3:3é"""'

w". ‘.

"‘ € ' '

.;-L’f:-.Q-. w
,

a}

3' I
' ’ v.4 ”ya/1.31:7:
”,3 '1 “3 "v5
..3

3",
#3
AI}? ' ' ' 2512'}?

9:253! 0'1, W7
.7; ’I
33:33"; (“r/74W ”far?

.‘Z’: '45?" ' 3- "

3;}? ".3: Incl. " " ’1'. [$4.72 ""1"; r "

.1

_,.1 - '1. 333333 1'
3.5}9'233/ 733'? 2 (355:
51%;:6: ’h'

r

.‘l‘ 4 H' 1‘

2.5-2”- .gr.
3- 1 r2"

{5.}; Cf" .-
J'I'ff'ff'l/é' .
"1'. ’

3:1."2
.,..2 .75
,2 2~-»~2.22.:.2
fax"??" '5' £511"):
. , .
1 .22. .22...
1.». "2’2”
,rm 333,137, 1271133,

I

”’47,,” £25732; 34$";
II 'If‘. - ' f '
515:3... 5'1" (I ‘ " {if/3.33“

1:4,, nu
'JV‘IEA‘
21,21
3:369:53; ‘. .3/ ”#735,
2.2.21”; 3'32
fail-'2‘"! Ila-€323: '5

u
I"

7'
Ij'".& I’, '
r, , fr .
3,33 533:1? ,3, 33,130,333; .93,
3,217 . , '23??un
31:}! @533?" " 331%}; A 3.);

“:5‘3
2:

2222222....

'-’ {.3+-~wf~23.“"‘
. 4-»?an i‘EL'; {.04 . .
E '* n .k ' a
sin ,_‘{.:t:~‘=k\>.
0'}.

,1\\.,. - ”‘3‘ 3333.021“

T?' ‘10:.
«39“

.1. _,
u
D‘ \

‘3‘ I

q u R's

,g'ysfbgfifl
9- ,

V . .p" A J.
ft]. 33:33, r:':;f"3 2! "5 "2’
f;/}’. 413;: ..:j}]

firf, 733:1}

i’lfrt;
(2,23,,

. ‘
‘1-‘3‘
_¢~a>’§3.3
‘~

.‘~. €33

o

7. . 597;.
[397.

“125" '77:} 2

\\:75:3
“8::

2 .322 22222 ’ 2:1:
, 3.221%! :3 732%}??? 5'33‘62251
1322-: 1322;322:2222 2
h!- ' .2.‘Mt.2"2:)

L5,“!- 3'3}:
5:2”N1?
7v;

4'
‘ 49 .ff’ ‘I
("I ,‘l;fl'{;

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

1293 00558 6619

LIBRARY
Michigan State
University

 

 

 

 

This is to certify that the
dissertation entitled

THE ISOLATION AND CHARACTERIZATION OF
BRADYRHIZOBIUM JAPONICUM DNA SEQUENCES THAT
ARE TRANSCRIBED SPECIFICALLY IN BACTEROIDS

presented by

John S. Scott-Craig

has been accepted towards fulfillment
of the requirements for

PhD degree mBotany and Plant
Pathology

 

 

Major professor

Date January 15, 1988

 

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

 

 

 

 

remove this checkout from
your record. FINES will
——— be charged if book is
returned after the date
stamped below.

MSU ‘ RETURNING MATERIALS:

Place in book drop to
LIBRARIES
”

 

 

 

 

ISOLATION AND CHARACTERIZATION OF BRAQXRHIZOBIUM JAPONICUM

DNA SEQUENCES THAT ARE EXPRESSED SPECIFICALLY IN BACTEROIDS

By

John Stewart Scott-Craig

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

DOCTOR OF PHILOSOPHY

Department of Botany and Plant Pathology
1988

(3/
Lx' 1'

f

7

/
o

m?

1,.
«9

ABSTRACT

ISOLATION AND CHARACTERIZATION OF BRADYRHIZOBIUM JAPONICUM

DNA SEQUENCES THAT ARE EXPRESSED SPECIFICALLY IN BACTEROIDS

By

John Stewart Scott-Craig

The formation of nitrogen-fixing nodules on the roots
of soybean (Glycine 22;) after infection by the Gram
negative bacterium Bradyrhizobium japonicum is a complex
process which involves differentiation by both partners in
the symbiosis. This interaction is a developmental process,
known to require the induction of gene transcription in both
organisms. A method was developed using differential
hybridization techniques which permits the isolation of
sequences from g. japonicum which are expressed specifically
in bacteroids, the differentiated form of the bacteria found
in nodules. Several sequences exhibiting bacteroid-specific
expression were examined by deletion analysis for effects on
the development of the symbiosis. Seven derivatives of
wild-type g. japonicum strain 110d which had deletions of
sequences exhibiting bacteroid-specific expression were

constructed. In all cases the mutant bacterial strains were

John Stewart Scott-Craig

fully competent for symbiosis as indicated by the ability to
form nodules and to reduce exogenously supplied acetylene to
ethylene. The growth rates of the deletion derivatives and
the parental strain were found to be identical in both rich
and minimal culture media. Expression of the
bacteroid-specific sequences under various physiological
conditions was examined and transcription of two sequences
was shown to be induced by lowered oxygen tension. The DNA
sequence of one of the regions exhibiting bacteroid-specific
expression was determined and two divergent transcriptional
initiation sites were localized by 81 nuclease protection
analysis. The DNA sequence of the promoter regions of these
two sequences does not resemble that of g. japonicum genes
known to be involved in the symbiosis and thus defines a new

type of symbiotically-regulated promoter.

for my wife, Pat, my son Thomas and my parents

ii

ACKNOWLEDGEMENTS

I will always be indebted to Barry Chelm for giving me
a chance to return to graduate school and for providing me
with excellent advice and guidance during my stay in his
lab. Barry was always generous with his time and spent much
of it demonstrating how to analyze scientific problems and
to devise experimental approaches for solving them. He also
had a great facility for assembling a congenial and
productive lab group which made his lab an exciting and
stimulating place in which to work and and learn. I feel
very fortunate to have known and worked with him.

I would also like to thank to other members of my
guidance committee - Chris Somerville, Mike Thomashow, Ken
Nadler and Dennis Fulbright - for their assistance and
support during my graduate career.

A special thank you goes to Mary Lou Guerinot who
participated in all stages of this thesis. She was
co-author of the grant which funded the research, a source
of advice and encouragement during the experimental portion
of the work and an invaluable reviewer of the thesis itself.
Sine Ma non.

Many other current and former members of the Chelm lab
willingly afforded me of their time and expertise during the
course of my thesis project. Rob McClung, Tom Adams, Todd
Carlson and Bryan Rahe all assisted me in getting to know
Bradyrhigobiug and in learning molecular biological
techniques. Bill Holben, John Somerville, Prudy Hall,
Elizabeth Verkamp, Todd Cotter and Greg Martin all helped me
with various aspects of sequencing, promoter mapping and
mutant analysis. Bill in particular was an indispensable
m13 guru, thesis proofreader and dart partner.

I gratefully acknowledge the expert assistance of
Stuart Pankratz in the ultrastructural analysis of mutant
strains of Bradyrhiggbium.

In addition I would like to thank my former advisor and
mentor, A1 Ellingboe, who introduced me to plant pathology
and taught me the value of "thinking genetically".

Finally, I would like to thank my wife, Pat Claire, and
my son, Thomas, whose love and support made it possible for
me to get a Ph.D., ”just like Dr. Bunsen Honeydew".

 

iii

TABLE OF CONTENTS

 

 

Page
List of Tables . . . . . . . . . vii
List of Figures . . . . . . . viii
CHAPTER 1. Introduction . . . y . . . . 1
CHAPTER 2. Isolation of Bradyrhizobigm japonicum DNA
Sequences Transcribed Specifically in Bacteroids . 14
Introduction . . . . . . . . 14
Materials and Methods . . . . . . 17
Chemicals and Enzymes . . . . . 17
Bacterial Strains . . . . . . 17
Plasmids and Phage . . . . . . 1?
Bacterial Growth Conditions . . . . 19
Nodule Bacteria Fractionation . . . . 19
RNA Methods . . . . . . . 20
DNA Methods . . . . . . . 21
cDNA Synthesis . . . . . . . 23
Solution Hybridization and Competition
Hybridization . . . . . . 25
Results . . . . . . . . . 27
Isolation and Characterization of the rDNA
Genes of g. japonicum . . . . 27
Preparation of rcDNA-depleted cDNA Probes . 30
Screening the Phage Library for Bacteroid—
specific Transcription . . . . 34

Identification of Cosmids Containing Sequences
Expressed Specifically in g. japonicum
Bacteroids . . . . . . 37

Discussion . . . . . . . . 40

iv

CHAPTER 3. Genetic Analysis of Bradyrhizobium japonicum
DNA Sequences That Are Transcribed Specifically in
Bacteroids
Introduction
Materials and Methods

Bacterial Strains

Plasmids

Bacterial Growth Conditions
DNA and RNA Methods
Gene-directed Mutagenesis .

Plant Tests and Acetylene Reduction Assays

Extraction of Poly-B-hydroxybutyrate (PHB)
From 5. japonicum Bacteroids .

Determination of the Concentration of PHB
Dissolved in CHCla . .

Electron and Light Microscopy
Results
Identification, Mapping, and Subcloning of
Sequences Transcribed Specifically in
Bacteroids

Construction of Deletion Strains . .

Verification of Bacteroid-specific Induction
of Deleted Regions . . . .

Characterization of Deletion Strains

Expression of the Bacteroid-specific Sequences
Under Various Physiological Conditions

Discussion

CHAPTER 4. Fine Structure Analysis of Two Bacteroid-
specific Promoters in Bradyrhizobium japonicum

Introduction

43

43

44

44

45

46

46

47

48

49

50

51

52

52

56

60

63

65

84

92

92

Materials and Methods . . . . . . 94

Bacterial Strains and Plasmids . . . 94

DNA Sequence Analysis . . . . . 94

Nuclease 81 Protection Analysis . . . 98

Results . . . . . . . . . 99
Discussion . . . . . . . . 104
CHAPTER 5. Discussion . . . .1 . . . 111
REFERENCES . . . . . . . . . 122

CHAPTER

2

TABLE

1

LIST OF TABLES
Page

List of recombinant lambda phage
containing B. japonicum sequences

that are expressed specifically in
bacteroids and the corresponding
members of the cosmid library of

B. japonicum containing these same
sequences . . . . . 39

Acetylene reduction by excised

nodules of B. japonicum deletion
strains and wild-type strain

BJllOd . . . . . 64

Poly-B-hydroxybutyrate content of

B. japonicum nodules incited by
deletion strains BJ241, BJ242, and
BJ243 and wild-type strain

BJllOd . . . . . . 71

 

LIST OF FIGURES
CHAPTER FIGURE Page
2 l Genomic Library of B. iaponicum

110d in lambda vector BFlOl probed
with 5 x 106 CPM of cDNA synthesized

 

 

from bacteroid RNA . . . . 29
2 2 Northern blot of total B. japonicum

110d RNA . . . . . . 29
2 3 Restriction map of the B. japonicum

rDNA operon . . . . . . 29
2 4 Determination of B. japonicum rDNA

operon copy number . . . . 29
2 5 Methods for lowering the hybridization

signal from rcDNA . . . . . 33
2 6 Rescreening of recombinant bacteriophage

containing B. Japonicum sequences that
are expressed specifically in
bacteroids . . . . . . 36

2 7 Restriction endonuclease digests of DNA
purified from seven recombinant
bacteriophage . . . . . 38

3 l Cosmids containing sequences exhibiting
bacteroid—specific expression . . 53

3 2 Restriction maps of five regions of the
B. japonicum genome . . . . 55

3 3 Verification of the structure of B.
japonicum deletion derivative BJ261 . 59

3 4 Slot blots of B. japonicum RNA probed
with nick-translated plasmids containing
control genes or one of the sequences

replaced in the deletion constructions. 62
3 5a,b Growth curve of B. japonicum deletion

strains in minimal medium . 66, 67
3 5c,d Growth curve of B. japonicum. deletion

strains in rich medium . . 68, 69

viii

CHAPTER FIGURE Page

 

3 6 Transmission electron micrographs of
sections of nodules incited by B.
iaponicum deletion strains . . . 70
3 7 Effect of developmental stage and

growth stage on expression of g.

japonicum DNA sequences that are

expressed at high levels in

bacteroids . . . . . . 74

 

3 8 Effect of carbon source and oxygen
tension on expression of §;_japonicum
sequences that are expressed at high

 

levels in bacteroids . . . . 76
3 9 Effect of lowered oxygen tension on

expression of bacteroid-specific

sequences . . . . . . 77
3 10 Expression of bacteroid-specific

sequences in two-week-old nodules
incited by BJ2101, a nifA deletion

strain of g. japonicum . . . . 79
3 11 Expression of bacteroid-specific

sequences in B. japonicum cultures
grown under nitrogen-limiting
conditions . . . . . . 82

3 12 Expression of the bacteroid-specific
sequences in a B. japonicum ntrC
insertion strain . . . . . 83

3 13 Expression of the bacteroid-specific
sequences in heat-shocked B. japonicum
cells . . . . . . . 85

 

4 l Nuclease 81 protection analysis of two
divergently transcribed g. japonicum
sequences . . . . . . . 100

4 2 The nucleotide sequence of the B.
japonicum 1000 bp PstI-HindIII DNA
fragment from plasmid pBJ296 . . 101

4 3 Restriction map of and sequencing
strategy for the 1000 bp PgtI-HindIII
fragment of B. japonicum DNA contained
in plasmid pBJ296 . . . . . 103

 

CHAPTER FIGURE

4

4

Nuclease 81 protection analysis of
two divergently transcribed B.
japonicum sequences which exhibit
bacteroid—specific expression

 

Precise localization of the
transcription initiation sites
for two divergently transcribed
symbiotically-regulated sequences
in B. japonicum

DNA sequence of the promoter regions
of two B. japonicum symbiotically-
regulated sequences .

Page

105

106

107

CHAPTER 1

Introduction

Biological nitrogen fixation is estimated to account
for more than fifty percent of annual global nitrogen
fixation (Hardy and Havelka, 1975; Burris, 1980). The
ever-increasing demand for nitrogenous fertilizer and the
high cost of industrial nitrogen fixation have prompted
investigations aimed at improving the efficiency of the
biological process (Hardy, 1980). Due to its direct impact
on world agriculture, the nitrogen-fixing symbiotic
association between leguminous plants and bacteria of the
genera Rhizobium and Bradyrhizobium has been the focus of
much research. An understanding of the molecular mechanisms
by which the plant and bacterial symbiotic partners form a
productive association is a necessary component of efforts
to increase the efficiency of biological nitrogen fixation.

The development of an effective nitrogen-fixing
symbiosis between the Gram negative soil bacterium
Bradyrhigobium japonicum and its plant host, Glycine max
(soybean), is a complex process involving the
differentiation of both organisms. Studies have revealed a
series of morphological, physiological and biochemical
changes which take place during the development of root
nodules, the specialized bacteria-containing "organelles”

within which nitrogen fixation occurs. The attachment of

1

2
bacterial cells to the soybean root is the earliest event in
the establishment of the symbiosis which exhibits
specificity (Turgeon and Bauer, 1982; Dazzo and Gardiol,
1984). In the presence of the bacteria, growth in emerging
root hairs is arrested at the tip causing a deformation of
the hairs into the characteristic ”shepherd’s crook",
effectively surrounding some sites of bacterial attachment
(Turgeon and Bauer, 1985). The entrapped bacteria than
directly penetrate the plant cell wall by a mechanism which
is not well understood. Although the bacteria seem to
partially degrade the host cell wall, they do not come into
direct contact with the host cytoplasm but cause instead an
invagination of the plant cell wall which elongates to form
a tubular structure known as the infection thread (Turgeon
and Bauer, 1985). The bacterial cells multiply within the
infection thread which extends through the epidermal cell
and ramifies into the cortical cell layers of the root. The
mechanism by which infection thread formation occurs is of
interest since the structure is not seen in invasion of
legumes by pathogenic bacteria.

Another early response of the host plant is the
dedifferentiation of certain cells in the inner cortex of
the root. Studies on the g. meliloti/alfalfa symbiosis have
shown that within 21-24 hours after inoculation, well prior
to the arrival of the infection thread, a group of cortical

cells begins to divide actively (Dudley gt 31., 1987).

When the infection thread reaches the meristematic region,
some of the bacteria are released into the plant cytoplasm
enclosed by the peribacteroid membrane which is of host
origin (Brewin gt gt., 1985). The cells of most bacterial
species now begin to visibly differentiate, increasing in
size and DNA content (Bisseling gt gt., 1977; van den Bos gt
gt., 1978). Among the physiological changes which have been
documented are a decrease in the thickness of the bacterial
cell wall (MacKenzie gt gt., 1973; van Brussel gt gt.,
1977), a rapid increase in the amount and activity of
nitrogenase (van den Bos gt gt., 1978), a decrease in
protein synthesis (Bisseling gt gt., 1979), and the
elaboration of a novel set of cytochromes (Appleby,
1969a,b). The accumulation of large amounts (up to 503 of
bacteroid dry weight) of the storage polymer poly-B
-hydroxybutyrate is perhaps the most conspicuous change
visible microscopically (Klucas and Evans, 1968; Wong and
Evans, 1980). The above-mentioned biochemical,
physiological and morphological alterations during the
development of the mature bacteroid-containing nodule led to
efforts to identify both plant and bacterial genes which are
involved in this process.

Solution hybridization studies by Auger gt gt. (1979;
1981) comparing mRNA populations from Bradyhizobium-infected
and uninfected soybean roots demonstrated the existence of
20-40 nodule-specific transcripts. The protein products of

such messages, termed nodulins, have been examined for a

number of legumes (e.g. Legocki and Verma, 1979, 1980;
Bisseling gt gt., 1983, 1984; Lang-Unnasch and Ausubel,
1985, Lang-Unnasch gt gt., 1986; Mellor gt gt., 1986;
Sengupta-Gopalan gt gt., 1986; Verma gt al., 1986; Moerman

_t _t., 1987). The biochemical functions of a few nodulins
have been determined and include enzymes involved in
nitrogen metabolism (Bergmann gt gt., 1983; Triplett, 1985;
Sengupta-Gopalan and Pitas, 1986), carbohydrate metabolism
(Thummler and Verma, 1987), membrane synthesis (Mellor gt
gt., 1986), and control of oxygen tension (Appleby, 1984).
Although some additional nodulins are thought to play a role
in maintenance of nodule structure (Franssen gt gt., 1987),
the majority of the nodulins have no known function (Verma
_t _t., 1986). Since techniques have not yet been devised
for re-introducing into plants specific mutations generated
tg ttttg in cloned plant genes, it is not known whether any
of the nodulins are in fact necessary or essential for the
establishment of the symbiosis.

Although the number of plant genes induced specifically
in root nodules is known, a good estimate of the number of
bacterial genes involved in the symbiosis has yet to be
obtained. Nonetheless, molecular genetic techniques have
provided much more precise information on the nature of
bacterial mutations which affect the symbiosis. Naturally
occurring and chemically-induced ineffective strains were

initially used in the development of a genetic system for

rhizobia (summarized in Vincent, 1980). The demonstration

by Beringer gt gt. (1978) that the transposable element Tug
could be transferred to and would function in rhizobia led a
number of groups to use random transposon mutagenesis for
the isolation of metabolic and symbiotic mutants (e.g.
Buchanan—Wollaston gt gt., 1980; Cen gt gt., 1982; Rolfe gt
gt., 1980; Duncan, 1981; Meade gt gt., 1982; Hon gt gt.,
1984; Rostas gt gt., 1984; Chua gt gt., 1985; Applebaum gt
gt., 1985; Regensburger gt gt., 1986; Wilson gt gt., 1987).
The Tng-mutagenized populations were then screened for the
desired mutant strains by a variety of techniques.

The most intensively studied mutant strains have been
those which were unable to fix nitrogen (Fix‘) or were
unable to form root nodules (Nod'). The gt: genes are a
subset of the it; genes which have homologues in the
well-studied free—living diazotroph glebgiellg pneggonigg.
Considerable sequence similarity exists between the gtt
genes in 5. pneumoniae (reviewed in Dixon, 1984) and their
homologues in Rhizobium and ttgdyrhigobigg. This similarity
permitted the use of cloned g. pneumoniae gt: genes as
hybridization probes to identify the structural genes
encoding nitrogenase (e.g. Ruvkun and Ausubel, 1980;
Hennecke, 1981; Corbin gt gt., 1982), as well as other gtt
genes (e.g. Fuhrmann gt gt., 1985; Ebeling gt gt., 1987) and
a nitrogen fixation specific regulatory gene (e.g. Szeto gt
gt., 1984; Buikema gt gt., 1985; Fischer gt gt., 1986).

Nodulation (92;) genes were isolated by direct

complementation of Nod“ mutants (Long gt gt., 1982; Downie

_t _t., 1983; Schofield gt gt., 1984; Kondorosi gt gt.,
1984). The discovery that ggg genes and other genes
essential for nitrogen fixation (gt: and ttg genes) tended
to be clustered (Corbin gt gt., 1982; Long gt gt., 1982)
resulted in the use of site—directed mutagenesis (e.g.
Ruvkun and Ausubel, 1981) of cloned DNA fragments to
identify other genes involved in the symbiosis. Many groups
studying a wide variety of rhizobia and bradyrhizobia have
now used cloned genes as heterologous hybridization probes
followed by site—directed mutagenesis to identify and
characterize numerous gtt, ttg and ggg loci (reviewed in
Ausubel, 1984b; Ausubel gt gt., 1985; Downie and Johnston,
1986; and Kondorosi and Kondorosi, 1986).

Leigh gt gt. (1985) screened a population of
transposon—mutagenized fl. meliloti strains with the
fluorescent stain Calcofluor to detect alterations in the
production of extracellular polysaccharides (EPS). Finan gt
gt. (1985) screened a similarly mutagenized collection of
Rhizobium strains for loss of reactivity with a monoclonal
antibody specific for the rhizobial cell surface and for
insensitivity to a series of phages which infect Rhizobium.
The mutant strains identified in these experiments failed to
produce EPS and, when inoculated on plants, produced
ineffective nodules which contained no bacteria. These
observations indicate that exopolysaccharides play an

important role in the development of the symbiosis and

demonstrated that plant and bacterial differentiation could
be effectively uncoupled (Finan gt gt., 1985).

Physiological experiments indicated that the bacterial
microsymbiont provided the heme moeity of leghemoglobin, the
abundant nodulin involved in maintenance of low oxygen
tension (Nadler and Avissar, 1977). This notion that heme
metabolism might play an important role in the development
of the symbiosis was confirmed by Leong gt gt. (1982) who
isolated the R. geliloti hemA gene, which catalyses the
first committed step in heme biosynthesis. Strains carrying
mutations at this locus produced ineffective nodules on
alfalfa. Since rhizobia are obligate aerobes, mutations in
components of the electron transport chain might well be
expected to affect symbiotic competence. Guerinot and
Chelm (1986) demonstrated, however, that a deletion in the
gggg gene in t. jagonicum produced 5-aminolevulinic acid
(ALA) auxotrophy, yet the bacterial enzyme was not essential
for the formation of effective nodules. These results may
reflect host-dependent differences in the availability of
plant ALA to the bacterium or may indicate the presence of a
second as yet undescribed bacterial gene responsible for ALA
synthesis. Recently, O’Brian gt gt. (1987) screened
transposon-mutagenized t. japonicum on plates containing
Nadi reagent (Marrs and Gest, 1973) to detect mutants with
altered cytochrome oxidase activity. One cytochrome aaa
mutant was obtained (O’Brien and Maier, 1987) as were other

mutant strains which formed ineffective nodules and may

contain an insertion in a gene affecting heme biosynthesis
(O’Brian gt gt., 1987).

Hybridization of cloned genes from the tgrobgcterigg
tumefaciens chromosomal virulence region (ggt genes) to g.
meliloti gene banks identified regions of R. meliloti DNA
that exhibited homology to these probes. These regions were
then cloned and the corresponding chromosomal copies deleted
from the R. meliloti genome (Dylan gt gt., 1986).
Interestingly, these deletion strains produced ineffective
nodules suggesting that Agrobgctertgg and Rhizobium possess
structurally related genes which are necessary for the
establishment of infection whether it ultimately leads to
pathogenesis or to symbiosis.

An additional class of bacterial symbiotic mutants was
discovered during investigations into the carbon metabolism
of Rhizobium. Although mutations in a number of bacterial
genes involved in carbon metabolism had no effect on the
development of symbiosis (Ronson and Primrose, 1979; Duncan,
1981; Glenn gt gt., 1984) some mutants unable to transport
Cc-dicarboxylic acids (ggt mutants) in g. trifolii (Ronson

gt gt., 1981), R. leguminosarum (Finan gt 1., 1983; Ronson

a1.

1984; Arwas gt gt., 1985) and R. meliloti (Bolton gt

m o
H H'

., 1986; Engelke gt _t., 1987) formed ineffective nodules.
Since carbon availability may limit nitrogen fixation
(Hardy, 1975; Guerinot and Chelm, 1987), this carbon

transport system may play a pivotal role in determining the

efficiency of this process. The induction of the dctA gene

by growth on dicarboxylic acids was exploited by Ronson gt
gt. (1987) to screen a Tng-mutagenized population of g.
meliloti cells for mutants unable to activate expression of
a ggtt-tggt gene fusion. These mutants were shown to
contain Tng insertions in the regulatory gene gtgt and to
produce ineffective nodules on alfalfa. The gtgt gene

1., 1985)

product (NtrA) is a sigma factor (Hirshman gt
which confers specificity on core RNA polymerase for a
particular set of promoters. These promoters also require
activation by one of a group of regulatory proteins related
to the nitrogen regulatory protein NtrC (Nixon gt gt., 1986;
Ronson gt gt, 1987).

A final class of symbiotic mutants was discovered in
studies of auxotrophic mutants of Rhizobium and
gtgdyrhigobigg. Strains carrying mutations at some steps in
presumed pathways for amino acid biosynthesis are
symbiotically defective while strains carrying mutations at
other steps in the same pathway or in other pathways are
fully competent for symbiosis (e.g. Denarie gt gt., 1976;
Pain, 1979; Wells and Kuykendall, 1983; Hon gt gt., 1984;
Sadowsky gt gt., 1986). The mechanisms by which auxotrophy
affects the development of the symbiosis are unknown.

As the review presented above indicates, a considerable
number of genes essential for effective symbiosis have been
identified. The complexity of the process is reflected in
the variety of metabolic pathways which, when interrupted,

produce a symbiotic phenotype. Despite the fact that in a

lO

few instances the biochemical functions of some of these
gene products are well characterized (e.g. the nitrogenase

1., 1985), little is known about how

subunits, Brooks gt
plant and bacterial gene products interact to promote or
disrupt the development of the symbiosis. Several types of
experiments have been undertaken which attempt to address
this question. In one group of experiments the expression
of nodule-specific plant genes (nodulins) has been examined
when the infecting rhizobial strains carry particular
mutations. Although there is significant variation in
nodulin gene expression when different host plant/bacterial
mutant combinations are compared, nodules incited by most
Nif’ and Fix’ strains contain the majority of the nodulin
gene products seen in wild-type nodules (Lang-Unnasch and
Ausubel, 1985; Covers gt gt., 1985; Gloudemans gt gt.,
1987). Ultrastructural analysis of of nodules incited by
Nif‘ and FiX' strains supports this observation (Hirsch and
Smith, 1987). Even the Ex0' mutants which form neither
infection threads nor bacteroids (Finan gt gt., 1985) incite
nodules which contain six of the eleven nodulins examined
(Lang-Unnasch gt gt., 1986). Only mutations in the ggg
genes appear to completely eliminate nodulin gene
expression. Interpretation of experiments involving two
interacting organisms is at best difficult, especially given
the large numbers of genes being studied. Nonetheless, it
has been possible to deduce an approximate temporal

framework for the interaction of some bacterial and plant

11

gene products (Govers gt gt., 1985; Moerman gt 1., 1987;

Morrison and Verma, 1987; Gloudemans gt gt., 1987; Covers gt
gt., 1987; Lullien gt gt., 1987).

A second type of experimental approach utilized in
examination of plant/bacterial interaction during symbiosis
evolved from the observation that extracts from plants can
affect bacterial gene expression (Mulligan and Long, 1985).
The expression of the nodABC operon was shown to increase
markedly when both a functional gggg gene was present and
exudate from plant roots or seeds was added to the culture
medium (Mulligan and Long, 1985; Rossen gt gt., 1985; Innes

t 1., 1985). The active compounds from several plant

exudates have been purified and identified as substituted

flavones, isoflavones or flavanones (Peters gt _t., 1986;
Redmond gt gt., 1986; Firmin gt gt., 1986; Kosslak _t gt.,
1987). Although these compounds are present in many

families of plants and cannot therefore account for the
specificity seen in the fltgdyrhiggbtgg and Rhizobium/legume
interactions, it is the first example of communication
between the two symbionts to be characterized at the
molecular level.

The identification of the nodule-specific plant genes
was based on their differential transcription when
Rhizobium—infected and uninfected roots were compared (Auger
and Verma, 1981). In contrast, most of the bacterial genes
involved in the symbiosis were isolated by complementation

of mutant strains. The transcription of symbiotically

12

important bacterial genes has been examined directly in the
case of the nif genes (Corbin gt gt., 1983). The low levels

of expression of nod genes and fix genes makes the

 

transcripts difficult to detect and these genes have been
monitored primarily by the use of tggt fusions (Jacobs gt
gt., 1985; Aguilar gt gt., 1985). Two recent reports by
Renalier gt gt. (1987) and David gt gt. (1987) describe an
analysis of transcription of the entire 3. meliloti gig
plasmid, yet there are no reports in the literature
identifying symbiotically important bacterial genes solely
on the basis of their nodule-specific transcription.

The major focus of this thesis project was the
development of a method for isolating bacterial genes which
are transcribed specifically in nitrogen-fixing soybean
nodules incited by t. jagonicum. Differential hybridization
techniques (Sargent, 1986) have been used to isolate
transcriptionally-regulated genes in a variety of eukaryotic
organisms including Saccharomyceg cerevigtgg (St. John and
Davis, 1979), Xenopus 1aevis (Sargent and Dawid, 1983),
Aspergillus nidulans (Zimmerman gt gt., 1980) and giggg
sativa (Hershey gt gt., 1984). The regulated genes were
identified solely on the basis of differential transcription
in distinct cell types or developmental stages. Since this
approach had not been used previously to study development
in prokaryotes, it was adopted in order to isolate new
classes of genes involved in the g. japonicum/soybean

symbiosis which might not have been found using the methods

13

employed by other researchers. It was anticipated that the
highly transcribed gtt genes would be among those isolated
by such a method. It was also expected that additional
symbiotically induced genes would be identified which are
essential for growth and thus can not be mutated or are
non-essential for development and therefore have no
discernible phenotype when mutated. The promoter regions
from such genes might well lead to the identification of

regulatory mechanisms important to the development of the

Bradyrhizobium japonicum/soybean symbiosis.

CHAPTER 2

Isolation of Bradyrhigobiug japonicgg DNA Sequences

Transcribed Specifically in Bacteroids

Introduction

One key to understanding the genetic regulation of
development in biological systems is the identification of
genes which play a role in the developmental process. Two
general approaches have been employed in the search for such
genes: the isolation of mutants exhibiting a defect in the
developmental program and the identification of genes whose
expression is altered during development. The first method
can detect only those genes which are dispensable for normal
vegetative growth and which provide a detectable phenotype
when mutated. The second method is less restrictive in that
it can also identify essential genes and genes which are
developmentally regulated but have no readily apparent
mutant phenotype.

The study of development in prokaryotic systems has
traditionally depended primarily on the isolation of
phenotypic mutants (Dworkin, 1985). In Bacillus subtilis
chemical mutagenesis has been the major tool used to
identify more than fifty genetic loci which specifically
affect endospore formation (reviewed in Losick and Youngman,

1984) and several of the genes corresponding to these loci

14

15

have been cloned (Dubnau gt gt., 1981; Ferrari gt gt., 1982;

1., 1983). Some understanding of the genetic

Shimotsu gt
control of the Rhizobium/legume symbiosis has also been
achieved, principally through the use of random transposon
mutagenesis (reviewed in Rolfe and Shine, 1984). The
development of gene fusion techniques (Casadaban and Cohen,
1979) extended the use of transposon mutagenesis to allow
the identification of genes which are transcribed only under
specific conditions. Fusion techniques have proven useful
for a wide variety of organisms including Bacillus subtilis
(Perkins and Youngman, 1986), Myxococcgg xanthus (Kroos and
Kaiser, 1984), quloggcter crescentgg (Bellofatto gt al.,

1984), tgrobggterigg tumefaciens (Stachel gt 1., 1985) and

Saccharomyces cerevisiae (Ruby gt gt., 1983). Gene fusion
techniques were also employed to isolate sets of
DNA-damage-induced (gtg) genes in g. cerevisiae (Ruby and
Szostak, 1985) and t. subtilis (Love gt gt., 1985) which
were analogous to those isolated earlier from t. ggtt
(Kenyon and Walker, 1980). Although both chemical and
transposon mutagenesis have proven to be useful in the study
of development, neither technique can be used to identify
non-auxotrophic genes which are essential for growth of a
haploid organism.

Another general method for identifying developmentally
regulated genes involves differential hybridization

techniques (St. John and Davis, 1979) which compare mRNA

populations from physically separable cell types or

l6

developmental stages. Radioactively—labelled probes
prepared from these mRNAs have been used to screen genomic
or cDNA libraries for sequences which have altered levels of
expression. This approach has been successfully employed in
the study of differentiation in Saccharomyceg cerevigtgg
(Clancy _t gt., 1983), Asgergillus nidulans (Zimmerman gt
gt,, 1980), and Xenopus laevis (Sargent and Dawid, 1983).
The lack of an efficient method for purifying mRNA from
bacteria has limited the use of of differential
hybridization in the study of development in prokaryotes.
The principal problem encountered when hybridization probes
are prepared from total bacterial RNA is the high level of
background hybridization. This background hybridization is
probably due to homology between the stable RNA (mostly
rRNA) of the organism being investigated and the rRNA of the
E. ggtt cells being used to propagate the library. The
experiments described in this chapter were designed to
determine whether removal of sequences derived from reverse
transcription of rRNA (referred to hereafter as rcDNA) from
cDNA probes made from total bacterial RNA would make these
probes suitable for differential hybridization experiments.
Experiments described here demonstrate that when depleted of
rcDNA sequences, such probes can be used to identify genes
which are transcribed at different levels in two bacterial
cell types. The model system compares gene expression
levels in free-living cells of tgggyrhiggbigg japonicum to

that in purified bacteroids, a highly differentiated form of

17

the soybean microsymbiont isolated from nitrogen-fixing

nodules.

Materials and Methods

Chemicals ggg Enzymes. Hydroxylapatite and Bio-gel P-60
were purchased from Bio—Rad (Richmond, CA); reverse
transcriptase from Life Sciences, Inc., (St. Petersburg,
FL); lysozyme, ribonulcease A and deoxyribonuclease I from
Sigma Chemical Co. (St. Louis, MO); deoxyribonucleic acid
(isolated from salmon testes) from Pharmacia P-L
Biochemicals (Piscataway, NJ); formamide GR from EM Science
(Gibbstown, NJ) and bovine serum albumin, Faction V from
U.S. Bichemical Corp. (Cleveland, OH). These products were

used according to the specifications of the manufacturer.

Bacterial Strgtgg. The tggherichig coli K-12 strain ED8654
tggt ggt hsdx supE gupF) (Borck gt gt., 1976) was used for
routine plasmid construction, maintenance and purification.
The lambda phage library was propagated on E. ggtt strain
K802 Ltggt ggt tggflt ggtfi gggg). BJllOd, a small colony
isolate of Bradyrhizobium japonicum USDA 311b110 (Kuykendall
and Elkan, 1976; Meyer and Pueppke, 1980), was used in all

experiments.

Plasmids and Phage. Plasmid pRJ676-l contains 2.1 kb of

upstream sequence and the 5’ portion of the coding region

18

for gttg, one of the nitrogenase structural genes (Adams and
Chelm, 1984). Plasmid pBJ53A contains the entire coding
region for glutamine synthetase, gtgt, (Carlson gt gt.,
1985). Plasmid pBJllO contains the entire coding region for
5-amino levulinic acid synthase gggt (Guerinot and Chelm,
1986). Plasmid pRNA404 contains the entire ribosomal RNA
operon from Anacystis nidulagg (Golden and Sherman, 1983).
Plasmid pLAFRI is a wide host range cosmid vector (Friedman
_t _t., 1982). The t. japonicum genomic library in
bacteriophage lambda was constructed as described (Carlson
gt _t., 1985). Total g. japonicum genomic DNA, partially
digested by restriction endonuclease gggI, was ligated with
lambda cloning vector BFlOl DNA (Maniatis gt gt., 1982)
which had been digested with restriction endonuclease gggHI.
After ligation, the DNAs were packaged tg yttgg into lambda
phage particles as described (Hohn, 1979) and plated on g.
ggtt strain K802. 1.4 x 105 plaques were obtained.

Assuming that the t. japonicum genome is no larger than 105
kb and that the minimum insert size in the recombinant phage
is 6 kb, the complete genome should be represented with a
probability of 0.99 by 7,700 phage. The t. japonicum
genomic library in cosmid vector pLAFRl was constructed as
described (Adams gt gt., 1984). Total t. japonicum genomic
DNA, partially digested with restriction endonuclease EggRI,
was ligated with EggRI-cleaved, phosphatase—treated pLAFRl
DNA at a ratio of 1:5 (Maniatis gt gt., 1982). After

ligation, the DNAs were packaged tg vitro into lambda phage

19

particles as described (Hohn, 1979) and E. ggtt strain
ED8654 was infected. Tetracycline resistant transductants
were replicated in an ordered array in 96-well polycarbonate
microtiter dishes containing LB medium with 15% glycerol
(v/v), incubated at 37°C for 24 h and frozen at -80°C. For
screening, the cosmid-containing isolates were replica
plated onto cellulose nitrate filters which had been placed

on agar-solidified media. Plates were incubated at 37°C.

Bacterial Growth Conditiong. t. japonicum was grown at 30°C
in yeast extract broth as described (Adams gt gt., 1984)
except that mannitol was substituted for xylose. t. ggtt
strain K802 was grown at 37°C in TB medium (Miller, 1972).
g. ggtt strain ED8654 was grown in LB medium (Davis gt gt.

1980) at 37°C.

Nodule Bacteria Egggtiongtion. The total t. jagonicum
population from frozen 5-6 week-old soybean nodules was
isolated as described by Adams and Chelm (1984). Three
developmental fractions representing bacteroids,
transforming bacteria and nodule bacteria were isolated on a
discontinuous sucrose gradient as described by Ching gt gt.
(1977) scaled up for use in a Beckman l4 Ti zonal rotor as
described by Carlson gt gt. (1985). In each of two separate
experiments, 150 grams of modules were macerated and the
bacterial cells subjected to two rounds of sucrose gradient

purification. This procedure yielded an average of 1.0, 1.1

20

and 3.5 grams of cells in the bacteroid, transforming
bacteria and nodule bacteria fractions respectively, which
in turn yielded an average of 0.7, 1.2 and 5.1 mg of total

RNA.

RNA Methods Total bacterial RNA was isolated as described by
Adams and Chelm (1984). Discontinuous CsCl gradients
consisting of a 5.0 ml cushion of 5.6 M CsCl overlayed with
no more than 13.4 ml of cell homogenate was subjected to
centrifugation at 18°C in a Beckman SW-28 rotor at 22,500
rpm for 36-40 hours. Gels for northern hybridizations were
prepared as described (Maniatis gt gt., 1982). RNA (1 ug)
was size fractionated by electrophoresis on a denaturing
0.9% agarose gel made using buffer containing 6X (v/v)
formaldehyde; 20 mM sodium 3-[N-morpholino]propanesulfonic
acid (MOPS), pH 7.0; 5 mM sodium acetate; and 1 mM
(ethylenedinitrilo)-tetraacetic acid (EDTA). The RNA
samples were dissolved in gel buffer which also contained
60% (v/v) formamide and 5x (v/v) glycerol. The samples were
heated to 55°C for 15 minutes prior to loading on the gel
which was subjected to electrophoresis at 100 volts (35 mA)
for four hours. During electrophoresis the gel buffer was
recirculated at a rate of 5 ml/min using a peristaltic pump.
After electrophoresis, a portion of the gel was stained with
ethidium bromide for visualization of the RNA. The RNA in
the remaining portion of the gel was transferred onto

Biodyne membrane (Pall Ultrafine Filtration Corp., Glen

21

Cove, N.Y.) by the method of Southern (1975) using 20X 886
buffer (1 x 880 = 0.15 M NaCl, 0.015 M sodium citrate [pH
7]). The filter was dried at 80°C tg ygggg for 60 min and
hybridized with radioactively labelled probes under the
following conditions which were specified by the
manufacturer. Hybridization buffer contained 50X formamide
(v/v), 5X SSC, 1X Denhardt’s solution (0.02% Ficoll [w/v],
0.028 polyvinylpyrrolidone [w/v], and 0.023 bovine serum
albumin Pentax Fraction V [w/v] in water), 30 mM Tris (pH
7.5), 0.1x sodium lauryl sulfate (SDS), 1 mM EDTA, and 100
ug/ml denatured salmon sperm DNA. The filter was
pre-hybridized with hybridization buffer (0.1 ml/cmz) for.
four h at 37°C. Fresh buffer and approximately 3 x 10° cpm
of 32P-labelled DNA probe was added and hybridized for 16-20
h at 37°C. After hybridization, the filters were washed in
200 m1 of 2X SSC, 0.1x SDS at room temperature for twenty
min with four changes of buffer followed by washing in 200
ml of 0.1x SSC, 0.13 SDS at 42°C for 1 h with four changes
of buffer. The radioactively labelled probes adhering to

the filters were then visualized by autoradiography.

Qflt Methods. Isolation and purification of recombinant
lambda phage and extraction of phage DNA were performed as
described (Davis gt gt., 1980). Plasmid DNA was isolated
from g. ggtt by CsCl-ethidium bromide centrifugation
(Clewell and Helinski, 1972). Total genomic DNA from t.

jagonicum was isolated by a modification of the method of

22

Marmur and Doty (1962). Five ml of a culture of g.
japonicum grown in the presence of 10 mM KNOa was made to
0.1% (w/v) with N-lauroylsarcosine, mixed briefly by
vortexing and the cells pelleted by centrifugation. The
cell pellet was resuspended in a solution of 25% sucrose
(w/v) to which was added 200 ul of a 10 mg/ml solution of
lysozyme in water. After incubation for 3 min at room
temperature, 3 m1 of 25% sucrose (w/v) 185 ul of 0.2 M EDTA
were added. The cells were lysed by the addition with
gently mixing of 0.5 ml of a 10% (w/v) solution of
N-lauroylsarcosine and RNA was hydrolysed by the
simultaneous addition of 1 ul of a 10 mg/ml solution of
ribonuclease A in water. After cell lysis was complete as
indicated by the clearing of the solution, the lysate was
extracted with 4 ml of phenol which had been redistilled and
equilibrated with water and 4 ml of chloroform. After
mixing the aqueous and organic phases were separated by
centrifugation at 3,000 x g for 10 min at 4°C. The aqueous
phase was retained and re-extracted with phenol/chloroform
as outlined above until a whitish interface was no longer
visible between the two phases. The aqueous phase was then
extracted twice with 4.0 m1 of chloroform alone and then the
nucleic acids precipitated by the addition of one tenth
volume of 3 M sodium acetate (pH 5.2) and two volumes of
100% ethyl alcohol. The DNA pellets were resuspended in an
appropriate volume of 10 mM Tris (pH 8) and 1 mM EDTA. DNA

was transferred to cellulose nitrate filters from agarose

23

gels by the method of Southern (1975), from phage lambda
plaques by the method of Benton and Davis (1977) and from t.
ggtt or t. japonicum colonies by the method of Grunstein and
Hogness (1975). Hybridization of radioactively-labelled
probes to the DNA affixed to the filters, washing of the
filters and autoradiography was performed as described
(Adams and Chelm, 1984; Adams gt gt., 1984). Before
hybridization, filters were incubated for 1 h at 65°C in 5X
Denhardt’s solution, 5X SSPE (1X SSPE = 0.18 M NaCl, 10 mM
NaPO¢ [pH 7.7], 1 mM EDTA), and 100 ug/ml sheared and
denatured salmon sperm DNA. 32P-labelled hybridization
probes were hybridized with the filters for 20 to 24 h at
65°C in 5X SSPE, 1.5x Denhardt’s solution and 100 ug/ml
salmon sperm DNA. After hybridization, filters were washed
for 30 min at room temperature in 200 ml of 2x SSPE
containing 0.1% SDS (w/v) with one change of buffer followed
by washing for 30 min at room temperature in 200 ml of 0.1x
SSPE containing 0.1% SDS (w/v) with one change of buffer.
The isolation of DNA fragments for use as hybridization

probes has been described (Adams and Chelm, 1984).

ggflt Synthesis. The DNA polymerase isolated from RNA tumor
virus particles is able to efficiently synthesize
complementary DNA (cDNA) if a suitable primer is hydrogen
bonded to an RNA (or DNA) template (Taylor gt gt., 1976).
Random oligonucleotide primers were prepared by DNase I

digestion of salmon sperm DNA as described elsewhere

24

(Maniatis gt gt., 1982). RNA (1 ug) and 125 ug of primer
DNA were combined in 12 ul of water, heated to 95°C for 10
minutes and then allowed to cool to room temperature. The
final reaction conditions for cDNA synthesis in a 25 ul
volume were: 50 mM Tris-H01 (pH 8.0), 1 mM dithioerythritol
(DTE), 5 mM MgClz, 40 mM KCl, 1 mM each of dATP, dGTP, dTTP
and 50 uM dCTP. 50 uCi of 32P-1abelled dCTP (15.17
TBq/mmole, Amersham Corp., Arlington Heights, IL.) and 30
units of reverse transcriptase (Life Sciences, St.
Petersburg, FL.) were then added and the reaction incubated
at 37°C for 60 min. The synthesis reaction was stopped by
the addition of 2 ul of 0.5 M EDTA (pH 8.0). The template
RNA in the RNAzDNA hybrids resulting from this reaction was
hydrolysed by the addition of 2.5 ul of 3.0 M KOH followed
by incubation at 37°C for 18 h. Unincorporated and newly
hydrolysed nucleotides were removed by chromatography on a 2
ml Bio-gel P-60 column equilibrated in 10 mM Tris-H01 (pH
8.0), lmM EDTA. A cDNA synthesis reaction typically yielded
approximately 0.5 ug of 32P-labelled cDNA having a specific
activity of 5 x 107 cpm/ug. When sized by electrophoresis on
5% acrylamide-8 M urea gels, the cDNA had an apparent size
range of 50 to 400 nucleotides. To assess the ability of
the cDNA to efficiently hybridize with RNA, RNA-driven
solution hybridization experiments were conducted as
described by Chelm and Hallick (1976). Approximately 85% of

the cDNA hybridized with an apparent Roti/z of 0.003

25

indicating that the cDNA represented primarily the abundant,

stable RNA species.

Solgtion Hybridigation ggg Qggpetition Hybridiggtion.
Approximately 0.5 ug of radioactively-labelled cDNA (3-4 x
107 cpm) was hybridized with 100 ug of pBJl42 plasmid DNA
which had been sheared by four ten-second bursts at 50 watts
on a Sonifier Cell Disruptor equipped with a microtip (Heat
Systems-Ultrasonics, Inc., Plainview, NY). A sample of the
sheared plasmid DNA was size fractionated on an agarose gel
and had an apparent modal size of 500 bp. The sheared
plasmid DNA and the cDNA in water were heated to 95°C in a
tightly capped microfuge tube for ten minutes, then placed
at 74°C in buffer containing 0.3 M NaCl, 5 mM Tris-HCl (pH
7.4), and 1 mM EDTA as described by Chelm gt gt. (1977).

The temperature used in the hybridizations (74°C) was 25°C
below the melting temperature (Tm) of the t. japonicum DNA
as estimated by the method of Bishop (1972) using a guanine
plus cytosine (G+C) content of 64% as determined by Elkan
and Usanis (1971). (Tm = 16.6 logM + 0.41[G+C] + 81.5 where
M = the concentration of salt in the reaction mixture). The
hybridization reaction was allowed to proceed to a Cot of 1
(molar nucleotides - seconds) in order to permit the
radioactively-labelled rcDNA sequences to hybridize to
complete kinetic termination with the excess unlabelled rDNA
plasmid DNA. In a typical solution hybridization reaction,

the driver rDNA plasmid DNA was present at a concentration

26

of 450 ug/ml of which only 31.5% (7.5 kb of one DNA strand
from 11.9 kb of double-stranded plasmid) is effective driver
DNA. This effective driver DNA concentration (142 ug/ml) is
equivalent to 4.3 x 10" molar nucleotides (1 ug/ml DNA =
3.03 x 10‘5 M nucleotides). To achieve a Cot = 1 (Co =
nucleic acid concentration in molar nucleotides, t = time in
seconds), the nucleic acids must be hybridized for 1/4.3 x
10'“ seconds = 38.8 min. Double-stranded nucleic acids were
then removed by chromatography on a 1.0 ml hydroxylapatite
column at 60°C. The column was washed six times with 1.0 m1
aliquots of 0.12 M sodium phosphate (pH 6.8) at 60°C to
elute the single-stranded nucleic acids. Aliquots of each
fraction were analysed by liquid scintillation counting and
the peak 32P-containing fractions were pooled.

For lambda library screening experiments, rDNA plasmid
DNA was used to block the hybridization of any rcDNA
sequences remaining in the probes with the Q. ggtt rDNA or
rRNA had been present in the plaques and was transferred to
the cellulose nitrate filters. The pBJl42 plasmid DNA was
sheared and denatured as outlined above and included at 50
ug/ml in both the pre-hybridization and the hybridization

solutions.

27

Results

Isolation and Characteriggtion gt the rDNA Genes gt t.

 

japonicum. A genomic library of t. japonicum DNA was
screened for recombinant phage carrying the rDNA genes by
plaque hybridization. The filters were probed using 5 x 106
CPM of cDNA synthesized using total RNA isolated from
purified bacteroids as template. As expected, the probe
hybridized to all plaques since g. ggtt rDNA is always
present. However, four individual plaques gave signals
which were well above this constant background (Fig. 1).
Restriction endonuclease digests of phage DNA purified from
these plaques revealed two common EggRI fragments and
several different EggRI fragments. One phage (LrDl) was
further characterized by hybridization of bacteroid cDNA to
Southern blots of phage DNA digested with various
restriction enzymes to better localize the region containing
the rDNA genes. The region hybridizing to the probe was
found to reside on a 7.6 kb figng fragment. This fragment
was subsequently subcloned into the tgng site of plasmid
pBR322 resulting in plasmid pBJl42.

To verify that the cloned DNA actually contained rDNA
sequences, hybridization probes were prepared by
nick-translation of fragments of the plasmid pRNA404 (Golden
and Sherman, 1983) which contains a copy of the rDNA operon

from Anacystis nidglans. These probes permitted

28

Figure l. Genomic library of B6 jgpgnicum 110d in lambda
vector BFlOl probed with 5 x 10 CPM of cDNA synthesized
from bacteroid RNA. Filter A is from the initial screening
of the library and Filter 3 is from a re-screening for
plaque purification. Arrows point to plaques containing the
rDNA operon.

Figure 2. Northern blot of total B. japgnicum 110d RNA (1
ug and 0.5 ug in adjacent lanes) probed with (A) the 1.3 kb
BglII fragment of plasmid pBJl42 which contains a portion of
the 165 RNA coding region or (B) the 1.4 kb PstI fragment of
plasmid pBJl42 which contains a portion of the 23S RNA
coding region. (See Fig. 3 for restriction map of plasmid
pBJl42.

Figure 3. Restriction map of the B. japoniggm rDNA operon
cloned on a 7.5 Kb figmHl fragment inserted into cloning
vector pBR322 to make plasmid pBJl42. A . _ygl, B - figmHl,
Bg - 39111, E - __QR1, K - Kp_l, P - Egtl, Rv - EQQRV, S -

 

§p_1,-_hgl.No restriction sites for ngl, §gll, or
Ital
X X
B AE S BQRVE P P A 89K A 89 B
I II I l l l l l l I l l I l
(235) (1651 | |
1Kb

Figure 4. Determination of B. jgpgnjggm rDNA operon copy
number. Total genomic DNA of 5. jgponigum digesteg with
various restriction enzymes and probed with 5 x 10 cpm of
randomly primed cDNA synthesized using RNA isolated from B.
jgponiggm mcells grown in culture. Arrows are placed beside
bands expected to be internal to the rDNA operon. Bg -
39111, E - EQQRI, Rv - figQRV, P - Egtl.

 

3O

localization of 16s and 23s RNA homologies within the t.
japonicum clone. The DNA fragments corresponding to these
two t. japonicum regions were then labelled and used to
probe northern blots of total t. jagonicum RNA (Fig. 2).
These results confirmed that pBJl42 contains rDNA coding for
16s and 23s RNA. A restriction map is shown in Figure 3.

To determine the rDNA operon copy number of t.
japonicum, total genomic DNA from g. jagonicum was digested
with restriction enzymes known to cut within the rDNA coding
region. These digests were blotted onto cellulose nitrate
filters by the method of Southern (1975) and probed with
plasmid pBJl42 labelled with 32F by nick translation. The_
existence of multiple rDNA operons would result in multiple
restriction fragments hybridizing to the rDNA probe by
virtue of the differences in DNA sequence flanking the rDNA
operon(s). Only rDNA containing fragments of predicted
sizes were detected indicating that g. japonicum contains a

single rDNA operon (Fig. 4).

Preparation gt rcDNA-deplgteg,gtflt Probes. To assess the
efficiency with which a particular protocol removed rcDNA
sequences from random-primed cDNA, probes prepared by
different methods were hybridized to digests of plasmids
containing known t. jagonicum genes which had been
transferred to cellulose nitrate filters by the Southern
transfer method. The three genes utilized were: (i) gttfl,

which encodes component II of nitrogenase (Adams and Chelm,

31

1984), (ii) the rDNA genes of g. jagonicum, (iii) either
gggg which encodes 5-amino levulinic acid synthase (Guerinot
and Chelm, 1986) or gtgg which encodes glutamine synthetase
(Carlson gt gt., 1985). The gttfl gene is known to be
expressed at high levels and the gggg and gtgt genes at low
levels in bacteroids (Adams and Chelm, in press; J.
Somerville, personal communication). Various protocols were
compared for their ability to reduce the hybridization
signal from the rDNA plasmid (pBJl42) to a level below that
seen from the other known genes, indicating that the probe
had been depleted of rcDNA sequences.

Three different methods for removing rcDNA sequences
from the cDNA pool were devised. In one method, the rDNA
plasmid (pBJl42) was sheared by sonication, denatured by
heating and included as a competitor in both the
pre-hybridization and hybridization solutions at 10 ug/ml or
50 ug/ml (Fig. 5A). The second method involved the
hybridization of cDNA, in solution, to an excess of sheared
and denatured pBJl42 (rDNA) plasmid to a Cot of 1. Since
the Rot1/2 of the fast-reacting component of the cDNA
(presumably the rcDNA) is 0.003, these conditions permitted
the rcDNA sequences to hybridize to complete kinetic
termination. This entire reaction mix was then used as
probe or, alternatively, aliquots were chromatographed on an
hydroxylapatite (HAP) column to remove the double-stranded
rDNA:rcDNA hybrids. when HAP column chromatography was

performed, 10-15% of the radioactive label eluted in the

32

Methods for lowering the hybridization signal from rcDNA.

Figure 5A. Method 1. Restriction endonuclease
digests of plasmids containing the rDNA (lane 1), glnA
(lane 2) and nifH (lane 3) genes hybridized with cDNA.
The control filter in panel A was hybridized with 5 x
10 cpm of randomly primed cDNA. The filters in
panels 8 and C were hybridized in an identical manner,
but with 10 and 50 ug/ml, respectively, of competitor
rDNA plasmid DNA included in the pre-hybridization and
the hybridization solutions.

Figure 58. Method 2. Restriction endonuclease
digests of plasmids containing the rDNA (lane 1), glnA
(lane 2), and gifH (lane 3) genes hybridized with
cDNA. The central filter in panel A was hybridized
with 5 x 10 cpm of unfractionated cDNA. The filter
in panel B was hybridized with 5 x 10 cpm of cDNA
which had been previously hybridized with sheared and
denatured pBJl42 DNA to a Cot - 1. In panel C the
filter was hybrigized with the sgngle-stranded
fraction (7 x 10 cpm) of 5 x 10 cpm of cDNA which
had been previously hybridized to plasmid pBJl42 DNA
to a Cot - l and then fractionated on a
hydroxylapatite (HAP) column.

Figure 5C. Method 3. Restriction endonuclease
digests of plasmids containing the rDNA (lane 1), hgmA
(lane 2), and gij (lane 3) genes hybridized with
cDNA. The gantrol filter in panel A was hybridized
with 5 x 10 cpm of unfractionated cDNA. The filter
in panel B was hybridized with cDNA which had been
previously hybridized in solution with rDNA plasmid
DNA and chromatographed on HAP as in panel C of Figure
58. The filter in panel C was hybridized with cDNA
treated as in panel B, but with the addition of 50
ug/ml of competitor rDNA plasmid DNA in the
pre-hybridization and the hybridization solutions.

The glnA gene was replaced by the hemA gene in this
figure since the latter gave a more readily detectable
signal. In Figure 5C only, the autoradiograms in
panels 8 and C were exposed six-fold longer than that
in panel A.

 

23 123

c

 

1 2 3 1
g I
5A.
rDNA a?
ram car

   
  

Control 10 uglml 50 uglml
53 1* ° °
I § ‘
rDNA W 0
mm w .' T‘ T
Control Cot :1 Cot=1+HAP
5c ‘

rDNA w

”0 V
. .4 .3"

Control CotflOHAP Cot sHHAP+50 uglml

hemA a?
mm ‘3'

34

low-salt, single-stranded fraction and was used as probe
(Fig. 58). Finally, since neither of the first two methods
lowered the hybridization signal to the rDNA plasmid to an
acceptable level, the three techniques of solution
hybridization, HAP chromatography and competition solution
hybridization were performed sequentially. This final
approach did succeed in lowering the hybridization signal
from the rDNA plasmid in Southern blot analyses to below
that of the gttfl signal and near that of the low abundance
gggg_signal (Fig. 5C, panel c). A method has thus been
developed which allows the detection of relatively low level
gene expression using randomly primed cDNA prepared from
total bacterial RNA. Artifactual hybridization bands are
visible in several lanes in Figure 5 (e.g. Fig. 5B, panel B,
lanes 2 and 3). Hybridization in solution of rcDNA with
sheared, denatured plasmid pBJl42 DNA produces some
double-stranded hybrids which also contain single-stranded
plasmid pBR322-derived sequences. The hybridization of
these vector sequences (and the attached, labelled rcDNA
sequences) to vector sequences on the filters produces the
artifactual signals. Removal of the labelled hybrids by HAP
chromatography results in removal of the artifactual signals

(e.g. Fig. 5B, panel C, lane 3).

§creenigg the Phage Library for figcteroid-gpecific
Trangcription. The t. japonicum genomic library was

screened for recombinant phage carrying g. jagonicum DNA

35

sequences expressed specifically in bacteroids.
Approximately 8,000 phage were screened by plaque
hybridization. The probe used was cDNA synthesized using
total bacteroid RNA as template. The background
hybridization signal resulting from high levels of rcDNA was
reduced by enriching for mRNA by sequential solution
hybridization, HAP column chromatography and competition
hybridization as described above. Thirty phage were
isolated which gave signals above background. These thirty
phage were dotted in an ordered arrays onto a TB medium
solidified with agar and seeded with t. ggtt strain K802,
incubated overnight at 37°C, and then transferred to
cellulose nitrate filters. Plaque hybridization was
performed on these filters using cDNA synthesized with
either total RNA from purified bacteroids or total RNA from
a pure culture of g. japonicum 110d grown in rich medium
(YEM) as template. 0f the 30 phage analysed, 16 appeared to
contain sequences which were expressed at higher levels in
bacteroids than in free living cells. An analysis of six
recombinant phage in which four give stronger signals when
probed with cDNA prepared from bacteroids than with cDNA
prepared from cells grown in culture is presented in Figure
6.

To further examine the expression of these sequences,
purified DNA was prepared from each of the 16 retained

phage, digested with appropriate restriction enzymes, size

fractionated by electrophoresis on agarose gels and then

 

Figure 6. Rescreening of recombinant bacteriophage
containing 8. sequences which are expressed
specifically in bacteroids. Arrays of recombinant phage
were dotted onto T8 medium which had been solidified with
agar and seededowith £. 9911 strain K802, the plates
incubated at 37 C overnight and DNA from the resultant
plaques transferred to cellulose nitrate. Two filters were
prepared from the same array of phage. Filter A was probed
with mRNA-enriched cDNA synthesized using 8.

isolated from purified bacteroids as template and filter 8
was probed with mRNA-enriched cDNA prepared from RNA
isolated from B. jgpgnigum cells grown in rich medium. The
plaques labelled a - f were produced by recombinant lambda
clones 816, 817, 822, 824, 826 and 827. Control plaques g
and h were produced by recombinant phage NH-l which contains
the 8. niffl gene and by lambda cloning vector
8F101. The first four recombinant phage in the array
presented above (816, 817, 822 and 824) were classified as
containing sequences induced in bacteroids and the latter
two (826 and 827) were not further characterized. (See
Table l for a list of phage and corresponding cosmids).

transferred onto cellulose nitrate. Duplicate filters were
probed with equal amounts of cDNA synthesized from either
bacteroid RNA or RNA from free-living cells. The cDNA
probes had been depleted of rcDNA sequences by solution
hybridization and HAP column chromatography. The results
confirmed that the sixteen phage analysed contained
sequences which are expressed at higher levels in bacteroids
than in free living cells. An analysis of seven recombinant
phage which contain sequences transcribed at higher levels
in bacteroids than in cells grown in culture is presented in

Figure 7.

Identificgtion gt Cosmids Containing Sequences Expressed

Specifiggllx 1g B.

japonicum Bacteroids. To facilitate
comparisons between the g. jagonicum sequences carried on
the recombinant phage and previously cloned genes, DNA from
each of the 16 phage described above was labelled by
nick-translation and the labelled phage pooled in groups of
four or five. These pools were used to probe an ordered t.
jagonicum genomic library constructed in the wide host range
cosmid vector pLAFRl. Thirty-nine cosmids which gave
positive hybridization signals were then replated in an
ordered array and colony hybridizations were performed using
each of the individual nick-translated phage singly as
probe. Ten groups of cosmids were thus identified which

contained g. japonicum sequences homologous to the phage

inserts (Table 1). All thirty-nine cosmids were also

-h-
fir.-

Figure 7. Restriction endonuclease digests of DNA purified
from seven recombinant bacteriophage was size-fractionated
on an agarose gel and transferred to cellglose nitrate. The
resultant filters were probed with 5 x 10 cpm of
mRNA-enriched cDNA synthesized using RNA isolated from
either (A) purified B. Jingniggm bacteroids or (B) B.
jgpgnjggm cells grown in culture. Lanes 1-7 on each filter
contain DNA from recombinant bacteriophage 815, 816, 817,
819, 820, 822 and 824. (See Table l for a list of phage and
corresponding cosmids).

39

Table 1. List of recombinant lambda phage containing 3.
japonicum sequences that are expressed specifically in
bacteroids and the corresponding members of the cosmid
library of g. jagonicum containing these same sequences.
The underlined cosmids were selected as representative of
each group.

 

 

 

 

 

Corresponding Plasmid
Phage t Cosmid(s) Subclones

81 7-39, 7-75

82 11-78

85 4-51, 5—35, 7—41 p8J273
8—10, 10-91

86 15-75

89 7-39 (same as 81)

815 none p8J27O

816 2-56, 3-61, p8J214
13-36, 15-68 p8J216

817 none

818' same as 816

819 14-8 p8J227

820 same as 85 i

821 8-17, 8-62

822 1-20

824 4-14, 5-14, 6-94, 7-36, 7-92

8-11, 8-68, 8-76, 10-90, 12-66
(contain previously identified
nifA and fixA genes)

825 12-37

830 same as 824

Number of recombinant phage containing sequences

exhibiting bacteroid-specific expression = 16
Number of unique phage = 12
Number of cosmid groups identified by unique phage = 10

screened by colony hybridization using plasmids containing
previously identified gtt genes (Hennecke, 1981; Adams gt
gt., 1984) and ttg genes (Lamb and Hennecke, 1986). One of
the ten groups of cosmids carrying sequences transcribed at
high levels in bacteroids was shown to contain the ttg
genes, while the remaining nine groups appear to define new
genes or operons which are expressed at higher levels in

bacteroids than in cells grown in rich medium.

Discussion

Differential hybridization is a powerful technique
which has been used successfully to isolate developmentally
regulated genes in Saccharomyceg cerevisiae (Clancy gt gt.,
1983), Asgergillus nidulans (Zimmerman gt gt., 1980) and
Xenopus laevis (Sargent and Dawid, 1983). Experiments
described in this paper extend the use of this technique to
the prokaryotic partner in the Bradyrhizobium
japonicum/soybean symbiosis. Differential hybridization has
traditionally been used primarily in eukaryotic systems
where techniques exist for separating messenger RNA from
rRNA and other stable RNA species (Aviv and Leder, 1972).
The hybridization kinetics of the cDNA synthesized in our
experiments to total bacterial RNA (85% of labelled cDNA
hybridizing with a rot1/2 = 0.003) confirm that the products

synthesized consist mostly of sequences homologous to

41

abundant stable RNA species. Removal of the sequences
homologous to the rDNA operon (rcDNA) by sequential solution
hybridization, HAP chromatography and competition
hybridization was sufficient to permit hybridization signals
from highly transcribed mRNAs to be visualized above
background. This result suggests that the rDNA homology
between t. japonicum and E. ggtt can account for most of the
high level of background hybridization which has previously
prevented the use of differential hybridization.

In a recent report Gianni and Galizzi (1986) described
the isolation of three genes expressed preferentially during
Bacillus subtilis spore outgrowth. These investigators used
excess vegetative RNA to compete out background.
hybridization when a lambda phage library was probed with
end—labelled RNA from outgrowing spores. This method
provides an example of how differential hybridization has
been used to isolate genes which are expressed only under
specific conditions in prokaryotes. To examine the possible
use of this methodology for investigating the g.
jagonicum/soybean symbiosis, the hybridization signals
produced by probes prepared by end-labelling RNA were
compared with those produced by the cDNA method outlined
above and were seen to be distinctly lower. Improvements in
the sensitivity of these methods, however, should allow the
identification of developmentally regulated genes.

We have isolated ten groups of cosmids which contain t.

japonicum sequences which are transcribed at higher levels

42

in nitrogen-fixing bacteroids than in cells grown in
culture. One group contains previously identified genes for
nitrogen fixation (Lamb and Hennecke, 1986), but the
remaining eight groups represent transcribed regions whose
functions are as yet unknown. In g. subtilis, mutational
analysis has identified more that fifty loci which are
necessary for endospore formation (reviewed in Youngman gt
gt., 1985). Since the differentiation of g. japonicum
bacterial cells into nitrogen-fixing bacteroids presumably
involves a large number of genes, the highly transcribed
sequences isolated in this study probably represent only a
small subset of the total number of genes which are
transcritionally regulated during development. A better
estimate of the number of genes involved in g. jagonicum
bacteroid development may be derived if additional methods
can be found which further reduce the background
hybridization signal resulting from rcDNA sequences.

The methods utilized to obtain these differentially
expressed DNA sequences permits the isolation of both
essential genes and genes which have no obvious phenotype
when mutated. To analyze the function of these sequences
genetically, g. jagonicum strains were constructed in which
several of the regions containing the differentially
expressed sequences were individually deleted from the t.
japonicum chromosome. In the following chapter the

properties of these deletion derivatives are described.

CHAPTER 3

Genetic Analysis of Bradyrhiggbium japonicum DNA Sequences

That Are Transcribed Specifically in Bacteroids

Introduction

Although natural variants and chemically-induced
mutants have been used in genetic analyses of the
Bradyrhizobium japonicum/soybean symbiosis (Vincent, 1980),
the most widely employed technique for identifying genes
required for symbiosis has been transposon mutagenesis
(Berg, 1977; Rolfe and Shine, 1984; de Bruijn and Lupski,
1984; Ditta, 1985). In Chapter 2 an alternative method for
isolating genes specifically expressed in bacteroids was
described. Differential hybridization was employed to
identify sequences which are expressed specifically in g.
japonicum bacteroids. This approach made the assumption
that the rate of transcription of at least some
symbiotically important genes changes during the
developmental process. Examples of both repression and
derepression of gene expression during development have been
reported in a number of different eukaryotic systems (e.g.
Firtel, 1972; Zimmerman gt gt., 1980; Putzer gt_gt., 1984;
Kurtz and Lindquist, 1984; Kopachik gt gt., 1985).

Auger and Verma (1981) demonstrated that in the case of

the Rhizobium/legume symbiosis, a set of plant genes (termed

43

nodulins) were specifically induced in developing nodules
and were not present in uninfected roots. Regarding the
bacterial partner of this symbiosis, it has been
demonstrated that the genes encoding the subunits of the
nitrogenase enzyme are highly induced during nodule
development (Corbin gt gt., 1982). Studies in which western
blots of total protein from bacteroids and free-living cells
were probed with polyclonal antibodies raised against
bacteroids demonstrated that a number of additional gene
products are specifically induced in bacteroids (Verma gt
gt., 1986). The likely importance of developmentally
regulated gene expression prompted efforts to adapt
differential hybridization techniques for use in studying
the gggdyrhigobigg_jagonicum/soybean symbiosis. Techniques
which permit the deletion of particular DNA sequences were
employed to examine the effect of removing from the g.
jagonicum genome several regions of DNA which are

transcribed at high levels specifically in bacteroids.

Materials and Methods

Bacterial Strains. The g. coli strains are described in
Chapter 2. BJ2101 is a derivative of BJllOd from which the
nifA-like gene has been deleted (Adams, 1986). 8J271 is a

derivative of BJllOd with an insertion of the neomycin

45

phosphotransferase gene (ngtII) in the ntrBC operon (C.

Martin and B.K. Chelm, unpublished).

Plasmids. The cosmid library of BJllOd DNA in vector pLAFRl
was described by Adams gt gt. (1984). Plasmid pRK2013 is a
helper plasmid derived from R82 (Ditta gt gt., 1980) which
provides conjugal transfer functions. pDS4101 is a
derivative of the 001K plasmid which helps mobilize pBR322
derivatives (Wolk _t _t., 1984). pRL16l, pRLl70 and pRL425
are positive selection cloning vectors (Elhai and Wolk,
submitted) and p807 is a cloning vector which contains the
ggttt gene, which encodes kanamycin resistance (Rao and
Rogers, 1979). The following plasmids contain g. jagonicum
DNA sequences: p8J33 contains the gttfi gene which encodes
one of the subunits of nitrogenase (Adams gt gt., 1984);

p8J53A contains the glnA gene encoding glutamine synthetase

I (Carlson gt gt., 1985); pBJ110d contains the hemA gene

 

encoding 5-amino levulinic acid synthase (Guerinot and
Chelm, 1986); pBJ196A contains the gtgtt gene encoding
glutamine synthetase II (Carlson and Chelm, 1986); pBJ220
contains the gggt gene (P. Hall, unpublished results).
Plasmid JKI contains a 5.4kb gtngII fragment which carries
the g. ggtt gggt gene and plasmid pJK5 contains an
EggRI-gtgl internal fragment of the t. ggtt gggt gene; both
plasmids were a gift of C. Georgopoulos. The plasmid pDNAKl
contains a gygI-gggRl internal fragment from the g. ggtt

dnaK gene subcloned from plasmid JKI.

Bacterial Growth Congjtions. g. japonicum was grown at 30°C
in YEM (0.04% yeast extract [w/v], 1% mannitol [w/v], 3 mM
K2POa, 0.8 mM MgSOa, 1.1 mM NaCl) or in YEGG (YEM with 0.5%
gluconate and 0.1% glutamate replacing the mannitol). The
minimal medium was described by Manian and O’Gara (1982) and
contained xylose at 0.3% (w/v). For the low oxygen
experiments, t. japonicum strain BJllOd was grown in YEM or
YEM plus lOmM KNOa in 10 1 fermenters sparged with a mixture
of 0.2% oxygen (v/v) and 99.8% nitrogen (v/v) as described
(Adams and Chelm, in press). For the nitrogen limitation
experiments, BJllOd was grown in minimal medium with 1 mM
NHaCl as sole nitrogen source as described (Carlson gt gt.,
1987). For the heat shock experiments, BJllOd was grown in
YEM at 30°C to an optical density of 0.3 at 420 nm at which
time the cultures were shifted to 37°C and shaken vigorously
for the designated lengths of time (5, 10, 30 or 60 min).
The heat-shocked cells were then rapidly cooled in an
ice-water bath, collected by centrifugation and frozen at

-80°C.

gyt ggg _Nt Methods. The methods for Southern transfer and
colony hybridizations and for the isolation of plasmid and
total genomic DNA were described in Chapter 2. The
isolation of total bacterial RNA and cDNA synthesis were
performed as described in Chapter 2. The cDNA used for

probing Southern blots was depleted of rcDNA sequences by

solution hybridization as described in Chapter 2, but since

47

the probes were hybridized to purified DNA, competitor rDNA
plasmid DNA was omitted from the hybridization solutions.
Slot blots of RNA samples were performed by the method of
Schloss gt gt. (1984). The RNA samples were denatured by
heating to 65°C in a buffer consisting of 50% v/v formamide,
6% v/v formaldehyde, and 20 mM NaP04 (pH 7.7) and then
chilled on ice. The denatured RNA was diluted in cold
phosphate buffer (10X SSC [1X SSC = 0.15 M NaCl, 0.015 M
sodium citrate], 3% v/v formaldehyde, 20 mM NaPOa [pH 7.7])
and aliquots were transferred to cellulose nitrate using a
Minifold II Slot-Blotter (Schleicher & Schuell, Keene,
N.H.). The filters were then baked tg_ygggg at 80°C for two
hours and probed with nick-translated DNA probes as
described (Schloss gt gt., 1984).

gene-directed Mutagenesis. Gene-directed mutagenesis was
performed as described by Guerinot and Chelm (1986). Seven
different plasmids containing g. jagonicum DNA inserts in
which the central portion had been deleted and replaced by
the ggttt_gene were constructed in t. ggtt using standard
recombinant DNA methods (Maniatis gt gt., 1982). The
restriction fragments containing the ggttt gene were
isolated from plasmids pKC7, pRL161 or pRLl70, depending on
the restriction site requirements. Donor strains for
conjugal matings were obtained by transforming the
deletion-containing plasmids (pBJ24l, p8J242, p8J243,

pBJ261, pBJZBl, pBJ282, pBJ283) individually into E. coli

48

strain H8101 which contained plasmid pDS4101. For the
tri-parental matings, cells were mixed at a ratio of
approximately 1 x 108 g. jagonicum recipients to 5 x 107
H8101 donors to 5 x 107 H8101/pRK2013 helpers. The mating
mix was spread on YEM plates and incubated for 4 days at
30°C. The cells were then resuspended in 5.0 ml of 0.01%
Tween-80 and 200 ul aliquots were plated on YEGG agar
containing kanamycin (150 ug/ml) and chloramphenicol (30
ug/ml), the latter to counterselect against the g. ggtt
donor and helper cells. Kanamycin resistant colonies were
then screened by colony hybridization for the absence of
vector (pBR322) sequences. Since the deletion construction
plasmids cannot replicate in g. 'a onicum, kanamycin
resistant strains arise only when the plasmid containing the
ggttt gene integrates into the g. jagonicum chromosome. An
homologous double recombination event involving the flanking
sequences on both sides of the ggttt gene (see Fig. 2)
effectively deletes a region of the g. jagonicum chromosome

and replaces it with the kanamycin resistance gene.

gtggt tggtg ggg Acetylene Reduction Assays. The ability of
the mutated g. japonicum strains to form an effective
symbiosis with soybeans (Glycine ggg, cv. Amsoy) as
determined by nodule formation and nitrogen fixation was
assessed as described (Guerinot and Chelm, 1986). Surface
sterilized soybean seeds were inoculated with one of the g.

jagonicum 110d deletion derivatives, with wild-type BJllOd

49

or were left uninoculated. The seeds were then planted in
modified Leonard jars (Vincent, 1970) using the
nitrogen-free medium of Johnson gt gt. (1966) and incubated
in a growth chamber under conditions described by Guerinot
and Chelm (1986). After twenty-eight days the nodules were
picked and their ability to reduce acetylene to ethylene was
measured by the method of Hardy gt gt. (1968). Six to eight
nodules from each plant were also surface sterilized and the
resident bacteria were extracted and checked for phenotype

as described by Guerinot and Chelm (1986).

Egtraction gt Poly:fi+hydroxybutytgtg (Egg) tggg_ t.
jagonicum Bacteroids. PHB is a storage polymer which can
account for up to 50% of the dry weight of g. jagonicum
bacteroids (Klucas and Evans, 1968). PHB was extracted from
frozen g. jagonicum nodules by the method of Wong and Evans
(1971) with the following modifications. Approximately 500
mg of nodules were ground in 5 m1 of cold 0.05M Tris (pH
8.4) using a mortar and pestle. The homogenate was squeezed
through four layers of cheesecloth into a 15 m1 Corex
centrifuge tube and subjected to centrifugation at 300 x g
for 10 min at 4°C. The supernatant was transferred
carefully to a clean 15 ml Corex centrifuge tube and
recentrifuged under the same conditions. The supernatant
was transferred again to a clean tube and the cells
collected by spinning at 8,000 x g for 10 min at 4°C. The

cell pellet was washed twice by being resuspended in 5.0 ml

50

cold (4°C) sterile water and again collected by
centrifugation. The cell pellet was resuspended finally in
1.0 ml 820 and transferred to a pre-weighed microcentrifuge
tube (each microcentrifuge tube was heated to 85°C for 30
minutes, cooled to room temperature and weighed to +/-
0.00005 mg on a Mettler model H10 balance). The cells were
pelleted again by centrifugation in a Fisher
Micro-Centrifuge (model 235C) at 4°C for 30 sec and the
supernatant carefully removed and discarded. The cell
pellets were dried to a constant weight at 85°C (usually for
12-16 h). Commercial bleach was added (200 ul/mg of cell
pellet) and cell hydrolysis allowed to proceed at room
temperature overnight. The lipid granules which remained
after cell hydrolysis were pelleted by centrifugation in a
microcentrifuge for 10 minutes at room temperature. The
supernatant was removed and discarded and the pellets were
washed sequentially with 500 ul of sterile 820 and 500 ul of
acetone and collected by centrifugation in a microcentrifuge
for 30 seconds at room temperature. The pellets were
dissolved in 1.0 ml of CHCla at 65°C for 10 minutes (with
intermittant vortexing) in the tightly capped
microcentrifuge tube and then cooled to room temperature.

The concentration of the dissolved PHB was then determined.

Deteryination gt the Concentration gt PHB Diggolved tg

fi‘

08013. The concentration of the PHB dissolved in CHCla was

 

determined by the method of Law and Slepecky (1961). This

51

assay is based on the fact that crotonic acid, the
hydrolysis product of PHB, has an extinction coefficient at
235nm of 1.55 x 10‘ M‘1 cm‘l. PHB solution (25 ul of the
PHB solution in chloroform) was allowed to dry down in a 16
x 150 mm culture tube at room temperature and than 10.0 ml
of concentrated H2804 was added. The tube was topped with a
marble and incubated at 100°C for ten min. After cooling to
room temperature, the optical density of the solution at 235
nm was measured. From these results the total dry weight of
P88 extracted per mg of dry weight of bacteroids was
calculated. As a control, known amounts of commercial PHB
were dissolved in CHCla and the concentration measured by

the protocol outlined above.

Electron ggg ttggt Microscopy. Nodules were harvested from
roots of four week old soybean plants, rinsed in sterile
deionized water, and thin (1 mm) cross-sectional slices were
hand out with a razor blade. Sections of the slices were
immediately fixed in 4% glutaraldehyde, 0.15 M sodium
cacodylate (pH 7.2) for 2 h tg ygggg at room temperature and
then incubated at 4°C overnight. Following fixation, the
tissue was rinsed for three 1 hr periods at room temperature
with Buffer A (Maupin and Pollard, 1983) containing 0.2%
(w/v) tannic acid and then for 15 min in Buffer A only.
Postfixation was in 1% (w/v) osmium tetroxide in Buffer A
for 1 hr at room temperature. The tissue was then

dehydrated through a graded ethanol series, cleared in

52

propylene oxide and embedded in VCD/HSXA ultra low viscosity
medium (Ladd Research Industries, Burlington, VT). Thin
sections were cut on an LED Ultrotome III with a Dupont
diamond knife. Sections were subsequently stained with
uranyl acetate and lead citrate and examined using a Philips
EM300 electron microscope. For light microscopy, 2-3 um

thick sections were stained with 0.5% toluidine blue.

Results

Identification, Magping, ggg Subcloning gt Sggyences
ttgggcribeg Specificalty tg Bacteroids. Nine cosmids
containing sequences that are specifically transcribed in
bacteroids were digested with the restriction endonuclease
tggRI, size fractionated on agarose gels and transferred to
cellulose nitrate. The DNA bound to the filters was then
hybridized with cDNA synthesized using total RNA from
bacteroids or from free-living cells grown on YEM. The cDNA
probes had been enriched for mRNA by the protocol outlined
in Chapter 2. Four DNA fragments from three different
cosmids (pRJcos3-61, pRJcos4-51 and pRJcosl4-8) which
produced the best differential signals (Fig. 1; Chapter 2,
Table l) were subcloned into appropriate plasmid vectors,
restriction mapped with a variety of restriction
endonucleases (Fig. 2) and probed again with the same cDNAs
to further localize the regions which were actively

transcribed in bacteroids. Recombinant lambda bacteriophage

A
..
“-1. .‘u .
- '- 3 an.
"' m
B_ ,

It

Figure l. Cosmids containing sequences exhibiting
bacteroid-specific expression. DNA from nine cosmids which
contain sequences transcribed at high levels in bacteroids
was digested with restriction enzymes and transferred onto
cellulose nitrase. The filters were hybridized with equal
amounts (2 x 10 cpm) of cDNA prepared using either RNA
isolated from bacteroids (panel A) or from cells grown in
culture (panel 8). Lane a - cosmid pRJcosZ-63 which contains
the nitrogenase structural genes (nifQK); lanes b-j are
cosmids isolated in this study - pRJcos 1-20, 3-61, 4-51,
7-39, 8-62, 11-78, 12—37, 14-8, and 15-75; lane k -
pRJcos7-36 and lane l - pRJcole-90 both of which contain
genes necessary for nitrogen fixation (fix genes).

54

Figure 2. Restriction maps of five regions of the 8.
japoniggm genome (p8J214, pBJ216, p8J227, pRJcos4-51,
p80270) which contain sequences which are specifically
transcribed in bacteroids. Dashed lines connect the
restriction maps to diagrams of contructions used to
generate seven deletions strains in which the transcribed
regions were replaced by the neomycin phosphotransferase
(nptll) gene. 8 - BgmHI, C - ngl, E - figgRI, H = flindIII,
Kh=flmhP=P_stI.S=ia11.Sm=§maI.St=§§tI.X=

X 01.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E SC H X P S E
p8J214 I II I I I -I I #1
' ’,” 5.3kb
1 ’I”
E “ 5“ E
3.1241 I {—1 4 I
p I ___L
E S P H S P S H X E
pBJZIG I I II I l I I L1
1 ’,” 5.1kb
I ,,”
H H
E E
. I A
.____——-I notII I—-———-3.2kb
85m P S Sm H P S P E
{18.1227 [1 I I I I I I I I
l ’,” 6.0kb
I I,”’
8 f j E
pBJZ43 I I ]_______l
, notII 3.9“
pRJcos4-51
E 8 E B K C E P H/C 8 P B E
I I I I I I I I I I I I;
\ \ 11.9kb
\ \
\ \
8.1 B B c C
p 261 -
"‘3‘ ”i 6.0Kb
8 c c a
p8J282 L______ not I
\ 4.9Kb
\
pRJcos4-51 \\ \
E 8 E 8 K C E P H/C B P B E
I I I I I I I I I I I I I
\ ,,/ 11.9kb
\ ’/
\ ,z
3.1 E E B a
p 233 I J-———n——L 4
_J‘er—T' 6.5Kb
E/St E P S H E/B E P B E/St
pBJZ70 I l l I I l I l I I
| ’x’ 4.2K!)
I ’z’
E/St E ‘1’ " BE/St
pBJ281

 

56

815, which contained sequences specifically transcribed in
bacteroids, showed no homology to any member of the cosmid
library (Chapter 2, Table l) and so represented a fifth
differentially expressed region of the g. japonicum genome.
A 4 kb SstI fragment of DNA was subcloned directly from this
phage to create plasmid pBJ270, which was subsequently

mapped and analysed as described above (Fig. 2).

Congtruction gt Deletion Strgins. Stable deletion
derivatives of BJllOd were constructed in order to examine

the effect of deleting sequences from the g. japonicum

 

genome which are highly transcribed specifically in
bacteroids. Seven regions of the g. japonicum genomic DNA,
ranging in size from 2.0 to 6.3 kilobases, were replaced tg
yttgg by DNA fragments containing the ggttt gene which
encodes neomycin phosphotransferase II (Kant). In each
construct the ggttt gene was flanked by colinearly oriented
regions of t. japonicum genomic DNA of at least 500 bp in
length in order to provide adequate substrates for
homologous recombination. The corresponding chromosomal
regions were then replaced by the deletion constructions
(depicted schematically in Figure 2) using the tri-parental
mating method described by Guerinot and Chelm (1986). To
confirm that the double crossover recombination event
necessary to produce the deletion had occurred, total
genomic DNA from several isolates of each derivative was

prepared. The DNA was digested with appropriate restriction

57

enzymes, size fractionated on agarose gels and then
transferred to cellulose nitrate. The resulting filters
were hybridized with nick-translated plasmids containing
vector pBR322 sequences only, vector pBR322 sequences and
the ggttt gene, or vector pBR322 sequences and DNA from the
g. jagonicum regions which had been deleted. The results
for deletion strain 8J26l are presented in Fig. 3 as an
example. As expected, the nick-translated vector failed to
hybridize the filters except in lanes where DNA from a
strain resulting from a single crossover event (which
results in the incorporation of vector sequences into the
genome) was included as a control (Figure 3A, lane 6). Also
as expected, the nick-translated probe containing the
deleted region hybridized to DNA from the single crossover
strain and 8J110d (Fig. 3C, lanes 6 and 7), but not to DNA
from double crossover derivative strains. The probe
containing the ggttt gene hybridized to the DNA of both the
mutant and the single crossover derivative strain resulting
in bands of the predicted size (Figure 3B, lanes l,2,4,5 and
6). The results for the hybridization of all three probes
to the DNA from the strain in lane 3 are anomalous and the
strain may have arisen by insertion of the deletion-bearing
plasmid into the g. jagonicum genome by non-homologous
recombination. Results similar to those for the correctly
constructed 8J261 deletion derivative strains (e.g. those in
Fig. 3, lanes 1,2,4 and 5) were obtained for all the other

mutant constructions except for strain BJ227 in which the

58

Figure 3. Verification of structure of B. jgpgnigum deletion
derivative 8J261. Plasmid p8J261 (see Figure 2) was
conjugated into wild—type strain BJllOd in which it is
unable to replicate. Kanamycin resistant colonies were
screened by colony hybridization for the absence of vector
sequences as an indication that a double crossover
recombination event had occurred. Total genomic DNA (2 ug)
isolated from five presumptive deletion derivatives (lanes
1—5), one single crossover derivative (lane 6) and wild-type
8. jgpogicgm 110d (lane 7) was digested with restriction
endonucleases, size fractionated on agarose gels and
transferred onto cellulose nitrate filters. Three
identically prepared filters were hybridized with
nick-translated DNA of vector pBR322 (panel A), pKC7 (panel
B) and plasmid pBJ172 which contains the region being
deleted from the chromosome (panel C). The lack of homology
to pBR322 or to the deleted region probe in lanes 1,2,4, and
5 confirm that the target and vector sequences are not
present. The presence of sequences homologous to plasmid
pKC7 indicates that the nptll gene has been inserted into
the chromosome in the Kan strains (panel 8, lanes 1-6) and
is absent in BJllOd (panel 8, lane 7). The derivative in
lane 3 may have arisen by non-homologous recombination. All
three probes hybridize to the DNA of the single crossover in
lane 6 and only the deleted region probe hybridizes to
BJllOd DNA as expected. The lane marked MN contains
molecular weight standards.

23450

1

 

6O

deleted region hybridized to numerous fragments in the
genomic digests. This latter result may indicate that a
repeated sequence analogous to those described by Kaluza gt
El- (1985) and Hahn and Hennecke (1987) may be present in

the deleted region of BJ227.

Verification gfi figcteroid-gpecific Induction g: Deleted
Regions. In order to confirm that the apparent induction of
the bacteroid-specific sequences was not an artifact of the
cDNA synthesis and fractionation, RNA from bacteroids and
cells grown in culture was transferred to cellulose nitrate
filters using a slot blot apparatus. The filters were
probed with nick-translated clones of the regions which had
been deleted in each of five of the constructed strains. In
all cases, the RNA corresponding to each deleted region is
present in greater amounts when bacteroids (Fig. 4, lane A)
are compared to cells growing exponentially in culture (Fig.
4, lane C) or cells in stationary phase in culture (Fig. 4,
lane B). The expression of a gene known to be induced in
bacteroids (alga, Fig. 4, panel A) and one known to have
similar levels of expression in all three cell types (glgé,
Fig. 4, panel B) were monitored as controls. Comparison of
signals seen in Figure 4 (panel A, lanes 1 and 2) suggest
that RNA homologous to the giffi gene is present at
approximately 20 to 30-fold higher levels in g. japonicum
bacteroids than in cells grown to stationary phase in

culture. Under similar conditions, the glnA gene shows a

61

Figure 4. Slot blots of a. japonicum RNA probed with
nick-translated plasmids containing control genes or one of
the sequences replaced in the deletion constructions. Lane
1 - bacteroid RNA, lane 2 = stationary culture (YEN) RNA,
and lane 3 - late exponential culture (YEN) RNA. In panels
A to G, RNA from each source was loaded in a vertical array
of six slots with (from top to bottom) 3.0, 1.0, 0.3, 0.1,
0.03 and 0.01 ug of RNA. In panel H the RNA concentrations
were lowered to cover the range 0.3 to 0.001 ug in order to
obtain more readily quantifiable signals.

The probes used were: panel A = pBJB3 (nifH), B = pBJ49
(glnA), C - pBJZl4, D - pBJ216, E = p8J227, F - pBJZSl, G =
pBJ285, H . pBJl42 (rDNA).

 

 

63

three to five—fold induction (Fig. 4, panel B). The
bacteroid-specific sequences isolated in this study all
showed between ten and thirty-fold induction (Fig. 4, panels
C,D,E,F, and G). The filter probed with the nick—translated
rDNA plasmid (Fig. 4, panel H) demonstrates that equal

amounts of total RNA were present in each sample.

Characterization g: Deletion §trains. At least six

 

individual soybean plants were inoculated with each of the
deletion mutant strains and grown in nitrogen-free medium in
modified Leonard jars. After four weeks, the plants were
examined for appearance, nodule formation and the ability of
excised nodules from each plant to reduce acetylene to
ethylene. In all cases the mutant g. japonicum strains were
similar to the wild-type strain BJllOd (Table 1). To ensure
that the phenotype seen on plants inoculated with deletion
strains was not due to contamination by wild-type g.
Japonicum, six nodules from each plant were crushed and the
antibiotic resistance phenotype of the resident bacteria was
determined. All of the bacteria isolated from "mutant”
nodules from each plant were resistant to kanamycin, and all
of the bacteria isolated from control plants inoculated with
BJllOd were sensitive to kanamycin indicating that the
"mutant" nodules indeed contained only deletion-bearing
bacteria.

The mutant strains were further examined by comparison

with BJllOd for growth on rich medium (YEGG) and on minimal

64

Table 1. Acetylene reduction by excised nodules of g.
iaponicum deletion strains and wild-type strain BJllOd.
Nodules were picked from the roots of four-week-old
soybean plants grown in Leonard Jars. The amount of
acetylene reduced is presented as a percentage of the
amount reduced by wild-type strain BJllOd.

 

Strain Acetylene reduced relative
___-___-___-______:9-!il§:ix2§_£§l ______

BJ241 108
BJ242 83
BJ243 133
BJ261 103
BJ281 105
BJ282 107
BJ283 100 ,

BJllOd reduced acetylene to ethylene at a rate of
5.6 i 0.4 umoles/g of nodules/h.

65

medium amended with 0.3% xylose. The growth of each of the
mutant strains was identical to that of wild-type DJllOd
under both conditions (Fig. 5). The ultrastructure of
nodules incited by each mutant strain was examined by
electron microscopy. No significant differences were
observed when each of the mutant strains was compared to
wild-type DJllOd (Fig. 6). Bacteria purified from nodules
incited by four of the mutant strains (DJ214, DJ216, 83227
and DJ261) were also assayed for levels of the storage
polymer poly-B-hydroxybutyrate and found to contain large
amounts (Table 2). Although there is some variability in
the data, since PHD is being actively accumulated it was
expected that deletion of a gene involved in PHD synthesis
would result in a marked decrease in PHD levels. The high
levels of PHD detected was in agreement with evidence from

the electron micrographs (Fig. 6).

Expresgion g: the Dacteroid-specific Sequences Under Various
Physiological Conditions. Since deletion of the
bacteroid-specific sequences from the genome of DJIlOd had
no apparent effect on modulation, nitrogen fixation, or
growth in culture, the expression of the sequences was
examined under a variety of conditions in an attempt to
determine how the expression of these presumptive genes was
regulated. In each of the experiments described below,
cosmid or plasmid DNAs were digested with appropriate

restriction endonucleases, the fragments size fractionated

DD

66

420
10.
‘5‘\
x/ 333
x //"
///g
6
1.0 ///*
X
/SE
1
/9
0.1 x
'k
2 * - 80241
+ - 30242
0 - 30243
, x - BJllOd
.01

 

 

H.

2 3 4 5 6 7 8 9 10 11
Days

Figure 5a. Growth curve of B. japgniggm deletion
strains in minimal medium with 0.3% xylose as the carbon
source. Growth was monitored by the change in optical
density at 420 nanometers.

67

 

 

OD420
10.
;”,,:--§--i
0/*
’//’X
1.0 2/
a////x
o
0.1
i
*/ X.BJ281
o + - 80282
i o - BJ283
/// * - BJllOd
.01 I
0 1 2 3 4 5 6 7 8 9 10 11
Days

Figure 5b. Growth curve of g. japgnigum deletion

strains in minimal medium with 0.3% xylose as the

carbon source. Growth was monitored by the change
in optical density at 420 nanometers.

 

68

 

OD420
10.
!./".“\.*
O
o,///
*
X
0
*
1.0 *
9
X
0.1
9
X
/// * - 30241
g x - 30242
+ - BJ243
o - BJllOd
.01 i, _
0 0.5 1.0 1.5 2.D**§.5 5.0 §.§ 4.0
Days

Figure 5c. Growth of B. jgpgniggm deletion strains in
rich medium (YEGG). Growth was monitored by the change
in optical density at 420 nanometers.

DD

69

420
10. *
a/x + .
x
./
*
o
» x
1.0
t
o
x
t
0.1
9
x - 8J281
+ + - 8J282
* o - 80283
0 * - BJllOd
x
.01

 

 

0 0.5 1.0 1.5 2f0 2.5 3.0 3.5 4.0 4.5
Days
Figure 5d. Growth of 8. japgnlgum deletion strains in

rich medium (YEGG). Growth was monitored by the change
in optical density at 420 nanometers.

 

Figure 6. Transmission electron micrographs of sections of
nodules incited by B. Jangniggm deletion strains and
wild-type strain BJllOd. Representative cross-sections of
deletion strains 8J261 (panel A), BJ282 (panel 8), 8J283
(panel C), and wildtype BJllOd (panel 0) are at 12-15,000
magnifications. HC - host cytoplasm, PM - peribacteroid
membrane, PHD - poly-fi-hydroxybutyrate granules. Bar equals
0. um.

71

Table 2. Poly-B-hydroxybutyrate content of D. japonicum
nodules incited by deletion strains DJ241, DJ242, and DJ243
and wild-type strain DJllOd. The results of two separate
experiments (designated as -l, or -2) are presented below.

 

Strain
00235 Ave. [PHB] Total mg of X PHD
ug/ml mg PHD cells dry wt.

DJ241-l .396 .461 2.56 1.02 1.8 56.9
.527

DJ241-2 .718 .682 3.79 1.51 2.2 68.6
.595
.733

DJ242—1 .258 .279 1.55 .62 0.9 68.9
.300

BJ242—2 .449 .478 2.65 1.06 2.8 37.9
.496
.488

DJ243-1 .426 .451 2.50 1.00 2.8 35.8
.441
.488

BJ243—2 .355 .373 2.07 .83 2.3 36.0
.370
.396

DJllOd-l .597 .672 3.72 1.49 3.1 48.1
.747

DJllOd-2 .768 .731 4.05 1.62 3.7 43.8
.705
.719

Blanks-l .016 s.d. .012

72

by agarose gel electrophoresis and then transferred to
cellulose nitrate filters. The DNA bound to the filters was
subsequently hybridized with cDNA made from total RNA
extracted from D. Japonicum cells of various genotypes grown
under particular cultural conditions or isolated from
soybean nodules.

In order to examine the possibility that the low level
of expression of the bacteroid-specific regions seen in
cells grown on YEM was due to the particular stage in the
growth curve at which the cells were harvested, samples of
cells were collected at middle and late logarithmic phase as
well as at three points in stationary phase. cDNA was
prepared from the RNA extracted from each sample and from
each of the three cell types isolated from the nodules of
5-week-old soybean plants. As shown in Fig. 7, with the
exception of sequences on cosmid pRJcosB-62 which are highly
expressed in each cell type, each cosmid contains one or
more bands which produce a greater signal when probed with
bacteroid cDNA when compared to cDNA prepared from cultured
cells. The sequences from cosmids 3-61, 4-51, and 14-8
which were selected for analysis are not expressed
appreciably at any phase in the growth curve yet are highly
expressed in bacteroids. Possible carbon source effects
were also examined and cDNA was prepared from cells grown on
minimal medium containing either xylose or formate as the
sole source of carbon. Neither growth on the "good" carbon

source (xylose) nor on the ”poor” carbon source (formate)

73

Figure 7. Effect of developmental stage and growth stage on
the expression of B. japonicum DNA sequences that are
transcribed at high levels in bacteroids. DNA from nine
cosmids was digested with restriction endonucleases, size
fractionated by agarose gel electrophoresis and transferred
onto cellulose nitrate filters. The filters were probed
with random primed cDNA prepared using RNA from eight
different 8. japgnicgm cell types (A-H). A - bacteroids; 8
- transforming bacteria; C - nodule bacteria; D - H - cells
grown in culture (YEN) and harvested at 00 0 of 0.5 (D),
0.8 (E), 2.5 (F), 3.5 (o) and 3.7 (H). Th3 B. Jgpgniggm
cell types A, 8, and C were isolated from 5-week old soybean
root nodules. Arrows in panel A indicate the bands from
particular cosmids which showed differential hybridization
signals when filters were hybridized with cDNA from
symbiotic cell types (panels A, 8, or C) as compared with
cDNA from cells grown in culture (panels 0, E, F, G, or H).

mn-m~
u-v~
nn-~_
o~-—~
~o-u
an-“
_m-v
_o.n
o~.~

ms-m~
o-v—
nn.~_
as-"—
No-0
on.h
—m-o
_o-n
c-~

0‘.

0".“

QC

 

75

appeared to induce transcription of the bacteroid-specific
regions (Figure 8).

Although Rhiggbiu! and Bradyrhizobium species are
obligate aerobes, it is a well known that the oxygen tension
in nodules is maintained at low levels, presumably through
the action of leghemoglobin, an oxygen-binding plant protein
which is specifically induced in nodules (Appleby, 1984).
This abundant molecule is thought to provide the high oxygen
flux necessary for the actively respiring rhizobia while
maintaining a free oxygen tension low enough to prevent
irreversible inactivation of nitrogenase (Mortenson and
Thornley, 1979). Several reports that some Dgggyrhigobigg
strains can be induced to fix nitrogen in free-living
microaerophilic culture (Keister, 1975: Pagan gt al., 1975;
Kurz and Larue, 1975; McComb £3 31., 1975; Tjepkema and
Evans, 1975) led to further studies of the effect of low
oxygen tension on the transcription of genes involved in the
symbiosis (Ditta gt al., 1987; Adams and Chelm, in press).
In order to determine whether the level of free oxygen might
be involved in the transcriptional regulation of the
sequences isolated in this study, cDNA was prepared using
total RNA (provided by T. Adams) isolated from D. Japonicum
cells grown in YEM with and without lOmM KNOa in fermenters
sparged with nitrogen containing 0.1x 02 (microaerobic
conditions). Transcripts homologous to p8J227 and pDJ270
were apparently induced under these conditions (Fig. 9).

Although the signals obtained were below levels seen in

O 0" F" O N O h In
D I l U 0 I m
I
-‘o
. . u ,3?
.--

Figure 8. DNA from nine cosmids identified as containing
sequences which are transcribed at high levels in bacteroids
was digested with restriction enzymes and blotted onto
cellulose nitrate. The filters were probed with random
primed cDNA synthesized using RNA isolated from 8. lgpgniggm
cells grown under the following conditions: (A) minimal
medium with 0.3% formate as the carbon source: (8) minimal
medium with 0.3% xylose as the carbon source: (C) YEN medium
at 0.2% oxygen; YEN with 10mN KNO at 0.2% oxygen. The
control filter showing hybridization patterns for cDNA
synthesized using bacteroid RNA is included in Figure 7.

78

bacteroids (compare Fig. 9A, lanes 4 and 6 to Figs. 90 and
9D, lanes 4 and 6), they were similar to those for p8J33
(Eiifl) (Fig. 9, lane 1). This result suggests that
transcripts originating from these two sequences may be
regulated in a manner similar to gt: genes, at least with
regard to their response to oxygen levels.

Another global control mechanism for gene regulation is
the gt; system which is associated specifically with
nitrogen metabolism. Initially identified and studied in
the Enterobggteriaceae, the system employs two positive
regulatory proteins (gttg and gttg) to activate numerous
genes involved in nitrogen assimilation (reviewed in
Magasanik, 1982). In Klebsiella pneumoniae nitrogen
fixation genes are activated specifically by the gttg and
the git; gene products (Ausubel, 1984a; Gussin gt gt.,
1986). The regulation of symbiotic nitrogen fixation is
similar to that in free-living diazotrophs, but differs in
that the gttg gene product is not necessary for activation
of the 31:; gene (Szeto gt gt., 1987). To determine if the
gttt gene product might be involved in the transcriptional
regulation of the sequences isolated in this study, RNA was
prepared from 8. japonicum cells extracted from nodules
incited by strain DJ2101 in which the gttg_gene had been
deleted (Adams gt gt., manuscript in preparation).
Complementary DNA (cDNA) synthesized from total nodule RNA
isolated from DJ2101 and DJllOd was hybridized to filters to

which digests of plasmids containing the bacteroid-specific

 

Figure 10. Expression of bacteroid-specific sequences in
two-week-old nodules incited by 802101, a nifA deletion strain of
8. Restriction endonuclease digests of plasmid DNAs
were size fractionated on agarose gels, transferred onto cellulose
nitrate and hybridized with cDNA synthesized using RNA isolated
from nodules incited by wild- -type strain 80110d (panel A) or nifA
deletion strain 802101 (panel B). The plasmids contained the
bacteroid- -specific sequences isolated in this study (lanes b- f -
p80214, p80216, p80227, p80270,1p80273) or various control genes
(lane a - n1ffl,- glnA, laneh - hemA, lane i - :g§A, lane
J - p80294 which contains a sequence expressed in all cell types,
lane k - glnll, and lanes l- -o - 0001, 001, .01, .1 ug of pBJl42
which contains the rDNA operon).

80

sequences had been transferred. Little or no signal was
detected on the filters probed with the git; deletion strain
cDNA (Fig. 10, panel 8). Sequences homologous to p80227
were present in small amounts as were sequences homologous
to various control genes (£325, glnA, rDNA). From the low
level of these signals it is apparent that much less intact
RNA was obtained from the nodules incited by the git;
deletion strain (802101). The nodules incited by 802101 do
not develop normally and the Rt: gene transcripts are not
detectable (Adams and Chelm, in press; Fisher gt gt., 1986).
Although this experiment does not provide evidence for a
direct involvement of the gttg gene product in the
transcriptional activation of the bacteroid-specific
sequences, it suggests that a regulatory mechanism or
pathway may be shared.

Two sets of experiments were conducted to examine
whether the global nitrogen regulatory (gtt) genes affect
the expression of the bacteroid-specific sequences. In their
studies of glutamine synthetase II regulation, Carlson gt
gt. (1987) grew cultures of DJllOd with a limited supply of
fixed nitrogen and took samples at various times. As
expected, the depletion of available nitrogen resulted in
the cessation of growth and the concomitant transcriptional
activation of 6811. RNA extracted from these same cultures
(a gift of T. Carlson) was used to prepare cDNA probe which
was hybridized to cellulose nitrate filters similar to those

used in the preceding experiments. The low signal levels

81

obtained indicate that induction of the gtg system alone is
insufficient to promote induction of the bacteroid-specific
sequences isolated in this study (Fig. 11). In addition,
when cDNA prepared from total RNA (a gift of C. Martin)
extracted from nodules incited by an gtgg' strain of BJllOd
(G. Martin, unpublished) was used as probe, all of the
bacteroid-specific sequences were expressed at levels
approximating those seen in wild-type nodules (Fig. 12,
lanes 2-6). As a control for this experiment, digests of
plasmid pDJ196A which contains an gtgg-regulated gene (GSII)
(Carlson gt gt., 1987) were included (Fig. 12, lanes 7,8)
and the lack of a hybridization signal to this DNA confirms
that the gtgg gene product was indeed inactivated (Fig. 12,
lanes 7,8).

Genes involved in the high temperature response (the
heat shock response) are known to be activated by a number
of diverse environmental stimuli (Neidhardt gt gt., 1984).
Although physiological conditions in soybean nodules are not
known to provide any of the characterized heat-shock
response inducing stimuli, the environment inside the plant
root differs greatly from the soil environment. It seemed
possible, therefore, that stresses imposed by the change in
environment might result in induction of stress—inducible
systems (e.g. heat-shock genes) and that the induction of
the bacteroid-specific sequences might be due to
physiological stresses encountered during nodule

development. This could explain why the highly expressed

 

0

Figure 11. Expression of the bacteroid-specific sequences in B.
cultures grown under nitrogen-limiting conditions.
Restriction endonuclease digests of plasmid DNAs were size
fractionated on agarose gels and transferred onto cellulose
nitrate filters. The filters were hybridized with cDNA
synthesized using RNA isolated from B. bacteroids (panel
A) or from aerobic cultures in which the supply of fixed nitrogen
was progressively depleted (panels 8, C, and D). The plasmids
contained niffl (lane 1), the bacteroid-specific sequences isolated
in this study (lanes 2-6 - p80214, p80216, p80227, p80270, p80273)

or glnll (lane 7).

12345678 12345670

 

Figure 12. Expression of the bacteroid-specific sequences
in a . ntrC insertion strain. Restriction
endonuclease digests of plasmid DNAs were size fractionated
on agarose gels and transferred onto cellulose nitrate
filters. The filters were hybridized with cDNA made using
RNA isolated from nodules incited by wild-type strain BJllOd
(panel A) or ntrg insertion strain 80262 (panel 8). The
plasmids contained niffl (lane 1), the bacteroid-specific
sequences isolated in this study (lanes 2-6 - p80214,
p80216, p80227, p80270, and p80273), or 91311 (lanes 7 and
8 .

84

bacteroid-specific sequences are not essential for effective
symbiosis. To test this hypothesis, cDNA was synthesized
from total RNA isolated from heat-shocked cultures of
DJllOd. Hybridization of this cDNA to cellulose nitrate
filters containing restriction digests of the cloned regions
of bacteroid-specific sequences indicated that none of the
presumptive genes are induced by these treatments (Fig. 13).
Hybridization of total g. jagonicum genomic DNA with
heterologous HSP70 (= gggg, a well characterized heat shock
gene) probes derived from either Q. ggtt (Georgeopolous gt
gt., 1982) or Arabadopsis thaliggg (Wu gt gt, manuscript in
preparation) did not detect a homologous sequence in g.
japonicum which could serve as a positive control for
heat-shock induction. Fortunately, the lack of an
appropriate control for heat induction was obviated by the
fortuitous induction in D. japonicum of the transcription of
a sequence which is also contained on one of the plasmids
used as a molecular weight marker. This sequence from g.
jgppnicum is not transcribed at detectable levels in
bacteroids (Fig. 13, panel A, lane 1), but its transcription
is induced when g. jagonicum cells are shifted from 30°C to

37°C for 10 min (Fig. 13, panel C, lane 1).
Discussion

An underlying assumption in this effort to isolate

genes which are transcribed specifically in Dradyrhizobigg

 

Figure 13. Expression of the bacteroid-specific sequences
in heat-shocked B. 1gpgniggm cells. Restriction
endonuclease digests of plasmid DNAs were size fractionated
on agarose gels and transferred onto cellulose nitrate
filters. The filters were hybridized with cDNA which was
prepared using RNA isolated from nodules incited by
wild-type figjgpgnjgum 110d (panel A), 05 from 8. Jangnigum
cells which had been heat-shocked at 37 C for 0 (panel B),
10 (panel C) or 30 (panel D) minutes. The plasmids
contained glnll (lane 2), the bacteroid-specific sequences
isolated in this study (lanes 3-7 - p80214, p80216, p80227,
p80270, and p80273), and the E. 9911 gggfi gene (lane 8).
Lane 1 in panels A and C contains a ladder of plasmids used
as molecular weight standards and the arrow in panel C
indicates the plasmid which contains a sequence which is
heat-shock inducible.

86

japonicum bacteroids was the notion that such genes would
have functions necessary for the proper development of the
g. iaponicum/soybean symbiosis. Although genes which are
induced specifically during development have been identified
in a number of different eukaryotic systems (e.g. Firtel,
1972; Zimmerman gt gt., 1980; Putzer gt gt., 1984; Kurtz and
Lindquist, 1984; Kopachik gt gt., 1985), there have been no
reports of the effect on development of mutations in these
genes. The prokaryotic system in which the relationship
between transcriptional regulation and development has been
most carefully examined is fruiting body formation in
Myxococcug xanthgg (Kaiser gt gt., 1979). Using a transposon
Tug-derived promoter probe (Tng tgg) which upon integration
into the chromosome fuses tgg; expression to exogenous
promoters (Kroos and Kaiser, 1984), Kroos gt. _t. (1986)
examined 2374 Tn§ _gg insertion-containing !. xanthus
strains for both development-specific expression and for
developmental phenotype. Thirty-six strains were identified
which had significantly increased levels of B-galactosidase
activity when placed under conditions known to induce
fruiting body formation. Only three of these 5. xanthus
strains (and eight of the original 2374) also caused
abnormal fruiting body development, apparently indicating
that far fewer genes are essential for development than are
regulated during development (Kroos gt gt., 1986).

In Rhizobium and Bradyrhizobium the best characterized

example of developmentally regulated genes are the nifH,D,K

87

genes which encode the subunits of nitrogenase (Krol gt gt.,
1980; Prakash gt gt., 1982; Corbin gt gt., 1982). The gtt
genes are induced at high levels specifically in bacteroids
(Corbin gt gt., 1982) and a mutation in any one of the three
structural genes prevents the reduction of dinitrogen to
ammonia (Ruvkun and Ausubel, 1981). The bacteroid-specific
sequences isolated and analysed in this study are all
expressed at levels approximately equal to those seen for
the gtt genes (Fig. l). The results of the deletion
analysis indicate that none of these highly transcribed
regions is necessary for modulation or nitrogen fixation.
The simplest explanation for these results is that the g.
japonicum genome contains an additional copy or copies of
these sequences. Examples of gene duplication have been
reported in Rhizobium (Renalier gt gt., 1987; Honma gt gt.,
1985; Quinto gt gt., 1982), but with the exception of the
sequence contained on plasmid p80227, hybridization of
nick-translated plasmid DNA from the deleted regions to
total g. japonicum genomic DNA did not reveal the existence
of other copies of the cloned regions (e.g. Fig. 3, panel
C). It is possible, however, that the functions carried out
by the products of these regions are duplicated by other
non-homologous regions of the genome. An example of such a
functional duplication in g. japonicum was described by
Carlson gt gt. (1985, 1986) who isolated two separate genes
encoding glutamine synthetase. Although the two gene

products perform the same enzymatic function, the two

88

proteins differ in physical properties and mode of
regulation (Carlson gt gt., 1987). One gene (gtgg) has
homology to the corresponding gene in g. ggtt while the
second gene (gtgtt) has homology to the corresponding gene
in eukaryotes, but they are not homologous to each other at
the DNA sequence level (Carlson and Chelm, 1986).
Mutational analysis revealed that deletion of either gene
alone did not result in glutamine auxotrophy and did not
affect symbiotic competence, but that a strain carrying
deletions in both genes required glutamine for growth and
formed ineffective nodules (Carlson gt gt., 1987). If the
highly transcribed bacteroid-specific sequences isolated in
this study were functionally duplicated by non-homologous
sequences located elsewhere in the g. jagonicum genome, then
the lack of a symbiotic phenotype in nodules incited by the
deletion strains would be expected.

A second possible explanation for the failure of the g.
jagonicum deletion strains to produce a symbiotic phenotype
is that the gene products encoded by the deleted regions are
not involved in the processes of modulation or nitrogen
fixation but are active at a later stage of the interaction
(e.g. senescence). The determination of the poly-B
-hydroxybutyrate content of nodules incited by the deletion
strains was an effort to examine this possibility, since PHD
is accumulated during nodule development and hydrolysed
during nodule senescence (Wong and Evans, 1971). No

evidence was found, however, for altered levels of PHD

89

accumulation in the deletion strains (Table 2). It would
seem unlikely, moreover, that all of the isolated sequences
would fall into the "late gene" category. The fact that the
bacteroid-specific sequences were actively transcribed in
nodules harvested at two (Fig. 10, panel A), four (Fig. 9,
panel A) and five (Fig. 7, panel A) weeks after infection
also argues against this explanation.

Examination of the expression of the bacteroid-specific

sequences in g. japonicum wild-type cells grown under

 

microaerobic conditions (0.2 X 02 +/- 10 mM KNOa) provides
insight into the mechanism by which two of the sequences are
regulated. As was previously demonstrated for the gtt genes
(Adams and Chelm, in press), lowered oxygen tension is
sufficient to induce transcription of the sequences
contained on plasmids p80227 and PDJZ7O (Fig. 9, lanes 5,6).
The control of free oxygen levels in soybean root nodules is
presumed to be mediated by leghemoglobin, the apo-protein
portion of which is not synthesized until the nodule
structure has begun forming, approximately ten days after
infection (Gloudemans gt gt., 1987). ,If active
leghemoglobin is necessary for the expression of
oxygen-regulated genes in g. jagonicum, then the gene
products encoded by these genes are most likely utilized,
like nitrogenase, after nodule development is well underway.
There is no evidence, however, that the bacteroid-specific
sequences are regulated differentially among the three

developmental cell types isolated from nodules (Fig. 7).

90

Only one of the bacteroid-specific sequences is
transcribed at a detectable level in nodules incited by

802101, the nifA deletion strain of g. jgponicum (Fig. 10,

 

panel 8). Given the pleiotropic nature of this mutation
(Adams, 1986; Fisher gt gt., 1986) it was important that in
total RNA extracted from two week old nodules incited by
802101, transcripts hybridizing to the gggg gene, plasmid
p80294 (a g. japonicum sequence expressed in all cell types
examined) and the rDNA plasmid were all present, albeit at
only approximately 10% of the levels seen in bacteroids
(Figure 10, panels A and D). This apparently low level of
intact RNA is in agreement with the observation that, as
visualized by the electron microscope, g. japonicum cells
which reach the developing nodules neither undergo release
into the plant cells nor differentiate into bacteroids
(Adams, 1986). While the failure to detect transcription of
four of the bacteroid-specific sequences in 802101-incited
nodules does not provide evidence for the direct activation
of the bacteroid-specific sequences by the gttg gene
product, interruption of the regulatory pathway for gtt gene
activation prevents the transcription of these four
sequences. This experiment also divides the
bacteroid-specific sequences into at least two classes,
since the sequences carried on p80227 do not require the
gttg gene product for their expression.

The level of expression of the bacteroid-specific

sequences in g. jagonicum nodules incited by the ntrC

91

insertion strain does not differ from those seen in nodules
incited by wild-type strain BJllOd (Fig. 12). This result
was anticipated since gtgg deletion strains of both 3.
meliloti (Szeto gt gt,, 1987) and g. jagonicum (G. Martin,
unpublished results) are not altered in symbiotic
competence. The failure of heat shock to induce expression
of the bacteroid-specific sequences indicates that the
increased transcription in nodules is not due to heat-shock
related physiological stresses imposed by the nodule
environment.

In summary, the role in development of the g.
Jagonicum/soybean symbiosis of the highly transcribed
sequences isolated in this study is not readily apparent.
The bacteroid-specific expression of these transcripts would
lead to the assumption that they are involved in some aspect
of metabolism related to nitrogen fixation or assimilation.
The most probable explanation for the unaltered symbiotic
competence of the strains specifically deleted for these
regions is that functional but non-homologous duplications
exist elsewhere in the g. Jagonicum genome. Despite this
functional redundancy, however, determination of the fine
structure of the promotors for these transcripts should
provide some insight into transcriptonal regulatory
mechanisms employed in the development of the g.

japonicum/soybean symbiosis.

 

Chapter 4

Fine Structure Analysis of Two Dacteroid-Specific

Promoters in Bradyrhizobium japonicum

Introduction

The regulation of developmental processes which involve
the interaction of two organisms is complex (VanEtten and
Kistler, 1984; Long, 1984). In the gggdyrhtgobigg and
Rhizobium/legume symbioses there is evidence that some
amount of developmental control is exerted at the level of
transcription (Verma gt gt., 1986). Elucidation of the
regulatory mechanisms by which the levels of certain
transcripts are modulated during nodule development may
provide important insights into the overall regulation of
symbiotic development. One approach to this problem is the
physical characterization of the promoter regions of genes
which are regulated during development. Analysis of
promoter structure might suggest possible mechanisms by
which promoter activity is modulated.

A number of bacterial genes which are essential for the
gggdyrhizobigg and Rhizobium/legume symbioses have been
identified. Three classes of genes have been characterized
at the molecular level: (i) The gtt genes which encode the
nitrogenase enzyme complex. (ii) The ggg genes which affect

the ability of the bacterial cells to invade the host plant.

92

93

(iii) The ttg genes which are essential for the process of
nitrogen fixation. The analysis of gtt, ggg, and ttg gene
promoters has revealed several regions of conserved DNA
sequence which have been implicated in transcriptional

control. In Rhizobium and Dragyrhizobium as well as in the

 

free-living diazotroph Klebgjellg pneggoniae, many genes
involved in nitrogen assimilation are positively regulated
by the gtgg gene product in combination with either the gtgg
or the gttt gene product (reviewed in Ausubel, 1984a). The
promoters of genes which are regulated by the level of
available nitrogen generally contain two conserved sequences
centered at -24 bp and -12 bp upstream from the
transcription initiation site (Ausubel, 1984a; Johnston and
Downie, 1984). Activator sequences which may interact with
the gttt gene product have been identified further upstream
of these conserved sequences (Duck gt gt., 1986). The
promoters of genes essential for modulation contain a 47 bp
conserved sequence approximately 200-240 bp upstream of the
translational start codon (Rostas gt gt., 1986; Spaink gt
gt., 1987). Disruption or mutation of the conserved regions

of these nif, fix, and nod promoter regions has been shown

 

to reduce or abolish the ability of the bacteria to form an
effective symbiosis with the host plant (Gussin gt gt.,
1986; Rostas gt gt., 1986). Although the importance of the
conserved regions in transcriptional regulation has been
demonstrated genetically, the underlying molecular

mechanisms are not well understood.

94

Determination of the transcription initiation sites of
additional genes which are transcriptionally regulated
during the development of the t. jagonicum/soybean symbiosis
is a first step in the characterization of the molecular
basis for this regulation. Partial DNA sequence of two
divergently transcribed bacteroid-specific sequences has
been obtained and the promoter regions and transcription

initiation sites localized.

Materials and Methods

Bacterial Strains ggg Plasmids. g. ggtt strain JM83 was
used as the host for clones constructed in plasmid vector
pUC8 (Vieira and Messing, 1982). g. ggtt strain JMlO3
(Messing gt gt., 1981) was used as the host for clones
constructed in the M13 phage mp8 and mp9 (Messing and
Vieira, 1982) or in M13 mp18 and mp19 (Yannish-Perron gt

gt., 1985). All other bacterial strains and plasmids used

were described in Chapters 2 and 3.

pg; Squence Anatggtg. A 1000 bp ggtI-fltngII g. jagonicum
DNA fragment from plasmid p80216 was cloned into vector pUCB
to make plasmid p80296. This plasmid was linearized by
digestion with restriction endonuclease HtngII or ggtI to
expose one end of the g. japonicum DNA insert. A set of

progressively larger deletions from either end of the DNA

95

insert was generated using exonuclease DAL-31 as described
by Poncz gt gt. (1982). The action of this nuclease results
in predominantly blunt-ended fragments. The overlapping
variable length inserts were separated from vector sequences
by digestion with the appropriate restriction endonuclease
(either Pgtl or gtngII) and then ligated into the M13
vector mp19 which had been previously digested with
restriction endonucleases fltchI (which produces a blunt
end) and either ggtl or HtngII. The products of the
ligation reaction were used to transform g. ggtt strain
JMlO3 and the transformants plated on L8 plates in 3 ml LD
top agar in the presence of 1.3 mM
isopropyl-beta-D-thiogalactopyranoside (IPTG) and
5-bromo-4-ch1oro-3-indolyl-beta-galactoside (X-gal, 40 ul of
2% w/v in dimethylformamide). In the presence of these
compounds, phage containing inserts will produce colorless
plaques while phage in which no fragment has been cloned
will produce blue plaques (Messing, 1983). M13 phage from
colorless plaques were picked with sterile toothpicks and
used to inoculate 2 ml of L broth containing 50 ul of a
stationary phase culture of E. ggtt strain JM103. After
incubation with shaking for 5 h at 37°C, the cells in 1.5 ml
of each culture were collected by centrifugation for l min
in a microcentrifuge at room temperature. The supernatants
were removed and saved for later purification of
single-stranded DNA. Plasmid DNA from the bacterial cell

pellet was isolated by the method of Holmes and Quigley

96

(1981), digested with appropriate restriction endonucleases
and size fractionated on agarose gels. After analysis of
the plasmid digests, a series of nested M13 clones was

selected which contained progressively smaller g. japonicum

 

DNA inserts differing in size by approximately 200 bp.
Single-stranded phage DNA suitable for use in DNA sequence
determination was purified by published methods from phage
particles present in the corresponding mini-lysate
supernatants (Messing, 1983).

The nucleotide sequence of the g. jagonicum DNA inserts
contained in the nested M13 clones described here was
determined by the dideoxy chain termination method of Sanger
_t _t. (1977) as described in the manual published by
Amersham (1983). The template and primer DNAs were combined
in a 10 ul reaction which contained: 0.5 - 1.0 ug of M13
template DNA, 0.2 pmoles of universal primer DNA
(dGTAAAACGACGGCCAGT, available from Bethesda Research
Laboratories [DRL], Gaithersburg, MD.), and 20 uCi of
[alpha32PJ-dCTP (400 Ci/mmole, Amersham Corp., Arlington
Heights, IL) in 1X Klenow reaction buffer (6 mM Tris.HCl [pH
7.5], 1 mM dithiothreitol, 50 mM NaCl, 9 mM MgClz). The
reaction mixture was heated to 55°C for 10 min and then
allowed to return slowly to room temperature to permit
annealing of the primer to the template DNA. Following the
addition of 0.5 units of the Klenow fragment of DNA
polymerase I (DRL, Gaithersberg, MD), four 2 ul aliquots

were removed and primer extension reactions initiated in

97

each aliquot by the addition of 2 ul of a solution
containing the four deoxynucleotides plus one each of the
dideoxynucleotides as described (Amersham Corp., 1983). To
destabilize DNA secondary structure during gel
electrophoresis, thereby reducing "compression" of bands
which complicates sequence analysis,
7-deaza-2’-deoxyguanosine-5’-triphosphate (c7dGTP) was
substituted for deoxyguanosine (Darr gt gt., 1986) at the
concentration recommended by the manufacturer (Doehringer
Mannheim, Indianapolis, IN). After incubation for 20 min at
room temperature, the reaction was completed by addition of
2 ul of a solution containing all four deoxynucleotides at a
concentration of 0.5 mM and further incubating for 15 min at
room temperature. The reaction was then stopped by the
addition of 5 ul of loading dye (0.1% bromophenol blue, 0.13
xylene cyanole, 95% formamide).

The DNA fragments in the stopped primer extension
sequencing reactions were denatured by immersion in water at
95°C for 5 min. Aliquots (2 ul) were size fractionated by
electrophoresis on 41 cm long denaturing gels made with
wedge-shaped (0.4 mm to 1.2 mm) spacers. The gels contained
6-BX acrylamide (depending on the desired conditions), 503
w/v urea in T88 buffer (89 mM Tris-base, 112 mM boric acid,
2.5 mM EDTA [pH 8.3)). Electrophoresis was carried out at
1100-1400 volts for 4.5, 8 or 11 h in order to resolve
fragments of increasing molecular weight (chain length).

The gels were then fixed by immersion in a solution of 10%

98

methanol and 10% glacial acetic acid in water for 20 min.
Fixed gels were then dried down onto a sheet of
chromatography paper (Whatman, 3MM) at 80°C, tg ggggg, for
60 min using a Bio-Rad Model SE11258 Slab Gel Dryer. The
labelled fragments on the gel were visualized by
autoradiography for 12 to 24 h at room temperature using

Kodak X-Omat AR film.

Nuclease gt Protection Analysig. Uniformly labelled
single-stranded DNA probes were synthesized by primer
extension as described by Holben gt gt., (1988). The 215 bp
gggI-gtll and the 272 bp §tyI~§gtI fragments isolated from
plasmid p80296 were used to prepare end-labelled
single-stranded DNA probes as described by Adams and Chelm
(1984). Both types of probes were used in SI nuclease
protection experiments to determine the direction of
transcription and the approximate location of the RNA
transcription initiation site. 81 protection analysis was
carried out by the method of Derk and Sharp (1977) as
adapted by Adams and Chelm (1984), except that RNA/DNA
hybridizations were performed at 58°C. Single-stranded DNA
probes for precise determination of the transcription
initiation site were synthesized using gene-specific
oligonucleotide primers as described by Adams and Chelm
(1988) and Carlson gt _t. (1987). The sequences of the two

primers used are: 5’ -TCATCGCGGTCTACGAG (Oligonucleotide #1)

and 5’ -GCAAGGACGTAATATGC (Oligonucleotide #2). The

99

single~stranded template DNA was derived from the 1000 bp

PstI—HindIII fragment from plasmid pDJ296 cloned into phage

 

 

vectors M13 mp8 and M13 mp9. DNA sequence comparisons were
performed using the MicroGenie software system (Beckman,

Palo Alto, CA).

Results

Initial Sl nuclease protection experiments using
uniformly labelled single-stranded DNA probes indicated that
total RNA isolated from soybean nodules incited by g.
igponicum strain 110d protected portions of both strands of
the DNA present on plasmid p80296 (Fig. 1, lanes 2 and 7).
The probes were specifically protected from $1 nuclease
digestion by total RNA isolated from bacteroids.

Conversely, no protection was detected when the probes were
hybridized with total RNA isolated from cultured cells of g.
japonicum grown aerobically in YEM medium (Fig. 1, lanes 3
and 8) or microaerobically in YEM medium with 10 mM KNOa
(Fig. 1, lanes 4 and 9). These results are in agreement
with the results of experiments presented in Chapter 3 in
which the bacteroid-specific expression of the same region
of DNA was examined using mRNA-enriched cDNA probes (Chapter
3, Figs. 7 and 8).

The DNA sequence of the 1000 bp 8. japonicum
EgtI-fltngII fragment contained in plasmid p80296 was

determined (Fig. 2 and Fig.3). The transcription initiation

 

Figure 1. Nuclease 51 protection analysis of two
divergently transcribed B. jgpgniggm sequences which exhibit
bacteroid-specific expression. Total RNA (10 ug) purified
from B. iapgnigum 110d was hybridized with two
uniformly-labelled probes synthesized using single-stranded
templates derived from the figtI-Egtl fragment of plasmid
pBJ296 cloned into M13 phage vectors mp18 and mpl9. The RNA
was isolated from root nodule bacteria (lanes 2 and 7),
cells grown in culture (lanes 3 and 8), or cells grown
microaerobically on YEM with 10 mM KNO (lanes 4 and 9).
Lanes 1 and 6 contain probe which was got treated with $1
nuclease. Lanes 5 and 10 contain probe which was hybridized
without RNA and then treated with $1 nuclease.

Figure 2.

The nucleotide sequence of the g.
bp PstI-HindIII DNA fragment from plasmid p80296.

101

transcription initiation sites for the two

symbiotically—regulated promoters are marked by asterisks.
Refer to the diagram in Fig.

transcription.

CTGCAGGGAC

AGCCAAGTTC

TCCAGCGCAA

TCATCAGCAA

TCCACCCTTA

CGCCGGCAGC

CGAGGGTCTC

TTGACTCTGC

TTGTTCATGC

*

GGCAAAAACT

TGGTAGGTGC

GAATTGGAGC

GGGTTGGCCA

CAACGAATGC

GCTTGCGGTG

AGAGCGCTGC

TGCTAAGTTC

CCAAACGGCA

TACTTTGGAA

TCTTCGTGCC

GTCGCCCATT

GGATAACTTT

CCACATACTT

AACGTGCATG

AGCTTAGAGA

CCCTTGGTAC

CTCTTGGGAG

CCAGCGCTCG

ATATATTCGA

CTTGCAGCCA

CCCTGTGTGT

AACTGGCGGG

CCGGGTCGCG

CGGGCAAGAA

ATTCTCTCCG

TTTCAACGAG

CCAGCGGCTT

GCGTCGCTCA

TGTCTGGTCG

GTCGCCGACG

ITCTCGAAGA

CAAAGGTGGC

CGACATCATG

GCGGCATATT

ATCGGGTCCG

CCGCGGGGAG

CGGGTAGGGA

CCCGCACTGC

GACTGCGTGG

CCAGGCGTGC

GTACTCCTCA

TCTATCGTGC

TGTGCGATTG

TGCTGGTGTC

AGGAGGGGAG

TACAACTTTG

CAGCCCCAGT

CCCACAGGGC

ACGTCCTTGC

CGGCCCCCCA

japonicum 1000

 

The

3 for the direction of

50
ATAAGCGCCG

100
CAGCGCCTGC

150
TTCCGATCAA

200
CTGAATCCCC

250
CGACGAGCGA

300
TCGCGGTCTA

350
TCCGCCTCGA

400
GCATGTTTTC

450
TCCAATTCAA

500
CCACCGATTT

* 550
TAGCCCTTTA

600
CTTTGCGACC

650
AGCTACAGGC

700
GGCCGCAAGC

750
TCGGTCTCTG

Figure 2 continued.

GGACTATCAG

GATGCTGGCC

CGACAGATCG

GGGCAACCAG

CAGTAGGTTG

ATCGCAGGGC

ATACTCGATT

CTGCGCGCAT

TTCGGGTAGG

TCCATCCGGT

102

GGCCACATTT

GCGTGGAACC

ACGCGCGCCT

ATGTCTGTGA

CCTGAACGAG

AACGTCGACG

CAACCTGCGG

CTAGAGGGGC

GACCATGAGG

GAAGCAACCT

800
AAGTTCTGGC

850
CCGAAAAGCG

900
CGCATGGGGG

950
TGTGAGCTTT

1000
GCCGAAGCTT

 

_t transcript #1 3_
7

J1
I l l

H
1

1|

200 400 600 8 oo

O—fir-Tj
O—WL—‘UJ

O 1 0 bp

A\

transcript #2

(b

 

0

 

0

Figure 3. Restriction map of and sequencing strategy for
the 1000 bp BgtI-fitngII fragment of B. iaponicum DNA
contained in plasmid pBJ296. The approximate locations of
the transcription initiation sites for two divergent,
symbiotically regulated transcripts are indicated. P =
_12_s_tI, A = A_C£I, St = gtyfl, S = figtI and H = _H_iti_dIII.
Beneath the restriction map is a schematic representation of
the series of nested deletion fragments cloned in M13 mp18
or mp19 which were subjected to DNA sequence analysis. The
sequence was determined from the end of the fragment
designated with a dot.

104

sites for the two bacteroid—specific sequences transcribed
from plasmid p80296 were localized by nuclease 81 protection
analysis using end-labelled DNA fragments as probes (Fig.
4). The transcription initiation sites were then precisely
identified by comparison of the protected portion of
end-labelled oligonucleotide-generated probes to a DNA
sequence ladder corresponding to the same region of DNA
(Fig. 5). Appropriate regions of the DNA sequences of the

two 8. japonicum promoters were compared with the conserved

 

regions of the E. ggtt consensus promoter (Hawley and
McClure, 1983) and some degree of similarity was found (Fig.
6). DNA sequence comparisons of the two promoters with that
of the promoters of the g. iaponicum nifDK, gggg, gtgg and
gtgtt genes as well as the entire Genbank data base did not
identify any regions of DNA with significant similarities.
The amino acid sequences of two polypeptides which could be
predicted from the DNA sequence of the two transcribed
regions were also compared with the sequences of proteins
recorded in the Genbank data base and no significant

simlarities were found.

Discussion

The bacteroid-specific expression of g. japonicum

sequences present on the recombinant phage 816 was

demonstrated initially by differential plaque hybridization

   

Figure 4. Nuclease 81 protection analysis of two
divergently transcribed B- 1gpgnignm sequences which exhibit
bacteroid-specific expression. Total bacterial RNA (10 ug)
purified from root nodules incited by a. 1gpggiggm strain
110d was hybridized with end-labelled, strand-separated DNA
probes prepared from the 272 bp fityI-figll (panel A) and the
215 bp AggI-fityl (panel 8) fragments from plasmid pBJ296.
Lane 1 contains DNA probe which was not digested with $1
nuclease. Lane 2 contains DNA probe which was hybridized
with RNA and then subjected to digestion with 81 nuclease.
Lane 3 contains DNA probe which was placed under
hybridization conditions without added RNA and then
subjected to digestion with $1 nuclease. The band in lane 2
indicated by the arrow is the probe DNA partially protected
by RNA. The band which migrates at the same size as the
untreated probe DNA in both lanes 2 and 3 represents full
length protection of the probe by DNA:DNA hybridization.

SiGATC SiGATC

 

Figure 5. Precise localization of the transcription
initiation sites for two divergently transcribed
symbiotically-regulated sequences in B. japonicum.
Single-stranded DNA probes were generated by primer
extension of two specific end-labelled 17 bp
oligonucleotides (see Materials and Methods). These probes
were hybridized with total bacterial RNA (10 ug) isolated
from root nodules incited by 5. japogigum strain 110d and
then subjected to digestion by S1 nuclease. The protected
DNA fragments were accurately sized by subjecting them to
electrophoresis next to a DNA sequencing ladder which had
been produced by primer extension of the same
oligonucleotide.

107

Figure 6. DNA sequence of the promoter regions of two 8.
jgpggiggm symbiotically—regulated sequences. The
transcription initiation sites are labelled above the
nucleotide(s) with an asterisk. The regions exhibiting
similarity to the E. ggli promoter consensus sequences at
~10 nt and ~35 nt upstream of the transcription initiation
sites are underlined. A nine base pair sequence which is
centered at position ~97 nt in transcript #1 and at position
—100 nt in transcript #2 is underlined with asterisks.

Transcript #1

~100 ~90 ~80 ~70 ~60
TCTCCAATTCAAGGCAAAAACTAGCTTAGAGAGTCGCCGACGAGGAGGGGA
*********
~50 ~40 ~30 ~20 ~10 *

GCCACCGATTTTGGTAGGTGCCCCTTGGTACTTCTCGAAGATACAACTTTGTAG

Transcript #2

~100 ~90 ~80 ~70 ~60
GCTCCAATTCTAAAGGGCTACAAAGTTGTATCTTCGAGAAGTACCAAGGGGCAC
*********
~50 ~40 ~30 ~20 ~10 *

CTACCAAAATCGGTGGCTCCCCTCCTCGTCGGGACTCTCTAAGCTAGTTTTTG

E. coli consensus promoter.
*

--------------- TTGAC-------~-~--------TATAAT--~-~-CA
Ntr consensus promoter.

------------------------ CTGGCAC--~--TTGCA----------*

108

experiments (Chapter 2, Fig. 6). Randomly primed cDNA
synthesized using total RNA from root nodules or from cells
grown in culture was hybridized to purified DNA from the
recombinant phage 816 which had been digested with
restriction endonucleases, size fractionated on agarose gels
and transferred to cellulose nitrate. The difference in the
hybridization signal levels produced by the cDNA probes
synthesized from nodule bacteria and cultured cell RNA
compares favorably with the results from the plaque
hybridization experiments (Chapter 2, Fig. 7). The data
obtained from the $1 nuclease protection experiments
eliminates the possibility that the cDNA which hybridized to
recombinant phage 816 was transcribed from elsewhere in the
genome since protection of a single-stranded DNA probe from
digestion by the nuclease can occur only by precise (i.e.
100% homologous) pairing of the DNA probe with RNA.
Therefore, the single protected bands seen in Fig. 2 could
only arise by protection of each of the two DNA probes by
RNA transcripts from those regions. These experiments also
confirmed that transcripts from the two regions examined are
present at significantly higher levels in g. japonicum cells
isolated from soybean root nodules than in g. japonicum
cells grown in culture either aerobically or
microaerobically (Fig. 1).

Since experiments described in this chapter confirmed
that the DNA region contained on plasmid p80216 and subclone

p80296 is transcribed specifically in bacteroids, the DNA

109

sequence of the two symbiotically-regulated promoter regions
present on the plasmid was determined and the transcription
initiation sites precisely localized. The regions 12
nucleotides and 24 nucleotides upstream of each of the
transcription initiation sites exhibit little similarity to
known nitrogen—regulated promoters from g. Japonicum (Fig.
5; Carlson gt gt., 1987). This result is not surprising
since transcription from this region was not induced in
cultures in which nitrogen had been depleted (Chapter 3,
Fig. 11).

The significance of the similarity between the E. ggtt
consensus promoter (Hawley and McClure, 1983) and the two
symbiotically regulated g. japonicum promoters is unclear.
The gtgt gene promoter is the only other characterized g.

jgponicum promoter which has similarity to the g. coli

 

promoter consensus sequence. Carlson gt gt. (1985) found
that the gtgg_gene, unlike the sequences isolated in this
study, was transcribed at a relatively constant rate in all
cell types and under all experimental conditions examined.
It seems unlikely, therefore, that the same promoter
sequences would be responsible for both types of regulation.
A sequence of potential importance in the
transcriptional regulation of these regions, however, is the
9 bp sequence which is centered at -97 and ~100 nucleotides,
respectively, from transcription initiation site in each
promoter (Fig. 5). Because these regions are transcribed

divergently and have only 83 bp between the initiation

110

sites, the reverse complement of the 9 bp conserved sequence
is also present either seven (transcript #1) or nine
(transcript #2) base pairs after transcription initiation
site. Further experimentation is necessary to determine if
and how this sequence is involved in bacteroid-specific
transcriptional regulation and whether it functions upstream

or downstream of the transcription initiation sites.

CHAPTER 5

Discussion

A developmental process has been defined as a complex,
ordered series of events which results in a new
morphological or biochemical condition (Dworkin, 1985).
Genetic analyses have been applied to such developmental
processes in a variety of experimental systems, the most
thoroughly examined of which are sporulation in Bacillus
subtilis (reviewed by Losick gt gt., 1986) and Saccharomyces
cerevisiae (reviewed by Esposito and Klapholtz, 1981),
fruiting body formation in Myxococcgg xanthus and
Dictyostelium discoideum (reviewed by Kaiser, 1986), and
conidiation in Aspergillgg nidulans (Timberlake, 1980). In
prokaryotic systems the genetic analyses have been based
primarily on the characterization of mutants exhibiting
impaired development (reviewed by Dworkin, 1985).

Transposon mutagenesis techniques have greatly facilitated
the isolation of such mutants and the cloning of the
corresponding genes (e.g. de Bruijn gt gt., 1986). In most
eukaryotic systems, however, the lack of similar methods for
the identification and isolation of genes has led to the use
of differential hybridization techniques (Sargent, 1986) to
study development. These techniques are more general than
mutagenesis in that they can identify genes which are

essential for growth of the organism and genes which are

111

112

developmentally regulated but produce no phenotype when
mutated. Despite the apparent utility of this approach,
differential hybridization has been employed only
infrequently to study prokaroytic development (e.g. Gianni
and Galizzi, 1986). The major focus of this thesis project
was the adaptation of differential hybridization techniques
to permit the isolation of developmentally regulated genes
from the bacterial partner in the Bradyrhigobium
japonicum/soybean symbiosis.

Differential hybridization was developed initially as a

method for isolation of inducible genes in g. cerevisiae

 

(St. John and Davis, 1979). Complementary DNA (cDNA) probes

were prepared from RNA isolated from g. cerevigtae cells

 

grown in the presence or absence of galactose. Libraries of
g. cerevtgjae DNA were then screened for recombinant phage
which contained sequences whose transcription was induced
specifically by the addition of galactose. The RNA used for
the cDNA synthesis had been enriched for messenger RNA
(mRNA) by oligothymidylic acid-cellulose (oligo-dT)
chromatography (Aviv and Leder, 1972), a technique which
utilizes the hybridization of long stretches of adenylic
acid residues (poly-[A]) present on the 3' end of many
eukaroytic mRNAs with the oligo-(dT)-cellulose to separate
mRNA from ribosomal, transfer and other stable RNA species.
This mRNA enrichment step is essential for the reduction of
background hybridization signals which would otherwise mask
the signal differences produced by regulated genes. There

have been a number of reports that poly-(A)-containing RNA

113

can be isolated from a variety of bacterial species
including g. ggtt (Nakasato gt gt., 1975; Srinivasan gt gt.,
1975), Bacillus brevis (Sarkar gt gt., 1978), g. subtilis
(Kerjan and Szulmajster, 1980; Gopalakrishna and Sarkar,
1982), Rhodospirillum rubrum (Majumdar and McFadden, 1984)

and Methanococcus vannielii (Brown and Reeve, 1985).

Efforts to isolate poly-(A)-containing RNA from g. japonicum

 

were unsuccessful. Alternative methods for obtaining
mRNA-enriched probes were therefore developed.

Synthesis of randomly primed cDNA from total bacterial
RNA is an efficient way to generate single-stranded DNA
probes of high specific activity (Taylor gt gt., 1976). The
use of template RNA which has not been enriched for mRNA,
however, results predominantly in labelled stable RNA
species since these account for more than 90% of the total
cellular RNA (Ingraham gt gt., 1983). The stable RNA
template molecules often have considerable sequence
similarity to the corresponding genes which encode stable
RNAs in g. ggtt. As a result, cDNA probes synthesized from
total bacterial RNA using random primers produce strong
background signals when hybridized to either phage or cosmid
libraries made in gt ggtt due to this heterologous
hybridization to t. ggtt stable RNA genes (e.g. Chapter 2,
Fig. 1). In the experiments reported here, the background
hybridization signal produced in plaque filter
hybridizations by the cDNA complementary to rRNA (rcDNA) was
reduced by a combination of techniques. The cDNA probes

were first hybridized in solution to an excess of unlabelled

114

plasmid DNA which contained the E. japonicum rDNA operon.
The resultant rcRNA:DNA hybrids were removed by
hydroxylapatite (HAP) chromatography and the single-stranded
fractions retained for use as a hybridization probe. An
additional reduction in the background signal level was
achieved by including 50 ug/ml of sheared, denatured rDNA
plasmid DNA in the pre-hybridization and hybridization
solutions when probing E. japonicug genomic libraries
propagated in E. ggtt. In combination, these techniques
permitted visualization of some hybridization signals above
background levels. The strong signals resulted from the
hybridization of the rcDNA-depleted cDNA probe with plaques
produced by recombinant phage which contained E. japonicum
DNA encoding mRNAs that were abundant in nitrogen-fixing
soybean root nodules. These experiments demonstrate that
cDNA probes suitable for use in differential screening of
phage libraries can be synthesized from total bacterial RNA
and used for isolating bacterial genes expressed at high
levels in particular cell types or developmental stages.

One limitation of the differential hybridization
methods described in this thesis is the level of sensitivity
achieved. If the sensitivity of detection could be improved,
the method might also be used to identify genes expressed
differentially, but at low levels. One obvious approach to
this problem would be to further lower the background
hybridization signals obtained in plaque hybridization
experiments. The reduction in background signal obtained

thus far was achieved solely by removal of, or competition

115

with, rDNA sequences. If additional E. japonicum genes are

 

identified which have considerable sequence similarity to
genes in E. ggtt, such sequences could also be removed from
the cDNA probes. This approach might well succeed since the
overall sequence similarity between the two organisms is low
(Heberlein gt gt., 1967) and even the highly conserved E.
,ggtt gggE_gene (the homologue to the heat shock gene HSP70)
fails to hybridize to E. japonicum genomic DNA under
conditions of low stringency (see Chapter 3).

The identification of differentially transcribed genes
might also be facilitated if the E. japonicum DNA libraries
were constructed using smaller DNA fragment sizes. The
phage library in vector BFlOl had 6 to 10 kb inserts and
thus each recombinant phage contained sufficient E.
japonicum DNA to encode several transcripts. The presence
of a constitutively transcribed gene (or genes) on a segment
of DNA could mask a differential hybridizational signal from
an adjacent regulated gene, particularly if the regulated
gene were expressed at comparatively low levels. The
technical problems involved with handling the increased
numbers of phage necessary for screening small-insert
libraries (approximately 45,000 recombinant phage for
library of 1 kb fragments from E. japonicum) would not be a
significant obstacle to such an approach. In fact, the
potential improvement in sensitivity might well be worth the
time necessary to synthesize and screen a small-insert

library.

116

The identification of developmentally regulated genes
by differential hybridization has both advantages and
disadvantages when compared with methods previously used for
isolating such genes. Two limitations of traditional
mutational analyses are that the genes of interest must be
dispensible for the growth of the organism and that such
genes must produce a detectable phenotype when mutated.
These limitations are avoided if genes are identified solely
on the basis of their transcriptional regulation. Thus, one
advantage of differential hybridization is that the
technique permits the identification and isolation of any
gene which is transcribed at detectably different rates in
any two cell types or developmental stages. Several classes
of differentially regulated genes might be isolated which
would not be identified by traditional mutagenesis. None of
the genes isolated in this study were essential for growth,
although only five distinct regions were examined. The
developmentally regulated sequences which were isolated,
however, could not have been identified by chemical or
transposon mutagenesis since they produced no symbiotic
phenotype when mutated.

One disadvantage of the technique described above is
the amount of time and effort required to characterize a
differentially transcribed region of DNA once the
recombinant phage containing it has been identified. The
region must be subcloned into plasmid vectors, mapped with
restriction endonucleases and the transcribed portions

localized more precisely by probing fragments of plasmids

117

generated by restriction endonuclease digestion with
rcDNA-depleted cDNA. The transcribed regions can then
deleted and replaced by a segment of DNA containing an
antibiotic resistance marker gene. These constructions may
then be introduced into E. japonicum by conjugation and
recombination into the chromosome and the strains resulting
from double-crossover events identified. With a relatively
slow-growing bacterium such as E. japonicum this entire
process can take several weeks. The examination of large
numbers of transcribed regions would thus be very time
consuming.

This problem has been overcome in several microbial
developmental systems by the construction of "promoter probe
transposons”. The promoter probes are transposons which
have a promoterless reporter gene (e.g. the tggE gene)
inserted near one end. An insertion of the transposon into
an actively transcribed gene in the correct orientation
causes the reporter gene to be transcribed and creates a
strain with an easily scorable phenotype. Such promoter
probe transposons have been successfully used in Myxococcus
xanthus (Kroos and Kaiser, 1984), Qggtgggcter cregcentgg
(Bellofatto gt gt., 1984) and Bacillus subtilis (Perkins and
Youngman, 1986) and have permitted the isolation of genes
which are expressed specifically during particular stages of
development in these organisms. Genes identified by this
method must be inessential for growth since they are mutated
by the insertion of the transposon, but the technique does

permit the screening of large numbers of insertionally

118

mutated strains for developmentally regulated genes. The
presence of an endogenous beta-galactosidase (the tggE gene
product) activity in soybean root cells and the sensitivity
of plant cells to the antibiotic kanamycin have been
obstacles to the use of available promoter probes for
analysis of the E. japonicum/soybean symbiosis. A

fast-growing Rhizobigg fredii strain was mutagenized using a

 

mini-mu phage (Mu-dI [Kan, tgg]) which had been integrated
into a plasmid which could be transferred to, but would not
replicate in, Rhizobium (Olson gt gt., 1985). The kanamycin
resistant transconjugants were analysed for tggE expression
in the presence of soybean root exudate, root extract,
nodule extract, and low oxygen tension. Several genes were
identified which were transcribed at higher levels under
these "symbiotic-like" conditions (Olson gt gt., 1985). The
direct analysis of E. japonicum or E. fredii gene expression
during the development of the symbiosis may be possible
using the tggAD gene from Vibrio harveyii (Baldwin gt gt.,
1984) which has been shown to function in E. japonicum
(Legocki gt gt., 1986). The E. ggtt beta-glucuronidase gene
(GUS) has also been developed as a gene-fusion marker
(Jefferson gt gt., 1986) and most higher plant species have
been shown to lack this activity (Jefferson gt gt., 1987).
If a promoter probe transposon could be devised utilizing
either of these reporter genes, the identification of
bacterial genes which are regulated during soybean root

nodule development would be greatly facilitated.

119

Although it may be possible to improve methods for

isolating developmentally regulated genes from E japonicum,

 

perhaps the most important and intriguing observation
resulting from this analysis of five regions containing such
genes was the fact that deletion of the regions did not
affect symbiotic competence. Similar results have also been
obtained in studies of development in E. xanthus and E.
subtilis in which only a small percentage of the promoter
probe transposon insertions in genes which are transcribed
specifically during development were found to be essential
for development (Kroos and Kaiser, 1986; P. Youngman,
personal communication). The fact that deletion or
disruption of genes which appear to be transcribed
specifically during development does not perturbthe process
may be explained in several ways: i) developmental
processes may consist of many branching pathways so that the
loss of any one pathway can be compensated for by others.

So little is known about the biochemical functions of genes
activated during development that it will require much more
research before useful models can be advanced. ii) there
may be considerable redundancy of function between
developmentally regulated genes and thus mutational analysis
will be difficult since a single mutation will not eliminate
a particular phenotype. iii) the methods used for
identifying the functions of the deleted genes may not be
sensitive enough to detect small changes (i.e. lowered
efficiency rather than complete loss of function). In any

case, it is clear that deletion analysis alone will be

120

insufficient.as a method for identifying the functions of
genes isolated by virtue of their transcriptional
regulation.

The results of the experiments described in this thesis
suggest several possible avenues for future research into
the molecular mechanisms of developmental regulation.

First, additional promoter regions should be identified, the
DNA sequences obtained, and the transcription initiation
sites localized. Comparative sequence analysis of these
promoter regions might reveal the presence of conserved
blocks of DNA sequence which are involved in regulation.
Secondly, direct evidence for the involvement of particular
portions of the promoter region in transcriptional
regulation might be obtained either by deletion analysis or
by oligonucleotide-directed mutagenesis. These types of
experiments are most conveniently carried out using
promoter-reporter gene fusions which have an easily
detectable phenotype. Promoter fusions could also be used
in a third type of experiment in which the expression of the
gene fusion is assayed in randomly mutagenized bacterial
strains. Second—site mutations which alter the level of
expression of the promoter-reporter gene fusion product
might result from mutated regulatory genes. Lastly, a
search for regulatory proteins which interact with the
promoter regions could be undertaken. The binding of DNA by
proteins from cell extracts can be detected by filter
binding, electrophoretic mobility shift, or nuclease

protection experiments (Hennighausen and Lubon, 1986).

121

Proteins which bind specifically to regulated promoters
would be candidates for purification and inclusion in tg
gttgg transcription/translation systems to analyse
expression from developmentally regulated promoters. The
results from these four general types of experiments either
singly or in concert may provide a overview of how
developmentally regulated promoters function in the E.
japonicum.

The results obtained during this thesis project are but
a small step toward an understanding of the molecular
mechanisms which regulate bacterial gene expression during
the development of nitrogen-fixing nodules in the
Bradyrhizobigg japonicum/soybean symbiosis. The isolation
and characterization of promoters of E. japonicum genes
which are regulated during symbiotic development should,
however, provide a starting point for research on the more
difficult and complex question of how the expression of such

promoters and genes is controlled.

REFERENCES

Adams, T.H. 1986. Transcriptional regulation of nitrogen
fixation and nitrogen assimilation genes in Bradyrhizobium
japonicum. PhD. Thesis, Michigan State Univ.

Adams, T.H. and D.K. Chelm. 1988. Effects of oxygen levels
on the transcription of nif and gln genes in Bradyrhiggbigg
japonicum. J. Gen. Microbiol. In press.

Adams, T.H. and D.K. Chelm. 1984. The nifH and nifDK
promoter regions from Rhizobium jgponicum share structural
homologies with each other and with nitrogen-regulated

promoters from other organisms. J. M01. and Appl. Genet.
2:392-405.

 

Adams, T.H., C.R. McClung and D.K. Chelm. 1984. Physical
organization of the Bragyrhizobiug japonicug nitrogenous
gene region. J. Bacteriol. 159:857-862.

Aguilar, O.M., D. Kapp, and A. Puhler. 1985.
Characterization of a Rhizobium meliloti fixation gene

(fixF) located near the common nodulation region. J.
Bacteriol. 164:245-254.

 

Amersham Corporation. 1983. M13 cloning and sequencing
handbook. (Amersham Corp., Arlington Heights, IL).

Applebaum, E.R., E. Johansen, and N. Chartrain. 1985.
Symbiotic mutants of USDA191, a fast-growing Rhizobium that
nodulates soybeans. Mol. Gen. Genet. 201:454-461.

Appleby, C.A. 1969a. Electron transport systems of Rhizobium
japonicum I. Haemoprotein P-450, other co-reactive pigments,
cytochromes and oxidases in bacteroids from Nz-fixing root
nodules. Biochim. Biophys. Acta. 172:71-87.

Appleby, C.A. 1969b. Electron transport systems of Rhizobium
jgponicum II. Rhizobium haemoglobin, cytochromes and
oxidases in free-living (cultured) cells. Biochim. Biophys.
Acta. 172:88-105.

 

 

Appleby, C. 1984. Leghemoglobin and Rhizobium respiration.
Annu. Rev. Plant Physiol. 35:443-478.

Arwas, R., I.A. McKay, R.R.P. Rowney, M.J. Dilworth, and
A.R. Glenn. 1985. Properties of organic acid utilization
mutants of Rhizobium leguminosarum strain 300. J. Gen.
Microbiol. 131:2059-2066.

Auger, 8., D. Baulcombe and D.P.S. Verma. 1979. Sequence
complexities of the poly(A)-containing mRNA in uninfected

122

123

soybean root and the nodule tissue developed due to the
infection by Rhizobium. Biochim. Biophys. Acta. 563:496-507.

Auger, S., and D.P.S. Verma. 1981. Induction and expression
of nodule-specific host genes in effective and ineffective
root nodules of soybean. Biochemistry 20:1300-1306.

Ausubel, F.M. 1984a. Regulation of nitrogen fixation genes.
Cell 37:5-6.

Ausubel, F.M. 1984b. Developmental genetics of the
Rhizobium/legume symbiosis. tg Microbial Development. R.
Losick and L. Shapiro eds., (Cold Spring Harbor Laboratory,
Cold Spring Harbor, NY) pp. 275-298.

Ausubel, F.M., W.J. Buikema, C.D. Earl, J.A. Klingensmith,
B.T. Nixon, and T. Szeto. 1985. Organization and regulation
of Rhizobium meliloti and Parasponia Bradyrhizobium nitrogen
fixation genes. tg Nitrogen Fixation Research Progress. H.J.
Evans, P.J. Bottomley, W.E. Newton, eds., (Martinus Nijhoff,
Dordrecht) pp. 165—171.

Aviv, H. and P. Leder. 1972. Purification of biologically
active globin messenger RNA by chromatography on
oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci.
U.S.A. 69:1408-1412.

Baldwin, T.O., T. Berends, T.A. Bunch, T.F. Holzman, S.K.
Rausch, L. Shamansky, M.L. Treat, and M.M. Ziegler. 1984.
Cloning of the luciferase structural genes from Vibrio
harveyii and expression of bioluminescence in Eggherichia
coli. Biochemistry. 23:3663-3667.

Barr, P.J., R.M. Thayer, P. Laybourn, R.C. Najarian, F.
Seela, and D.R. Tolan. 1986.
7-deaza-2’-deoxyguanosine-5’-triphosphate: enhanced
resolution in M13 dideoxy sequencing. Biotechniques.
4:428-432.

Bellofatto, V., L. Shapiro, and D.A. Hodgson. 1984.
Generation of a Tng promoter probe and its use in the study

of gene expression in Caulobacter crescentus. Proc. Natl.
Acad. Sci. U.S.A. 81:1035-1039.

Benton, W.D., and R.W. Davis. 1977. Screening Agt
recombinant clones by hybridization to single plaques tg
situ. Science 196:180-182.

Berg. D.R. 1977. Insertion and excision of the transposable
kanamycin resistance determinant Tng. tg DNA Insertion
Elements, Plasmids and Episomes. A.I. Bukhari, J.A. Shapiro
and S.L. Adhya eds. (Cold Spring Harbor Laboratory, Cold
Spring Harbor, New York) pp. 205-212.

124

Bergmann, H., E. Preddie, and D.P.S. Verma. 1983.
Nodulin-35: a subunit of specific uricase (uricase II)

induced and localized in the uninfected cells of soybean
nodules. EMBO J. 2:2333-2339.

Beringer, J.E., J.L. Benyon, A.V. Buchanan-Wollaston, and
A.R.B. Johnston. 1978. Transfer of the drug resistance
transposon TnE to Rhizobium. Nature 276:633-634.

 

Berk, A.J., and P.A. Sharp. 1977. Sizing and mapping early
adenovirus mRNAs by gel electrophoresis of $1
endonuclease-digested hybrids. Cell 12:721-732.

Bishop, 0.0. 1972. DNA-RNA hybridization. tg Karolinska
Symposia on Research Methods in Reproductive Endochrinology.

5th Symposium. (Karolinska Institute, Stockholm). pp.
247-273.

Bisseling, T., R.C. van den Bos, A. van Kammen, M. van der
Ploeg, P. van Duijn, and A. Houwers. 1977.
Cytofluorometrical determination of the DNA contents of
bacteroids and corresponding broth-cultured Rhizobium
bacteria. J. Gen. Microbiol. 101:79-84.

Bisseling, T., R.C. van den Bos, M.W. Weststrate, M.J.J.
Hakkaart and A. van Kammen. 1979. Development of the
nitrogen-fixing and protein-synthesizing apparatus of
bacteroids in pea root nodules. Biochim. Biophys. Acta.
562:515-526. .

Bisseling, T., C. Been, J. Klugkist, A. van Kammen, K.
Nadler. 1983. Nodule-specific host proteins in effective and
ineffective root nodules of Pisum sativum. EMBO J.
2:961-966.

Bisseling, T., F. Govers, R. Wyndaele, J-P. Nap, J.W.
Taanman, A. Van Kammen. 1984. Expression of nodulin genes
during nodule development from effective and ineffective
root nodules. tg Advances in Nitrogen Fixation Research. C.
Veeger and N.E. Newton, eds., (Nijhoff/Junk, The Hague) pp.
579-586.

Bolton, 8., D. Higgisson, A. Harrington and F. O’Gara. 1986.
Dicarboxylic acid transport in Rhizobium meliloti: isolation
of mutants and cloning of dicarboxylic acid transport genes.
Arch. Microbiol. 144:142-146.

Borck, K., J.D. Beggs, W.J. Brammar, A.S. Hopkins, and N.E.
Murray. 1976. The construction in vitro of transducing
derivatives of phage lambda. Mol. Gen. Genet. 146:199-207.

Brewin, N.J., J.G. Robertson, E.A. Wood, 8. Wells, A.P.
Larkins, G. Galfre, and G.W. Butcher. 1985. Monoclonal

125

antibodies to antigens in the peribacteroid membrane from
Rhizobium-induced root nodules of pea cross-react with
plasma membranes and Golgi bodies. EMBO J. 4:605—611.

Brooks, S.J., J. Imperial, and N.J. Brill. 1985. Biochemical
genetics of nitrogen fixation in Klebsiella pneumoniae. tg
Nitrogen Fixation and C02 Metabolism. P.R. Ludden and J. E.
Burris eds. (Elsevier, New York), pp. 65-74.

Brown, J.N. and J.N. Reeve. 1985. Polyadenylated, noncapped
RNA from the archaebacterium Methanococcus vanniellii. J.
Bacteriol. 162:909-917.

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

 

 

Duck, M., S. Miller, M. Drummond, and R. Dixon. 1986.
Upstream activator sequences are present in the promoters of
nitrogen fixation genes. Nature 320:374-378.

Buikema, N.J., W.W. Szeto, P.V. Lemley, R.H. Orme-Johnson,
and F.M. Ausubel. 1985. Nitrogen fixation specific
regulatory genes of Klebsiella pneumoniae and Rhizobium
meliloti share homology with the general nitrogen regulatory
gene ntrC of E. pneumonia. Nucleic Acids Res. 13:4539—4555.

 

Burris, R.H. 1980. The global nitrogen budget - science or
seance? tg Nitrogen Fixation Volume I: Free-living Systems
and Chemical Models. N.E. Newton and R.H. Orme-Johnson eds.
(Univ. Park Press, Baltimore) pp. 7-16.

Carlson, T.A. and D.K. Chelm. 1986. Apparent eukaryotic
origin of glutamine synthetase II from the bacterium
Bradyrhizobigg japonicum. Nature 322:568-570.

Carlson, T.A., M.L. Guerinot and D.K. Chelm. 1985.
Characterization of the gene encoding glutamine synthetase I'

(glnA) from Bradyrhizobiug japonicum. J. Bacteriol.
162:698-703.

Carlson, T.A., G.B. Martin, and D.K. Chelm. 1987.
Differential transcription of the two glutamine synthetase

genes of Bradyrhizobium japonicum. J. Bacteriol.‘
169:5861-5866.

Casadaban, M.J. and S.N. Cohen. 1979. Lactose genes fused to
exogenous promoters in one step using a Mu-lac
bacteriophage: tg vivo probe for transcriptional control
sequences. Proc. Natl. Acad. Sci. U.S.A. 76:4530-4533.

126

Cen, Y., G.L. Bender, M.J. Trinick, N.A. Morrison, K.F.
Scott, P.M. Gresshoff, J. Shine, and B.H. Rolfe. 1982.
Transposon mutagenesis in rhizobia which can nodulate both

legumes and the nonlegume Parasponia. Appl. Environ.
Microbial. 37:233-236.

Chelm, D.K. and R.B. Hallick. 1976. Changes in the
expression of the chloroplast genome of Euglena gracilis
during chloroplast development. Biochemistry 15:593-599.

 

Chelm, D.K., P.J. Hoben and R.B. Hallick. 1977. Expression
of the cholorplast ribosomal RNA genes of Euglena gracilis
during chloroplast development. Biochemistry 16:776-781.

Ching, T.M., S. Hedtke and W. Newcomb. 1977. Isolation of
bacteria, transforming bacteria, and bacteroids from soybean
nodules. Plant Physiol. 60:771-774.

Chua, K-Y., C.E. Pankhurst, P.E. MacDonald, D.H. Hopcroft,
D.W. Jarvis, and B.D. Scott. 1985. Isolation and
characterization of transposon TnE—induced symbiotic mutants
of Rhizobium loti. J. Bacteriol. 162:335-343.

Clancy, M.J., B. Buten—Magee, D.J. Straight, A.K. Kennedy,
R.M. Partridge and P.T. Magee. 1983. Isolation of genes
expressed preferentially during sporulation in the yeast

Saccharagyceg cerevigiae. Proc. Natl. Acad. Sci. U.S.A.
80:3000-3004.

Clewell, D.B. and D.R. Helinski. 1972. Effect of growth
conditions on the formation of the relaxed complex of
supercoiled CalEl deoxyribonucleic acid and protein in
Escherichia coli. J. Bacteriol. 110: 1135-1146.

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

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.

David, M., 0. Domergue, P. Pognonec, and D. Kahn. 1987.
Transcription patterns of Rhizobiug meliloti symbiotic
plasmid pSym: identification of nifA-independent ttg genes.
J. Bacterial. 169:2239-2244.

Davis, R.H., D. Botstein and J.R. Roth. 1980. Advanced
bacterial genetics. (Cold Spring Harbor Laboratory, Cold
Spring Harbor, N.Y.)

Dazzo, F.B. and A.E. Gardiol. 1984. Host specificity in
Rhizobium-legume interactions. tg Genes Involved in

127

Microbe-Microbe Interactions. D.P.S. Verma and Th. Hahn.
eds. (Springer-Verlag, New York) pp. 3-31.

De Bruijn, F.J. and J.R. Lupski. 1984. The use of transposon
TnE mutagenesis in the rapid generation of correlated
physical and genetic maps of DNA segments cloned into
multicopy plasmids - a review. Gene 27:131-149.

Denarie, 0., G. Truchet, and B. Bergeron. 1976. Effects of
some mutations on symbiotic properties of Rhizobium. tg
Symbiotic nitrogen fixation. P.S. Nutman ed. (Cambridge
Univ. Press, New York) pp. 47-61.

Ditta, G. 1985. TnE mapping of Rhizobium nitrogen fixation
genes. Methods Enzymol. 118:519-528.

 

Ditta, G., S. Stanford, D. Corbin, and D.R. Helinski. 1980.
Broad host range DNA cloning system for Gram-negative

bacteria: construction of a gene bank of Rhizobium meliloti.
Proc. Natl. Acad. Sci. U.S.A. 77:7347-7351.

Ditta, G., E. Virts, A. Palomares, and C-H. Kim. 1987. The
nifA gene of Rhizobium meliloti is oxygen regulated. J.
Bacterial. 169:3217-3223.

Dixon, R.A. 1984. Genetic complexity of nitrogen fixation.
J. Gen. Microbial. 130:2745-2755.

Downie, J. A., G. Hombrecher, Q.S. Ma, C.D. Knight, 8.
Wells, and A.W.B. Johnston. 1983. Cloned nodulation genes of

Rhigabium leguminosgrum determine host-range specificity.
Mol. Gen. Genet. 190:359-365.

 

Downie, J.A. and A.W.B. Johnston. 1986. Nodulation of
legumes by Rhizobium: the recognized root? Cell 47:153-154.

Dubnau, E., N. Ramakrishna, K. Cabane and 1. Smith. 1981.
Cloning of an early’sporulatian gene in Bacillus subtilis.
J. Bacteriol. 147:622-632.

Dudley, N.E., T.A. Jacobs, and S.R. Long. 1987. Microscopic
studies of cell divisions induced in alfalfa roots by
Rhizobium meliloti. Planta 171:289-301.

Duncan, M.J. 1981. Properties of TnE—induced carbohydrate
mutants in Rhizobium meliloti. J. Gen. Microbial. 122:61-67.

Dworkin, M. 1985. Developmental Biology of the Bacteria.
(Benjamin/Cummings Publishing Company, Inc., Reading, Mass.)
pp. 22-49.

Dylan, T., L. Ielpi, S. Stanfield, L. Kashyap, C. Douglas,
M. Yanofsky, E. Nester, D.R. Helinski, and G. Ditta. 1986.

128

Rhizobium meliloti genes required for nodule development are
related to chromosomal virulence genes in Egrobacteriug
tumefaciens. Proc. Natl. Acad. Sci. U.S.A. 83:4403-4407.

 

Ebeling, 8., M. Hahn, H-M. Fischer, and H. Hennecke. 1987.
Identification of nifE-, nifN- and nifS-like genes in
Bradyrhizobiug japanicum. Mol. Gen. Genet. 207:503-508.

Elkan, G.H. and R.A. Usanis. 1971. Theoretical
deoxyribonucleic acid homology between strains of Rhizobium
japanicum. Internatl. 0. System. Bacterial. 21:295-298.

Engelke, T., M.N. Jagadish, and A. Puhler. 1987. Biochemical
and genetical analysis of Rhigobium meliloti mutants

defective in Ca-dicarboxylate transport. J. Gen. Microbial.
133:3019-3029.

 

Esposito, R.E. and S. Klaphalz. 1981. Meiosis and ascospore
development. tg The Molecular Biology of the Yeast
Saccharomyces. J.N. Strathern, E.W. Jones, and J.R. Broach.
(Cold Spring Harbor Laboratory, Cold Spring Harbor, NY) pp.
211-287.

Ferrari, E.A., D. Lang, E. Ferrari, and J.A. Hach. 1982.
Molecular cloning of the 52008 sporulation locus in
bacteriophage lambda. J. Bacteriol. 147:809—814.

Finan, T.M., A.M. Hirsh, J.A. Leigh, E. Johansen, G.A.
Kuldau, S. Deegan, G.C. Walker, and E.R. Signer. 1985.
Symbiotic mutants of Rhizobium meliloti that uncouple plant
from bacterial differentiation. Cell 40:869-877.

Finan, T.M., J.M. Wood, and D.C. Jordan. 1983. Symbiotic
properties of Ca-dicarboxylic acid transport mutants of
Rhizobium leguminosarum. J. Bacteriol. 154:1403-1413.

Firmin, J.L., R.E. Wilson, L. Rossen and A.W.B. Johnston.
1986. Flavanoid activation of nodulation genes in Rhizobium

reversed by other compounds present in plants. Nature
324:90-92.

Firtel, R.A. 1972. Changes in the expression of single-copy
DNA during development of the cellular slime mold
Dictyostelium discoideum. J. Mol. Biol. 66:363-377.

Fisher, R.M., A. Alvarez-Morales, and H. Hennecke. 1986. The
pleiotropic nature of symbiotic regulatory mutants:
Bradyrhtgobigg japonicum nifA gene is involved in control of

nif gene expression and formation of determinate symbiosis.
EMBO J. 5:1165-1173.

Friedman, A.M., S.R. Long, S.E. Brown, W.J. Buikema, and
F.M. Ausubel. 1982. Construction of a broad host range

129

cosmid vector and its use in the genetic analysis of
Rhizobium mutants. Gene 18:289-296.

Franssen, H.J., J-P. Nap, T. Gloudemans, W. Stiekema, H. van
Dam, F. Govers, J. Louwerse, a. van Kammen, and T.
Bisseling. 1987. Characterization of cDNA for nodulin-75 of
soybean: a gene product involved in early stages of root
nodule development. Proc. Natl. Acad. Sci. U.S.A.
84:4495-4499.

Fuhrmann, M., H-M. Fischer, and H. Hennecke. 1985. Mapping
of Rhiggbium japoniggg nifB-, fixBC-, and fixA-like genes
and identification of the fixA promoter. Mol. Gen. Genet.
199:315-322.

Georgopaulas, C., K. Tilly, D. Drahas, and R. Hendrix. 1982.
The 866.0 protein of Escherichtg coli is the product of the
dnaK gene. J. Bacteriol. 149:1175-1177.

Gianni, M. and A. Galizzi. 1986. Isolation of genes
preferentially expressed during Bacillis gubtilis spore
outgrowth. J. Bacteriol. 165:123-132.

Glenn, A.R., I.A. McKay, R. Arwas, and M.J. Dilworth. 1984.
Sugar metabolism and the symbiotic properties of carbohydate

mutants of Rhizobium leguminosarum. J. Gen. Microbiol.
130:239-245.

Gloudemans, T., S. de Vries, H-J. Bussink, N.S.A. Malik,
H.J. Franssen, J. Louwerse, and T. Bisseling. 1987. Nodulin
gene expression during soybean (Glycine max) nodule
development. Plant Mol. Biol. 8:395-403.

 

Golden, 5.8. and L.A. Sherman. 1983. A hybrid plasmid is a
stable cloning vector for the cyanobacterium Anacystis
nidulans R2. J. Bacterial. 155:966-972.

Gopalakrishna, Y. and N. Sarkar. 1982. Characterization of
polyadenylate-containing ribonucleic acid from Bacillus
subtilis. Biochemistry 21:2723-2729.

Govers, F., T. Gloudemans, M. Moerman, A. van Kammen, and T.
Bisseling. 1985. Expression of plant genes during the
development of pea root nodules. EMBO J. 4:861-867.

Govers, F., J-P. Nap, M. Moerman, H.J. Franssen, A. van
Kammen and T. Bisseling. 1987. cDNA cloning and
developmental expression of pea nodulin genes. Plant Mol.
Biol. 8:425-435.

Grunstein, M. and D. Hogness. 1975. Colony hybridization: a
method for the isolation of cloned DNAs that contain a
specific gene. Proc. Natl. Acad. Sci. U.S.A. 72:3961-3965.

130

Guerinot, M.L. and D.K. Chelm. 1986. Bacterial
5-aminolevulinic acid synthase activity is not essential for
leghemoglobin formation in the soybean/Bradyrhizobigg

japanicum symbiosis. Proc. Natl. Acad. Sci. U.S.A.
83:1837-1841.

Guerinot, M.L. and D.K. Chelm. 1987. Molecular aspects of
the physiology of symbiotic nitrogen fixation in legumes. tg
Plant-Microbe Interactions Molecular and Genetic
Perspectives Volume 2. T. Kosuge and E.W. Nester eds.
(Macmillan Publ. Co., New York) pp. 103-146.

Gussin, G.N., C.W. Ronson, and F.M. Ausubel. 1986;
Regulation of nitrogen fixation genes. Annu. Rev. Genet.
55:567-589.

Hahn, M. and H. Hennecke. 1987. Conservation of a symbiotic
DNA region in soybean root nodule bacteria. Appl. Environ.
Microbiol. 53:2253-2255.

Hardy, R.H. 1980. The global carbon and nitrogen economy. t_
Nitrogen Fixation Volume I: Free-living Systems and Chemical
Models. W.E. Newton and W.R. Orme-Johnson eds. (Univ. Park

Press, Baltimore) pp. 3-5.

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.B. Holsten, E.R. Jackson, and R.C. Burns.
1968. The acetylene-ethylene assay for nitrogen fixation:

laboratory and field evaluation. Plant Physiol.
43:1185-1207.

Hawley, D.K. and W.R. McClure. 1983. Compilation and
analysis of Escherichia coli promoter DNA sequences. Nucleic
Acids Res. 11:2237-2255.

Heberlein, G.T., J. de Ley, and R. Tijtgat. 1967.
Deoxyribonucleic acid homology and taxonomy of

Egrabacterium, Rhtgpbiug and Chromobacterium. J. Bacteriol.
94:116-124.

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

Hennighausen, L. and H. Lubon. 1987. Interaction of protein
with DNA tg vitro. Methods Enzymol. 152:721-735.

Hershey, H.P., J.T. Colbert, J.L. Lissemore, R.F. Barker,
and P.R. Quail. 1984. Molecular cloning of cDNA for Avena
phytochrame. Proc. Natl. Acad. Sci. U.S.A. 81:2332-2336.

131

Hirsch, A.M. and C.A. Smith. 1987. Effects of Rhizobium
meliloti nif and fix mutants on alfalfa root nodule
development. J. Bacteriol. 1691137—1146.

Hirshman, J. P.-K. Wong, K. Sei, J. Keener, and S. Kustu.
1985. Products of nitrogen regulatory genes ntrA and ntrC of
enteric bacteria activate glnA transcription tg vitro:
evidence that the ntrA product is.a sigma factor. Proc.
Natl. Acad. Sci. U.S.A. 82:7525-7529.

 

Hahn, B. 1979. tg vitro packaging of lambda and cosmid DNA.
Methods Enzymol. 68:299-309.

 

Holben, W.E., J.K. Jannson, D.K. Chelm, and J.M. Tiedje.
1988. DNA probe method for the detection of specific
microorganisms am the soil bacterial community. Appl.
Environ. Microbiol. In press.

Holmes, D.S., and M. Quigley. 1981. A rapid boiling method
for the preparation of bacterial plasmids. Anal. Biochem.
114:193-197.

Ham, S.S.M., S.L. Uratsu, and F. Haang. 1984. Transposon
TnE-induced mutagenesis of Rhizobium japanicum yielding a
wide variety of mutants. J. Bacteriol. 159:335-340.

Honma, M.A., C.A. Smith, A.M. Hirsch, N. Lang-Unnasch, and
F.M. Ausubel. 1985. Evidence for additional nodD-like genes
in E. meliloti. tg Nitrogen Fixation Research Progress. H.J.
Evans, P.J. Bottomley, and W.E. Newton, eds., (Martinus
Nijhoff Publishers. BV, Dordecht, The Netherlands). p. 120.

Ingraham, J.L., O. Maalae, and F.C. Neidhardt. 1983. Growth
of the Bacterial Cell. (Sinauer Assoc. Inc., Sunderland,
MA). p. 3.

Innes, R.W., P.L. Kuempel, J. Plazinske, H. Canter-Cremers,
B.G. Rolfe, and M.A. Djordjevic. 1985. Plant factors induce
expression of nodulation and host-range genes in Rhizobium

trifolii. Mol. Gen. Genet. 201:426-432.

Jacobs, T.W., T.T. Egelhoff, and S.R. Long. 1985. Physical
and genetic map of a Rhizobium meliloti nodulation gene

region and nucleotide sequence of nodC. J. Bacteriol.
162:469-476.

Jefferson, R.A., S.M Burgess, and D. Hirsh. B-Glucuronidase
from Escherichig coli as a gene-fusion marker. Proc. Natl.
Acad. U.S.A. 83:8447-8451.

Jefferson, R.A., T.A. Kavanagh and M.W. Bevan. 1987.
Beta-glucuronidase (GUS) as a sensitive and versatile gene

132

fusion marker in plants. J. Cell. Biochem. Supplement 118.
p. 57.

Johnston, A.W.B. and J.A. Downie. 1984. What is a nif
promoter? Trends Biochem. Sci. 9:367-368.

 

Johnson, G.V., H.J. Evans, and T.M. Ching. 1966. Enzymes of
the glyoxylate cycle in rhizobia and nodules of legumes.
Plant Physiol. 41:1330-1336.

Kaiser, D. 1986. Control of multicellular development:
Dictyastelium and Myxococcus. Annu. Rev. Genet. 20:539-566.

Kaiser, D., C. Maniol, and M. Dworkin. 1979. Myxobacteria:
cell interactions, genetics, and development. Annu. Rev.
Microbiol. 33:595-639.

Kaluza, K., M. Hahn and H. Hennecke. 1985. Repeated
sequences similar to insertion elements clustered around the

nif region of the Rhizobiug japonicum genome. J. Bacteriol.
162:535-542.

Keister, D.L. 1975. Acetylene reduction by pure cultures of
Rhizobia. J. Bacteriol. 123:1265-1268.

Kenyon, 0.0. and G.C. Walker. 1980. DNA-damaging agents
stimulate gene expression at specific loci in Escherichig
coli. Proc. Natl. Acad. Sci. U.S.A. 77:2819-2823.

 

Kerjan, P. and J. Szulmajster. 1980. Isolation and
characterization of polyadenylated RNA species from
sporulating cells of Bacillug subtilis. Biochem. Biophys.
Res. Commun. 93:201-208.

Klucas R.V. and H.J. Evans. 1968. An electron donor system
for nitrogenase-dependent acetylene reduction by extracts of‘
soybean nodules. Plant Physiol. 13:1458-1460.

Kondorosi, E., Z. Banfalvi, and A. Kondorosi. 1984. Physical
and genetic analysis of a symbiotic region of Rhizobiug
meliloti: identification of nodulation genes. Mol. Gen.
Genet. 193:445-452.

Kondorosi, E. and A. Kondorosi. 1986. Nodule induction on
plant roots by Rhizobium. Trends Biochem. Sci. 11:296-299.

Kopachik, W., L.G. Bergen, and S.L. Barclay. 1985. Genes
selectively expressed in proliferating Dictyogteligg
amoebae. Proc. Natl. Acad. Sci. U.S.A. 82:8540-8544.

Kosslak, R.M. R. Bookland, J. Darkei, J.E. Paaren, E.R.
Applebaum. 1987. Induction of Bradyrhiggbium japonicum

 

133

common nod genes by isoflavones isolated from Glycine max.
Proc. Natl. Acad. Sci. U.S.A. 84:7428-7432.

 

Krol, A.J.M., J.G.J. Hontelez, R.C. Van den Dos, and A. Van
Kammen. 1980. Expression of large plasmids in the

endosymbiatic form the Rhizobium leguminosarug. Nucleic
Acids Res. 8:4337-4347.

Kraos, L. and D. Kaiser. 1984. Construction of TnE tgg, a
transposon that fuses lacZ expression to exogenous
promoters, and its introduction into Myxococcus xanthus.
Proc. Natl. Acad. Sci. U.S.A. 81:5816-5820.

Kraos, L., A. Kuspa, and K. Kaiser. 1986. A global analysis
of developmentally regulated genes in Myxococcus xanthus.
Devel. Biol. 117:252-266.

Kurtz, S. and S. Lindquist. 1984. Changing patterns of gene
expression during sporulation in yeast. Proc. Natl. Acad.
Sci. U.S.A. 81:7323—7327.

Kurz, W.G.W. and T.A. Larue. 1975. Nitrogenase activity in
rhizobia in absence of plant host. Nature 256:407-409.

Kuykendall, L.D. and G.H. Elkan. 1976. Rhizobium japgnicum
derivatives differing in nitrogen-fixing efficiency and
carbohydrate utilization. Appl. Environ. Microbiol.
32:511-519.

Lamb, J.W. and H. Hennecke. 1986. In Bradyrhizobigg
japonicum the common nodulation genes, nodABC are linked to
nifA and fixA. Mol. Gen. Genet. 202:512-517.

 

Lang-Unnasch, N. and F.M. Ausubel. 1985. Nodule-specific
polypeptides from effective alfalfa root nodules and from

ineffective nodules lacking nitrogenase. Plant Plysiol.
77:833—839.

Lang-Unnasch, N., K. Dunn, and F.M. Ausubel. 1986. Symbiotic
nitrogen fixation: developmental genetics of nodule
formation. tg Cold Spring Harbor Symposium on Quantitative
Biology. (Cold Spring Harbor Laboratory, Cold Spring Harbor,
N.Y.) Vol. 50. pp. 555-563.

Law, J.R. and R.A. Slepecky. 1961. Assay of poly-3
-hydroxybutyric acid. J. Bacteriol. 82:33-36. '

Legocki, R.P. and D.P.S. Verma. 1979. A nodule-specific

plant protein (Nodulin-35) from soybean. Science
205:190-193.

134

Legocki, R.P. and D.P.S. Verma. 1980. Identification of
"nodule-specific" host proteins (nodulins) involved in the
development of rhizobium-legume symbiosis. Cell 20:153-163.

Legocki, R.P., M. Legocki, T.O. Baldwin, and A.A. Szalay.
1986. Bioluminescence in soybean root nodules: Demonstration
of a general approach to assay gene expression tg vivo by

using bacterial luciferase. Proc. Natl. Acad. Sci. U.S.A.
83:9080-9084.

Leigh, J.A., E.R. Signer, and G.C. Walker. 1985.
Exapolysaccharide-deficient mutants of Rhizobium meliloti
that form ineffective nodules. Proc. Natl. Acad. Sci. U.S.A.
82:6231-6235.

Leong, S.A., G.S. Ditta, and D.R. Helinski. 1982. Heme
biosynthesis in Rhigobium. Identification of a cloned gene

coding for d-aminolevulinic acid synthetase from Rhizobium
meliloti. J. Biol. Chem. 257:8724-8730.

 

Long, S.R. 1984. Genetics of Rhizobium nodulation. t_
Plant-Microbe Interactions Molecular and Genetic
Perspectives Vol. I., T. Kosuge and E.W. Nester eds.
(Macmillan Publ. Co., New York). pp. 265-306.

 

Long, S.R., W.J. Buikema and F.M. Ausubel. 1982. Cloning of
Rhizobium meliloti nodulation genes by direct
complementation of Nod‘ mutants. Nature 298:485-488.

Losick, R. and P. Youngman. 1984. Endospore formation in
Bacillus. tg Microbial Development. R. Losick and L. Shapiro
eds., (Cold Spring Harbor Laboratory, Cold Spring Harbor,
N.Y.) pp. 63-88.

Losick, R., P. Youngman, and P.J. Piggat. 1986. Genetics of
endospore formation in Bacillus subtilis. Annu. Rev. Genet.
20:625-670.

Lave, P.E., M.J. Lyle and R.E. Yasbin. 1985.
DNA-damage-inducible (din) loci are transcriptionally

activated in competent Bacillus subtilis. Proc. Natl. Acad.
Sci. U.S.A. 82:6201-6205.

Lullien, V., D.G. Barker, P. de Lajudie, and T. Huguet.
1987. Plant gene expression in effective and ineffective
root nodules of alfalfa (Medicggg satigg). Plant Mol. Biol.
10:469-478.

MacKenzie, C.R., W.J. Vail, and D.C. Jordan. 1973.
Ultrastructure of free-living and nitrogen-fixing forms of

Rhizobium meliloti as revealed by freeze-etching. J.
Bacteriol. 113:387-393.

135

Magasanik, B. 1982. Genetic control of nitrogen assimilation
in bacteria. Annu. Rev. Genet. 16:135-168.

Majumdar, P.K. and D.A. McFadden. 1984. Polyadenylated mRNA
from the photosynthetic procaryote Rhodogpirillgg rubrum. J.
Bacteriol. 157:795-801.

Manian, 8.8. and F. O’Gara. 1982. Induction and regulation
of ribulosebisphosphate carboxylase activity in Rhtgpbium

japonicum during formate dependent growth. Arch. Microbiol.
131:51-54.

 

Maniatis, T., E.F. Fritsch and J. Sambrook. 1982. Molecular
cloning. (Cold Spring Harbor Laboratory. Cold Spring Harbor,
N.Y.)

Marmur, J., and P. Doty. 1962. Determination of the base
composition of deoxyribonucleic acid from its thermal
denaturatian temperature. J. Mol. Biol. 5:109-118.

Marrs, B. and H. Gest. 1973. Genetic mutations affecting the
respiratory electron-transport system of the photosynthetic

bacterium Rhodopsegdomonas capsulata. J. Bacteriol.
114:1045-1051.

Maupin, P. and T.D. Pollard. 1983. Improved preservation and
staining of HeLa cell actin filaments, clathrin-coated
membranes, and other cytoplasmic structures by tannic
acid-g1utaraldehyde-saponin fixation. J. Cell. Biol.
9651—62.

McComb, J.A., J. Elliot, and M.J. Dilworth. 1975.
Nitrogenase activity in rhizobia in absence of plant host.
Nature 256:409-410.

Meade, R.M., S.R. Long, G.B. Ruvkun, S.E. Brown, and F.M.
Ausubel. 1982. Physical and genetic characterization of
symbiotic and auxtrophic mutants of Rhizobigg meliloti

induced by transposon TnE mutagenesis. J. Bacteriol.
149:114-122.

Mellor, R.B., T.M.I.E. Christensen, and D. Werner. 1986.
Choline kinase II is present only in nodules that synthesize

stable peribacteroid membranes. Proc. Natl. Acad. Sci.
U.S.A. 83:659-663.

Messing, J. 1983. New M13 vectors for cloning. Methods in
Enzymal. 101:20-78.

Messing, J., R. Crea, and P.A. Seeburg. 1981. A system for
shotgun DNA sequencing. Nucleic Acids Res. 9:309-321.

136

Messing, J., and J. Vieira. 1982. A new pair of M13 vectors
for selecting either DNA strand of double-digest restriction
fragments. Gene 19:269-276.

Meyer, M.C. and 8.6. Pueppke. 1980. Differentiation of
Rhizobium japonicum strain derivatives by antibiotic
sensitivity patterns, lectin binding, and utilization of
biochemicals. Can. J. Microbiol. 26:606-612.

Miller, J.H. 1972. Experiments in molecular genetics. (Cold
Spring Harbor Laboratory, Cold Spring Harbor, N.Y.).

Moerman, M., J-P. Nap, F. Govers, R. Schilperoort, A. van
Kammen, and T. Bisseling. 1987. Rhizobium nod genes are
involved in the induction of two early nodulin genes in
Vicia sativa root nodules. Plant Mol. Biol. 9:171-179.

 

Morrison, N. and D.P.S. Verma. 1987. A block in the
endocytasis of Rhizobium allows cellular differentiation in
nodules but affects the expression of some peribacteroid
membrane nodulins. Plant Mol. Biol. 9:185-196.

Mortenson, L.E. and R.N.F. Thornley. 1979. Structure and
function of nitrogenase. Annu. Rev. Biochem. 48:387-418.

Mulligan, J.T. and S.R. Long. 1985. Induction of Rhiggbiug
meliloti nodC expression by plant exudate requires nodD.
Proc. Natl. Acad. Sci. U.S.A. 82:6609-6613.

 

Nadler, K.D. and Y.J. Avissar. 1977. Heme synthesis in
soybean root nodules. I. On the role of bacteroid
5-amino-levulinic acid dehydratase in the synthesis of the
heme of leghemoglobin. Plant Physiol. 60:433-436.

Nakazato, H., S. Venkatesan, and M. Edmonds. 1975.

Polyadenylic acid sequences in E. coli messenger RNA. Nature
256:144-146.

Neidhardt. F.C., R.A. VanBogelen, and V. Vaughn. 1984. The
genetics and regulation of heat-shock proteins. Annu. Rev.
Genet. 18:295-329.

Nixon, B.T., C.W. Ronson, and F.M. Ausubel. 1986.
Two-component regulatory systems responsive to environmental
stimuli share strongly conserved domains with the nitrogen
assimilation regulatory genes ntrB and ntrC. Proc. Natl.
Acad. Sci. U.S.A. 83:7850-7854.

O’Brien, M.R., P.M. Kirshbom, and H.J. Meier. 1987.
TnE-induced cytochrome’mutants of Bradyrhtgobium japonicum:

effects of the mutations on cells grown symbiotically and in
culture. J. Bacteriol. 169:1089-1094.

137

O’Brien, M.R. and R.J. Meier. 1987. Isolation of a
cytochrome aaa gene from Bradyrhizobium jgponicum. Proc.
Natl. Acad. Sci. U.S.A. 84:3219-3223.

Olson, E.R., M.J. Sadowsky, and D.P.S. Verma. 1985.
Identification of genes involved in the Rhizobium-legume
symbiosis by Mu-dI (Kan, tgg)-generated transcription
fusions. Bio/technology 3:143-149.

 

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

Pain, A.N. 1979. Symbiotic properties of
antibiotic-resistant and auxotrophic mutants of Rhtgpbium
leguminosarum. J. Appl. Bacterial. 47:53-64.

 

Perkins, J.D. and P.J. Youngman. 1986. Construction and
properties of TN917-lac, a transposon derivative that
mediates transcriptional gene fusions in Bacillus subtilis.
Proc. Natl. Acad. Sci. U.S.A. 83:140-144.

 

Peters, N.K., J.W. Frost and S.R. Lang. 1986. A plant
flavane, luteolin, induces expression of Rhizobium meliloti
nodulation genes. Science 233:977-980.

Poncz, M., D. Solowiejczyk, M. Ballantine, E. Schwartz, S.
Surrey. 1982. "Nonrandom" DNA sequence analysis in
bacteriophage M13 by the dideoxy chain-termination method.
Proc. Natl. Acad. Sci. U.S.A. 79:4298-4302.

Prakash, R.H., A.A.N. Van Brussel, A. Quint, A.M. Mennes,
and R.A. Schilperoort. 1982. The map position of syn-plasmid
regions expressed in the bacterial and endosymbiatic form of
Rhigabium leguminosarug. Plasmid 7:281-286.

Putzer, H., C. Verfuerth, M. Claviez, and T. Schreckenbach.
1984. Phatomorphogenesis in Physarum: Induction of tubulins

and sporulation—specific proteins and of their mRNAs. Proc.
Natl. Acad. Sci. U.S.A. 81:7117-7121.

Quinta, C., H. de la Vega, M. Flores, L. Fernandez, T.
Ballado, G. Soberan-Chavez, and R. Palacios. 1982.
Reiteration of nitrogen fixation gene sequences in Rhizobium
phaesoli. Nature (London) 299:724-726.

Raa, R.N. and 8.6. Rogers. 1979. Plasmid pKC7: a vector
containing ten restriction endonuclease sites suitable for
cloning DNA segments. Gene 7:79-82.

Redmond, J.W., M. Batley, M.A. Djordjevic. R.W. Innes, P.L.
Kuempel and D.G. Rolfe. 1986. Flavones induce expression of
nodulation genes in Rhizobium. Nature 323:632-635.

138

Regensburger, D., L. Meyer, M. Filser, J. Weber, D. Studer,
J.W. Lamb, H-M. Fischer, M. Hahn, and H. Hennecke. 1986.
Bradyrhizobium japanicum mutants defective in root-nodule
bacteroid development and nitrogen fixation. Arch.
Microbial. 144:355-366.

Renalier, M.R., J. Batut, J. Ghai, B. Terzaghi, M. Gherardi,
M. David, A-M. Garnerone, J. Vasse, G. Truchet, T. Huguet,
and P. Boistard. 1987. A new symbiotic cluster on the pSym
megaplasmid of Rhiggbium meliloti 2011 carries a functional

fix gene repeat and a nod locus. J. Bacteriol.
169:2231-2238.

Rolfe, D.G., P.M. Gresshoff, and J. Shine. 1980. Rapid
screening for symbiotic mutants of Rhizobium and white
clover. Plant Sci. Lett. 19:277-284.

Rolfe, D.G. and J. Shine. 1984. Rhizobium-Leguminosae
symbiosis: the bacterial point of view. tg Genes Involved in
Microbe-Plant Interactions. D.P.S. Verma and Th. Hahn eds.
(Springer-Verlag, New York) pp. 95-128.

Ronson, C.W., P.M. Astwood, and J.A. Downie.'1984. Molecular
cloning and genetic organization of Ca-dicarboxylate
transport genes from Rhtgobium leguminosarum. J. Bacteriol.
160:903-909.

Ronson, C.W., P.M. Astwood, B.T. Nixon, and F.M. Ausubel.
1987. Deduced products of Ca-dicarboxylate transport
regulatory genes of Rhigobium leguminosargg are homologous

to nitrogen regulatory gene products. Nucleic Acids Res.
15:7921-7934.

Ronson, C.W., P. Lyttleton, and J.G. Robertson. 1981.
Ca-dicarboxylate transport mutants of Rhigobium trifolii
form ineffective nodules on Trifolium repens. Proc. Natl.
Acad. Sci. U.S.A. 78:4284-4288.

 

Ranson,C.W., B.T. Nixon, L.M. Albright, and F.M. Ausubel.
1987. Rhizobium ggliloti gtrA (rpoN) gene is required for
diverse metabolic functions. J. Bacteriol. 169:2424-2431.

Ronson, C.W. and 8.8. Primrose. 1979. Carbohydrate
metabolism in Rhigobium trifolii: identification and

symbiotic properties of mutants. J. Gen. Microbiol.
112:77-88.

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

139

Rostas, K., P.R. Sista, J. Stanley, and D.P.S. Verma. 1984.
Transposon mutagenesis of Rhizobium japonicug. Mol. Gen.
Genet. 197:230-235.

Rostas, H., E. Kondorosi, B. Horvath, A. Simoncsits, and A.
Kondorosi. 1986. Conservation of extended promoter regions
of nodulation genes in Rhizobium. Proc. Natl. Acad. Sci.
U.S.A. 83:1757-1761.

Ruby, S.W. and J.W. Szostak. 1985. Specific Saccharogyces
cerevisiae genes are expressed in response to DNA-damaging
agents. Mol. Cell. Biol. 5:75-84.

Ruby, S.W., J.W. Szostak, and A.W. Murray. 1983. Cloning
regulated yeast genes from a pool of lacZ fusions. Methods
Enzymal. 101:253-268.

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

Ruvkun, G.B. and F.M. Ausubel. 1981. A general method for
site-directed mutagenesis in prokaryotes. Nature 289:85-88.

Sadowsky, M.J., K. Rostas, P.R. Sista, H. Bussey, and D.P.S.
Verma. 1986. Symbiotically defective histidine auxotrophs of
Bradyrhizobiug japonicgg. Arch. Microbiol. 1442334-339.

Sanger, F., S. Nicklen, and A.R. Coulson. 1977. DNA
sequencing with chain termination inhibitors. Proc. Natl.
Acad. Sci. U.S.A. 74:5463-5467.

Sargent, T.D. 1986. Isolation of differentially expressed
genes. Methods Enzymal. 152:423-432.

Sargent, T.D. and 1.8. Dawid. 1983. Differential gene

expression in the gastrula of Xenopus laevis. Science
222:135~l39.

Sarkar, N., D. Langley, and H. Paulus. 1978. Isolation and
characterization of polyadenylate-containing RNA from
Bacillus brevis. Biochemistry 17:3468-3474.

Schloss, J.A., C.D. Silflow, and J.L. Rosenbaum. 1984. mRNA
abundance changes during flagellar regeneration in
Chlggydomonas reinhardtii. Mol. Cell. Biol. 4:424-434.

Schafield, P.R., R.W. Ridge, D.G. Rolfe, J. Shine, and J.M.
Watson. 1984. Host-specific nodulation is encoded on a l4kb
DNA fragment in Rhizobium trifolii. Plant Mol. Biol. 3:3-11.

140

Sengupta-Gopalan, C. and J.W. Pitas. 1986. Expression of
nodule-specific glutamine synthetase genes during nodule
development in soybeans. Plant Mol. Biol. 7:189-199.

Sengupta-Gopalan, C., J.W. Pitas, D.V. Thompson, and L.M.
Hoffman. 1986. Expression of host genes during root nodule
development in soybean. Mol. Gen. Genet. 203:410-420.

Shimotsu, H., F. Kanamura, Y. Kobayashi and H. Saito. 1983.
Early sporulation gene spoF: nucleotide sequence and

analysis of gene product. Proc. Natl. Acad. Sci. U.S.A.
80:658-662.

Southern, E. 1975. Detection of specific sequences among DNA
fragments separated by gel electrophoresis. J. Mol. Biol.
98:503.

Spaink, R.P., R.J.H. Okker, C.A. Wijffelman, E. Pees, and
D.J.J. Lutenberg. 1987. Promoters in the nodulation region
of the Rhizobium leguminosarum Sym plasmid pRLlJI. Plant
Mol. Biol. 9:27-39.

Srinivasan, P.R., M. Ramanarayanan, and E. Rabbani. 1975.
Presence of polyriboadenylate sequences in pulse-labeled RNA
of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A.
72:2910-2914.

St. John, T.P. and R.W. Davis. 1979. Isolation of
galactose-inducible DNA sequences from Saccharomyces

cerevisiae by differential plaque filter hybridization. Cell
16:443-452.

 

Stachel, S.E., G. An, C. Flores, and E.W. Nester. 1985. A
TnE lacZ transposon for the random generation of
b-galactosidase gene fusions: application to the analysis of
gene expression in Eggabacterium. EMBO J. 4:891-898.

Szeto, W.W., B.T. Nixon, C.W. Ronson, and F.M. Ausubel.
1987. Identification and characterization of the Rhizobium
meliloti ntrC gene: E. meliloti has separate regulatory
pathways for activating nitrogen fixation genes in

free-living and symbiotic cells. J. Bacteriol.
169:1423-1432.

Szeto, W.W., J.L. Zimmerman, V. Sundarsen, and F.M. Ausubel.
1984. A Rhizobium meliloti regulatory gene. Cell
36:1035-1043.

Taylor, J.M., R. Illmensee, and J. Summers. 1976. Efficient
transcription of RNA into DNA by avian sarcoma virus
polymerase. Biochim. Biophys. Acta 442:324-330.

141

Thummler, F. and D.P.S. Verma. 1987. Nodulin-100 of soybean
is the subunit of sucrose synthase regulated by the

availability of free heme in nodules. J. Biol. Chem.
262:14730-14736.

Timberlake, W.E. 1980. Developmental gene regulation in
Aspergtllug nidulans. Devel. Biol. 78:497-510.

Tjepkema J., and H.J. Evans. 1975. Nitrogen fixation by
free-living Rhizobium in a defined liquid medium. Biochim.
Biophys. Res. Comm. 65:625-628.

Triplett, E. 1985. Intercellular nodule localization and
nodule specificity of xanthine dehydrogenase in soybean.
Plant Physiol. 77:1004-1009.

Turgeon, D.G., and W.D. Bauer. 1982. Early events in the
infection of soybean by Rhizobigg japanicgg. Time course and
cytology of the initial infection process. Can. J. Bot.
60:152-161.

Turgeon, D.G. and W.D. Bauer. 1985. Ultrastructure of
infection-thread development during the infection of soybean
by Rhizobium japonicum. Planta 163:328-349.

van Brussel, A.A.N., K. Planque, and A. Quispel. 1977. The
wall of Rhizobium leguminasarum in bacteroid and free-living
forms. J. Gen. Microbiol. 101:51-56.

van den Bos, R.C., T. Bisseling, and A. van Kammen. 1978.

Analysis of DNA content, nitrogenase activity and tg viva

protein synthesis of Rhizobium leguminasarum bacteroids on
sucrose gradients. J. Gen. Microbiol. 109:131-139.

 

VanEtten, B.D., and H.C. Kistler. 1984. Microbial enzyme
regulation and its importance for pathogenicity. tg
Plant-Microbe Interactions Molecular and Genetic
Perspectives. Vol. I., T. Kosuge and E.W. Nester eds.,
(Macmillan Publ. Co., New York) pp. 42-68.

Verma, D.P.S., M.G. Fortin, J. Stanley, V.P. Mauro, S.
Purohit, and N. Morrison. 1986. Nodulins and nodulin genes
of Glycine max. Plant Mol. Biol. 7:51-61.

Vieira. J., and J. Messing. 1982. The pUC plasmids, an
Ml3mp7-derived system for insertion mutagenesis and

sequencing with synthetic universal primers. Gene
19:259-268.

Vincent, J.M. 1970. A Manual for the Practical Study of Root
Nodule Bacteria. (Blackwell, Oxford). p. 86.

142

Vincent, J.M. 1980. Factors controlling the legume-Rhizobium
symbiosis. tg Nitrogen Fixation Volume II: Symbiotic
Associations and Cyanobacteria. W.E. Newton and
W.H.0rme-Johnson eds. (University Park Press, Baltimore) pp.
103-130.

 

Wells, S.E. and D.L. Kuykendall. 1983. Tryptophan auxotrophs
of Rhtgpbium japonicum. J. Bacteriol. 156:1356-1358.

Wilson, K.J., V. Anjaiah, P.T.C. Nambiar, and F.M. Ausubel.
1987. Isolation and characterization of symbiotic mutants of
Bradyrhizobium sp. (Arachis) strain N092: mutants with
host-specific defects in nodulation and nitrogen fixation.
J. Bacteriol. 169:2177-2186.

Walk, C.P., A. Vonshak, P. Kehoe, and J. Elhai. 1984.
Construction of shuttle vectors capable of conjugative
transfer from Escherichia coli to nitrogen-fixing

filamentous cyanobacteria. Proc. Natl. Acad. Sci. U.S.A.

Wong, P.P. and H.J. Evans. 1971. Poly-s-hydroxybutyrate
utilization by soybean (Glycine ggg Merr.) nodules and
assessment of its role in maintenance of nitrogenase
activity. Plant Physiol. 47:750-755.

Yannish-Perron, C., J. Vieira, and J. Messing. 1985.
Improved M13 phage cloning vectors and host strains:

nucleotide sequences of the Ml3mp18 and pUC19 vectors. Gene
33:103-119.

Youngman, P, P. Zuber, J.D. Perkins, K. Sandman, M. Igo and
R. Losick. 1985. New ways to study developmental genes in
spore-forming bacteria. Science 228:285-291.

Zimmerman, C.R., W.C. Orr, R.F. Leclerc, E.C. Bernard and
W.E. Timberlake. 1980. Molecular cloning and selection of
genes regulated in Aspergillus development. Cell 21:709-715.

“TTTTTTTTTTTTTT