:i@%.
yawn“.

,3 .

fl

. .
“.fimmflhv
w n. vawfi
. «. ~lm We.
$3,qu a; 1.2
‘ -u. in? 1. .
.vafimfififli .

$. 1‘!
vw.%\zh~nd
_ _
.N

tits
.l.§. ~ Q...
23

c".

at"

.tW" '
H‘\1 In fi.‘t|.‘rt8.'$."l“
‘ _ I‘mu-i‘u‘f
..

 

llllllllfllliljllilllllll

This is to certify that the

dissertation entitled

CARBON, NITROGEN, AND OXYGEN LIMITATION-INDUCED LOCI OF
RHIZOBIUM MELILOTI ISOLATED BY TN5-LUXAB MUTAGENESIS AND
THEIR ROLE IN COMPETITION AND SURVIVAL

presented by

Daniel Martin Ragatz

has been accepted towards fulfillment
of the requirements for

Ph.D. Genetics

degree in

 

 

 

 

Date 216/3}

M5 U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY
Michigan State
UnIversIty

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINE return on or before date due.
MAY BE RECAILED with earIier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

we W44

CARBON, NITROGEN, AND OXYGEN LIMITATION-INDUCED LOCI OF RHIZOBIUM
MELILOTI ISOLATED BY TNS-LUXAB MUTAGENESIS AND THEIR ROLE IN
COMPETITION AND SURVIVAL

By

Daniel Martin Ragatz

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

DOCTOR OF PHILOSOPHY
Graduate Program in Genetics

1997

ABSTRACT
CARBON, NITROGEN, AND OXYGEN LIMITATION-INDUCED LOCI OF RHIZOBIUM

MELILOTI ISOLATED BY TN5-LUXAB MUTAGENESIS AND THEIR ROLE IN
COMPETITION AND SURVIVAL

By

Daniel Martin Ragatz

Soil bacteria, such as Rhizobium meliloti, live in environment where
nutrients are scarce, and slow growth or no grth of bacteria over time is the rule,
rather than the exception. Competition for available resources is very high, and
changing environmental conditions, such as exposure to microaerobiosis (or
anoxia) due to oxygen gradients within the soil pore matrix, requires the bacteria
to sense and respond in a timely manner in order to survive. In addition, the
processes of microbial competition and colonization of plant roots are of great
importance when contemplating the use of beneficial microbes in agriculture or for
bioremediation purposes. In order to examine this problem, a Tn5-1uxAB
transposon was used to generate insertions in R meliloti, which could then be
examined for induction by environmental conditions.

A large collection of genes (69) was isolated in response to three conditions:
nitrogen and carbon limitation, and microaerobiosis at 1% 02 concentration. Many
of the tagged loci were similar to known genes, and some of these genes were even
known to be regulated by the stress of interest. However, most of the tagged loci
were novel, and thus seem to be specific in R. meliloti to the starvation conditions
(or to microaerobiosis). This supports the hypothesis that soil microorganisms, such
as R. meliloti, regulate gene expression in respon se to soil and rhizosphere

conditions differently than non-soil bacteria such as Escherichia coli.

The importance of these tagged loci in competition and persistence
experiments was tested and several R. meliloti strains harboring nutrient regulated
fusions were significantly reduced in numbers when coinoculated with the wild-
type 1021 in nutritionally poor soil. However, competition for modulation
experiments with these same strains yielded the surprising finding that none of the
Tn5-1063 tagged loci were necessary for nodule competitiveness versus the wild-

type strain 1021.

It’s harder to believe than not to.

- Steve Taylor

iv

ACKNOWLEDGMENTS

This thesis would not have been possible without the support and
patience of my supervisor, Frans de Bruijn. I would like to thank him and the
rest of my committee members, Peter Wolk, Mike Thomashow, and Eldor Paul,
for their help and patience in completing this project.

I am also indebted to the members of the de Bruijn lab who have
helped me at various stages of my thesis research, as well as by acting as a
moral support for me. In particular, I am grateful to Anne Milcamps for all of
her suggestions and criticisms which helped me to be a more critical thinker.
Also, to Brian McSpadden Gardener for many hours of statistical examination
of my data, as well as to discuss future experiments.

I would like to thank all those in the Center for Microbial Ecology who
have given me suggestions and use of their labs.

I also want to acknowledge my parents, Dan and Nancy, for always
believing in me and supporting my decisions.

Finally, I cannot express in words the thanks I feel to my wife, Matinga,
for all of her hours of patience and understanding which helped me to get
through the most challenging period in my life.

This research was supported by NSF STC grant DEB#9120006.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ ix
LIST OF FIGURES ....................................................................................................... x
CHAPTER 1: Introduction ....................................................................................... 1
Introduction ................................................................................................... 2
Bacteria in the soil ......................................................................................... 2
Bacteria in the rhizosphere .......................................................................... 5
Bacterial response to environmental stresses ............................................ 6
Scope of this thesis ..................................................................................... 14
References .................................................................................................. 1 7

CHAPTER 2: Isolation and Characterization of Rhizobium meliloti
Genes Whose Promoters are Induced by Nitrogen

Deprivation ..................................................................................... 2 4
Abstract ....................................................................................................... 2 5
Introduction ................................................................................................ 2 6
Results ......................................................................................................... 2 9
Discussion ................................................................................................... 44
Materials and Methods .............................................................................. 48
References .................................................................................................. S 3

CHAPTER 3: Isolation and Characterization of Rhizobium meliloti
Genes Whose Promoters are Induced by Carbon
Deprivation ..................................................................................... 59

Abstract ....................................................................................................... 60

Introduction ................................................................................................ 6 1
Results ......................................................................................................... 6 3
Discussion ................................................................................................... 7 2
Materials and Methods .............................................................................. 7 5
References .................................................................................................. 7 8

CHAPTER 4: Isolation and Characterization of Rhizobium meliloti
Genes Whose Promoters are Induced by Oxygen

Limitation ........................................................................................ 8 2
Abstract ....................................................................................................... 8 3
Introduction ................................................................................................ 84
Results ......................................................................................................... 90
Discussion ................................................................................................... 9 5
Materials and Methods ............................................................................ 100
References ................................................................................................ 1 O 1

CHAPTER 5: Rhizobium meliloti Genes Induced by Nutrient
Deprivation are Important for Competition in

Nutrient Poor Soil ......................................................................... 106
Abstract ..................................................................................................... 1 07
Introduction .............................................................................................. 1 08
Results ....................................................................................................... 1 1 3
Discussion ................................................................................................. 1 1 8
Materials and Methods ............................................................................ 1 2 1
References ................................................................................................ 1 2 3

CHAPTER 6: Rhizobium meliloti Genes Induced by Nutrient
Deprivation and Affected in Soil Competition are
Not Involved in Competition for Nodulation ............................. 1 2 6

Abstract ..................................................................................................... 1 2 7

Introduction .............................................................................................. 1 2 8
Results ....................................................................................................... 1 3 1
Discussion ................................................................................................. 1 34
Materials and Methods ............................................................................ 13 7
References ................................................................................................ 139
CHAPTER 7: Conclusions and Future Directions ............................................ 141
Conclusions and Future Directions ........................................................ 142
Overlap of loci induced by multiple stresses ................................. 143

Importance of Tn5—1063 tagged genes regulated by
nutrient deprivation for persistence and competition

in the soil and in nodule occupancy ............................................... 147
Search for regulators responsive to stress conditions .................. 148
References ................................................................................................ 1 SO

viii

LIST OF TABLE

TABLE 2-1. Phenotypes of selected auxotrophic Tn5-1063 insertion
mutants of R. meliloti 1021 ............................................................. 3 1

TABLE 2-2. Phenotypes of selected auxotrophic Tn5-1063 insertion
mutants of R. meliloti 1021 induced by nitrogen deprivation 36

TABLE 3-1. Phenotypes of selected auxotrophic Tn5-1063 insertion
mutants of R meliloti 1021 induced by carbon deprivation ....... 64

TABLE 4-1. Phenotypes of selected auxotrophic Tn5-1063 insertion
mutants of R. meliloti 102 1 induced by microaerobiosis ............. 91

FIGURE 1—1.

FIGURE 2-1.

FIGURE 2-2.

FIGURE 2-3.

FIGURE 2-4.

FIGURE 2-5.

FIGURE 3-1.

FIGURE 3-2.

FIGURE 4- 1.

FIGURE 4- 2.

FIGURE 4-3.

FIGURE 5-1.

FIGURE 5-2.

LIST OF FIGURES

Section through a soil matrix ........................................................... 4

Tn5-1063 structure, mutagenesis protocol, and target
junction DNA sequencing .............................................................. 30

Luciferase activity dermination of fusions induced
by nitrogen deprivation ................................................................ 3 3

Temporal lux expression pattern of selected
nitrogen deprivation-induced gene fusions ................................ 3 7

Codon-usage analysis of Tn5-1063 tagged target
gene DNA sequences ..................................................................... 40

Similarity of the amino-acid sequence deduced from
selected Tn5-1063 tagged ORFs induced by
nitrogen deprivation ...................................................................... 4 1

Similarity of the amino—acid sequence deduced from
selected Tn5-1063 tagged ORFs induced by
carbon deprivation ........................................................................ 66

Chemotactic determination of strain C27 in a diffusion

gradient chamber (DGC) ............................................................... 70
Diagram of an indeterminate nodule of alfalfa .......................... 86
Comparative models of nif and fix gene regulation ................... 87
Map of the exo gene cluster .......................................................... 98

Persistence and competition of selected Tn5-1063
induced mutant strains in soil .................................................... 1 14

Persistence and competition of other selected Tn5- 1063
induced mutant strains in soil .................................................... 1 1 6

FIGURE 6-1. Competition for nodulation of selected Tn5- 1063
induced mutant strains versus the wild-type 1021 ................. 1 32

FIGURE 7-1. Ven diagrams illustrating the overlap of loci induced by multiple
stresses .......................................................................................... 144

xi

CHAPTER 1

INTRODUCTION

Parts of this chapter have been submitted for publication in Molecular Microbiology
(Ragatz et 21]., 1 997).

INTRODUCTION

Over the last two decades, interest in microbial ecology has increased
dramatically as new molecular tools have become available for detecting and
tracking microbial populations in natural environments, and as interest has grown
in the use of genetically engineered microorganisms (GEMS) for a variety of
purposes, such as biodegradation. Of particular interest has been an
understanding of how bacteria sense and respond to changing environmental
conditions. In the first part of this chapter I will give an overview of one such
natural environment, the bacterial soil system, and in the second part I will discuss
two stresses to which microorganisms are subjected in such an environment. This
chapter will not be an attempt to give a comprehensive review, but to highlight

findings that are definitive and to provide an overview of soil microbial ecology.

NOTE: The word ‘starvation’ has been widely used in the literature, but as this
implies an understanding of the physiological response of the bacterium, I have

chosen to use the word ‘deprivation’ in this thesis.

BACTERIA IN THE SOIL

Soil bacteria in the environment are frequently exposed to nutrient
deprivation conditions. In fact the majority of non-rhizosphere soil is so oligotrophic
that it has been called a “nutritional desert” (Metting, 1985). Consequently,

bacterial growth is extremely limited and non-growth, or very low growth, may be

3

considered the norm (Kjelleberg et 21., 1987). The mean generation times of
bacteria in the soil have been estimated to be from less than 1 to 80 per year
(Shields et 211., 1973; Gray, 1976; Lynch, 1988; Matin, 1991), for bacteria whose
generation times may be hours or less in a laboratory setting.

Not only is the soil environment largely oligotrophic, it is an extremely
complex mixture of variously sized organic and mineral particles, living organisms,
and their remains. It has been described as the most complex microbiological
habitat (Stotzky, 1972; Metting, 1993). Bacteria inhabit soil pores of varying sizes
and compositions, and the analysis of the effects of soil factors should actually be
judged at the level of the soil pore rather than soil as a whole (Smiles, 1988). Figure
1-1 illustrates microbial microcommunities in the soil pores of a complex soil
matrix. Microbial microcommunities in soil pores are dynamic due to the rapid flux
of water, solutes, and other environmental cues, as well as the particular make-up
of the microbial communities. Environmental factors will vary greatly both in time
and across short distances within microhabitats (Metting, 1993).

Bacteria are not randomly scattered among or in soil pores, but are
generally adsorbed to soil particles (Hattori and Hattori, 1976), often preferentially
to organic or organic/clay complexes (Hisset and Gray, 1976). Paul and Clark
(1989) showed that bacteria occupy less than 1% of the soil pore space; therefore,
bacteria are more numerous within microhabitats such as soil pores than total soil
counts would suggest. Competition for available resources, as well as the ability to
sense and respond to environmental changes, are therefore very important to the
survival of a given species of microorganism.

The types of bacteria present in bulk soil vary according to the conditions of

the soil and environment, but are largely gram negative rods (Alexander, 1977;

Microcolony

\ \\§\\
K ‘3
Quartz \¢§‘\i ‘ \
Sis

s¢
\A\\\\‘D‘&\ . \' \ ‘
r? ‘ \ -'-'
":2 _ - . .4‘ .
_ * -. ’ w.-

      

  
    

  
     
   

/

  

Organic

. .......

Figure 1-1. Section through a soil matrix showing
microhabitats and patchy distribution of bacterial
microcolonies. Air and water pores are also indicated
(Brock, 1979).

Atlas and Bartha, 1989). Bacillus Sp. (a gram-positive species) are sometimes
present in large numbers (Alexander, 1977), however, their population size is
misleading since viable counts do not distinguish between colonies developed from
a spore or a vegetative unit. Frequently 60 to 100 percent of Badllus cells in the
soil exist as spores (Mishustin and Mirsoeva, 1967; Siala et al., 1974).
Understanding how bacteria survive and persist under oligotrophic
conditions is important both in terms of increasing our fundamental knowledge of
microbial ecology, as well as in applied contexts, such as the release of genetically
engineered microorganisms (GEMS) into the environment. In addition, isolation of

environmentally regulated promoters will aid in designing GEMS to be used in

5
specific environmental situations, such as in bioremediation experiments, where

nutrient regulated promoters can act as switches to provide a greater level of

regulation of gene expression.

BACTERIA IN THE RHIZOSPHERE

In contrast to bacterial numbers in bulk soil, bacterial numbers in the
rhizosphere—an area encompassing the surface of the root (the rhizoplane) to a
few millimeters away from it—are much higher. This rhizosphere effect is often
expressed as a ratio of the microbial population in the rhizosphere to that in the
bulk soil (the R/S ratio). Both bacterial numbers and diversity are increased in the
rhizosphere (Foster, 1986; Foster and Rovira, 1978) with R/S ratios in the tens to
hundreds (Bolton et al., 1993). This effect is due to an increase in soluble carbon
and nutrients found close to the plant roots.

These compounds encompass a wide variety of substances from low-
molecular weight compounds such as sugars and amino-acids, to mucilage and
lysates, and become available through root leaching, as well as active secretion by
the plant root (Bolton et al., 1993). In fact, the presence of microorganisms in the
rhizosphere has been Shown to increase active root exudation (Barber and Martin,
1976). Many other factors affect root exudation including temperature, irradiance,
soil moisture content, plant age and nutrient status, and stresses (Rovira, 1959;
Vancura, 1988).

Although more nutrients are available, the higher number of
microorganisms in the rhizosphere makes competition for these nutrients even

more important. This is exacerbated by the fact that bacteria are not randomly

6
distributed on the root surface, but limited to microcolonies covering only 4-10% of

the root surface (Rovira et al, 1974; Rovira, 1979). The position of these
microcolonies correlates with the presence of soil organic matter on the root surface
and often coincides with epidermal cell junctions (Bowen and Rovira, 1976; van
Vuurde and Schippers, 1980). In addition, the higher concentration and diversity
of microorganisms can increase the level of deleterious compounds, such as
antibiotics, produced by competing microorganisms. Also, oxygen levels can be
severely depleted due to root and microbial respiration leading to conditions
favorable for anaerobic or microaerobic growth (Alexander, 1977). Therefore, the
growth of microorganisms even in the relatively nutrient-rich rhizosphere can still
be limited (van Elsas and van Overbeek, 1993).

As in bulk soil in general, rhizosphere microorganisms are predominantly
short, gram negative rods (Alexander, 1977). In fact, the percentage occurrence of
short, gram-positive rods, so called “coccoid” rods, and spore-forming bacteria
declines (Alexander, 1977). The reason for this observation is not clear; there does
not appear to be a selective stimulation or inhibition of gram-variable rods.
Pseudomonas, Flavobacterium, Alcaligenes, and Agrobacterium Species are
especially common (Alexander, 1977). In contrast to the rhizosphere effect on
bacteria, total counts of fungi in the rhizosphere are not significantly altered
compared to bulk soil, however the predominance of specific fungal genera may be

Stimulated (Alexander, 1977; Sivasithamparam et al., 1979).

7
BACTERIAL RESPONSE TO ENVIRONMENTAL STRESSES

Bacteria in the soil are subjected to a large number of stresses which affect
their physiological state. The intensities and interactions among these stresses vary
over time, as well as over very small distances from one microcommunity to the
next. Stress factors which are most important include nutrient deprivation, oxygen
availability, water limitation, temperature and pH extremes, and UV irradiation.
This chapter will only discuss the two stresses that are relevent to the research
described in later chapters, i.e., nutrient deprivation and oxygen availability. For
an in-depth consideration of the soil environment, the reader is referred to texts
by Alexander (1977), Lynch (1988), van Elsas and van Overbeek (1993), and

Metting (1993).

Nutrient deprivation

The manner in which bacteria deal with the problem of nutrient deprivation
varies between classes of microorganisms. Some bacteria, such as Bacillus Sp. and
Myxococcus Sp., can differentiate into stress resistant endospores and myxospores,
respectively, when confronted with nutrient deprivation (Losick et al., 1986; Kaiser,
1986). This process has been well studied and involves a complicated cascade of
gene expression, temporally regulated by an alternative Sigma factor (sigma B),
which in turn is post-translationally regulated by a multi-component network of
regulatory proteins (Hecker and volker, 1990; Antelmann et al., 1996; Akbar et al.,
1997). However, the majority of bulk soil bacteria are actinomycetes and gram

negative, non-sporulating rods (Rovira, 1956; Alexander, 1977; Lynch and Wood,

8
1988; Atlas and Bartha, 1989), most of which are not known to differentiate into

protective forms, and thus, must deal with nutrient deprivation by other means.
These bacteria include Agrobacterium, Azospirillum, Azotobacter, Erwinia,
Pseudomonas, Rhizobium, and Xanthomonas species (Miller and Wood, 1996).

Matin et al. (1989) have described two fundamental ways in which gram
negative bacteria confront nutrient deprivation, both of which involve the synthesis
of new proteins in response to the deprivation condition: (i) by altering the
bacterium to make it a more efficient scavenger of scarce nutrients, or (ii) by
making it more resistant to stresses in general. Thus, understanding which genes
are regulated before, during, and after the onset of deprivation conditions is
important for the elucidation of the mechanisms by which bacteria survive and
compete under oligotrophic conditions.

Contrary to earlier assumptions, so-called non-differentiating bacteria
undergo an elaborate process of molecular realignment when they are exposed to
deprivation conditions, which lead to the development of a resistant cellular state.
For example, when E. coli is starved it forms small, nearly spherical cells and
develops enhanced resistance to a variety of stresses including oxidation,
hyperosmosis, heat, low pH, and disinfectant agents (Matin, 1990). Similar states of
enhanced resistance have been observed in Salmonella typhimurium (Spector and
Cubit, 1992), Pseudomonas putida (Givskov et al., 1994), and Rhizobium
leguminosarum (Crockford et al., 1996). In Vibrio sp., starved cells form
ultramicrocells and develop resistance against growth inhibiting agents and
conditions (Holmquist et al., 1994). Interestingly, it has been found that carbon,
rather than nitrogen or phosphorus, deprivation is the determinant in forming this

resistant state (Holmquist and Kjelleberg, 1993; Nystrdm et al., 1992). This is in

9
contrast to E. coli where cells will form their protective state regardless of the

Specific nature of the deprivation condition (Matin, 1990).

Most of the knowledge regarding deprivation responses in non-
differentiating bacteria derives from work in E. coli, S. typhimurium, and Vibrio Sp.
In E. coli, over 70 polypeptides that are initiated or increased during carbon,
phosphorus, or nitrogen deprivation have been visualized using two-dimensional
polyacrylamide gel electrophoresis (2D-PAGE) (Matin, 1991). Many of these
polypeptides are uniquely induced by deprivation of a single nutrient; however,
several of them are induced by two deprivation conditions and a group of fifteen
proteins constitutes a core set which is induced under all deprivation conditions
examined. Similar results have been observed using 2D-PAGE protein analysis of S.
typhimurium (Spector and Cubitt, 1992), where Six polypeptides are induced by
three or more conditions, and Vibrio Sp. (Nystrom et al., 1990) where three
proteins are commonly induced during nutrient deprivation conditions.

Although much is known about bacterial gene expression patterns from
laboratory studies involving growth of bacterial cultures under ideal conditions in
terms of temperature, pH, and nutrients, comparatively little is understood about
their behavior under unfavorable conditions, such as those found in the soil
environment. In fact, few investigations on nutrient deprivation-induced gene
expression have been carried out on common soil isolates, which may have
different strategies for dealing with nutrient deprivation than enteric or marine
bacteria. Only recently, the effect of nutrient deprivation on P. putida, a common
gram negative soil bacterium, has been reported (Givskov et al., 1994). A large
number of proteins induced by carbon and nitrogen deprivation have been

detected using 2D-PAGE protein analysis. These proteins are temporally regulated,

10
as in the case of E. coli and Vibrio sp., and include a group of eight proteins

induced by multiple Stresses. Another common soil isolate, Anabaena sp., has been
studied under nitrogen deprivation conditions using a luxAB reporter (Cai and
Wolk, 1997). Several genes were isolated and some were Similar to genes required
for nitrite and nitrate uptake and utilization.

To date, only a few roles have been assigned to the large collection of
deprivation-induced proteins in any of the bacterial species studied (Givskov et al.,
1994). However, some progress has been made in identifying the core proteins
induced under multiple conditions. In E. coli, for instance, three of the 15 core
proteins have been identified: DnaK, GroEL, and Hth. These proteins had first
been identified in response to heat shock, but they are also increased during
deprivation (Matin, 1991). These proteins appear to have a role in protein folding
and macromolecular assembly (chaperone function).

Regulation of the genes encoding these proteins has not been examined in
most of the organisms studied, except for E. coli where the regulatory mechanisms
have been at least partially elucidated. Many of the core proteins appear to be
regulated by the heat-Shock alternative sigma factor on, but are not dependent on
cyclic AMP/CRP. In contrast, most of the carbon-deprivation induction specific
proteins are regulated by cAMP/CRP. The putative Sigma factor, KatF, has been
found to regulate most of the carbon deprivation-induced proteins, as well as Six of
the core proteins (McCann et al., 1991). The core proteins appear to have a distinct
role in the development of cellular resistance, since the deletion of rpoH (encoding
oi2 ) in E. coli results in an approximately 50% reduction in strain viability under
carbon deprivation conditions (Jenkins et al., 1991). Recently, two carbon

deprivation regulated genes in P. putida have been isolated using mini-Tn5

1 1
transposon mutagenesis (Kim et al., 1995). These genes were found to be

regulated by another Sigma factor, 6“. This sigma factor has been shown to be
required for the expression of many genes such as flagellin in Pseudomonas sp.
(Totten et al., 1990). However, this is the first time it has been implicated in the
regulation of deprivation-induced genes (Kim et al., 1995).

A few other deprivation enhanced (induced) proteins have been elucidated:
a protease (Reeve et al., 1984), the ribosomal modulation factor Rmf (Wada et al.,
1990; Yamagishi et al., 1993), and the product of the uspA gene, the universal
stress protein (Nystrém and Neidhardt, 1992, 1993). The latter protein is induced
by the cessation of cell growth, regardless of the condition inhibiting growth, and its
regulation is independent of all global regulators tested thus far including RpoS,
PhoB, RelA, SpoT, AppY, OmpR, H-ns, Lrp, and RpoH (Nystrom and Neidhardt,
1992). Deletions in UspA result in mutants which are highly susceptible to a
variety of stresses including osmotic shock and carbon deprivation (Nystrom and
Neidhardt, 1994). UspA is thought to modulate the flow of carbon in central
metabolic pathways, however its regulation and the subsequent pleiotrophic
phenotype diplayed in E. coli is not understood. There is no evidence to indicate
that UspA is a global regulator of gene expression; however, overproduction of
UspA seems to affect the pl of other proteins, which may indicate a possible role of

UspA in post-translational modification (Nystrom and Neidhardt, 1994).

Oxygen limitation

The availability of oxygen in natural environments, such as soil, is one of the

most important parameters affecting growth and competition of microorganisms.

1 2
Soils are frequently subjected to varying oxygen conditions due to rainfall and

drying. When soils are flooded a unique oxidation-reduction profile is created with
a thin, oxidized soil surface layer, and an underlying thicker, reduced layer
deficient in oxygen. Even within soil particles, steep oxygen gradients exist within a
few millimeters of the soil surface. Soils with a sufficient supply of decomposable
organic matter may be depleted of oxygen within one day, due to growth of
heterotrophic microorganisms that scavenge the available oxygen (Foth, 1984).
Thus, in order to survive and compete, soil microorganisms must be able to adapt
to changing oxygen conditions by, (i) altering their ability to bind oxygen as
terminal electron acceptor via alternative cytochromes, (ii) by using alternative
electron acceptors such as nitrate, or (iii) changing to anaerobic metabolism.

Genes regulated in response to low oxygen or anaerobiosis have been
described in many systems. Among the most well studied organisms are the
facultative anaerobes, such as E. coli, Bacillus sp., S. typhimurium, Pseudomonas
aeruginosa, Vibrio fischeri, and Rhodobacter sphaeroides. In E. coli, two global
regulatory systems have been described for sensing and responding directly to
molecular oxygen: a two-component sensor/ regulator system comprised of
Ach/ArcA (Iuchi and Lin, 1988, 1992, 1993), and a transcriptional sensor-
regulator protein, FNR (Lambden and Guest, 1976; Chippaux et al., 1978; Spiro
and Guest, 1990; Lin and luchi, 1991). Genes responding to these regulators are
part of a group of four global regulatory systems required during anaerobic
metabolism, and a significant overlap occurs among the regulatory systems. Many
other bacterial systems have been identified with FNR homologues (Fischer, 1994;

Spiro, 1994).

1 3
In addition to genes responding to the absence of oxygen, Specific

microaerobic (microoxic) respiratory chains, utilizing a set of genes which are
different than those expressed during anaerobic conditions, are induced during
microoxic conditions in E. coli (Lin and luchi, 1991; Gunsalus, 1992). These
microoxic conditions also induce gene expression by aerobic rhizobia. The well—
studied two-component FixL/Fix] system (Figure 1-2) senses and responds to low
oxygen levels and initiates a complex cascade of gene regulation events (Agron et
al., 1994; Fischer, 1994). The FixL/Fix] regulatory cascade is extremely complex,
and varies among the rhizobial genera (Figure 1-2). This system is responsible for
the microaerobically-regulated gene expression patterns described in rhizobia to
date, including the genes required for nitrogen-fixation (nif & fix; David et al.,
1988), as well as genes necessary for rhizopine synthesis (mos; Murphy and Saint,
1991), nodulation competitiveness (nfe; Sanjuan and Olivares, 1989, 1991; Soto et
al., 1993, 1994), and melanin production (melA and melC; Hawkins and Johnston,
1988; Hawkins et al., 1991). In Bradyrhizobium japonicum, it has also been Shown
to control genes required for anaerobic nitrate respiration (Anthamatten and
Hennecke, 1991; Fischer, 1994). Recently, however, Zhulin et a1. (1995)
demonstrated that FixL and Fix] were not involved in oxygen response and
regulation of R. meliloti behavior (chemokinesis and aerotaxis) in oxygen gradients,
indicating that at least one other oxygen sensing mechanism must exist in rhizobia.

Although much has been learned about the nature of the genes, promoters,
and transcriptional control of these regulons, the nature of the sensing systems
involved in responding to fluctuation in concentration of molecular oxygen is still
poorly understood. The mechanism of Oz-sensing appears to be different in

different systems. The FixL mediated pathway is a little better understood since the

1 4
FixL protein has been found to contain a heme moiety which is thought to interact

directly with oxygen, resulting in an alteration of the Fe oxidation state, which in
turn results in the autophosphorylation of FixL (Gilles-Gonzalez et al., 1991).
However, it is not known whether the regulation occurs intramolecularly through
conformational changes or intermolecularly by promoting or preventing
dimerization (Agron et al., 1994). FNR also contains a metal cofactor, Fe(II) (Spiro,
et al., 1989) which is required for FNR to act as a transcriptional activator (Green et
al., 1991, 1993). It is thought that FNR senses oxygen by participating in a redox
reaction that results in the conversion of FNR to a transcriptionally inactive State
(Unden et al., 1994). Unlike FixL and FNR, Ach lacks Oz-reactive sites such as
hemes, iron-sulfur clusters or metal ions which could interact directly with oxygen
(Iuchi et al., 1990). Spiro and Guest (1991) have suggested that the source of the
signal to Ach is in the electron transport chain and that Ach detects changes in
the ratio of the oxidized to reduced forms of an electron carrier such as a flavin,
heme, quinone, or iron-containing component. However, there are no indications
that redox mediators, such as quinones or redox enzymes, interact with Ach

(Unden et al., 1994).

SCOPE OF THIS THESIS

In this thesis, I am interested in answering several questions related to the

background presented in this chapter.
1) How many loci are induced in R. meliloti during deprivation of
nutrients and oxygen, and does this compare to genes induced in non-

soil genera?

1 5
2) What is the nature of these genes and are they different from those

found in non-soil genera?

3) Are genes induced by nutrient deprivation and oxygen limitation
important for survival and competition in the soil and for competition
for nodulation?

Except for recent analysis of Pseudomonas putida (Kragelund et al., 1995)
and Anabaena Sp. (Cai and Wolk, 1997), little work has been done with common
soil isolates using molecularly tagged genes, making it difficult to hypothesize how
this method would compare to 2D—PAGE analysis. Nevertheless, based on the P.
putida data, I propose that the number of genes isolated will be Similar to those
found in non-soil genera. The nature of these genes, however, I propose to be
different for soil isolates versus non-soil isolates. Further, I propose that these
genes will be very important for competition in the soil, as well as for competition
for nodulation.

Since few soil isolates have been Studied to date, I have chosen to use R.
meliloti for the research described in this thesis. R. meliloti is a common soil isolate
with many advantages for studying environmentally regulated gene expression.
First, It is a common gram negative soil bacterium. Second, it has been extensively
analyzed using molecular genetic techniques, and both physical and genetic maps
are available (Honeycutt et al., 1993). Third, it is amenable to transposon Tn5
mutagenesis. Finally, it has two distinct niches, free-living in the soil and in a highly
Specific symbiotic interaction with its plant-host (alfalfa), which allows a
comparative study of the role of nutrient deprivation-induced genes in bulk soil, in
the rhizosphere, in the infection process, and within the infected cells (nodules) of

the plants.

16
Chapter 2 of this thesis describes the isolation and characterization of 22

genes tagged by Tn5-lux mutagenesis which are induced under nitrogen
deprivation. Some of the cloned and sequenced genes are identical or similar to
genes in bacteria known to be regulated by nitrogen deprivation. However, most of
the R. meliloti genes induced by nitrogen deprivation appear to be novel. Chapter
3 describes the isolation and characterization of 12 genes tagged by Tn5-lux
mutagenesis which are induced under carbon deprivation. One of the cloned and
sequenced genes is Similar to a gene in bacteria known to be regulated by carbon
deprivation. Most of the R. meliloti genes induced by carbon deprivation also
appear to be novel. Chapter 4 describes the isolation and characterization of 34
genes tagged by Tn5-qu mutagenesis which are induced under oxygen limitation.
One of the cloned and sequenced genes is identical to a gene in rhizobia known to
be microaerobically regulated. Most of the R. meliloti genes induced by
microaerobiosis again appear to be novel.

Chapters 5 and 6 describe experiments designed to examine the
importance of tagged loci for the survival and competition in soil and nodulation,
respectively, of selected strains. The results of chapter 5 support the hypothesis
that these genes are important in competition; however, the results of chapter 6 do
not support the hypothesis that these genes are important for competition for
nodule occupancy.

Chapter 7 discusses the overall collection of stress induced loci in R. meliloti
from this Study compared to Stress induced loci from other bacteria (particularly E.
coli), as well as other conclusions drawn from this study. In addition, directions for

further research are discussed.

17
REFERENCES

Agron, P.G., Manson, E.K., Ditta, GS, and Helinski, DR (1994) Oxygen regulation

of expression of nitrogen fixation genes in Rhizobium meliloti. Res Microbial
145: 454—459.

Akbar, S., Kang, C.M., Gaidenko, TA, and Price, CW. (1997) Modulator protein
RSbR regulates environmental Signalling the the general stress pathway of
Bacillus subtilis. Mal Mia'abial 24: 567—578.

Alexander, M. (1977) In Introduction to Sail Microbiology, 2nd ed. New York: John
Wiley & Sons, pp. 16—51.

Anthamatten, D., and Hennecke, H. (1991) The regulatory status of the fixL—like
and fixl-like genes in Bradyrhizabium japanicum may be different from that in
Rhizobium meliloti. M01 Gen Genet 225: 38—48.

Antelmann, H., Engelmann, S., Schmid, R., and Hecker, M. (1996) General and
oxidative stress responses in Bacillus subtilis: cloning, expression, and mutation
of the alkyl hydroperoxide reductase operon. J Bacterial 178: 6571—6578.

Atlas, RM, and Bartha, R. (1989) Microbial Ecology: Fundamentals and
Applications, 3rd ed. Reading, MA: Benjamin/Cummings Publishing Company,
INC, pp. 271-285.

Barber, D.A., and Martin, J.K. (1976) The release of organic substances by cereal
roots into soil. New Phytol 76: 69—80.

Bolten Jr., H., Fredrickson, J.K., and Elliott, LE. (1993) Microbial ecology of the
rhizosphere. In Sail Microbial Ecology: Applications in Agricultural and

Environmental Management. Metting, F.B. (ed). New York: M. Dekker, pp. 27—
63.

Cai, Y. and Walk, GP. (1997) Nitrogen deprivation of Anabaena Sp. strain PCC7120
elicits rapid activation of a gene cluster that is essential for uptake and utiliztion
of nitrate. J Bacterial 179: 258—266.

Chippaux, M., Giudici, D., Aboujaoude, A., Casse, F. and Pascal, M. (1978) A
mutation leading to the total lack of nitrite reductase activity in Escherichia coli
K-12. Mol Gen Genet 182: 477—479.

Crockford, A.J., Behncke, C., and Williams, H.D. (1996) The adaptation of
Rhizobium leguminosarum bv. phaseali to oxidative stress and its overlap with
other environmental stress responses. Microbiology 142: 331-336.

David, M., Daveran, M.L., Batut, J., Dedieu, A., Domergue, 0., Ghai, J., Hertig, C.,
Boistard, P. and Kahn, D. (1988) Cascade regulation of nif gene expression in
Rhizobium meliloti. Cell 54: 671—683.

18

de Bruijn, F.J., Rossbach, S., Schneider, M., Ratet, P., Messmer, S., Szeto, W.W.,
Ausubel, F.M., and Schell, J. (1989) Rhizobium meliloti 1021 has three
differentially regulated loci involved in glutamine biosynthesis, none of which is
essential for symbiotic nitrogen fixation. J Bacteriol 171: 1673-1682.

Espin, G., Moreno, S., Wild, M., Meza, R., and Iaccarino, M. (1990) A previously
unrecognized glutamine synthetase expressed in Klebsiella pneumoniae from
the glnT locus of Rhizobium leguminasarum. Mol Gen Genet 223: 513—516.

Fischer, H.-M. (1994) Genetic regulation of nitrogen fixation in rhizobia. Microbial
Rev 58: 352—386.

Foth, H.D. (1984) Fundamentals of Soil Science, 7th Ed. New York: John Wiley &
Sons, pp. 219-221.

Foster, RC. (1986) The ultrastructure of the rhizoplane and rhizosphere. Annu Rev
Phytapath0124: 211—234.

Foster, RC, and Rovira, AD. (1978) The ultrastructure of the rhizosphere of
Trifalium subterraneum L. In Microbial Ecology. Loutit, M.W., and MilerS, J.A.R.
(eds). Berlin: Springer-Verlag, pp. 282—290.

Givskov, M., Eberl, L., and Molin, S. (1994) Responses to nutrient starvation in
Pseudamanas putida KT2442: two-dimensional electrophoretic analysis of
Starvation- and stress-induced proteins. J Bacterial 176: 4816—482 4.

Gilles-Gonzalez, M.A., Ditta, GS, and Helinski, DR. (1991) A haemoprotein with
kinase activity encoded by the oxygen sensor of Rhizobium meliloti. Nature 350:
1 7 0— 1 7 2.

Gray, T.R.G. (1976) Survival of vegetative microbes in soil. Symp Gen Microbiol 26:
327—364.

Green, J., Trageser, M., Six, S., Unden, G., and Guest, J.R. (1991) Characterization
of the FNR protein of Escherichia coli, and iron-binding transcriptional
regulator. Prac R Soc Land 244: 1 37—144.

Green, J., Sharrocks, A.D., Green, 3., Geisow, M., and Guest, J.R. (1993) Properties
of FNR proteins substitued at each of the five cysteine residues. Mol Microbiol 8:
61-68.

Gunsalus, RP. (1992) Control of electron flow in Escherichia coli: coordinated
transcription of respiratory pathway genes. J Bacteriol 174: 7069—7074.

Hattori, T., and Hattori, R. (1976) The physical environment in soil microbiology: an
attempt to extend principles of microbiology to soil microorganisms. CRC Crit
Rev Microbial 4: 42 3—461.

19

Hawkins, F.K.L., and Johnston, A.W.B. (1988) Transcription of a Rhiabium
leguminasarum biovar phaseali gene needed for melanin synthesis is activated
by nifA of Rhiabium and Klebsiella pneumoniae. Mal Microbial 2: 331-337.

Hawkins, F.K.L., Kennedy, C., and Johnston, A.W.B. (1991) A Rhizobium
leguminasarum gene required for symbiotic nitrogen fixation, melanin
synthesis and normal growth on certain growth media. J Gen Microbiol 137:
1721—1728.

Hecker, M., and valker, U. (1990) General stress proteins in Bacillus subtilis. FEMS
Microbiol Ecol 74: 197-213.

Hisset, R., and Gray, T.R.G. (1976) Microsites and time changes in soil microbe
ecology. In The Role of Terrestrial and Aquatic Organisms in Decomposition
Processes. Anderson, J.M., and MacFadyen, A. (eds). London: Oxford University
Press, pp. 23-39.

Holmquist, L., and Kjelleberg, S. (1993) Changes in viability, respiratory activity
and morphology of the marine Vibrio Sp. strain S14 during starvation of
individual nutrients and subsequent recovery. FEMS Microbiol Ecol 12: 2 1 S-
224.

Holmquist, L., Nelson, DR, and Kjelleberg, S. (1994) The DnaK homologue of the
marine Vibrio Sp. strain S14 binds to the unprocessed form of a carbon
starvation-specific periplasmic protein. Mol Microbiol 11: 861—868.

luchi, S., and Lin, E.C.C. (1988) Arm (dye), a global regulatory gene in Escherichia
coli mediating repression of enzymes in aerobic pathways. Proc Natl Acad Sci
USA 85: 1888-1892.

luchi, S., and Lin, E.C.C. (1992) Purification and phosphorylation of the Arc
regulatory compnents of Escherichia coli. J Bacterial 174: 5617-5623.

luchi, S., and Lin, E.C.C. (1993) Adaptation of Escherichia coli to redox
environments by gene expression. Mal Microbial 9: 9-1 5.

luchi, S., Matsuda, Z., Fujiwara, T., and Lin, E.C.C. (1990) The ach gene of
Escherichia coli encodes a sensor-regulator protein for anaerobic repression of
the arc modulon. Mal Microbial 4: 715—727.

Jenkins, D., Auger, E., and Matin, A. (1991) Role of RpoH, a heat shock regulator
protein, in carbon starvation protein synthesis and survival of Escherichia coli. J
Bacteriol 173: 1992-1996.

Kaiser, D. (1986) Control of multicellular development: Dictyostelium and
Myxococcus. Ann Rev Genet 20: 539-566.

Kim, Y., Watrud, LS, and Matin, A. (1995) A carbon starvation survival gene of
Pseudamanas putida is regulated by 05“. J Bacteriol 177: 1850-1859.

20

Kjelleberg, S., Hermansson, M., Mardén, P., and Jones, G.W. (1987) The transient
phase between growth and nongrowth of heterotrophic bacteria, with emphasis
on the marine environment. Ann Rev Microbial 41: 25-49.

Lin, E.C.C., and luchi, S. (1991) Regulation of gene expression in fermentative and
respiratory systems in Escherichia coli and related bacteria. Annu Rev Genet
25:361-387.

Losick, R., Youngman, P., and Piggot, P.J. (1986) Genetics of endospore formation in
Bacillus subtilis. Ann Rev Genet 20: 62 5-669.

Lynch, J .M. (1988) The terrestrial environment. In Microorganisms In Action:
Concepts and Applications In Microbial Ecology. Lynch, J.M. and Hobbes, J .E.
(eds). Oxford: Blackwell Scientific, pp. 224-2 48.

Lynch, J.M., and Wood, M. (1988) Interactions between plant roots and
microorganisms. In Russell’s Soil Conditions and Plant Growth. Wild, A. (ed).
New York: Wiley, 526-563.

Matin, A. (1990) Molecular analysis of starvation stress in Escherichia coli. FEMS
Microbiol Ecol 74: 185-196.

Matin, A. (1991) The molecular basis of carbon-starvation-induced general
resistance in Escherichia coli. Mal Microbial 5: 3-10.

Matin, A., Auger, E.A., Blum, RH, and Schultz, J.E. (1989) Genetic basis of
starvation survival in nondifferentiating bacteria. Ann Rev Microbiol 43: 293-
316.

McCann, M., Kidwell, J., and Matin, A. (1991) The putative a factor KatF has a
central role in Pex protein synthesis and development of starvation-mediated
general resistance in Escherichia coli. J Bacteriol 173: 4188-4194.

Metting, B. (1985) Soil microbiology and biotechnology. In Biotechnology:
Applications and Research. Ovellette, RP. and Cheremisinoff, P.A. (eds).
Lancanster, PA: Technomic Publishers, pp. 196-214.

Metting, EB. (1993) Structure and physiological ecology of soil microbial
communities. In Sail Microbial Ecology: Applications in Agricultural and
Environmental Management. Metting, F.B. (ed). New York: M. Dekker, pp. 3-25.

Miller, K.J., and Wood, J.M. (1996) Osmoadaptation by rhizosphere bacteria. Annu
Rev Microbiol 50: 101-1 36.

Mishustin, E.N., and Mirsoeva, VA. (1967) In The ecology of soil bacteria. Gray,
T.R.G., and Parkinson, D. (eds). Liverpool: Liverpool Univ. Press, pp. 458-473.

Murphy, P.J., and Saint, GP. (1991) Rhizopines in the lugume-Rhizabium
symbiosis. In Molecular Signals in Plant-Microbe Communications. Verma, D.P.S.
(ed). London: CRC Press, pp. 378-390.

21

Nystrom, T., Flardh, K., and Kjelleberg, S. (1990) Responses to multiple-nutrient
starvation in marine Vibrio Sp. strain CCUG15956. J Bacterial 172: 7085-7097.

Nystrt'im, T., Olsson, RM, and Kjelleberg, S. (1992) Survival, stress resistance, and
alterations in protein expression in the marine Vibrio sp. strain 814 during
starvation for different individual nutrients. Appl Environ Microbial 58: 55-65.

Nystrom, T., and Neidhardt, EC. (1992) Cloning, mapping and nucleotide
sequencing of a gene encoding a universal stress protein in Escherichia coli. Mal
Microbiol 6: 3187-3198.

Nystrtim, T., and Neidhardt, EC. (1993) Isolation and properties of a mutant of
Escherichia coli with an insertional inactivation of the uspA gene, which encodes
a universal stress protein. J Bacterial 175: 3949—3956.

Nystrom, T., and Neidhardt, EC. (1994) Expression and role of the universal Stress
protein, UspA, of Escherichia coli during growth arrest. Mal Microbial 11: 537-
544.

Paul, E.A., and Clark, BE. (1989) Soil Biology and Biochemistry. New York: Academic
Press.

Reeve, C.A., Bockman, AT, and Matin, A. (1984) Role of protein degradation in the
survival of carbon-Starved Escherichia coli and Salmonella typhimurium. J
Bacteriol 157: 758-763.

Rovira, AD. (1956) A Study of the development of the root surface microflora
during the initial stages of plant growth. J Appl Bacterial 19: 72—79.

Rovira, AD. (1959) Root excretions in relation to the rhizosphere effect. IV.
Influence of plant Species, age of plant, light, temperature, and calcium
nutrition on exudation. Plant Soil 1 1: 53-64.

Rovira, AD. (1979) Biology of the soil-root interface. In The Soil Root Interface.
Harley, J .L., and Scott Russell, R. (eds). New York: Academic Press, pp. 145—160.

Rovira, A.D., Newman, E.I., Bowen, H.J., and Campbell, R. (1974) Quantitative
assessment of the rhizoplane rnicroflora by direct microscopy. Sail Biol Biachem
6: 21 1-216.

Sanjuan, J., and Olivares, J. (1989) Implication of nifA in regulation of genes
located on a Rhiabium meliloti cryptic plasmid that affect nodulation efficiency.
JBacteriol 171: 4154-4161.

Sanjuan, J., and Olivares, J. (1991) NifA-NtrA regulatory system activates
transcription of nfe, a gene locus involved in nodulation competitiveness of
Rhizobium meliloti. Arch Microbiol 155: 543-5 48.

Shields, J.A., Paul, E.A., and Lowe, W.E. (1973) Turnover of microbial tissue in soil
under field conditions. Soil Biol Biachem 5: 753-764.

22
Siala, A., Hill, LR, and Gray, T.R.G. (1974) J Gen Microbial 81: 183-190.

Sivasithamparam, K., Perker, C.A., and Edwards, CS. (1979) Rhizosphere
microorganisms of seminal and nodal roots of wheat grown in pots. Sail Biol
Biochem 11: 155-160.

Smiles, DE. (1988) Aspects of the physical environment of soil organisms. Biol Fertil
Soils 6: 204—2 1 5.

Soto, M.J., Zorzano, A., Mercado-Blanca, J., Lepek, V., Olivares, J., and Toro, N.
(1993) Nucleotide sequence and characterization of Rhiabium meliloti
nodulation competitiveness genes nfe. J Mal Biol 229: 570—576.

Soto, M.J., Zorzano, A., Garcia-Rodriguez, F.M., Mercado—Blanca, J., Lopez-Lara, I.M.,
Olivares, J., and Taro, N. (1994) Identification of a novel Rhizobium meliloti
nodulation efficiency nfe gene homolog of Agrabacterium ornithine
cyclodeaminase. MPMI 7: 703-707.

Spector, MP. and Cubitt, CL. (1992) Starvation-inducible loci of Salmonella
typhimurium: regulation and roles in starvation-survival. Mal Microbial 6:
1 467-1 476.

Spiro, S. (1994) The FNR family of transcriptional regulators. Antonie van
Leeuwenhoek 66: 23-36.

Spiro, S., and Guest, J.R. (1990) Pm and its role in oxygen-regulated gene
expression in Escherichia coli. FEMS Microbiol Rev 75: 399-428.

Spiro, S., and Guest, J.R. (1991) Adaptive responses to oxygen limitation in
Escherchia coli. TIBS 16: 310-314.

Spiro, S., Roberts, RE, and Guest, J.R. (1989) FNR-dependent repression of the
ndh gene of Escherichia coli and metal ion requirement for FNR-regulated gene
expression. Mal Microbial 3: 601-608.

Stotzky, G. (1972) Activity, ecology, and population dynamics of microorganisms in
soil. CRC Crit Rev Microbiol 2: 59-137.

Totten, P.A., Lara, J.C., and Lory, S. (1990) The rpoN gene product of Pseudamanas
aeruginasa is required for expression of diverse genes, including the flagellin
gene. J Bacteriol 172: 389-396.

Unden, G., Becker, 8. Bongaerts, J., Schirawski, J., and Six, 8. (1994) Oxygen
regulated gene expression in facultatively anaerobic bacteria. Antonie van
Leeuwenhoek 66: 3—23.

van Elsas, J.D., and van Overbeek, LS. (1993) Bacterial responses to soil Stimuli. In
Starvation in Bacteria. Kjelleberg, S. (ed). New York: Plenum Press, pp. 55—79.

23

van Vuurde, J.W.L., and Schippers, B. (1980) Bacterial colonization of seminal
wheat roots. Sail Biol Biochem 12: 599-565.

Vancura, V. (1988) Plant metabolites in soil. In Developments in Agricultural and
Managed-Forest Ecology 1 7. Soil Microbial Associations: Control of Structures
and Functions. Vancura, V. and Kunc, F. (eds). New York: Elsevier, pp.57-144.

Wada, A., Yamazaki, Y., Fujita, N., and lshihama, A. (1990) Structure and probable
genetic location of a “ribosome modulation factor” associated with 1008
ribosomes in stationary-phase Escherichia coli cells. Proc Natl Acad Sci USA 87:
2657-2661.

Yamagishi, M., Matsushima, H., Wada, A., Sakagami, M., Fujita, N., and lshihama, A.
(1993) Regulation of the Escherichia coli rmf gene encoding the ribosome
modulation factor: growth phase- and growth rate-dependent control. EMBO J
12: 625—630.

Zhulin, I.B., Lois, A.F., and Taylor, BL. (1995) Behavior of Rhizobium meliloti in
oxygen gradients. FEBS Lett 367: 180-182.

CHAPTER 2

Isolation and Characterization of Rhizobium meliloti Genes

Whose Promoters are Induced by Nitrogen Deprivation.

Parts of this chapter have been presented at the 6th International Microbial
Ecology Symposium (Lim et al., 1993) and the 10th International Nitrogen Fixation
Symposium (de Bruijn et al., 1995), and have been submitted for publication in
Molecular Microbiology (Ragatz et al., 1997).

NOTE: Parts of this chapter were contributed by PyungOK Lim (the generation of
the mutant collection, Southern blot analysis of auxotrophs, and screening for
nitrogen mutants), Michael Renner (the screening and characterization of
auxotrophs), and Anne Milcamps (the gene-replacement of strain N5).

24

2 5
ABSTRACT

Soil bacteria are subject to constantly changing environmental conditions,
and the ability of bacteria to sense these changing conditions and respond
accordingly is of vital importance to their survival and persistence in the soil and
rhizosphere. Nitrogen is particularly limited in the soil and rhizosphere, and
consequently it has been of interest to understand how Rhizobium meliloti, a
common soil isolate, senses and responds to this nutrient deprivation condition. To
investigate this problem, R. meliloti 1021 was mutagenized using a derivative of
Tn5 which creates transcriptional fusions to promoterless luxAB genes.
Subsequently, 5000 insertion mutants were screened for gene fusions induced by
nitrogen deprivation. The isolation of twenty-two gene fusions induced by nitrogen
deprivation is described. These fusions were found to be temporally regulated and
induced to a range of intensities. The strains harboring Tn5-luxAB fusions were
tested for nodulation and nitrogen-fixation phenotypes and two Strains were found
to be FiX'. Cloning and partial DNA sequence analysis of the transposon tagged loci
revealed a variety of novel genes, as well as R. meliloti genes with sequence
Similarity to known bacterial loci. Genes identical to already described genes for
exopolysaccharide synthesis (ean and eon) were found, as well as genes with

Significant Similarity to assimilatory nitrate and nitrite reductases.

2 6
INTRODUCTION

Bacteria in the soil and rhizosphere are subject to constantly changing
environmental conditions to which they must adapt or be out—competed. Rhizobia
in particular, must compete for a niche both in the soil as well as the rhizosphere of
their host plant in order to be successful in nodulation. As mentioned in Chapter 1,
the soil is very oligotrophic, and consequently nutrient deprivation is a common
environmental factor. Available nitrogen is particularly limited in the soil and
rhizosphere, and consequently it has been of interest to understand how
Rhizobium meliloti, a common soil isolate, senses and responds to this nutrient
deprivation condition.

Some genes have already been isolated in rhizobia and other bacteria that
respond to nitrogen deprivation, including the nitrate and nitrite reductase genes
(Goldman et al., 1994; Cai and Walk, 1997) and genes involved in formation of
exopolysaccharides (Gray et al., 1990; Ozga et al, 1994). In addition, rhizobia have
been found to contain a two component nitrogen-regulatory system, consisting of
ntrB and ntrC that is responsible for controlling nitrogen metabolism genes (Dusha
and Kondorosi, 1993; Arcondeguy et al., 1997), as well as participating in the
regulation of nitrogen fixation genes (although this varies between rhizobial
genera; Fischer, 1994; see also chapter 4, Figure 4-2). This ntr system regulates
two of the three distinct loci involved in glutamine sythesis (glnII and gInT) in R.
meliloti which are differentially regulated in response to varying ammonia levels
(Szeto et al., 1987; de Bruijn et al., 1989).

In most systems described thus far, identification of nitrogen deprivation—

induced proteins has been accomplished via 2D-PAGE analysis. This analysis has

27
some advantages, Since it generates an overall view of the deprivation response at

the protein level. However, it has disadvantages as well, since it fails to identify the
regulated genes/promoters and corresponding mutations to ascertain the role of
the genes.

Therefore, I have taken a different approach for investigating nitrogen
deprivation-induced loci, by mutagenizing the genome of R. meliloti with a Tn5
derivative containing the promoterless luxAB genes (Walk et al., 1991), encoding
luciferase as gene reporter. This transposon creates transcriptional fusions,
bringing the luxAB genes under the control of the tagged gene promoter. Under
appropriate conditions, these gene fusions will express luciferase, resulting in
bacterial bioluminescence (Meighen and Dunlap, 1993). The use of Tn5 has been
shown to be an extremely useful tool for random mutagenesis in many gram
negative bacterial systems including R. meliloti (Meade et al., 1982; De Vos et al.,
1986; Ditta, 1986; de Bruijn, 1987; Simon et al., 1989; Sharma and Signer, 1990;
de Bruijn and Rossbach, 1994). The use of luciferase as a reporter has proven to be
useful in studies of bacterial gene expression (Carmi et al., 1987; Heitzer et al.,
1994). In fact, recently we have used Tn5-1063 successfully in Pseudamanas
fluorescens to isolate gene fusions whose bioluminescence were induced under
nitrogen and phosphate deprivation (Kragelund et al., 1995). And it has also been
used in Anabaena Sp. to isolate gene fusions induced by nitrogen deprivation (Cai
and Walk, 1997). In addition, lux or luc expression has been a useful marker in
studies involving the tracking of microorganisms in the soil and rhizosphere (De
Weger et al., 1991; Beauchamp et al., 1993; Boelens et al., 1993; Moller et al.,

1994; Jansson, 1995).

28
In this chapter, the generation and characterization of a collection of 22 R.

meliloti mutants generated by random Tn5-1063 mutagenesis, carrying gene
fusions induced by nitrogen deprivation, is described. These fusions were found to
be temporally regulated and induced to different levels of intensity. The strains
harboring Tn5-luxAB fusions were tested for nodulation and nitrogen-fixation
phenotypes and two strains were found to be Fix‘. Cloning and partial DNA
sequence analysis of the transposon tagged loci revealed a variety of novel genes,
as well as R. meliloti genes with sequence Similarity to known bacterial loci,
including the exopolysaccharide genes exoY and exaF, and assimilatory nitrate and

nitrite reductases.

2 9
RESULTS

Tn5-1063 mutagenesis of R. meliloti 1021

Plasmid pRL1063a, carrying the promoterless quAB transposon Tn5-1063
(Figure 2-1A), was used to generate a collection of 5000 R. meliloti 1021 insertion
mutants. Insertion of Tn5-1063 into a R. meliloti gene can result in the creation of
a transcriptional fusion, whereby the luxAB genes become controlled by the
resident R. meliloti promoter (Figure 2-1B). Due to the presence of an E. coli origin
of replication in Tn5-1063, the interrupted gene can be easily excised from the
genome, ligated to form a self-replicating plasmid, and recovered by
electroporation or calcium chloride transformation into E. coli (Figure 2-1C).

The insertional Specificity of the Tn5-1063 transposon in R. meliloti 1021
was examined by screening the collection of 5000 for auxotrophs and determining
the nature of the auxotrophies. Out of 5000 insertion mutants, 62 were found to
be auxotrophic, a frequency of 1.2%, distributed among thirteen phenotypic
groups (Table 2-1).

The insertional specificity of Tn5-1063 was also examined by Southern blot
analysis. Total DNA of 69 R. meliloti strains, including auxotrophs and selected
prototrophs, was isolated, digested with EcoRI or ClaI (neither enzyme cuts Tn5-
1063), blotted, and hybridized with pRL1063a plasmid DNA as a probe. The
analysis of the Southern blots revealed different sized bands for each isolate (not
shown), suggesting random insertion of Tn5-1063 in R. meliloti 1021. In addition,
the number of hybridizing fragments was examined, to determine whether the

Tn5-1063 insertion occurred via a Single transpositional event, or involved other

Figure 2-1.

 

DNA sequencing
from primers A & B

Tn5-1063 structure, mutagenesis protocol, and target junction DNA
sequencing. The structure of the suicide plasmid carrying Tn5-1063
(Walk et al., 1991) is Shown in Panel A. In Panel B, the transcriptional
fusions generated by Tn5-1063 in target genes is schematically
diagrammed. In Panel C, the cloning of Tn5-1063 tagged loci is
diagrammed and the position of the DNA sequencing primers (A, B),
as well as the direction of the sequencing reactions are indicated.

31

Table 2-1. Phenotypes of selected auxotrophic Tn 5-1063 insertion mutants of R. meliloti 1021.

 

Nutritional ‘ Hybridizing b _—Symbi°fi° Phemtype
Strains Requirements Fragments Nod ‘ Fixd

 

1
ll

4 C osine or Uracil

a
E
i

>
g—A
+
+

10 Serine or Cl cine

>

1

E
i

l Serine or I cine
17 Histidine

>
l—I
5

D.

n.d

19 C osine or Uracil

>
p—A
+
+

1 T tohan

BF:
3

6 orune
T to han
T to ban

[3

i
E
E

3»
83
+
+

39 C tosine or Uracil

>
+
P
9-

T tohan

>
It
+
.9
e

H oxanthine

>
8
l
.5
P-

49 T osine

>
p-I
+
+

1 Adenine or H oxanthine
6

EH;

61 Methionine or C steine

>
r-I
+
+

3.
3

Methionine

l—l
+
+

3.
$
+
+

' tidine

 

a) As determined by the Holliday (1956) method.

b) Number of genomic EcoRI fragments hybridizing with pRL1063a sequences.

c) Ability to nodulate alfalfa plants (de Bruijn at al., 1989). n.d. = not determined.

d) Nitrogen fixation ability 0 alfalfa nodules as determined by the acetylene reduction
assay (de Bruijn et al., 1989)

in

32
mechanisms, such as multiple transposition or cointegration of the plasmid

pRL1063a. A single hybridizing fragment was found in 61 of the 69 (88%) Strains
tested (not Shown), including those which were further characterized (Table 2-2b),
indicating a single transposition event in most cases.

The auxotrophic mutants were examined for their symbiotic phenotype by
testing their ability to form effective nodules on alfalfa seedlings. Nitrogenase
activity of plant nodules was determined by the acetylene reduction assay (see
Materials and Methods). All auxotrophic strains, except for two leucine (Leu')
requiring Strains, induced nodules on alfalfa (Nod*). All leucine mutants induced
nodule-like structures which were ineffective in nitrogen fixation (FiX'). In addition,
all adenine or hypoxanthine requiring (Pur') Strains, one methionine mutant, and
one cytosine or uracil (Pyr) requiring strain, failed to produce effective nodules
(Nodt FIX‘). However, three of the Six Pyr auxotrophs did produce effective nodules

(Nodt Fixt; see Table 2—1).

Isolation of R. meliloti mutant Strains carrying Tn5-1063 fusions induced by

nitrogen deprivation

The collection of 5000 random Tn5-1063 insertion mutants was screened
for luxAB expression induced by nitrogen deprivation, as described in Materials
and Methods. The bioluminescent images of colonies on filters before and after the
transfer to nitrogen deprivation conditions were compared. Twenty-two mutants
were isolated whose bioluminescence was consistently activated under nitrogen

deprivatidn conditions (Figure 2-2A).

Figure 2-2.

33

Ludferase activity determination of fusions induced by nitrogen
deprivation. A screen of nitrogen deprivation-induced lux gene
fusions by using the Hamamatsu camera is Shown. In Row A, the
analysis of 21 nitrogen deprivation-induced fusions is depicted 0, 7,
and 24 hours after transfer to nitrogen-free medium, respectively.
The strain designations are the same in all panels and are outlined in
Row B (first block). The induction patterns of the fusions on 0.2%
nitrate, and 0.2% glutamine as sole nitrogen sources are depicted in
Rows B & C, respectively. For further details, see text.

 

34

 

®©®@
ONCE
®®®®
®®©O@
©®®®O

 

35
The Strains were examined for growth on nitrate (KNO3) and glutamine, a

poor nitrogen source and a good nitrogen source for R. meliloti, respectively. All of
the Strains grew well on glutamine; however, five of them failed to grow on KN03
(Table 2-2b). In addition, the strains were examined for induction and repression
of their gene fusions by KNO3 and glutamine. Ten of the gene fusions were
activated in the presence of KNO3, but none of the gene fusions were induced in
the presence of glutamine as the sole nitrogen source (Table 2-2a; Figure 2-2B).
Two of the gene fusions (N1 19 and N150) appeared to be expressed at a low level
in the presence of glutamine (Figure 2-2C). This appears to be a background
expression level, since a similar level of bioluminescence was observed on GTS
(minimal media) plates lacking glutamine (not Shown).

The nitrogen deprivation-induced gene fusions were examined for temporal
expression patterns, as described in the Materials and Methods. The level of
expression in response to nitrogen deprivation (Figure 2-3) ranged from very high
(A), to medium (B), to low (C). In addition, the temporal regulation of the
expression of the fusions varied from a very rapid induction pattern within a few
hours (Strains N21, N110, N3, N104, and N12), to a late induction pattern after 10
to 15 hours (strains N119, N161, N113, and N4). Some fusions (strains N161 and
N183) were induced only after 18 to 24 hours of nitrogen deprivation conditions
(data not Shown). These fusions were induced at such comparatively low levels that
they could not be graphically displayed with the other fusions.

The varied temporal patterns of nutrient deprivation regulated loci are
similar to what has been seen in E. coli (Matin, 1991), S. typhimurium (Givskov, et
al., 1994), and P. fluorescens (Kragelund et al., 1995). These data suggest that the

fusions are under the control of different promoters.

36

Table 2-2. Phenotypes of Tn 5—1063 insertion mutants of R. meliloti 1021 induced
by nitrogen deprivation.

 

Lux Fusion Expression ' b Symbiotic c
in the Presence of: Growth on: Phenotype: ,
Hybridizingd Genomee Gene

Strains -N +N03- +Gln +NO3- +Gln Nod Fix Fragments Location Similarity

 

 

Rm N1 + + — + + + + 1 n.d. NS

 

 

a) Induction of Tn5-1063 fusions by growth of mutants on G'IS without nitrogen (-N),
with 0.2% nitrate (+NO 3-), and with 0.2% glutamine (+Gln).
b) Growth of mutants on GTS with 0.2% nitrate or 0.2% glutamine as a sole nitrogen source.
c) Symbiotic phenotype of alfalfa roots inoculated with mutant strains for presence of nodules
on roots (Nod +) and presence of nitro ogenase as determined by acetylene reduction (Fix +).
d) Number of fragments of genomic DN cut with 15le which hybridize to a probe of
whole pRL1063a.
e) Genomic location of Tn 5-1063 insertions (Honeycutt, 1993; 'from personal correspondence
with R. Honeycutt).
f) Genes sharing significant similarity with R. meliloti ORFs: nasAB (Lin et al.,1993);
aonY (Miiller etal.,1993); n. d.= not determined, NS= no significant similarity to
known genes in GenBank

37

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

60000
'2 500004
3 400004 +N21
o ---G-- N30
2 3°°°°‘ +N110
3 20000— "43-- N12
0: +cv2
'U +N15
5 ---0-- N111
8 ---l-- N4
2 --~I3-- N113
3 —¢—cv2
n:
navy-"'9'"
1’ +N104
1: 600 ‘-
8 500 a ""G" N161
° ' —I—N119
to ~ \
\ 400 ‘. ---o-- N3
3 300 7" \ 9-‘_
J .. l " I, "'\__“ "_ 1
m 200 ,. p" y -°-._‘__ _°_-o-" -- .l
100 ‘§ ' __-o—'E-'----
o x """" T I 1 1
0 5 10 15 20 25 30
Time of Induction (hours)
Figure 2-3. Temporal qu expression pattern of selected nitrogen deprivation-

induced gene fusions. The temporal lux expression patterns of 1 4
Tn5-1063 induced gene fusions after transfer to nitrogen free
medium is shown in Panels A-C. On the Y-axis, relative light units per
second are indicated (RLU/ sec). On the X-axis, five hour intervals are
indicated (see bottom of Panel C). CV2 is included as a control.

3 8
In order to examine the physical structure of the transposon tagged Strains,

total DNA of the Tn5-1063 containing strains was isolated, digested with either
EcoRI or ClaI (neither enzyme cuts the transposon), blotted, and hybridized with
plasmid pRL1063a DNA as a probe. Two of the 22 strains displayed two
hybridizing fragments for EcoRI and C131 restricted DNA (Table 2-2d), indicating
that either an integration of the plasmid pRL1063a or a multiple transposition
event had occurred. These two Strains were not analyzed further except to test for
nodulation and nitrogen-fixation ability (see Table 2-2).

The strains carrying fusions induced by nitrogen deprivation were
examined for their symbiotic phenotype by testing their ability to form effective
nodules on alfalfa seedlings. Nitrogenase activity of plant nodules was determined
by the acetylene reduction assay (see Materials and Methods). All strains except for
two (Strains N112 and N149; described below), induced nodules (Nodt) on alfalfa
that were nitrogen-fixation competent (Fixt) (Table 2-2).

In order to identify the nitrogen deprivation-induced genes, Tn5-1063
mutated loci were cloned from the R. meliloti genome. Total genomic DNA of strains
carrying Tn5-1063 fusions was digested with EcoRI or ClaI restriction enzymes
which do not cut the transposon (Figure 2-1C). The resulting fragments were self-
ligated and transferred into E. coli DH501 or HB101 by electroporation or CaClz
transformation. The DNA sequence of the R. meliloti DNA fragments flanking the
Tn5-1063 inverted repeats was determined by using unique primers (Figure 2-1C)
homologous to the left and right ends of the Tn5-1063 (see Materials and
Methods). A nine base-pair duplication of the target site of Tn5-1063 was observed
in each case, and the DNA sequences flanking the Tn5-1063 were fused by

omitting the duplicated target sequence.

39

Significant open reading frames (ORFS) were found for all of the isolated
gene fusions, as determined by analyzing the DNA sequence with a codon
preference program based on the codon usage of R. meliloti (Figure 2-4). This
program also indicates start and stop codons, as well as potential non-coding
regions (Figure 2-4A). In addition, potential frame-Shifts (Figure 2-4B) caused by
compressions of R. meliloti 1021 GC-rich Stretches during DNA sequencing could
be recognized and resolved by further sequencing (Figure 2-4C).

The partial DNA sequences of the Tn5-1063 tagged loci from strains N4 and

N41 were found to be identical, thus indicating that the two strains are clonal.

Sequence Similarities of R. meliloti Tn5-1063 tagged lad induced by nitrogen
deprivation

The amino—acid sequences deduced from the primary DNA sequence were
compared to sequences in the non-redundant protein databases of GenBank (Table
2-2f). The sequence data indicated that the Tn5-1063 tagged locus in Strain N30
possessed Significant Similarity (Figure 2-5A) to assimilatory nitrate reductases of
Klebsiella pneumoniae (NasA; Lin et al., 1993), Oscillataria chalybea (NarB; Unthan
et al., 1996), and Synechocaccus Sp. (NarB; Omata et al., 1993), and to the formate
dehydrogenases FdnG (Berg et al., 1991) and FdoG (Plunkett et al., 1993) of E. coli.
Since this particular mutant Strain is unable to grow with KNO3 as sole nitrogen
source (Table 2-2d), it is likely that the gene tagged encodes the assimilatory
nitrate reductase of R. meliloti. In addition, this gene fusion is induced in the
presence of its substrate, KNO3 (Figure 2-ZB; Table 2-2b), which has also been

observed for the K. pneumoniae nasA gene (Goldman et al., 1994). The alignment

Figure 2-4.

 

 

 

 

 

 

 

 

 

 

N9lsgrrect;19 ,, ,

+1

 

 

 

+2

 

 

(laden-usage analysis of Tn5-1063 tagged target gene DNA
sequences. In Panel A, the codon-usage analysis of the Tn5—1063
tagged strain N3 is Shown. The three open reading frames (ORFS) in
the direction of lux reporter gene transcription are indicated by +1,
+2, and +3, according to C. Hailing (see Materials and Methods). The
DNA coordinates (in intervals of 100 bp) are indicated on the
horizontal axes at the bottom of Panels A-C. In Panels B & C, the
codon-usage analysis of mutant Strain N9 is shown, before (B) and
after (C) correcting for a frameshift in the DNA sequence (see text).
The significance of the codon usage is indicated by the Shaded areas.
Start and stop codons are indicated by long and short vertical lines,
respectively, at the bottom of the panels. Horizontal lines connecting
the Start & Stop codons indicate potential ORFS.

B I9 In
- Ill-D ll

IirB IC
Inna Bl

 

..-.41....._.

....41.......

 

 

 

 

 

 

 

 

83338. 588883

 

 

 

I v E ---------- RYLER vK RVGLMEIRRQ -Moov K; F fl
1 as KQLVLVGNGMAG Rmiermtsv DEFQ ------ TIFGA P P - a
1 s VRLAIIGNGMVGH.FIED LDK DAANF- ----- TVFCE P 1 - u
1 _T PVLVLVGHGMVGHHF QCVSRDLHQQYR ----- VVFCE YA. - M
to DRF r o v p 51: -RvschxHErHPM[ZITvcEIQSG ~|Dfi0 . I
u NRI L a“ G TDIKDITLN wo YEENIIQLmTNE H KVD E K O
a DRVHL v HT--AE--ELSLV EGFYEKHHIKV v02” TINRQ In
As IV.HL Y .GRS--. --s LVEEGDF TQH IELRL EHVAS . on
I7 r HIv svo LVIATGSVPFIIPVPGK 1111101.: :1!
01 IA R PrDlfiLILATGs P I PI GA K. 1.-.rr<| :
01 v H SA R V; DKLIMATGSVPéIPPI GS QéJCFVVRTIE 1

fl v RDAF HETH v. . v- P- ' H E CFv :T 1 ‘fl

 

....

H!
m
mmt/HXJ
<< R
Ono
X!”
‘7‘0'0
*Eb’
<r-r—

z

4

-<
1-‘r‘r‘
>>O

 

a S
e
rrr
pvt
'51»
—<
2722?")
MOO
rrr
~32
>13)
awn
n
mm <
_, mx<
can 0
1131»
>>
<<<
on,
'R’ M
“Eh «<4
rrr -rM)
an: >) Fri
n n I 44 xx xx-
IK 'u'oxi-<
3:: «<1 a,“
tn-Id II
nnn‘ HF“
<<< “ ‘UV
090 <-<
In —1 < 0102'!
TA U 33:!
f x
.. :‘.“ 22
as: no
<‘Hfi 1
fimfi 1'14 .
< 4 I'm l
0 OO 1),)
COO ) r222
mum -—
Eh .000
“>9 xxx
rrr ~UOO
>00 006
<—— 444
rn—I )u <<<
fl worn
S3! "I

 

SI!
<<<
<<<
'U'V‘
’l,

w_ I

ram

-44

'_<

i I I-(X

"1

rrr

> x— a
>

0))

00

—<<

<

a

In
On:
r—<

‘
r
<
XXX
(I<
066'!
9‘61 2::
000
U”,
00
PF
1171 r'
“G
> _
x
I
mfln
000
I“

 

 

 

Similarity of the amino-acid sequence deduced from selected Tn5-1063 tagged
ORFS induced by nitrogen deprivation. In panel A, the amino-acid sequence
Similarity of the putative protein product of the Tn5-1063 tagged ORF of strain
N30 to nitrate reductases of K. pneumoniae (NaSA), Oscillatoria chalybea
(NarB), Synechococcus Sp. (NarB), and to formate dehydrogenases of E. coli
(FdnG, FdoG) is shown. Identical amino-acids are Shaded and conserved amino-
acid residues are boxed (based on the PAMZSO matrix). The amino-acid residue
coordinates are given at the beginning and end of each line. The triangles mark
a conserved four cysteine consensus [4Fe-4S] cluster-binding motif. In panel B,
similarity of the putative protein products from the Tn5-1063 tagged ORF of
Strain N9 to nitrite reductases from Bacillus subtilis (NasD), A. vinelandii
(NaSA), and K. pneumoniae (NasB) is shown. In panel C Similarity of the
putative protein products from the Tn5-1063 tagged ORF of strain N5 to nitrite
reductases from Bacillus subtilis (NasD), K. pneumoniae (NasB) and E. coli
(NirB) is Shown. The Tn5-1063 of Strains N9 and N5 is located within the same
ORE. The solid bar in panel B indicates the Start of an NAB-binding domain. In
panel C, the conserved cysteines of the 4Fe-4S / siroheme domain found in
nitrite reductases and sulfite reductases are indicated with triangles.

42
of the deduced amino—acid sequence of the Tn5-1063 tagged locus of strain N30,

with the nitrate reductases and formate dehydrogenases is shown in Figure 2-5A.
A conserved four cysteine consensus [4Fe-4S] cluster-binding motif (Breton et al.,
1994), marked by triangles in the amino-terminal region, has been postulated to
contribute to iron-sulfur binding (Berg et al., 1991).

The Tn5-1063 tagged loci in strains N5 and N9 Share significant Similarity
with the Bacillus subtilis assimilatory nitrite reductase NasD (Ogawa et al., 1995),
the K. pneumoniae assimilatory nitrite reductase NaSB (Lin et al., 1993), and the E.
coli NADH-dependent nitrite reductase NirB enzymes (Peakman et al., 1990)
(Figure Z-SB). These isolates did not grow on minimal medium with nitrate as the
sole nitrogen source (Table 2-2d), confirming that their assimilatory nitrite
reduction pathway was inactivated. As in the case of Strain N30, the gene fusions
were induced in the presence of KNO3 (Figure 2-ZB; Table 2-2b). In mutant strain
N5, it was observed that in addition to assimilatory nitrite reductase inactivation, a
non-motile phenotype was present. Gene replacement of the wild-type R. meliloti
gene with the mutated analogous gene restored the motility, but growth with
nitrate as sole nitrogen source was still abolished. This confirmed that the tagged
gene was indeed in the assimilatory nitrite reductase locus, and that the loss of
motility in the original strain was not linked to the Tn5-1063 mutation.

The Tn5-1063 tagged loci of strains N112 and N149 were identified as the
exoY and exaF genes, respectively, due to a 100% homology with the sequences
obtained from GenBank (Muller et al., 1993). These genes are involved in
exopolysaccharide production, and are organized as contiguous genes in an operon
in R. meliloti (Glucksmann et al., 1993). The phenotype of ean and exaF mutants

includes an absence of exopolysaccharide production, as well as the presence of

43
nodule-like structures which are unable to fix nitrogen (Gray et al., 1990; Leigh

and Walker, 1994). Mutant strains N112 and N149 were indeed found to lack
exopolysaccharide production under nitrogen deprivation conditions, based on
visual inspection of the plates for mucoid colonies typical for the N-starved wild—
type Strain (data not Shown). When these mutants were inoculated on alfalfa roots
(Table 2-2c), nodule-like structures were formed which did not fix nitrogen (FiX'),
confirming the phenotype described for these mutants (Gray et al., 1990; Leigh
and Walker, 1994).

Even though Significant ORFS were identified in each of the remaining
nitrogen deprivation-induced loci (data not Shown), no Significant similarity of
these loci to sequences in the non-redundant protein databases at GenBank was

found, suggesting that they represent novel genes (Table 2-2f).

4 4
DISCUSSION

Insertional Specificity of Tn5-1063: Characterization of auxotrophic mutants

In the collection of 5000 R. meliloti Tn5-1063 mutants examined,
auxotrophic mutants were isolated with a frequency of 1.2%. This frequency is
higher than has been observed with Tn5 derivatives in R. meliloti thus far (Meade
et al., 1982). However, it is lower than frequency observed with nitrous acid
mutagenesis (Kerppola and Kahn, 1988). With regard to the type of auxotrophies,
Meade et al. (1982) also found a high number of auxotrophs which could be
supplemented by methionine (9 of 20), but none supplemented by tryptophan. In
Rhizobium Ieguminasarum, however, a high number of tryptophan and methionine
auxotrophs was also observed (Pain, 1979). The high number of tryptophan and
methionine requiring auxotrophs may indicate non-randomness of Tn5-1063
insertion, or a high number of genes involved in tryptophan and methionine
synthesis.

At least eight genes are needed for tryptophan synthesis from the precursor
chorismate, which are tightly arranged in an operon in E. coli (Yanofsky et al.,
1981) and S. typhimurium (Yanofsky and van Cleemput, 1982). At least nine
genes, involved in the biosynthesis and transport of general aromatic amino-acids,
are necessary for tryptophan synthesis in E. coli and S. typhimurium (Pittard and
Wallace, 1966; Gollub et al., 1967). Four of the genes from the tryptophan
auxotrophs in this study (aux 35, 37, 42, and 47) were mapped by Honeycutt et al.

(1993), and reside within a Single cluster, on linkage group Pme #3 (1040 kb),

45
indicating that the R. meliloti genes are also tightly linked, similar to E. coli and S.

typhimurium, and may be similarly arranged.

Methionine metabolism also requires a large number of genes for its
synthesis and transport. In most bacteria examined, at least eight genes are
involved in methionine metabolism, and nine genes in E. coli (Rowbury, 1983).

The inability of leucine, methionine, purine, and pyrimidine auxotrophs of
R. meliloti to form effective nodules has been reported previously (Dénarié et al.,
1976; Meade et al., 1982; Kerppola and Kahn, 1988). Kerppola and Kahn (1988)
also found that methionine auxotrophs failed to produce effective nodules;
however, all of their methionine auxotrophs formed ineffective nodules, unlike
what was observed in this thesis (Table 2-1). In addition, they found that purine
and pyrimidine requiring mutants formed ineffective nodules. I also observed that
purine requiring mutants formed ineffective nodules; however, many of the
pyrimidine auxotrophs were able to form effective nodules.

These results suggest that the ability of Specific host plants to provide the
necessary nutrients to their endosymbionts varies. Many studies have shown that a
large number of factors affect the types and quantities of root exudates released
by plants including: the Species of plants used, the age of the plants, and
environmental factors such as temperature, moisture, irradiance, and other plant
stresses (Rovira, 1956 and 1959; Vancura, 1988). It is my belief that a
combination of these factors accounts for the variation of the types of auxotrophies

which lead to deficient (FiX‘) nodules.

46

Characterization of Tn5-1063 tagged loci induced by nitrogen deprivation.

In this study, the isolation of 22 R. meliloti genes involved in nitrogen-
deprivation was reported. Sequence analysis of the tagged loci induced by
nitrogen-deprivation revealed genes involved in nitrate and nitrite assimilation
(nasA and nasB) and exopolysaccharide synthesis (ean and exaP). It is known that
bacterial nitrate and nitrite assimilation genes are induced by nitrogen limiting
conditions (Peakman et al., 1990; Lin et al., 1993; Goldman et al., 1994). In
addition, it has been observed that exopolysaccharide synthesis decreases in the
presence of ammonia, although no Specific regulator has been described
(Sutherland, 1979; Ozga et al., 1994). Thus, the methods used to isolate and
screen for nitrogen deprivation-induced genes did generate a diverse set of loci
with predicted functions, as well as a number of novel genes, and suggests that the
conditions for nitrogen deprivation applied were sufficient and adequate for the
isolation of nitrogen deprivation-induced loci.

The number of genes isolated by nitrogen deprivation is Similar to what has
been observed in other bacteria (Matin, 1990; Givskov et al., 1994; Kragelund et
al., 1995 ), as was hypothesized (chapter 1); however, this does not imply that
saturation was reached, only that at least as many loci exist in R. meliloti induced
by nitrogen deprivation as exist in E. coli. In addition, the large number of novel
genes (fourteen), supports the hypothesis that soil bacteria may have different
genes for dealing with stress than enteric or marine bacteria. More work needs to
be carried out to characterize the novel loci before this hypothesis can be fully

tested. This important hypothesis will be examined in greater detail following the

47
analysis of genes induced by carbon deprivation (chapter 3) and oxygen limitation

(chapter 4).

The nitrate and nitrite reductase genes have not yet been isolated in R.
meliloti . The partial sequence analysis of both loci indicates a high similarity with
the K. pneumoniae nas genes. In this organism, the nas genes are organized in an
operon and their regulation occurs through the ntr system responding to nitrogen
availability, and the nasR gene responding to nitrate and nitrite availability
(Goldman et al., 1994). The limited sequence data does not warrant a prediction
about the organization of these genes in R. meliloti, but their regulation clearly

appears to be similar.

4 8
MATERIALS AND METHODS

Bacterial strains and plasmids

R. meliloti 1021 (Sm’) has been described by Meade et al., (1982), E. coli
DHSa by Hanahan (1983), and E. coli HB101 by Boyer and Roullaud-Dussoix
(1969). Plasmid pRL1063a (Walk at al., 1991) was kindly provided by P. Wolk
(Michigan State University). Plasmid pRK2013 has been described by Ditta et al.

(1980).

Media and growth conditions

R. meliloti was grown at 28°C in TY (Beringer, 1974) or in GT5 medium (Kiss
et al., 1979). E. coli Strains were grown at 37°C in LB medium (Silhavy et al., 1984).
Antibiotics were used at the following final concentrations: 250 ug/ml streptomycin
(Sm) for R. meliloti; 200 ug/ml kanamycin (Km) for R. meliloti; 20 ug/ml Km for E.

coli.

Transposon mutagenesis

The following procedure was used to mobilize plasmid pRL1063a, carrying
transposon Tn5-1063, to R. meliloti 1021 using the helper plasmid pRK2013. E.
coli DHSa, carrying pRL1063a or pRK2013, and R. meliloti 1021 were grown in LB-
szo and TY-Sm250 respectively, washed twice with TY, and concentrated five-fold in

TY medium. Equal amounts of donor, helper, and recipient cells were Spotted on TY

49
plates. After one day at 28°C, the mating mixtures were resuspended in sterile

distilled water and plated on selective plates. Five thousand (5000) Kmr colonies
were isolated, purified, grown in liquid TY medium, and stored in microtiter plates

at -70°C.

Identification of auxotrophic mutants

Auxotrophic mutants were identified by screening the transconjugants for
their inability to grow on GTS minimal medium. They were purified and tested for
their auxotrophic requirements using the supplementation test described by
Holliday (1956). Amino-acids were added at a final concentration of 20—80 mg/ ml,

nucleotides at 10 mg/ml, and vitamins at 5 mg/ ml (see Holliday, 1956).

Screening of the Tn5-1063 insertion mutants for nitrogen deprivation-induced

gene fusions

The R. meliloti strains carrying Tn5-1063 were spotted on membrane filters
(Nucleopore, Cambridge, MA) and incubated on solid GTS media for 48 hours.
Luminescent colonies on the plate were visualized by Spreading 50 pl of N—decanal
inside the top of a glass petri-dish, placing the glass petri-dish over the plate,
exposing the cells to N-decanal for 60 seconds, and measuring light emission using
the Hamamatsu photonic system model C1966-20, as described by Walk et al.
(1991).

In order to screen the mutants for nitrogen deprivation-induced gene

fusions, the membrane filters bearing the cells were transferred to both GTS

50
medium and GTS medium lacking nitrogen, and incubated for an additional 7 or

24 hours. Strains carrying Tn5-1063 whose luminescence was weak or absent
before transfer, but Showed Strong luminescence afterwards, were selected for

further analysis and rescreened.

DNA isolation and manipulation

Plasmid DNA was prepared by the alkaline method as described in
Kragelund et al. (1995). Total DNA was isolated from R. meliloti strains according to
de Bruijn et al. (1989).

All restriction enzyme digestion, ligation, and Southern blotting experiments
were carried out as described in Sambrook et al. (1989). Labeling of probes and
DNA hybridizations was performed using a non-radioactive DNA labeling and
detection kit (Boehringer Mannheim, Mannheim, FRG), according to the

manufacturer’s recommendations.

DNA Sequence analysis

Sequencing of double-stranded plasmid DNA was performed using the di—
deoxy method of Sanger et al. (1977) with sequenase kits (U.S. Biochemicals,
Cleveland, OH). To determine the R. meliloti DNA sequence on both Sides of the
transposon insertion, two Tn5-1063 derived oligonucleotides were synthesized by
the Macromolecular Synthesis Facility at Michigan State University and used as
sequencing primers. One primer (corresponding to position 110—86 of the Tn5-

1063 DNA sequence: 5'-TACTAGATTCAATGCTATCAATGAG-3') was designed to

51

determine the upstream sequence from the Tn5-target site in the antisense
direction. The other primer (corresponding to positions 7758—7781 of the Tn5-
1063 DNA sequence: 5 ' -AGGAGGTCACATGGAATATCAGAT-3 ') was designed for
determining the downstream sequence from the Tn5-target Site in the sense
direction. These primers were modified from previously described sequencing
primers for Tn5-1063 and other Tn5 derivatives (Black et al., 1993; Fernandez-
Pinas et al., 1994). DNA sequences were analyzed using the software packages of
the University of Wisconsin Genetics Computer Group (Devereux et al., 1984), and
DNASIS (Hitachi Software Engineering Co., Ltd.). Deduced amino-acid sequences
were analyzed using PROSIS (Hitachi Software Engineering Co., Ltd.). Database
searches were carried out using the program Blastx (Gish and States, 1993), by
screening the non-redundant GenBank database Release 1.4.9MP (8/17/97). The
codon preference profiles were determined by the program CodonUse 3.1 (Codon
window Size = 33, logarithmic range = 3), kindly provided by Conrad Hailing (U. of

Chicago, Chicago, IL 60637).

Nodulation and nitrogen-fixation assays

R. meliloti strains carrying Tn5-1063 insertions were screened for their
symbiotic phenotype by inoculation on alfalfa (Medicaga sativa) seedling roots.
Alfalfa seeds were sterilized by soaking for three minutes in 95% ethanol, followed
by three minutes in 0.1% HgClz and rinsed thoroughly with Sterile distilled water.
The seeds were placed on a piece of sterile Whatman 3MM filter paper in test tubes

containing 20 ml of sterile nitrogen-free B+D liquid medium (Broughton and

5 2
Dilworth, 1971). Saturated cultures (late logarithmic stage) of strains carrying Tn5-

1063 were diluted with sterile H20 (1:5) and one ml aliquots were added to tubes
containing one week old germinated seedlings. Inoculated plants were grown for Six
to seven weeks in a growth chamber (16 hrs light, 28°C), and examined for the
presence or absence of nodules (Nod phenotype).

Nitrogenase activity was measured in glass test tubes containing the
nodulated plants by capping each tube with a Stopper, injecting acetylene (1/10
volume), and withdrawing a one ml sample from the tube after 30 minutes,
followed by a measurement of acetylene reduction to ethylene by gas

chromatographic analysis.

5 3
REFERENCES

Arcondeguy, T., Huez, 1., Tillard, P., Gangneux, C., de Billy, F., Gojon, A., Truchet, G.,
and Kahn, D. (1997) The Rhizobium meliloti PII protein, which controls bacterial
nitrogen metabolism, affects alfalfa nodule development. Genes Dev 11: 1 194-
1 2 06.

Beauchamp, C.J., Kloepper, J.W., and Lemke, PA. (1993) Luminometric analyses of
plant root colonization by bioluminescent pseudomonads. Can J Microbial 39:
434—441.

Berg, B.L., Li, J., Heider, J., and Stewart, V. (1991) Nitrate-inducible formate
dehydrogenase in Escherichia coli K-12. 1. Nucleotide sequence of the fdnGHl
operon and evidence that opal (UGA) encodes selenocysteine. J Biol Chem 266:
22380—22385.

Beringer, J.E. (1974) R factor transfer in Rhizobium leguminasarum. J Gen
Microbial 84: 188-198.

Black, T.A., Cai, Y., and Walk, GP. (1993) Spatial expression and autoregulation of
hetR, a gene involved in the control of heterocyst development in Anabaena.
Mal Microbial 9: 77 -8 4.

Boelens, J., Zoutman, D., Campbell, J., Verstraete, W., and Paranchych, W. (1993)
The use of bioluminescence as a reporter to study the adherence of the plant
growth promoting rhizopseudomonads 7NSK2 and ANPIS to canola roots. Can J
Microbial 39: 329-334.

Boyer, H.W., and Roullaud-Dussoix, D. (1969) A complementation analysis of the
restriction and modification of DNA in Escherichia coli. J Mol Biol 41: 459-472.

Breton, J., Berks, B.C., Reilly, A., Thomson, A.J., Ferguson, SJ, and Richardson, DJ.
(1994) Characterization of the paramagnetic iron-containing redox centres of
Thiasphaera pantatropha periplasmic nitrate reductase. FEBS Lett 345: 76-80.

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

Cai, Y. and Walk, GP. (1997) Nitrogen deprivation of Anabaena Sp. strain PCC7120
elicits rapid activation of a gene cluster that is essential for uptake and utiliztion
of nitrate. J Bacterial 179: 258-266.

Carmi, O.A., Stewart, G.S., Ulitzur, S., and Kuhn, J. (1987) Use of bacterial luciferase
to establish a promoter probe vehicle capable of nondestructive real-time
analysis of gene expression in Bacillus Spp. J Bacterial 169: 2165-2170.

de Bruijn, P.J. (1987) Transposon Tn5 mutagenesis to map genes. Methods in
Enzymology 154: 175-196.

54

de Bruijn, F.J., Graham, L., Milcamps, A., and Ragatz, D. (1995) Use of the luciferase
(Tn5-lux) reporter system to Study Rhizobium meliloti genes responding to
N/C/O2 limitation or plant factors and their role in rhizosphere competition. In
Nitrogen Fixation: Fundamentals and Applications. Tikhonovich, I.A., Provorov,
N.A., Romanov, V.I., and Newton, W.E. (eds). Boston: Kluwer Academic
Publishers, pp. 195-200.

de Bruijn, F.J., and Rossbach, S. (1994) Transposon mutagenesis. In Methods for
General and Molecular Biology. Gerhardt, P., Murray, R.G.E., Wood, W.A., and
Krieg, N.R. (eds). American Society for Microbiology: Washingtion, D.C., pp.
387-405.

de Bruijn, F.J., Rossbach, S., Schneider, M., Ratet, P., Messmer, S., Szeto, W.W.,
Ausubel, F.M., and Schell, J. (1989) Rhizobium meliloti 1021 has three
differentially regulated loci involved in glutamine biosynthesis, none of which is
essential for symbiotic nitrogen fixation. J Bacterial 171: 1673-1682.

de Vos, G., Walker, G.C., and Signer, ER. (1986) Genetic manipulations in
Rhizobium meliloti utilizing two new transposon Tn5 derivatives. Mal Gen Genet
204: 485-491.

de Weger, L.A., Dunbar, P., Mahafee, W.F., Lugtenberg, B.J.J., and Sayler, GS.
(1991) Use of biolumenescence markers to detect Pseudamanas bacteria in the
rhizoshpere. App Env Microbial 57: 3641-3644.

Dénarié, J., Truchet, G., and Bergeron, B. (1976) Effects of some mutations on
symbiotic properties of Rhizobium. In Symbiatic Nitrogen Fixation. Nutman, P.S.
(ed). Cambridge: Cambridge University Press, pp.47—63.

Devereux, J., Haeberli, P., and Smithies, O. (1984) A comprehensive set of
sequence analysis programs for the VAX. Nucleic Acids Res 12: 387-395.

Ditta, G. (1986) Tn5 mapping of Rhizobium nitrogen fixation genes. Meth Enzym
118: 519-528.

Ditta, G., Stanfield, S., Corbin, D., and Helinski, DR (1980) Broad host range DNA
cloning system for gram-negative bacteria: construction of a gene bank of
Rhizobium meliloti. Prac Natl Acad Sci USA 77: 7347-7351.

Dusha, I and Kondorosi, A. (1993) Genes at different regulatory levels are required
for the ammonia control of nodulation in Rhizobium meliloti. Mol Gen Genet
240: 435-444.

Fernandez-Pinas, F., Leganes, F., and Walk, GP. (1994) A third genetic locus
required for the formation of heterocysts in Anabaena Sp. strain PCC 7120. J
Bacteriol 176: 5277 -5283.

Fischer, H.-M. (1994) Genetic regulation of nitrogen fixation in rhizobia. Microbial
Rev 58: 352—386.

SS

Gish, W., and States, DJ. (1993) Identification of protein coding regions by
database similarity search. Nat Genet 3: 266-272.

Givskov, M., Eberl, L., and Molin, S. (1994) Responses to nutrient starvation in
Pseudamanas putida KTZ 442: two-dimensional electrophoretic analysis of
starvation- and Stress-induced proteins. J Bacteriol 176: 4816-4824.

Glucksmann, M.A., Reuber, TL, and Walker, G.C. (1993) Family of glycosyl
transferases needed for the synthesis of succinoglycan by Rhizobium meliloti. J
Bacterial 175: 7033-7044.

Goldman, B.S., Lin, J.T., and Stewart, V. (1994) Identification and structure of the
nasR gene encoding a nitrate- and nitrite-responsive positive regulator of
nasFEDCBA (nitrate assimilation) operon expression in Klebsiella pneumoniae
MSal. J Bacterial 176: 5077-5085.

Gollub, E.G., Zalkin, H., and Sprinson, DB. (1967) Correlation of genes and
enzymes, and studies on regulation of the aromatic pathway in Salmonella. J
Biol Chem 242: 5323-5328.

Gray, J.X., Djordjevic, M.A., and Rolfe, B.G. (1990) Two genes that regulate
exopolysaccharide production in Rhizobium Sp. Strain NGR234: DNA sequences
and resultant phenotypes. J Bacteriol 172: 193-203.

Gray, T.R.G. (1976) Survival of vegetative microbes in soil. Symp Gen Mia'abiol 26:
327—364.

Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J
Mol Biol 166: 557-580.

Heitzer, A., Malachowsky, K., Thonnard, J.E., Bienkowski, P.R., White, D.C., and
Sayler, GS. (1994) Optical biosensor for environmental on-line monitoring of
naphthalene and salicylate bioavailability with an immobilized bioluminescent
catabolic reporter bacterium. Appl Environ Microbial 60: 1487-1494.

Holliday, R. (1956) A new method for the identification of biochemical mutants of
micro-organisms. Nature 178: 987.

Honeycutt, R.J., McClelland, M. and Sobral, B.W. (1993) Physical map of the genome
of Rhizobium meliloti 1021. J Bacterial 175: 6945-6952.

Jansson, J. (1995) Tracking genetically engineered microorganisms in nature. Curr
Opin Biotech 6: 275—283.

Kerppola, T.K. and Kahn, ML. (1988) Symbiotic phenotypes of auxotrophic
mutants of Rhizobium meliloti 104A14. J Gen Microbial 134: 913-919.

Kiss, G.B., Vincze, E., Kalman, Z., Forrai, T., and Kondorosi, A. (1979) Genetic and
biochemical analysis of mutants affected in nitrate reduction in Rhizobium
meliloti. J Gen Microbiol 113: 105—1 18.

S6

Kragelund, L., Christoffersen, B., Nybroe, O. and de Bruijn, P.J. (1995) Isolation of
lux Reporter Gene Fusions in Pseudamanas fluorescens DFS7 Inducible by
Nitrogen or Phosphorus Starvation. FEMS Microbial Ecol 17: 95-106.

Leigh, J.A. and Walker, G.C. (1994) Exopolysaccharides of Rhizobium: synthesis,
regulation and symbiotic funtion. Trends Genet 10: 63-67.

Lim, P.O., Ragatz, D., Renner, M., and de Bruijn, P.J. (1993) Environmental control
of gene expression: isolation of Rhizobium meliloti gene fusions induced by N-
and C— limitation. In Trends in Microbial Ecology. Guerrero, R., and PedroS-Alio,
C. (eds). Spanish Society for Microbiology, pp. 97-100.

Lin, J.T., Goldman, BS, and Stewart, V. (1993) Structures of genes nasA and nasB,
encoding assimilatory nitrate and nitrite reductases in Klebsiella pneumoniae
M5al. JBacterial 175: 2370-2378.

Matin, A. (1991) The molecular basis of carbon-starvation—induced general
resistance in Escherichia coli. Mol Microbiol 5: 3-10.

Meade, H.M., Long, S.R., Ruvkun, G.B., Brown, SE, and Ausubel, FM. (1982)
Physical and genetic characterization of symbiotic and auxotrophic mutants of

Rhizobium meliloti induced by transposon Tn5 mutagenesis. J Bacterial 149:
1 1 4- 1 2 2.

Meighen, EA. and Dunlap, P.V. (1993) Physiological, biochemical and genetic
control of bacterial bioluminescence. Adv Microb Phys 34: 1-67.

Moller, A., Gustafsson, K., Jansson, J.K. (1994) Specific monitoring by PCR
amplification and bioluminescence of firefly luciferase gene-tagged bacteria
added to environmental samples. FEMS Microbial Ecol 15: 193—206.

Miiller, P., Keller, M., Weng, W.M., Quandt, J., Arnold, W., and Puhler, A. (1993)
Genetic analysis of the Rhizobium meliloti eanFQ operon: ExoY is homologous
to sugar transferases and Eon represents a transmembrane protein. Mal
Plan t-Micrabe Interact 6: 5 5-65.

Ogawa , K., Akagawa, E., Yamane, K., Sun, Z.W., LaCelle, M., Zuber, P., and Nakano,
MM. (1995) The nasB operon and nasA gene are required for nitrate/nitrite
assimilation in Bacillus subtilis. J Bacterial 177: 1409-1413.

Omata, T., Andriesse, X., and Hirano, A. (1993) Identification and characterization
of a gene cluster involved in nitrate transport in the cyanobacterium
Synechocaccus Sp. PCC7942. Mal Gen Genet 236: 193-202.

Ozga, D.A., Lara, J.C., and Leig, J.A. (1994) The regulation of exopolysaccharide
production is important at two levels of nodule development in Rhizobium
meliloti. Mal Plant-Microbe Interact 7: 758—765.

Pain, A.N. (1979) Symbiotic properties of antibiotic-resistant and auxotrophic
mutants of Rhizobium leguminasarum. J Appl Bacterial 47: 53-64.

57

Peakman, T., Crouzet, J., Mayaux, J.E., Busby, S., Mohan, S., Harborne, N., Wootton,
J., Nicolson, R., and Cole, J. (1990) Nucleotide sequence, organisation and
structural analysis of the products of genes in the nirB-cysG region of the
Escherichia coli K-12 chromosome. EurJ Biochem 191: 315-323.

Pittard, J. and Wallace, B.J. (1966) Distribution and function of genes concerned
with aromatic biosynthesis in Escherichia coli. J Bacterial 88: 61 1-619.

Plunkett, G., Burland, V., Daniels, D.L., and Blattner, ER. (1993) Analysis of the
Escherichia coli genome. 111. DNA sequence of the region from 87.2 to 89.2
minutes. Nucleic Acids Res 21: 3391—3398.

Ragatz, D.M., Milcamps, A., Lim, P.O., Berger, K.A., and de Bruijn, F.J. (1997)
Isolation of carbon and nitrogen starvation-induced loci of Rhizobium meliloti
by Tn5-luxAB mutagenesis. Mal Microbial. submitted.

Rowbury, R]. (1983) Methionine biosynthesis and its regulation. In Amino Acids:
Biasynthesis and Genetic Regulation. Herrmann, KM. and Somerville, R.L. (eds).
Don Mills, Ontario: Addison-Wesley Publishing Company, pp.191-21 1.

Rovira, AD. (1956) A study of the development of the root surface rnicroflora
during the initial Stages of plant growth. J Appl Bacterial 19: 72-79.

Rovira, AD. (1959) Root excretions in relation to the rhizosphere effect. IV.
Influence of plant Species, age of plant, light, temperature, and calcium
nutrition on exudation. Plant Soil 1 1: 53—64.

Sambrook, J., Fritsch, ER, and Maniatis, T. (1989) Molecular cloning: a laboratory
manual, 2nd ed. New York: Cold Spring Harbor Press.

Sanger, F., Nicklen, S., and Coulson, AR. (1977) DNA sequencing with chain-
terminating inhibitors. Prac Natl Acad Sci USA 74: 5463-5 467.

Sharma, S. B. and Signer, E. R. (1990) Temporal and spatial regulation of the
symbiotic genes of Rhizobium meliloti in planta revealed by transposon Tn5-
gusA. Genes and Develop 4: 344-356.

Silhavy, T.J., Berman, M.L., and Enquist, L.W. (1984) Experiments with Gene
Fusions. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press.

Simon, R., Quandt, J., and Klipp, W. (1989) New derivatives of transposon Tn5
suitable for mobilization of replicons, generation of operon fusions and
induction of genes in gram—negative bacteria. Gene 80: 161-169.

Sutherland, I.W. (1979) Microbial exopolysaccharides: control of synthesis and
acylation. In: Microbial Palysaccharides and Palysaccharases. Berkeley, R.G.W.,
Gooday, G.W., and Ellwood, D.C. (eds). New York: Academic Press, pp. 1—2 8.

58

Szeto, W.W., Nixon, B.T., Ronson, G.W., and Ausubel, FM. (1987) Identification and
characterization of the Rhizobium meliloti ntrC gene: R. meliloti has separate
regulatory pathways for activation of nitrogen fixation genes in free-living and
symbiotic cells. J Bacteriol 169: 1423-1432.

Unthan, M., Klipp, W., and Schmid, G.H. (1996) Nucleotide sequence of the nar
beta gene encoding assimilatory nitrate reductase from the cyanobacterium
Oscillataria chalybea. Biachirn Biophys Acta, Gene Struct Expr 1305: 19-2 4.

Vancura, V. (1988) Plant metabolites in soil. In Developments in Agricultural and
Managed-Forest Ecology 1 7. Soil Microbial Associations: Control of Structures
and Functions. Vancura, V. and Kunc, F. (eds). New York: Elsevier, pp.57-144.

Wolk, P.C., Cai, Y., and Panoff., J.-M. (1991) Use of a transoposon with luciferase as
a reporter to identify environmentally responsive genes in a cyanobacterium.
Prac Nat Acad Sci USA 88: 5355-5359.

Yanofsky, C., Platt, T., Crawford, I.P., Nichols, B.P., Christie, G.E., Horowitz, H.,
vanCleemput, M. and Wu, AM. (1981) The complete nucleotide sequence of
the tryptophan operon of Escherichia coli. Nucleic Acids Res 9: 6647-6668.

Yanofsky, C. and vanCleemput, M. (1982) Nucleotide sequence of trpE of
Salmonella typhimurium and its homology with the corresponding sequence of
Escherichia coli. J Mol Biol 155: 235-2 46.

CHAPTER 3

Isolation and Characterization of Rhizobium meliloti Genes

Whose Promoters are Induced by Carbon Deprivation.

Parts of this chapter have been presented at the 6th International Microbial
Ecology Symposium (Iim et al., 1993) and the 10th International Nitrogen Fixation
Symposium (de Bruijn et al., 1995), and have been submitted for publication in
Molecular Microbiology (Ragatz et al., 1997).

59

60
ABSTRACT

Carbon availability is an important condition for life, but is severely limited
in most natural systems. Limitation is particularly pronounced in the soil, where
bacteria grow so Slowly they may only divide a few times per year. Consequently, it
has been of interest to understand how Rhizobium meliloti, a common soil isolate,
senses and responds to this nutrient deprivation condition. To investigate this
problem, the 5000 R. meliloti 1021 Tn5 insertion mutants that were screened for
nitrogen deprivation (see chapter 2) were utilized to screen for gene fusions
induced by carbon deprivation. The isolation of twelve gene fusions induced by
carbon deprivation is described. The strains harboring Tn5-quAB fusions were
tested for nodulation and nitrogen-fixation phenotypes. No strains were found to
be Fix: Cloning and partial DNA sequence analysis of the transposon tagged loci
revealed a variety of novel genes, as well as R. meliloti genes with sequence
Similarity to known bacterial loci. Genes with significant Similarity to ferredoxin

reductases, and to ribose transport/chemotaxis were found.

6 1
INTRODUCTION

As is the case for combined nitrogen, carbon is also severely limited in
natural environments. Organic nutrient concentration can range from between 0.4
to 10% in temperate soils (Atlas and Bartha, 1987). However, much of this carbon
is inaccessible to microbial populations, and the rest is quickly scavenged by
indigenous populations (Alexander, 1977; Atlas and Bartha, 1987). Thus the soil
environment, though richer in organic carbon than the marine environment,
remains largely oligotrophic.

Even in the rhizosphere, where carbon is relatively more abundant, the
higher numbers of microorganisms present necessitates the ability of rhizosphere
microorganisms to adapt to varying carbon conditions in order to compete
successfully (see Chapter 1 for an overview). Therefore, it was of interest to
understand how Rhizobium meliloti, a common soil and rhizosphere isolate, senses
and responds to carbon deprivation.

Many carbon deprivation-induced proteins have been observed to be
induced in bacteria ranging from E. coli to Vibrio (see chapter 1). It has been one of
the most studied of the environmental stresses. Some proteins have even been
identified; particularly those from E. coli, such as the heat-shock proteins DnaK,
GroEL, and Hth (Matin, 1991) and the product of the uspA gene, the universal
stress protein (Nystrom and Neidhardt, 1993; see Chapter 1 for examples in other
bacteria). However, the sensing of, and regulation in response to, limiting carbon is
still poorly understood, especially in common soil isolates.

Few investigations on nutrient deprivation-induced gene expression have

been carried out with common soil isolates, which may have different strategies for

62
dealing with nutrient deprivation than enteric or marine bacteria. This could be

due to differences in the types of carbon available (e.g., enteric vs. soil
environments) as well as to the total levels of organic carbon available, which is
higher for enterics and is available in regular cycles, and is lower in marine
environments. Only recently, the effect of nutrient deprivation on Pseudamanas
putida, a common gram negative soil bacterium, has been reported (Givskov et al.,
1994). Many proteins induced by carbon and nitrogen deprivation have been
detected using 2D-PAGE protein analysis. These proteins are temporally
regulated—as in the case of E. coli and Vibrio sp., and the nitrogen deprivation-
induced fusions described in chapter 2-and include a group of eight proteins
induced by multiple stresses. To date, only a few roles have been assigned to the
large collection of deprivation-induced proteins in any of the bacterial species
identified by 2D-PAGE analysis (Givskov et al., 1994; see Chapter 1).

Therefore, as described in chapter 2 for nitrogen deprivation-induced
fusions, the Tn5-1063 mutagenesis approach was taken to generate and
characterize a collection of 12 R. meliloti mutants carrying gene fusions induced by

carbon deprivation.

6 3
RESULTS

Isolation of Strains carrying Tn5-1063 gene fusions induced by carbon deprivation

A screening procedure, modified from the one described in Chapter 2, was
used to isolate carbon deprivation-induced genes, Since the bioluminescence of the
fusions was too weak to detect when cells were incubated without a carbon source
for several hours (see Materials and Methods). This is due to the fact that light

expression via luciferase requires FMNHz (Meighan, 1991), the generation of

which is closely associated with carbon metabolism. With the modified screening
procedure, we isolated 12 gene fusions whose bioluminescence was consistently
increased by carbon deprivation (Table 3-1).

Strains carrying fusions induced by carbon deprivation were tested for
growth and luciferase expression in the presence of a number of alternative carbon
sources (Table 3-1b). All of the strains were able to grow on the carbon sources
tested, in liquid and on solid media, with no discernable difference (data not
shown); however, growth rates were not determined. After incubation on filters
placed on media containing the selected carbon source for 48 hours, the induction
of the luxAB fusions they carried was examined under the photonic camera. None
of the gene fusions were induced when the bacteria that harbored them were
incubated only on the citric acid cycle derivatives succinate or fumarate (Table 3-
1b). One strain (C47) carried a fusion that was induced in the presence of malate

as sole carbon source. Many of the gene fusions were induced in the presence of C6

64

Table 3—1. Phenotypes of Tn 5-1063 insertion mutants of R. meliloti 1021 induced by
carbon deprivation.

 

Relative Increase in Lux Fusion Expression After Inoculation
in the Presence of:

 

. a myo—b Hybridizationc Gened
Shams 4°C succinate malate glucose sucrose fructose maltose trehalose inositol Fragments Similarity

_ _ — I NS
.ar.-—a...:...:~:.— -g.--.,_.-.x. .-...-,---- 1-< _ -,r ..‘.wna. 'V'Of- -

 

En‘C41v’1‘ +12 L .- '.' 1.31:. . . H...“ IT'S- — . --'.-~'_ ,. f“. L-t-i . ."I- - . .. 1‘. .L. if - . . . ". . .'.' L‘ 'i I; .‘ Cl.‘ .1.
Rm C18 - - - - + - + + —
+

E-fiI

Rm C22 _- _ - -__-_.-._.-_:_--_-__T_- . t . 1 _._--- .
l
1

 

 

Rm ( 35 _ _ _. + _. .. +
Eif-ats-i-I- - -.A'- n‘-'-'a‘- A” C... -'-'- - » . . .‘..'.'._... .1. ..'. ._I am - t'--.' -.-.‘.I.--'-'.' .‘u.\ .' .‘.‘.‘.'-'- ' .’-‘- .'-.'.‘« a {a - h -'-Z '-,t i - -- uh. -_rll.‘.h..*.-.k‘ ’.‘-‘.‘~’ . AB. .‘- .‘ .‘ ‘o'.'.‘-‘.’.'.'.‘.'—‘.‘.'-'-'-'- -
Rm C47 -
nm-‘nmvnnwee-a-ru-in:A\‘o rm'nvw rww win-ream - \‘b'fv‘v'p‘flv- . ---r'u-v'v-—- ,, . .-... ....-. ..._. - . . - 3. . n,- ...\- “a ., ,~,-.-- -,~;-.-,- 1;: ,o,- .y- -.--_~_‘ pp. 4.. . ,., )‘4 .- ,.
I‘ n ‘-"‘.‘-:—’J— ~‘O‘V -\'o‘v'§‘.'-‘Q‘.'g'n'OV.‘.\T-~'- - -‘_ >‘a’. u‘dh .T- .“n'- y’x’n‘. - a" . I.‘-'s - A ‘ ‘ a p Q n ‘ a-& e - m - u-A - - -..- . A --.e _‘ n ‘ ‘n - Q A A- A-b‘ A.-> - ‘.~ 5 I ‘1‘ I--_- u l.- I I.‘ A A \n‘a'a'a'—‘g'.‘«‘h‘—F-'-‘ 5. .‘

 

 

a) Strains were incubated at 4°C for 6 hours.

b) Expression of Tn 5-1063 fusions was determined by growing strains on minimal
media with 0.2% of the carbon source described. Number of plusses indicates
relative stren h of induction compared to induction without a carbon source
as determine by photon counts.

c) Number of bands of genomic DNA digested with EcoRI or ClaI hybridizing to
the pRL1063a probe.

d) Genes sharing significant Similarity with R. meliloti ORFS. n.d = not determined.
NS = no Significant similarity to known genes in GenBank.

65
carbon sources, including glucose which is not a preferred substrate for R. meliloti

(based on glucose versus succinate Chemotactic data, D.M. Ragatz and P.J. de
Bruijn, unpublished; see also Appleby, 198 4; Finan et al., 1988; McKay et al., 1988;
Day and Copeland, 1991). The gene fusion in strain C47 was particularly sensitive,
responding strongly to incubation on all of the carbon sources, except for succinate
and fumarate. The induction patterns varied quantitatively and temporally for
each strain tested, suggesting that the insertions were in distinct and/or distinctly
regulated loci. Strains C4, C22, C67, and C101 carried fusions that were not
induced in the presence of any of the carbon sources tested, and therefore seem to
be induced by a general carbon deprivation condition, or are too weakly induced to
be observed under the conditions used.

Cold- and heat-shock treatments were applied to the Tn5—1063 tagged
strains in order to determine if other stresses could activate the carbon
deprivation-induced fusions. None of the fusions were induced by incubation on
filters at 4°C for 6 hours (Table 3-1). However, the data obtained after heat-Shock
incubation were inconclusive, as the bacterial luciferase encoded by luXAB (derived
from Vibrio fischeri) proved to be exceedingly sensitive to higher temperatures,
becoming completely inactivated at 37°C within five to ten minutes of exposure
(data not shown).

In order to test the structural nature of the transposition events generating
the Tn5-1063 tagged Strains, total DNA was isolated, digested with EcoRI or ClaI
which do not cleave Tn5-1063 (Wolk et al., 1991), and hybridized with plasmid
pRL1063a DNA as a probe. Only one of the twelve strains (C55) harbored more

than one hybridizing fragment (Table 3-1c; Figure 3-1), indicating that a single,

265

276
323

313

SDE
ST
KRA

R
E
V

032

L
l

 

- M

N
Q -
R

 

A,

GIA
GIV
VLC'

N
R
A

CA

CAG
RC .
GA.

 

Q

 

 

 

lulRttcwx E00
VR svo -v00
RLES NA

GL
PIGM

 

66
VS‘PR
PGKA
P.

 

 

 

 

 

GD
GD

A
l
L

 

 

 

 

 

F I Y A
NIEIA
HVI—

S
1:
V

T
T

 

266
275
31‘
324

Rm
Rm
Rn

C18

Tth R-
C18

Tth R:
HopA Bc
C18 Rm
Tth R:

C18
Tth Re

HopA Be
HopA Be

A.

 

 

 

 

 

 

 

Rm
AraG Bo
RbsA 8:

C27
' RbaA EC

B

 

 

 

 

 

 

 

Km
RbaA Ec
AraG Ec
RbeA Ba

C27

 

 

 

 

 

Era]
1K
KE

KIIEK
MYAEA
LNY A
MKALA

K
L
RK

116
118
114

C77 Rm
RbeA Ec
AraG Ec
RbeA Be

 

 

 

(MopA) of Burhalderia cepacia is Shown. The solid bar above the sequence
indicates a conserved FAD-binding domain. The amino-acid residue coordinates

 

 

Similarity of the amino-acid sequence deduced from selected Tn5-1063 tagged
ORFS induced by carbon deprivation. In panel A, the amino-acid sequence
Similarity of the putative protein product of the Tn5-1063 tagged ORF of
mutant strain C18 to rhodocoxin (Tth) of Rhodacoccus sp., and a reductase
are given at the right of each line. In panel B, Similarity of the putative protein

product of the Tn5-1063 tagged ORF of mutant strain C27 to ribose transport
dehydrogenases from P. denitrificans (MoxF) and Methylaphilis W3a1(MedH)

protein product of the Tn5-1063 tagged ORF of mutant Strain C47 to PQQ-
dependant dehydrogenase protein from P. denitrificans (XoxF) and methanol
are shown.

proteins from E. coli (RbSA) and B. subtilus (RbSA), and to an E. coli arabinose
transport protein (AraG) are Shown. In panel C, Similarity of the putative

Rm

HoxF Pd
Hedfl HI
C47

XOxF Pd
Hoxr Pd
HedH He

C47
' XQXF Pd

Figure 3-1.

C

67

simple transposition event had occurred in the majority of the strains isolated (see
also chapter 2).

The Strains harboring gene fusions induced by carbon deprivation were
examined for their symbiotic phenotype by testing their ability to form effective
nodules on alfalfa seedlings. Nitrogenase activity of plant nodules was determined
by the acetylene reduction assay (see chapter 2, Materials and Methods). All of the
Tn5-1063 tagged strains were capable of forming nitrogen-fixation competent
nodules (Nod+ Fix") comparable to those induced by the wild-type R. meliloti 1021

(summarized in Table 3-1d).

Sequence Similarities of carbon deprivation-induced Tn5-1063 tagged R. meliloti
loci

The carbon deprivation-induced luxAB gene fusions were cloned from the
genome as described in chapter 2 (Figure 2-2), and a partial DNA sequence of the
tagged locus was determined. The amino-acid sequences deduced from the
primary DNA sequence were compared to sequences in the non-redundant protein
databases at GenBank. By analyzing the codon usage in the primary sequence (see
chapter 2, Figure 2-5) Significant ORFS were identified in all of the tagged loci (not
Shown). Three of the Tn5-1063 tagged loci were found to Share Significant
Similarity with known genes (Table 3-1c). In addition, one of the Tn5-1063 tagged
loci (C37) shared a degree of Similarity with an unknown ORF from E. coli K-12
(ORF_f375; GenBank Accession# U28377), which maps between approximately 65

and 68 minutes on the chromosome (Table 3-1c). DNA sequences derived from the

68
Tn5-1063 tagged loci in Strains C4 and C22 were found to overlap, indicating that

the transposons had inserted within the same locus. In addition, the locus tagged
by Strains C4 and C22 was found to be identical to the nitrogen deprivation-
induced locus tagged in strain N104 (see chapter 2), and the locus tagged in strain
C32 (not shown) was found to be identical to the nitrogen deprivation-induced
locus tagged in strain N4 (see chapter 2).

The partial DNA sequence of the Tn5—1063 tagged locus of strain C18 was
found to share a high degree of similarity with several ferredoxin reductase genes.
The highest similarity was observed with the rhodocoxin reductase gene tth from
Rhodacoccus sp. (Nagy et al., 1995), and the mapA reductase gene of Burkholderia
cepacia (Saint and Romas, 1996). The Similarity was particularly high in the
carboxy-terminal region due to the presence of a highly conserved FAD-binding
amino acid consensus sequence, TXXXX(I/V) (r/Y)A(A/V/I)GD (Figure 3-1A).
This sequence is uniquely characteristic of FAD-binding oxidoreductases, such as
those mentioned here (Eggink et al., 1990). Therefore, it is likely that the locus
tagged in strain C18 encodes an FAD-binding oxidoreductase, possibly involved in
energy scavenging during carbon deprivation.

The Tn5-1063 tagged locus of strain C27 was found to Share Significant
sirnilariy with the E. coli ribose transport gene rbsA (Bell et al., 1986), the Bacillus
subtilis ribose transport gene rbsA (GenBank Accession# 225798), and the
arabinose tranport gene araG (Scripture et al., 1987). The highest Similarity was
found to occur in the NH2 -terminal region (Figure 3-1B), which includes an ATP-
binding motif, as well as membrane-Spanning regions of the rbsA and araG gene

products.

69
The Tn5—1063 tagged locus of strain C47 was found to be surprisingly

similar to the predicted amino-acid sequence of xoxF, located in the Paracaccus
denitrificans operon encoding cytochrome c553i (RaS et al., 1991; Figure 3-1C).
This ORF corresponds to a gene induced by growth on choline as sole carbon
source, and has been predicted by Ras et al. (1991) to encode a quinoprotein
dehydrogenase, based on sequence Similarity with the product of the maxF genes
of P. denitriiicans (Harms et al., 1987) and Methylabacterium arganophilum
(Machlin and Hanson, 1988). In addition, a high degree of Similiarity with the
product of the methanol dehydrogenase medH gene of Methylaphilis W3a1 was
observed (MEDLINE identifier: 93054513; Figure 3-1C).

Although Significant ORFS were identified in each of the remaining carbon
deprivation-induced loci, no significant similarity of these loci to sequences in the
non-redundant protein databases at GenBank was found, suggesting that the

tagged genes are novel.

Characterization of the Chemotactic phenotype of Strain C27 harboring a gene

fusion with similarity to ribose transport genes

Strain C27, which carries a Tn5-1063 tagged locus with a high degree of
similarity to ribose transport genes, was tested for its Chemotactic response in a
diffusion gradient chamber containing a 0.15% agarose gel across which a 4mM
gradient of ribose was established (see Materials and Methods). Two diffusion
gradient chambers were set up identically with ribose gradients and inoculated
with 10 uL of late log cells (OD600=1.0), from Strains C27 or the wild-type, forming a

column in the center of the gels (appears as a single point when viewed from

 

 

 

L______.mmm _ -1 ,_.E

I Rm 1021
I

 

 

 

 

 

 

0 mM ribose
4 mM ribose

Figure 3-2. Chemotactic determination of strain C27 in a diffusion gradient
chamber (DGC). In panel A, R. meliloti 1021 is inoculated in a 0.15%
agarose gel to in which a 4mM gradient of ribose was established from
left (high) to right (low), and grown for 3 days. In panel B, a separate
DGC was inoculated with R. meliloti strain C27 carrying a Tn5—1063
tagged fusion with similarity to ribose transport genes, under the same
conditions as panel A. Arrows indicate the point of inoculation.

7 1
above; Figure 3-2). The cells were allowed to grow and move for three days, at

which point they were photographed. Strain C27 was able to grow with ribose as its
sole carbon source, but did not move up the ribose gradient (Figure 3-2B), unlike
the wild-type R. meliloti 1021 Strain (Figure 3-2A). This provides proof that the
Tn5-1063 insertion occurred in a gene necessary for ribose chemotaxis, probably
the ribose transport gene rbsA (see sequence comparison results presented above

and in Figure 3-1B).

7 2
DISCUSSION

Characterization of Tn5-1063 generated luxAB fusions induced by carbon

deprivation.

Stress induced genes and gene products have been studied mainly through
two dimensional gel electrophoresis in several bacteria (Spector and Cubitt, 1992;
Matin, 1991; Givskov et al., 1994). In this way, many proteins have been isolated,
induced by one or multiple deprivation conditions. For some of these proteins, the
correponding genes have been isolated. The use of reporter systems to isolate
genes induced by deprivation conditions, however, has only been applied to a few
bacterial systems including Salmonella typhimurium, resulting in the isolation of 7
lacZ gene fusions (Spector et al., 1988), Pseudomonas putida (Kragelund et al.,
1995), and Anabaena Sp. (Cai and Walk, 1997).

In this chapter, the use of luxAB as a reporter system in R. meliloti has
resulted in the isolation of 12 loci that appear to be involved in the carbon
deprivation response. The absence of a carbon source appears to cause the cell to
induce the production of a number of carbon-transport proteins in its cellular
membrane, making it more sensitive to carbon substrates in its environment, as
well as more efficient at carbon uptake.

The total number of carbon deprivation-induced loci identified appears to
be lower than was expected, given the large number of carbon deprivation—induced
proteins in E. coli and other systems. This may be due to several factors. First,
identifying carbon deprivation-induced luxAB fusions with this system was Slightly

awkward (due to FNMHZ limitation) and may have reduced the ability of

73
deprivation promoters to be detected at the same level as those induced by

nitrogen deprivation conditions. Second, it may be that the carbon deprivation-
induced loci are required for normal growth and persistance, and therefore Tn5
insertions would lead to lethal mutations. This possibility is increased due to the
lack of Tn5-1063 fusions with homology to E. coli (and other bacterial) ’core’
proteins induced by numerous stresses, such as the heat Shock proteins described
by Matin (1991). Third, the number of Strains tested (5000) is likely be too small to
adequately cover the genome (see chapter 7 for calculations). The occurrence of
Tn5-1063 insertions within the same loci in two independent screenings—one
induced by nitrogen deprivation (Tn5-1063 tagged loci in strains N5 and N9; see
chapter 2) and one induced by carbon deprivation (Tn5-1063 tagged loci in strains
C4 and C22)—Suggests that this is not the case. However, chapter 7 presents data
to support this third line of reasoning. Fourth, a large number of carbon-
deprivation—induced loci may be post-transcriptionally regulated, rendering our
approach unable to detect most of the carbon deprivation-induced loci. These
possibilities will be examined in more detail in Chapter 7.

Of course, it may be that the number of carbon deprivation-induced loci in
R. meliloti is Simply smaller than the number of proteins induced in other bacteria,
due to inherent differences in the kinds of proteins induced. This would support
the hypothesis that soil bacteria such as R. meliloti have a different means of
sensing and responding to environmental stresses than non-soil bactera. However,
sequencing of the Tn5-1063 tagged loci, as well as physiological characterization of
the strains carrying novel loci, needs to be completed before such conclusions can

be drawn.

74
In Spite of the possible limitations of the screening procedure used, the

isolation of a locus from strain C47 with striking similarity to an ORF known to be
induced by a poor carbon source (choline) is a good indication that we were
successful in isolating carbon deprivation-induced loci. Also, the locus tagged in
strain C27 is Similar to rbsA, a gene involved in ribose transport. This gene, in
addition to functioning as membrane-spanning carbon transport protein, is known
to have a dual role as the primary receptor for chemotaxis toward ribose in E. coli
(Bell et al., 1986). Examination of Chemotactic behavior in this Tn5-1063 tagged
strain (C27) demonstrated it to be non-Chemotactic for ribose. These data strongly
suggest that the conditions applied for carbon deprivation were sufficient, if

perhaps not completely adequate, to isolate carbon deprivation-induced loci.

7 5
MATERIALS AND METHODS

For bacterial strains and plasmids used, media and growth conditions,
transposon mutagenesis, DNA isolation and manipulation, DNA sequence analysis,
and nodulation and nitrogen-fixation assays see chapter 2 (Materials and

Methods).

Screening of the Tn5-1063 insertion mutants for carbon deprivation-induced gene

fusions

The R. meliloti strains carrying Tn5-1063 were Spotted on a membrane
filter (Nucleopore, Cambridge, MA) and incubated on solid GTS media for 48
hours. Luminescent colonies on the plate were visualized by Spreading 50 ul of N-
decanal inside the top of a glass petri—dish, placing the glass petri-dish top over the
bottom of the plate, exposing the cells to N-decanal for 60 seconds, and measuring
light emission using the Hamamatsu photonic system model C1966-20, as
described by Wolk et al., (1991).

Since the generation of FMNHZ, a cofactor required by luciferase, is closely
correlated with the carbon status of the cells, and the levels of FMNHZ appeared to
be a limiting factor for luciferase activity in carbon starved cells, a screening
procedure modified from that used in Chapter 2 was used. Colonies were
transferred on filters to both GTS and GT5 in which carbon sources had been
omitted. The membrane filters on GTS without carbon were then transferred to
regular GTS medium (containing succinate and glucose) and incubated for 30

minutes before visualization of luminescent colonies. This treatment was found to

76
provide sufficient energy for luciferase activity, while not interfering with the

isolation of carbon deprivation-induced gene fusions (see the Results and

Discussion sections above).

Testing of the Tn5-1063 tagged loci for induction by alternative carbon sources

The R. meliloti strains carrying Tn5-1063 were grown in 5 ml TY-Km200
media to saturation density (36 hrs), spotted on a membrane filter, and incubated
on solid GTS media, containing 10 mM of each alternative carbon source tested, for
48 hours at 28°C. Colonies on the plates were then examined for light induction
under the photonic camera as described above. It was not necessary to move the
colonies to fresh plates, since all of the Strains grew on all of the carbon sources

indicating that energy was available for the luciferase reaction.

Use of a diffusion gradient chamber (DGC) to test chemotards

The DGC was kindly provided by David Emerson (Michigan State University),
and set up according to the methods described in Emerson et al. (1994). The
chamber was filled with 0.15% agarose, and a gradient of ribose was established
ranging from 4mM to 0mM within the agarose chamber. A basal feed of GTS-Km100
media (without a carbon substrate) was flushed through the other two reservoirs to
ensure that no other nutrients were limiting or formed a gradient. The gradient
was allowed to form for two days, to ensure that it was complete, before inoculation

with bacteria.

77
R. meliloti strains were grown to saturation (36 hrs) in 5 ml TY—szoo, and a

10 ul portion of cells were inoculated by inserting the tip of a micropipette nearly
to the bottom of the agarose and smoothly releasing the cells as the tip was
removed, forming a Single column of cells in the center of the chamber. R. meliloti
strains in the agarose were allowed to grow for five days before the entire chamber

was visualized photographically.

78
REFERENCES

Alexander, M. (1977) In Introduction to Sail Microbiology, 2nd ed. New York: John
Wiley & Sons, pp. 16—51.

Appleby, CA. (1984) Leghemoglobin and Rhizobium respiration. Ann Rev Plant
Physiol 35: 443-478.

Atlas, RM, and Bartha, R. (1989) Microbial Ecology: Fundamentals and
Applications, 3rd ed. Reading, MA: Benjamin/Cummings Publishing Company,
INC, pp. 271-285.

Bell, A.W., Buckel, S.D., Groarke, J.M., Hope, J.N., Kingsley, DH, and Hermodson,
MA. (1986) The nucleotide sequences of the rbsD, rbsA, and rbsC genes of
Escherichia coli K12. J Biol Chem 261: 7652-7658.

Beringer, J.E. (1974) R factor transfer in Rhizobium leguminasarum. J Gen
Microbial 84: 188-198.

Boyer, H.W., and Roullaud-Dussoix, D. (1969) A complementation analysis of the
restriction and modification of DNA in Escherichia coli. J Mol Biol 41: 459-472.

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

Cai, Y. and Walk, GP. (1997) Nitrogen deprivation of Anabaena Sp. strain PCC7120
elicits rapid activation of a gene cluster that is essential for uptake and utiliztion
of nitrate. J Bacterial 179: 258—266.

Day, DA. and Copeland, L. (1991) Carbon metabolism and compartmentation in
nitrogen-fixing legume nodules. Plant Physiol Biochem 29: 185-201.

de Bruijn, F.J., Graham, L., Milcamps, A., and Ragatz, D. (1995) Use of the luciferase
(Tn5-lux) reporter system to Study Rhizobium meliloti genes responding to
N/C/O2 limitation or plant factors and their role in rhizosphere competition. In
Nitrogen Fixation: Fundamentals and Applications. Tikhonovich, I.A., Provorov,
N.A., Romanov, V.I., and Newton, W.E. (eds). Boston: Kluwer Academic
Publishers, pp. 195-200.

de Bruijn, F.J., Rossbach, S., Schneider, M., Ratet, P., Messmer, S., Szeto, W.W.,
Ausubel, F.M., and Schell, J. (1989) Rhizobium meliloti 1021 has three
differentially regulated loci involved in glutamine biosynthesis, none of which is
essential for symbiotic nitrogen fixation. J Bacterial 171: 1673-1682.

Devereux, J., Haeberli, P., and Smithies, O. (1984) A comprehensive set of
sequence analysis programs for the VAX. Nucleic Acids Res 12: 387—395.

79

Ditta, G., Stanfield, S., Corbin, D., and Helinski, DR. (1980) Broad host range DNA
cloning system for gram-negative bacteria: construction of a gene bank of
Rhizobium meliloti. Proc Natl Acad Sci USA 77: 7347-7351.

Eggink, G., Engel, H., Vriend, G., Terpstra, P., and Witholt, B. (1990) Rubredoxin
reductase of Pseudomaans aleovorans: Structural relationship to other
flavoprotein oxidoreductases based on one NAD and two FAD fingerprints. J
Mol Biol 212: 135-142.

Emerson, D., Worden, RM, and Breznak, J.A. (1994) A diffusion gradient chamber
for studying microbial behavior and separating microorganisms. Appl Env
Microbial 60: 1269-1 278.

Erickson, 8D. and Mondello, P.J. (1992) Nucleotide sequencing and transcriptional
mapping of the genes encoding biphenyl dioxygenase, a multicomponent
polychlorinated-biphenyl-degrading enzyme in Pseudamanas strain LB400. J
Bacterial 174: 2903—2912.

Finan, T.M., Orsnik, I., and Bottacin, A. (1988) Mutants of Rhizobium meliloti that
are defective in succinate metabolism. J Bacterial 170: 3396-3 403.

Givskov, M., Eberl, L., and Molin, S. (1994) Responses to nutrient starvation in
Pseudamanas putida KTZ 442: two-dimensional electrophoretic analysis of
starvation- and stress-induced proteins. J Bacterial 176: 4816—482 4.

Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J
Mol Biol 166: 557-580.

Harms, N., de Vries, G.B., Maurer, K., Hoogendijk, J. and Stouthamer, AH. (1987)
Isolation and nucleotide sequence of the methanol dehydrogenase structural
gene from Paracaccus denitrificans. J Bacteriol 169: 3969-3975.

Kiss, G.B., Vincze, E., Kalman, Z., Forrai, T., and Kondorosi, A. (1979) Genetic and
biochemical analysis of mutants affected in nitrate reduction in Rhizobium
meliloti. J Gen Microbial 1 13: 105—118.

Kragelund, L., Christoffersen, B., Nybroe, O. and de Bruijn, F.J. (1995) Isolation of
lux Reporter Gene Fusions in Pseudamanas fluorescens DF57 Inducible by
Nitrogen or Phosphorus Starvation. FEMS Microbial Ecol 17: 95-106.

Lim, P.O., Ragatz, D., Renner, M., and de Bruijn, P.J. (1993) Environmental control
of gene expression: isolation of Rhizobium meliloti gene fusions induced by N-
and C- limitation. In Trends in Microbial Ecology. Guerrero, R., and Pedros-Alio,
C. (eds). Spanish Society for Microbiology, pp. 97-100.

Machlin, SM. and Hanson, RS. (1988) Nucleotide sequence and transcriptional
start Site of the Methylabacterium organaphilum XX methanol dehydrogenase
structural gene. J Bacteriol 170: 4739—4747.

80

Matin, A. (1991) The molecular basis of carbon-starvation-induced general
resistance in Escherichia coli. Mal Microbial 5: 3—10.

McKay, I.A., Dilworth, J.J., and Glenn, AR, (1988) C4-dicarboxylate metabolism in
free-living and bacteroid forms of Rhizobium leguminasarum MNF3 481. J Gen
Microbial 134: 1433-1440.

Meade, H.M., Long, S.R., Ruvkun, G.B., Brown, SE, and Ausubel, RM. (1982)
Physical and genetic characterization of symbiotic and auxotrophic mutants of

Rhizobium meliloti induced by transposon Tn5 mutagenesis. J Bacterial 149:
114-122.

Meighen, EA. (1991) Molecular biology of bacterial bioluminescence. Microbial Rev
5 5: 1 2 3-1 42.

Meighen, EA. and Dunlap, P.V. (1993) Physiological, biochemical and genetic
control of bacterial bioluminescence. Adv Microb Phys 34: 1-67.

Nystrom, T., and Neidhardt, EC. (1993) Isolation and properties of a mutant of
Escherichia coli with an insertional inactivation of the uspA gene, which encodes
a universal stress protein. J Bacteriol 175: 3949-3956.

Ragatz, D.M., Milcamps, A., Lim, P.O., Berger, K.A., and de Bruijn, F.J. (1997)
Isolation of carbon and nitrogen Starvation-induced loci of Rhizobium meliloti
by Tn5-luxAB mutagenesis. Mal Microbial. submitted.

RaS, J., Reijnders, W.N., Van Spanning, R.J., Harms, N., Oltmann, LE. and
Stouthamer, A. H. (1991) Isolation, sequencing, and mutagenesis of the gene
encoding cytochrome c553i of Paracaccus denitrificans and characterization of
the mutant strain. J Bacteriol 173: 6971-6979.

Sambrook, J., Fritsch, BF, and Maniatis, T. (1989) Molecular cloning: a laboratory
manual, 2nd ed. New York: Cold Spring Harbor Press.

Scripture, J.B., Voelker, C., Miller, S., O’Donnel, R.T., Polgar, L., Rade, J.,
Horazdovsky, BE, and Hogg. R.W. (1987) High-affinity L—arabinose transport
operon: Nucleotide sequence and analysisi of gene products. J Mol Biol 197: 37-
46.

Silhavy, T.J., Berman, M.L., and Enquist, L.W. (1984) Experiments with Gene
Fusions. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press.

Spector, MR and Cubitt, CL. (1992) Starvation-inducible loci of Salmonella
typhimurium: regulation and roles in starvation-survival. Mal Microbial 6:
1467-1476.

Spector, M.P., Park, Y.K., Tirgari, S., Gonzalez, T. and Foster, J.W. (1988)
Identification and characterization of Starvation-regulated genetic loci in
Salmonella typhimurium by using Mu d-directed lacZ operon fusions. J Bacteriol
170: 345-351.

81

Taira, K., Hirose, J., Hayashida, S., and Furukawa, K. (1992) Analysis of bph operon
from the polychlorinated biphenyl-degrading Strain of Pseudamanas
pseudaalcaligenes KF707. J Biol Chem 267: 4844-4853.

Tan, H.M., Tang, H.Y., Joannou, C.L., Abdel-Wahab, N. H., and Mason, J.R. (1993)
The Pseudamanas putida MLZ plasmid-encoded genes for benzene dioxygenase
are unusual in codon usage and low in G+C content. Gene 130: 33-39.

Walk, P.C., Cai, Y., and Panoff., J.-M. (1991) Use of a transoposon with luciferase as
a reporter to identify environmentally responsive genes in a cyanobacterium.
Prac Nat Acad Sci USA 88: 5355—5359.

Zylstra, G.J. and Gibson, D.T. (1989) Toluene degradation by Pseudamanas putida
F1: Nucleotide sequence of the tadCl CZBADE genes and their expression in
Escherichia coli. J Biol Chem 264: 14940-14946.

CHAPTER 4

Isolation and Characterization of Rhizobium meliloti Genes

Whose Promoters are Induced by Oxygen Limitation.

82

83
ABSTRACT

Oxygen levels in the soil environment vary drastically from microsite to
microsite, as well as temporally due to changing water conditions within soil pores,
and range from atmospheric levels to anaerobic conditions (van Elsas and van
Overbeek, 1993). The ability to sense and respond to changing oxygen levels is one
of the most important abilities of a soil microorganism, such as Rhizobium meliloti,
which is also exposed to microaerobic (microoxic) conditions as an endosymbiont
within the nodule of its host plant, alfalfa. Consequently, it has been of interest to
understand how R. meliloti senses and responds to this microoxic condition. To
investigate this question, the 5000 R. meliloti 1021 Tn5-lax insertion mutants that
were screened for nitrogen and carbon deprivation (see chapters 2 and 3) were
utilized to screen for gene fusions induced by microaerobiosis. The isolation of
thirty-four gene fusions induced by microaerobiosis (1% 02) is described. The
strains harboring Tn5-luxAB fusions induced by microaerobiosis were tested for
nodulation and nitrogen-fixation phenotypes. None were found to be Fix: Cloning
and partial DNA sequence analysis of the transposon tagged lad revealed a variety
of novel genes, as well as R. meliloti genes with sequence Similarity to known
bacterial loci. Genes identical to already described genes for exopolysaccharide
synthesis (exoO), and for nitrogen fixation (fixN) were found, as well as genes with
Significant similarity to aldehyde dehydrogenases, amino acid transport proteins,

and cytochrome oxidases.

8 4
INTRODUCTION

Rhizobia are non-fermenting aerobic bacteria which primarily use oxygen as
their terminal electron acceptor (Long, 1989; Hirsch, 1992; Batut and Boistard,
1994). However, they are exposed to microoxic conditions during two distinct
phases of their life-cycle, which comprise two distinct habitats: as free-living soil
bacteria and as nitrogen-fixing plant endosymbionts. Both niches present a
metabolic challenge to bacteria, in addition to other stresses they encounter in the
environment such as the nutrient deprivation conditions described earlier
(chapters 2 and 3). Thus, rhizobia have had to evolve mechanisms to sense and

respond to changing oxygen conditions both in the soil and nodule environments.

Oxygen environment in the soil

As was Stated in Chapter 1, the level of oxygen in the soil environment can
change Significantly over time with changing weather patterns, resulting in
gradients of oxygen in the water films surrounding soil pores (Smiles, 1988;
Metting, 1993). A microorganism in this environment can experience conditions
ranging from standard atmospheric oxygen concentrations to virtually anaerobic
conditions within a day (Foth, 1984). In addition, oxygen levels in the rhizosphere
can be even lower than in the bulk soil, due to respiration by plant roots and
microbial communities in the rhizosphere (Alexander, 1977). It is no surprise,
then, that adaptation to changing oxygen conditions is one of the most important

challenges of a soil microorganism, such as R. meliloti.

85

Oxygen environment in the nodule

The symbiotic association between rhizobia and their host plants results in a
Specialized ecological niche, the nodule (Figure 4-1). This Specialized plant
Structure provides the physiological conditions required for the energy intensive
nitrogen-fixation process (including high levels of bacterial respiration), while
maintaining an intracellular free oxygen concentration less than 10nM, so that the
oxygen sensitive nitrogenase enzyme is protected from irreversible oxygen damage
(Batut and Boistard, 1994; Fischer, 1994; Kim and Rees, 1944).

The rhizobial nitrogen fixation genes are regulated in response to limited
oxygen concentration via the two-component FixL/FixJ system (Fischer, 1994;
Figure 4—2). FixL is a membrane-bound, oxygen-regulated, hemoprotein kinase and
phosphatase that binds oxygen, resulting in an alteration of the Fe oxidation State,
which in turn results in the autophosphorylation of FixL (Gilles-Gonzalez et al.,
1991), which is then able to phosphorylate FixJ (de Philip et al., 1992; Fischer,
1994). Once phosphorylated, FixJ is able to act as a trascriptional activator in the
induction of nifA and fer (David et al., 1988; de Philip et al., 1990), which in turn
regulate all of the nif and fix genes in R. meliloti. Kahn and Ditta (1991) have
proposed that a conformational change occurs in FixJ when phosphorylated such
that previously masked transcriptional activation sites in the C—terminus become
unmasked. In vitra evidence by Da Re et al. (1994) supports this hypothesis.

Figure 4-2 illustrates the FixL/FixJ regulatory cascades (as well as the
NtrB/NtrC regulatory cascade; see chapter 2) in species representative of the three

rhizobial genera: Rhizobium, Azarhizabium, and Bradyrhizobium.

86

Zone III
1

I l

Nitrogen-fixing
Zone IV
senescent zone

Ineliicient
zone Interzone Il—III

 
  

Zone I
meristem

 

 

P 1' ‘ Invasion
re Ixmg

Vascular zone zone
bundle Nodule Nodule

endodermis cortex

 

Nodule parenchyma Zone ll

Figure 4-1. Diagram of an indeterminate nodule of alfalfa. The four zones of
development are labeled, as well as various tissues. The characteristic
meristem delineates the region where new cells are being added,
causing the nodule to elongate. The central shaded region indicates the
zones containing rhizobia. (Hirsch, 1992) For further details, see text.

A. Rhizobium meliloti

 

 

 

an U.“
j 9
ea L-
I no.
I iv
0 lit 9 in
9 9 G
9e. H“
g ‘Mr 1
It@
z:
a :3 a 0V T:

 

 

 

 

~0— fflm in?“
._- if I
.‘ r.
e u -m.
I
-———in
M- 9m
9? I
-_I L@
"
I... I -

Figure 4-2. Comparative models of nif and m gene regulation. Homologous
regulatory proteins are symbolized identically in panels A, B, and C.
Solid arrows and open arrowheads with dashed lines indicate positive
and negative regulation, respectively. Dotted lines with open arrow-
heads denote NtrC-mediated activation of the R. meliloti niHDKE,
nithde, and fixABOf promoters which is only observed in free-living
cells, and is not relevant for symbiotic nitrogen fixation (from Fischer,

 

 

 

 

 

1994)., See text for further details.

 

. Bradyrhizobium japanicum
I
Gt
.., I ‘
0 H 0 man
“T": I T:
E] “3
I I,”
9 fit. -.-—'-
I l E...
5‘9 .33" a"

88
Note that, although the niffix genes are regulated differently in the three systems

(primarily due to different NifA regulatory pathways), microaerobically induced
gene expression occurs through the same FixL/FixJ two-component mechanism in
all three genera.

FixK in rhizobia is part of a protein family of transcriptional regulators in
bacteria, called the FNR family, which are responsible for many oxygen-regulated
genes (Spiro, 1994; Fischer, 1994). The role of this transcriptional activator and
repressor varies among bacterial genera. In most bacteria, regulation falls into one
of two categories: regulation of the expression of genes involved in anaerobic
metabolism (Spiro and Guest, 1991; Saffarini and Nealson, 1993; Zimmermann et
al., 1991; Cuypers and Zumft, 1993; Dispensa et al., 1992), or regulation of genes
necessary for nitrogen-fixation (Batut et al., 1989; Colonna—Romano et al., 1990;
Anthamatten et al., 1992; Kaminski et al., 1991). However, FNR homologues have
been described in other bacteria which are responsible for regulating other
processes, such as bioluminescence (Meighen and Dunlap, 1993; Spiro, 1994) and
cyanide synthesis (SawerS, 1991), and there are additional FNR homologues with
unknown functions (Upadhyaya et al., 1992; Irvine and Guest, 1993).

The symbiotic pattern of expression of Nz-fixation genes is tightly coupled to
the physiology of the nodule, and expression has been Shown to be induced at a
Sharply defined region of the nodule called interzone II-III (Soupéne et al., 1995;
Figure 4-1). Figure 4-1 illustrates the different regions of an indeterminate nodule,
such as those found on alfalfa. Note that bacteroids are present (indicated by
shading) only inside of and within interzone II-III (i.e., in zones II-III, III, and IV).
Evidence by Soupene et al. (1995) has verified the hypothesis that nitrogen

fixation genes are regulated by oxygen according to a Spatial program that is

89
related to the histological organization of the mature nodules. They found a steep

decline in 02 concentration approaching interzone II-III, attributed to the presence
of an O; diffusion barrier. It is not known, however, whether interzone II-III itself
restricts OZ diffusion, or what is the nature of the diffusion barrier.

To date, all microaerobically induced genes in rhizobia have been Shown to
be regulated via the FixL/FixJ system. However, as was Stated in chapter 1,
evidence from Zhulin et al. (1995) demonstrates that Chemotactic behavior of
Rhizobium meliloti in oxygen gradients is not regulated by the FixL/FixJ system.

Because of the success isolating nutrient deprivation-induced loci via the
Tn5-luxAB method, I undertook the same approach to isolate gene fusions induced
by nricroaerobic conditions. In this chapter, the generation and characterization of
a collection of 34 R. meliloti mutants generated by random Tn5-1063 mutagenesis,
carrying gene fusions induced by 1% oxygen, is described. Cloning and partial DNA
sequence analysis of the transposon tagged loci revealed a variety of potentially
novel genes, as well as R. meliloti genes with sequence Similarity to known bacterial
loci. Genes identical to already described genes for exopolysaccharide synthesis
(exoO), and for nitrogen fixation (fixN) were found, as well as genes with significant

Similarity to dehydrogenases, carbon transport proteins, and cytochrome oxidases.

9 0
RESULTS

Isolation of strains carrying Tn5-1063 gene fusions induced by microaerobiosis

The screening procedure used for isolating microaerobically induced genes
was modified, Since it was unnecessary to move the filters bearing the colonies to
new plates in order to “starve” them for oxygen. Preliminary experiments,
however, made it clear that the results were more consistent and clear (Stronger
light emission) when the bacteria were plated on nucleopore filters, rather than
directly on the media. A 1% oxygen mixture was chosen because it is as low as one
can go before R. meliloti ceases to grow. Moreover, this concentration of oxygen has
been Shown to induce the FixL/FixJ regulated nitrogen fixation genes in rhizobia
(Ditta et al., 1987; Gilles-Gonzalez et al., 1991; Reyrat et al., 1993; Page and
Geurinot, 1995). Using 1% oxygen induction conditions, 34 Tn5-1063 generated
luxAB gene fusions were isolated whose bioluminescence was consistently increased
at 1% 02 (Table 4-1).

In order to test the nature of the transposition events involved in generating
these Strains, total DNA was isolated, digested with EcoRI or ClaI (which do not cut
within the transposon; Wolk et al., 1991), and hybridized with plasmid pRL1063a
DNA as a probe. Only two of the thirty-four mutants displayed multiple bands
(Table 4-1a), indicating that a Single, Simple transpositional event occurred in the
majority of the Strains, as observed with strains carrying fusions induced by
nitrogen and carbon deprivation (chapters 2 and 3). The two strains with multiple
hybridizing fragments were not further examined except as part of the nodulation

and nitrogen fixation assays (Table 4-1b).

91

Table 4-1. Phenotypes of Tn5-1063 insertion mutants of R. meliloti 1021

induced by microaerobiolsis.

 

Symbiotic Phenotype a

 

Hybridizingb

Strain Nod Fix Fragments

Gene C
Similarity

 

one

0?“ /C101_..__...

0 o'vwdh-y -‘..

0x3 .. ,, one
9x4
0256,, .. ..

one

one

+ +‘.+:+ '+'+‘:+§_i+;+}+i+"+ §+§+ ‘+f?+?.+;+ ,+ + _+_§+ i+ +f+_i+i.+ +;+:;+j+,+f+'+

 

;+§+?+ .+,+i+'+;j-1-j+ ,'+:+ +£+ +9+ig+g+ q+ +++ + +’3+"+:+ +;+;+ ~‘+;+§+:j+
"
:r
S

666:; _
.one. .;

one

....................................

NS

, , NS?” 5, -._-_a -13.. .~.,._..--. ~. “rs-.9 '

' acaD, aldehyde dehydrogenase

' ’ .FtZtN nitrogen fixation

braF, amino acid transport

szN nitrogen fixation
NS _ g
.' N3,; 1.

n.d.

~_.,. , _. '.\ ‘-_---A._.‘.«_.~ _~_._... .1 _.‘-.-¢...... .. ,.. ,..-t....
'. ._ ITI'IS_.._.>....I....... A "'

NS

' (NS ”

s'

NS

exofiisucmnoglycansynthesm
NS

.2, N5.
n.d.

-------

NS

. N513" .

NSW

' if ’NSW

ORF6 in cbbR 3' end i. _ ,

NS ,., ..

NS

NS” ' ' ’
- - - I). '- , , 0 _.‘.‘.,'.l o‘.'_':'.' ' '2 - I r s I

NS

" 7 7 .1 57.ICyp'CLifytgshrdmééXidaSe]
NS

’ NS 7:]: .;

NS

.i'ffj'NSfinff.[....fl ___i;.;;.. jg.
NS

 

a) Symbiotic phenotype of alfalfa roots, inoculated with R. meliloti strains carrying
Tn5-1063 tag gged genes induced by microaerobiosis, for the presence of nodules

on roots (No
reduction (Fix+).

+) and the presence of nitrogenase as determined by acetylene

b) Number of bands of genomic DNA cut with EcoRI or ClaI hybridizing to a probe of

whole pRL1063a.

c) Genes sharing significant similarity with R. meliloti ORFS: acoD (Priefert et al., 1992);
fixN (Preisig etal.,1993); braF (Hoshino 8: Kose, 1990); exoO (Becker et al., 1993);
cyoC (Chepuri etal.,1990);NS=no significant Similarity to known genes;

n. d. = not determined.

92
The strains harboring gene fusions induced by 1% oxygen levels were

examined for their symbiotic phenotype by testing their ability to form effective
nodules on alfalfa seedlings. Nodule nitrogenase activity was measured by the
acetylene reduction assay (see chapter 2, Materials and Methods). All of the strains
were capable of forming nitrogen—fixation competent nodules (Nod+ Fix”)
comparable to those induced by wild-type R. meliloti 1021 (summarized in Table 4-

1b).

Sequence similarities of microaerobically induced Tn5-1063 tagged R. meliloti loci

The microaerobically induced luxAB fusions were cloned from the genome,
and partial DNA sequences of the Tn5-1063 tagged loci were determined. The
amino-acid sequences deduced from the primary DNA sequence were compared to
sequences in the non-redundant protein databases at GenBank. By analyzing the
codon usage in the primary sequence (see chapter 2, Figure 2-4) significant ORFS
were identified in all the tagged loci (data not shown), and seven of the 34 loci
analyzed were found to share significant similarity with known genes (Table 4-1b).
However, the majority of the loci (27 out of 34) appear to be novel since they did
not share significant similarity with GenEMBL sequences (as of 8/ 1/ 97).

The fusion in strain 0x1 was found to be the same as the previously
identified carbon deprivation-induced fusion in strain C101 (chapter 3). Testing
strain 0x1 for carbon deprivation induction revealed that it was indeed induced by
both microaerobiosis and carbon deprivation. In addition, strain 0x221 was found
to carry a fusion that had been identified both in the nitrogen deprivation-induced,

and in the carbon deprivation-induced set of strains (strains N104 and C4/C22,

93
chapters 2 and 3). This interesting strain was thus the only one found to carry a

luxAB fusion induced by all three stresses tested (see Discussion below). Sequence
data from the Tn5-1063 tagged loci of strains 0x4 and 0x19 were identical, as
were sequence data from the Tn5-1063 tagged loci of strains 0x81 and 0x86, thus
indicating that these two sets of two strains each are clonal. Two of the Tn5-1063
tagged loci (in strains 0x4 and 0x106) correspond to already seqenced genes of R.
meliloti (fiXN and exoO, see Discussion below).

The Tn5-1063 tagged locus in strain 0x3 was found to be most similar (P(N)
= 2.0 x 10‘”) to the aldehyde dehydrogenase gene acoD from Alcaligenes
eutrophus (Priefert et al., 1992), as well as to aldehyde dehydrogenases from
Rhodococcus sp. (thcA; Nagy et al., 1995), E. coli (aldB; Sofia, et al., 1994; Xu and
Johnson, 1995), and Vibrio cholerae (aldA; Parsot and Mekalanos, 1991)

The Tn5-1063 tagged locus in strain 0x4 (as well as in the identical strain
0x19) was found to match the fixN gene, which has been completely sequenced
(Kahn et al., 1993), and encodes a component of an alternative cytochrome-c
containing heme/ copper cytochrome oxidase required for nitrogen-fixation in
rhizobia (Preisig et al., 1993). There are two copies of this gene in R. meliloti, such
that a single mutation does not result in a Fix‘ phenotype. This observation was
confirmed in this study (Table 4-1b).

The Tn5-1063 tagged locus in strain 0x6 was found to be most similar
(P(N)=4.2 x 10'”) to the Pseudamanas aeruginasa braF gene encoding a branched-
chain amino-acid transport protein (Hoshino and Kose, 1990), as well as to amino-
acid transport proteins in E. coli (IivG; Adams et al., 1990; Sofia et al., 1994), and

Salmonella typhimurium (livG; Matsubara et al., 1992). The highest region of

94
similarity occurred in an ATP-binding motif common to active-transport proteins

(Hoshino and Kose, 1990).

The Tn5-1063 tagged locus in strain 0x106 was found to match the R.
meliloti exoO gene, which is part of an operon (eonMONP) involved in
succinoglycan (exopolysaccharide) biosynthesis (Glucksmann et al., 1993; Becker et
al., 1993). Interestingly, the Tn5-1063 tagged insertion mutant in strain 0x106
displayed a Nod‘ Fix” phenotype (Table 4-1b), in contrast to other exo mutants of
R. meliloti, which result in an empty nodule (Nod+ FiX’) phenotype (Finan et al.,
1985; Leigh et al., 1987; Gray and Rolfe, 1990; Hirsch, 1992).

The Tn5-1063 tagged locus in strain 0x193 was found to be highly similar
(P(N)=8.S x 10‘”) to an unknown 0RF located 3’ to CbbR, a LysR-type
transcriptional activator in Xanthobacter flavus (van den Bergh et al., 1993). This
protein is required for expression of the autotrophic C02-fixation enzymes of X.
flavus.

The Tn5—1063 tagged locus in strain 0x219 was found to have an extremely
high similarities to cytochrome quinol oxidases from Paracaccus denitrificans
(cytochrome ba(3) III, P(N)=2.2 x 10"“; Richter et al., 1994), E. coli (cyoC, P(N)=1.8
x 10'“; Chepuri et al., 1990), and Acetobacter aceti (cytochrome a1 chain 111,
P(N)=3.9 x 10'“; Fukaya et al., 1993). Therefore, it seems likely that the Tn5-1063

tagged locus in strain 0x219 encodes a cytochrome quinol oxidase.

9 5
DISCUSSION

Characterization of Tn5-1063 generated luxAB fusions induced by microaerobiosis

Because of the success in isolating nitrogen deprivation-induced loci and
carbon deprivation-induced loci in R. meliloti using Tn5-1063 tagged mutants, a
similar approach was used to isolate microaerobically induced loci. In this chapter,
the use of luxAB as a reporter system in R. meliloti resulted in the isolation of 34
loci involved in microaerobiosis, including genes involved in nitrogen fixation (fixN)
and succinoglycan synthesis (exoO), as well genes with similarity to
dehydrogenases, amino acid transport proteins, and cytochrome oxidases. The
significance of the similarity of tagged loci to other genes is discussed below.

The number of genes isolated by microaerobiosis was higher than in the
other two deprivation conditions tested (chapters 2 and 3). This may indicate that a
larger number of genes is responsive to this testing condition, or it may be that the
potential problems of using a transposon to tag genes of interest (discussed in
chapters 2 and 3) do not apply to genes regulated by low oxygen concentration. In
addition, the large number of novel genes tagged (2?), similarly to the large
number of novel genes described in chapters two and three, support the
hypothesis that soil bacteria may have different genes than non-soil bacteria for
dealing with environmental limitation conditions. As with the tagged loci described
in chapters two and three, more work needs to be done characterizing the novel
loci and their regulation, as well as complete sequencing of the tagged genes,
before this hypothesis can be fully examined (see chapter 7 for further discussion

of this important hypothesis).

96
The similarity of the locus in strain 0x3 to aldehyde dehydrogenases may

indicate that a gene encoding an enzyme in a redox pathway, that is induced due to
microaerobiolsis, has been tagged. The aldehyde dehydrogenases to which this
locus has similarity are located on the chromosome near cytochrome oxidases and
are believed to participate in the metabolism of various substrates (Nagy et al.,
1995). Therefore, this gene may function in cooperation with other redox enzymes
and cofactors, in conjunction with alternative cytochrome oxidases, to provide the
cell with the reducing equivalents it needs during microaerobiosis. This is
particularly important while in the nodule, where the high ATP demands of
nitrogen-fixation make it necessary to increase respiration rates (Fischer, 1994).
Interestingly, expression of the V. cholerae gene (aldA) is regulated by ToxR, a
transcriptional activator which regulates several virulence genes (Parsot and
Mekalanos, 1991). This is reminiscent of the transcriptional activator FixK in R.
meliloti which regulates several nitrogen fixation genes (Figure 4—2).

The microaerobic induction of the putative amino acid transport protein
gene tagged in strain 0x4 is less simple to explain. Plant roots are known to exude
many amino acids. Therefore, a microaerobically induced transport system may be
involved in sensing the lower oxygen potential near the root and inducing
transport genes which would help R. meliloti to compete for the available limited
resources. Another possibility, suggested by the presence of an ATP-binding motif,
is that the gene tagged in strain 0x4 is simply an active-transport protein of an
unknown substrate. Best matches of the tagged loci to amino acid transporters may
be coincidental, since these proteins belong to a family of proteins, called ABC

transporters, which all bind ATP in order to actively transport a variety of

97

substrates. Complete sequencing of this locus and further comparison to GenBank
sequences should resolve this question.

The interesting finding that the R. meliloti exoO gene is regulated by
microaerobiosis constitutes the first report of any exopolysaccharide gene in R.
meliloti being regulated by microaerobiosis. As mentioned in chapter 2, it is known
that exopolysaccharide synthesis is decreased in the presence of ammonia (Ozga et
al., 1994), although it is not known how this regulation occurs. In addition, two of
the exo genes characterized in this study (exoYP, see chapter 2) were isolated by
nitrogen deprivation induction of their tagged genes, thus confirming the
regulation of this operon in response to available nitrogen. Figure 4-3 illustrates
the complete exo regulon, with the eon and exoO genes, tagged in this study,
indicated. R. meliloti has apparently evolved two divergent operons responding to
two environmental conditions it would find in the rhizosphere; microaerobiosis and
nitrogen deprivation. This complicated regulatory cascade could help the
bacterium compete in the soil by preventing the induction of unnecessary (and
perhaps energy intensive) genes, while at the same time providing a mechanism to
sense the presence of rhizosphere conditions leading to successful nodulation.
Figure 4-3 also illustrates why an insertion in exoO (or any of the genes in this
operon) is not FiX' due to a duplication of the eonKLAMONP operon, unlike the
FiX' phenotype of the eon mutants isolated in chapter 2.

The similarity of the locus in strain 0x193 to an 0RF downstream from CbbR
in X. flavus presents some interesting possibilities. CbbR is a LysR—type
transcriptional activator that is required in X. flavus for the expression of
autotrophic C02 fixation enzymes (van den Bergh et al., 1993), suggesting that the

Tn5-1063 tagged locus in strain 0x193 might be in an operon regulated by a

98

 

 

 

 

0x106 N112 N149
<7 I ‘- 4——-> +( +4—
- r — l
P N O M A L K H I T W V U X Y F 0 2 B
2 Kb
Duplicated Operon I——-l

Figure 4-3. Map of the exo gene cluster on the second symbiotic megaplasmid of
R. meliloti Rm1021. Boxes represent the open reading frames of the
exo genes, and the arrows indicate the orientation of transcription. A
duplicated region is indicated by a hatched box (adapted from
Glucksmann et al., 1993). For additional details, see the text.
transcriptional activator encoded nearby. Further sequence analysis of this tagged
fusion, both upstream and downstream of the insertion site, would determine if
this assertion is true.

Isolation of a known oxygen—regulated gene (fsz), as well as another
putative cytochrome oxidase (in the Tn5—1063 tagged locus of strain 0x219),
suggests that the conditions for oxygen limitation chosen were both adequate and
sufficient for the identification of microaerobically induced loci. It was hypothesized
that changes in oxygen levels would require the bacteria to respond by altering its
ability to bind oxygen via induction of alternative cytochrome oxidases. Therefore,
isolation and characterization of a tagged gene in R. meliloti with very high
similarity to cytochrome oxidases was expected. What is, perhaps, unexpected is

that more cytochrome oxidases were not found. This may be a true reflection of the

induction of cytochrome oxidases by oxygen limitation, or may be related to

99
problems discussed in chapters 2 and 3 regarding the nature of the Tn5-1063

system. In addition, screening of 5000 Tn5-1063 insertion mutants is probably not

sufficient to saturate the genome (see chapter 7 for calculations).

1 00
MATERIALS AND METHODS

For bacterial strains and plasmids used, media and growth conditions,
transposon mutagenesis, DNA isolation and manipulation, DNA sequence analysis,
and nodulation and nitrogen-fixation assays see chapter 2 (Materials and

Methods).

Screening of the Tn5-1063 insertion mutants for microaerobically induced gene

fusions

The R. meliloti strains carrying Tn5-1063 were spotted on a membrane
filter (Nucleopore) and incubated on solid GTS minimal media for 36 hours at 28°C.
The plates were then placed in a sealed oxygen chamber, that was continuously
flushed with a 1% (i 0.1) oxygen-mixture for seven hours, in a growth chamber set
to a temperature of 28°C. The concentration of oxygen was monitored using an
oxygen electrode. At the end of the incubation period, the plates were removed
from the oxygen chamber and briefly exposed to standard oxygen concentrations
before the period of luminescence measurement in order to provide substrate (02)
for the luciferase enzyme.

Luminescent colonies on the plate were visualized by spreading 50 u] of N—
decanal inside the top of a glass petri-dish, placing the glass petri-dish top over the
plate, exposing the cells to N—decanal for 60 seconds, and measuring light emission
using the Hamamatsu photonic system model C1966—20, as described by Wolk et

al. (1991).

101
REFERENCES

Adams, M.D., Wagner, L.M., Graddis, T.J., Landick, R., Antonucci, T.K., Gibson, A.L.,
and 0xender, D.L. (1990) Nucleotide sequence and genetic characterization
reveal six essential genes for the LIV-I and LS transport systems of Escherichia
coli. JBiol Chem 265: 11436—11443.

Alexander, M. (1977) In Introduction to 501'] Microbiology, 2nd ed. New York: John
Wiley & Sons, pp. 16—51.

Anthamatten, D., Scherb, B., and Hennecke, H. (1992) Characterization of a fixL] -
regulated Bradyrhizobium japonicum gene sharing similarity with the
Escherichia coli fnr and Rhizobium meliloti fixK genes. J Bacteriol 174: 21 1 1—
2120.

Batut, J., and Boistard, P. (1994) Oxygen Control in Rhizobium. Antonie van
Leeuwenhoek 66:1 29-1 50.

Batut, J., Daveran-Mingot, M.L., David, M., Jacobs, J., Gamerone, A.M., and Kahn,
D. (1989) fixK, a gene homologous with fnr and cm from Escherichia coli,
regulates nitrogen fixation genes both positively and negatively in Rhizobium
meliloti. EMBO J 8: 1279—1286.

Becker, A., Kleickmann, A., Keller, M., Arnold, W., and Puhler, A. (1993)
Identification and analysis of the Rhizobium meliloti eonMONP genes involved
in exopolysaccharide biosynthesis and mapping of promoters located on the
eonKLAMONP fragment. Mol Gen Genet 241: 367—379.

Chepuri, V., Lemieux, L., Au, D.C., and Gennis, RB. (1990) The sequence of the cyo
operon indicates substantial structural similarities between the cytochrome o
ubiquinol oxidase of Escherichia coli and the aa3-type family of cytochrome c
oxidases. JBiol Chem 265: 11185-11192.

Colonna—Romano, S., Arnold, W., Schliiter, A., Boistard, P., and Priefer, U.B. (1990)
An fnr-like protein encoded in Rhizobium leguminasarum biovar viciae shows
structural and functional homology to Rhizobium meliloti FixK. Mol Gen Genet
223: 138—147.

Cuypers, H., and Zumft, W.G. (1992) Regulatory components of the denitrification
gene cluster of Pseudamanas stutzeri. In Pseudamanas Molecular Biology and
Biotechnology. Galli, 13., Silver, S. and Witholt, B. (eds). Washington, DC:
American Society for Microbiology, pp. 188-197.

Da Re, S., Bertagnoli, S., Fourment, J., Reyrat, J.-M., and Kahn, D. (1994)
Intramolecular signal transduction within the FixJ transcriptional activator: in
vitro evidence for the inhibitory effect of the phosphorylatable regulatory
domain. Nucleic Acids Res 22:1 555—1 561.

102

David, M., Daveran, M.-L., Batut, J., Dedieu, A., Domergue, 0., Ghai, J., Gertic, C.,
Boistard, P., and Kahn, D. (1988) Cascade regulation of nif gene expression in
Rhizobium meliloti. Cell 54:671—683.

de Philip, R, Batut, J., and Boistard, P. (1990) Rhizobium meliloti FixL is an oxygen
sensor and regulates R. meliloti nifA and fixK genes differently in Escherichia
coli. J Bacterial 172:4255—4262.

de Philip, R, Soupene, E., Batut, J., and Boistard, P. (1992) Modular structure of the
FixL protein of Rhizobium meliloti. Mal Gen Genet 235:49-5 4.

Dispensa, M., Thomas, C., Kim, M.—K., Perrotta, J.A., Gibson, J., and Harwood, GS.
(1992) Anaerobic growth of Rhadaspeudamanas palustris on 4-
hydroxybenzoate is dependent on AadR, a member of the cyclic AMP receptor
protein family of transcriptional regulators. J Bacterial 174: 5803-5813.

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

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

Fischer, H.-M. (1994) Genetic regulation of nitrogen fixation in rhizobia. Microb Rev
58:352-386.

Foth, H.D. (1984) Fundamentals of Soil Science, 7th Ed. New York: John Wiley &
Sons, pp. 219—221.

Fukaya, M., Tayama, K., Tamaki, T., Ebisuya, H., 0kumura, H., Kawamura, Y.,
Horinouchi, S., and Beppu, T. (1993) Characterization of a cytochrome al that
functions as a ubiquinol oxidase in Acetabacter aceti. J Bacterial 175: 4307—
43 14.

GfllesGonzalez, M.A., Ditta, G.S., and Helinski, DR. (1991) A haemoprotein with
kinase activity encoded by the oxygen sensor of Rhizobium meliloti. Nature 350:
1 7 0— 1 7 2 .

Glucksmann, M.A., Reuber, T .L., and Walker, G.C. (1993) Family of glycosyl
transferases needed for the synthesis of succinoglycan by Rhizobium meliloti. J
Bacterial 175: 7033—7044.

Gray, J.X., and Rolfe, B.G. (1990) Exopolysaccharide production in Rhizobium and
its role in invasion. Mal Microbial 4: 1425—1431.

Hirsh, A. (1992) Tansley review No. 40: Developmental biology of legume
nodulation. New Phytol 1 22: 2 1 1-237

103

Hoshino, T., and Kose, K. (1990) Cloning, nucleotide sequences, and identification
of products of the Pseudamanas aeruginasa PAO bra genes, which encode the
high-affinity branched-chain amino acid transport system. J Bacterial 172:
5531-5539.

Irvine, A.S., and Guest, J.R. (1993) Lactobacillus casei contains a member of the
CRP-FNR family. Nuc Acid Res 21: 753.

Kahn, D., and Ditta G. (1991) Modular structure of FixJ: homology of the
transcriptional activator domain with the -35 binding domain of sigma factors.
Mal Microbial 5:987-997.

Kahn, D., Batut, J., Daveran, M.-L., and Fourment, J. (1993) Structure and
regulation of the fixNOQP operon from Rhizobium meliloti. In New Horizons in
Nitrogen Fixation. Palacios, R., Mora, J., and Newton, W.E. (eds). Dordrecht, The
Netherlands: Kluwer Academic Publishers, p. 474.

Kaminski, P.A., Mandon, K., Arigoni, F., Desnoues, N., and Elmerich, C. (1991)
Regulation of nitrogen fixation in Azarhizabium caulinadans: identification of a
fixK -lil<e gene, a positive regulator of nifA. Mal Microbial S: 1983—1991.

Kim, j., and Rees, DC. (1994) Nitrogenase and biological nitrogen fixation.
Biochemistry 33: 389—397.

Leigh, J.A., Reed, J.W., Hanks, J.F., Hirsch, A.M., and Walker, G.C. (1987) Rhizobium
meliloti mutants that fail to succinylate their calcofluor-binding
exopolysaccharide are defective in nodule invasion. Cell 51: 579—587.

Long, SR. (1989) Rhizobium-legume nodulation: Life together in the underground.
Cell 56: 203—214.

Matsubara, K., 0hnishi, K., and Kiritani, K. (1992) Nucleotide sequences and
characterization of liv genes encoding components of the high-affinity

branched-chain amino acid transport system in Salmonella typhimurium. J
Biochem 112: 93-101.

Meighen, EA. and Dunlap, P.V. (1993) Physiological, biochemical and genetic
control of bacterial bioluminescence. Adv Microb Phys 34: 1-67.

Metting, RB. (1993) Structure and physiological ecology of soil microbial
communities. In Sail Microbial Ecology: Applications in Agricultural and
Environmental Management. Metting, F.B. (ed). New York: M. Dekker, pp. 3-25.

Nagy, 1., Schoofs, G., Compernolle, F., Proost, P., Vanderleyden, J., and de Mot, R.
(1995) Degradation of the thiocarbamate herbicide EPTC (S-ethyl
dipropylcarbamothioate) and biosafening by Rhodacoccus sp. strain NI86/ 21
involve an inducible cytochrome P-450 system and aldehyde dehydrogenase. J
Bacterial 177: 676—687.

104

Ozga, D.A., Lara, J.C., and Leig, J.A. (1994) The regulation of exopolysaccharide
production is important at two levels of nodule development in Rhizobium
meliloti. Mal Plant-Microbe Interact 7: 758—765.

Page, KM, and Guerinot, ML. (1995) Oxygen control of the Bradyrhizabium
japanicum hemA gene. J Bacterial 177:3979-398 4.

Parsot, C. and Mekalanos, J.J. (1991) Expression of the Vibrio cholerae gene
encoding aldehyde dehydrogenase is under control of ToxR, the cholera toxin
transcriptional activator. J Bacterial 173: 2842—2851.

Preisig, 0., Anthamatten, D. and Hennecke, H. (1993) Genes for a microaerobically
induced oxidase complex in Bradyrhizabium japanicum are essential for a
nitrogen-fixing endosymbiosis. Prac Natl Acad Sci USA 90: 3309-3313.

Priefert, H., Kriiger, N., Jendrossek, D., Schmidt, B., and Steinbiichel, A. (1992)
Identification and molecular characterization of the gene coding for
acetaldehyde dehydrogenase II (acoD) of Alcaligenes eutraphus. J Bacterial
174: 899—907.

Reyrat, J.M., David, M., Blonski, C., Boistard, P., and Batut, J. (1993) Oxygen-
regulated in vitro transcription of Rhizobium meliloti nifA and 1“er genes. J
Bacterial 175: 6867—6872.

Richter, 0.M., Tao, J.S., Turba, A., and Ludwig, B. (1994) A cytochrome ba3
functions as a quinol oxidase in Paracaccus denitrificans. Purification, cloning,
and sequence comparison. J Biol Chem 269: 23079—2 3086.

Saffarini, D.A., and Nealson, K.H. (1993) Sequence and genetic characterization of
etrA, an fnr analog that regulates anaerobic respiration in Shewanella
putrefaciens MR-l. J Bacterial 175: 7938—7944.

Sawers, R.G. (1991) Identification and molecular characterization of a
transcriptional regulator from Pseudamanas aeruginasa PA01 exhibiting
structural and functional similarity to the FNR protein of Escherichia coli. Mal
Microbial 5: 1469-1481.

Smiles, DE. (1988) Aspects of the physical environment of soil organisms. Biol Fertil
Sails 6: 204—2 1 5.

Sofia, H.J., Burland, V., Daniels, D.L., Plunkett, G., and Blattner, FR (1994) Analysis
of the Escherichia coli genome. V. DNA sequence of the region from 76.0 to
81.5 minutes. Nuc Acid Res 22: 2576-2586.

Soupene, F... Foussard, M., Boistard, P., Truchet, G., and Batut, J. (1995) Oxygen as
a key developmental regulator of Rhizobium meliloti Nz-fixation gene
expression within the alfalfa root nodule. Prac Natl Acad Sci USA 92:3759-
3763.

105

Spiro, S. (1994) The FNR family of transcriptional regulators. Antonie van
Leeuwenhoek 66: 23-36.

Spiro, S., and Guest, J.R. (1991) Adaptive responses to oxygen limitation in
Escherchia coli. T185 16: 310—314.

Upadhyaya, N.M., Scott, K.F., Tucker, W.T., Watson, J.M., and Dart, P.J. (1992)
Isolation and characterisation of Rhizobium (IC3342) genes that determine leaf
curl induction in pigeon pea. MPMI S: 129-143.

Van den Bergh, E.R., Dijkhuizen, L., and Meijer, W.G. (1993) CbbR, a LysR-type
transcriptional activator, is required for expression of the autotrophic C02
fixation enzymes of Xanthabacter flavus. J Bacterial 175: 6097—6104.

van Elsas, JD. and van Overbeek, LS. (1993) Bacterial responses to soil stimuli. In
Starvation in Bacteria. Kjelleberg, S. (ed). New York: Plenum Press, pp. 55-79.

Walk, P.C., Cai, Y., and Panoff., J.-M. (1991) Use of a transoposon with luciferase as
a reporter to identify environmentally responsive genes in a cyanobacterium.
Prac Nat Acad Sci USA 88: 5355-5359.

Xu, J. and Johnson, RC. (1995) aldB, an RpoS-dependent gene in Escherichia coli
enconding an aldehyde dehydrogenase that is repressed by Fis and activated
by Crp. JBacterial 177: 3166-3175.

Zhulin, I.B., Lois, A.F., and Taylor BL. (1995) Behavior of Rhizobium meliloti in
oxygen gradients. FEBS Lett 367:180—182.

Zimmermann, A., Reimmann, C., Galimand, M., and Haas, D. (1991) Anaerobic
growth and cyanide synthesis of Pseudamanas aeruginasa depend on anr, a
regulatory gene homologous with fnr of Escherichia coli. Mal Microbial 5: 1 483-
1 490.

CHAPTER 5

Rhizobium meliloti Genes Induced by Nutrient Deprivation

are Important for Competition in Nutrient Paar Sail.

Parts of this chapter have been presented at the 6th International Microbial
Ecology Symposium (Lim et al., 1993) and the 10th International Nitrogen Fixation
Symposium (de Bruijn et al., 1995), and have been submitted for publication in
Molecular Microbiology (Ragatz et al., 1997).

106

107
ABSTRACT

Competition between bacteria in the soil and rhizosphere for changing
supplies of scarce nutrients is an important factor that distinguishes the soil
microenvironment from other microhabitats. Little is known about how soil bacteria
sense changing stress conditions, such as nutrient deprivation, and what role is
played by the resulting genetic cascade in competition for survival in this harsh
environment. Consequently, it has been of interest to address this question using
R. meliloti strains with genes, tagged by Tn5-1063, induced under conditions of
carbon or nitrogen deprivation. Three strains harboring Tn5-1063 insertions
induced by carbon deprivation, and one strain harboring Tn5—1063 insertions
induced by nitrogen deprivation, were selected for examination of persistence and
competition with the wild-type strain 1021 in two soil types: a nutrient rich and a
nutrient poor soil system. All four mutant strains exhibited a decline in bacterial
numbers when inoculated in competition with the wild-type strain in nutrient poor
soil, but exhibited no relative decline in numbers during persistence in this soil. No
relative difference was found for either the mutant strains or the wild-type strain

when inoculated in nutrient rich soil.

1 0 8
INTRODUCTION

The soil environment is highly oligotrophic, and bacterial grth rates are
extremely slow under these conditions (see Chapter 1). Nutrients are either scarce
or in a recalcitrant form making them inaccessible to the microflora (Alexander,
1977; Foth, 1984; Metting, 1993). When nutrients do become available, there is a
burst of microbial cell growth by the indigenous populations which quickly
scavenge any available nutrients and return the soil to its normal oligotrophic state
(van Elsas and van Overbeek, 1993). Therefore, it would seem that the ability to
sense and respond to environmental signals, such as nutrient deprivation, would
be of benefit to the competition and survival of a species in the sail environment.
This is the hypothesis to which this chapter is devoted.

Although much is known about bacterial gene expression patterns from
laboratory studies involving growth of bacterial cultures under ideal conditions in
terms of temperature, pH, and nutrients, comparatively little is understood about
their behavior under unfavorable conditions, such as are found in the soil
environment. Investigation of gene expression induced by nutrient deprivation
conditions is rare, and has been largely studied in non-soil isolates. However, it is
my contention that soil isolates may have different strategies for dealing with
nutrient deprivation than non-soil isolates, such as enteric or marine bacteria.
Recently, the effect of nutrient deprivation on Pseudamanas putida, a common
gram-negative soil bacterium, has been reported (Givskov et al., 1994), using 2D-
PAGE protein analysis, the most common method for analyzing bacterial responses
to nutrient deprivation, as has been pointed out in previous chapters. However,
bacterial mutations in loci encoding these proteins carrying useful markers for

tracking them in the soil have not been available thus far (Givskov, et al., 1994;

109
Matin, 1991). Therefore, almost nothing is known about how mutations in the loci

encoding stress induced proteins affect survival and competition in the soil.

One of the problems associated with studying the relative persistence or
competition of (mutant) bacteria of interest has been how to track them in the soil
environment. Traditionally, genetic markers such as antibiotic resistance genes
have been used; however, the potential for lateral dissemination of genes encoding
these antibiotic resistance proteins precludes extensive examination in studies of
natural ecosystems (Tiedje et al., 1989; Gustafsson and Jansson, 1993).

Consequently, many other marker systems have been developed to detect
GEMS in the soil including genes encoding metabolic enzymes such as lacZY
(encoding B-galactosidase and lactose permease; Kluepfel, 1993), xylE (encoding
catechol 2,3 oxygenase; Saunders et al., 1995), and gusA (encoding [3-
glucuronidase [GUS]; Wilson et al., 1994, 1995) which are detected by their
phenotypes when grown on appropriate media. Although these systems have some
advantages in particular instances, two major problems of their use in soil systems
are the potential of background marker enzyme activity in the indigenous
microflora, and the requirement for growth and cultivation in the detection
methods. New molecular techniques such as nucleic acid probes (Zeph and Stotzky,
1989), PCR amplification of DNA (Akkermans et al., 1991; Smalla et al., 1993) and
rRNA (Hahn et al., 1990), and a novel hybrid gene phaE-caa which can be detected
using PCR, or immunological methods (Zaat et al., 1994), have thus become
invaluable tools for tracking bacteria of interest in the soil.

Other groups have focused on the use of visible markers such as the Vibrio
sp. lux genes encoding luciferase (Prosser, 1994), and the beetle luc genes
(Cebolla, et al., 1993), which encode firefly and other beetle luciferases. The
luciferase encoded by luc may be more sensitive (e.g., induced to a higher level)

than the one encoded by lux, as well as having more versatility due to the

1 10
availability of luc derivatives that emit different wavelengths of light (Cebolla et al.,

1995; Jansson, 1995). New types of phenotypic markers continue to be discovered.
Most recently, green fluorescent protein (GFP), cloned from jellyfish, has been
employed as an environmental marker (Chalfie et al., 1994) because it requires no
exogenous substrate, unlike other light emitting proteins. See Jansson (1995) for
an excellent review of these and many other marker systems being used to track
bacteria in the soil and other environments.

The potential to use luxAB expression as a visible marker for soil and
rhizosphere studies was one of the driving motivations for this study. After the
isolation and characterization of strains carrying fusions induced by environmental
stresses or other conditions, it was hoped that these strains could be used directly
in the soil environment to assess the importance of their tagged genes for
persistence and competition experiments. Studies with R. meliloti strains, carrying
Tn5-1063 luxAB fusions expressed constitutively at a high level, in different soil
environments were undertaken to test the sensitivity of a fiber optic-CCD camera
system (D.M. Ragatz and P.J. de Bruijn, unpublished). Results of this study
indicated that available equipment (including the fiber-optic system) for the
detection of luxAB expression in R. meliloti within the soil matrix, was not sensitive
enough to be used in detecting luxAB expression as an ecological marker.
Therefore, it was necessary to use the antibiotic resistance markers present in the
fusions for small, contained soil competition experiments, with the idea that they
could be replaced later (via gene-replacement techniques) with a better marker
(such as luc or GFP mentioned above), for larger, more extensive ecological studies
in natural soil environments.

In general, when tagged bacteria have been studied in the soil environment,
it has been to compare the persistence of these introduced strains versus the

indigenous strains. For example, Tn5-tagged Pseudamanas fluorescens populations

1 l 1
were introduced into a sandy soil and showed a progressive decline in numbers

(van Elsas et al., 1991). The general assumption in experiments of this type is that
the bacteria are not adversely affected by the introduction of the marker into the
genome. In fact, one might imagine that the insertion of any type of foreign
element into the bacterial genome might have a deleterious effect, however minor.
However, more work has to be done with different markers in many different
systems to establish whether or not this hypothesis holds true. Curiously, one
report of E. coli growing in batch cultures revealed that the presence of 1850, the
insertion sequence at the ends of the Tn5 transposon, improved the growth of cells
relative to their otherwise isogenic counterparts (Hartl et al., 1983). This result did
not depend on either the initial position, or the actual transposition of the ISSO
element, or a transposon containing it. Thus, the presence of the 1850 element,
itself, appears to mediate a physiological effect in the cells, rather than via
inactivation of genes during transposition. Whether or not the 1850 insertion
sequence and/or Tn5 transposons would have a beneficial effect outside of the
chemostat is not known, but is of obvious interest.

In this chapter, the use of R. meliloti with Tn5-tagged loci induced by carbon
or nitrogen deprivation is reported as a model system to measure persistence and
competition (survival) in soil systems. Sterile soil systems, containing either a
nutrient rich soil or a nutrient poor soil, were examined first to minimize the
influence of the indigenous microflora. The competition of four R. meliloti Tn5-
1063 mutant strains (carrying tagged fusions induced by carbon or nitrogen
deprivation) with an otherwise isogenic R. meliloti 1021 wild-type parental strain
was examined over time. In addition, persistence of the R. meliloti mutant strains,
as well as the wild-type strain, in the soil was examined. All four mutant strains
exhibited a significant decline in bacterial numbers when inoculated in competition

with the wild-type strain in nutrient poor soil, but exhibited no significant decline

112
in numbers during persistence in this soil. No significant decline in bacterial

numbers over time was found for either the mutant strains or the wild-type strain

when inoculated in nutrient rich soil.

113
RESULTS

Effect of mutations in deprivation-induced loci on persistence in the sail

Soil studies were initiated in order to examine the hypothesis that nutrient
deprivation-induced genes are important for persistence and competition of R.
meliloti in the environment. Mutant strain C101, carrying a Tn5-1063 fusion
induced by carbon deprivation, and mutant strain N5 carrying a Tn5-1063 fusion
induced by nitrogen deprivation, were inoculated in sterile, nutrient-rich (NR) and
nutrient-poor (NP) soils, separately or in combination with the wild-type strain (see
Materials and Methods). The soils were thoroughly mixed, and samples were taken
weekly and plated on selective media to determine bacterial numbers. The results
of these plate counts are presented in Figure 5-1, a grapth representative of two
independent experiments for each mutant strain tested. No soil counts were taken
at time zero (only culture concentration), therefore no zero points are plotted.

The Tn5-1063 tagged bacteria were tested for their ability to persist
independently of wild-type bacteria by inoculating them in either NR or NP soil,
and examining their plate counts over time. The mutant strains examined (C101
and N5), as well as the wild-type strain 1021, behaved similarly, showing little or
no decline in plate counts (Figure 5-1A & C), although a difference (approximately
ten-fold) could be observed between the relative numbers of bacteria growing in
the NP soil versus NR soil.

In order to test the effect of the Tn5-1063 induced mutations on bacterial
competition, C101 and N5 mutant strains were coinoculated with the wild-type
strain at a 1:1 ratio in NR and NP soils. The mutant strains did not show any
deviation nor decline in numbers vs. the wild-type in NR soil (Figure 5-1B & D; solid

symbols and lines).

, 8U My“;

 

'7‘

 

l
l
1
; A 1E+ 0 §
1 ;
‘ l
; lE+09 3 — ~- —— -------
' i I: N
l = r
I O r
' ;1E+08'fa‘." .b‘- “A"'7"A'
' 1’. ia' ~ \ '61-.
l a) 1. S-- ‘3
1 a 1£+07 i ...............
i ° i
! 1E+06 % —-
l 1
l .
l t
l 1E+OS 1. —~4— .+.___.+.___,__ 4
i 10 20 30 4o 50 60
t Days after inoculation
‘ i
i i
“”9 if:;>4:; “
n
E1E+08f_ ' :;ob"'.-.A-»_;;-.A.
I "f l X"“~-<3-----e-.~
l a:
‘ 3 1E+07 )5 ---------------
u, .:
o
‘ 1E+06 - — » -
a l
l 1£+os 4L 4444-- 44 ..-_ 44 — 4
ll 10 20 30 40 so 60
Days after inoculation
Figure 5-1.

1

 

p—____

114

COMPETITION

 

A .1
l

 

 

 

i
l
B i
G
1508 —-- —— — 4 - ___g __ /_ 1,,
E Nil-“~13- ‘A j --B--C101(NP)T1|
- 1 ' -, I +0101 (NR)
3' l D, - "an 1021 (NP) \
a W: """""" ' " 04—1921 mm,
o i g
1E+06 I ,;,
f o
1E+05 l -—+-—--4——+»——~+--——-—+
10 20 30 40 SO 60
Days after inoculation
1E+09 ' .
Ti: ‘ w. - -- -1, -
alE+08fO. ------- 2 -------- —- [--O~-N5(NP)
‘3 T ' ‘5 ----- -A +N5 (NR) I
a i - nay-1021mm!
‘ 154.07 . . _ 924----.0 . .. .- ‘ |
a i """"" e L+__I021<_:43>1
o i -
1E+06 i _..
l
i
1E+05 .4 ._ 444 4. — l .— 4
10 20 30 40 SO 60

Days after inoculation

 

 

 

 

 

 

 

 

 

 

 

 

 

Persistence and competition of selected Tn5-1063 induced mutant

strains in soil. In Panels A & C, the persistence of strains C101
(boxes), N5 (circles) and the wild-type strain Rm1021 (triangles) as
single inocula in NR soil (closed symbols) and NP soil (open symbols;
hatched lines) is shown. The X-axis indicates days after inoculation in
soil, and the Y-axis indicates the colony forming units (CFUs) per
gram of dried soil. In Panels B & D, the competitive ability of strain
C101 (boxes) versus 1021 (diamonds), and strain N5 (circles) versus
1021 (diamonds) in dual 1:1 inoculation experiments is shown.

 

 

 

 

115

In contrast, an observed reduction in survival of the mutant strains versus the
wild-type strain occurred in NP soil (Figure 5—1B & D; open symbols and dashed
lines). This reduction was not as pronounced in strain N5 where the wild-type
strain decreased as a similar rate. Weekly plate counts from serial dilutions of soil
samples were carried out for a total of 56 days for mutant strains C101 and N5, as
well as for the wild-type strain 1021.

Two other mutants, strains C18 and C27, carrying Tn5-1063 fusions
induced by carbon deprivation, were examined for their persistence and
competition in soil using the same experimental procedures. The functions of the
loci harboring the Tn5-luxAB fusions in these two strains have been partially
elucidated by sequence similarities (see Chapter 3), and are likely to encode an
oxidoreductase (strain C18) and a ribose transport/Chemotactic protein (strain
C27). The results of the plate counts are presented in Figure 5-2, a graph
representative of two independent experiments with each mutant strain tested. No
soil counts were taken at time zero (only culture concentration), therefore no zero
points are plotted. Weekly plate counts from serial dilutions of soil samples were
carried out for a total of 42 days for mutant strains C18 and C27, as well as for the
wild—type strain 1021.

Results similar to those presented in Figure 5-1A & C for strains C101 and
N5 were observed with strains C18 and C27. In simple inoculation experiments in
NR soil, no deviation or decline in plate counts over the time period tested were
observed (Figure 5-2A & C). Again, an approximately ten-fold decrease of total
counts could be observed comparing bacteria growing in NP versus NR soil.

As had been observed with strains C101 and N4, neither strain C18 nor
C27 showed any decline in plate counts relative to the wild-type in coinoculation

(competitive) experiments in NR soil (Figure 5-2B & D; solid symbols and lines).

SURuVAL

116

couPETITION

 

 

 

 

 

 

 

 

 

l T “ "‘ TLTA ml‘
l 1E+10 ----------- 1E+10 E ................ -
' l l l 4... ’
) 15,091, -- - - - -. ' 1E+09§ . -» ’
_ i g = 9‘1 ------ A l
i .6 l ‘s 'A l 3 I Q A 7_ ‘ 777 ___ l
a + - ---"A -— — lE 08£-—~,- --------------- '_-G__C18 NP ‘
l E: 'E 08' A S """ B “““ B i '5' + f 'B“-- ' C18(NR) '
P. i i a. * ~37 l+ ( ) I
' 3 " . -_. I \ 1E+07 ' , ___ _ _ g ‘rs‘ _ _ g g, ln'A“ 1021 (NP)
' 3 ”+07 i ' ' a i ‘ t—f—iom (NR)_J
l l l . _ _--_ _ g
o r 7 o l .7
' 1£+os ', ; iE+06 L— 4 _b .
l
1 f ' i
'| lE+OS #44444 4444444444 l lE+OS I A. . +--~+—~—~—4
l o 10 20 30 40 50 l a 10 20 30 40 50
o o I
‘ Days after inoculation I Days after Inoculation
rL if
I
1£+1o ------ - -------- lE+lO f ___________ - ........
c , la
' m?_ I 1E+09 ' _ W:
lE+09 - . - — - 4 4 I
_ ' _ .‘7
E ”.45 l E 0. 3... 7 7 7 _7
grace . gram-34‘? _-_O -— 3.1508 f v; " ‘ "-‘a-‘---°‘A' l--o-- c27 (NP) '
u i '9" ‘3, ¥ °-.. l+c27 (NR)
. a,
3 1E+07- —. ---------------- l 515.07 _— ————————— 9 '4- "-6 —- ‘M‘A 1021(NP)
2 i g +1621<18L
o l T
isms: ----------------- 15:-06; - _
’ l . h h
1£+os l—---—#——+~—L—+— 4 lE+OS . . . . .
O 10 20 3O 4O 50 O 10 20 30 40 50
Days after inoculation Days after inoculation
Figure 5-2. Persistence and competition of other selected Tn5-1063 induced

mutant strains in soil. In Panels A & C, the persistence of strains C1 8
(boxes), C27 (circles) and the wild-type strain Rm1021 (diamonds) as
single inocula in NR soil (closed symbols) and NP soil (open symbols;
hatched lines) is shown. The X-axis indicates days after inoculation in
soil, and the Y—axis indicates the colony forming units (CFUs) per
gram of dried soil. In Panels B & D, the competitive ability of strain
C18 (boxes) versus 1021 (diamonds), and strain C27 (circles) versus
1021 (diamonds) in dual 1:1 inoculation experiments is shown.

117

However, a large reduction in relative plate counts was observed in NP soil (Figure
S-ZB & D; open symbols and dashed lines).

Overall, the relative reduction in bacterial plate counts was most
pronounced for the strains harboring fusions induced by carbon deprivation,
resulting in a decline to approximately 1.0% (strains C101 and C27) to less than
0.1% (strain C18) of their original numbers. The mutant strain N5 carrying a
nitrogen deprivation-induced Tn5-1063 fusion showed a decline to approximately

10% of its original cell numbers in NP soil.

1 1 8
DISCUSSION

The large relative decline in bacterial numbers of mutant strain C101 only
in the presence of wild-type bacteria in nutrient poor (NP) soil suggests a role for
this induced locus; namely, as a mechanism to compete for, and utilize, unusual or
‘poor’ substrates as energy sources critical to the long-term survival of R. meliloti in
competition with other microbes in the soil. Since the Tn5-1063 fusion in strain
C101 is not induced by alternative carbon sources (see chapter 3, Table 3-1b), it
may encode a protein that is near the top of a regulatory pathway, such as a
sensor, making strain C101 especially susceptible to deprivation conditions in
competition with the wild-type Rhizobium. In fact, the Tn5-1063 fusion in strain
C101 is also induced by low oxygen levels (see Chapter 4), and thus makes sense as
a regulatory protein. Alternatively, the gene fusion in mutant C101 may be
involved in competition in some way that is not directly related to carbon
assimilation, but to the deprivation condition itself. To date, no genes have been
isolated in R. meliloti that are directly involved in competition in the soil. However,
it is known that mutations in some of the nod genes, and in nifA, reduce both
nodulation rate and competitiveness (Downie and Johnston, 1988; Sanjuan and
Olivares, 1991). In addition, two genes have recently been isolated by Tn5
mutagenesis which appear to be involved only in nodule competitiveness
(Onishchuk et al., 1994). The nature and regulation of these genes is unknown.

The mutant strains C18 and C27 were also remarkably affected by the
presence of the wild-type bacteria in NP soil. The fusion induced in strain C18 is
also responsive to alternative carbon sources (Table 3-1b) and appears to be an
oxidoreductase, suggesting that it is being outcompeted by the wild-type strain
1021 for the scarce nutrients available in NP soil. It might also be sensitive to

secretions from the wild-type owing to a mutation in a gene required for

1 1 9
degradation of potentially harmful bacterial waste products. The fusion induced in

strain C27, in contrast, was shown to be similar to a ribose transport/chemotaxis
operon. Unlike strain C18, strain C27 is only induced by one other carbon source
tested, glucose. It would appear, then, that it was outcompeted for available carbon
substrates in NP soil.

In comparison, strain N5 was less affected by the presence of the wild-type
bacteria in competition for survival in NP soil, but the trend is still quite different
than when strain N5 is inoculated in persistence experiments. Because the gene
from this mutant has been identified as being involved in nitrate utilization, the
most likely explanation of these results is that the mutant strain N5 is able to
obtain some fixed nitrogen, perhaps as ammonia, even in the NP soil; however, not
enough to completely alleviate the problem, which is most pronounced when the
wild-type strain 1021 is also present. Presumably, the wild-type Rhizobium utilizes
the soil nitrate, a nitrogen source inaccessible for the mutant strain NS. The
observation (made after this set of experiments involving strain N5) made in
chapter 2 that strain N5 is non-motile presents an alternative explanation for the
reduction in plate counts of strain NS versus the wild-type strain in NP soil. Lack of
motility has been observed to be a competitive disadvantage to bacteria in the soil
and rhizosphere (Lauffenburger et al., 1982; Dowling and Broughton, 1986;
Triplett and Sadowsky, 1992; Moens and Vanderleyden, 1996). Thus, the
argument that the motility‘ phenotype (not coupled to the Tn5-1063 insertion, see
chapter 2) is responsible for lower bacterial plate counts of strain N5 versus the
wild-type 1021, is at least equally valid as the explanation presented initially. Only
by repeating the experiment with the motility restored N5 strain (see chapter 2)
could this question be answered.

The ability of these strains, as well as all other mutants tested, to survive in

the soil and successfully inoculate alfalfa on their own is an indication that it is not

1 20
the presence of the Tn5 transposon itself which is putting an insurmountable

burden on the bacteria, but the lack of a functional protein necessary to compete
with the wild-type strain during stressful (nutrient limiting) conditions.
Examination of these and other deprivation-induced loci in more natural, i.e., non-
sterile soils is the next step in understanding the roles of these loci, most of which
do not have a phenotype or genetic similarity to known genes. This could lead to an
enhanced understanding of soil microbial communities which would aid in the
design and tracking of GEMS, as well as to a better understanding of how

introduced chemicals and pollutants affect the microbial ecology of the soil.

121
MATERIALS AND METHODS

For bacterial strains and plasmids used, media and growth conditions,
transposon mutagenesis, DNA isolation and manipulation, DNA sequence analysis,
and nodulation and nitrogen-fixation assays see chapter 2 (Materials and

Methods).

Soil Persistence and Competition Experiments

Two types of soil were used for seeding with bacteria; a nutritionally rich soil
consisting of a 5:3:1 ratio of Metromix 510:sand:vermiculite, and a nutrient poor
soil consisting only of sand. Metromix 510 is a professional plant growth medium
consisting of Sphagnum, peat moss, ash bark, and vermiculite, and is rich in carbon
and nitrogen. Soils were sifted through a metal strainer to increase homogeneity,
placed in small glass jars (100 grams per jar), and sterilized by autoclaving three
times for 25 minutes, with 24 hours between each autoclaving.

Bacteria were grown to saturation (36 hours) in 5 ml of 'IY medium at 28°C.
a 100 pl portion of culture was subcultured in 5 ml of TY medium and grown to a
cell density of approximately OD600=1.O. The cell density was adjusted by
spectrometric measurement of optical density and subsequent centrifugation and
resuspension (dilution or concentration) of the cultures in sterile distilled H20 to a
final cell density of OD600=1.0, before 20 ml of cell culture was inoculated into the
soil. For competition experiments, the bacterial cultures were centrifuged, and the
pellets washed twice with sterile distilled H20, before resuspending them to a final
cell density of OD600=1.0. Ten ml of each culture to be examined was combined in a
sterile tube and briefly vortexed before inoculation into the soil. The soil was

thoroughly mixed with a sterile spatula. Experiments were carried out in duplicate.

1 2 2
Plate counts were determined by placing approximately one gram of soil into

sterile 100 ml dilution blanks and determining the exact weight of the soil. Serial
dilutions were carried out and two dilutions were plated three times on TY-plates
containing 1011200 and 8mm, respectively. Initial counts were determined in
triplicate with three soil samples from each jar, but this was found to be
unnecessary, since the levels of error were found to be higher between dilution

series than between soil replications (data not shown).

123
REFEREN‘

Akkermans, A.D.L., Hahn, D., and Mirza, MS. (1991) Molecular ecology of Frankia:
advantages and disadvantages of the use of DNA probes. Plant Soil 13?: 49—5 4.

Alexander, M. (1977) In Introduction to Sail Microbiology, 2nd ed. New York: John
Wiley & Sons, pp. 16—51.

Cebolla, A., Ruiz-Berraquera, F., and Palomares, A.J. (1993) Stable tagging of
Rhizobium meliloti with the firefly luciferase gene for environmental
monitoring. App Env Microbial S9: 251 1—25 19.

Cebolla, A., Vazquez, ME, and Palomares, A.J. (1995) Expression vectors for the
use of eukaryotic luciferases as bacterial markers with different colors of
luminescence. App Env Microbial 61: 660-668.

Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W., and Prasher, DC. (1994) Green
fluorescent protein as a marker for gene expression. Science 263: 802-805.

de Bruijn, F.J., Graham, L., Milcamps, A., and Ragatz, D. (1995) Use of the luciferase
(Tn5-lux) reporter system to study Rhizobium meliloti genes responding to
N/C/O2 limitation or plant factors and their role in rhizosphere competition. In
Nitrogen Fixation: Fundamentals and Applications. Tikhonovich, I.A., Provorov,
N.A., Romanov, V.I., and Newton, W.E. (eds). Boston: Kluwer Academic
Publishers, pp. 195—200.

Dowling, D.N. and Broughton, W.J. (1986) Competition for nodulation of legumes.
Ann Rev Micriobiol 40: 131-157.

Downie, J.A., Johnston, A.W.B. (1988) Nodulation of legumes by Rhizobium. Plant
Cell Environ 11: 403—412.

Foth, H.D. (1984) Fundamentals of Soil Science, 7th Ed. New York: John Wiley &
Sans, pp. 219-221.

Givskov, M., Eberl, L., and Molin, S. (1994) Responses to nutrient starvation in
Pseudamanas putida KT2 442: two-dimensional electrophoretic analysis of
starvation- and stress-induced proteins. J Bacterial 176: 4816-482 4.

Gustafsson, K. and Jansson, J.K. (1993) Ecological risk assessment of the deliberate
release of genetically modified microorganisms. Ambia 22: 236—242.

Hahn, D., Starrenburg, J.J.C., and Akkermans, A.D.L. (1990) Oligonucleotide
probes that hybridize with rRNA as a tool to study Prankia strains in root
nodules. App Env Microbial 56: 1 342-1 346.

Hartl, D.L., Dykhuizen, D.E., Miller R.D., Green, L., and Framond, J (1983)
Transposable element ISSO improves growth rate of E. coli cells without
transposition. Cell 35: 503—510.

Jansson, J. (1995) Tracking genetically engineered microorganisms in nature. Curr
Opin Biotech 6: 275-283.

124

Kluepfel, DA. (1993) The behaviour and tracking of bacteria in the rhizosphere.
Ann Rev Phytapathal 31: 441—472.

Lauffenburger, D., Aris, R., and Keller, K. (1982) Effects of cell motith and
chemotaxis an microbial population growth. Biophys J 40: 209-219.

Lim, P.O., Ragatz, D., Renner, M., and de Bruijn, P.J. (1993) Environmental control
of gene expression: isolation of Rhizobium meliloti gene fusions induced by N-
and C- limitation. In Trends in Microbial Ecology. Guerrero, R., and Pedras-Alié,
C. (eds). Spanish Society for Microbiology, pp. 97-100.

Matin, A. (1991) The molecular basis of carbon-starvation-induced general
resistance in Escherichia coli. Mal Microbial 5: 3—10.

Metting, RB. (1993) Structure and physiological ecology of soil microbial
communities. In Sail Microbial Ecology: Applications in Agricultural and
Environmental Management. Metting, F.B. (ed). New York: M. Dekker, pp. 3-25.

Moens, S. and Vanderleyden, J. (1996) Functions of bacterial flagella. Crit Rev
Microbial 22: 67—100.

Onichchuk, O.P., Sharypava, LS, and Simarav, BB. (1994) Isolation and
characterization of the Rhizobium meliloti Tn5-mutants with impaired
nodulation competitiveness. Plant Sail 167: 267-274.

Prosser, J]. (1994) Molecular marker systems for detection of genetically
engineered miicro-arganisms in the environment. Microbiology 140: 5-1 7.

Ragatz, D.M., Milcamps, A., Lim, P.O., Berger, K.A., and de Bruijn, P.J. (1997)
Isolation of carbon and nitrogen starvation-induced loci of Rhizobium meliloti
by Tn 5-luxAB mutagenesis. Mal Microbial. submitted.

Sanjuan, J., and Olivares, J. (1991) NifA-NtrA regulatory system activates
transcription of nfe, a gene locus involved in nodulation competitiveness of
Rhizobium meliloti. Arch Microbial 155: 543—5 48

Saunders, J.R., Pickup, R.W., Morgan, J.A., Winstanley, C., and Saunders, V.A.
(1995) XylE as a marker for microorganisms. In Molecular Microbial Ecology
Manual. Akkermans, A.D.L., van Elsas, J.D., and de Bruijn, P.J. (eds). Dardrecht:
Kluwer Academic Publishers.

Smalla, K., Cresswell, N., Mendonca-Hagler, L.C., Walters, AC, and van Elsas, JD.
(1993) Rapid DNA extraction protocol from soil for polymerase chain reaction-
mediated amplification. J Appl Bacterial 74: 78—85.

Tiedje, J.M., Colwell, R.K., Grassman, Y.L., Hodsan, R.E. Lenski, R.E., Mack, R.N., and
Regal, P.J. (1989) The planned introduction of genetically engineered

organisms: ecological considerations and recommendations. Ecology 70: 298—
31 S.

Triplett, B.W. and Sadowsky, M.J. (1992) Genetics of competition for nodulation of
legumes. Annu Rev Microbiol 46: 399—428.

125

van Elsas, JD. and van Overbeek, LS. (1993) Bacterial responses to soil stimuli. In
Starvation in Bacteria. Kjelleberg, S. (ed). New York: Plenum Press, pp. 55-79.

van Elsas, J.D., van Overbeek, L.S., Feldman, A.M., Dullemans, A.M., and de Leeuw,
O. (1991) Survival of a genetically engineered Pseudamanas fluorescens in soil
in competition with the parent strain. FEMS Microbial Ecol 85: 53-64.

Wilson, K.J., Sessitsch, A., and Akkermans, A. (1994) Molecular markers as tools to
study the ecology of microorganisms. In Beyond the Biomass: Compositional and
Functional Analysis of Aail Microbial Communities. Ritz, K., Dighton, J., and
Giller, K.E. (eds). Chichester: John Wiley, pp. 149-156.

Zaat, S.A.J., Slegtenharst-Eegdeman, K., Tommassen, J., Geli, V., Wijffelman, GA.
and Lugtenberg, B.J.J. (1994) Construction of phaE-caa, a novel PCR— and
immunologically detectable marker gene for Pseudamanas putida. Appl Env
Microbial 60: 3965-3973.

Zeph, LR. and Stotzky, G. (1989) Use of a biatinylated DNA probe to detect
bacteria transduced by bacteriophage P1 in soil. Appl Env Microbial 55: 661-
665.

CHAPTER 6

R. meliloti Genes Induced by Nutrient Deprivation and
Affected in Soil Competition are Not Involved in

Competition for Nodulation.

126

127
ABSTRACT

The rhizophere environment is characterized by plant root exudates, and
higher bacterial numbers and diversity than in bulk soil (Bolton et al., 1993),
creating conditions where it is highly advantageous for a microorganism to respond
quickly to the changing conditions in order to compete for available microsites.
Rhizobia are particularly interesting in that they are competing not only for
nutrients, but for sites that can lead to infection thread formation resulting in a
microhabitat that is free from outside competition. Five mutant strains harboring
Tn 5—1063 insertions induced by carbon deprivation, as well as a constitutively
expressed control, were selected and examined for competition for nodulation
versus the wild-type 1021 at three different starting concentrations: 1:10, 1:1, and
10:1 mutant versus wild-type strains in two independent experiments. None of the
mutant strains exhibited a loss of nodule competitiveness at any of the
concentrations tested. Rather, two strains (C101 and C22) were significantly better
than the wild-type strain at nodulation efficiency in both of the independent

experiments.

1 2 8
INTRODUCTION

Rhizobia are unique soil bacteria that live in two distinct habitats: in the soil
and rhizosphere, and in nodules induced on their host plant (Vincent, 1970). The 1
unique nodule environment is separated from the ‘outside’, and thus presents an
ideal biological system for examining bacterial competition in the absence of the
complex microbial communities in the soil and rhizosphere. Because of this ideal
system, and because of the importance of understanding and controlling the
efficiency of biological nitrogen-fixation, more work has been carried out in the
elucidation of the molecular mechanisms underlying competition for nodulation
than in soil persistence and competition experiments (see e.g., McLaughlin et al.,
1987; Lagares et al., 1992; Milner et al, 1992; Onishchuk et al., 1994; Sharypova et
al, 199 4)

For example, Jiménez-Zurdo et al. (1995) obtained an R. meliloti Tn5
insertion mutant in a proline dehydrogenase—which allows R. meliloti to utilize the
amino acids ornithine and proline as sole carbon and nitrogen sources—which did
not alter its ability to fix nitrogen, but drastically reduced its ability to compete for
nodule occupancy (the ability of a particular rhizobial strain to occupy its host
nodule; see further explanation below). In this case, competition could be
understood in terms of the type of mutation involved. In other cases, however,
nothing is known about the mutation in question except that it results in lowered
nodulation competitiveness (Onishchuk et al., 1994; McLaughlin et al., 1987) or
sometimes enhanced symbiotic effectiveness (Sharypova et al., 1994). In
Rhizobium meliloti and Bradyrhizabium japanicum, genes have been isolated on
the megaplasmids, termed nfe genes (nodule formation efficiency), which are
involved in nodulation efficiency and competitiveness (Tara and Olivares, 1986;

Sanjuan and Olivares, 1989; Chun and Stacey, 1994). In R. meliloti, expression of

1 2 9
the nfe genes has been found to be regulated by the NifA-NtrA regulatory system,

which also controls the nitrogen fixation genes (Sanjuan and Olivares, 1991).

Examination of nodule competition is not without complications, however.
Many factors have been shown to influence competitiveness including: genetic
determinants of nitrogen fixation, production of polysaccharides, as well as
environmental conditions such as temperature and soil pH (Noel and Brill, 1980;
Boonker et al., 1978; Beattie et al., 1989). In addition, multiple-occupancy of a
single nodule by more than one rhizobial strain has been observed in selected
symbiosis, particularly in those giving rise to determinate nodules such as in R.
trapici-bean (Wilson et al., 1995; Sessitsch et al., 1996). In other cases (e.g., R.
meliloti-alfalfa) single-occupancy in a discrete nodule is the norm (Olivares et al.,
1980; Soto et al., 1992; Goldman et al., 1994). It may be that outside factors
influence multiple-occupancy as well.

Because of the limited amount of nodules which are formed on the plant
roots (Vincent, 1970; Amarger and Lobreau, 1982; Hirsch, 1992; regulated by
unknown pathways), rhizobia are said to compete for nodule occupancy when two
strains which can nodulate a particular host are both competing for the available
sites which lead to nodulation. This competition occurs even when multiple-
occupancy is the norm.

In this chapter the use of R. meliloti strains harboring Tn5-tagged loci
induced by carbon deprivation as a model for measuring competition for
nodulation is reported. Some of these strains (C18, C27, and C101) carry Tn5—
1063 insertions in genes important for competition versus the wild-type strain
1021 in nutrient poor soil (see chapter 5). It is thought that because of the
competitive nature of the rhizophere environment (see chapter 1), and because R.
meliloti is an endosymbiont of alfalfa, genes responding to nutrient deprivation

would be of benefit in competing for microsite occupancy on the roots of alfalfa

1 30
leading to subsequent root hair curling and nodule invasion. Therefore, mutations

in such genes due to insertional inactivation by Tn5-1063 might lead to reduced
nodulation effectiveness, or at the very least to no significant change in nodulation
competitiveness phenotype. However, we found no competitive advantage for
nodulation of wild-type strain 1021 versus any of the mutant strains tested,
whereas two of the strains, C22 and C101, appear to be enhanced in their

nodulation competitiveness phenotype.

131
RESULTS

Efiect of mutations in deprivation-induced loci on competition for nodulation

Nodule competition studies were initiated in order to test the importance of
the nutrient deprivation-induced genes for competition of R. meliloti for
nodulation. Mutant strains C18, C27, and C101, carrying Tn5-1063 fusions
induced by carbon deprivation (and oxygen limitation in the case of the locus
tagged in strain C101), and shown to be outcompeted by the wild-type in nutrient
poor soil (see Chapter 5) were chosen for co-inoculation experiments versus the
wild-type strain 1021 on plant roots. Mutant strains C22 and C47, also carrying
carrying Tn5-1063 fusions induced by carbon deprivation (and in the case of
strain C22, carrying a fusion induced by nitrogen and oxygen limitation as well),
were included in the nodule occupancy experiments., Strain CV1, bearing a
constitutively expressed fusion, was also included in the analysis as an uninduced
control. All six strains, as well as the wild-type strain, were grown in 5 ml TY liquid
medium to saturation (36 hrs), reinoculated in larger volumes of TY, and grown to
an approximate OD600 of 1.0 as determined by spectrophotometer. Cells were
centrifuged, washed twice with ddHZO, and resuspended in a volume of ddHZO
calculated to yield an OD600 =- 1.0. Bacterial cultures were combined into test-tubes
at three ratios of mutant to wild-type: 1:10, 1:1, and 10:1, and diluted 1:5 with
ddHZO before inoculating one week old alfalfa seedlings in large tubes with 1 ml of
each condition (five tubes per condition tested). The results of two independent
experiments, separated temporally, are presented in Figures 6-1A and 6-1B,
respectively.

In the experiment shown in Figure 6-1A, none of the five mutant strains

tested showed a relative decrease in nodule occupancy at any of the three

132

I

.>

Nodule Competition #1

Ratio of Mutantzwild-type (1021)

 

   

C22 C27 C47 C101 CV1
Strain

PUT?—

Nodule Competition #2

Ratio of Mutant:wild—type (1021)

 

 

   

Strain

Figure 6-1. Competition for nodulation of selected Tn5-1063 induced mutant
strains versus the wild-type 1021. In Panels A & B, the percentage
number of nodules harboring mutant strains vs. the wild-type is
shown in two independent experiments, respectively. Mutants were
inoculated at 1:10 (blue bars), 1:1 (red bars), and 10:1 (yellow bars)
ratios versus the wild-type. Nodules were individually harvested and
plated to determine which strain was inhabiting the nodule.

133
inoculation ratios tested. In fact, the nodulation efficiency for all of the mutant

strains was significantly higher than expected (<50% expected) at the 1:1
inoculation ratio, whereas the control strain CV1 was within normal limits at this
ratio (Figure 6-1A). In addition, strains C22, C47, and possibly C101 (a large
standard of error was observed) were significantly enhanced for nodule occupancy
versus the wild-type strain at the 1:10 ratio of mutant strain to wild-type strain
inoculation. Strain C22 was also enhanced for nodule occupancy versus the wild-
type strain at the 10:1 ratio of mutant strain to wild-type strain inoculation (no
nodules harboring the wild-type strain were recovered), making this strain the best
competitor overall. No other strains deviated significantly from expected results at
the 10:1 ratio, except for a slightly high value for the control strain, CV1 (no
nodules harboring the wild-type strain were recovered).

The experiment was repeated using the same strains and conditions
described above. The results, showing a similar pattern of nodule occupancy to the
first experiment, are shown in Figure 6-1B. As before, none of the five mutant
strains tested demonstrated a relative decrease in nodule occupancy at any of the
three inoculation ratios tested. Unlike the first experiment, however, nodulation
efficiency of only strain C101 (and possibly strains C27 and C47) was significantly
higher than expected at the 1:1 innoculation ratio. In addition, nodulation
efficiency of strain C22, but not strains C47 and C101, was again greater than
expected at the 1:10 and 10:1 inoculation ratios. No other strains, including the
control strain CV 1, deviated significantly from their expected values at inoculation

ratios of 1:10 and 10:1.

1 3 4
DISCUSSION

The data in these competition for nodule occupancy experiments clearly
indicate that no competitive disadvantage is incurred by the insertion of Tn 5-1063
in the loci of these selected mutants harboring fusions induced by carbon
deprivation, under the conditions used in this experiment. In addition, the results
presented in this chapter are very similar to those found by Anne Milcamps for
competitive nodulation of three of the same R. meliloti Tn5-1063 tagged strains
(C18, C27, and C101) versus the wild-type in a separate study (Ragatz et al., 1997)
using similar conditions. In this study, experiments were done in triplicate, and
statistical analysis of the nodule occupancy data demonstrated that no significant
difference between the mutant strains and the wild-type strain existed.

The data from both experiments is particularly surprising given that three
of these mutants (C18, C27, and C101) were all shown to be at a competitive
disadvantage in nutrient poor soils (see Chapter 5). However, unlike the soil
system, bacteria inoculated on one week old alfalfa roots would not be starved due
to the presence of nutrient exudates released by both the seeds and the plant
roots into the B + D solution. This system then, may more resemble the nutrient
rich soil system described in chapter 5, than the nutrient poor soil system in which
the Tn5-1063 tagged loci were found to be important in competition.

Clearly, the hypothesis that the presence of Tn5 in the genome would be
detrimental in and of itself is not true in this system. One might even suspect it was
advantageous, much like the IS50 element in E. coli batch cultures (Hartl et al.,
1983), due to the number of strains with increased nodulation competitiveness
versus the wild-type strain. However, the contitutively expressing control, CV1, was

nearly ideal in terms of its nodule occupancy at the various inoculation ratios.

135

Nevertheless, the increase in nodule competitiveness of most of the strains at one
or more of the inoculation ratios is a surprising result.

Strains C22 and C101, in particular, were enhanced for nodulation
competitivess in both of the experiments, and at more than one ratio. Neither of
the cloned loci from these mutants show any similarity to known genes (see chapter
3, Table 3-1). In addition, the Tn5-1063 tagged genes they harbor are regulated
by multiple stresses (see Chapter 4). This suggests that the Tn5-1063 tagged loci
encode regulatory proteins of some sort, and that the presence of these genes (or
regulons) in the genome presents a kind of energetic burden to the cell, such that it
must be compensated for by some mechanism which improves competition for
survival during deprivation conditions. It would be interesting to observe these
strains in similar nodulation competitiveness experiments after they were exposed
to deprivation conditions for a period of time. Nodule occupation competitiveness
might be exprected to drop for these strains under such conditions, if the genes
encoded by their tagged loci are indeed important for competition for survival
prior to their exposure to plant roots.

As far as I know, this is the first report of Tn5-tagged loci with improved
nodulation competitiveness in R. meliloti. Sharypova et al. (1994) isolated ten Tn5-
induced mutants in R. meliloti with enhanced symbiotic effectiveness (Eff*+) as
measured by an increase in the host plant’s growth (dry weight), at the expense of
nitrogen fixation in the root nodules. However, these strains were not tested for
nodulation competitiveness versus the wild-type strain.

The observation that strains C22 and C101 are also induced by
microaerobiosis suggests a regulatory pathway for their induction via the NifA-NtrA
regulatory pathway. As was mentioned previously (chapter 4), part of this pathway

is oxygen regulated (NifA) via the two-component FixL/FixJ system, and are the

136
same transcriptional regulators of the nfe locus involved in nodulation

competitiveness of R. meliloti (Sanjuan and Olivares, 1991).

Overall, the results of these experiments do not appear to support the initial
hypothesis, stated in chapter 1, that genes regulated by nutrient deprivation might
be important for competition for nodulation. However, It may be that pre-starving
the bacteria for nutrients, or inoculation of the bacteria into the B + D media (which
contains no carbon or nitrogen sources) at a time prior to plant germination, would

yield completely different results from those obtained in this experiment.

1 37
MATERIALS AND METHODS

For bacterial strains and plasmids used, media and growth conditions,
transposon mutagenesis, DNA isolation and manipulation, DNA sequence analysis,
and nodulation and nitrogen-fixation assays see chapter 2 (Materials and

Methods).

Nodule Competition Experiments

Alfalfa seeds (Medicaga sativa bv. cardinal) were sterilized by soaking for
three minutes in 95% ethanol, followed by three minutes in 0.1% HgClz and rinsed
thoroughly with sterile distilled water. Three seeds were placed on a piece of sterile
Whatman 3MM filter paper in test tubes containing 20 ml of sterile, nitrogen-free
B+D liquid medium (Broughton and Dilworth, 1971). Five tubes were used for each
strain at each of the three concentrations tested.

Bacteria were grown to saturation (48 hours) in 5 ml of TY medium at 28°C.
100 pl of culture was subcultured in 5 ml of TY medium and grown to a cell density
of approximately OD600=1.0. The cell density was adjusted by spectrophotometric
measurement of optical density and subsequent centrifugation and resuspension
(dilution or concentration) of the cultures in sterile distilled H20 to a final cell
density of OD600=1.0. Mutant cultures were combined with the wild-type 1021 at
ratios of 1:10, 1:1, and 10:1, diluted to a 1:5 concentration, and applied to one
week old alfalfa roots at a final volume of 1ml. Each culture was briefly vortexed
before inoculation into the tubes.

Nodule occupancy was determined by collecting the nodules from two to
three tubes (yielding approximately 20—40 nodules per condition) containing 8-10

week old plants, surface sterilizing them in 10% bleach for 3 minutes in individual

138
1 ml microtubes, and rinsing them three times with sterile ddHZO. Nodules were

crushed with sterile micro-pestles and plated on TY medium containing szso or
Km200 and grown for three days at 28°C. Cultures which grew on both antibiotics

were streaked in TY plates with szso, and individual colonies were picked for

plating on TY-sz00 plates.

139
REFERENCES

Amarger, N. and Lobreau, JP. (1982) Quantitative study of nodulation
dompetitiveness in Rhizobium strains. App Env Microbial 44: 583—588.

Beattie, G.A., Clayton, M.K., and Handelsman, J. (1989) Quantitative comparison fo
the laboratory and field competitiveness of Rhizobium leguminasarum biovar
phaseali. App EnvMicrabiol 55: 2755-2761.

Bolten Jr., H., Fredrickson, J.K., and Elliott, LP. (1993) Microbial ecology of the
rhizosphere. In Sail Microbial Ecology: Applications in Agricultural and
Environmental Management. Metting, F.B. (ed). New York: M. Dekker, pp. 27—
63.

Boonker, N.D., Weber, BF, and Bezdicek, DP. (1978) Influence of Rhizobium
japanicum strains and inoculation methods on soybeans grown in thizobia-
populated soil. Agran J 70: 547—549.

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

Chun, J.Y. and Stacey, G. (1994) A Bradyrhizabium japanicum gene essential for
nodulation competitiveness is differentially regulated from two promoters. Mol
Plant Microbe Interact 7: 248-255.

de Bruijn, F.J., Graham, L., Milcamps, A., and Ragatz, D. (1995) Use of the luciferase
(Tn5-lax) reporter system to study Rhizobium meliloti genes responding to
N/C/Oz limitation or plant factors and their role in rhizosphere competition. In
Nitrogen Fixation: Fundamentals and Applications. Tikhonovich, I.A., Provorov,
N.A., Romanov, V.I., and Newton, W.E. (eds). Boston: Kluwer Academic
Publishers, pp. 195-200.

Goldman, A., Lecoem, L., Message, B., Delarue, M., Schoonejans, E., and Tepfer, D.
(1994) Symbiotic plasmid genes essential to the catabolism of proline betaine,
or stachydrine, are also required for efficient nodulation by Rhizobium meliloti.
FEMS Microbial Lett 115: 305-312.

Hartl, D.L., Dykhuizen, D.E., Miller R.D., Green, L., and Framond, J (1983)
Transposable element IS50 improves growth rate of E. coli cells without
transposition. Cell 35: 503-510.

Hirsh, A. (1992) Tansley review No. 40: Developmental biology of legume
nodulation. New Phytol 122: 21 1—237

Jiménez—Zurdo, J.I., van Dillewijn, P., Soto, M.J., de Felipe, M.R., Olivares, J., and
Toro, N. (1995) Characterization of a Rhizobium meliloti proline dehydrogenase
mutant altered in nodualtion efficiency and competitiveness on alfalfa roots.
MPMI 8: 492—498.

Lagares, A., Caetano—Anollés, G., Niehaus, K., Lorenzen, J., Ljunggren, H.D., Piihler,
A., and Favelukes, G. (1992) A Rhizobium meliloti Lipopolysaccharide mutant
altered in competitiveness for nodulation of alfalfa. J Bacterial 174: 5941—5952.

140

McLaughlin, T.J., Merlo, A.O., Satola, S.W., and Johansen, E. ( 1987) Isolation of
competition-defective mutants of Rhizobium fredii. J Bacterial 169: 410—413.

Milner, J.L., Araujo, R.S., and Handelsman, J. (1992) Molecular and symbiotic
characterization of exopolysaccharide-deficient mutants of Rhizobium trapici
strain CIAT899. Mol Microbiol 6: 3137—3147.

Noel, K.D. and Brill, W.J. (1980) Diversity and dynamics of indigenous Rhizobium
japanicum populations. App Env Microbial 40: 931—938.

Olivares, J., Casadesus, J., and Bedmar, B.J. (1980) Method for testing degree of
infectivity of Rhizobium meliloti strains. App Env Microbial 56: 389—393.

Onichchuk, O.P., Sharypava, L.S., and Simarav, B.V. (1994) Isolation and
characterization of the Rhizobium meliloti Tn5—mutants with impaired
nodulation competitiveness. Plant Sail 167: 267-274.

Ragatz, D.M., Milcamps, A., Lim, P.O., Berger, K.A., and de Bruijn, F.J. (1997)
Isolation of carbon and nitrogen starvation-induced loci of Rhizobium meliloti
by Tn 5-luxAB mutagenesis. Mal Microbial. submitted.

Sanjuan, J ., and Olivares, J. (1989) Implication of nifA in regulation of genes
located on a Rhizobium meliloti cryptic plasmid that affect nodulation efficiency.
JBacterial 171: 4154—4161.

Sanjuan, J., and Olivares, J. (1991) NifA-NtrA regulatory system activates
transcription of nfe, a gene locus involved in nodulation competitiveness of
Rhizobium meliloti. Arch Microbial 155: 543-5 48

Sessitsch, A., Wilson, K.J., Akkermans, A.D.L., and de Vos, WM. (1996)
Simultaneous detection of different Rhizobium strains marked with either the

Escherichia coli gusA gene or the Pyracaccus furiasus celB gene. App Environ
Microbiol 62: 4191-4194.

Sharypava, l..A., Onishchuk, O.P., Chesnokova, O.N., Fomina-Eshchenko, J.G., and
Simarav, B.V. (1994) Isolation and characterization of Rhizobium meliloti Tn5

mutants showing enhanced symbiotic effectiveness. Microbiology 140: 46 3-
470.

Soto, M.J., Zorzano, A., Olivares, J., and Taro, N. (1992) Nucleotide sequence of
Rhizobium meliloti GR4 insertion sequence ISRm3 linked to the nodulation
competitiveness locus nfe. Plant Mol Biol 20: 307-309.

Taira, K., Hirose, J., Hayashida, S., and Furukawa, K. (1992) Analysis of bph operon
from the polychlorinated biphenyl-degrading strain of Pseudamanas
pseudaalcaligenes KF707. J Biol Chem 267: 4844-4853.

Vincent, J.M. (1970) A manual for the practical study of root-nodule bacteria. IBP
Handbook 1 5. Oxford: Blackwell Scientific Publications.

Wilson, K.J., Sessitsch, A., Corbo, J.C., Giller, K.E., Akkermans, A.D.L., and Jefferson,
R.A. (1995) B—Glucuronidase (GUS) transposons for ecological and genetic
studies of rhizobia and other gram-negative bacteria. Microbiology 141: 1691-
1705.

CHAPTER 7

Conclusions and Future Directions

141

1 4 2
CONCLUSIONS AND FUTURE DIRECTIONS

Throughout this thesis it has been stated, and reiterated, just how harsh
and oligotrophic the soil environment can be, and how bacteria need to sense and
respond to environmental signals in order to persist and compete in their
environment. Microbial ecologists have traditionally taken two approaches to
studying microorganisms in the soil: (i) an ecological approach marked by the
examination of bacteria directly and how they compete with other microorganisms
for resources and survive in the midst of harsh environmental conditions, and (ii) a
molecular approach marked by the examination of individual enzymes necessary
for the cell to survive and compete successfully in its environment and how the
individual genes for these enzymes are regulated. I have tried to combine the two
approaches in the hope that they would complement one another in the elucidation
of not only genes and their regulators, but in the understanding of the role of such
genes for competition and survival in the natural environment.

At the outset of the research described in this thesis, very little was known
about environmentally regulated gene expression in soil bacteria, and only a little
more was known in non-soil bacteria. Since that time, nearly seventy loci have been
cloned and sequenced in R. meliloti which respond to nutrient or oxygen limitation
conditions (chapters 2-4), and some of the Tn5-1063 tagged loci have been
examined for their importance in the persistence and competition of the strains
harboring them in the soil and in nodulation competitiveness (chapters 5 and 6).

The large number of genes isolated in response to nutrient and oxygen
limitation is comparable to the number of proteins found to be induced by stresses
in other systems (Matin, 1991; Givskov et al., 1994). However, the lack of overlap
of genes induced by multiple-stresses is decidedly different than what had been

seen in non-soil isolates, and supports the hypothesis, presented in chapter 1, that

1 4 3
soil bacteria have a different means of regulating gene expression in response to

environmental stresses than non-soil bacteria. This interesting hypothesis is

discussed in greater detail below.

Overlap of loci induced by multiple stresses

In this collection of 69 mutants with gene fusions induced by nutrient
deprivation and microaerobiosis, only one locus (tagged in strains C4 and C22, see
chapters 3 and 4) was found to be induced by all three stresses, and only two other
loci (tagged in strains N4 and C101, see chapters 2 and 3) were found to be
induced by two stresses, in contrast to what has been seen in other systems, such
as E. coli and S. typhimurium, in which the presence of deprivation-induced
proteins was screened via the use of 2D-PAGE (Matin, 1991).

Figure 7-1 shows a ven diagram adapted from Matin (1991) that compares
the overlap of proteins induced by three stresses in E. coli with the overlap of Tn5-
tagged loci regulated by three stresses in R. meliloti. It is immediately apparent
that there is a large difference in the number of overlapping proteins in E. coli
compared to the very few overlapping genes in R. meliloti. This difference could be
due to an inherent difference in deprivation-induced gene regulation between R.
meliloti and other non-soil type bacteria, or due to the difference in selecting for
Tn5-tagged promoter induction versus protein induction. Two lines of reasoning
suggest the latter. First, the use of reporter systems lacZ and luxAB in S.
typhimurium and P. fluorescens, respectively, yielded similar results to ours with
the isolation of few mutants induced by multiple stresses (Spector et al., 1988;
Kragelund et al., 1995). Secondly, the similarity between the number of proteins

induced by carbon deprivation in P. putida, a common soil isolate, to other systems

144

Figure 7-1 . Ven diagrams illustrating the overlap of lad induced by multiple
stresses. The overlap of proteins in E. coli induced by three stresses,
and the Tn5-1063 tagged loci in R. meliloti induced by three stresses,
is shown. Each number represents either an induced protein, in the

case of E. coli, or a Tn5-tagged locus in the case of R. meliloti . E. coli
data are from Matin (1991).

145

.v. .9}:

M 50365 33535

., .ammwaI H.328: NO

R335 /
38 as 3

EN EN SN 3. /
9. am. we. 3.. /

.56than
5on Z

‘\
\ mm...c..o2
.\ o... .R. .2.

               

. m...~......
. o: .3. .9. 3.. .mm. .3. .8.
.. 3.2. 8..cw..w.~;
N.,...x $.32 \
/ 3... 3. a. .o.
/ 0.4 \
. \
/. ._
.\

/ E .5. .2.
/ S. .3. .R \

//.., o. .w. ._ \

cougtnmomonumu

So. .532 88a 6833. ..

II I II I I, I: EWIIIVIIItiJ

AMI :8 Snuimfimm.

 

 

=oum>tmog ,, ..\I. I cosmzumoa
coonZ _, ,. ,, 2.5.38.5

\ / ,

. \ x/ .,/

\ .Ne...o. _\ 2.62 / 02.8. /

_... 3. 6: mm. \ mesa. / .,,,

\ Edie: \ ., mm. 3.
E.A.,»: \III/flp, $35.8. ,
m...~...c..\/ 2.2.x... ..

S. 8. \ ,,, v2.5.5 . 8;:

/ E on 24.33.. . , . .

/ S \3... .33, 3.2.... __\ S. / w: \\
/, . mo. 8. .0... f. 8.5m. .. \
4 N« R w. .. 8. .818. .,,.\

x/I 2,\\

VIII, ,
... 332$:saazafidm. .
, mm. .2. .9: .w: .3 a:
// STBATNMSdN \
/ 8.8 .m. .N. .w... \\
/ \.

// III\\
gouging. c0980

1 46
such as E. coli (Givskov et al., 1994), suggests that the use of Tn5-luxAB versus 2D-

PAGE protein analysis is the main reason for these differences. This may be
explained in several ways: (i) The loci may be induced at such a low level that they
could not be visualized with luciferase as a reporter; (ii) insertion of Tn5 in these
loci may be lethal; (iii) the number of mutants generated may not cover the full
extent of the genome; or, (iv) regulation of these proteins may be primarily at the
post-transcriptional level.

Any of these reasons, or a combination of them, is plausible. In this study,
many of the fusions were induced at a low level, but were not retained for further
study because low-level induction was hard to reproduce accurately. It is also likely
that the 5000 mutants generated was not enough to completely saturate the
genome. Based on the size of the R. meliloti genome (6.5 Mb; Honeycutt et al.,
1993), and the average number of genes found in bacterial genomes thus far
(1743 predicted coding regions in Haemophilus influenzae [1.8 Mb; Fleischmann et
al., 1995]; 1738 in Methanacaccus jannashii [1.6 Mb; Bult et al., 1996]),
approximately 1 Kb per coding region in R. meliloti is estimated, yielding an
estimate of 6500 total loci. Given a random distribution of Tn5-1063 insertions in
the genome, and the fact that the Tn5-1063 must insert in the proper orientation
to generate luxAB fusions, It is estimated that 43,290 individual Tn5-1063
insertions would be necessary to achieve a 99% confidence level (z=2.33) of one hit
per gene in the proper orientation, thus approaching saturation of the genome.
However, given the highly polar nature of Tn5 mutations, and the high occurance
of multiple lad regulated by a single promoter (regulons), many fewer insertions
would be necessary to ensure that all regulons were mutagenized

In addition, post-transcriptional regulation of gene expression might occur
during many deprivation-induced conditions. The observed instability of bacterial

mRNA has led to the suggestion that this would provide the bacterial cell with a

1 47
very sensitive method of dealing with sudden environmental changes (Gros et al.,

1961; Jacob and Monod, 1961), and during non-deprivation conditions the mRNA
would be quickly recycled. Although much less studied in bacteria, more and more
proteins, including a lipoprotein encoded by lpp, the outer membrane protein
OmpR, and thioredoxin TrxA, are being shown to be regulated via mRNA
degradation (O’Hara et al., 1995). In addition, mRNA stability has been shown to
be regulated via polyadenylation in ribosomal protein 815 encoded by rpsO
(Hajnsdorf et al., 1995) and RNA 1, which encodes the antisense repressor of
replication of ColEl-type plasmids (Xu et al., 1993; Xu and Cohen, 1995). Thus,
mRNA stability is probably highly underestimated as a mechanism of protein

regulation in bacteria (Hajnsdorf et al., 1994).

Importance of Tn5-1063 tagged genes regulated by nutrient deprivation for
persistence and competition in the soil and in nodule occupancy

Perhaps the overall guiding theme of this thesis has been the recurring
contrast of soil microorganisms versus non-soil microorganisms. Because of the
harsh, oligotrophic conditions of the soil environment, it seems clear that bacterial
genes regulated in response to environmental stresses might be very important to
the microorganism when competing for available resources and niches. In addition,
the hypothesis that such genes might be different, and/or differently regulated in
soil bacteria compared with non-soil enterics or aquatic microorganisms, is
compelling. In chapter 5 of this thesis, the competition of selected strains—
harboring Tn5-1063 insertions in nutrient regulated loci—versus the wild-type
strain 1021 supports the hypothesis that the tagged genes are important for
competition and survival when the soil is lacking in nutrients. The nutrient poor soil

used in this study was a sandy soil that is typical of many native soils such as in

1 4 8
Eastern Michigan. Thus, conditions even within this restricted environment reflect

real soil environments, and lend even more weight to the observed results.

In contrast to the soil, competition for nodulation experiments yielded
unexpected and surprising results which go counter to the hypothesis presented in
chapter 1. However, the conditions of the experiment may not have presented a
sufficient environmental challenge to the tested strains, and more work needs to
be done with pre-deprivatian conditions before any final conclusions can be drawn.
Nevertheless, the apparent increase in nodulation competitiveness of strains C22
and C101 (both harboring gene fusions regulated by multiple stress conditions)
versus the wild-type strain suggests that genes involved in global regulation of
stress induced pathways may carry a competitive burden under certain
circumstances that is apparently outweighed by its potential to protect the cell

while in harsh, oligotrophic, naturally occurring soil and rhizosphere environments.

Search for Regulators Responsive to Stress Conditions

The next obvious step in this research is to isolate and characterize the
regulators of the tagged loci described in this thesis. Carbon, nitrogen and oxgyen
limitation responsive regulators have already been described in many bacteria
including R. meloliti (see chapters 2-4; Szeto et al., 1987; Matin, 1991; Fischer,
1994; Arcondeguy et al., 1997). It would be of interest to see if the Tn5-1063
tagged loci described in this thesis are also regulated by any of the the known two-
component (e.g., FixL/FixJ or NtrB/NtrC) regulators or, in the case of carbon-
deprivation-induced genes, the cAMP/CRP regulators. Tn5-1063 tagged loci not
regulated by any of these described systems would then be the most interesting
candidates for a secondary mutagenesis approach in the isolation of the unknown

regulatory proteins.

149
Researchers in the Frans de Bruijn lab (Anne Milcamps and Mary Ellen

Davies) have already begun to look for regulators of selected carbon and nitrogen
deprivation-induced loci, using the mini-Tn3 transposon. Random insertion of this
transposon within the genome of the strain of interest, should lead to loss of luxAB
induction (no light production) if a positive regulator controlling this promoter
fusion is inactivated via the insertion of mini-Tn3. Of course, insertion of the mini-
Tn3 within the luxAB genes would also lead to this phenotype, so Southern analysis
using luxAB and mini-Tn3—labeled probes would be necessary to distinguish
between these possibilities. Another regulator one might expect to see using this
method is a repressor of the tagged gene of interest. Insertional inactivation of a
repressor would lead to a constitutively expressing phenotype (always producing
light). A last possiblility is that such an insertion would occur in a regulator at the
top of a cascade of regulators leading to a pleiotropic phenotype that would be
hard to characterize, but extremely interesting.

The elucidation, via sequence comparison, of many of the Tn5-1063 tagged
genes has provided clues into the physiological response of R. meliloti while
subjected to stress conditions, and may in turn help lead to a better understanding
of how soil microoganisms survive in such harsh environments. In addition, the
tagged genes and their promoters may be useful markers to test the nutrient
conditions of a particular environment of interest, or to regulate genes of interest
(such as for biodegradation or plant growth) in the soil and rhizosphere.

Taken together, the data presented in this thesis answer many of the
questions posed in chapter 1, and provide a definite direction for the continued
examination of the overall hypothesis that soil bacteria have a different set of
genes and regulators required for sensing and responding to environmental

stresses.

150
REFERENCES

Alexander, M. (1977) In Introduction to Sail Microbiology, 2nd ed. New York: John
Wiley & Sons, pp. 16-51.

Arcondeguy, T., Huez, l., Tillard, P., Gangneux, C., de Billy, F., Gojon, A., Truchet, G.,
and Kahn, D. (1997) The Rhizobium meliloti PII protein, which controls bacterial
nitrogen metabolism, affects alfalfa nodule development. Genes Dev 11: 1 194—
1 2 06.

Bult et al. (1996) Complete genome sequence of the methanogenic arcaeon,
Methanacaccus jannaschii. Science 273: 1058—1073.

Fischer, H.-M. (1994) Genetic regulation of nitrogen fixation in rhizobia. Microb Rev
58:352-386.

Fleischmann et al. (1995) Whole-genome random sequencing and assembly of
Haemophilus influenzae Rd. Science 269: 496-512.

Givskov, M., Eberl, L., and Molin, S. (1994) Responses to nutrient starvation in
Pseudamanas putida KT2 442: two-dimensional electrophoretic analysis of
starvation- and stress-induced proteins. J Bacterial 176: 4816-4824.

Gros F., Hiatt, H., Gilbert, W., Kurland, C.G., Risebrough, R.W., and Watson, JD.
(1961) Nature 190: 581—585.

Hajnsdorf, 13., Braun, F., Haugel-Nielsen, J., and Regnier, P. (1995)
Polyadenylylation destabilizes the rpsO mRNA of Escherichia coli. Prac Natl Acad
Sci USA 92: 3973—3977.

Hajnsdorf, E., Steier, O., Coscoy, L., Teysset, L., and Regnier, P. (1994) Roles of
RNase E, RNase II and PNPase in the degradation of the rpsO transcripts of
Escherichia coli: stabilizing function of RNase II and evidence for efficient
degradation in an ams pnp rnb mutant. EMBO J 13: 3368—3377.

Honeycutt, R.J., McClelland, M. and Sobral, B.W. (1993) Physical map of the genome
of Rhizobium meliloti 1021. J Bacterial 175: 6945-6952.

Jacob, F., and Monod, J. (1961) Genetic regulatory mechanisms in the synthesis of
proteins. J Mol Biol 3: 318-356.

Kragelund, L., Christoffersen, B., Nybroe, O. and de Bruijn, P.J. (1995) Isolation of
lux Reporter Gene Fusions in Pseudamanas fluorescens DF57 Inducible by
Nitrogen or Phosphorus Starvation. FEMS Microbial Ecol 17: 95—106.

Matin, A. (1991) The molecular basis of carbon-starvation-induced general
resistance in Escherichia coli. Mal Microbial 5: 3-10.

O’Hara, F..B., Chekanova, J.A., Ingle, C.A., Kushner, Z.R., Peter, E., and Kushner, SR.
(1995) Polyadenylylation helps regulate mRNA decay in Escherichia coli. Prac
Natl Acad Sci USA 92: 1807—181 1.

151

Spector, M.P., Park, Y.K., Tirgari, S., Gonzalez, T. and Foster, J.W. (1988)
Identification and characterization of starvation-regulated genetic loci in
Salmonella typhimurium by using Mu d-directed lacZ operon fusions. J Bacterial
170: 345—351.

Szeto, W.W., Nixon, B.T., Ronson, C.W., and Ausubel, FM. (1987) Identification and
characterization of the Rhizobium meliloti ntrC gene: R. meliloti has separate
regulatory pathways for activation of nitrogen fixation genes in free-living and
symbiotic cells. J Bacterial 169: 1423-1432.

Xu, F., and Cohen, SN. (1995) RNA degradation in Escherichia coli regulated by 3'
adenylation and 5' phosphorylation. Nature 374: 180—183.

Xu, F., Lin-Chao, S., and Cohen, SN. (1993) The Escherichia coli pan gene
promotes adenylylation of antisense RNAI of ColEl-type plasmids in vivo and
degradation of RNAI decay intermediates. Prac Natl Acad Sci USA 90: 6756—
6760.