rpm to minimize shearing of plasmid DNA.
Using these procedures, plasmid DNA of 36 to about 325 Mdal could be
detected in the plasmid preparations after: 1) sucrose gradient cen-
trifugation; 2) agarose gel electrophoresis; 3) CsC1~EtBr density
equilibrium centrifugation; and 4) electron microscopy.

Sucrose Gradient Centrifugation. In alkaline environments, super-
coiled plasmid molecules collapse into condensed structures with sedi-
mentation coefficients 3 to 4 times those of similarly-sized linear DNA

(156). This allows plasmid DNA to be separated from chromosomal DNA on

43

44

alkaline sucrose gradients. Strains of Rhizobium were cultured in the
presence of [3H]thymidine and lysed using the alkaline lysis

procedure. Lysates were layered onto alkaline sucrose gradients and
the gradients centrifuged to separate plasmid DNA from chromosomal DNA.
Gradients were fractionated and the amount of radioactivity contained
in each fraction was determined. A peak of 3H—labeled plasmid DNA
(fractions 24-26) which sedimented faster than the chromosomal DNA was
observed in the sedimentation profile of an alkaline sucrose gradient

of R; leguminosarum strain 128C53 DNA (Figure 4). A similar peak of

 

radioactivity was not observed in alkaline lysates of 3H-labeled DNA
from R; "cowpea" strain 32H1 (Figure 4) or strain CB756 (data not
shown).

Plasmid DNA can also be separated from chromosomal DNA by centri-
fugation through neutral sucrose gradients. R;_trifolii strain T37 was
cultured in the presence of [3H]thymidine. The cells were lysed
using the alkaline lysis procedure and the lysate neutralized. The
lysate was layered on a linear 5 - 20% neutral sucrose gradient and
centrifuged. The sedimentation profile of a neutral lysate of
3H-1abeled R_. trifolii strain 137 is shown in Figure 5. The shoulder
of radioactivity in fractions 29-34 contained supercoiled plasmid DNA.

The detection of plasmids on sucrose gradients was hampered by the
low incorporation of label into bacterial DNA. In addition, the cell
lysates were briefly sheared in order to reduce the viscosity of the
lysates which probably resulted in the loss of some plasmid DNA.

The size and number of plasmids could not be readily determined
from the gradient profile of strains in which plasmids were detected.

A single peak of radioactivity corresponding to plasmid DNA was

45

Fi ure 4. Alkaline sucrose gradient sedimentation profile of [3H]thymi-
dine-labeled DNA from cells lysed using the alkaline lysis procedure.
Gradients were fractionated from the bottom, 100 01 fractions were
collected and radioactivity was determined. (0), R; leguminosarmn
strain 128C53; ([]), 3;."cowpea" strain 32H1.

46

 

 

 

 

 

 

la,
Cl? 1%
O:
m.
I
Ola/0,0, 0
r6
0
.4.
u .0
. 2
i q u q q 4 n a v .1 -
40 3m

238 £38853 I m

Number

Fracii on

47

F’ ure 5. Neutral sucrose gradient sedimentation profile of
I§Hlthymidine labeled cell lysate of R;_trifolii strain T37. Cells
were labeled as described in Methods and lysed using the alkaline lysis
procedure. The lysate was neutralized and layered on a linear 5 - 20%
neutral sucrose gradient. The gradient was centrifuged, fractions (l00

ul) were collected and radioactivity was determined as described in the
Methods.

48

 

 

‘ \OOO‘O
«so.
' 'o/
O
[0].,
O’clolo’
o, o, .0
0'
M‘O'.
I
.....
.P
’0
..
PI.
’0
N
- 1 I.
d u - q-«ddq. q - .quqdqd 1

40' ' "e‘e

r T V T I

Fraction Number

20

T I r I

 

 

em .0.
253 335835 In

49

detected in sucrose gradients of R; leguminosarum strain 128C58 (Figure
4) and R; trifolii strain T37 (Figure 5). .3; trifolii strain T37 con-
tains at least three plasmids (see Chapter II) as does R;_leguminosarum
strain 128C53 (Figure 6). Therefore, while sucrose gradient centrifu-
gation can be used to confirm the presence of plasmid DNA in Rhizobium,
the size and number of plasmids can not be determined.

Agarose Gel Electrophoresis of Plasmid DNA. Agarose gel electro-
phoresis was used successfully to identify two plasmids in crude
extracts of plasmid DNA from A; tumefaciens strain C58(RP4) using the
Currier-Nester plasmid isolation procedure (Figure 6, lane 1). Plasmid
DNA could not be detected in R;_"cowpea" strain 32H1 or strain CB756,
or in R;_trifolii strain T37 by agarose gel electrophoresis of plasmid
preparations from these strains using the Currier-Nester procedure.
Plasmid DNA from R; trifolii strain T37, however, was detected on
CsCl’EtBr gradients using this procedure.

A third, high molecular weight plasmid was detected in A;

tumefaciens strain C58(RP4) using the more gentle alkaline lysis proce-

 

dure. The estimated molecular weight of this plasmid, pAtC58, was
about 325 x 106 (Figure 6, lane 2). Freeze-thawing of the A;

tumefaciens strain C58(RP4) plasmid sample resulted in the loss of

 

pAtC58 (Figure 6, lane 3). Using the alkaline lysis procedure, two
large plasmids were detected in R; trifolii strain T37 and three plas-
mids were detected in R;_leguminosarum strain 128C53. As with the
Currier-Nester procedure, repeated attempts to detect or isolate plas-
mid DNA in the R; "cowpea" strains using the alkaline lysis procedure
were unsuccessful. This apparent absence of plasmids in the two 3;

"cowpea" strains may reflect the fact that these two strains do not

 

Figure 6. Agarose gel electrophoresis of plasmid DNA. Lanes 1-3,
A; tumefaciens strain C58(RP4); lane 4, 3; trifolii strain T37; lane 5,
R. Ie uminosarum strain 128C53. The plasmid DNA in lane 1 was isolated
E? t e urrier- ester procedure. Samples in lanes 2-5 were isolated
using the alkaline lysis procedure. Plasmid DNA in lane 3 was
subjected to several freeze-thaw cycles. Molecular weight estimates
were obtained from the relative mobility of plasmids of known size in
ggaroselgels: pBR3l3, 5.8 Mdal; RP4, 35 Mdal; pTiC58, 130 Mdal; pAtC58,
25 Mda .

51

contain plasmids. However, this does not seem likely, since Nuti gt
al. (137) have detected plasmid DNA in a strain of’R;_"cowpea". A more
likely possibility is that these strains contain large plasmids which
are not detected using the alkaline lysis procedure. Plasmids were not
detected in several strains of R; meliloti using this procedure (146).
Subsequent analysis of the R;_meliloti strains using the Ekhardt aga-
rose gel technique (discussed in Chapter II) resulted in the detection
of extremely large plasmids (>>300 Mdal).

CsCl'EtBr Density Equilibrium Centrifugation. Plasmid DNA was
separated from chromosomal DNA on CsCl'EtBr density equilibrium
gradients. Less intercalating dye is bound by intact, CCC plasmids
than by linear DNA, resulting in a smaller decrease in the density of
the supercoiled plasmid-dye complex (152). Supercoiled plasmid DNA was
isolated from R; leguminosarum strain 128C53 and A;_tumefaciens strain
C58(RP4) using the Currier-Nester procedure and the alkaline lysis pro-
cedure (Figure 7). A very faint CCC plasmid band was also observed for
fi;_trifolii strain T37 using the Currier-Nester procedure. A more
intense plasmid band reflecting a higher concentration of plasmid DNA
was observed with R; trifolii strain T37 plasmid preparations obtained
using the alkaline lysis procedure. Supercoiled plasmid DNA was
observed in CsC1°EtBr gradients of R; “cowpea” strain 32H1 and strain
CB756 DNA using either plasmid isolation procedure.

CsCl°EtBr density equilibrium centrifugation is useful both for
the detection of plasmid DNA in Rhizobium and for the purification of
plasmid DNA. Two major limitations are apparent with this technique.
First, no information is obtained regarding the size and number of

plasmids present. Only a more dense band corresponding to supercoiled

52

 

Figure 7. CsCl‘EtBr density equilibrium gradient of plasmid DNA
isolated by the alkaline lysis procedure. (A) R_. le uminosarum strain
128C53; (B) A. tumefaciens C58(RP4). The lower, fainter Band contains
supercoiled BTasmid DNA; the upper band contains nicked plasmid and
linear chromosomal DNA.

53

plasmid DNA is observed. Second, the detection of plasmids in strains
of Rhizobium using this technique is limited by the lysis procedure
used.

Electron Microscopy of Plasmid DNA. Plasmid DNA from CsCl°EtBr
density equilibrium gradients was examined by electron microscopy. An

open-circular (0C) molecule of RP4 from A; tumefaciens strain C58(RP4)

 

is shown in Figure 8. The molecular weight was calculated to be 35 i 2
Mdal (mean 1 3.0.) based on the contour length measurements of 14
molecules. This value agreed with the reported molecular weight of RP4
(41,52). Plasmid DNA molecules from CsCl'EtBr gradients of R;
leguminosarum strain 128C53 plasmid DNA preparations were found to have
a molecular weight of 119 1 8 Mdal (Figure 9). This corresponds to the
fastest migrating (smallest) plasmid of Rgnlgguminosarum strain 128C53
(Figure 6), which migrated about the same distance into the gel as the
125-130 Mdal plasmid, pTiC58. The large size of the other two 3;
leguminosarum strain 128C53 plasmids and of the two plasmids of R;
trifolii strain T37 would increase the probability of shearing of the
plasmid DNA during preparation of the samples for electron microsc0py.
Thus, circular molecules of these plasmids were not observed when these
samples were examined in the electron microscope.

Electron microscopy, like agarose gel electrophoresis can be used
to determine both the size and the number of plasmids present in a
strain. However, plasmids with molecular weight greater than 119 x
106 were not observed, probably due to shearing of the plasmid DNA
molecules during preparation of the samples.

Plasmid Curing Experiments. Concurrent with the development of

plasmid isolation techniques, experiments were performed with the

54

Fi ure 8. Electron micrograph of an open circular (0C) molecule of
R154. Plasmid DNA from A. tumefaciens strain C58(RP4) was isolated on a
CsCl-EtBr density equil'i'Eri'um gradient. The molecular weight was
calculated to be 35 t 2 Mdal based on an average contour length
measurement of 16.9 :1: 1 nM (14 molecules). Magnification: 2.64 x

10 -

55

 

56

Figure 9. Electron micrograph of an open circular plasmid molecule
from R; leguminosarum strain 128C53. Plasmids were purified by
CsCl‘EtBr density equilibrium centrifugation. The molecular weight
was calculated to be 119 1 8 Mdal from an average contour length
mggsurement of 57.7 t 3.3 um (8 molecules). Magnification: 1.28 x

57

 

58

intent of eliminating plasmids from strains of Rhizobium. The symbi-
otic properties (Nod+ Fix+) of isolates obtained from these experi-
ments were then examined in order to correlate the loss of these prop-
erties with the absence of a specific plasmid.

Elimination of plasmids from R; "cowpea" strain 32H1 was attempted
by growth at elevated temperature. The use of this technique has
resulted in the loss of plasmids from Staphylococcus (124), Proteus
(191), and several strains of Rhizobium (146,219). 3; "cowpea" strain
32H1 was cultured for two days at 30°C and then shifted to an elevated
temperature (37°C) for another 12 days of incubation. The viable cell
density decreased only slightly over the period of growth at 37°C,
although the turbidity (Klett units) increased for 4-5 days before
leveling off (Figure 10). At the end of the treatment period, an ali-
quot was diluted and plated on YEM agar, and 50 colonies were picked.

The ability to induce dinitrogenase activity in cultures of free-
living Rhizobium has been reported (110,193,144). This property was
examined in wild-type R;_"cowpea" strain 32H1 and in isolates grown for
12 days at 37°C. wild-type strain 32H1 showed substantial activity
after several days growth on CS7 medium (Figure 11). A loss of
dinitrogenase activity was observed in 4% of the 50 isolates obtained
from the culture grown at 37°C. All 50 isolates of strain 32H1
obtained prior to the temperature shift still retained dinitrogenase
activity. Since the existence of plasmids in R; "cowpea“ strain 32H1
could not be demonstrated, the loss of dinitrogenase activity in 4% of
the “heat-treated" isolates could not be correlated with the loss of a
specific plasmid. The loss of dinitrogenase activity was probably due

to a mutation and not to the elimination of a plasmid.

59

 

 

 

 

150‘ ’0/0 3
,0
1..
e 2’4 ‘5'
oCPCI “Cf-'13
3 100' agfi’g ;
3: [if’ _ 'E;
g / 90 o
A ’A‘A"A-b‘. '- .9
w , ‘1-A : .3
_ '5;
‘1 -ao .18
LP . . . .

4 a 12
Days after inoculation

Figure 10. Growth of R; "cowpea" strain 32H1 at elevated tem erature.
Two OO-ml YEM cultures were inoculated with approximately 10 cells
of R. "cowpea" strain 32H1. The cultures were grown at 30°C with
shaFTng until the cell density was about 2 x 10 cells/ml (30 Klett
units). On day 2, one flask (E3) was shifted to 37°C 0?). The flasks
were incubated another 12 days at 30°C (0) or 37°C ([3). Viablity of
cells grown at 37°C (A) was determined by plating suitable 10-fold
serial dilutions on YEM agar.

 

 

 

 

 

- D
:3
E150 -
8 O
O
100- -
:5; a
2‘
2;, 50‘ o/ -
_ o/
E E g_._ _, r s

3 5
Days after addition of acetylene

Figure 1 . Acetylene reduction assays of free-living Rhizobium grown
on CS7 medium. R. "cowpea" strain 32H1, (0); R. "cowpea" strain CB756,

(E1); and R. trifElii strain T37, (A).

61

Acridine dyes are effective in eliminating the F sex factor from

Escherichia coli (88), and extrachromosomal DNA from yeast (183) and

 

eukaryotic cells (21,186). The elimination is apparently due to inter-
ference by the dye in the replication of the plasmid (91,209). Non-
nodulating mutants of Rhizobium have been isolated following treatment
with acridine dyes (60,85,216). The Nod' phenotype was not demon-
strated to be due to the elimination of a plasmid, however.

Three strains of Rhizobium reportedly capable of nitrogen fixation
in a free-living state were treated with the intercalating dye acridine
orange. Cultures were grown for 24 h in the presence of various con-
centrations of acridine orange (Figure 12). Increasing concentrations
of acridine orange resulted in a decrease in the number of viable cells
present after 24 h of growth. In 20 ug/ml acridine orange, the cell
concentration decreased below the concentration after inoculation for
R; "cowpea" strain 32H1 and R; trifolii strain T37, indicating cell
death. Isolates from cultures treated with acridine orange were tested
for the loss of symbiotic properties. 3; “cowpea" strain 32H1 and
strain CB756 both reduced acetylene in culture on CS7 medium (Figure
11). Treatment of R; "cowpea" strain CB756 with 10 and 20 ug/ml acri-
dine orange resulted in a frequency of loss of dinitrogenase activity
similar to that observed for isolates not treated (Table 3). In con-
trast, 10 and 16% of the isolates of strain 32H1 treated with 10 and 20
ug/ml acridine orange, respectively, had lost the ability to reduce
acetylene (Fix‘) in culture. The Fix' isolates of strain 32H1
treated with acridine orange appeared larger and more mucoid than the

Fix+ colonies.

62

Figure 12. Treatment of strains of Rhizobium with various

concentrations of acridine orange. Cell densities (cells/ml) in flasks
cont ining acridine orange at the start of he 24 h incubation were: 8
x 10 , R. "cowpea" stragn 32H1, (0); 4 x 10 , R. "cowpea“ strain

CB756, TTJ); and 5 x 10 , R. trifolii strain T37, (A).

Viable Cells . ml’I

IO

5

63

 

1 111111

L

01

1 1 llfll

 

\A

 

lo 15 2'0
Acridine Orange Gag/ml)

 

64

Table 3. In_vitro nitrogen fixation by Rhizobium "cowpea" strains
grown in the presence of acridine orange.

 

 

 

Acridine Number of
Strain orange colonies % Fix“
ug/ml tested
0 185 3
32H1 10 88 10
20 150 16
O 91 2
CB756 10 158 4
20 99 2

 

Data were pooled from three experiments. Acetylene reduction assays
were carried out as described in Methods.

65

R; "cowpea" strain 32H1 isolates were also examined for the abili-
ty to nodulate cowpea. Isolates (5 FixT, 5 Fix') from both
untreated cultures and those containing 20 ug/ml acridine orange were
used to inoculate cowpea plants. All untreated Fix+ isolates formed

T isolates from

effective nodules on cowpeas, as did 4 of the 5 Fix
the culture treated with 20 ug/ml acridine orange. All five acridine
orange-treated Fix” isolates and four of the five untreated Fix‘
isolates were unable to nodulate cowpeas.

R;_"cowpea“ strain 32H1 is apparently more susceptible to acridine
orange than strain CB756, since a higher frequency of isolates with
mutant phenotypes were obtained upon treatment of strain 32H1 with
acridine orange as compared with acridine orange treated isolates of
strain CB756.

The Fix‘ isolates of strain 32H1 were generally also Nod",
suggesting that genes coding for nodulation and nitrogen fixation are
linked in this strain. Elimination of a plasmid from these isolates of
strain 32H1 could not be demonstrated. Therefore, the mutant phenotype
may result from a deletion of either plasmid or chromosomal DNA, and
may not be due to the elimination of a plasmid.

R;_trifolii strain T37 cultures were also assayed for dinitro-
genase activity, however, no activity was detected in cultures grown on
CS7 medium (Figure 11). Cultures grown on a modified medium in which
malic acid and glutamate were substituted for arabinose, succinic acid
and glutamine (193) also lacked dinitrogenase activity. Reduction of
the oxygen concentration in the vials to 1% failed to elicit dinitro-
genase activity. Therefore, in lieu of assaying dinitrbgenase activity

in culture, the ability of acridine orange-treated isolates of R;

66

trifolii strain T37 to nodulate clover was determined. All untreated
isolates nodulated clover plants within two weeks (Table 4). Half of
the isolates treated with 20 ug/ml acridine orange and 35% of the iso-
lates treated with 5 ug/ml did not nodulate clover plants two weeks
after inoculation. After four weeks, however, all of the plants inocu-
lated with acridine orange-treated isolates had formed effective
nodules. Uninoculated plants did not have nodules; thus, the delayed
nodulation was not due to contamination. Isolates exhibiting delayed
nodulation and those nodulating normally were examined for plasmid DNA
(Figure 13). No differences in the plasmid profiles were observed.
Thus, the delay in nodulation was probably caused by an acridine
orange-induced mutation and not as a result of the elimination of a
plasmid. Whether this mutation was located on a plasmid or on the

chromosome could not be determined.

67

Table 4. Nodulation of white clover by R; trifolii strain T37 grown in
the presence of acridine orange.

 

 

 

Acridine % colonies forming
orange Number of nodules at:
condentration colonies _—
ug/ml tested 2 weeks 4 weeks
0 24 100 100
5 20 65 100
20 4 50 100

 

Plants were grown in enclosed tubes on Jensen agar slants as described

in Materials and Methods. Uninoculated plants were not nodulated after
4 weeks. Nodules on all plants reduced acetylene.

68

 

Figure 13. Agarose gel electrophoresis of plasmid DNA isolated using
the alkaline lysis procedure from R; trifolii strain T37 isolates
treated with acridine orange. Lane 1, E; tumefaciens strain C58(RP4);
lanes 2 and 3, R. trifolii strain T37 isolates exhibiting normal

nodulation; lanE§ 4 and 5, strain T37 isolates exhibiting delayed
nodulation.

CHAPTER II

The Identification of a Rhizobium trifolii strain T37 Plasmid
Coding for Nitrogen Fixation and Nodulation Genes and its
Interaction with pJB5JI, a Rhizobium leguminosarum plasmid.

 

INTRODUCTION

The genetic and biochemical events involved in the establishment
and maintenance of the Rhizobium - legume symbiosis are not well under-
stood. Numerous reports, however, have indicated that many species of
Rhizobium contain large plasmids, with molecular size > 100 megadaltons
(Mdal), which encode symbiotic functions. Nitrogenase (Elf) structural
genes have been located on large plasmids in R; leguminosarum (137,
146), R;_meliloti (5,164), and R; phaseoli (14). Genes involved in
nodulation ability (Egg) have also been found to be plasmid encoded
(95,100,179,215), and are located on the same plasmid as the 31: genes
in R;_me1iloti (5,164), 3; leguminosarum (94,147) and R;_phaseoli (94).

Johnston gt al. (100) identified a plasmid coding for pea nodula-
tion in R; leguminosarum. This bacteriocinogenic plasmid transferred
at high frequency ( 10'2) into other Rhizobium and conferred the
ability to nodulate peas to a Nod‘ strain of R;_leguminosarum and to
heterologous species of Rhizobium. Similar results have been obtained
by Brewin £5 21. (23) who reported that a non-bacteriocinogenic plasmid
from a strain of R; 1eguminosarum transferred at low frequency ( 10'5)
and conferred pea nodulation ability.

Transfer of clover nodulation ability from a NodTFix+ R;
trifolii strain to Nod‘ strains has been achieved by transfer of a
cointegrate plasmid composed of a broad-host-range plasmid and the R;

trifolii nodulation plasmid (179). Transfer of a Tn §fcontaining

69

7O

derivative of a‘R; trifolii nodulation plasmid to other species of
Rhizobium and to Agrobacterium tumefaciens has also been reported (95).
Incompatibility between two Rhizobium plasmids has been used to identi—
fy a plasmid coding for symbiotic functions 1".3; phaseoli (14). Loss
of the ability of R; phaseoli to nodulate beans correlated with the
absence of a 190 Mdal plasmid in isolates into which a R;_leguminosarum
plasmid had been transferred.

Results presented in this chapter identify pRtT37a as the plasmid
in R. trifolii strain T37 which codes for the 21f genes. The loss of
pRtT37a or formation of "hybrid" plasmids of various molecular weights

from pRtT37a and pJBSJI, a.R; leguminosarum plasmid coding for pea nod-

 

ulation, was observed upon transfer of pJBSJI to R; trifolii strain
T37. Symbiotic properties and plasmid profiles of strains generated
upon transfer of R. leguminosarum plasmid pJBSJI to R; trifolii strain

T37 were examined.

MATERIALS AND METHODS

Materials. Tryptone, yeast extract and Bacto agar were purchased
from Difco Laboratories, Detroit, MI. Rifampicin, kanamycin sulfate,
streptomycin sulfate, lysozyme, RNase (Type 1A), DNase I, polyethylene
glycol (6000), pronase (Type XIV), ethidium bromide, Ficoll (400,000),
bovine serum albumin (BSA, Fraction V), and all vitamins were obtained
from Sigma Chemical Company, St. Louis, MO. Proteinase K was purchased
from Calbiochem, La Jolla, CA. DNA polymerase I, and restriction endo-
nucleases fligd III, §§m_Hl and Bgl II were purchased from Bethesda
Research Laboratories, Gaithersburg, MD. .539 I was purchased from P.
L. Biochemicals, Milwaukee, WI. .EEQ R1 was a generous gift from Dr.
A. Revzin, Michigan State University, East Lansing, MI. Polyvinyl-
pyrolidone was obtained from GAF Corporation, New York, NY.
[a-3ZPJLabeled deoxyadenosine triphosphate, deoxyguanosine triphos-
phate, deoxycytidine triphosphate, and thymidine triphosphate were pur-
chased from Amersham Corporation, Arlington Heights, IL (800 Ci/mmole)
or from New England Nuclear, Boston, MA (600 Ci/mmole). Agarose for
routine use was obtained from Sigma (Type VI) or Bethesda Research
Laboratories (gel electrophoresis grade). Agarose (low-mr) used for
the analysis of plasmid profiles by the Ekhardt agarose gel technique
was purchased from BioRad Laboratories, Richmond, CA. Low melting
temperature agarose (Sea plaque) was purchased from FMC Corporation,

Rockland, ME. All other chemicals were reagent grade or were obtained ‘

71

72

from sources described in Chapter I. Nitrocellulose (BA85, 0.45 um;
BA83, 0.2 pm) and a microfilter assembly were purchased from Schleicher
and Schuell, Keene, NH. GF/C glass fiber filters (24 mm) and 3 MM fil-
ter paper were obtained from Nhatman, Clinton, NJ. Filters (HANP; 0.45
um) to be used as a support for bacterial matings were purchased from
Millipore Corporation, Bedford, MA. X-Ray film was purchased from
Kodak, Rochester, NY. Dupont Cronex Lightning Plus intensifying
screens were purchased from Picker Corporation, Detroit, MI.

Bacterial Strains, Plasmids, and Phagi. The bacterial strains and
plasmids used in this work are listed in Table 5 and 6, respectively.
The nomenclature employed for designation of naturally occurring
Rhizobium plasmids conformed to the format recommended by Novick gt al.
(136). '3; trifolii plasmids are designated by pRt followed by the
strain number and serial letters in cases of multiple plasmids:
pRtT37a, pRt0403c.

Bacteriophage A::Tn§_was obtained from B. Chelm, Michigan State
University, East Lansing, MI.

.!£212° YEM and YE2 media were described in Chapter I. TY medium
(10) contained: 5.0 g tryptone, 3.0 g yeast extract; and 0.8 g CaClz
per liter redistilled water. The minimal medium for growth of
Rhizobium was RM medium (119). RM medium consisted of : 10.0 g manni-
tol, 0.22 g KZHP04, and 1.1 g sodium glutamate per liter redis-
tilled water (pH 6.8). After autoclaving, 1.0 ml each of the vitamins
and minor salts stock solutions was added per liter of medium. The
vitamin stock solution consisted of a filter-sterilized solution of 10
mg thiamine-HCl and 20 mg biotin in 100 m1 of 50 mM potassium phos-
phate, pH 7.0. The minor salts stock solution contained: 0.04 g A

73

Table 5. Bacterial strains.

 

Bacteria Relevant Properties Source

 

Rhizobium trifolii

T37 w.t., Nod+Fix+ on clover (a)
0403 w.t., NodTFix+ on clover (a)
6001-6140 T37 (pJBSJI) transconjugants; (kanr) (b)
7001-7051 0403 (pJBSJI) transconjugants; (kanr) (b)

Rhizobium leguminosarum

 

T83K3 phe trp rifr strr NodTFix+ on pea (c)
contains pJBSJI (kanr)
726 .gifir NodTFix+ on pea (c)

Agrobacterium tumefaciens
C58(RP4) contains RP4, pTiC58, pAtC58 (d)
Pseudomonas aeruginosa

PA02 contains pMGl (e)
PA02 contains pMG5 (e)

Escherichia coli

E08654 gl_e_t_ ggl hsd.< 3" _M"’ su E sllpF A::Tn_5_ host (f)
AB2880 contains pSA30 (tet (g)
HBlOl contains pRmR2 (tetr) (9)

 

(a)F.B. Dazzo, Michigan State University, East Lansing, MI
(b)This work

(c)A.N.B. Johnston, John Innes Institute, Norwich, England
(d)T.C. Currier, Kansas State University, Manhattan, KS
(e)R.H. Olsen, University of Michigan, Ann Arbor, MI

(f)B. Chelm, Michigan State University, East Lansing, MI
(g)F.M. Ausubel, Massachusetts General Hospital, Boston, MA

74

Table 6. Plasmids

 

 

 

 

 

Plasmids Relevant Properties Reference
pJBSJI encodes R; leguminosarum nif genes, pea
‘ggg genes; contains Tn.§ (kanr); 130 Mdal (100)
pSA30 Klebsiella pneumoniae nif genes; 10 kb (34)
pRmR2 R;_meliloti nif genes; 8 kb (166)
RP4 kanr tetr ampr; 35 Mdal (41)
pTiC58 tumor-inducing plasmid; 130 Mdal (41)
pAtC58 325 Mdal (35)
pMGl 312 Mdal (75)

pMGS 280 Mdal (75)

 

75

CaClz, 0.033 g FeCl3'6H20, 0.1 g MgSO4°7H20, and 0.83 ml 12 N HCl in
99 m1 redistilled water.

§;_ggli and Pseudomonas aeruginosa strains were grown at 37°C on
LB medium (130) composed of: 10 g tryptone, 5 g yeast extract, and 5 9
NaCl per liter redistilled water. For the propagation of A::Tn§, LB
medium was supplemented with l M MgSO4'7H20 to a final concentra-
tion of 10 mM.

Media were solidified with 1.5% Bacto agar. Soft LB agar con-
sisted of LB medium plus 0.7% Bacto agar.

Plasmid Transfer. Bacterial matings were carried out essentially
as described by Beringer 33 31. (13). Bacteria were grown for three
days on TY medium agar slants. The bacteria were washed from the
slants with 2.5 m1 sterile 10 mM MgSO4 and matings were performed by
mixing 0.5 m1 (about 5 x 108 cells) of donor and recipient cells.

The mixed cell suspensions were pipetted onto 0.45 um Millipore filters
on non-selective TY medium agar plates. The plates were incubated for
24 h at 30°C. Filters were transferred to 3-dram vials and cells were
resuspended in 2.5 ml sterile 10 mM MgSO4. Appropriate lO-fold

serial dilutions in 10 mM MgSO4 were plated on selective medium.
Transfer frequencies were calculated per recipient cell plated. Trans-
conjugant colonies were restreaked on selective medium to obtain single
colony isolates.

Ekhardt Agarose Gel Electrophoresis Technique. Plasmid profiles
of donor, recipient and transconjugant strains were obtained using a
modification of the Ekhardt agarose gel electrophoresis technique (63).
Bacteria were grown without shaking in 5 ml YEZ medium for 48 h at

30°C. Cells (107 - 108) were centrifuged, washed with 0.1% sarkosyl

76

in TE buffer and resuspended in 20 ul of modified Ekhardt gram-negative
bacteria lysis buffer. The buffer contained 2.0 mg lysozyme, 5.0 mg
bromphenol blue, 2.0 mg Ficoll (400,000) and a lO-fold greater amount
of RNase (0.5 mg) in 10 ml of TBE electrophoresis buffer. Immediately
after resuspension, cells were carefully layered beneath 150 pl of
Ekhardt SDS solution (0.2% SDS, 10% Ficoll in TBE buffer) in a well (3
mm x 6 mm) of a 3 mm thick vertical 0.7% agarose gel. Plasmid DNA was
separated by electrophdresis for l h at 8 mA followed by 10 - 12 h at
20 mA. Alternatively plasmid DNA was rapidly separated by electro-
phoresis for l h at 8 mA followed by 3 h at 40 mA. The sizes of the
plasmids were determined from their relative mobilities in agarose gels
(129) using RP4 (35 Mdal), pTiC58 (l30 Mdal), pJBSJI (l30 Mdal), pMGS
(280 Mdal), pMGl (312 Mdal), and pAtCSB (325 Mdal) as size markers.

Nodulation Tests. The symbiotic properties of strains of

 

Rhizobium were tested on white clover (Trifolium repens var. Ladino)

and pea (Pisum sativum var. Wisconsin Perfection). Clover nodulation

 

tests were carried out in enclosed tubes as described in Chapter 1.

Pea nodulation tests were carried out on partially enclosed plants
(201). Pea seeds were surface sterilized as described for clover seeds
and germinated on water agar plates. Pea seedlings were transferred to
sterile, cotton-plugged, 25 x 200 mm culture tubes containing Vermicu-
lite wetted with 30 m1 of Fahraeus medium containing:. 0.2 g CaClz,
0.22 g MgS04'7H20, 0.1 g KH2P04, 0.15 g NaZHP04, 0.005 9

iron citrate and 1 ml BMM medium trace element stock solution (see
Chapter 3) per liter redistilled water (pH 6.5). Pea seedlings were
inoculated as described for clover seedlings and the plants were grown

under the same conditions. After approximately one week, the shoot of

77

the pea plant was gently pulled past the cotton plug. Additional
sterile Fahraeus medium was added three weeks after inoculation. After
five weeks, plants were scored for root nodules and were tested for
acetylene reduction ability.

Acetylene Reduction Assay. Dinitrogenase activity was assayed by
measuring the amount of ethylene produced from the reduction of acety-
lene (78). -Acetylene reduction assays on pea plants were carried out
essentially as described for clover and cowpea plants (Chapter 1). The
shoot of the pea plant was excised and the culture tubes were sealed
with serum caps. Air (5.0 ml) was removed and replaced with an equal
volume of acetylene and samples (0.2 ml) were taken after 15 min incu-
bation at room temperature. Samples were analyzed for ethylene by gas
chromatography as described in Chapter I.

Isolation of Bacteria from Root Nodules. Rhizobium were isolated
from clover and pea root nodules essentially as described by Vincent
(201). Nodules were clipped from the plant with a small portion of the
root still attached. The detached nodules were placed in sterile 3-
dram vials and washed several times with 10 mM M9504. The nodules
were surface sterilized by covering with 95% ethanol for 30 seconds,
and then treating with 5% (v/v) hydrogen peroxide (300 01) for 2 min.
The hydrogen peroxide solution was replaced with 150 pl sterile 10 mM
M9504 and the nodules were thoroughly crushed with the flattened end
of a sterile glass rod. The nodule suspension was spread onto the sur-
face of YEM agar plates (containing antibiotic, if desired) and then
incubated at 30°C for 4-5 days. Bacteria were restreaked to obtain

single colony isolates.

78

Fahraeus Slide Technique. The infection of clover root hairs by
.3; trifolii bacteria was studied using a simple glass slide procedure
(65). Bacteria were grown for 3 days on TY agar plates. Cells were
aseptically washed from the plates with l M NaCl, 10 mM EDTA and
pelleted by centrifugation. The bacteria were washed with sterile
Fahraeus medium and resuspended to a density of about 10 Klett units
(108 cells/ml) in sterile Fahraeus medium. An aliquot (0.4 ml) of
the cell suspension was pipetted onto one end of a sterile microscope
slide (25 mm x 75 mm). Two l-day-old clover seedlings were placed on
the microscope slide with roots (1 cm in length) projecting into the
cell suspension. A sterile cover slip (25 mm x 40 mm) was carefully
placed over the clover roots leaving the clover hypocotyl uncovered.
The microsc0pe slide was placed in a sterile covered tube containing 7
m1 of Fahraeus medium. The tubes were incubated in a growth chamber as
previously described for clover plants. The clover roots were examined
with a phase contrast microscope at 24 h intervals.

Isolation of Total DNA. Starter cultures of bacteria were grown
in 10 ml YEZ medium for 3 days. An aliquot (2 ml) was used to inocu-
late ZOO-ml cultures of YE2 medium. After growth for 24 h, the cells
were pelleted by centrifugation, washed with 0.1% (w/v) Sarkosyl in TE
buffer, and resuspended in 2.5 m1 STE buffer. Lysis was carried out by
the addition of 2 ml “sucrose mix“ solution (1.6 M sucrose, 0.55 M
Tris, pH 8.0, 0.1 M EDTA), 1 ml 5 mg/ml lysozyme, and 22 ml 10 mM EDTA
(177). The lysate was incubated on ice for 20 min at which time 15 ml
2.5% (w/v) Sarkosyl was added. The lysate was incubated for 20 min on
ice, extracted twice with phenol, reextracted twice with chloroform,

and digested for 2 h at 37°C with 20 ug/ml RNase (1 mg/ml in 0.4 M

79

sodium acetate, pH 4.0; boiled 10 min). The RNase-digested lysate was
reextracted with phenol and chloroform. The aqueous phase was
recovered and the DNA was ethanol precipitated. The DNA was pelleted
by centrifugation and resuspended in TE buffer. The DNA concentration
was determined by the diphenylamine assay.

Determination of DNA Concentration. The concentration of DNA was
determined by absorbance at 260 nm or by the diphenylamine method (70).
DNA samples (500 pl) containing 10% (v/v) perchloric acid were mixed
with an equal volume of 4% (w/v) diphenylamine in glacial acetic acid.
An aliquot (25 pl) of aqueous acetaldehyde (1.6 mg/ml) was added and
the samples were incubated for 10 - 12 h at 37°C. The DNA concentra-
tions in the samples were determined by measuring the absorbance at 595
nm. The assay was linear between 0 - 20 pg calf thymus DNA. The con-
centrations of small amounts of DNA ((250 ng/ml) were estimated by the
comparison of the intensity of fluorescence of the EtBr-stained unknown
DNA with that of EtBr-stained DNA standards in agarose gels (121). The
unknown DNA sample and a series of standard DNA solutions were electro-
phoresed on an agarose minigel. The gel was stained with ethidium bro-
mide and photographed. The quantity of DNA in the unknown sample was
estimated by eye from the gel photo.

‘Agarose Gel Electrophoresis. DNA samples were separated electro-
phoretically on vertical agarose gels as described in Chapter 1.
Horizontal agarose gels (12.5 cm x 18 cm x 1 cm) were electrophoresed
at 60 volts for 15 h. Small amounts of DNA (20 - 50 ng) could be
analyzed rapidly by electrophoresis on a mini agarose gel apparatus

(121). Electrophoresis of DNA samples on a mini agarose gel (8.5 cm x

80

6.5 cm x 0.3 cm) was carried out at 100 volts for 45-60 min. Gels were
stained and photographed as described previously (Chapter I).

Growth of Bacteriophage A::Tn 5. §;_ggli strain E08654 was used
as a host for bacteriophage 1::Tn.§. Overnight cultures of strain
E08654 grown on LB medium were centrifuged at 3000 x g for 10 min at
4°C. The pellet of bacteria was resuspended in 0.4 volume sterile 10
mM MgSO4. This cell suspension was used to inoculate broth cultures
and to determine bacteriophage titers. Titers were determined by
mixing 0.1 ml of ten-fold serial dilutions of phage solution with 0.1
ml of the cell suspension. The bacteria/phage mixture was incubated at
37°C for 20 min, then combined with 3 ml soft LB agar at 45°C and
plated on LB agar plates. After 12 h at 37°C, the plates were scored
for plaques.

A plate lysate stock solution of bacteriophage 1::Tn §_was pre-
pared by plating approximately 105 plaque-forming units (pfu) with
0.1 ml of the bacterial cell suspension, as described above. Phage
from the confluent lysis were recovered by adding 10 ml SM phage buffer
(10 mM Tris, pH 7.4, 100 mM NaCl, 10 mM MgC12, 0.05% gelatin) to the
plate and storing at 4°C for 12 h. The SM buffer containing the phage
was decanted from the plate and 0.1 ml chloroform was added. After
brief vortexing, the SM buffer was centrifuged at 4000 x g for 10 min
at 4°C. The supernatant fluid was recovered, and chloroform was added
to a final concentration of 0.3%. The titer of the plate lysate stock
solution of A::Tn§ was determined to be 5 x 109 pfu/ml.

Large scale growth of phage 1::Tn§ was carried out as described
(18,210). An overnight culture (about 109 cells/m1) of strain E08654

in 10 ml LB medium was supplemented with 1 M MgS04 to a final

81

concentration of 10 mM MgSO4. Phage plate lysate stock solution (0.2
ml) was added and the mixture was incubated at 37°C for 20 min. The
phage/bacteria mixture was transferred to a 500-ml culture of LB medium
+ 10 mM M9504 equilibrated to 37°C. The culture was incubated at

37°C with shaking for 10 h, at which time lysis was evident. Chloro-
form (5 ml) was added to the lysate and the flask was shaken 15 min.
Cellular debris was removed by centrifugation at 3000 x g for 30 min at
4°C. The supernatant fluid (titer = 8x1010 pfu/ml) was decanted

into a 1-liter flask and digested with DNase (1 pg/ml) and RNase (1
pg/ml) for l h at 4°C. PEG 6000 (50 g) and NaCl (29.22 g) were added
and the solution was stirred for 15 min. The solution was allowed to
stand for 2 h and then centrifuged at 3000 x g for 40 min at 4°C. The
phage pellet was resuspended in 2.5 ml TM buffer and was mixed until
well dispersed. The phage suspension was extracted with an equal
volume of chloroform to remove the PEG, and the aqueous phase contain-
ing the phage was recovered. The phage solution was layered on a 5% to
40% glycerol step gradient (197) and centrifuged at 35,000 rpm for 60
min at 4°C in a Beckman SN41 rotor. The pellet of phage particles was
resuspended in 1 ml TM buffer. The phage was further purified on a
CsCl step gradient (;= 1.7 g/ml, 1.5 g/ml, 1.45 g/ml). The gradient
was centrifuged in a Beckman SN41 rotor at 37,000 rpm for 2 h at 4°C.
The gradient was fractionated and the absorbance at 260 nm was used to
determine which fractions contained phage.

Isolation of DNA from Phage Particles. Cesium chloride gradient
fractions containing phage were pooled and dialyzed for 2 h at room
temperature against 0.1 M Tris (pH 8.0), 0.15 M NaCl, 1 mM MgC12 (2
liter). DNA was extracted from the phage particles by digestion at

82

68°C for l h in a solution containing 50 pg/ml Proteinase K, 0.1% SDS,
and 10 mM EDTA. The phage DNA solution was extracted twice with an
equal volume of phenol/chloroform (1:1,v/v), twice with chloroform, and
once with ether. The phage DNA was ethanol precipitated and resus-
pended in 1 ml TE buffer. From two 500-m1 cultures, 720 pg of 1::Tn.§
DNA was obtained. 1
Cleared-Lysate Procedure for Plasmid Isolation. The small ampli-
fiable plasmids, pSA30 and pRmR2, were isolated using the cleared-
lysate technique (37). A lO-ml culture of LB medium containing the
appropriate antibiotic was inoculated with the E; ggli strain harboring
either pSA30 or pRmR2. The culture was incubated overnight at 37°C
with shaking and was used to inoculate a 1-1iter culture of LB medium +
antibiotic. The culture was grown at 37°C until the 00500 was
between 0.4 and 0.6. Chloramphenicol (150 mg) was added to amplify the
plasmid (36) and the incubation was continued for 15-18 h. Bacterial
cells were harvested by centrifugation at 4000 x g for 10 min at 4°C.
The cells were washed with 20 ml 25% (w/v) sucrose in TE buffer, and
resuspended in 50 ml 25% sucrose in TE buffer. Lysozyme (7 m1 of a 5
mg/ml solution in 0.25 M Tris, pH 8.0) was added and the solution was
incubated on ice. After five min, 15 ml 0.25 M EDTA was added. The
solution was mixed gently and incubated on ice for 5 min. Cells were
lysed by the addition of 8 ml 10% (w/v) SDS and the solutions were
mixed by gentle inversion. The cleared-lysate was adjusted to 1 M NaCl
with 20 m1 5 M NaCl. The solution was incubated on ice for 90 min and
was centrifuged at 30,000 x g for 30 min at 4°C. The supernatant fluid
containing plasmid DNA was decanted, extracted with an equal volume of

phenol (equilibrated with TE buffer) and reextracted with chloroform.

83

The aqueous phase was recovered and DNA was precipitated by the addi-
tion of 0.1 volume 3 M sodium acetate and 2 volumes 95% ethanol. The
DNA was pelleted by centrifugation at 12,000 x g for 20 min at 4°C.
The DNA pellet was resuspended in 1 ml TE buffer and was further puri-
fied by CsCl°EtBr density equilibrium centrifugation as described in
Chapter 1.

Isolation of DNA Fragments from Agarose Gels. DNA fragments were
isolated from agarose gels by electr0phoresis onto dialysis membranes
(71,121). An incision was made in the agarose gel directly in front of
the desired DNA band. A piece of Whatman 3 MM filter paper and a piece
of dialysis membrane were inserted into the incision with the filter
paper between the dialysis membrane and the DNA band. Electrophoresis
was carried out until the DNA had migrated into the 3 MM paper. The
dialysis membrane and the 3 MM paper were inserted into a 500-pl micro-
centrifuge tube with a hole in the bottom. The tube was inserted into
a 1.5-ml microcentrifuge tube and centrifuged for 15 sec to force the
buffer containing the DNA fragment into the 1.5-ml tube. The filter
paper and dialysis membrane were washed three times with 100 pl elution
buffer (0.2 M NaCl, 50 mM Tris, pH 7.6, 1 mM EDTA, and 0.1% $05). The
eluates were pooled and extracted with equal volumes of phenol,
phenol/chloroform (1:1, v/v) and chloroform. The DNA was ethanol pre-
cipitated and pelleted by centrifugation. The DNA was resuspended in
20 to 50 pl TE buffer and contaminating traces of agarose were removed
by filtration of the DNA solution through a BA83 nitrocellulose filter
using a microfilter apparatus.

Alternatively, DNA fragments were isolated from low-melting tem-

perature agarose (Sea Plaque) gels (121,204). The segment of gel

84

containing the desired DNA fragment was cut out of the gel and was
melted at 68°C in 5 volumes TE buffer. The melted agarose was
extracted with equal volumes of phenol, phenol/chloroform, and chloro-
form. The DNA was recovered by ethanol precipitation and traces of
agarose were removed as described above.

Preparation of nif Probe DNA. The 3.4 kb Egg RI-Hjfld III fragment
of plasmid pSA30 containing the region of K. pneumoniae nif DNA homolo-
gous to the gifi structural genes of other Nz-fixing bacteria or the
2.0 kb Egg_RI-thbl fragment of pRmR2 containing R;_meliloti 31f was
used as a probe for Rhizobium nif DNA (Figure 14). Plasmid DNA was

 

isolated from bacteria by the cleared-lysate technique and purified by
ethidium bromide-cesium chloride centrifugation. Plasmid DNA was
digested with restriction endonucleases according to manufacturer's
specifications and the resulting fragments were separated by agarose
gel electrophoresis. The desired fragments were recovered from the gel
and 1 pg amounts were 32P-labeled by nick translation.

Isolation ofprtT37a andngBSJI. Crude extracts of plasmid were
isolated from R; trifolii strain T37 and R; leguminosarum strain T83K3
using the alkaline lysis procedure (Chapter I). Plasmid DNA was
separated from chromosomal DNA and other plasmids by electrophoresis on
vertical 0.7% low-melting-temperature agarose gels. Segments of the
gel containing pRtT37a and pJBSJI were excised and the DNA was
recovered as described above. The DNA was 32P-labeled by nick
translation and used as a hybridization probe.

32P-Labeling of DNA by Nick Translation. DNA was labeled _i_n
11339 with 32P-labeled deoxynucleotide triphosphates (dNTPs) by the
nick translation method (122,158). Labeled dNTPs (100 pCi) were

85

 

 

 

R H B R
A. pSA30 , . L .
I I
R X X X R
B. pRmR2 1 l 1 I
L____J
2kb
l_______l

Figure 14. Partial restriction map of gjj_DNA cloned in pSA30 and
pRmR2. A, endonuclease cleavage sites for Egg RI (R), fligd III (H),
and Bag H1 (B) are given for K. pneumoniae nif DNA cloned in pSA30
(34). 8, Eco RI (R) and Xho T'(X) cleavage—ETtes are indicated for R;

meliloti gEfTDNA cloned ifi‘ihmaz (166). The region of gjf_homology is
indicated by the open bar.

86

pipetted into a 500-p1 Eppendorf tube. DNA (1 ug) and 2.5 pl 10X nick
translation buffer (0.5 M Tris, pH 7.5, 0.1 M MgS04, 10 mM DTT, 500
pg/ml BSA) were mixed with the dNTPs. The volume of the reaction mix-
ture was adjusted to 23 pl with redistilled water. DNase I (1 pl of a
104-fold dilution in redistilled water of 1 mg/ml stock solution) and
1 pl of §;_gglj DNA polymerase (l unit/pl) were added. The reaction
was incubated for 2 h at 16°C. The reaction mixture was brought to
room temperature and 2.5 pl 5 M NaCl, 2.5 p1 0.5 M EDTA, 1 pl 10% SDS,
and 2.5 pl 5 mg/ml proteinase K in redistilled water were added. The
reaction was incubated at 68°C for 30 min. Single stranded salmon
sperm DNA (100 p1 of 10 mg/ml redistilled water), yeast tRNA (100 pl of
1 mg/ml redistilled water) and 0.1 volume 3 M sodium acetate were
added. The DNA was precipitated with the addition of 2 volumes of
ethanol to separate the labeled DNA from the unincorporated nucleo-
tides. Alternatively, unincorporated nucleotides were removed by the
"spun column“ procedure (121). A Sephadex G-50 column (0.9 ml bed
volume) equilibrated with STE buffer was prepared in a disposable plas-
tic syringe (1 ml). The nick translation reaction mixture was brought
to 100 pl with STE buffer, and centrifuged through the column. The
effluent was collected in an Eppendorf tube and contained the labeled
DNA. Unincorporated nucleotides remained in the column. An aliquot (1
p1) of the labeled DNA solution was spotted on a GF/C filter and radio-
activity was determined by liquid scintillation spectrometry. Specific
activities of 5 to 6 x 107 dpm/pg DNA were routinely obtained.

Southern Hybridization. Transfer of DNA from agarose gels to
nitrocellulose filters was carried out essentially as described by

Southern (185). Prior to transfer, DNA was partially depurinated by

87

treatment with 300 ml 0.25 M HCl for 15 min (202), denatured for 30 min
in 300 ml 0.3 M NaOH, 1.5 M NaCl, and neutralized for 60-90 min in 300
ml 0.5 M Tris (pH 7.0), 3 M NaCl. DNA transfer was carried out for 48
h using 6X SSC (1X SSC = 0.15 M NaCl, 0.015 M sodium citrate). Filters
were rinsed with 6X SSC, air dried and vacuum-baked for 2 h at 80°C.
Filters were pretreated in sealed plastic pouches for at least 6 h at
68°C with hybridization buffer: 6X SSC, 0.5% SDS, 1 mM EDTA, 2X
Denhardts solution (51), and 100 pg/ml denatured salmon sperm DNA. The
pretreatment solution was removed and fresh hybridization buffer con-
taining denatured 32P-labeled probe DNA was added. The pouches

were resealed and incubated at 68°C for 24 h. The filters were washed
at 68°C in 300 m1 2X SSC, 0.5% (w/v) SDS (3 changes). The filters were
rinsed a final time in 2X SSC at room temperature and then air dried.
The filters washed under stringent conditions were treated with 300 ml
0.1x SSC, 0.5% $05 (300 m1) at 68°C (3 changes), rinsed with 300 ml
0.1X SSC and air dried. Autoradiography was carried out at -80°C for
3-5 days using Kodak X-Omat AR film with a Dupont Cronex Lightning Plus

intensifying screen.

RESULTS

5; trifolii strain T37 and strain 0403 were analyzed for plasmid
DNA using the Ekhardt agarose gel technique. At least three large
plasmids were present in strain T37. Four large plasmids were detected
in strain 0403.

The molecular weights of these plasmids were estimated from the
relative mobilities of the plasmids in agarose gels. The relative
mobilities of plasmids of known molecular weight were determined and a
standard curve was constructed (Figure 15) (129). Two linear regions
were observed in the curve: from 36 x 106 to 130 x 106 and from
280 x 106 to 325 x 105. The molecular weights of plasmids greater
than 325 x 106 were estimated by extrapolation of the curve. The
molecular weight of the smallest plasmid in R; trifolii strain T37,
pRtT37a, was about 275 x 106, while pRtT37b and pRtT37c had molecular
weights of about 300 x 106 and 380 x 105, respectively. The molec-
ular weight of pRt0403a was about 290 x 106 and the molecular weights
of pRt0403b, c, and d were estimated to be 335 x 105, 350 x 105,
and 390 x 105, respectively.

Transfer Frequency of pJBSJI. The frequencies of transfer of
Tn Q-encoded kanamycin resistance located on pJB5JI from R;
leguminosarum strain T83K3 to three strains of Rhizobium were deter-
mined (Table 7). Comparable frequencies of transfer of pJBSJI into R;
trifolii strain T37 and strain 0403 were observed. Transfer of pJB5JI

88

89

Po mcoPapoa eechP mzP

maPoPPPeee osPoePae .mmoe<a .. “Page .. "mesa ._- “mmoPPe .o “Pawnee .a mammomNPPaa .nu mememmPPea .. m a

21.1.1.1 .m>e:o on» we meoPuePoaoeuxo mrllll .a>e:u we»
.mxoeem on» Pa umPeuPucP mew moeo cPeeum new PmP :Peeum PPPoPPeu .m.Po mePEmaPa Po

 

  

.x .umm mmoemmm :P chEmqu Po PPPPPnoE m>Paume moP mamsm> <zo cPEmoPa Po usumz eszuwPo: .mP we: Pm

90

Pesos. 2:22

no...

 

 

N. 3 O... m..0
{X
. Q0
x. n _
/
39.0 ms
.333»... .N
x / eBHoea
/, as; s
“Pl/awelh: “r 1
as» e .. 1 1
nets... 4 . w
ewes...

 

 

(8_o|x) iubgeMuolnoelow

91

Table 7. Transfer of pJBSJI from R; leguminosarum strain T83K3

 

 

Recipient Strain Frequency

8.: trifolii strain 137 2.5 x 10-5
_l_2_._ trifolii strain 0403 2.4 x 10-5
R;_leguminosarum strain 726 5 x 10'3

 

 

Transconjugants were selected on RM medium containing 200 pg/ml
kanamycin. The transfer frequency was determined per recipient cell.
The frequency of Spontaneous resistance of all three recipient strains
to 200 pg/ml kanamycin was < 10'3. The frequency of reversion to
auxotrophy of R;_leguminosarum strain T83K3 was < 10‘3.

92

into R; leguminosarum strain 726 occurred at a frequency ZOO-fold
greater than the frequencies observed for the R;_trifolii strains. The

frequency of transfer of pJBSJI into R;_leguminosarum strain 726 was

 

similar to frequencies of transfer of pJBSJI into other species of
Rhizobium (100). R; leguminosarum strain T83K3 contains a second
transmissible plasmid (100 Mdal) which transfers at a frequency of
about 10'6 (101). This plasmid is one of two plasmids in the

fastest migrating band of the donor strain (23) and was observed to
cotransfer with pJBSJI into R; trifolii strain T37 and strain 0403 at a
frequency of 7 to 8 x 10'6 per recipient cell.

Characterization of R. trifolii Strain 0403 (pJBSJI) Transconju-
gaLts. Fifty kanamycin-resistant colonies from crosses between R;
leguminosarum strain T83K3 and R; trifolii strain 0403 were restreaked
on selective medium and a single colony from each plate was selected.
The strains (numbered consecutively from 7001-7050) were analyzed for
plasmid DNA using the Ekhardt agarose gel electrophoresis technique.

The four plasmids indigenous to R; trifolii strain 0403 were pre-
sent in all of the transconjugants and no alterations in the electro-
phoretic mobilities of these plasmids were observed (Figure 16). All
of the transconjugants (except strain 7004) contained pJB5JI and 14
strains also contained the 100 Mdal plasmid (pR1T83K3a) of the donor
strain. Strain 7004 was kanamycin-resistant yet lacked a plasmid band
correspoding to pJBSJI. However, the intensity of fluorescence of the
100 Mdal plasmid in agarose gels of strain 7004 was increased, suggest-
ing that approximately 30 Mdal of DNA had been deleted from pJBSJI
(Figure 16, lane 4).

93

1234

 

Fi ure l6. Ekhardt agarose gel electrophoresis of plasmid DNA of _R_.
trifolii strain 0403 (pJBSJI) transconjugants. (1), R. trifolii strain
0403;12), strain 7001, harboring pJB5JI; (3), strain-70M, harboring
pJB5JI and pRlT83K3a; (4), strain 7004, harboring pRlT83K3a and
plasmid- -deletion mutant of pJBSJI.

94

The symbiotic properties of 14 of the transconjugants were exam-
ined on pea and clover plants. Thirteen of the transconjugants con-
tained pJBSJI in addition to the four indigenous plasmids of R;
trifolii strain 0403. Three of these strains also harbored pRlT83K3a.
The other strain (7004) contained a derivative of pJBSJI in which about
30 Mdal of plasmid DNA was apparently deleted. Southern hybridization
analysis using a heterologous-git probe demonstrated that the nif
structural genes were also located on this region of deleted DNA
(Figure 17, lane 4).

All of the transconjugants formed effective nodules on clover
plants. Nodulation of the clover plants was not delayed when compared
to R;_trifolii strain 0403, and occurred within two weeks. The number
of nodules produced on the plant roots were also similar.

All of the transconjugants (except strain 7004) formed nodules on
pea plants. Fewer nodules were produced, however, and the nodules were
small, white and unable to reduce acetylene (Fix‘). No nodules were
produced on pea plants inoculated with strain 7004, suggesting that a
gene(s) coding for pea nodulation ability were located on the 30 Mdal
segment of pJB5JI DNA that was deleted.

Plasmid Analysis of R. trifolii Strain T37 (pJBSJI) Transconju-
.HEEEE- Initially, 40 kanamycin-resistant transconjugants from crosses
between R; leguminosarum strain T83K3 and R; trifolii strain T37 were
analyzed for plasmid DNA as described for the R; trifolii strain 0403
(pJB5JI) transconjugants. In contrast to the low number of different
plasmid profiles obtained upon transfer of pJBSJI to strain 0403, a
wide variety of plasmid profiles were observed for the strain T37

(pJB5JI) transconjugants (Figure 18). To further investigate the

95

1111104031— I. - p. .-
plBSll— -

a

1.

Figure 17. Autoradiogram of 32P-labeled pRmR2 to a Southern filter
of plasmid DNA from R; trifolii strain 0403 (pJB5JI) transconjugants.
Plasmid DNA from an agarose gel similar to the one shown in Figure 15
gas transferred to a nitrocellulose filter and hybridized to .
P-labeled R; meliloti gif DNA. Lane 1, R; trifolii strain 0403;
2, strain 7001, harboring pJBSJI; 3, strain 7014, harboring pJB5JI and
pRlT83K3a; 4, strain 7004, harboring pRlT83K3a and a deletion
derivative of pJBSJI.

96

 

Figure 18. Ekhardt agarose gel electrophoresis of plasmid DNA of R;
trifolii strain T37 (pJBSJI) transconjugants. (1), R;_le uminosarum
strain T83K3; (2), R. trifolii strain T37; (3-10), strain l37 lpJBSJI)
transconjugant strains: 6032, 6060, 6012, 6076, 6067, 6130, 6135,

6137. The plasmid pJB5JI is noted with the symbol (0) and pRtT37a,
pRtT37b, and pRtT37c are marked with the symbol (>).

97

occurrence of the different plasmid profiles, another 100 transconju-
gants were selected, restreaked and analyzed for plasmid DNA (Table 8).

The two largest plasmids of R; trifolii strain T37, pRtT37b and
pRtT37c were present in all transconjugants with no observable differ-
ences in electrophoretic mobility in agarose gels.

Approximately 30% of the transconjugants (Class 1) contained a 130
Mdal plasmid corresponding in size to pJBSJI (Figure 18, lane 3).

These transconjugant strains lacked a plasmid band corresponding to
pRtT37a. Another 40% of the transconjugant strains (Class II) lacked a
plasmid corresponding to pJBSJI. These transconjugants contained plas-
mids ranging in size from 140 Mdal to approximately 270 Mdal (Figure
18, lanes 4-7). The remaining 30% of the transconjugants (Class III)
did not contain plasmids corresponding to pJBSJI or pRtT37a nor were
plasmids of molecular weight intermediate to pJBSJI and pRtT37a
observed (Figure 18, lane 8). Four transconjugants (Class IV) were
obtained which harbored two plasmids with molecular weights intermedi-
ate to that of pRtT37a and pJBSJI (Figure 18, lanes 9 and 10). In
addition, the smallest plasmid of the donor strain, pRlT83K3a, randomly
cotransferred with pJBSJI into approximately 33% of the transconjugants
of each class. No additional plasmid rearrangements were observed in
the isolates harboring this plasmid.

Symbiotic Properties of R. trifolii Strain T37 (pJB5JI) Transcon-
jugants. The symbiotic properties of strains from each class were
examined by inoculation of white clover plants, the normal 3;_trifolii
host, and pea plants, the host plant specified by pJB5JI-encoded genes.
Differences in Nz-fixation (Fix) and nodulation abilities (Nod) of

the transconjugant stains were evident upon examination of both plant

98

 

 

 

 

 

Table 8. Diagrammatic representation of plasmids of strain T37
(pJBSJI) transconjugants after agarose gel electrophoresis.
Class Class Class Class
I II III IV
---a -..- -.... -.... --- --- --- -....
6016 6024* 6012 6045 6001 6002 6030 6088
6017* 6041 6063 6003 6005* 6135* 6137
6018* 6046 6076* 6004 6006*
6022* 6060 6084 6020* 6007
6027* 6074* 6106 6025* 6008*
6032 6090* 6047 6009
6038 6109* 6049* 6010*
6042* 6111 6051* 6014
6043* 6131* 6058* 6015
6044* 6133 6059* 6021
6048* 6134 6062* 6023
6052* 6064* 6026
6054 6065* 6028
6055 6066* 6033*
6069 6067 6034
6070 6075 6036
6073* 6088* 6037*
6078 6092 6039*
6080 6093 6057*
6082* 6096 6061
6086 6101 6068*
6087* 6102 6071
6091 6103 6072
6095 6104* 6077
6097* 6107* 6079
6099* 6108* 6081
6100* 6110 6083
6105 6113 6085
6117 6115 6089*
6126* 6116 6094
6132 6119* 6098*
6138* 6120 6112
6121 6114*
6122* 6118
6124 6123
6127* 6125
6128 6130*
6129 6136*
6140

 

 

aDashed lines represent pRlT83K3a which was present

designated with an asterisk.

only in strains

99

hosts (Table 9). Transconjugant strains in Class I were unable to nod-
ulate clover plants while all the transconjugants of Class II and III
(except strain 6023) formed effective (Fix+) nodules on clover roots.
Nodulation of clover plants by the transconjugant strains was not
delayed when compared to the wild-type 3;_trifolii strain T37.

All transconjugants strains tested from the three classes nodu-
lated peas (Table 9). All transconjugants in Class I and 5 out of 6
transconjugants in Class III formed pink, effective (FixT) nodules on
pea plants. Nodules produced from Class II transconjugants were white
and were ineffective (Fix‘).

Bacteria were isolated from the root nodules of pea and clover
plants and analyzed for plasmid DNA. The nodule isolates were found to
have the same plasmid profiles as the transconjugant strains used as
inocula (Figure 19). In addition, bacteria isolated from pea root nod-
ules were used to inoculate clover plants. No change in the expected
phenotype was observed. Similar results were obtained from clover nod-
ule isolates used to inoculate pea plants.

Thus, the plasmid alterations occurring upon transfer of pJBSJI to
“R. trifolii strain T37 were stable, even after passage of the transcon-
jugant strains through root nodules on both host plants.

Fahraeus Slide Analysis of Infection Process. The infection pro-
cess of three Nod' transconjugant strains on white clover root hairs
was examined microscopically using the Fahraeus slide technique (65).
The three Nod' transconjugant strains were: Class I strains 6032 and
6017, and Class III strain 6023. R;_trifolii strain T37 and
R;_leguminosarum strain T83K3 were used as positive and negative

controls, respectively.

100

Table 9. Nodulation and nitrogen fixation abilities of the four
classes of transconjugant strains on white clover and peas.

 

 

 

 

 

 

Nhite Clover Pea
Class No. Nod+a No. Fix+b No. Nod+ No. Fix+
No. tested No. Nod+ No. tested No. Nod+
Class I 0/10 - 6/6 6/6
Class II 10/10 10/10 6/6 0/6
Class III 9/10C 9/9 6/6 5/6d
Class IV 3/3 3/3 3/3 3/3

 

Plants were grown as described in Materials and Methods and were
examined for nodules and tested for acetylene reduction ability after
five weeks. No nodules were observed on peas inoculated with R;
trifolii strain T37, clover plants inoculated with E; leguminosarum
strain 183K3, or uninoculated plants.

aThe number of transconjugants which formed nodules on clover plants
out of the number tested.

The fraction of transconjugants which formed nodules that actively
reduced acetylene.

cR. trifolii strain 6023 did not nodulate clover but formed
TTectlve nodules on peas.

Strain 6130 formed ineffective nodules on peas and effective
nodules on clover.

101

plBSJl‘—‘
pR1183K31———

 

 

Figure 19. Ekhardt agarose gel electrophoresis of bacteria isolated
from pea nodules. Lane 1, Class III strain 6130; 2, Class II strain
6049; 3, Class 11 strain 6131; Class I strain 6080; 4, R; leguminosarum
strain T83K3. ~

102

Bacterial cells from all of the cultures used to inoculate white
clover plants bound to the tips of clover root hairs (Figure 20, c and
d). Slight curling or deformation and some branching of root hairs
were observed in plants inoculated with transconjugant strains 6032 and

6017, and R; leguminosarum strain T83K3. In general, though, the root

 

hairs were relatively unaffected (Figure 20,c).

In contrast, marked curling of clover root hairs was observed on
plants inoculated with the Nod' transconjugant strain 6023 and with
.3; trifolii strain T37 (Figure 20, a and b). Infection threads were
observed to have formed four days after inoculation of the clover
plants with R; trifolii strain T37 (Figure 21). Even though root hairs
of plants inoculated with strain 6023 were markedly curled, no infec-
tion threads were observed. This suggests that DNA coding for an
essential step in the formation of infection threads has been deleted
in strain 6023.

The response observed for plants inoculated with strain 6032 and
6017 was very similar to that obtained when R;_leguminosarum strain
T83K3 was used as the inoculum. The absence of marked curling of root
hairs and of the development of infection threads in plants inoculated
with these two transconjugant strains indicated that DNA encoding genes
essential for these processes had been deleted from these strains.

Tn 5 Hybridization Analysis. The absence of a plasmid correspond-
ing in size to pJBSJI in the Class II and Class III transconjugant
strains indicated a possible loss of pJB5JI DNA. The selection was
based on resistance to kanamycin, however, which suggests that Tn §_may
have transposed from pJBSJI to a new site in the R. trifolii genome.

Transposition of TnԤ would result in new restriction endonucleases

103

Figure 20. Fahraeus slide analysis of root hairs of clover plants
inoculated with strains of Rhizobium. (a) R. trifolii strain T37; (b),

Class III strain 6023; (c) Class I strain 6032; (d) R; leguminosarum
strain T83K3.

 

 

104

 

105

Figure 21. Infection thread in a root hair of a clover plant
inoculated with R1 trifolii strain T37.

 

 

 

 

106

 

107

sites in the sequences flanking Tn.§ DNA. To investigate this, the
flanking regions of Tn §_DNA were characterized for several transconju-
gant strains. A partial restriction map of Tn‘é DNA is shown in Figure
22. Tn.§ does not contain an Egg_RI restriction endonuclease cleavage
site. Thus, Egg RI-digested total DNA from the transconjugant strains
should contain only one Egg_RI fragment that will hybridize to
32P-labeled Tn‘g probe. If Tnig has transposed, the fragment con-
taining Tn.§ would be different in molecular weight. A single hybridi-
zation band of approximately 11 kilobases (kb) was observed in an auto-
radiogram of a Southern filter of Egg RI-digested total DNA from donor,
recipient, and several transconjugant strains probed with 32P-labeled
X::Tn §_DNA (Figure 23). No hybridization was observed to the recipi-
ent strain DNA (Figure 23, lane 7). To further confirm that Tn.§ had
not transposed from pJB5JI DNA, total DNA from the transconjugants was
digested with Bgl II restriction endonuclease and analyzed by Southern
hybridization. Tn.§ is cleaved twice by Bgl II, once each in the
inverted terminal repeats of the transposon. Thus, three Bgl II frag-
ments should hybridize to the Tnig probe. Hybridization to 9.6, 2.65,
and 2.5 kb fragments in the Bgl II-digested DNA from the donor and
several transconjugant strains was observed (Figure 23).

The hybridization data above indicated that Tn.§ had not trans-
posed from pJBSJI. Since pJBSJI transfers at high frequency and no
conjugative plasmids have been observed in R;_trifolii strain T37, the
correlation of the transfer of kanamycin resistance with the transfer
of a specific plasmid from the strain T37 (pJBSJI) transconjugants to
R;_lgguminosarum strain 726 would indicate which plasmid contained Tn§,

and thus pJBSJI DNA. Therefore, the transfer of kanamycin resistance

108

 

 

Tn.§
Hind 111 891 11 Bam HI BglII Hind 111
d . Pb
kanr
1kb
l_.—..—-..J

Figure 22. Partial restriction map of Tn 5. Heavy lines indicate
inverted repeats. Coding region for kanamycin resistance is indicated
by the arrow. Data were obtained from Jorgensen gt 31. (102).

109

a 11
1231511 12:15

(it)
2:-—
u—"“"' “"

5.5-
1.1—

P..-
. . ,
‘ C

Figure 23. Autoradiogram of 32P-label ed X::Tn ,5 DNA hybridized to
restrictlon endonuclease-digested DNA. Panel a, Egg RI-digested DNA
from: (1), R;_trifolii strain 6002; (2), strain 6032, (3), strain
6067; (4), strain 6007; (5), 3; le uminosarum strain T83K3; (6), strain
6012; (7), R. trifolii strain T37. Panel b, B l II-digested DNA from:
(1), R. trifolii 6032; (2), R. trifolii 6093; 3 , R;_leguminosarum

T83K3E—(4l, R; trifolii 60233—(51, E; trifolii 6012.

110

from a number of strain T37 (pJBSJI) transconjugants to
R; leguminosarum strain 726 was determined. Plasmid profiles of the
resulting transconjugants were also analyzed.

The transfer frequencies varied from the normal value observed for
pJBSJI of approximately 10"3 for the transconjugant strains of
Class I, to very low values (10‘6 to 10'7) observed for some
strains (Table 10).

Several transconjugants from each cross were restreaked and ana-
lyzed for plasmid DNA (Figure 24). The 130 Mdal plasmid corresponding

to pJB5JI transferred to R; leguminosarum strain 726 from Class I

 

strains 6032 and 6069. The fluorescence of the fourth smallest plasmid
(the smallest plasmid is a doublet as in R; leguminosarum strain 183KB)
of transconjugants from the latter crosses was more intense in
EtBr-stained Ekhardt agarose gels than the corresponding plasmid in
gels of transconjugants obtained from other crosses (Figure 24, lane
5). This suggested that a plasmid of the same molecular weight (about
300 x 105) had transferred from the Class III strains into R;
leguminosarum strain 726.

The presence of Tn‘g on these plasmids was confirmed by hybridiza-
tion of 32P-labeled 1::Tn.§ DNA to a Southern filter of an Ekhardt
gel similar to the one shown in Figure 18. The Tn.§ probe hybridized
to pJBSJI in the donor strain and to a plasmid of the same molecular
weight in Class I transconjugants (Figure 25; lanes a,c). In Class II
transconjugants, Tn.§ hybridized to the plasmid ranging in molecular
size between those of pBJSJI and pRtT37b (lanes d-g). The Tn.§ probe
hybridized to plasmid DNA in the region of pRtT37b in Class III trans-

conjugants (lane h).

111

Table 10. Frequency of transfer of kanamycin resistance from R.
trifolii strain T37 (pJB5JI) transconjugants to R. leguminosarum strain
726

 

Kanamwcin Resistance

 

 

 

Class Donor Strain Transfer Frequency
5032 9.0 x 10-4
5055 2.1 x 10‘
5070 2.9 x 10'
I
5017* 1.3 x 10-3
5022* 7.2 x 10-4
5044* 3.1 x 10'-3
5012 1.3 x lo- -5
5050 1.2 x 10-7
5057 3.1 x 10-3
5075 1.4 x 10-3
5092 4.1 x 10-4
11
5049* 8.1 x 10-5
5059* 2.3 x 10-3
5054* 2.2 x 10-4
5074* 3.8 x 10-7
5109* 4.0 x 10-4
5002 8 10-4
5007 0 10-4
5023 4 10-4
111
5005* 9 2 10-6
5033* l 3 10-7
5114* 4.7 x 10-5

 

Transconjugants were selected on 40 pg/ml rifampicin and 200 pg/ml
kanamycin. The transfer frequency was determined per recipient cell
plated. The frequency of spontaneous resistance of the donor strains
to 40 pg/ml rifampicin and of the reciBient strain to 200 pg/ml
kanamycin was 1. x 10' and < 6 x 10' , respectively.

*These strains also contain pRlT83K3a.

112

 

Figure 24. Ekhardt agarose gel electrophoresis of isolates of
bacterial crosses of R_. trifolii strain T37 (pJB5JI) transconjugants
with .13.: 1e uminosarum strain 72 . Lane 1, R_. leguminosarum strain 726
(recipientl; lanes 2-5, transconjugants from crosses with Class I
strain 6032, Class II strain 6060, Class II strain 6067, and Class III
strain 6023, respectively. The plasmid transferred is indicated by >.

113

e
'71:
.11131c_ a
HUI...
filma— F. .

 

W~ Autoradiogram of 32P-labeled X::Tn §_ DNA hybridized to
a Southern filter of an Ekhardt agarose gel. Lane a, R.

1e uminosarum
strain T83K3; (b), 3;_trifolii strain T37; (c), Class T—strain 6032,
(d-g), Class 11 strains 6060, 6012, 6076, 6067, respectively; (h),

Class III strain 6130. The positions of certain plasmids are noted.

114

Tn_§ was also located on a specific plasmid in the Class IV trans-
conjugants strains (Figure 26). The Tn §_probe hybridized to the
smallest of the two plasmids of molecular weight intermediate to the
molecular weights of pJBSJI and pRtT37a in strains 6135 and 6137.

nif Hybridization Analysis. Plasmids containing the 31: structur-
al genes were also identified in the donor, recipient, and transconju-
gant strains. Southern blot analysis of an Ekhardt agarose gel probed
with 32P-labeled R_. meliloti m DNA showed hybridization with
pJB5JI and pRtT37a (Figure 27), lanes a and b). The plasmids in Class
I, II, and III transconjugant strains which hybridized with Tn.§ also
hybridized with the gif_DNA probe (Figure 26, lanes c-h). The 51:
probe hybridized with the higher molecular weight "hybrid” plasmid in
Class IV transconjugant strains 6135 and 6137 (Figure 26).

The R;_meliloti g1: DNA is homologous to both R;_1eguminosarum and

R;_trifolii nif DNA. Thus, it was not possible to distinguish between

 

the presence in the transconjugants of 31: DNA from only one strain or
both R; trifolii and 3;.leguminosarum. To resolve this, total DNA from
.3; trifolii strain T37 and R; leguminosarum strain T83K3 was digested
with restriction enzymes and examined for the presence of the Eli
structural genes by Southern blot analysis. The R;_trifolii strain T37
31: structural genes were located on a 5.0 kb Egg RI fragment (Figure
28, lane e), while the 31: genes (located on pJB5JI) in R; legumino-
m strain T83K3 were present on a 2.65 kb _E_c_:9_ RI fragment (Figure
28, lane a). This gave an easy method of distinguishing between the
‘31: DNA of R; trifolii strain T37 and R;_leguminosarum strain T83K3 and
therefore, which g1: DNA was present in the transconjugant strains.

Egg_RI-digested DNA from transconjugant strains was examined in the

115

 

Figure 26. Hybridization analysis of plasmids of Class IV
transconjugant strains 6135 and 6137. Lane 1, EtBr-stained plasmid DNA
in strain 6137; Lanes 2 and 3, Southern filtggs of the gel in lane 1
probed with 2P-labeled X::Tn _5 (lane 2) or P-labeled R;

meliloti _rljj DNA (lane 3). Lane 4, EtBr-stained plasmid DNA in strain
6 5; lanes 5 and 6, Southern filterg of the gel in lane 4 probed with
3 P-labeled XzzTn g DNA (lane 5) or zP-labeled 3; meliloti _n_ij

DNA (lane 6).

116

abodeigh

 

Figure 27. Autoradiogram of 32P-labeled R_. meliloti fl. DNA
hybridized to Southern filters of an Ekhardt agarose gel similar to
Figure 13. Lane a, & 1e uminosarum strain T83K3; (b), _R_. trifolii
strain T37; (c), Class I strain 6032; (d-g), Class 11 strains 6060,
6012, 6076, 6067, respectively; (h), Class III strain 6130. The
mobilities of the three R. trifolii strain T37 plasmids and pJBSJI
are indicated at the rigfit and left of the autoradiogram. (0), pJB5JI;
(>), pRtT37a; (A) transconjugant plasmids encoding if DNA.

117

a b c d 13

(kb)

 

Figure 28. Autoradiogram of 32P-labeled K; pneumoniae nif DNA
hybridized to Southern blots of Eco RI-digested DNA from Rhizobium

strains. (a), 3;_leguminosarumԤ??hin T83K3; (b), Class I strain 6032;
(c), Class 11 strain 6012; (d), Class III strain 6007; (e), R;_trifolii
strain T37.

 

 

118

same manner. Three patterns of hybridization of the g1: probe to £29
RI-digested DNA from transconjugants were observed. The 21: probe
hybridized only to a 2.65 kb Egg RI fragment of DNA from Class I
transconjugants (Figure 28, lane b). This fragment corresponded to the

.3; leguminosarum nif-containing DNA fragment. In Class II transconju-

 

gants, only a 5.0 kb fragment corresponding to the R; trifolii strain
T37 g1: DNA hybridized (Figure 28, lane c), whereas, both fragments
hybridized to the Eli probe in Class III strains (Figure 28, lane d).
Similar analysis of DNA from the Class IV strains 6088, 6135, and 6137
indicated that both R;_leguminosarum and R; trifolii 21: DNA was pre-
sent as noted for Class III strains (data not shown). Since only one
plasmid in these strains hybridized to the g1: DNA probe, both the R;
trifolii and the R; leguminosarum g1: DNA must be on the same "hybrid"
plasmid. In strains 6088 and 6137, this plasmid did not hybridize to
the Tn.§ probe, a further indication that recombination between pJB5JI
and pRtT37a had occurred.

pJBSJI and pRt737a Hybridization Studies. In order to determine
the amount of pJBSJI and pRtT37a still present in the transconjugant
strains, the plasmids were isolated, labeled 15.!itrg, and hybridized
to Southern transfers of Egg RI-digested total DNA from several strains
from each class of transconjugants.

Initially, pJB5JI was isolated by the alkaline lysis procedure and
purified on CsCl-EtBr density equilibrium gradients. Plasmid DNA puri-
fied in this manner would be free of chromosomal DNA, but would also
contain pRlT83K3a/b DNA and, to a lesser extent, the higher molecular
weight plasmids of strain T83K3. The presence of pRlT83K3 a/b DNA

119

would not affect the hybridization results because total DNA was iso-
lated from transconjugant strains which did not contain pRlT83K3a.

Hybridization of pJBSJI DNA purified by CsCl-EtBr gradient centri-
fugation to a Southern filter of Egg RI-digested total DNA from several
transconjugant strains is shown in Figure 29. The hybridization pat-
tern was complex due to the large number of restriction fragments which
are homologous to pJBSJI. Basically the same pattern of hybridization
was obserVed in all of the strains examined. Several differences were
detected, however. Hybridization to 4.3 kb and 6.0 kb Egg_RI restric-
tion fragments was not observed in Class I transconjugant DNA but was
observed to these fragments in the DNA of Class II strains (Figure 29).
This indicates that some pJB5JI sequences have been deleted from Class
I strains.

No major differences in the hybridization pattern were observed
for Class II transconjugants which contained "hybrid" plasmids of vari-
ous molecular weights. This indicated that a similar amount of pJBSJI
DNA was present in these transconjugant strains. Hybridization of
32P-labeled pJBSJI to §_c_o_ RI-digested DNA from 3_._ leguminosarum
strain T83K3 was not carried out due to the presence of pRlT83K3a/b in
both the probe and in strain 183K3 DNA. This would have allowed the
determination of the number of Egg_RI restriction fragments in pJBSJI
and thus whether some pJBSJI sequences had been deleted from the
strains examined.

To address this problem, and to determine the amount of pRtT37a
which is present in the transconjugant strains, hybridization to Egg_RI
digested total DNA from transconjugant strains was carried out with

pJBSJI and pRtT37a isolated from low-melting-temperature agarose gels

120

123458

 

 

Figure 2 . Autoradiogram of 32P-labeled pJB5JI DNA hybridized to a
Southern filter of Egg RI-digested total DNA from transconjugant
strains. The pJBSJI DNA was isgéated by CsCl'EtBr density
equilibrium centrifugation and P-labeled 1g vitro by the nick
translation method. Lane 1, Class I strain 6032; 2, Class I strain
6069, 3-6, Class II strains 6084, 6060, 6012, 6093, respectively.

121

as described in the Methods. Recovery of plasmid DNA from agarose gels
allowed isolation of a specific plasmid, unlike CsCl-EtBr gradients
where total, supercoiled plasmid DNA was isolated.

The results of hybridization of 32P-labeled pJBSJI to Egg
RI-digested DNA from donor, recipient and transconjugant DNA are shown
in Figure 30. In general, less hybridization was observed to DNA from
the transconjugants than to R; leguminosarum T83K3 (positive control).
Differences in the pattern of hybridization of pJB5JI to DNA from
Classes 1, II, and III were also observed. Thus, varying amounts of
pJBSJI appear to have been deleted during the recombination event with
pRtT37a. No hybridization was observed to DNA from R; trifolii strain
T37 under these exposure conditions. However, exposure of the film to
the filter for about 7 days allowed detection of Egg RI restriction
fragments of R; trifolii strain T37 which were homologous to pJBSJI
(Figure 31).

Hybridization of 32P-labeled pRtT37a to Southern filters of
Egg RI-digested transconjugant DNA was carried out in a similar manner
(Figure 32). A large number of fragments in the transconjugant strains
were homologous to pRtT37a, but the amount of pRtT37a present in the
strains could not be determined quantitatively due to the complexity of
the hybridization pattern and the high background. To further compli-
cate the interpretation of the results, pRtT37a hybridized to R;
leguminosarum strain T83K3 DNA (Figure 32, lane 1). These sequences
are quite homologous, since the filters were washed under stringent
conditions (0.1X SSC, 68°C). Due to the complexity of pRtT37a probe,
long exposures of the film to the Southern filters were necessary for

the detection of restriction fragments of pRt137a. These long

122

l 2 3 4 5 6 1 I I

 

Figure 30. Autoradiogram of 32P-labeled pJBSJI DNA to a Southern
filter of £29 RI-digested total DNA from donor, recipient and
transconjugants. The BJBSJI DNA was isolated from low-melting-tempera-
ture agarose gels and 2P-labeled jg vitro by the nick translation
method. Lane 1, R. 1e uminosarum strain T83K3 (donor); (2), Class I
strain 6032; (3-6): Class II strains 6060, 6012, 6076, and 6067,
respectively; (7-8), Class III strains 6002 and 6023, respectively;
(9), R. trifolii strain T37 (recipient).

 

123

‘5

21"
IJ"

 

Figure 31. Autoradiogram of 32P-labeled pJBSJI DNA hybridized to a
Southern filter of Ec_o RI-digested DNA from donor and recipient
strains. A, R_._ leguminosarum strain T83K3 (5 pg); B, R_._ trifolii
strain T37 (5 pg). Film was exposed for 12 h (A) or 7 days (B).

124

123456789

(kb)

23— g

9.5— F

5.11—

1.3— ,
1'; z

2 2— “9'

2.1— ,. e

1.35—' (I
311

Fi ure 32. Autoradiogram of 32P-labeled pRtT37a DNA hybridized to
a Soutfiern filter of Egg RI- -digested DNA from donor, recipient and

transconjugant strains. Lane 1, R. leguminosarum strain T83K3 (donor);
(2), Class I strain 6032; (3- 7), CTass 11 strains 6060, 6012, 606

6086, and 6093, respectively; (8), Class III strain 6023; (9), R.
trifolii strain T37 (recipient).

125

exposures, however, allowed detection of the fragments of pJBSJI which
are homologous to pRtT37a. Thus, the varying amounts of pRtT37a pre-
sent in the transconjugant strains could not be determined.

DNA from Class I transconjugant strain 6032 which contained a
plasmid of molecular weight similar to that of pJBSJI, hybridized with
32P-labeled pRtT37a to a greater extent than the DNA in the donor
strain (Figure 32, lanes 1 and 2). Thus, the plasmid in Class I
strains which corresponded in molecular weight to pJBSJI contained some

DNA from pRtT37a.

DISCUSSION

At least three large plasmids were present in 3; trifolii strain
T37; four large plasmids were observed in Rg_trifolii strain 0403.
Transfer of the conjugative R; leguminosarum plasmid pJBSJI into 3;
trifolii strain 0403 did not alter the plasmid profile of the four
resident plasmids of this strain as determined by Ekhardt agarose gel
electrophoresis. In contrast, dramatic changes in the plasmid profile
were evident upon transfer of pJB5JI into R; trifolii strain T37. The
observed changes in the plasmid profiles of the strain T37 (pJB5JI)
transconjugants apparently resulted from the recombination between
pJB5JI and pRtT37a. The hybrid plasmids which resulted from this
recombination event contained variable amounts of pJBSJI and pRtT37a
based on an analysis of the symbiotic properties of the transconjugants
on both clover and peas and on hybridization studies using
32P-labeled X::TnԤ and heterologous g1: DNA.

Johnston 35 gl. (100) reported that transconjugants resulting from
the transfer of pJB5JI into heterologous species of Rhizobium retained
the ability to nodulate their normal host, although fewer nodules were
formed and they were slower to appear. These transconjugants also nod-
ulated pea plants but the nodules produced were not always effective.
The presence of genetic information necessary for the nodulation of a
second host plant in a Rhizobium strain was suggested to impair the

nodulation of the normal host plant.

126

127

The symbiotic properties of R; trifolii strain 0403 (pJBSJI)
transconjugants on the normal host plant, white clover, were not
altered. The host range of these transconjugants was extended, how-
ever, to include pea plants as a result of pJBSJI transfer. The root
nodules induced on peas by these transconjugants were unable to reduce
acetylene (Fix‘) which was similar to results obtained for other
strains of Rhizobium into which pJBSJI had been transferred (100). A
derivative of pJBSJI with a 30 Mdal deletion was also observed in one
transconjugant. This transconjugant was unable to nodulate peas sug-
gesting that the 30 Mdal region of deleted DNA contained gene(s) essen-
tial for pea nodulation. Southern hybridization analysis demonstrated
that the R; leguminosarum nif structural genes had also been deleted
(Figure 17). Spontaneous deletion of pJB5JI DNA resulting in the loss
of pea nodulation ability has been observed previously in pJBSJI trans-
fers (56).

The R; trifolii strain T37 (pJBSJI) transconjugants obtained in
this study varied in their symbiotic properties on the two host plants,
pea and white clover. Class I transconjugants were missing a plasmid
corresponding in size to pRtT37a and did not nodulate white clover
plants. Little curling or deformation of root hairs of clover plants
inoculated with two Class I transconjugant strains (strain 6032 and
6017) was observed (Figure 20), suggesting that gene(s) encoding the
curling inducer(s) had been deleted during the recombination event. In
contrast, strain 6023, a Nod' Class III strain, induced "marked"
curling of root hairs (Figure 20). Class 11, Class III, and Class IV
transconjugants contained "hybrid" plasmids and formed effective nod-

ules on clover. No delay in the nodulation of clover plants by Class.

128

II and III transconjugants as compared to wild-type strain T37 was ob-
served. The R; trifolii g1: genes were missing in Class I strains but
were present in Classes II, III, and IV transconjugants. These find-
ings suggest that pRtT37a codes for genes essential for nodulation of
clover as well as the Rg_trifolii nif genes and that the hybrid plas-
mids in strains of all four classes of transconjugants contain some of
pRtT37a.

All four classes of R; trifolii T37 (pJBSJI) transconjugants nodu-
lated peas although nodulation was delayed several days and fewer nod-
ules were formed than the number induced by R;_leguminosarum T83K3, the
pJBSJI donor strain. All three classes contained g1: genes but only
Class I, Class III, and Class IV strains (which contained 3; legumino-
Igggggngif genes) formed effective nodules on peas. Nodules induced by
Class II transconjugants which lacked the R; leguminosarum nif genes
were white and ineffective. Although the R; trifolii g1: genes were
present in these transconjugants and were functional in clover nodules,
they did not complement the deleted 3; leguminosarum nif genes in pea
nodules.

There are several possible explanations for this apparent anomaly.
The biochemical signal necessary for activating the expression of the
Rg_leguminosarum nif genes in pea nodules may be different from that
required for expression of the R; trifolii g1: genes in clover nodules.
If true, the Rg_trifolii.gif genes may not be expressed in pea nodules
and this could account for the £157 phenotype of this class of trans-
conjugants on peas. Alternatively, the R; trifolii g1: genes may be
transcribed and translated in pea nodules but may not lead to active

gene products. Possibly these transconjugants lack essential factors

129

which fail to allow the recognition of signals from pea plants which
would permit differentiation of the bacteria into functional bacteroids
in pea nodules. Another explanation is that these transconjugants lack
factors which would elicit the expression of plant proteins, such as
apoleghemoglobin, which are essential for an effective symbiosis.

These factors could be encoded on parts of pJBSJI deleted during forma-
tion of the hybrid plasmids or on parts of thele_leguminosarum genome
not transferred to strain T37. The latter is unlikely since Class I
and III transconjugants form effective nodules on peas.

A "switch mechanism" was postulated to explain the instability of
the Fix+ phenotype of R; trifolii strains into which pJBSJI had been
transferred (56). Vacillation between Fix+ and Fix‘ phenotypes on
pea and clover plants was observed. No changes in the indigenous 3g
trifolii plasmids or recombination events between pJBSJI and the _R_:_
trifolii plasmids were reported.

In contrast, we have observed recombination between pRtT37a and
pJB5JI which resulted in alterations in the plasmid profiles of the
transconjugant strains. The recombination resulted in the deletion of
either the 3g trifolii 311 and ggg genes (Class I) or the R; legumino-
§ggggugifi genes (Class II). In Class III and Class IV transconjugant
strains, both sets of gym genes were stably maintained and were func-
tional on both host plants. No oscillation between Fix+/Fix‘
phenotypes was observed.

The recombination event between pJBSJI and pRtT37a may occur by
several mechanisms. The altered plasmid profiles may result from
recombination between homologous regions of DNA on the two plasmids, in

which varying amounts of plasmid DNA are deleted. Prakash gg'gl. (147)

130

using non-stringent hybridization and washing conditions, observed
extensive DNA homology between gig-containing plasmids of R; legumino-
ggggg, R; phaseoli, and R; trifolii. Conservation of DNA on_g1:-con-
taining plasmids in these strains of Rhizobium was proposed.

In fact, there are regions of homology on pJBSJI and pRtT37a.

Both plasmids hybridize to 32P-labeled g1: DNA (Figure 27, a and

b). However, between species, the homologous region of £1: DNA is only
about 1.5 kb (166). Pea nodulation genes are encoded on pJB5JI (100)
and clover ggg genes have been reported to be plasmid encoded in R_._
trifolii (95,215). Results presented here suggest that genes essential
for clover nodulation are encoded on pRtT37a. The degree of homology
between pea and clover ggg genes is not known.

Homology between pJBSJI DNA and pRt137a DNA was observed even
under stringent hybridization and washing conditions (Figure 31; Figure
32, lane 1). This region of DNA common to both plasmids may have
allowed recombination to occur. However, recombination between pJB5JI
and plasmids of Rg_trifolii strain 0403 or other strains has not been
observed (Figure 16, reference 56). Thus, if the recombination event
occurs by this mechanism, the DNA sequences involved must be unique to
pRtT37a and not other Rg_trifolii plasmids.

Another mechanism for the recombination between pJBSJI and pRtT37a
may be the action of transposable elements and/or insertion sequences.
The formation of cointegrates has been proposed as an intermediate step
in the transposition and replication of transposable elements and
insertion sequences (33,181). Evidence reported herein suggests,
however, that the recombination is not mediated by Tn §_transposition.

Southern blot analysis of donor and transconjugant DNA flanking Tn.§

131

demonstrated that Tn.§ had not transposed (Figure 23). Also genetic
analysis of Tn 5 transposition in Rg_leguminosarum (12) and Rg_meliloti
(126) indicated that transposition occurs only at very low frequency
(approximately 10‘8) in Rhizobium. Indigenous insertion elements

have been identified in Rg_lggigi_(l49) and in one strain of R;
meliloti (167), but have not as yet been reported in 3g_trifolii.

A model based on the recombination between homologous regions of
pJBSJI and pRtT37a is shown in Figure 33. Conjugal transfer of pJBSJI
from R; leguminosarum strain T83K3 to R; trifolii strain T37 occurs
first (step 1). This is followed by recombination between pJBSJI and
pRtT37a (step 2), accompanied by the deletion of varying amounts of
plasmid DNA. In some cases (Class I and II) ggg and/or g1: genes are
deleted.

Class IV strains, which contain two "hybrid" plasmids, could arise
by a double crossover between pJBSJI and pRtT37a. During this event,
the R; leguminosarum nif genes (and possibly the ggg genes) would be
recombined into pRtT37a. This would explain the hybridization data in
which Tn 5 was localized on a different plasmid than the gif genes
(Figure 26). The plasmid in Class IV strains on which the pea ggg
genes were located was not determined. The g1: and ggg genes have been
shown to be linked on pJBSJI (94) and also on plasmid in other strains
of Rhizobium (5,164). This suggests that the pea ggg genes may be
located on the same plasmid as the g1: genes.

This model indicates how the plasmids observed in the transcon-
jugant strains could result, and explains the Tn.§ and g1: hybridiza-
tion data and also the observed symbiotic phenotypes of these strains

inoculated on pea and clover plants.

132

Figure 33. A model for the generation of the "hybrid plasmids of R;
trifolii strain T37 (pJBSJI) transconjugants. Step 1, transfer of
pJBSJI from R; leguminosarum strain T83K3 to R; trifolii strain T37.
Step 2, recombination event between pJB5JI and pRtl37a resulting in the
four classes of transconjugants. 1, Class I transconjugants in which
Rg_trifolii strain T37 g1: and nod genes (hatched boxes) have been
deleted; Rg_leguminosarum strain-T83K3‘ggg and g1: genes (open boxes)
are present. II, C ass II transconjugants in which strain T83K3 g1:
genes have been deleted; strain T83K3 ggg and R; trifolii nif and ggg
genes are present. III, Class III transconjugants which contain both
sets of g1: and ggg genes; DNA from one or both plasmids may be
deleted. IV, Class IV transconjugants which contain two "hybrid"
plasmids. Tn 5 is located on a different plasmid from the g1: genes.
The location a? the R; leguminosarum nod genes is not definite.

 

 

 

133

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CHAPTER III

Characterization of a Spontaneous Non-nodulating Mutant of
Rhizobium trifolii Strain 0403: Linkage of Nodulation
and Nitrogen Fixation Genes.

INTRODUCTION

Bacteria of the genus Rhizobium are capable of forming a species-
specific symbiosis with leguminous plants. The presence of large plas-
mids (>100 Mdal) appears to be a general feature of strains of
Rhizobium (35,72,137,146). Several reports have shown that genes
required for nodulation (Egg) of legumes and for nitrogen fixation
(git) are plasmid-encoded (5,11,79,94,100,123,147,164). Historically,
many investigators have reported instability of nodulation and nitrogen
fixation phenotypes in strains of Rhizobium following long term storage
or growth on certain media (2,58,117,139,154,175). More recently,
non-nodulating (Ngg') isolates of R; leguminosarum and Rg_meliloti
have been obtained which contain deletions of plasmid DNA (5,11,164).

Results presented in this chapter describe the isolation and
characterization of a mutant of Rg_trifolii strain 0403 which contains
a large deletion of DNA in the smallest plasmid. Evidence presented
herein indicated that the g1: structural genes and gene(s) essential

for nodulation are located on this region of deleted plasmid DNA.

134

MATERIALS AND METHODS

Materials. All of the materials used in this chapter have been
described in Chapters I and II.

Bacterial Strains and Plasmids. The bacterial strains and plas-

 

mids used in this chapter have been described in Chapter II. Plasmid-
deletion mutants of R; trifolii strain 0403 were detected in a culture
of strain 0403 stored for about one year at 4°C on a BMM medium (7)
agar slant. The culture was restreaked on BMM agar plates and 40 iso-
lates were picked. These isolates were numbered sequentially starting
with 2001.

‘Mggig. In addition to the media described in Chapters I and II,
BMM medium (7) was used to culture 3; trifolii strains. BMM medium
consisted of: 10.0 g mannitol, 0.23 g KZHP04, 0.1 g MgSO4°7H20,
1.1 g sodium glutamate, and 1.0 ml each of vitamin and trace element
stock solutions per liter of redistilled water (pH 6.8). The trace
element stock solution contained: 0.5 g CaClz, 14.5 mg H3803, 12.5 mg
FeSO4-7H20, 7.0 mg CoSO4*7H20, 0.5 mg CuSO4-7H20, 0.5 mg MnC12°4H20,
11 mg ZnSO4'7H20, 12.5 mg Na2M004 per 100 ml redistilled water. The
vitamin stock solution contained (per 100 ml redistilled water): 12.0
mg gyg-inositol and 2 mg each of riboflavin, p-aminobenzoic acid,
nicotinic acid, biotin, thiamine'HCl, pyridoxine°HCl, and calcium

pantothenate.

135

136

Methods. The characterization of the plasmid deletion mutant of
R;_trifolii strain 0403 was performed using the techniques described in

Chapter II.

RESULTS

Bacteria from a culture of R; trifolii strain 0403 stored at 4°C
for about one year were observed to have an altered plasmid profile
(data not shown). In addition to the four indigenous plasmids of R;
trifolii strain 0403, a smaller plasmid was present in the bacteria
from this culture. The intensity of fluorescence associated with this
plasmid and with pRt0403a, the smallest plasmid of wild-type strain
0403, was less than that observed for the other three plasmids. This
culture was restreaked and forty single colony isolates were obtained.

Plasmid Analysis. Plasmid profiles of these isolates were
examined using the Ekhardt agarose gel technique (Figure 34). Nineteen
isolates were found to have the same plasmid profile as wild-type
strain 0403 (Figure 34, lane A). The plasmid profiles of the other
twenty-one isolates were altered. The three high molecular weight
plasmids, pRt0403b, c, and d, had the same electrophoretic mobility in
agarose gels as the corresponding plasmids of wild-type strain 0403.
The smallest plasmid in these twenty-one isolates migrated faster than
pRt0403a. Apparently, a large segment of DNA had been deleted from
pRt0403a to generate the new, smaller plasmid.

To determine the approximate size of the deletion, the molecular
weights of pRt0403a and the plasmid-deletion derivative, termed
pRt0403aA, were estimated. A standard curve was constructed using

plasmids of known molecular weight as described in Chapter II (Figure

137

138

 

Fi ure 34. Ekhardt agarose gel electrophoresis of R; trifolii strain
0403 and strain 2016. Lane A, strain 0403; 8, strain 20l6 containing
pRt0403aA. Abbreviations of plasmid designations are indicated on the
left. -

139

35). The molecular weight of pRt0403a was estimated to be 290 x 105.
The molecular weights of the other three plasmids of R; trifolii strain
0403 were estimated by extrapolation of the curve above 325 x 105.

The estimated molecular weights were 335 x 105, 350 x 106 and 390 x

106 for pRt0403b, c, and d, respectively.

Unfortunately, the log relative mobility of pRt0403aA is located
on the non-linear portion of the curve. Since the shape of the curve
through this region is not known, only an approximate molecular weight
can be obtained. By extrapolation of both linear segments of the
curve, upper and lower estimates of the molecular weight of pRt0403aA
of 250 x 106 and 185 x 105, respectively, were obtained. This
gives an average molecular weight of about 220 x 105, indicating that
approximately 70 Mdal of DNA had been deleted from pRt0403a.

Clover Nodulation Test. The symbiotic properties of the isolates
from the mixed culture of strain 0403 were examined on the host plant,
white clover (Table 11). The twenty-one isolates harboring pRt0403aA
did not nodulate clover plants, whereas, all eighteen isolates possess-
ing the wild-type plasmid profile formed nitrogen-fixing nodules
(Nod+Fix+) on clover roots. Thus, the deleted region of plasmid
DNA apparently codes for gene(s) required for nodulation ability.

Fahraeus Slide Analysis of Infection Process. To further charac-
terize the plasmid deletion mutation, the infection process of wild-
type strain 0403 and a plasmid-deletion mutant (strain 2016) was
examined microscopically using the Fahraeus slide technique.

Bacteria of both strain 2016 and strain 0403 were observed to bind
to the tips of clover root hairs (Figure 36, panels a and b). Attach-

ment of RhizobiUm to root hair tips has been described as one of the

140

 

.37 d

'05“ l c b

3? - 1 a

._.‘4 “---~.‘-“~J!~ 1' 11¢;

4- . 3|P~P.~_

.523 - l-- l

5.. I

.12 “~~"X

3

\

 

 

 

685 090 0395 1.0
Log Relative Mobility

Fi ure 35. Molecular weight of plasmid DNA versus log relative
mobility of plasmids in agarose gels. X, pTiC58, II, pMGS; A, pMGl; 0,
pAtC58. The log relative mobilities of the four indigenous plasmids of
R; trifolii strain 0403 (a, b, c, and d) and the deletion derivative of
pRt0403a (denoted aA) are indicated by the arrows. The line drawn
through pTiC58 (X) was constructed as described in Figure 15.

141

Table 11. Comparison of the symbiotic properties of isolates ofjig
trifolii strain 0403 containing pRt0403a or pRt0403aA.

 

 

Plasmid Present Number Number
in Isolate Tested NodTFix+
pRt0403a 18 18
pRt0403aA 21 0

 

White clover seedlings were grown aseptically on Jensen agar slants in
enclosed tubes. Plants were inoculated with the respective culture and
after five weeks growth, scored for nodules (NodT) and tested for
acetylene reduction ability (FixT).

142

Figure 36. Attachment of R; trifolii to clover root hair tips. Panel
a, strain 0403 containing pRt0403a; panel b, mutant strain 2016
containing pRt0403aA.

 

 

 

 

143

 

144

initial steps in the infection process (6,43,44,48). Curling of root
hairs of clover plants inoculated with either strain 2016 or strain
0403 was also observed (Figure 37, panels a and b). Infection threads
were induced in very tightly curled root hairs of plants inoculated
with R; trifolii strain 0403 (Figure 38). However, neither tightly
curled root hairs nor infection threads were observed in plants inocu-
lated with strain 2016. Presumably, DNA coding for attachment of bac-
teria to clover root hair tips and for the "curling inducer(s)" were
not deleted from pRt0403a or are encoded on another region of the
genome. A gene(s) coding for an essential process in the development
of infection threads was apparently located on the DNA deleted from
pRt0403a.

Nif Hybridization Analysis. To determine which plasmid in R;
trifolii strain 0403 and strain 2016 codes for the 91: genes,
32P-labeled 3g meliloti g1: DNA was hybridized to a Southern filter
of an Ekhardt agarose gel of plasmid DNA from R; trifolii strain 0403
and strain 2016 (Figure 39). The gif probe was observed to hybridize
only to pRt0403a in strain 0403. The probe did not hybridize to
pRt0403aA, or to the other plasmids in strain 2016, indicating that the
‘gif structural genes were probably deleted.

To confirm that the 31: genes were indeed deleted from strain 2016
and had not recombined into the chromosome, total DNA from strain 0403
and strain 2016 was isolated, digested with restriction endonuclease
Egg_RI and examined for the presence of the g1: structural genes by
Southern blot analysis using K; pneumoniae g1: DNA as a heterologous
hybridization probe. The 32P-labeled‘fig pneumoniae g1: probe

hybridized to a 5.0 kb Egg_RI restriction fragment of R; trifolii

145

Figure 37. Root hair curling of clover plants inoculated with R;
tri 0 ii strains. Panel a, strain 0403 containing pRt0403a; panel b,
mutant strain 2016 containing pRt0403aA.

 

146

 

147

Fi ure 38. Infection thread in a curled root hair of a clover plant
inoculated with R. trifolii strain 0403. The arrow indicates the

infection threadT— Infection threads were not observed in plants
inoculated with Rg_trifolii strain 2016.

 

148

 

149

strain 0403 DNA. The probe did not hybridize to any fragments of
strain 2016 DNA. Thus, the 5.0 kb Egg_RI restriction fragment contain-
ing the R. trifolii nif structural genes which is normally present on

pRt0403a has been deleted in strain 2016.

150

‘I-l'

 

8H.

Figure 39. Hybridization of a heterologous L'f probe to plasmid DNA
from R; trifolii strain 0403 and strain 2016 separated on an Ekhardt
agarose gel. Lanes 1 and 2 show the plasmids present in strains 0403
and 2016, respectively. Lanes 3 and 4 show autoradiographs of Southern
blots of the plasmid DNA from lanes 1 and 2, respectively, hybridized
with 32P-labeled pRmR2. The symbol, >, denotes pRt0403a.

151

5.3- -
eo- -.
4.3- u
3.5- «-
11.4- .—
20- a

Figure 40. Autoradiogram of a heterologous 1le probe hybridized to a
Southern blot of Egg RI-digested total DNA of 1_1_._ trifolii. Lane 1, a
mixture of 25 ng of Egg RI digested, 25 ng Eco RI+Eag Hl-digested and
25 ng of Ec_o R1 + Hind III-digested pSA30 DNA: Lanes 2 and 3, fig
RI-digested total W (5 pg) from R_._ trifolii strain 0403 and strain
2016, respectively. The heterologous gj_f_ probe was the 3.4 k E
RI-Efld III Fragment of pSA30 (see Figure 14).

DISCUSSION

Four large plasmids are present in 3; trifolii strain 0403. Long
term storage of this strain resulted in a deletion of about 65 Mdal of
DNA from the smallest resident plasmid, pRt0403a. Isolates harboring
the approximately 220 Mdal deletion derivative of pRt0403a were unable
to nodulate clover plants.

Analysis of the infection process of strain 0403 and strain 2016
by the Fahraeus slide techique, showed that cells from both strains
would bind to the tips of clover root hairs and would induce curling
and deformation of the root hairs. Strain 2016 could not induce the
formation of infection threads. Thus, DNA coding for genes essential
for attachment of the bacterial cells to clover roots, and for the
curling inducer(s) were either not deleted from pRt0403a or were
located on another region of the genome. In contrast, DNA coding for
gene(s) essential for the formation of infection threads were deleted
from pRt0403a.

Hybridization of a heterlogous £1: probe to plasmid DNA from
strain 0403 and strain 2016 showed that the gif structural genes,
normally present on pRt0403a, had also been deleted. Thus, at least
some g1: and ggg genes are linked and are present on the approximately
70 Mdal region deleted from pRt0403a.

Deletion of DNA coding for genes required for the symbiosis

between Rhizobium and legumes have been reported for other species of

152

153

Rhizobium (5,11,164). Beringer gg.gl. (11) isolated a Nod‘ strain of
.3; leguminosarum following UV-mutagenesis. The mutant phenotype was
found to be the result of a large deletion in one of the plasmids.
Complementation tests with pRllJI, a transmissible plasmid coding for
both ggg and git genes, and symbiotically defective derivatives (Nod'
or Fix‘) of pRllJI demonstrated that both gif and ggg genes were
located in the deleted region of DNA (22,29). Spontaneous and heat-
induced deletions in the "megaplasmid" of several strains of _R_._
meliloti have been reported (5,164). These deletions were variable in
size and were also found to span both the g1: strucutral genes and the
ggghgenes.

It is interesting to note that most of the observed plasmid dele-
tions have occurred in the region of the ggg and gjj_genes. Few dele-
tions which produce auxotrophs (i.e., chromosomal) or which occur in
plasmid DNA not coding for the Egg and g1: genes have been observed.
For the latter, this may be due to the lack of genetic markers other
than g1: and ggg.

The relatively high frequency of loss of the g1: and ggg genes
suggests there may be a "hot spot" for deletion formation in the region
of these symbiotic genes (5). Insertion sequences (IS elements) have
been shown to mediate the deletion of variable, but not random, amounts
of DNA adjacent to the IS element (33,155). The frequency of deletion
formation mediated by the insertion element 131 can be 100- to 1000-
fold above the background rate of spontaneous deletions (155). Both
the location and orientation of the IS element affect the frequency of
deletions (33,155). The mechanism of deletion formation appears to be

closely related to the replitation/transposition process (33).

154

Insertion elements indigenous to Rhizobium have been identified in
Eg_lgpjgi_(149) and E; meliloti (167). The E; meliloti insertion ele-
ment (ISle) is a multicopy element which transposes at high frequency
into Eg_meliloti g1: DNA. The occurrence of deletions of Egg and gif
genes in the strain containing ISle was not reported.

While the mechanism for the deletion of g1: and ggg genes in
Rhizobium is at present unknown, an explicitly speculative model for
IS-mediated deletion of plasmid DNA containing the g1: and ggg genes
can be postulated (Figure 41, panel B). This model is based on a model
(Figure 41, panel A) proposed for deletion formation during transposi-
tion of IS elements and transposons (33,181). The insertion element
transposes to another site on the plasmid (Figure 41, step 1). During
the cleavage, replication, and ligation events necessary for production
of the new IS element, two circular DNA molecules are generated (Figure
41, step 2). However, with the orientation shown for the origin of
replication (ggi), the IS element, the symbiotic genes, and the target
site, the circular DNA molecule containing the gij and ggg genes (8)
would not contain an origin of replication and would be diluted out of
the culture by cell division. The other plasmid molecule (A) from
which the symbiotic genes have been deleted contains an origin of
replication and would be stably maintained. The variation in the size
of the deletion would be dependent upon the location of the target site
of transposition of the IS element.

To date, IS elements have not been detected in E; trifolii strain
0403, or in any other 3; trifolii strains. Until IS elements are

demonstrated in Eg_trifolii strain 0403, the speculative model

155

presented for the mechanism of deletion of symbiotic genes from

pRt0403a cannot be verified.

156

Figure 41. A speculative model for IS-mediated deletion of plasmid DNA
containing nif and nod genes in Rhizobium. Panel A, model proposed by
Shapiro (18TT—for the generation of delet tions by insertion elements.
Panel 8, model for deletion of plasmid DNA encoding nif and nod genes
in Rhizobium. Step 1, transposition of insertion element to— target
site. Step 2, replication of the insertion element at the target site
resulting in the production of two plasmid DNA molecules, (A) and (B).
Plasmid (A) contains the origin of replication (ori) and will replicate
normally. Plasmid B, which contains the nif and nod genes, lacks an
ori site and does not replicate. Plasmid T8T'will— be diluted out of
the bacterial population by cell division and plasmid segregation.

 

d
0
j 13
° b
«101101
18 IS

157

normally

 

 

replicate

SUMMARY

Large plasmids coding for genes essential for the Eg_trifolii -
clover symbiosis have been identified in two strains of.Eg'trifolii.l
The smallest of the three plasmids of E; trifolii strain T37 and of the
four plasmids of R; trifolii strain 0403 were shown to code for the
nitrogenase (Elf) structural and nodulation (ggg) genes. A non-nodu-
lating mutant of strain 0403 containing a spontaneous deletion of about
70 Mdal of plasmid pRt0403a DNA was isolated. This mutant was able to
attach to and induce the curling of root hairs but was unable to induce
the formation of infection threads. Hybridization analysis using a
heterologous gii probe indicated that the g1: structural genes, which
are encoded on pRt0403a, had been deleted. Thus the gifi genes and
genes essential for nodulation are encoded within a 70 Mdal region of
pRt0403a.

Transfer of pJBSJI, a Eg_leguminosarum plasmid coding for pea nodu-
lation and giE genes, into E;_trifolii strain T37 generated transconju-
gants containing a variety of plasmid profiles. The altered plasmid
profiles resulted from the recombination of pJBSJI with the smallest
plasmid, pRtT37a, of strain T37. The plasmid profiles of the transcon-
jugants and the symbiotic properties exhibited on both peas and clover
were stably maintained even after reisolation of the transconjugants
from root nodules. The 140 transconjugants were grouped into four

classes based on the plasmid profile and the symbiotic properties

158

159

exhibited on both host plants.
Class I transconjugants were unable to nodulate clover but formed
effective nodules on peas. The Eg_leguminosarum nif genes were present

but the Eg_trifolii nif genes had been deleted. These strains con-

 

tained a plasmid of molecular weight corresponding to pJBSJI. Southern
analysis of total DNA using pJBSJI as a hybridization probe indicated
that the Class I strains contained less pJBSJI sequences than the

pJB5JI donor, 3; leguminosarum strain T83K3. Similar analysis with

 

pRtT37a indicated that some pRt137a sequences were also present.
The Class II transconjugant strains formed nodules on both host
plants. However, Nz-fixing nodules were only formed on pea plants.

Hybridization analysis indicated that the E; leguminosarum nif genes

 

had been deleted. Although the Eg_trifolii.gif genes were present,
these genes could not complement the deleted 3; leguminosarum nif genes
in pea nodules, suggesting that the regulation of Eg_leguminosarum nif
genes may be different than that for the gifi genes in Eg_trifolii.
Strains of Classes III and IV formed effective nodules on both host
plants. Both sets of symbiotic genes were present in these strains and
were functional. Class III strains contained a single "hybrid" plasmid
larger then pRtT37a, while Class IV strains contained two plasmids of
molecular weights intermediate to the molecular weights of pJB5JI and
pRtT37a.

Several unique findings were observed during this research.
Transfer of pJBSJI to E; trifolii strain T37 resulted in recombination
between pJBSJI and pRtT37a, both of which encode genes essential for
the Rhizobium - legume symbiosis. Approximately 30% of the transconju-

gants (Class I) were stably converted to a Eg_leguminosarum strain.

160

Other transconjugants possessed a host range extended to include pea
plants, a first step toward extension of the host range of Rhizobium to
include non-leguminous plants.
Several of the strains isolated during this research may be useful

for further research on the symbiosis. The E; trifolii strain 0403
non-nodulating mutant containing a large deletion of plasmid DNA could
be utilized in the study of genes required for the symbiosis by comple-
mentation analysis. The Class II transconjugants may be beneficial in

the study of the regulation of g1: expression in pea nodules.

REFERENCES

2.

3.

5.

6.

10.

11.

12.

REFERENCES

. 1983. Gas prices to alter methanol, ammonia uses.

 

Chem. Eng. News 16:16-17.

Allen, E.K. and 0.N. Allen. 1950. Biochemical and symbiotic
properties of the rhizobia. Bacterial. Rev. 14:274-330.

Appelby, C.A. 1974. Leghemoglobin, p. 522-554. lg_A. Quispel
(ed.), The biology of nitrogen fixation. North Holland,
Amsterdam.

Ausubel, F.M. and F.C. Cannon. 1981. Molecular genetic analysis
of Klebsiella pneumoniae nitrogen fixation (git) genes. Cold
Spring Harbor Symp. Quant. Biol. 45:487-492.

 

Banfalvi, Z., V. Sakanyan, C. Konec, A. Kiss, I. Dusha, A.
Kondorosi. 1981. Location of nodulation and nitrogen fixation
genes on a high molecular weight plasmid of Rhizobium meliloti.
Mol. Gen. Genet. 184:318-325.

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

Bergersen, F.J. 1961. The growth of Rhizobium in synthetic

Bergersen, F.J. 1974. Formation and function of bacteroids. p.
474-498. lg A. Quispel (ed.), The biology of nitrogen fixation.
North-Holland, Amsterdam.

Bergersen, F.J. and D.J. Goodchild. 1973. Cellular location and
concentration of leghemoglobin in soybean root nodules. Aust. J.
Biol. Sci. 26:741-756.

Beringer, J.E. 1974. R factor transfer in Rhizobium
leguminosarum. J. Gen. Microbiol. 84:188-198.

Beringer, J.E., N.J. Brewin, and A.W.B. Johnston. 1980. The
genetic analysis of Rhizobium in relation to symbiotic nitrogen
fixation. Heredity 45:|6l-|86.

Beringer, J.E., J.L. Beynon, A.V. Buchanan-Wollaston, and A.W.B.

Johnston. 1978. Transfer of the drug resistance transposon Tn.§
to Rhizobium. Nature 276:633-634.

161

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

162

Beringer, J.E., S.A. Hoggan, and A.W.B. Johnston. 1978. Linkage
mapping in Rhizobium lgguminosarum by means of R plasmid mediated
recombination. J. Gen. Microbiol. 104:201-207.

Beynon, J.L., J.E. Beringer, and A.W.B. Johnston. 1980.
Plasmids and host range in Rhizobium leguminosarum and Rhizobium
ghaseoli. J. Gen. Microbiol. 120:421-429.

Bhuvaneswari, T.V. and W.D. Bauer. 1978. Role of lectins in
plant-microorganism interactions. 111. Binding of soybean lectin
to root cultured rhizobia. Plant Physiol. 62:71-74.

Bhuvaneswari, T.V., S.G. Pueppke, W.D. Bauer. 1977. Role of
lectins in plant-microorganism interactions. I. Binding of
soybean lectin to rhizobia. Plant Physiol. 60:486-491.

Bisseling, T., R.C. van den B05, and A. van Kammen. 1978.
Effect of ammonium nitrate on synthesis of nitrogenase and
leghemoglobin in pea root nodules induced by Rhizobium
leguminosarum. Biochim. Biophys. Acta 539:1- .

Blattner, F.R., B.G. Williams, A.E. Blechl, K.
Denniston-Thompson, H.E. Faber, L.-A. Furlong, D.J. Grumwald,
D.O. Kiefer, D.D. Moore, E.L. Sheldon, and 0. Smithies. 1977.
Charon phages: safer derivatives of bacteriophage A for DNA
cloning. Science 196:161-169.

Bogusz, 0., Houmard, J. and J.-P. Aubert. 1981. Electron
transport to nitrogenase in Klebsiella pneumoniae. Eur. J.
Biochem. 120:421-426.

Bohlool, B.B. and E. L. Schmidt. 1974. Lectins: a possible
basis for specificity in the Rhizobium-legume root nodule
symbiosis. Science 185:269-271.

Brachet, J. 1968. Effect of acridines on morphogenesis. Nature
220:488-489.

Brewin, N.J., J.E. Beringer, A.V. Buchanon-Wollaston, A.W.B.
Johnston and P.R. Hirsch. 1980. Transfer of symbiotic genes
with bacteriocinogenic plasmids in Rhizobium leguminosarum. J.
Gen. Microbiol. 116:261-270.

Brewin, N.J., J.E. Beringer, and A.W.B. Johnston. 1980.
Plasmid-mediated transfer of host-range Specificity between two
strains of Rhizobium leguminosarum. J. Gen. Microbiol.
120:413-420.

Brewin, N.J., T.M. DeJong, D.A. Phillips, and A.W.B. Johnston.
1980. Co-transfer of determinants for hydrogenase activity and
nodulation ability in Rhizobium leguminosarum. Nature
288:77-79.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

163

Brill, W.J. 1977. Biological nitrogen fixation. Scientific
Amer. 236:68-81.

Brill, W.J. 1980. Biochemical genetics of nitrogen fixation.
Microbiol. Rev. 44:449-467.

Broughton, W.J. 1978. Control of specificity in legume-
Rhizobium associations. J. Appl. Bacterial. 45:165-194.

Broughton, W.J. and W.J. Dilworth. 1971. Control of
leghemglobin synthesis in snake beans. Biochem. J. 125:1075-
1080.

Buchanon-Wollaston, A.V., J.E. Beringer, N.J. Brewin, P.R.
Hirsch, and A.W.B. Johnston. 1980. Isolation of symbiotically
defective mutants of Rhizobium leguminosarum by insertion of the
transposon Tn.§ into a transmissible plasmid. Mol. Gen. Genet.
178:185-190.

Buchanon-Wollaston, V., M.C. Cannon, and F.C. Cannon. 1981. The
use of cloned g1: (nitrogen fixation) DNA to investigate
transcriptional regulation of-gif expression in Klebsiella
pneumoniae. Mol. Gen. Genet. 184:102-106.

 

Bulen, W.A., and J.R. LeComte. 1966. The nitrogenase system
from Azotobacter. Proc. Natl. Acad. Sci. U.S.A. 56:979-986.

Caldwell, B.E. and Vest, H.G. 1977. Genetic aspects of
nodulation and dinitrogen fixation in legumes: The
macrosymbiont. p. 557-575. 'Eg R.W.F. Hardy and W.S. Silver
(eds.), A treatise on dinitrogen fixation. Wiley Interscience,
London.

Calos, M.P. and J.H. Miller. 1980. Transposable elements. Cell
20:579-595.

Cannon, F.C., G.E. Riedel, and F.M. Ausubel. 1979. Overlapping
sequences of Klebsiella pneumoniae nif DNA cloned and
characterized. Mol. Gen. Genet. 174?59-66.

Casse, F., C. Boucher, J.S. Julliot, M. Michel, and J. Denarie.
1979. Identification and characterization of large plasmids in
Rhizobium meliloti using agarose gel electrophoresis. J. Gen.
Microbiol. ll3z229-242.

Clewell, D.B. 1972. The nature of Col El plasmid replication in
Escherichia coli in the presence of Chloramphenicol. J.
Bacteriol. 110:667-676.

Clewell, D.B. and D.R. Helinski. 1969. Supercoiled circular
DNA-protein complex in Escherichia coli: purification and induced
conversion to an apen circular DNA form. Proc. Natl. Acad. Sci.
U.S.A. 62:1]59‘1166.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

164

Clowes, R.C. 1972. Molecular structure of bacterial plasmids.
Bacterial. Rev. 36:361-405.

Corbin, D., L. Barran, and G. Ditta. 1983. Organization and
expression of Rhizobium meliloti nitrogen fixation genes. Proc.
Natl. Acad. Sci. U.S.A. 80:3005-3009.

Corbin, 0., G. Ditta, and D.R. Helinski. 1982. Clustering of
nitrogen fixation (gif) genes in Rhizobium meliloti. J.
Bacteriol. 149:221-228.

Currier, T.C. and E.W. Nester. 1976. Isolation of covalently
closed circular DNA of high molecular weight from bacteria.
Anal. Biochem. 76:431-441.

Cutting, J.A. and H.M. Schulman. 1971. The biogenesis of
leghemoglobin. The determinant in the Rhizobium-legume symbiosis
for leghemoglobin specificity. Biochim. Biophys. Acta

229:58-62.

Dart, P.J. 1974. The infection process. p. 381-329. Eg_A.
Quispel (ed.), The biology of nitrogen fixation. North Holland,
London.

Dart, P.J. 1977. Infection and development of leguminous
nodules. p. 367-472. Eg_R.W.F. Hardy and W.S. Silver (ed.), A
treatise on dinitrogen fixation, section III, Wiley Interscience,
London.

Dart, P.J. and F.V. Mercer. 1964. The legume rhizosphere.
Arch. Microbiol. 47:344-378.

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

Dazzo, F.B. and W.J. Brill. 1979. Bacterial polysaccharide
which binds Rhizobium trifolii to clover root hairs. J.

Bacteriol. 137:1362-7 .

Dazzo, F.B. and D.H. Hubbell. 1975. Crossreactive antigens on
lectins as determinants of symbiotic specificity in Rhizobium

trifolii clover association. Appl. Microbiol. 30:10 - .

Dazzo, F.B., M.R. Urbano, and W.J. Brill. 1979. Transient
appearance of lectin receptors on Rhizobium trifolii. Curr.
Microbiol. 2:15-20.

de Brui, F.J., and F.M. Ausubel. 1981. The cloning and
tranSposon Tn E_mutagenesis of the gig R, a gene involved in the
regulation of the g1: and Egg operons. Mol. Gen. Genet.
183:289-297.

51.

52.

53.

54.

55.

56.

57.
58.

59.

60.

61.

62.

63.

64.

65.

165

Denhardt, R.T. 1966. A membrane-filter technique for the
ggtggtigz6of complementary DNA. Biochem. Biophys. Res. Comm.

DePicker, A., M. van Montagu, and J. Schell. 1977. Physical map
of RP4. p. 678-679. Eg_A.I. Bukhari, J.A. Shapiro, and S.L.
Adhya (eds.), DNA: Insertion elements, plasmids and episomes.
Cold Spring Harbor Laboratory.

Dilworth M.J. 1969. The plant as the genetic determinant of
leghemoglobin production in the legume root nodule. Biochem.
Biophys. Acta 184:432-441.

Dilworth, M.J. and D.K. Kidby. 1968. Localization of iron and
leghemoglobin in the le ume root nodules by electron microscope
autoradlography. Exp. ell Res. 49:148-159.

Dixon, R.A., R.R. Eady, G. Espin, S. Hill, M. Iaccarino, D. Kahn,
and M. Merrick. 1980. Analysis of regulation of Klebsiella .

neumoniae nitro en fixation (nif) gene cluster with gene qulon.
Nafure 286:128-1 2 (1980)

Djordjevic, M.A., W.Zurkowski, and B.G. Rolfe. 1982. Plasmids
and stability of symbiotic properties of Rhizobium trifolii. J.
Bacteriol. 51:560-568.

Drummond, M. 1979. Crown gall disease. Nature 281:343-347.

Drummond, M., J. Clements, M. Merrick, and R. Dixon. 1983.
Positive control and autogenous re ulation of the g1: LA promoter
in Klebsiella pneumoniae. Nature 01:302-307.

Dunican, L.K. and Cannon, F.C. 1971. The genetic control of
symbiotic roperties in Rhizobium: evidence for plasmid control.
P ant and 01 . Special Volume:73-79.

Dun'can, L.K. and A.B. Tierney. 1974. Genetic transfer of
ni from Rhizobium trifolii to Klebsiella aerogenes. Biochem.
'BTEphys. Res. Camm. 57:62-72.

 

Ead , R.R., R. Issack, C. Kennedy J.R. Pastgate. H.D. Ratcliffe.
197 . Nitrogenase synthesis in lebsiella pneumoniae: comparison
p34agmgnggg and oxygen regulation. J. Gen. Micrfibial.

Eady R.R. and J.R. Pastgate. 1974. Nitrogenase. Nature
249:805-810.

Ekhardt, T. 1978. A rapid method for the identification of
plasmid desaxynucleic acid in bacteria. Plasmid 1:584-588.

Espin, G., A. Alvarez-Morales and M. Merrick. 1981.
Complementation anal sis of In A-linked mutations which affect
nitrogen fixation. 1. Gen. enet. 184:213-

Fahraeus, G. 1957. The infection of clover root hairs by nodule
bacteria studied bg a simple glass slide technique. J. Gen.
Microbiol. 16:374- 81.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

166

Fishbeck, K., H.J. Evans, and L.L. Boersma. 1973. Measurement
of nitrogenase activity of intact legume symbionts in situ using
the acetylene reduction assay. Agronomy J. 65:429-433.

Freifelder, D. 1971. Electron microscope study of ethidium
bromide-DNA complex. J. Mol. Biol. 60:401-403.

Furhrmamm, M. and H. Hennecke. 1982. Coding properties of
cloned nitrogenase structural genes from Rhizobium japonicum.
Mol. Gen. Genet. 187:419-425.

Gibbons, A.M. and K.F. Gregory. 1972. Relatedness among
Rhizobium and Agrobacterium species determined by three methods
of nucleic acid hybridization. J. Bacteriol. 111:129-141.

 

Giles, K.W. and A. Meyers. 1965. An improved diphenylamine
method for the estimation of deoxyribonucleic acid. Nature
206:93.

Girvitz, S.C., S. Bacchetti, A.J. Rainbow, and F.L. Graham.
1980. A rapid and efficient procedure for the purification of
DNA from agarose gels. Anal. Biochem. 106:492-496.

Gross, D.C., A.K. Vidaver, and R.V. Klucas. 1979. Plasmids,
biological properties and efficacy of nitrogen fixation in

Rhizobium a onicum strains indigenous to alkaline soils. J.
Gen. Microbiol. ll4:257-266.

Guerry, P., D.J. LeBlanc, and S. Falkow. 1973. General method
for the isolation of plasmid deoxynucleic acid. J. Bacteriol.
116:1064-1066.

Hamblin, J. and S.P. Kent. 1973. Possible role of
phytohaemogglutinin in Phaseolus vulgaris L. Nature New Biology
245:28-30.

Hansen, J.B. and R.H. Olsen. 1978. Isolation of large bacterial
plasmids and characterization of the P2 incompatibility group
plasmids pMGl and pMG5. J. Bacteriol. 35:227-238.

Hardy, K.G. 1975. Colicinogeny and related phenomena.
Bacterial. Rev. 39:464-515.

Hardy, R.W.F. and U.D. Havelka. 1975. Nitrogen fixation
research: A key to world food Science. 188:633-643.

Hardy, R.W.F., R.D. Holston, E.K. Jackson, and R.C. Bruns. 1968.
The acetylene-ethylene assay for nitrogen fixation: laboratory
and field evaluation. Plant Physiol. 43:1185-1207.

Haugland, R. and D.P.S. Verma. 1981. Interspecific plasmid and
genomic DNA sequence homologies and localization of g1: genes in

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

167
effective and ineffective strains of Rhizobium japonicum. J.
Mol. Appl. Gen. 1:205-217.

Heberlein, G.T., J. de Ley, and R. Tijtgat. 1967.
Deoxyribonucleic acid homology and taxonomy of A robacterium,
94'll6-l24.

Rhizobium and Chromobacterium. J. Bacteriol.

Helinski, D.R. and D.B. Clewell. 1971. Circular DNA. Ann. Rev.
Biochem. 40:900-942.

Hennecke, H. 1981. Recombinant plasmids carrying nitrogen
fixation genes from Rhizobium japonicum. Nature. 291:354-355.

Hennecke, H. and K.T. Shananagan. 1979. Temperature control of
nitrogen fixation in Klebsiella ppeumoniae. Arch. Microbiol.
123:259-265.

Higashi, S. 1966. Electron microscopic studies on the infection
thread developing in the root hair of Trifolium re ens L.
infected with Rhizobium trifolii. J. Een. Appl. lcrobiol.
12:147-156.

Higashi, S. 1967. Transfer of clover infectivity of Rhizobium
trifolii to Rhizobium haseoli as mediated by an episomic factor.
en. Appl. Mlcrofilo . : 1-403.

Hill, S. and E. P. Kavanagh. 1980. Roles of Lif F and Lif J gene
products in electron transport to nitrogenase“ in Klebsiella
pneumoniae. J. Bacteriol. 141:470-475.

Hill, 5., C. Kennedy, P. Kavanagh, R. B. Goldberg, and R. Hanaw.
1981. Nitrogen fixation gene (Lif L) involved in oxygen
regulation of nitrogenase syntheSls in Klebsiella pneumoniae.
Nature. 290: 424-426.

Hirota, Y. 1960. The effect of mating type factors in
Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 46:57-64.

Hirsch, P.R. 1979. Plasmid-determined bacterocin production by
Rhizobium leguminosarum. J. Gen. Microbiol. 113:219-228.

Hirsch, P.R., M. van Montagu, A.W.B. Johnston, N. Brewin, and J.
Schell. 1980. Physical identification of bacteriocinogenic,
nodulation, and other plasmids in strains of Rhizobium
leguminosarum. J. Gen. Microbiol. 120:403-4 2.

Hahn, B. and D. Korn. 1969. Cosegregation of a sex factor with
the Escherichia coli chromosome during curing by acridine orange.
J. Mol. Biol. 455385-395.

Hall, F.B. and T.A. LaRue. 1976. Genetics of legume plant
hosts. pp. 391-399. Eg_W.F. Newton and C.J. Nyman (eds.), Proc.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

168

First International Symposium on Nitrogen Fixation, Vol. II.
washington State Univ. Press, Pullman, HA.

Holsters, M., B. Silva, C. Genetello, F. Engler, F. van Vliet, M.
DeBlock, R. Villaroel, M. van Montagu, and J. Schell. l978.
Spontaneous formation of cointegrates of the oncogenic Ti plasmid
and the wide host range plasmid RP4. Plasmid l:456-467.

Hombrecher, G., N.J. Brewin, and A.W.B. Johnston. l981. Linkage
of genes for nitrogenase and nodulation ability on plasmids in R;
leguminosarum and R; phaseol'. Mol. Gen. Genet. 182:133-l36.

 

Hookyaas, P.J.J., A.A.N. van Brussel, M. Den Dulk-Ras, G.M.S. van
Slogtern and R.A. Schilperoot. l98l. Sym plasmid of Rhizobium

trifolii expressed in different rhizobial species and

Agrobacterium tumefaciens. Nature 29l:35l-353.

Hrabak, E.M., M.R. Urbano, and F.B. Dazzo. l98l.
Growth-phase-dependent immuno-determinants of Rhizobium trifolii
lipopolysaccharide which bind Trifolin A, a white clover lectin.
J. Bacteriol. l48z697-7ll.

Hubbell, D.H. 1970. Studies on the root hair "curling factor"
of Rhizobium. Bot. Gaz. l3lz337-342.

Inman, R.B. and M. Schnas. l97D. Partial denaturation of
thymine- and S-bromouracil-cantaining A DNA in alkali. J. Mol.

Janssan, P.E., B. Lindberg and H. Ljunggren. l979. Structural
studies of the Rhizobium trifolii extracellular polysaccharide.
Carbohydr. Res. 75:207-220.

Johnston, A.W.B., J.L. Beynon, A.V. Buchanan-Hollaston, S.M.
Setchell, P.R. Hirsch, and J.E. Beringer. l978. High frequency
transfer of nodulating ability between strains and species of

Rhizobium. Nature. 276:635-636.

Johnston, A.W.B., G. Hombrecker, N.J. Brewin, M.C. Cooper. l982.
Two transmissible plasmids in Rhizobium leguminosarum strain 300.

Jorgensen, R.A., S.J. Rothstein, and H.S. Reznikoff. l979. A
restriction enzyme cleavage map of Tn.§ and location of a region
coding for neomycin resistance. Mol. Gen. Genet. l77z65-72.

Kamberger, H. l979. An ouchterlony double diffusion study an
the interaction between legume lectins and rhizobial cell surface
antigens. Arch. Microbial. l21:83-9D.

Klein, G.E., P. Jemison, R.A. Haak, and A.G. Matthysee. 1975.
Physical evidence of a plasmid in Rhizobium japonicum.
Experientia 31:532-533.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

169

Kleinschmidt, A.K. 1967. Monolayer techniques in electron
microscopy of nucleic acid molecules. p. 361-376. .lfl L.
Grossman and K. Moldave (eds.), Methods in Enzymology, Vol. XII
B. Academic Press, New York.

Kawalczuk, E. and Z. Lorkiewicz. 1977. Transfer of Rldrd
plasmid from Escherichia coli to Rhizobium trifolii by
conjugation. Acta Microbiol. Polan. 26:9-l8.

 

Krol, A.J.M., J.G.J. Hontelez, R.C. van den 305, and A. van
Kammen. l980. Expression of large plasmids in the endasymbiotic
form of Rhizobium leguminosarum. Nuc. Acids Res. 8:4337-4347.

Krol, A.J.M., J.G.J. Hantelez and A. van Kammen. l982. Only one
of the large plasmids in Rhizobium le uminosarum strain PRE is
strongly expressed in the endosymBioEic sfafe?‘ J. Gen.
Microbiol. 128zl839-lB47.

Kumarasinghe, R.M.K. and P.S. Nutman. 1977. Rhizobium
stimulated callose formation in clover root hairs and its
relation to infection. J. Exp. Bot. 28:96l-976.

Kurtz, w.a.w. and T.A. Larue. l975. Nitrogenase activity in

Rhizobium in absence of host plant. Nature 256:407-408.

Ledeboer, A.M., A.J.M. Krol, J.J.M. Dons, F. Spier, R.A.
Schilperoort, I. Zaenen, N. van Larebeke, and J. Schell. 1976.
On the isolation of Ti plasmid from Agrobacterium tumefaciens.
Nuc. Acids Res. 3:449-463.

Legocki, R.P. and D.P.S. Verma. l979. A nodule-sgecific plant
protein (nodulin-35) from soybean. Science 205:19 -l93.

Legocki, R.P. and D.P.S. Verma. l980. Identification of "nodule
s ecific" host proteins (nodulins) involved in the development of

R izobium-legume symbios1s. Cell 20:153-l56.

Libbenga, K.R. and R.J. Bo ers. l974. Root nodule
morphogenesis. pp. 43l-47 . In A. Quispel (ed.). The biology
of nitrogen fixation. North-Holland, Amsterdam.

Liener, I.E. l976. Phytohemagglutinins. Ann. Rev. Plant
Physiol. 27:291-3l9.

Long, S.R., H.J. Buikema, and F.M. Ausubel. l982. Cloning of
Rhizobium meliloti nodulation genes by direct complementation of
N297 mutants. Nature 298:485-488.

Langley, B.J., T.D. Berge, J.M. van Lamen, and I.L. Baldwin.
l937. Changes in the infective ability of rhizobia and
Phytomonas tumefaciens induced by culturing in media containing
glycine. J. Bact. 35:29-30.

Lorkiewicz, 2., R. Russa, and T. Urbanik. l978. Nitrogen
fixation by Rhizobium in pure culture. Acta Microbiol. Polan.
27:5‘90

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

170

Maier, R.J. and H.J. Brill. l976. Ineffective and
non-nodulating strains of Rhizobium japonicum. J. Bacteriol.
l27z763-769.

Maier, R.J. and N.J. Brill. 1978. Involvement of Rhizobium

flagonicum O-antigen in soybean nodulation. J. Bacteriol.
: -1299.

Maniatis, T., E.F. Fritsch, and J. Sambrook. l982. Molecular
cloning: a laboratory manual, p. 168-l69. Cold Spring Harbor
Laboratory, Cold Spring Harbor, N.Y.

Maniatis, T., A. Jeffrey, and D.G. Kleid. l975. Nucleotide
sequence of the rightward operator of phage A. Proc. Natl. Acad.
SCio U.S.A. 72:]184-11880

Masterson, R.V., P.R. Russell, and A.G. Atherly. l982. Nitrogen
fixation (nif) genes and large plasmids of Rhizobium japanicum.
J. BacteriETT l52:928-93l.

May, J.H., R.H. Houghton, and C.J. Perret. l964. The effect of
growth at elevated temperature on some heritable properties of
Staphylococcus aureus. J. Gen. Microbiol. 37:l57-l69.

McLean, P.A. and R.A. Dixon. 1981. Requirement of flif_v gene
for production of wild-type nitrogenase enzyme in Klebsiella
pneumoniae. Nature. 292:655-666.

Meade, H.M., S.R. Long, G.B. Ruvkun, S.E. Brown, and F.M.
Ausubel. l982. Physical and genetic characterization of
symbiotic and auxotrophic mutants of Rhizobium meliloti induced
by transposon TnI§ mutagenesis. J. Bacteriol. l49:ll4-l22.

Merrick, M., S. Hill, H. Hennecke, M. Hahn, R. Dixon, C. Kennedy.
l982. Repressor properties of the git L gene product in
Klebsiella pneumoniae. Mol. Gen. Genet. l85:75-8l (l982).

 

Meyer, M.C. and S.G. Pueppke. l980. Differentiation of
Rhizobium japonicum strain derivatives by antibody sensitivity
patterns, lectin binding, and utilization of biochemicals. Can.
J. Microbiol. 26:606-6l2.

Meyers, J.A., D. Sanchez, L.P. Elwell, and S. Falkow. l976.
Simple agarose gel electrophoretic method for the identification
and characterization of plasmid deoxyribonucleic acid. J.
Bacteriol. 127:l529-1537.

Miller, J.H. l972. Experiments in molecular genetics. p.
352-355. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

Mortenson, L.E. and R.N.F. Thorneley. l979. Structure and
function of nitrogenase. Ann. Rev. Biochem. 48:387-4l8.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

171

Nadler, K.D. and Y.J. Avissar. l977. Heme synthesis in soybean
root nodules. I. On the role of bacteroid.£-levulinic acid
synthetase and i-acid dehydrase in the synthesis of the heme of
leghemoglobin. Plant Physiol. 60:433-436.

Napoli, C.A. and P. Albersheim. 1980. Rhizobium leguminosarum
mutants incapable of normal extracellular polysaccharide
production. J. Bacteriol. l4l:l454-l456.

Napoli, C.A. and D.H. Hubbell. 1975. Ultrastructure of
Rhizobium-induced infection threads in clover root hairs. Appl.
Microbial. 30:1003-1009.

Nieva-Gomez, D., G.P. Roberts, S. Kleviciks, and H.J. Brill.
1980. Electron transport to nitrogenase in Klebsiella
pneumoniae. Proc. Natl. Acad. Sci. U.S.A. 77:2555-2558.

Navick, R.P., R.C. Clowes, S.N. Cohen, R. Curtiss, N. Datta, and
S. Falkow. 1976. Uniform nomenclature for bacterial plasmids: a
proposal. Bact. Rev. 40:168-169.

Nuti, M.P., A.M. Ledeboer, A.A. Lepidi and R.A. Schilperoort.
1977. Large plasmids in different Rhizobium species. J. Gen.
Microbiol. 100:241-248.

Nuti, M.P., A.A. Lepidi, R.K. Prakash, R.A. Schilperoort, and
F.C. Cannon. 1979. Evidence for nitrogen fixation (Elf) genes
on indigenous Rhizobium plasmids. Nature 282:533-535.

Nutman, P.S. 1946. Variation within strains of clover nodule
bacteria in the size of the nodule produced and in the
effectivity of the symbiosis. J. Bacterial. 51:411-412.

Nutman, P.S. 1969. Genetics of symbiosis and nitrogen fixation
in soybeans. Proc. Royal Soc. B. 172:417-437.

Olivares, J., E. Montoya, and A. Palomares. 1977. Some effects
derived from the presence of extrachromosomal DNA in Rhizobium
meliloti. pp. 375-385. In H. Newton, J.R. Pastgate. and C.
Roariguez-Barruesco (eds.7: Recent developments in nitrogen
fixation. Academic Press, New York.

0w, D.H., and F.M. Ausubel. 1983. Regulation of nitrogen_
metabolism genes by 21: A gene product in Klebsiella pneumoniae.
Nature. 301:307-313.

0w, D.H., V. Sundareson, D.M. Rothstein, S.E. Brown, and F.M.
Ausubel. 1983. Promoters regulated by the gln G (Q£E_C) and nif
A gene products share a heptameric consensus sequence in the -
region. Proc. Natl. Acad. Sci. U.S.A. 80:2524-2528.

144.

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

172

Pagan, J.D., J.J. Child, M.R. Scowcroft, and A.H. Gibson. 1975.
Nitrogen fixation by Rhizobium on a defined medium. Nature
256:406-407.

Planque, K. and J. Kijne. 1977. Binding of pea lectins to a
glucan-type polysaccharide in the cell walls of Rhizobium
leguminosarum. FEBS Lett. 73:64-66.

Prakash, R.K., P.J.J. Hooykaas, A.M. Ledeboer, J. Kijne, R.A.
Schilperoort, M.P. Nuti, A.A. Lepidi, F. Casse, C. Boucher, J.S.
Julliot and J. Denarie. 1979. Detection, isolation and
characterization of large plasmids in Rhizobium. In Proceedings
of the 3rd International Symposium on Nitrogen Fixation. H.E.
Newton and D.H. Drme-Johnson. University Press, Madison, WI.

Prakash, R.K., R.A. Schilperoort, and M.P. Nuti. 1981. Large
plasmids of fast-growing Rhizobia: homology studies and location
of structural nitrogen fixation Lnif) genes. J. Bacteriol.
145:1129-1136.

Prakash, R.K., A.A.N. van Brussel, A. Quint, A.M. Mennes, and R.
A. Schilperoort. 1982. The map position of sym-plasmid regions
expressed in the bacterial and endasymbiotic form of Rhizobium
leguminosarum. Plasmid 7:281-286.

Priefer, U.B., H.J. Burkhardt, H. Klipp, and A. Puhler. 1980.
ISRI: an insertion element isolated from the soil bacterium
Rhizobium lupin‘. Cold Spring Harbor Symp. Quant. Biol.

Puhler, A. and H. Klipp. 1981. Fine structure analysis of the
nif gene region for Nz-fixation (311) of Klebsiella pneumoniae.

'Tfi—“Biology of inorganic nitrogen and sulfur",. Bathe and A.

Trebst (eds.) p. 276-278. Springer Verlag. Berlin.

Pull, S.P., S.G. Pueppke, T. Hymowitz, and J.H. Drf. 1978.
Soybean lines lacking the 120,000-dalton seed lectin. Science
200:1277-1279.

Radloff, R., N. Bauer, and J. Vinagrad. 1967. A dye-bouyant
density method for the detection and isolation of closed circular
duplex DNA: the closed circular DNA in HELA cells. Proc. Natl.
Acad. SCio U.S.A. 57:1514'15210

Rawlings, J., V.K. Shah, J.R. Chisnell, H.J. Brill, R. Zimmerman,
E. Munck, and R.H. Orme-Johnsan. 1978. Novel metal cluster in
an iron-molybdenum cofactor of nitrogenase. J. Biol. Chem.
253:1001-l004.

Reid, J.J., E.B. Fred, I.L. Baldwin. 1935. The effects of
storage on rhizobia. J. Bacteriol. 29:75-76.

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

173

Reif, H.J. and H. Saedler. 1975. 151 is involved in deletion
formation in the gal_region of E; coli K12. Mol. Gen. Genet.
137:7-15.

Rhoades, M., L.A. MacHattie, and C.A. Thomas, Jr. 1968. The P22
bacteriophage DNA molecule. 1. The mature form. J. Mol. Biol.
37:21-40.

Riedel, G.E., F.M. Ausubel, and F.C. Cannon. 1979. Physical map
of chromosomal nitrogen fixation 121:) genes of Klebsiella
pneumoniae. Proc. Natl. Acad. Sci. U.S.A. 76:2866-2870.

 

Rigby, P.N.J., M. Dieckman, C. Rhodes, and P. Berg. 1976.
Labeling deoxynucleic acid to a high specific activity in vitro
by nick translation with DNA polymerase I. J. Mol. BioTT
113:237-251.

Roberts, G.P. and N.J. Brill. 1980. Gene product relationship
of the nif regulon of Klebsiella pneumoniae. J. Bacteriol.
144:2103216.

Roberts, G.P. and H.J. Brill. 1981. Genetics and regulation of
nitrogen fixation. Ann. Rev. Microbiol. 35:207-235.

Roberts, G.P., T. MacNeil, D. MacNeil, and N.J. Brill. 1978.
Regulation and characterization of protein products coded by the
£1: (nitrogen fixation) genes of Klebsiella pneumoniae. J.
Bacteriol. 136:267-279.

Robertsen, B.K., P. Aman, A.G. Darvill, M. McNeil, and P.
Albersheim. 1981. Host symbiont interactions. V. The structure
of acidic extracellular polysaccharides secreted by Rhizobium
.lgguminosarum and Rhizobium trifolii. Plant Physiol.
67:389-400.

Robertson, J.G., P. Lyttleton, and G.E. Pankhurst. 1981.
Preinfection and infection processes in the legume-Rhizobium
symbiosis. p. 280-291. In A.H. Gibson and H.E. Newton eds.),
Current perspectives in nitrogen fixation. Griffin Press
Limited, Natley, Australia.

Rosenberg, C., P. Boistard, J. Denarie, and F. Casse-Delbart.
1981. Genes controlling early and late functions in symbiosis
are located on a megaplasmid in Rhizobium meliloti. Mol. Gen.
Genet. 184:326-333.

Roslycky, E.B. 1967. Bacteriocin production in the rhizobia
bacteria. Can. J. Bacteriol. 13:431-433.

Ruvkun, 6.8. and F.M. Ausubel.‘ 1980. Interspecies homology of
nitrogenase genes. Proc. Natl. Acad. Sci. U.S.A. 77:191-195.

167.

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

174

Ruvkun, G.B., S.R. Long, H.M. Meade, R.C. van den 805, and F.M.
Ausubel. 1982. ISle: a Rhizobium meliloti insertion sequence

that transposes preferentially into nitrogen fixation genes. J.
Mol. Appl. Gen. 1:405-418.

Ruvkun, G. B., V. Sundaresan, and F. M. Ausube1.l982. Directed
transposon Tn 5 mutagenesis and complementation analysis of

Rhizobium meliloti symbiotic nitrogen fixation genes. Cell.
29: 551- 559.

Sahlman, K., and G. Fahraeus. 1963. An electron microscope
study of root-hair infection by Rhizobium. J. Gen. Microbial.
33:425-531.

St. John, R.T., H.M. Johnston, C. Seidman, D. Garfinkel, J.K.
Gordon, V.K. Shah, and N.J. Brill. 1975. Biochemistry and
genetics of Klebsiella pneumoniae mutant strains unable to fix
N20 Jo BaCteriO]o 121: - a

St. John, R.T., V.K. Shah, and N.J. Brill. 1974. Regulation of
nitrogenase synthesis by oxygen in Klebsiella pneumoniae. J.
Bacteriol. 119:266-269.

Sanders, R.E., R.H. Carlson, and P. Albersheim. 1978. A
Rhizobium mutant incapable of nodulation and normal
polysaccharide secretion. Nature 271:240-242.

Schubert, K.R. 1982. Symbiotic nitrogen fixation in nodulated
leguminous and nonleguminous angiosperms. p. 265-285. ln_0.R.
Zaborsky, CRC handbook of biosolar resources, Vol. 1. CRC Press,
Boca Raton, Florida.

Schubert, K.R. and H.J. Evans. 1976. Hydrogen evolution: a
major factor affecting the efficiency of nitrogen fixation in
nodulated symbionts. Proc. Natl. Acad. Sci. U.S.A.
73:1207-1211.

Schwinghamer, E.A. 1969. Mutations to auxotrophy and
pratotrophy as related to symbiotic effectiveness in Rhizobium

1e uminosarum and Rhizobium trifolii. Can. J. MicrobioI.
15:611-616.

Schwinghamer, E.A. 1975. Properties of some bacteriocins
produced by Rhizobium trifolii. J. Gen. Microbiol. 91:403-413.

Schwinghamer, E.A. 1980. A method for improved lysis of some
gram-negative bacteria. FEMS Microbiol. Letters. 7:157-162.

Schwinghamer, E.A. and R.P. Belkengren. 1968. Inhibition of
rhizobia by a strain of Rhizobium trifolii. Archiv. Microbial.
64:130-145.

179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

190.

191.

192.

175

Scott, D.B. and C.N. Ronson. 1982. Identification and
mob111zation by cointegrate formation of a nodulation plasmid in
Rhizobium trifolii. J. Bacteriol. 151:36-43.

 

Shah, V.K. and N.J. Brill. 1977. Isolation of an iron
molybdenum cofactor of nitrogenase. Proc. Natl. Acad. Sci.
U.S.A. 74:3249-32530

Shapiro, J.A. 1979. Molecular model for the trans osition and
rep 1cation of bacterioBhage Mu and other transposa 1e elements.
Proc. Natl. Acad. Sci. .S.A. 76:1933-1937.

Skinner, K.J. 1976. Nitrogen fixation. p. 22-33. Chemical and
Engineering News. October 4, 1976.

Slonimiski, P.P, G. Perrodin, and J.H. Croft. 1968. Ethidium
bromide induced mutation of yeast mitochondria: complete
transformation of cells into res iratory deficient
ggnéggrgggsomal "petites“. Bioc em. Biophys. Res. Comm.

Solheim, B., and J. Raa. 1973. Characterization of the
substances causing deformation of root hairs of Trifolium repens
when inoculated with Rhizobium trifolii. J. Gen. Microbiol.
77:241-247.

Southern, E.M. 1975. Detection of specific sequences among DNA
fragments separated by gel electrophoresis. J. Mol. Biol.
98:503-517.

Steinert, M. and S. van Assel. 1967. The loss of kineto lastic
DNA in two species of trypanosomatidae treated with acrif avin.
J. Cell Biol. 34:489-503.

Strijdam, B.H. and 0.N. Allen. 1969. Properties of strains of
Rhizobium trifolii after cultivation on medium supplemented with
amfino acids. Pfiifophylactica 1:147-152.

Su, L.-C., S.G. Preppke, and H.P. Friedman. l980. Lectins and
the soybean-Rhizobium symbiosis. I. Immunological investigations
of soybean lines, the seeds of which have been reported to lack
the 120,000 daltan soybean lectin. Biochim. Biophys. Acta
629:292-304.

Sundareson, V., J.D.G. Jones, D.H. 0w, and F.M. Ausubel. 1983.
Klebsiella pneumoniae gif_A product activates the Rhizobium

meliloti nitrogenase promoter. Nature 301:728-732.

Sutton, H.D. 1974. Some features of the DNA of Rhizobium
bacteroids and bacteria. Biochim. Biophys. Acta 366:1-10.

Terawaki, Y-. H- Takayasu, and T. Akiba. 1967. Thermosensitive
gngggagggn of a kanamycin resistance factor. J. Bacteriol.

Thornton, H.G. 1945. Effective and ineffective strains of
legume nodule bacteria. Nature 156:654-655.

193.

194.

195.

196.

197.

198.

199.

200.

201.

202.

203.

204.

176

Tjepkema, J. and H.J. Evans. 1975. Nitrogen fixation by free
living Rhizobium in a defined liquid medium. Biochem. Biophys.
Res. Comm. 65:625-628.

Toms, G.C. 1971. Phytohemagglutinins. p. 367-462. lfl.J-B°
Harborne, D. Boulter, and B.L. Turner, (eds.), Chemotaxonomy of
Legumes. Academic Press, New York.

Tshitenge, G., N. Luyindula, P.F. Lurquin, and L. Ledoux. 1975.
Plasmid deoxynucleic acid in Rhizobium vigna and Rhizobium
trifolii. Biochem. Biophys. Acta 414:357-361.

Tsien, H.C. and Schmidt, E.L. 1980. Accumulation of soybean
lectin-binding polysaccharide during growth of Rhizobium
Japonicum as determined by hemagglutination assay. Appl.
Environ. Microbiol. 39:1100-1104.

Vande Noude, G.E., M. Dskarsson, L.w. Enquist, S. Nomura, and

P.J. Fischinger. 1979. Cloning of integrated moloney sarcoma
proviral DNA sequences in bacteriophage 1. Proc. Natl. Acad.

Sci. U.S.A. 76:4464-4468.

van Lorebecke, N., G. Engler, M. Holsters, S. van den Elsacker,
I. Zaenen, R.A. Schilperoort, and J. Schell. 1974. Large
plasmid in Agrobacterium tumefaciens essential for crown
gall-inducing ability. Nature 252:169-170.

Verma, D.P.S. and A.K. Bal. l976. Intracellular site of
synthesis and localization of leghemoglobin in root nodules.
Proc. Natl. Acad. Sci. U.S.A. 73:3843-3847.

Verma, D.P.S., V. Kazazian, V. Zagbi and A.K. Bal. 1978.
Isolation and characterization of the membrane envelope enclosing

the bacteroids in soybean root nodules. J. Cell Biol.
78:919-936.

Vincent, J.M. 1970. International Biological Programme
Handbook, no. 15. A manual for the practical study of root
nodule bacteria. Blackwell Scientific Publications, Oxford.

Hahl, G.M., M. Stern, and G.R. Stark. 1979. Efficient transfer
of large DNA fragments from agarose gels to diazobenzyloxymethal-
paper and rapid hybridization by using dextran sulfate. Proc.
Natl. Acad. Sci. U.S.A. 76:3683-3687.

Watson, 8., T.C. Currier, M.P. Gordon, M.-D. Chilton, and E. H.
Nester. 1975. Plasmid required for virulence of Agrobacterium
tumefaciens. J. Bacteriol. 123:255-264.

Neislander, L. 1979. A simple method to recover intact high
molecular weight RNA and DNA after electrOphoretic separation in
low gelling temperature agarose gels. Anal. Biochem. 98:305-

205.

206.

207.

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

177

Netmur, J.G., N. Davidson, and J.V. Scarletti. 1966.‘ Properties
of DNA of bacteriophage N1, a DNA with reversible circularity.
Biochem. Biophys. Res. Comm. 25:684-688.

Wittenberg, J.B. 1980. Utilization of leghemoglobin-bound
oxygen by Rhizobium bacteroids. p. 53-67. ‘IQ'N.E. Newton and
H.H. Orme-Johnson 1eds.), Nitrogen Fixation II. University Park
Press, Baltimore.

Holpert, J.S. and P. Albersheim. l976. Host-symbiont
interactions. I. The lectins of legumes interact with the
O-antigen containing lipopolysaccharides of their symbiont
rhizobia. Biochem. Biophys. Res. Comm. 70:729-737.

Hang, P.P. 1980. Interactions between rhizobia and lectins of
lentil, pea, broad bean and jack bean. Plant Physiol.
65:1049-1052.

Yamagata, H. and H. Uchida. 1969. Effect of acridine orange on
sex factor multiplication in Escherichia coli. J. Mol. Biol.
46:73-84.

Yamamoto, K.R., B.M. Alberts, R. Benzinger, L. Lawhorne, and G.
Treibner. 1970. Rapid bacteriophage sedimentation in the
presence of polyethylene glycol and its application to large
scale virus purification. Virology 40:734-744.

Yao, P.Y. and J.M. Vincent. 1969. Host specificity in root hair
“curling factor" of Rhizobium spp. Aust. J. Biol. Sci.
22:413-423.

Yao, P.Y. and J.M. Vincent. 1976. Factors responsible for the
curling and branching of clover root hairs by Rhizobium. Plant
and Soil 45:1-16.

Zaenen, I., N. van Larebeke, H. Teuchy, M. van Montagu, and J.
Schell. 1974. Supercoiled circular DNA in crown gall inducing
Agrobacterium strains. J. Mol. Biol. 86:109-127.

Zhu, J. and H.J. Brill. 1981. Temperature sensitivity of the
regulation of nitrogenase synthesis by Klebsiella pneumoniae. J.
Bacteriol. 145:1116-1118.

Zurkowski, w. 1981. Conjugal transfer of the
nodulation-conferring plasmid pHZ2 in Rhizobium trifolii. Mol.
Gen. Genet. 181:522-524.

Zurkowski, M., M. Hoffman and Z. Lorkiewicz. 1973. Effect of
acriflavin and sodium dodecyl sulfate on infectiveness of

Rhizobium trifolii. Acta Microbiol. Polan. 5:55-60.

Zurkowski, N. and Z. Lorkiewicz. 1976. Plasmid DNA in Rhizobium

trifolii. J. Bacteriol. 128:481-484.

178

218. Zurkowski, w. and Z. Lorkiewicz. 1978. Effect method for the
isolation of non-nodulating mutants of Rhizobium trifolii.

Genet. Res. 32:311-314.

219. Zurkowski, N. and Z. Lorkiewicz. 1979. Plasmid mediated control
of nodulation in R; trifolii. Arch. Microbiol. 123:195-201.