 

 

 

 

 

 

2. ,. .25...
. 5.1.. . 3..
«3......» gr aim. arr;
35.5.... ._.

. . 1;...aenuduun3h

v 11 I013}. .11..

’ I
a Ivy... .qu .u.

(f
1 11...: I a

at

8.3..

at")...
535?.
xtvziti
a!!!) Y: .
39...!
\ 7...: #2.?
. .. :
:‘I. a: r...“
...t«.\; )1.

‘1‘
E

14.5... ..
a 31.2.: it.
3.. .. 1. ”Want... u L
. ..

)1»
lizltxﬂ-
., 1.3.1355;
...uf...i.f 5:! Z
. 3

.. v.3
...

i! U.
'33..“
345...... 15:11
7. ..

. ~ .
r )9. 2.331%}:
.L.aa...$r.}0:-u.tlru natal,
.3? I. .5"
3L1}. ’11:: 11.. 9 «3’.le
.1335: 5:! SI . . f‘ﬂigiﬂﬁo‘smilh
. 2...“. ... i. a: I
at J... I D

:‘l
|.S\.rl¢| A-
)tllztsi’st.
(57.1.1.5.1 :l \l
.2... 3.}.1.
nt!.(::.

13L |t|zl
7:. 2......

 

LlBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled
Impact of host population and catabolic
plasmid on the fitness of 2,4-dichloro-
phenoxyacetic acid degrading bacteria

presented by

Kirsti Maria Ritalahti

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degree in My

 

 

 

Date 11 May, 2000

 

 

M Ico’wa/OJJMV

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

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m cICIRCIDleDuepGS—p."

 

 

IMPACT OF HOST POPULATION AND CATABOLIC PLASMID ON
THE FITNESS OF 2,4-DICHLOROPHENOXYACETIC ACID
DEGRADING BACTERIA

lBy

Kirsti Maria Ritalahti

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Microbiology
Center for Microbial Ecology

hAay,2000

ABSTRACT
IMPACT OF HOST POPULATION AND CATABOLIC PLASMID ON

THE FITNESS OF 2,4-DICHLOROPHENOXYACETIC ACID
DEGRADING BACTERIA

By
Kirsti Maria Ritalahti

The acquisition of catabolic plasmids and genes carried by these
plasmids affect the competitive ability of bacterial populations. Competition
experiments between two different 2,4-dichlorophenoxyacetic acid (2,4-D)
degrading bacteria (Burkholderia cepacia strain DBO] and Ralstonia
eutropha strain JMP228), carrying either of two different 2,4-D degradative
plasmids (pJP4 and pEMT3), were performed. The two plasmids that
carried different gene alleles for the 2,4-D degradative pathway endowed
their bacterial hosts with different growth rates on 2,4-D. The grth rate
and half-saturation constants determined for pure cultures were used to
predict the competitively superior strain. Both host-associated and plasmid-
associated traits inﬂuenced the outcome of competition for 2,4—D under
different growth conditions. Host-associated traits had a pronounced effect
on competitive outcome. In batch culture competition experiments
characterized by a high initial 2,4-D concentration, Ralstonia populations
became dominant. When hosting the same plasmid, Ralstonia exhibited a

faster growth rate than Burkholderia when 2,4-D was provided as the

limiting resource. In contrast, Burkholderia was more successful under
continuous culture conditions characterized by very slow growth rates,
where the populations were forced to grow at less than half of their
maximum growth rates on 2,4-D. Plasmid-associated characteristics also
inﬂuenced the competitive outcome, particularly in the batch culture
environment. Harboring a particular plasmid was of importance when its
presence provided the increased beneﬁt of a higher growth rate.
Consistently in intra-species competitions, the presence of pJP4 endowed a
competitive advantage over strains with pEMT3 under batch culture
conditions, and the pJP4 carrying strain became dominant. However, intra-
species competition experiments in continuous culture demonstrated that
identical strains bearing different plasmids coexisted, suggesting that p] P4
confers considerably less of a selective advantage under low nutrient
conditions. Different environments impose different selective pressures.
The populations that compete under different environmental conditions
adapt to lifestyles that are advantageous in one habitat, and less desirable for
another. The ability to compete can be inﬂuenced by the presence of

different catabolic pathways for the utilization of a speciﬁc substrate.

I dedicate this work to the memory of my father, Tapani J. Ritalahti,
and to my mother Susan, who encouraged me to "follow my bliss". Their
inspiration has instilled in me a drive to never quit. I want to thank them

both for believing in my ability to succeed.

iv

ACKNOWLEDGMENT

I would like to thank the Center for Microbial Ecology and
Department of Microbiology at Michigan State University and NSF grant
#DEB912006 to the Center for Microbial Ecology, the sources which have
funded my Ph.D. dissertation research. I want to thank Dr. Terry L.
Marsh and Dr. Frank E. Ldfﬂer, for scientiﬁc discussion, critical
proofreading and moral support. I would like to thank the members of
my Graduate committee, Dr. Mike Klug, Dr. Pat Oriel, Dr. Rosetta Reusch
and Dr. Rich Lenski for their review of the work, and Dr. Larry J.
Fomey, for opening the door to exploring in the Research on Microbial
Evolution (ROME) Lab group. I’d like to extend thanks to Dr. Michael
Roberts, who has always lent an ear and answered my innumerable
questions. And ﬁnally, thanks to all of my colleagues in the ROME lab,

who I worked with, over the past years.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................ viii

LIST OF FIGURES ...................................................................... x

INTRODUCTION ......................................................................... 1

Role of Catabolic Genes and Host Genotype in Competition ............... 1

CHAPTER 1:

Competition Theory and 2,4-D Degradation ..................................... 9
Competition for Limited Resources ........................................ 9

2,4-dichlorophenoxyacetic Acid Degradation as a Model
System for Studying Interactions among Bacterial

Populations ....................................................................... 20
References ........................................................................... 38
CHAPTER 2:

The Impact of Similar Degradative Genes on Different Plasmids to
the Fitness of Isogenic Ralstom’a eutropha Strains Competing for

2,4-Dichlorophenoxyacetic Acid ........................ , ............................ 44
Abstract ............................................................................... 44
Introduction ......................................................................... 46
Materials and Methods .......................................................... 52
Results ................................................................................ 67
Discussion ........................................................................... 78
References ........................................................................... 81

CHAPTER 3:

Ralstonia eutropha and Burkholderia cepacia Host Populations as
Reservoirs for 2,4-Dichlorophenoyacetic Acid Degradative Plasmids... 83

Summary ............................................................................. 83
Introduction ......................................................................... 85
Materials and Methods .......................................................... 90
Results ................................................................................ 108
Discussion ........................................................................... 127
References ........................................................................... 136

CHAPTER 4:
Two Hosts and Two Plasmids in Simple Environments: The inﬂuence of
2,4-D Degradative Plasmids to the Competitive Fitness of a Host

Strain ........................................................................................... 137
Abstract ............................................................................... 137
Introduction ......................................................................... 139
Materials and Methods .......................................................... 148
Results ................................................................................ 160
Discussion ........................................................................... 194
References ........................................................................... 207

CONCLUDING REMARKS ......................................................... 211

LIST OF TABLES

CHAPTER 1
Table 1.1. Experimental evidence for microbial competition..........

Table 1.2. 2,4-D degrading populations carry mosaics of catabolic
genes ..................................................................

Table 1.3. Diversity of 2,4-D degrading isolates. ........................

CHAPTER 2

Table 2.1. Genotypic and phenotypic comparison of WpJP4, pEMTl
and pEMT3... .. ..

Table 2.2. Bacterial strains used to determine the inﬂuence of the
plasmid backbone on competition ..........................

Table 2.3. Plasmids used to determine the outcome of competition
in isogenic populations .........................................

Table 2.4. Transconjugants obtained with the ﬁlter mating
technique ............................................................

Table 2.5. Growth rates of parent strains and transconjugants on
succinate and 2,4-D ..............................................

Table 2.6. Carrying capacity of R. eutropha strains in batch and
chemostat conditions ............................................

CHAPTER 3
Table 3.1. Bacterial Strains used in the competition experiments .....

Table 3.2. Plasmids used in the competition experiments........... . . . .
Table 3. 3. Transconjugants used in the competition experiments ........
Table 3. 4. Testing the antibiotic resistance proﬁles of strains. ..
Table 3.5. BIOLOG substrate utilization by Ralstonia eutropha
JMP228 and Burkholderia cepacia DBOI ................

Table 3.6. Growth rates of strains supplied with succinate or 2,4-D
under batch culture conditions ...............................

Table 3.7. One-tailed t-test on growth rates of strains determined
under batch culture conditions ...........................................

13

31
34

47

54

55

56

69

76

93

94
96
108

111

115

117

Table 3.8. Parameters used to calculate expected grth curves
based on the Monod model, using Stella Research
Software ............................................................................ 122

CHAPTER 4

Table 4.]. Pure culture characteristics for strains grown on mineral 148
salts medium with succinate or 2,4-D under batch
culture conditions .................................................

Table 4.2. Competition expenments 152
Table 4.3. Values for constants in the Stella model ........................ 177

n ‘l-Au1

 

LIST OF FIGURES

CHAPTER 1

Figure 1.1. Competitive edge due to a lower resource threshold
concentration(R*)...

Figure 1.2. 2 ,4-D degradation pathway of plasmid pJP4 of R.
eutropha JMP.134. ............................................................

CHAPTER 2

Figure 2.1. Bgl H restriction analysis of plasmids isolated from
nansconjugant strains. .......................................................

Figure 2.2. Schematic representation of the chemostat used in the
competition experiments under low nutrient
conditions .............................................................................

Figure 2. 3. Growth of purem cultures carrying plasmid pJP4 or "pEMTl
on 2 ,4—D....

Figure 2.4. Competition among R. eutropha strains JMP228/pJP4, and
JMP228/pEMTl on 200 ppm 2,4-D.............

Figure 2.5. Ralstonia eutropha hosting pJP4 or pEMTl are equal
competitors against Burkholderia cepacia hosting
pJP4 ..............................................................................

Figure 2.6. R. eutropha with pJP4 or pEMTl in pure culture or in
competition. ........................................................................

CHAPTER 3
Figure 3.1. Restriction of plasmids isolated from transconjugants and
cut with enzyme EcoRl ..........................................

Figure 3.2. 2, 4-D depletion curves used to calculate rates of 2, 4-D
degradation for use in calculating KS for strain

Figure 3.3. Lineweaver-Burke plot” to determine the KS for each of the
four host/plasmid combinations.............

Figure 3.4. Growth model developed on the Stella Research Program v
5.0 to generate expected population growth curves and
2,4-D depletion curves for pure cultures........ . . .

Figure 3.5. Average population sizes for selected strains grown in
batch culture under 24 and 72 hours transfer
conditions ............................................................................

18

23

57

66

70

74

75

77

97

103

104

107

119

Figure 3.6. Allocation of ring-labeled 14C-2,4-D in growing cultures
of JMP228/pJP4, JMP228/pEMT3, DBOl/pJP4 and
DB01/pEMT3 ........................................................

Figure 3.7. Chemostat growth of JMP228R/pJP4, JMP2288/pJP4,
JMP228R/pEMT3 and JMP228S/pEMT3 ..................

Figure 3.8. Chemostat Growth of DBOlR/pJP4, DBOlT/pJP4,
DBO 1 R/pEMT3 and DBO l T/pEMT3 ........................

CHAPTER4
Figure 4.1. An overview of the genetic contribution of different genes
in the competition experiments................... . . . . .. . .. . . . . . . .

Figure 4.2. Growth of JMP228/pEMT3 and JMP228/pJP4 competing
for 5 mM succinate as a limiting resource ...................

Figure 4.3. Plasmids pEMT3 and pJP4 do not govern ﬁtness of R.
eutropha competing for succinate .....................................

Figure 4. 4. JMP228 is a superior competitor to DB01 when succinate
issuppliedasthelimitingresouroe..."

Figure 4. 5.1n of the population sizes of JMP228/DB01 when succinate

is supplied as the limiting resource...

Figure 4. 6. Effect of antibiotic resistance markers to ﬁtness of
JMP228/pJP4 and JMP228/pEMT3....

Figure 4. 7. Competition between JMP22.8/I.){.I.)4 and JMP228/pEMT3
for 2 ,4—D 1n batch culture... ..

Figure 4.8. Effect of antibiotic resistance markers on ﬁtness of
DB01/pJP2 and DB01/pEMT3 ..........................................

Figure 4.9. Competition between DB01/pJP4 and DB01/pEMT3 for
2,4—D in batch culture ............................................................

Figure 4.10. Competition between JMP228/pJP4 and DB01/pJP4 for
2 ,4—D in batch culture .........................................................

Figure 4.11. Competition between JMP228/pEMT3m and DBOl/ pEMT3
for 2 ,4—D 1n batch culture... .... ..

Figure 4.12. Competition between JMP228/pJP4. and DB01/pEMT3
for 2 ,4—D in batch culture...

Figure 4.13. Competition between JMP228/pEMT3 and mDB01/pJP4
for 2 ,4-D in batch culture... ... .

122

124

125

142
162
162
164
164
165
167
168
169
171
172
173

174

Figure 4.14. The model that was used by the Stella Research Program
to generate expected population growth curves for two
competing populations.............. . . .

Figure 4.15. Chemostat competition between JMP228S/pJP4 and

JMP228R/pEMI‘3 .............................................................. .
Figure 4.16. Chemostat competition between JMP228R/pJP4 and
JMP2288/pEM'I‘3 ................................................................
Figure 4.17. Chemostat competition between DB01T/pJP4 and
DB01R/pEMT3 ...................................................................
Figure 4.18. Chemostat competition between JMP2288/pJP4 and
DB01R/pJP4 ........................................................................

Figure 4.19. Different KS values for DB01 inﬂuence competition
between JMP228S/pJP4 and DB01R/pJP4 in

chemostats ...........................................................................

Figure 4.20. Chemostat competition between JMP228S/pJP4 and
DBOlR/pJP4 ........................................................................

Figure 4.21. Chemostat competition between JMP228R/pEMT3 and
DB01T/pEMT3 ....................................................................

Figure 4.22. Chemostat competition between JMP228R/pEMT3 and
DB01T/pJP4 .........................................................................

Figure 4.23. Chemostat competition between JMP228R/pEMT3 and
DBOlT/pEMT3 ....................................................................

176

180

181

182

184

186

188

189

191

192

INTRODUCTION

ROLE OF CATABOLIC GENES AND HOST GENOTYPE IN COMPETITION

Competition for limiting resources in the environment has captured
the attention of ecologists and microbiologists. In both disciplines
competition has been implicated as a force driving speciation, promoting
extinction, and inﬂuencing the distribution of populations that rely on the
same set of resources. The introduction of man-made compounds to the
environment potentially changes the balance of functional ecological
systems at the microbial and macro community level. Microbial community
changes may be due to negative (e.g. toxicity) or positive (e.g. enrichment)
effects of the introduced chemical. When a toxic compound is applied,
susceptible populations decline in numbers relative to resistant ones. The
increase in the relative abundance of resistant populations produces a shift
in the community composition. In most environments, microbial
communities are responsible for the mineralization of many natural and
anthropogenic compounds, and many compunds are used as a source of
carbon and/or energy. Populations that are capable of utilizing
anthropogenic compounds gain a distinct growth advantage over

populations that cannot use the resource. The result may be observed as an

enrichment of populations that are capable of beneﬁting from the
compound. Some of the populations that grow with the introduced
compound may become involved in a competitive interaction with other
populations that can degrade the same compound. Even among very simple
communities comprised of only a few populations, our interpretation of
changes in population dynamics is still rudimentary. When changes in
community composition occur, an comprehensive view of the factors that
inﬂuence the dominant population, include the host population, its genetics
and the surrounding environment.

The rate at which a population grows and the efﬁciency at which it
utilizes a resource are phenotypic features that inﬂuence competitive ﬁtness.
Among sexually reproducing organisms it has been suggested that some
components of competitive ﬁtness relate to the manner in which populations
reproduce. The mode of reproduction, including grth rate, age of
reproductive maturity and the amount of time during which a female is
reproductive, as well as the allocation of energy to producing and raising
offspring, contn'bute to the “life history” of a population. Asexually
reproducing populations, like bacteria, face a similar set of criteria when the
relationship of life history to competitive ﬁtness is considered. The rate of
reproduction, (grth rate), the time at which division begins (lag time) and
the efﬁciency of resource use (allocation to reproduction) can be used to

deﬁne competitive properties of microbial populations.

2

Bacteria are prime candidates for use as model organisms to test
competition theory among asexual populations. Bacterial populations have
advantageous properties because (i) they attain high cell densities relative to
sexually reproducing organisms, (ii) they exhibit very short generation
times, and (iii) they are distinguished by easily selectable phenotypic
markers, which may be naturally present or genetically introduced.

The many phenotypic differences in the growth characteristics of
bacterial populations are due to differences in their genetic makeup. Genes
residing on the chromosome contribute to general metabolism, substrate
range and reproduction. While bacteria are considered asexual, novel
genetic arrangements can be created by lateral gene transfer. Plasmids are
mobile, extrachromosomal genetic elements that can disburse or maintain a
variety of different traits in bacteria. They can supply catabolic pathways,
such as for the degradation of herbicides, pesticides and aromatic solvents,
or can convey antibiotic resistance. The acquisition of broad host range
plasmids and the catabolic genes carried by these plasmids (such as for 2,4-
dichlorophenoxyacetic acid, toluene and 3-chlorobenzoate degradation)
affects the competitive ability of different bacterial populations when the
novel substrate is provided as a limiting resource. Since there exists a
correlation between a population's genotype and its ﬁtness, it is reasonable
to assert that when new catabolic genes are acquired, the ﬁtness of its host

may become altered. Recruiting different catabolic plasmids among

3

 

_4___

members of the same host species results in genetically distinct populations.
Subpopulations with different plasmids, or different genes may be
competitively equal in the absence of selection, but may result in a dominant
population when selection for a speciﬁc plasmid-encoded catabolic pathway
occurs.

Bacteria are ideal model systems for examining aspects of inter-
speciﬁc and intra-speciﬁc competition. This work focuses on the well-
characterized degradation pathway for the herbicide 2,4-
dichlorophenoxyacetic acid (2,4-D) and two different 2,4-D-degrading
bacterial species. Ralstonia eutropha strain JMP228 and Burkholderia
cepacia strain DB01 were used as host populations for different catabolic
plasmids carrying alternative gene alleles for 2,4-D degradation. Previous
research indicated that the application of 2,4-D to agricultural sites resulted
in the enrichment of speciﬁc 2,4-D degrading bacterial populations. These
ﬁndings suggested that different 2,4-D genotypes contributed to the
development of the dominant population in soils where 2,4-D had been
applied. Since it is likely that alternative degradative genotypes for a
particular substrate impact different bacterial populations in different ways,
we can determine the ﬁtness advantage of carrying different catabolic
plasmids to both inter-speciﬁc and intra-speciﬁc competition, when the

ability to utilize the limiting substrate resides on a catabolic plasmid.

 

Chapter 1 reviews information about population growth, competition, and
the characteristics of the 2,4-D degradative pathway.

The ﬁrst hypothesis to be tested (Chapter 2) is that in host strain, R.
eutropha, different plasmid backbones with similar alleles for 2,4-D
degradation, result in competitively equivalent populations. The two
plasmids, pJP4 and pEMTl, carry homologous alleles for 2,4-D degradative
genes in a different plasmid background. Within the R. eutropha host,
similar 2,4-D degradative alleles on different plasmid backbones resulted in
similar growth and 2,4-D degradation rates. Competition studies with the
two strains under batch culture conditions demonstrated that the strains
coexisted, which suggested that the inﬂuence of the plasmid backbone was
minimal under non-limiting resource conditions (2,4-D was in excess).

The second hypothesis tested whether possessing different plasmids
and degradative alleles affected competition within the same host population
under batch and chemostat conditions. Both R. eutropha strain JMP228 and
B. cepacia strain DB01, each of which carried the two different 2,4-D
degradative plasmids (pJP4 or pEMT3) were examined. Plasmids pJP4 and
pEMT3 encode for different alleles of the 2,4-D degradation pathway,
expressed from similar plasmid backbones. Prior to carrying out
competition experiments, pure culture growth data were obtained, and the

differences between R. eutropha and B. cepacia were explored (Chapter 3).

 

IE an;

While the growth rates on 2,4-D were different for all four strains, there
was no measured difference in the half-saturation constants.

In batch competition experiments of Chapter 4, the population with
the fastest growth rate became numerically dominant. Among members of
the same host species, plasmid pJP4 conferred dominance to its host, and
the pEMT3 carrying strain was out competed. In controlled, low growth
rate conditions of a chemostat, members of the same host population with
either set of alleles coexisted and the selective advantage of the fast growth
rate conferred by pJP4 was not as signiﬁcant.

The third hypothesis was to determine if carrying the same plasmid
and allele combination affected competition among members of different
host species under two different grth conditions. Based on the pure
culture grth parameters determined in Chapter 3, R.. eutropha strains
were predicted to dominate in batch culture due to having the faster growth
rate. Experimental results in Chapter 4, indicate that when R.. eutropha and
B. cepacia both carried the same plasmid, be it pJP4 or pEMT3, R..
eutropha dominated under batch conditions. Conversely, B. cepacia became
dominant under chemostat growth conditions. Collectively, these results
suggested that the host population played a signiﬁcant role in controlling
population dynamics under different growth conditions.

The fourth hypothesis was to test if the presence of either plasmid or

allele combination would change the results of competition so that B.

6

' 'I 1.0“

 

cepacia could out compete R.. eutropha in batch culture, or if R.. eutropha
could out compete B. cepacia in the chemostats (Chapter 4). In batch
culture, the order of dominance correlated with the expected pattern
predicted by the growth rates: R. eutropha (pJP4) > B. cepacia (pEMT3)
and B. cepacia (pJP4) > R. eutropha (pEMT3). As in the previous
example, faster growth rates lead to dominance in batch culture
competitions. In contrast, in the chemostats with a low dilution rate (0.03 h‘
l), and low nutrient concentrations, the B. cepacia strains became
dominant, supporting evidence that the B. cepacia population is more
competitive under the low grth rate selected growth regime.

The data presented in the following chapters support the hypothesis
that phenotypic characteristics of a population inﬂuence success under
different environmental conditions. Different genotypes conferring
catabolic pathways with similar function, such as 2,4-D utilization, do
inﬂuence ﬁtness when competition occurs for that limiting substrate. Under
the conditions that were tested, dominance due to plasmid—bome alleles was
observed in the conditions where a faster grth rate was advantageous.
Experimental results described herein indicated that the host genotype in
concert with the appropriate environment played a signiﬁcant role in
determining dominance. The outcome of carrying different alleles for 2,4-D

degradation on a catabolic plasmid was important to intra-speciﬁc

[mg—ﬂ
1 3‘

competition. In highly similar host backgrounds, the strain carrying pJP4
became dominant.

Based on physiological differences, R.. eutropha and B. cepacia
species have slightly different life history strategies. While both species are
widely distributed opportunistic organisms, R.. eutropha uses a rapid
response to environmental stimuli and a fast growth rate to gain an
advantage under high nutrient conditions (e.g. batch transfers). On the
other hand, B. cepacia can use a lower resource threshold and hence, has
an ability to maintain high population densities at low resource

concentrations, (e.g. chemostat conditions).

 

CHAPTER 1:

COMPETITION THEORY AND 2,4-D DEGRADATION

COMPETITION FOR LIMITED RESOURCES

Populations and Community Diversity

Who wins and why has been a key question in ecological research for
decades. Some populations gain a prominent position within a community,
while others follow a road to extinction. Competition for limiting resources
has been considered a driving force in determining the distribution and
abundance of populations within a community. It has been implicated as a
force of evolution, and for driving the divergence of one species into many.
(Darwin, 1859; Arthur, 1982; Connell, 1978; Vadas, 1990). Much of the
foundation for research on competitive interactions was initiated in
eukaryotic systems. Darwin set the stage by suggesting that competition
was central to the struggle for existence (Darwin, 1859). His work on
ﬁnches was an example of evolution, where one ancestor gave rise to many
different lines of ﬁnches; each ﬁnch suited a particular habitat. In the years
that followed, experimental studies comparing ﬂuctuations of competing

populations were observed directly, and spread to include microbial

_ r..h‘£

 

competitions as well, (Table 1.1). Microbial systems are ideally suited to
study both interspeciﬁc and intraspeciﬁc competition. Several studies
demonstrated that limiting nutrient conditions are important factors in

determining dominance.

The inﬂuence of interspeciﬁc competition among animals and plants
was demonstrated in many ﬁeld studies, as summarized in reviews by
Schoener (1983) and Connell (1983). Both supported the notion that
competition was widespread, at least among the species under investigation.
Connell suggested that in many of the studied cases where interspeciﬁc
competition was demonstrated, a high possibility of ﬁnding the interaction
could have inﬂuenced the choice of the research subjects. In other words,
where the inﬂuence of competition was expected, it was often afﬁrmed.
Finally, the absence of present-day competition does not mean that
competition did not take place in the past. The composition of current
communities may be a reﬂection of the winners of prior competition, with
the losing populations driven to extinction, and the existing populations

coexisting due to minimal or balanced competitive interactions.

The “ghost of competition past” (Connell, 1983) is evidenced in our
current evaluation of microbial populations that degrade the herbicide 2,4-
dichlorophenoxyacetate (2,4-D). A common set of catabolic genes for

utilizing this herbicide exists in many of the dominant 2,4-D degrading

10

"I

populations. The presence of these alleles correlated with the prior
application history of the herbicide at a particular site (Ka et al., 1994a,
Dunbar et al., 1997). Further addition of the herbicide tended to enrich for
greater numbers of the populations carrying the favorable genotypes.The
catabolism of 2,4-D, and the populations that compete for this resource,

will be addressed later in this chapter.

Competitive Exclusion or Coexistence?

The research on competitive interactions in Paramecium lead to the
development of two key concepts in community interactions (Gause, 1934).
The ﬁrst principle was called “Competitive Exclusion”. Gause observed
that if a single resource is supplied to two populations of Paramecium, they
will compete for it. The one that uses the resource most rapidly will
displace the population that is less efﬁcient. Not every competitive
interaction inevitably leads to a complete extinction of a population, but may
lead instead to a signiﬁcant reduction in numbers. Simply stated, a
competitive interaction is one that decreases the growth and reproductive
rate, or increases the death rate, of each population due to a shared demand
on a common resource base. Competition may be due to exploitation or
interference. Competition by exploitation is due to consumption of a

resource, so that it is not available to a competitor population. Interference

11

competition involves a direct interaction of populations and may include the

use of toxins or force to eliminate an antagonist.

One principle of classical ecological theory states that two different
species can not coexist while competing for the same set of limiting
resources in a similar manner, (Gause, 1934). In order to avoid
competition, habitat differentiation can occur leading to coexistence and
increased diversity. Research on eukaryotic populations noted that instances
of cohabitation by two species of the same genera in the same habitat was
rare (Elton, 1946). In cases where cohabitation does occur, the populations
usually coexisted by utilizing different portions of the environment. Gause
(1934), noted that two different populations of Paramecium could coexist,
as long as one fed on yeast cells in the bottom of the culture vessel, and the
other fed on bacteria growing in the aerobic, upper portion of the vessel. If
both populations of Paramecium fed on both food types, competitive

exclusion would occur, assuming the resources were limiting.

12

Table 1.1. Experimental evidence for microbial competition.

Populations

Limiting
resource

details

Reference

 

Various species of Yeast, E. coli.

Paramecium.

Chromatium
weissei and
Chromatium
vinosum

Light in
batch and
chemostat

Heterotrophic and Ammonia and

autotrophic
nitriﬁers

High mutating
E. coli vs low
mutating E. coli

E. coli vs P.
aeruginosa

Methylosinus
trichosporium
OBBb and
Methylomonas
albus BGS

Ancestral vs
glucose evolved
lines of E. coli

oxygen
in chemostats

Glucose
in chemostats

Tryptophan
in chemostats

Methane
copper
nitrate

in chemostats

Glucose

Competitive exclusion among
some competitive pairs, Niche
differentiation among others

Continuous light selects for C.
vinosum, while intermittent light
allows for coexistence, since C.
weissei has a faster rate of sulﬁde
oxidation in the dark

Most cases showed no
interaction. Under high C/N, or
low oxygen, the heterotrophic
populations dominated.

High mutating strain wins in
chemostat if present at greater
than 7 x 10-5 starting ratio, due
to development and replacement
by faster growing mutants

Measurements of u, Ks and
calculated J, predicted the
outcome of competition.

M. albus dominant under
methane limitation, M.
Trichosporium under nitrate
limitation.

Derived strains had less lag, and
faster growth rates on glucose

13

Gause, 1934

Van Gemerden,
l 974

VanNiel, 1993

Chao and Cox,
1983

Hansen and

Hubbell 1980

Graham et a1,

1993

Lenski et al.
1 998

Microbial population dynamics

Microbial populations, especially the well studied bacterium
Escherichia coli (Levin, 1972; Lenski et al., 1994; Lenski et al., 1998.
Dykhuizen and Hartl, 1983), have served as USBﬁlI tools for testing specific
cases associated with the ﬁtness of populations and competition theory.
Frequently the behavior of populations can be modeled using some fairly
simple assumptions. For example, the logistic growth equation describes
population growth without the inﬂuence of a competitor. In early stages of
growth, growth proceeds exponentially, with the population increasing at a
maximum rate. As resources become scarce, there is an increase in
competitive pressure from individuals within the same population.
Population growth slows, and as a result, the population size plateaus at a
level that has been described as the carrying capacity of the environment.
Natural populations are often maintained well below their carrying capacities
due to predators or environmental perturbations.

The growth of a microbial population can be based on measurable
physiological traits. The two traits of particular relevance are the maximum
growth rate (umax), and the half saturation constant (K,) for the compound
supplied as a sole resource (Van Niel et al., 1993; Gottschal, 1992; Duetz et
al. 1994). The half saturation constant represents the resource
Concentration that allows a population to grow at half the maximum growth

rate. Often, the outcome of competition between two competing

14

populations was correlated with the relationship of these parameters (Duetz
et al., 1994). Using substrate-limited growth kinetics derived by Monod
(1949), it became increasingly common to perform single-resource
microbial competitions with outcomes in agreement with theory. The
Monod equation and resulting models are presented in Chapter 4.

As a life history strategy, some populations take advantage of a fast
maximum growth rate (a high pm), or a high reproductive output, when
resource concentrations are plentiful. Other populations maximize the
efﬁciency at which they utilize low concentrations of a substrate, and
therefore can grow and maintain viable population sizes even when
resources are limited. In the laboratory, the typical test tube or shake ﬂask
simulates a high growth-rate favored habitat. The ﬂask has a high initial
concentration of resource that is readily utilized. The population that grows
faster utilizes a greater proportion of the resource and increases its
population size accordingly. In contrast, a population with a very high
afﬁnity but a low growth rate on a substrate would maintain a viable
population size when nutrients are meager. A constant and low substrate
concentration, which leads to a low growth rate, can be simulated in the
laboratory with the use of a chemostat. Generally, chemostats are used to
maintain a continuously growing culture, where the ﬂow rate of the

Substrate into the culture vessel is constant.

15

In a continuous ﬂow environment, waste products are removed as
new resources are provided. The rate at which the populations divide is
controlled by the rate at which the substrate is provided (Gottchal, 1992;
Dykhuizen and Hartl, 1983). A population will divide at the maximum rate
allowed by the substrate provided, as long as the dilution rate (the rate at
which the culture volume is replaced) does not exceed the culture’s growth
rate. When the dilution rate exceeds the growth rate, a population will
become diluted out over time. Under chemostat conditions suited to a low
growth rate, the substrate concentration also tends to be extremely low,
because incoming nutrients are consumed immediately by the culture.
Consequently, Ks becomes more signiﬁcant than pm”. For two populations
with the same maximal growth rate, the one with a lower Ks value may
have a faster growth rate at a lower substrate concentration. The population
with a lower Ks value will reduce the concentration of substrate in the
environment to a level that will not allow its competitor to grow. For
example, if population A reduces the concentration of substrate below that
capable of supporting pouplation B, A will eventually become dominant, as
its competitor B starves, cannot reproduce, and is diluted out of the
chemostat. Thus, at low substrate concentrations, afﬁnity (Ks) for the
resource is crucial. Additionally, with low substrate concentrations, the
energy and carbon provided would ﬁrst have to be used for maintenance of

cell integrity, and less would be allocated to growth and cell division.

16

 

Expanding on the work of Monod, growth rate and Ks were cleverly
combined into an equation to describe the J parameter, and experiments
were carried out in mixed continuous cultures with multiple bacterial
strains, (Hansen and Hubbell, 1980). J represents the minimum threshold
concentration of a growth limiting substrate that will support a viable
population size. Similar experiments were carried out with diatoms, but the
parameter was termed R*, (Tilman, 1981; Smith, 1990; Smith, 1993).
Among competitors, the population that can grow at resource
concentrations lower than those which can support a competitor will have
an advantage, and therefore should dominate. The minimum threshold
concentration (J or R*) for the substrate can be calculated. Figure 1.1
describes how R* can be used to predict the winner of a single resource
competition. The dilution rate for a chemostat determines the population
growth rate. In this example, population A has an R* of 0.2 g L'1 at a
dilution rate of 0.1 h", and will outcompete population B which has an R“
of 1 g L". Population A will dominate (win), because the substrate in the
chemostat will reach a steady state concentration of 0.2 g L". At this
substrate concentration, B will have a grth rate of 0.025 h", which is
less than the dilution rate of the chemostat. Consequently, population B will

be washed out.

17

 

 

   
 
 
 

 

o 25 Population A
Speciﬁc 0'2 Population B
growth 0.15
rate
0‘ h-l) 0'1 ‘— dilution rate (d1 114)
0.05 R“ for population A R* ___ dKS
b- R* for population B W

 

o 1 2' 3 i 3
Substrate concentration, (S g H)

Figure 1.1. Competitive edge due to a lower resource threshold
concentration(R*). Graphical representation and equation of the concept
of resource threshold (R*). Calculation of R* uses the values for um,
(maximum growth rate), Ks (half saturation constant) and d (dilution rate
of the chemostat). “d” is equal to the speciﬁc growth rate, a. Population A
will outcompete B because R"‘A < R*B. Growth rate b in the diagram
represents the growth rate of population B when the substrate
concentration is reduced by the presence of population A, at a dilution rate
of 0.1 h‘ .

In an experiment to determine the validity of the theory, Tilman
demonstrated the value of R* in predicting the outcome of competition
among two species of freshwater algae competing for two limiting
substrates. In a case of interspeciﬁc competition, coexistence was obtained
between the two algal species (Asterionella formosa and Cyclotella
meneghiniana), when silicate and phosphate were limiting (Tilman 1977).
In the results of a long term study of 76 competition experiments, the

species-speciﬁc characteristics involved in the utilization of silicate and

18

 

phosphorous as resources inﬂuenced the competitive outcome of two
populations. When both populations were phosphate limited, Asterionella
predominated; when both were silicate limited, Cyclotella dominated, and
when each was growth-rate-limited by a different resource, both could
coexist. These results matched predictions based on the Monod model,
which, as mentioned previously, describes the dependence of the growth

rate on the concentration of limiting resource.

An organism’s capacity to utilize resources present in the
environment has a profound impact on its competitive ability. When a
bacterial population detects the presence of a compound that may be used
as a source of carbon and energy, ﬁrst the compound must be transported
into the cell. Following this, the organism must transform the substrate into
useful metabolic intermediates, which can be utilized for energy and
biomass production. Differences in the detection limit for a compound, the
rates of uptake and metabolism are all important to the reproductive
potential of a population. In addition to properties directly involved with
resource utilization, other biological attributes, such as motility, attachment,
substrate range and resilience to environmental perturbations may also play

a signiﬁcant role.

19

2,4-D DEGRADATION AS A MODEL SYSTEM FOR STUDYING
INTERACTIONS AMONG BACTERIAL POPULA T IONS

2,4-dichlorophenoxyacetic acid degradation

The bacterial utilization of the broad-leaf herbicide, 2,4-D was chosen
as the model system for testing competition theory. This herbicide has been
widely studied since its initial use during the 1940’s in both agricultural and
domestic capacity. Shortly after application, it was found that 2,4-D could
be readily degraded both by the combined efforts of microbial communities
as well as by numerous autocthenous bacterial populations in the soil (Ka et
al., 1994; Fulthorpe et al., 1994; Dunbar et al., 1996). To date, many
microorganisms capable of growth on 2,4-D as their sole source of carbon
and energy have been isolated, most often from sites with prior application
histories. Research has been focused on the enzymes involved in the

degradation of 2,4-D and the evolution of these pathways.

Competition experiments, described in chapters 2 and 4, take
advantage of a speciﬁc genotype which used 2,4-D as a sole source of
carbon and energy. Within the 2,4-D degrading bacterial community, there
is diversity at the allelic, vector (plasmid), and host levels that can be
exploited to address questions of inter- and intra-species competition among
2,4-D degrading populations. In addition to comparing ﬁtness differences

among different microbial species, the effect of different gene alleles to the

20

A n.u.l'

 

!ﬁ_ﬁiT

ﬁtness' of a host species can also be tested within the same genetic

background.

2,4-D catabolic pathway and types of tfd genes

The 2,4-D degradation pathway contains allelic variation. Initially this
variation was described relative to the genes characterized for the most
thoroughly studied 2,4-D catabolic plasmid pJP4 from Ralstonia eutropha
JMP134 (formerly Alcaligenes eutrophus JMP134) (Don and Pemberton,
1985; Don and Pemberton, 1981; Don et al., 1985). The genes carried on
pJP4 have been sequenced and characterized, and the operon structure has
been determined (Streber et al., 1987; Ghosal et al., 1988, Perkins et al.,
1990). This plasmid carries ﬁve of the genes encoding for enzymes
responsible for the degradation of 2,4-D to 2-chloromaleylacetate (CMA)
(Fig.1.2) (Don et al., 1985; Perkins et al., 1990; Fukumori and Hausinger,
1993). The enzymes responsible for the subsequent transformation of
CMA to intermediates of the Krebs cycle, were found to be chromosomally
encoded (Don et a1. 1985; Kukor et al., 1989; Bollrner et al., 1993). The
tfdA gene is an a-ketoglutarate dependent dioxygenase, which cleaves the
acetate moiety from the 2,4-D molecule, resulting in 2,4-dichlorophenol and
glyoxylate. Glyoxylate is transformed to succinate, which is shunted into
the bacteria’s metabolic pathways (Fukumori and Hausinger, 1993). The

transcriptional units of the genes tdeCDEF encode for enzymes involved in

21

the catabolic transformation of 2,4-dichlorophenol to CMA, which is then
degraded by host encoded enzymes. The tde and tde genes code for
regulatory proteins of tfdA, tde and tdeFEF gene cassettes. (Don and
Pemberton, 1985; Don et al., 1985; Fukumori and Hausinger, 1993;
Streber et al., 1987; Harker et al., 1989; Kaphammer et al., 1990;
Kaphammer and Olsen, 1990; Perkins et al., 1990; Matrubuthan and
Harker, 1994). The gene tde has been implicated in the transport of 2,4-D
into the cell (Leveau et al. 1998). The genes of pJP4 have been extensively
analyzed, and are an important tool for analyzing and comparing the 2,4-D
degradative genes among bacterial isolates.

Different 2,4-D alleles have been identiﬁed among 2,4-D degrading
isolates. Most of the differences have been found using the tfd genes of
pJP4 as a probe at different hybridization stringencies against the catabolic
plasmids, and chromosomes of strains that grow on this compound. One
of the plasmids isolated using exogenous plasmid-trapping experiments, is
pEMT3 (Top et al., 1994). Plasmid pEMT3 was the chosen representative
of a set of plasmids, pEMTZ-pEMT7, all of which exhibited similar
restriction patterns and hybridization proﬁles to each other, yet differed
from those of pJP4. Only the gene for tfdA was detected under high
stringency conditions, whereas tde could be seen under medium, and tde
under low stringency conditions. (Top et. al. 1994). The tfdA, -B and -C
genes of pEMT3 are still related, but different from those of pJP4.

22

Furthermore, when the sequence and enzyme differences of the tde and -C
genes of pEMT3 were determined, it was found that differences in the two
sets of alleles, pJP4 and pEMT3, are in their speciﬁcity toward chlorinated
and non-chlorinated substrates. These ﬁndings were conﬁrmed by studying
the substrate speciﬁcities of the individual enzymes (Maltseva, 1996). - /
Enzymes encoded by the pJP4-like genes have higher activity toward
chlorinated substrates while those of pEMT3 are better suited to the non-

chlorinated counterpart.

0015:0011 0H
1
GludACd-Hctﬁmf (:‘jc—p- ”01:9.“

ltfd D

OOH HOOC C=O
C1
* C1 * Cl
smﬁcicr} ads W tde \0/

Figure 1.2. 2,4-D degradation pathway of plasmid pJP4 of R.
eutropha JMP.134. deA: oc-ketoglutarate dependent dioxygenase,
tde: 2,4-DCP hydroxylase, tde: chlorocatechol-l,2-dioxygenase, tde,
chloromuconate cycloisomerase, tde, trans-chlorodienelactone isomerase,
tde, chlorodienlactone hydrolase. The mar gene, chloromaleylacetate
reductase, is chromosomally encoded.

24D

23

Another plasmid, pEMT8, was isolated and carried only a tfdA gene
that could complement a pathway in which tfdA had been knocked out.
The tfdA gene of pEMT8 shares high sequence similarity to that of pJP4.
Another 2,4-D degrading isolate, Burkholderia cepacia strain RASC carried
another tfdA gene that did not hybridize to pJP4, while the tde and tde
genes did. Sequencing tfdA(RASC) indicated the gene was only 77%
similar to that of pJP4 (Suwa et al., 1994). Although the tfdA(pJP4) and
tfdA(RASC) have a 23% difference in their DNA sequence, they still encode
for proteins with 81.9% amino acid identity and 96.5% similarity.
Curiously, the tfdA, tde and tdeDEF genes are located sequentially, in the
same type of grouped cluster as those on plasmid pEMT3 (Top et al.,
1994). The hybridization patterns indicated sequence differences, but with
a familiar spatial organization. Hence, it is likely that the tfdA gene has
evolved for some period of time. Some strains have genes much like pJP4,
others like that of Burkholderia cepacia strain RASC, and yet others
without clear homology to either (Table 1.2).

A distinct pattern of non-homology is seen in many of the
Sphingomonas strains that are phylogenetically distant from other isolated
2,4-D degraders (Ka et al. 1994a). tfdA, -B and -C from pJP4 do not
generally hybridize to DNA isolated from these strains. Despite not having

sequence similarity to tde, this strain exhibits chlorocatechol dioxygenase

24

 

activity when grown on 2,4-D (Fulthorpe et al., 1994). Although no tfdA-
like activity has yet been detected, there may be another oxygenase that may
fulﬁll this function.

Since many of the 2,4-D degrading isolates do not carry sequences
with recognizable homology to the previously identiﬁed 2,4-D degradative
genes, there is reason to believe that other pathways for the catabolism of
this herbicide exist (Matheson et al., 1996; Ka et al.,1994; Fulthorpe et al.,
1995). At this time, preliminary evidence for alternative pathways for the
catabolism of 2,4-D have been demonstrated. There is evidence for
degradation through a hydroxyquinone pathway in Nocardia species
(Haugland et al., 1990). Some isolates have been shown to lack tfdA
activity while other strains were shown to use an alternative ether-cleavage
enzyme that degrades 4—NPA to nitrophenol (Sassanella 1994). Isolates
from low nutrient, and pristine environments, Kamagata et al. (1997)
detected a novel ether-cleavage activity which did not exhibit the oc-keto
acid dependence as does tfdA of pJP4. Since many of these novel isolates
have not been examined, other distinctive mechanisms and pathways for

2,4-D degradation may exist.

25

2,4-D degradative plasmids and mosaics of catabolic genes

The 2,4-D degradation genes from various bacterial isolates are located on
plasmids (Don and Pemberton 1981, Mae et al. 1993, Ka and Tiedje 1994,
Top et al. 1994) or on the chromosome (Mae et al. 1993, Fulthorpe et al.
1994, Ka and Tiedje, 1994, Matheson et al. 1994, Suwa et al. 1994).
Among isolates capable of utilizing 2,4-D, by way of catabolic plasmids,
often the degradative and regulatory genes occur in different spatial
arrangements. An analysis of the catabolic genes from plasmids of 2,4-D
degrading bacteria indicated that homologous tfd genes are located on
different plasmids, or in a different spatial order from those of pJP4. In
Australia, Don and Pemberton (1981) obtained identical plasmids from both
Ralstonia eutropha JMP134 (pJP4) and from an Alcaligenes paradoxus
strain (pJP3). Several years later, Alcaligenes species EML159 was isolated
from Oregon soil, and this isolate contained a plasmid similar to pJP4 based
on restriction analysis and cross-hybridization experiments. Despite the
similarities in their encoding functions, the two plasmids did exhibit different
host ranges for the transfer of 2,4-D degradative ability (Amy et al., 1985).
The latter result suggests that the degradative plasmid may have a different
transfer or replication function, but the features leading to incompatibility

and host range were not included in the analysis of the 2,4-D degrading

phenotype.

26

In addition to pJP4-like plasmids, other plasmids carry 2,4-D
degrading genes. 2,4-D catabolic plasmids can range from moderately sized
(45 kb) to large (160 kb) and are represented among different
incompatibility groups. Using incompatibility tests and restriction proﬁles
with hybridization, some plasmids such as pJP4 and pEMT3 were shown to
belong to the broad host range IncPl plasmid group (Don and Pemberton,
1981; Top et al., 1994). Other plasmids, such as pTFD41, pRClO and
pEMTl belong to a different, yet unknown, incompatibility group (Ka and
Tiedje, 1994; Top et al., 1994). The plasmid, pRClO, isolated from
F lavobacterium sp. strain 5000] by Chaudry and Huang (1988) encodes for
many of the same functions as pJP4 and hybridization of pJP4 tfd genes
indicates high sequence similarity. However, no homology was noted for
pRClO and pJP4 for those regions that encode for functions of
maintenance, replication, conjugation or incompatibility (Don et al., 1985).
Similarly, the large conjugative plasmid (160kb) of Ralstonia sp. strain
TFD41 possesses tfd with nearly 100% nucleotide sequence identity to
those of pJP4, yet the plasmid backbone and gene order is different.
Plasmid pTF D41 maintains a duplicate copy of the tde and a partial
duplication of the tdeDEF operons. The tfdA and tde genes are located
in close proximity on pTFD41, and these are grouped farther away from the

rest of the operon as in the pathway carried on pJP4 (Nakatsu et al., 1994).

27

Different types of tfd genes have been described, and the mosaic
structure of the 2,4-D degradative pathway of several populations are
depicted in Table 1.2 (data adapted from Fulthorpe et al., 1995 sources
listed as references). As further evidence for the assortment of 2,4-D
degradative genes, an experiment was carried out to characterize diverse
2,4-D degradative plasmids occurring in soil. Top used R. eutropha
JMP228 (JMP134, rifR, cured of pJP4) as a genetic sink to trap catabolic
plasmids. Examination of the recovered 2,4-D degrading JMP228
transconjugants showed that seven different plasmids, designated pEMTl-
pEMT7, had been captured (Top et al., 1994). All plasmids were digested
with restriction enzymes, and the resulting ﬁ'agments were hybridized with
the pJP4 tfdA, -B, -C, -D, -E, -F and -R probes. Plasmids pEMT2- pEMT7
were similar to pEMT3, described above. An EcoRl digest showed pEMTl
to be very different from pJP4 and the other six captured plasmids. As
with pTFD41, all the tfd genes of pJP4 hybridized strongly to fragments of
pEMTl under high stringency conditions. Although the individual ﬁ'agment
sizes carrying these tfd genes did not match those of pJP4 and the distances
between pEMTl genes varied, the spatial arrangement of the tfd genes
remained the same as for the well studied pJP4. Therefore very similar gene
sequences are found on different plasmids.

The mosaic gene pattern for the tfd genes that arose among the

analyzed 2,4-D. degrading isolates, is probably due to plasmid transfer

28

followed by recombination. It is known that catabolic plasmids can move
between soil populations (Mergeay et al., 1994; Top et al., 1994; Kinkle et
al., 1993; Ka and Tiedje, 1994). Due to the conjugative nature of these
plasmids, they serve as vectors for shuttling catabolic pathways between
microbial populations. Plasmids carrying homologous gene sequences and
different spatial arrangements of tfd genes could recombine with
homologous sequences of other plasmids or chromosomes, exchanging tfd
gene cassettes. In this way, genes carried on different replicons could
become arranged in new combinations. This rearrangement could explain
why genes with nearly identical hybridization proﬁles are found in different
sites, at different times and within different organisms (Don and Pemberton,
1981; Top et al., 1994).

Although variations in gene sequences, arrangement and location have
been found, the implications of these differences to the ﬁtness of the 2,4-D
degrading populations have not been determined. Many isolates have genes
with varying degrees of similarity to those of pJP4, yet most seem to utilize
similar biochemical steps as the canonical pathway. It is not known how
the ﬁtness of the host population is affected by the efﬁciencies of these
enzymes with observed differences in gene sequence. Slight differences in
amino acid sequence can have profound repercussions as to the function
and structure of a protein. Unfortunately, there is very little data to indicate

if a particular combination of tfd genes confers a competitive advantage to a

29

 

bacterial strain when compared to the same strain carrying a different
permutation of tfd genes. It is not yet understood what the ﬁtness effect of
different tfd gene sequences or spatial arrangements are among competitors

utilizing 2,4-D as a substrate.

2,4-D organism diversity

Although a few pure cultures of microorganisms with the ability to
grow on 2,4-D as the sole source of carbon have been obtained from
pristine soils without previous exposure to 2,4-D, most of the bacteria
capable of catabolizing the compound have been isolated from sites with
prior exposure to the herbicide (Fulthorpe et al., 1994; Kamagata et al.,
1995). Multiple bacterial populations with this ability can be isolated from a
single soil sample (Fulthorpe et al., 1995; Tonso et al., 1994; Don and
Pemberton; 1981, Top et al., 1994, Dunbar et al., 1994) (Table 1.3).
Most frequently, 2,4-D degrading bacterial isolates are members of the
gram negative Proteobacteria, however, some gram positive bacteria and

fungi can also degrade 2,4-D (Haggblom ,1992).

30

lmwﬂ _ _' ‘ "a

Table 1.2. 2,4-D degrading populations carry mosaics of catabolic
genes. Southern hybridization analysis of 2,4-D degrading strains with the
tfdA, tde and tde gene probes from plasmid pJP4, are shown with
asterixes, *** is high stringency hybridization (>90% sequence similarity),
** is medium stringency (75-90%), and * is low stringency (60-75%
sequence similarity). Other notations are: -(hs) no hybridization under high
stringency, (rasc) low stringency hybridization to tfdA of pJP4, high
sequence similarity to tfdA from strain RASC. Fulthorpe et al, 1995

 

similarity to pJP4
genus strain plasmid 2,4-D tfdA tde tde Reference
4genes
Ralstoniaeutropha JMP134 pJP4 Pa *** *** *** Don 1981

Ralstonia eutropha JMP228 pEMTl P *** *** *** Top 1995
Ralstonia eutropha JMP228 pEMT2-7 P *** ** * Top 1995
Alcaligenes 2811P pKA2 P *** -(hs) —(hs) Ka 1994

paradoxus

Alcaligenes TF D9 *** ** * F ulthorpe 1995

xylosoxidans

Burkholderia sp. TFD2 * * - Fulthorpe 1995

Burkholderia sp. RASC Cb * *** *** Suwa 1996

Pseudomonas mallei BRI6001 C (rasc) *** *** Greer 1990

Pseudomonas K712 pKA4 P *** ** * Fulthorpe 1995
andropogonis

Variovorax sp. * Kamagata 1997

Rhodoferax TF D23 *** * * Fulthorpe 1995
fermentans

Rhodoferax TFD31 — - - Fulthorpe 1995
fermentans

Rhodoferax B6-9 ** - * Fulthorpe 1995
fermentans

Sphingomonas 1443 C? - - - Ka 1994
paucimobilis

Sphingomonas sp. K1443 C - * - Fulthorpe 1995

Sphingomonas sp. EML146 C - * - Fulthorpe 1995

Sphingomonas sp. TFD44 C - * - Fulthorpe 1995

Sphingomonas sp. B6-10 C - - - Fulthorpe 1995

Halomonaddaceae I-l8 *** *** *** Maltseva 1996

sp.
a P indicates plasmid— borne 2,4-D genes.

indicates chromosomally located 2,4-D genes.

31

n_‘ ___-

 

m‘.‘

 

In addition to genus level bacterial diversity, genetic diversity has also
been demonstrated to occur within a single species (Dunbar et al., 1994;
Dunbar et al., 1995). Consistent with the results of prior hybridization
experiments, Dunbar isolated 136 different bacterial isolates belonging to the
same species, but that could be distinguished into 32 different taxonomic
groups based on 168 rDNA restriction proﬁles (ARDRA). Groups
separated by ARDRA share a similar phylogeny. The restriction groups
were further subdivided by REP-PCR (DeBruijn et al., 1992). In another
similar set of experiments, using a variety of phenotypic and genotypic
methods for evaluating the diversity found among the 2,4-D utilizing groups
of microorganisms, Tonso et al. (1994) also characterized 46 isolates of
2,4-D degrading bacteria. The isolates were obtained from seven
geographically distinct sites and were divided into ﬁve taxonomic groups
based on carbon source utilization patterns (BIOLOG) and fatty acid
proﬁles (FAME). Genetic ﬁngerprints by REP-PCR indicated that these
groups could be further sub-divided based on identical, similar or unique
banding patterns (Tonso et al., 1994). Additionally, when the 16S rDNA
phylogenies of 21 different 2,4-D degrading isolates were compared with
known reference strains, the degraders were largely represented among 0t-,
B- and y- proteobacteria (Fulthorpe et al., 1994). Since so many 2,4-D

degrading populations occur in natural environments, inter-speciﬁc and

32

 

intra-speciﬁc competition may play a relevant role in determining the

microbial community structure.

Consequence of the high diversity in 2, 4-D degradative genes

As mentioned above, several different gene clusters and arrangements
for the known 2,4-D degradative alleles have been described. A single
species of bacteria may have different sets of alleles, depending on the
selective pressures of its environment. Consequently, similar sets of alleles
may be found in different species of bacteria, especially when the pathway
is found on a catabolic plasmid. When a plasmid is beneﬁcial, it propagates,
both by assisting in the growth and reproduction of its host, as well as by
promoting its own lateral transfer. The result is a skewed distribution of the
2,4-D degradative alleles toward the ﬁttest genotype when selection

pressure is applied.

33

Table 1.3. Diversity of 2,4-D degrading isolates.

 

group genus strain source Reference
B—Proteobacteria
Alcaligenes eutrophus J MP 1 34 Australia Don 1981
Alcaligenes paradoxus 2811P KBS, Michigan Ka 1994
Alcaligenes TFD9 Saskatchewan F ulthorpe 1995
xylosoxidans
Burkholderia sp. TF D2 KBS, Michigan Fulthorpe 1995
Burkholderia sp. RASC activated sludge Suwa 1996
Burkholderia mallei TF D6 KBS, Michigan Matheson et al.
(1996)
Burkholderia mallei BRI6001 Peat, Nova Scotia Greer 1990
Pseudomonas picketii 712 KBS, Michigan Ka 1994)
Pseudomonas K712 KBS, Michigan Fulthorpe 1995
andropogonis
Variovorax sp. Hawaiian soil Kamagata 1997
Rhodoferax fermentans TFD23 Michigan Fulthorpe 1995
Rhodoferax fermentans TFD31 Saskatchewan Fulthorpe 1995
Rhoda/crux fermentans B6-9 Ontario Fulthorpe 1995

a—Proteobacteria

Braayrhizobium sp.
Sphingomonas 1443
paucimobilis

Sphingomonas sp. EML146
Sphingomonas sp. TF D44
Sphingomonas sp. B6-10

Halomonadaceae Halomonaddaceae sp. I-18

F lavobacterium

Gram Positive

thgi

F lavobacterium sp. M H

Nocardioides simplex 3E
Arthrobacter sp.

Arthrobacter sp.

C orynebacterium sp.

Aspergillus niger

34

Hawaiian soil
KBS, Michigan

Oregon
Montana
Ontario

Kamagata 1997
Ka (1994

Fulthorpe 1995
Fulthorpe 1995
Fulthorpe 1995

Oregon, Alkali Lake Maltseva 1996

2-(2,4-DCP)PA
USSR, 2,4,5-T
soil

soil

soil

soil

Chaudry 1988

Kozyreva 1993
Loos 1967

Tiedje 1969

Rogoﬂ‘ 1956
Shailubhai, 1984

 

The maintenance of a high diversity of degradative genes and vectors
within populations is ecologically advantageous. A more diverse gene pool
will allow for the formation of more host-plasmid combinations, that may
offer advantages under variable environmental conditions. From an
evolutionary standpoint, an understanding of the true evolutionary origin of
the 2,4-D phenotype would be useful. Since the herbicide 2,4-D is
structurally similar to the plant hormone auxin, the pathway may have
developed as a means of degrading this type of natural compound. Auxins
however, are not degraded by the 2,4-D pathway. The majority of the 2,4-
D degradation pathway utilizes tde, (hydroxylase) and tdeDEF (ring
cleavage) genes, speciﬁc for chloroaromatic compounds.

Different degradation mechanisms may have evolved to degrade the
assortment of phenolic products that result from the breakdown of lignin.
The fact that the 2,4-D genes of plasmid pJP2 have undergone a duplication
event, is indirect evidence that different pathways may arise by subsequent
duplication and modiﬁcation. Acts of further evolution of the duplicated
genes may lead to divergent evolution of the pathway. In addition to the
genes for hydroxylation and ring cleavage, 2,4-D must ﬁrst be converted to
a chlorophenol, which frequently occurs as a result of the enzyme encoded
by the gene tfdA. The function of the tfdA gene is the initial step in 2,4-D
degradation, and several different homology groups of the gene have been

identiﬁed, (pJP4, RASC and pEMT8). While the fundamental ﬁrnction of

35

 

the ancestral tfdA gene is not known, Hogan et al. (1998) found the
presence of genes highly similar to tfdA in 70% of heterotrophic soil isolates
from Michigan Agricultural soil. None of the populations were capable of
utilizing 2,4-D as a sole source of carbon and energy, and the function of
tfdA in these isolates is unknown.

It is unclear whether the gene sequences encoding for catabolic
enzymes, the characteristics of the plasmids carrying these genes, or the
organism’s species-speciﬁc properties determine competitive superiority
among populations utilizing the same limiting resource. The natural
diversity of 2,4-D degradative populations, vectors and alleles provide a
series of variables that can be used to investigate intra- and inter-species
competition. Because the canonical pathway for 2,4-D utilization is well
characterized, it makes a good system for determining whether competitive
outcome among organisms is due to the inﬂuence of the organism, or the
tfd gene sequences. The variables of allele and plasmid to host strain ﬁtness
can be determined for a single species, which facilitates intra-speciﬁc
competition experiments comparing different 2,4-D genotypes. The
variables of allele and plasmid to the ﬁtness' of two competing host species
allow for inter-speciﬁc comparisons. Previously, Ka et al. (1994) had noted
a correlation between the plasmid carried by B. cepacia DB01 and its
growth rate on 2,4-D, although the two plasmid-bearing strains were not

subjected to conditions of direct competition. Thus, there is reason to

36

 

1r I“!

believe that the catabolic genes encoding for enzymes of 2,4-D degradation

pathways can inﬂuence the outcome of competition in bacteria.

2,4-D degradative genes may inﬂuence competitive ﬁtness

To determine how various tfd gene alleles affect competition,
isogenic hosts (spontaneous antibiotic resistant mutants of one natural
isolate) differing in their plasmid-encoded genes for the utilization of an
aromatic substrate (2,4—D) will be competed under different growth
conditions. Two of the bacterial species that can degrade 2,4-D when they
contain a 2,4-D plasmid, are Burkholderia and Ralstonia. By comparing
the results of competition experiments among the same species carrying
two different types of plasmids with different sets of alleles, it can be
determined if there is a signiﬁcant beneﬁt to carrying one of the allelic
groups. By forcing competition among different host species with the same
sets of degradative alleles, the contribution of host phenotype to ﬁtness can

be compared.

37

Al -‘ L‘-

 

References:

Amy P.S., J.W. Schulke, L. M. Frazier and RJ. Seidler. 1985. Characterization of
aquatic bacteria and cloning of genes specifying partial degradation of 2,4-
dichlorophenoxyacetic acid. Appl. Environ. Microbiol. 49: 1237-1245.

Arthur, W. 1982. The evolutionary consequences of inter-speciﬁc competition.
Adv. Ecol. res. 122127-187

Bollmer, M.D., K. Stadler-Fritzche, and M. Schlomann. 1993. Conversion of 2-
cloromaleylacetate in Alcaligenes eutrophus JMP134. Arch. Microbiol.
159:182-188

Chaudhry, GR. and G.H. Huang. 1988. Isolation and characterization of a new
plasmid from a Flavobacterium sp. which carries the genes for degradation
of 2,4-dichlorophenoxyacetate. J. Bacteriol. 170:3897-3902.

Chao, L. and E. C. Cox, 1983. Competition between high and low mutating
strains of E. coli. Evolution 37: 125-134.

Connell, J. H. 1978. Diversity and the coevolution of competitors, or the ghost of
competition past. Oikos 35: 131-138

Connell, J.H. 1983 On the prevalence and relative importance of interspeciﬁc
competition: evidence from ﬁeld experiments. American Naturalist,
122:661-696

Darwin, C. 1859 The Origin of Species by Means of Natural Selection. Penguin
Books, Harmondsworth (reprint 1968), p116

Don, R. H. and J.M. Pemberton. 1985. Genetic and physical map of the 2,4-
dichlorophenoxyacetic acid-degradative plasmid pJP4. J. Bacteriol.
161:466-468,

Don, R. H. and J.M. Pemberton. 1981. Properties of six pesticide degradation
plasmids isolated from Alcaligenes eutrophus and Alcaligenes paradoxus.
J. Bacteriol. 145:681-686

Don, R. H., A]. Weightman, H.J. Knackmuss and K. N. Timmis. 1985b.
Transposon mutagenesis and cloning analysis of the pathways for
degradation of 2,-4dichlorphenoxyacetic acid and 3-chlorobenzpate in
Alcaligenes eutrophus JMP134 (pJP4). J. Bacteriol. 161:85-90.

Duetz W. A., C. DeJong, P.A. Williams, and J. G. Van Andel. 1994. Competition
in chemostat culture between Pseudomonas strains that use different

38

pathways for the degradation of toluene. Appl. Environ. Microbiol.
60:2858-2863.

Dunbar, J ., D.C. L. Wong, M.J. Yarus, and L. J. F omey. 1996. Autoradiographic
mothod for isolation of diverse microbial species with unique catabolic
traits. Appl. Envrion. Microbiol. 62:4180-4185.

Dunbar, J., S. White, and L. F omey. 1997. Genetic diversity through the looking
Glass: effect of enrichment bias. Appl. Envrion. Microbiol. 6321326-
1331.

Dykhuizen, D. E. and D. L. Hartl, 1983. Selection in chemostats Microbiol. Rev.
47:150-168.

Elton, C. 1946. Competition and the structure of ecological communities. Journal
of Animal Ecology, 15:54-68

Fukumori F. and RP. Hausinger. 1993. Alcaligenes eutrophus JMP134 “2,4-D
dichlorophenoxyacetate monooxygenase” is an a-ketoglutarate-dependent
dioxygenase. J. Bacteriol. 175:2083-2086.

Fukumori F. and RP. Hausinger. 1993. Puriﬁcation and characterization of 2,4-D
dichlorophenoxyacetate/a-ketoglutarate dioxygenase. J. Biol. Chem.
268:2431 1-24317.

Fulthorpe, R.R., C. Mcgowan, O.V. Maltseva, W.E. Holben, and J .M. Tiedje.
1995. 2,4-Dichlorophenoxyacetic acid-degrading bacteria contain mosaics
of catabolic genes. Appl. Environ. Microbiol. 61:3274-3282

Gause, GR 1934. The struggle for existence. Hafner, New York. 1964. pp 163-.

Ghosal D. and I.-S.You. 1988. Nucleotide homology and organization of
chlorocatechol oxidation genes of plasmids pJP4 and pAC27. Mol. Gen.
Genet.211:113-120.

Ghosal D., I.-S.You, D.K. Chatterjee and AM. Chakrabarty. 1988. Genes
specifying degradation of 3-chlorobenzoic acid in plasmids pAC27 and
pJP4. Proc. Nat. Acad. Sci. 82:1638-1642.

Graham, D. W. , J. A. Chaudhary, R. S. Hanson and R. G. Arnold. 1993. Factors
affecting competition between Type I and Type II methanotrophs in two-
organisms, continuous-ﬂow reactors. Microbial Ecology, 25:1-17

Gottschal, J. C. 1992. Continuous Culture. In “Encyclopedia of Microbiology”,
Vol.1. Academic Press Inc. London. pp559-572.

39

Hansen S. R. and S. P. Hubbell. 1980. Single-nutrient microbial competition:
qualitative agreement between experimental and theoretically forecast
outcomes. Science. 207: 1491-1493.

Harker A.R., R.H. Olsen and RJ. Seidler. 1989. Phenoxyacetic acid degradation
by the 2,4-dichlrophenoxyacetic acid (TFD) pathway of plamid pJP4:
mapping and characteriztion of the TFD regulatory gene tde. J. Bactriol.
171:314-320.

Haugland, R. A., D. J. Schlemm, R. P. Lyons III, P. R Sferra and A. M.
Chakrabarty. 1990. Degradation of the chlorinated phenoxyacetate
herbicides 2,4-dichlorophenocyacetic acid and 2,4,5-
trichlorophenoxyacetic acid by pure and mixed bacterial cultures. Appl,
Environ. Microbiol. 56: 1357-1362.

Haggblom, M. M. 1992. Microbial breakdown of halogenated aromatic pesticides
and related compounds. FEMS microbiology rev. 103: 29-72.

Ka, 1.0.: ,W.E. Holben, and J.M. Tiedje. 1994a. Analysis of Competition in Soil
among 2,4-dichlorophenoxyacetic acid-degrading bacteria. Appl. Environ.
Microbiol. 60:1121-1128

Ka, J0. and J.M. Tiedje 1994b. Integration and Excision of a 2,4-
dichlorophenoxyacetioc acid-degradative plasmid in Alcaligenes paradoxus
and Evidence of its natural intergeneric transfer. J. BActeriol. 17625284-

5289,

Kamagata, Y., R.R. Fulthorpe, K. Tamura, H. Takami, L.J. Fomey and J.M.
Tiedje (1997). Pristine environments harbor a ne wgroup of oligotrophic
2,4-Dichlorophenoxyacetic acid-degrading bacteria. Appl. Environ.
Microbiol. 63:2266-2272

Kaphammer, B., J.J. Kukor and RH. Olsen. 1990a. Regulation of tdeDEF by
tde of the 2,4-dichlorophenoxyacetic acid degradation plasmid pJP4. J.
Bacteriol. 172:2280-2286.

Kaphammer, B., and RH. Olsen. 1990b. Cloning and characterization of tde, the
repressor-activator gene of tde, from the 2,4-dichlorophenoxyacetic acid
degradation plasmid pJP4. J. Bacteriol. 172:5856-5862.

Kinkle, B.K., M.J. Sadowsky, E.L. Schmidt and W. C. Koskinen. 1993. Plasmids

pJP4 and r6845 can be transferred between populations of Bradyrhizobia
in non-sterile soil. Appl. Environ. Microbiol. 59:1762-1766.

40

 

Kozyreva L. P., Y. V. Shurukhin, Z. I. Finkel’shtein, B. P. Bashunov, and L. A.
Golovleva. 1993. Metabolism of the Herbicide 2,4-D by a Nocardioides
simplex strain. Microbiology. 62:78-85.

Kukor, J.J, R.H. Olsen and IS. Siak. 1989. Recruitment of a chromosomally
encoded maleylacetate reductase for degradation of 2,4-
dichlorophenoxyacetic acid by plasmid pJP4. J. Bacteriol. 171:3385-3390.

Lenski, R.E., S.C. Simpson, and T.T. Nguyen. 1994. Genetic analysis of a
plasmid-encoded host genotype-speciﬁc enhancement of bacterial ﬁtness.
Journal of Bacteriology.

Lenski R. E., J. A. Mongold, P. D. Sniegowski, M. Travisano, F. Vasi, P.J.
Gerrish and T. M. Schmidt. 1998 Evolution of competitive ﬁtness in
experimental populations of E. coli: What makes one genotype a better
competitor than another. Antonie von Leeuwenhoek. 73:35-47

Leveau, J.H.J., A.J.B. Zehndera nd J. R. van der Meer. 1998. The tde gene
product facilitates uptake of 2,4—dichlorophenoxyacetate by Ralstonia
eutropha JMP134. J. Bacteriol. 176:2348-2353

Levin, B. 1972. Coexistence of two asexual strains on a single resource. Science.
175:1272-1274

Mae AA, 0. M. Reet, N.R., Ausmees, V.M.Koiv and AL. Heinaru. 1993.
Characterization of a new 2,4-dichlrophenoxyacetic acid degrading plasmid
pEMT4011: physical map and localization of catabolic genes. J. Gen.
Microbiol. 139:3165-3170.

Maltseva, O. 1996. Personal Communication.

Matheson, V.G., L. J. Fomey, Y. Suwa, Ch. H. Nakatsu, A.J. Sextone and W. E.
Holben. 1996. Evidence for acquisition in nature of a chromosomal 2,4-
dichlrophenoxyacetic acid/a-ketoglutarate dioxygenase gene by different
Burkholderia spp. Appl. Environ. Microbiol. 62:2457-2463.

Matrubuthan, U. and AR. Harker. 1994. Analysis of duplicated gene sequences

associated with tde and tde in Alcaligenes eutrophus JMP134. J.
Bacteriol. 176:2348-2353.

41

Mergeay, M., H. De Rore, E. Top, D. Srinagael, M. Hofte, D. Van der Lelei, R.
Dijkmans, and W. Verstraete. 1994 Plasmid biology and risk assessment:
Elements for modelling approach of plasmid-mediated gene release in soil
environments. In Environmental Gene Release: Models, experiments and
risk assessment. Eds. J. J. Bazin, and J. M. Lynch. Chapman & Hall,
London. pp.67-75.

Monod, J. 1949. The growth of bacterial cultures. Annu. Rev. Microbiol. 3:371-
394.

Nakatsu, C. 1994. Genetic organization and molecular evolution of the 2,4-D

catabolic pathway. In JRDC-MSU Joint Workshop on Microbial Evoultion
and Biodegradation. Organizing Committee: L. Fomey, M. Fukuda, T.
Kudo and Y. Tanaka.. JRDC-MSU. pp.51-59.

Perkins E.J., G.P. Milton, 0. Caceres and RR. Lurquin. 1990. Organization and
sequence anlaysis of the 2,4-dichlorophenol hydroxylase and
dichlorocatechol oxidative operons of plasmid pJP4. J. Bacteriol. 2351-
2359.

Rogoff, M. H and J. J. Reid. 1956. Bacterial decomposition of 2,4-
dichlorophenoxyacetic acid. J. Bacteriol. 712303-307

Sassanella 1994, personal communication

Schoener, T.W. 1983. Field experiments on interspeciﬁc competition. American
Naturalist, 122:240-285

Shailubhai, K., S. R. Sahasrabudhe, K. A. Vora and V. V. Modi. 1984. Degradation
of chlorobenzoates by Aspergillus niger. Experentia, 40: 406-407.

Smith, V. 1993, Implications of Resource-Ratio Theory for microbial ecology.
Advances in Microbial Ecolgy. 13:1-37.

Smith, V. 1993, Applicability of Resource-Ratio Theory to microbial ecology.
Limnol. Oceanogr. 38:239-249.

Streber W. R., K. N. Timmis and M.H. Zenk. 1987. Analysis, cloning and high-
level expression of 2,4-Diclrophenoxyacetate monnoxygenase gene tfdA of
Alcaligenes eutrophus JMP134. J. Bacteriol. 169:2950—2955.

Suwa, Y. A.D. Wright, F. Fukumori, K.A. Nummy, R.P. Hausinger, WE. Holben
and L]. Fomey. 1996. Characterization of a Chromosoamlly encoded 2,4-
Dichlorphenoxyacetic acid/alpha-Ketoglutarate dioxygenase from
Burkholderia sp. Strain RASC. Appl. Environ. Microbiol. 62:2464-9.

42

Top, E., 1. De Smet, W. Verstraete, R. Dijkmans, and M. Mergeay. 1994.
Exogenous isolation of mobilizing plasmids from polluted soils and
sludges. Apple. Environ. Microbiol. 60:831-839.

Top E.M., WE. Holben and L.J. Fomey (1995) Characterization of diverse 2,4-
Dichlorophenoxyacetic acid-degrading plasmids isolated from soil by
complementation. Appl. Environ. Microbiol. 61 :3274-3281

Tilman, D. 1977. Resource competition between planktonic algae: an
experimental and theoretical approach. Ecology: 58:338-348.

Tilman, D. 1981. Tests of resource competition theory using four species of Lake
Michigan algae. Ecology. 62:802-815.

Van Gemerden, H. 1974. Coexistence of organisms competing for the same
substrate: an example among the purple sulfur bacteria. Microbial
Ecology. 1:104-119.

Vadas R. L. 1990. Competitive exclusion, character convergence, or optimal
foraging: which should we expect? Oikos 00: 123:228

Van Niel, E.W.J. , L. A. Robertson, and J. G. Kuenen. 1993. A mathematical
description of the behaviour of mixed chemostat cultures of an autotrophic
nitriﬁer and a heterotrophic nitriﬁer/arerobic denitriﬁer: a comparison with
experimental data

43

 

CHAPTER 2

THE IMPACT OF SIMILAR DEGRADATIVE GENES ON DIFFERENT
PLASMIDS TO THE FITNESS OF ISOGENIC RALSTONIA E UT ROPHA STRAINS
COMPETING FOR 2,4-DICHLOROPHENOXYACETIC ACID.

ABSTRACT

Genes encoding the 2,4-D degradative pathway have been identiﬁed
on different plasmid backbones. Ralstonia eutropha strains acquire the
ability to degrade 2,4-D when they possess the appropriate catabolic genes.
This research examines the hypothesis that for R. eutropha strain JMP228,
2,4-D degradative alleles with high sequence similarity that reside on
different plasmid backbones, will result in populations of equal
competitiveness when 2,4-D is supplied as the limiting resource. Both
strains carrying pJP4 and pEMTl, exhibit very similar growth
characteristics in terms of growth rate on 2,4-D, and identity based on
colony morphology and time of colony appearance. Previously it was
shown that the two plasmids, pJP4 and pEMTl, carry homologous alleles
for the genes encoding the 2,4-D degradative pathway on different
replicons. The genes of pEMTl hybridize to probes generated from the
degradative genes of pJP4 under high stringency conditions. To test the

hypothesis, pure strains of R. eutropha, carrying either pJP4 or pEMTl,

I 4" COT

were grown with 2,4-D as their sole limiting substrate. Similar growth
rates (p=0.19) and rates of 2,4-D degradation by the host, were observed in
R. eutropha containing either pJP4 or pEMTl. Growth parameters derived
from pure culture studies indicated that plasmids pJP4 and pEMTl
conferred very similar growth rates and degradation rates in this host. Pair-
wise combinations of R. eutropha carrying pJP4 and pEMTl were
competed in batch culture conditions with 10 % of the culture transferred
daily for 5 days. During this period, the strains coexisted at nearly equal
ratios, which suggested that the inﬂuence of the plasmid backbone was
minimal. The slope of the line generated from natural logarithm of
population sizes did not vary signiﬁcantly from zero (P>0.1). In
chemostats, pure cultures of R. eutropha carrying either pJP4 or pEMTl
reached similar population densities of 3-6 x 108 CFU/ml at a dilution rate
of 0.033 h". When competed in the chemostats, the two populations
coexisted, and the total cell numbers were similar to those attained by either

pure culture.

45

YK——4 4f‘04—‘1
K, ‘
. l

INTRODUCTION

Among bacteria, plasmid transfer is a common mechanism for
acquiring new catabolic functions. Some of the prevalent broad-host range
plasmids of the proteobacteria can carry genes encoding for the
degradation of various aromatic compounds. One of the best studied
catabolic traits to be distributed by conjugative plasmids is the ability to
degrade 2,4-D. Self-mediated horizontal transfer of plasmid pJP4 has been
shown to occur in soils, as well as in the laboratory. (DiGiovanni et al.,
1996,; Neilson et al., 1994; Friedrich et al., 1983; Kinlde et al., 1993; Top
etaL,l998)

The propensity of these plasmids to propagate among the
proteobacteria allowed Top et a1. (1995) to use Ralstonia eutropha
JMP228 as a recipient strain for capturing several naturally occurring 2,4-
D degradative plasmids from the environment. Transconjugants that
gained the ability to degrade 2,4-D were isolated, and their plasmids were
characterized using gene probes targeting internal fragments derived from
the well-studied plasmid pJP4 (Holben et al., 1992; Top et al., 1995). All
of the 2,4-D genes of plasmid pEMTl hybridized under high stringency
conditions to those of pJP4. Upon examination of the plasmid backbone
with restriction analysis and a gene probe generated from the 750 bp Hae I I

fragment of vector pULB2420, carrying the inc-P/rep function of the Inc-

46

Inc-P1 plasmid RK2 it was concluded that pEMTl was not an Inc-P
plasmid like pJP4 (Top et al., 1995). The gene cassettes of pJP4 and
pEMTl were found to be in a different arrangement, with different
distances between the cassettes (Table 2.1). Similar results were seen in a
bacterial isolate identiﬁed as Pseudomonas putida, where the catabolic
genes were found in a different arrangement, on a plasmid backbone that
did not hybridize to the one of pJP4 (Mae et al., 1993). Consequently,
cassette shuffling and lateral transfer could lead to the persistence of highly
homologous alleles residing in slightly different genetic backgrounds even

within the same species of bacteria.

Table 2.1. Genotypic and phenotypic comparison of pJP4, pEMTl
and pEMT3. Characteristics of plasmids pJP4 and pEMTl used in
experiments of this chapter, and characteristics of plasmid pEMT3 used in
Chapters 3 and 4 (Top et al., 1995)

 

 

pJP4 pEMT] pEMT3
Size (Kb) 88 96 63
tfd gene High to ydA (pJP4) High to tfdA (pJP4)
homology High to tdeCDEF (pJP4) Moderate to tde (pJP4)
Low to tdeDEF (pJP4)
de gene order tfdA, tfdA, tfdA,
JdCDFE y'aCDFE tde, tde,
€de €de $10,!de
Distance from Ca. 16 Kb Ca. 7 Kb Ca. 2 Kb
tfdA to tde
Inc group Inc-PB Non-Inc-P Inc-PB
Host range or, B, y -proteobacteria (all)

 

47

. u_--_—t‘-ﬁ

1'

Since the same 2,4-D degradative genes appear to be located in
several different plasmid backgrounds, it raises the question of whether a
host strain population carrying these different plasmids would behave in a
competitively equivalent manner. Previous evidence supports the idea that
genetically similar bacterial populations do coexist in natural environments
(Dunbar et al., 1996). Analysis of a set of phenotypically identical, but
genetically distinct subspecies of 2,4-D degrading Variovorax paradoxus
populations were isolated from a single small soil sample. The great
diversity, indicated the coexistence of intra-speciﬁc variants (Dunbar et al.,
1996). In soil competition experiments among denitn'fying bacteria added
to soils, or among epiphytic bacteria, generally both populations would
coexist. When one population dominated, inoculum density or prior
colonization served as the primary determinants of the interaction (Murray
et al., 1992; Wilson and Lindow, 1994).

While several broad host range plasmids have been found to carry
homologous cassettes for the degradation of 2,4-D, the magnitude of
sequence differences is not precisely known. Plasmid pEMTl carries a set
of 2,4-D degradative alleles that hybridize to those of pJP4 under high
stringency conditions, indicating that they are >90 % similar in sequence,
but are the genes identical? (Top et al., 1995; Fulthorpe et al., 1995).

Based on this result it is both possible that the genes are identical, or that

48

 

they may contain some slight differences. Very small differences in
nucleotide composition, perhaps due to 3rd codon nucleotide variation in
the DNA sequence, may not signiﬁcantly affect the enzyme activities of the
proteins carrying out the consecutive degradative steps of the pathway.
Conversely, there may be differences that are not accounted for, if they
occur in the portion of the gene that was not used as the probe.

When a wild type bacterial population obtains a single genetic
mutation to generate different strains with a phenotypic marker that can
readily be identiﬁed in the laboratory, the resulting mutants are often
termed isogenic. While the populations are of the same species as the wild
type strain, and have nearly identical genomes, they now have a slight
genetic difference. The small differences in phenotype that arise as a
consequence of mutations utilized for genetic markers may or may not
impact the competitive ﬁtness of the bacterial host under different
environmental conditions. When utilized as strains for competition
experiments, it is desirable to obtain the mutation in a gene that does not
inﬂuence the ﬁtness of the host under the competition conditions.
Evidence from previous research has established the possibility that among
isogenic strains of bacteria, slight differences in genotype or the presence
of a plasmid may or may not affect the relative competitive abilities of
bacterial populations under particular laboratory or environmental

conditions (Levin, 1972; Bouma and Lenski, 1988; Rutter and Nedwell,

49

1994) . Dominance is not only affected by growth kinetics (Chapterl) but
also by a large number of factors, ranging from inoculum density to the
environment or culture conditions used for the competition experiments.
Therefore, it is possible that the same genes carried on different
conjugative plasmids could result in strains with very similar phenotypes,
and with very little observed differences in ﬁtness under limiting resource
conditions. On the other hand, the converse could prove to be true. There
may be factors carried on the plasmids that do not allow for the evolution
of compatible host-plasmid combinations, and thus strains which are
expected to display similar ﬁtness may prove to be signiﬁcantly different.
The research presented in this Chapter examined whether there was a
measurable difference in the phenotype and ﬁtness of two isogenic
Ralstonia. eutropha JMP228 strains, when they served as hosts to plasmids
pJP4 and pEMTl and utilized 2,4-D as the limiting resource. Antibiotic
resistant mutants of R. eutropha JMP228, bearing either streptomycin
resistance or rifampicin resistance were used as the isogenic host
populations for competition experiments. Presumably, the rifampicin and
streptomycin resistance markers are neutral with regard the 2,4-D degrading
phenotype, but to control for the inﬂuence of the genetic markers, both
plasmids pJP4 and pEMTl were introduced to both isogenic strains. Pure
culture grth characteristics in both batch cultures and chemostats, were

determined for the strains before competition experiments. To determine

50

experiments. To determine how the two isogenic populations compared to
a member of a different species (non-isogenic), the pJP4 and pEMTl
carrying R. eutropha JMP228 strains were competed with the pJP4
carrying Burkholderia cepacia DB01 strain in batch culture. If the two
plasmids conferred equivalent ﬁtness upon R. eutropha, then they should
also boast similar population dynamics in competition with the B. cepacia
strain. Fitness differences due to the presence of the different plasmid
backbones of pJP4 and pEMTl were addressed with the isogenic host
populations bearing the different plasmids. If the plasmids are neutral with
respect to succinate metabolism, isogenic populations with the different
plasmids should exhibit very similar grth rates on a non-selective carbon
source, such as succinate. If the ﬁtness differences are inﬂuenced by the
plasmid backbone, then the phenotypes and/or competitive ﬁtness of the
pJP4 and pEMTl carrying strains may differ from each other. If, on the
other hand, the different plasmid backbones have no effect on the 2,4-D

degradative phenotype, no differences in competitive ﬁtness should be

detected.

51

MA T ERIALS AND METHODS

Bacterial strains. Ralstonia eutropha JMP228, is the pJP4-cured
strain of R. eutropha JMP134 (Don 1981). Spontaneous mutants of R.
eutropha JMP228, resistant to rifampicin (250 pgml") (Sigma, St. Louis,
MO) and streptomycin (250 pg ml") (Sigma, St. Louis, M0) were used to
distinguish between plasmid bearing populations (JMP228R and
JMP228S). The rifampicin resistant isolate was obtained by Top (1995)
and the streptomycin resistant strain was generated for these competition
experiments as follows. Strain JMP228S was a spontaneous mutant
prepared by growing a culture of R. eutropha JMP228 ) to late log phase in
minimal medium containing sodium succinate (5 mM) (Sigma, St. Louis,
MO. Cells (1 ml of culture ﬂuid) were collected by centrifugation at 6000
rpm in a microcentrifuge (Beckman, Westbury, NY), washed once with
sterile phosphate buffered saline (PBS) and resuspended in 100 pl PBS.
The suspension was serially diluted in ten-fold increments and 100 pl was
spread over the surface of a minimal medium agar plate containing 2,4-D
(200 pg ml") (Sigma, St. Louis, MO) and streptomycin (100, 150, 200
and 250 pg ml"). Three of the colonies which grew on the 10'2 dilution
plate with 250 pg ml" streptomycin were streaked out to the same medium
three times. All spontaneous mutants were equivalent to one another and

the parental strain, in terms of growth rates on succinate and colony

52

morphology on LB and R2A (Difco, Detroit, MI) agar plates (Sambrook et
al., 1989). In the absence of selection for streptomycin resistance, the
marker was not readily lost. When JMP228S was grown in the presence
or absence of streptomycin, then diluted and plated to R2A agar with or
without the antibiotic, equal numbers of colonies were obtained for both
the non-selective and selective plates. B. cepacia strain DB01, naturally
moderately resistant to streptomycin (100 pg m1"), served as the non-
isogenic host strain. A mutant resistant to rifampicin (100, 150, 200, 250
pg ml") was obtained in the same manner as for R. eutropha JMP228.
Three colonies from the highest antibiotic concentration were chosen and
puriﬁed by three consecutive transfers to R2A agar containing rifampicin.
The mutants did not readily lose the rifampicin marker under non-selective
conditions, and the resistant Burkholderia strain was designated DB01R.
Ralstonia and Burkholderia strains can also be distinguished based on their
different colony morphologies on R2A agar plates. Bacterial strains used in

this experiment are listed in table 2.2.

53

Table 2.2. Bacterial strains used to determine the inﬂuence of the
plasmid backbone on competition

 

 

Species Strain Antibiotic Reference
resrstance

Ralstonia eutropha JMP228 rif‘S strS Don and Pemberton 1981

Ralstonia eutropha JMP228R riP‘ strs Top et al. 1995

Ralstonia eutropha JMP228S r115 suR This study

Burkholderia cepacia DB01 rifs m's Holben et a1. 1993

Burkholderia cepacia DBO l R rilﬂ triS This study

 

Plasmids and transconjugant construction. Plasmid pJP4, was
isolated from R. eutropha JMP134, and carries genes encoding the best
characterized 2,4-D degradative pathway (Don 1995). Plasmid pEMTl
was obtained in a previous study as a result of an exogenous plasmid
trapping experiment using strain JMP228R as a recipient (Top 1995).
Plasmid pEMTl carries homologous tfdA tde and tde genes to those of
pJP4, but on a different replicon backbone, as determined by hybridization
experiments using the replication origin for Inc-P plasmids as a probe (Top

1995). The plasmids used in this study are described in Table 2.3.

54

 

Table 2.3. Plasmids used to determine the outcome of competition in
isogenic populations

 

Hybridization witha
tfd genes of pJP4 IncP replicon

 

Plasmid Size (kb) tfdA tde tde rep R Reference
pJP4 88 *** *** ##1! *1“! Don and
Pemberton 198]
pEMTl 96 *** *** *** none Top et al. 1995

 

a . . . . . . . . .
”* indicates hybridization under high stringency conditions and correlates to
>90% sequence similarity.

Plasmids pJP4 and pEMTl were transferred from JMP228/pJP4 or
JMP228R/pEMT1 into the R. eutropha JMP228R and JMP228S strains
using a ﬁlter mating technique. Donor cells (1x107) were combined with
recipient cells (2x106) on a 13 mm, 0.02 pm Poretics ﬁlter (Fisher,
Pittsburgh, PA) which had been placed on LB agar plates for 8 hours. Cell
material from the ﬁlters was suspended by placing the ﬁlter in 1 ml A+N
minimal medium broth (composition below), and vortexing the ﬁlter for 30
seconds. The suspended cells were serially diluted in 10-fold increments,
and plated to A+N minimal media plates containing antibiotics to counter-
select the donor strain, and with 2,4-D to assure the presence of the
degradative plasmid in resulting transconjugants. The resulting

transconjugants are listed in Table 2.4.

55

 

 

Table 2.4. Transconjugants obtained with the filter mating
technique
Transconjugant Donor Recipient 2,4-D Antibiotic
utilization a resistanceb
JMP228/pJP4 ++ rifs sirg
JMP228R pEMT3 + rifR strs
JMP228R pEMTl ++ rifR strS
JMP228R - rifR str‘s
JMP228S - rifs strR
0130112 - iii“
JMP228R/pJP4 JMP228/pJP4 JMP228R ++ rifR Suﬁ
JMP228S/pJP4 JMP228/pJP4 JMP228S ++ rifs strR
JMP228R/pEMT1 indigenous JMP228R ++ riiﬂ str-s
JMP228S/pEMT1 JMP228R pEMTl JMP2288 ++ rifs strR
JMP228R/pEMT3 indigenous JMP228R + ritR strs
JMP228S/pEMT3 JMP228R pEMT3 JMP228S + rifs strR
DB01R/pJP4 JMP228/pJP4 DBOlR ++ rifR

 

a A minus indicates that no growth with 2,4-D occurred, a plus indicates slow growth,
two plusses indicate fast growth.
The antibiotics rifampicin (rif) and streptomycin (str) were used at a ﬁnal
concentration of 250 pg ml"

Plasmid pJP4 was introduced into B. cepacia DBOlR as described
for R. eutropha JMP228. The presence of the plasmid in the
transconjugants was veriﬁed by isolation of plasmid DNA using the
QIAGEN midi-prep column procedure as described in the manual
(QIAGEN, Chatsworth, CA). The plasmid was restricted by combining
0.5 pg of plasmid with 3 pl buffer REact 3, lpl of enzyme Bgl II and
water to a total volume of 25 p1 (Gibco, Grand Island, NY), and incubated

at 37 °C for 3 hours. Digestion products were separated and viewed on a

56

3% agarose gel in 1X TAE buffer stained with ethidium bromide. (ﬁg.
2.1). For storage, the transconjugant and recipient cultures were mixed
with sterile 60% glycerol to a ﬁnal glycerol concentration of 15%, and

frozen in Corning cryovials (Fisher, Pittsburgh, PA).

12345678 91011

a

I l.|
('-
1

Figure 2.1. Bgl II restriction analysis of plasmids isolated from
transconjugant strains. Lane 1: Hind III lambda marker, Lane 2:
JMP228R/pJP4, Lane 3: JMP228R/pEMT3 (not used in this study), Lane 4:
JMP228R/pEMT1, Lane 5: JMP228S/pEMT3 (not used in this study), Lane
6: JMP228S/pEMTl, Lane 7: JMP228S/pJP4, Lanes 8 and 9: Hind III
lambda marker, Lane 10: JMP228R/pJP4, Lane 11: DB01R/pJP4.

   

57

 

Cultivation of cells. Cells were grown in a mineral salts medium

supplemented with carbon sources and antibiotics. Per liter, the A+N
minimal medium consisted of 1.71 g KZHPO4-3H20, 0.34 g NaH2P04-H20,

0.33 g (NH4)2SO4, 0.25 g MgS04-7HZO, and 10 ml of a trace metals

solution. The trace metal solution contained, per liter: 12.0 g
NazEDTA4°2H20, 2.0 g NaOH, 0.4 g ZnSOﬂHzO, 0.4 g MnSO4-4H20, 0.1

g CuSO4-5HZO, 3.0 g FeSO4°7HzO, 5.2 g Na2SO4,0.1 g NaMoO4-2H20.
(Wyndham et a1. 1986). The pH was adjusted with 10 N H2804 to 7.2 prior
to autoclaving. Agar (1.7%, difco, Detroit, M1) or sodium succinate (5
mM) was added prior to autoclaving. The substrate 2,4-D and the
selective antibiotics, rifampicin and streptomycin were added after
autoclaving and cooling the medium to 55 °C from ﬁlter-sterilized (0.22
pm) stock solutions. The 2,4—D stock solution consisted of 20 mg ml"

2,4-D dissolved in 0.1 M NaH2P04 (pH 7.2), the rifampicin stock was 25

mg ml" in 100% methanol, and the streptomycin stock was 25 mg ml" in
water. Stock solutions were added to ﬁnal concentrations of 200 pg ml"
(2,4-D) and 250 pg ml" rifampicin or streptomycin. LB agar solid media
consisted (per liter) of 10 g yeast extract (Difco, Detroit, MI), 5 g tryptone
(Difco), 5 g NaCl (Baker, VWR, Plainsﬁeld, NJ) and 17 g agar. For each
independent experiment, cultures were started from a loopful of frozen

culture stock. The cell cultures were grown to late log phase in 16 ml test

58

[W uni-25am“.

grown to late log phase in 16 ml test tubes with 3 ml A+N minimal media
with 2,4-D (200 pg ml") and the appropriate antibiotic. A 2.5 %
inoculum to A+N medium supplied with succinate (SmM) or 2,4-D (200
pg ml") as the sole source of carbon and energy, was used to start the
cultures for growth and competition experiments.

Growth rate with 2,4-D. The growth rates of the four strains were
determined with 2,4-D (200 pg ml") as the carbon source using two
methods. In a more traditional method, three replicate cultures containing
100 m1 A+N minimal medium in 500 ml Erlenmeyer ﬂasks were inoculated
with a 1% inoculum, and grown at 30 °C on a rotary shaking incubator
(200 rpm). During growth with 2,4-D, samples were periodically removed
to determine optical density, and the concentration of 2,4-D. In addition,
the viable population size was determined by plating 100 pl of culture ﬂuid
onto A+N minimal medium agar plates. The optical density of the growing
cultures were monitored spectrophotometrically with a Hewlett Packard
8452A diode array spectrophotometer (Hewlett Packard , Waldbronn,
Germany) at 600 nm. The maximum growth rates were determined by
regression analysis of the linear portion of the logarithmic growth phase. A
one-tailed t-test was conducted with the calculated growth rates of replicate
samples using Excel 5.0 (Microsoft, Seattle, WA).

A second method for measuring grth was used. The cultures were

grown in Falcon 96-well microtiter plates (Fisher, Pittsburgh, PA), and

59

 

optical density measurements were made with a BIO-TEC Microtiter
EL312e microplate reader (BIO-TEK instruments Inc., Wenooski, VT). A
1 % inoculum from cultures grown in A+N minimal medium with succinate
(5 mM) was transferred into 25 ml of A+N minimal medium with succinate
(5 mM) in a 125 ml Erlenmeyer ﬂask. The cells were grown to stationary
phase (24 h, 30 °C, 200 rpm in a shaking rotary incubator) in order to
exhaust the resource, and to become nutrient limited. . A 1 ml aliquot of
the cell culture was spun down gently (4,000 rpm, 20 min in a
microcentrifuge), and suspended in sterile A+N medium without a carbon
source to an OD of 0.5 at 600 nm. Aliquots (20 pl) of this cell suspension
were then added to wells of a 96-well microtiter plate, that contained 230 pl
A+N medium per well. The medium was supplemented with either 2,4-D
(100 or 200 pg ml") or succinate (5 mM). The Kcalc program (BIO-TEK
instruments Inc., Wenooski, VT) was used to program the microtiter reader.
Microtiter plates were incubated at the ambient temperature inside the
reader (~25°C). Every hour for 60 hours the microtiter plate was shaken at
high speed for 45 seconds, and the optical density was determined
spectrophotometrically at 590 nm. Microtiter plates were covered and
paraﬁlm sealed during the course of the growth experiment. No evaporation
of media was observed for the inoculated wells. The data was saved to a
computer disk and analyzed using Microsoft Excel 5 .0. Comparison of the

growth rates was carried out using a one-tailed t-Test.

6O

High Performance Liquid Chromatography (HPL C). To measure
the 2,4-D concentration by HPLC, 1 ml samples from growing cultures
were centrifuged at 4 °C for 15 minutes at 13,000 rpm in an eppendorf
microcentrifuge (Brinkmann, Westbury, NY) to remove bacterial cells. The
supernatant was ﬁltered through a 13 mm minispike 0.45 pm Acrodisc®
ﬁlter (Fisher, Pittsburgh, PA), and the ﬁrst 0.5 ml were discarded. One
ﬁlter was used for 5 consecutive samples so that by discarding the ﬁrst 0.5
ml, 2,4-D remaining from the previous sample was washed out of the ﬁlter
unit. The remaining 0.5 ml was transferred an HPLC vial, crimp sealed and
frozen until all the samples could be run consecutively using an
autosampler. Prior to analysis, all samples were acidiﬁed by adding 10 pl of
10% phosphoric acid . The protonated and deprotonated forms of 2,4-D
have slightly different retention times, and acidiﬁcation to below the pKa of
2,4-D (pKaMD = 2.34) resulted in a single peak. Samples were analyzed
with a Hewlett Packard Series 1050 HPLC equipped with a 250 mm x 4.6
mm Lichrosorb RP-l8, 10 pm particle size, chromatography column
(Alltech, Bannockbum, IL), and a UV detector which was set to 260 nm.
A mixture of. 70% methanol/30% phosphoric acid (0.1%) (vol/vol) was
used as the eluent at a ﬂow rate of 1.5 ml per minute. In order to quantify
the concentrations of 2,4-D in the samples, the resulting peak areas were
compared with those of a standard curve generated in triplicate, with 5, 10,

25, 50, 100 and 250 pg ml" 2,4-D. A linear relationship was observed

61

through the entire concentration range that was examined, and
concentrations as low as 2.5 pg ml" could be detected.

Batch Culture Competitions. All cultures were grown in succinate
(SmM) media as described above, rinsed with sterile media without a
carbon source, then resuspended to ODéoo = 0.1 in minimal media without
2,4-D. Competition experiments were carried out using equal numbers of
competitor cells (4-6 x 107 CFU ml"). Bacterial growth was monitored by
measuring the optical density at 600 nm. Viable cell counts on A+N agar
medium with 2,4-D (250 pg ml") were performed in quadruplicate. Equal
volumes (500 pl each) of the rinsed and resuspended JMP228S/pJP4 and
JMP228R/pEMT1 cells were combined in two Eppendorf tubes and mixed
to yield similar starting population densities. As reciprocal pairs, two tubes
of JMP228R/pJP4 and JMP228S/pEMTl mixtures were also prepared. In
six replicate wells, 20 pl of each competition mixture was added to 1080 pl
of A+N amended with 2,4-D (200 pg ml"), in the alternating wells of
Falcon, ﬂat-bottom, 24-well tissue culture plates (Becton Dickinson,
Lincoln Park, NJ) to avoid cross contamination between adjacent wells.
Plates were sealed with paraﬁlm and incubated on a rotary shaker at 30 °C.
For 5 days, the CFU ml" was determined daily by dilution and plating to
A+N minimal media agar with 2,4-D (250 pg ml") and the selective
antibiotic to distinguish between competing populations. 10% of the mixed

culture volume was transferred to fresh medium in a new, sterile microtiter

62

culture volume was transferred to fresh medium in a new, sterile microtiter
plate to continue the competition. Analysis of the relative ﬁtness of the
competing populations was determined by plotting the ratios of the natural
logarithms of the population sizes as described by Lenski et al. (1992). A
deviation in the slope of this line away from zero, indicated that one strain
is competitively dominant to another. Competitive pairs consisted of
combinations of strains with different antibiotic susceptibilities, and
carrying different plasmids. Intra-species competitions relied on the
antibiotic resistance markers to identify the competing populations. In the
inter-speciﬁc competition experiments between R. eutropha and B. cepacia,
microtiter plates were inoculated in the same manner as the inter-species
competitions described above. F our replicate competitions were carried out
for each of the two pairs: JMP228R/pJP4 with DB01R/pJP4 and
JMP228R/pEMT1 with DB01R/pJP4. Differences in the colony
morphologies were used to distinguish between the R. eutropha and B.

cepacia populations.

Chemostat growth. Chemostats modiﬁed ﬁ'om the design of Chao et
al. (1977) were constructed and used to compare the growth characteristics
of strains JMP228S/pJP4, JMP228R/pEMT1 and JMP228S/pEMT1, at low
dilution rates. A schematic representation of the chemostat is shown in

Figure 2.2. The starter cultures were initiated in the same manner as for the

63

._._ . _ ii _

 

attained mid to late log phase, which correlated with an optical density of ~
0.1 at 600 nm (2-3 x 107 CFU ml"), the cells were harvested (4000 rpm, 20
minutes in a Sorvall SS34 rotor). After rinsing with A+N minimal media,
the pellets were suspended to a ﬁnal OD600 of 0.2 in A+N with 200 pg ml"
2,4-D. Either 75 ml of the pure culture or 37.5 ml of each competitor was
placed in a chemostat and 2,4-D was added from a 20 mg/ml stock solution
to a ﬁnal concentration of 200 pg ml". Samples (3 ml) were taken from
each chemostat at the onset of the experiment. The chemostats were
incubated for 8 hours at 30° C, before constant ﬂow conditions were
initiated. The dilution rate was set to 1.25 volume exchanges per day (2.5
ml h"), with a reservoir 2,4-D concentration of 200 pg ml". The
chemostats were operated for 120 h. Since the dilution rate (d) of a
chemostat determines the growth rate (p) of the populations, if d= 2.5 ml h’
1 for a 75 ml volume then p = 0.033 b", which correlates to a 17.5 h
doubling time. This growth rate was well below the maximum growth rate
for the strains under batch culture conditions. To sample the chemostats,
the port was swabbed with 70% ethanol. Using a 3 ml Hamilton syringe
ﬁtted with a sterile spinal tap needle (4 mm) (Fisher, Pittsburgh, PA), a 3 ml
sample was removed. The sample was used for obtaining measurements of
the 2,4-D concentration, optical density and the CFU ml" of each
competing strain on A+N agar with 200 pg ml-l 2,4-D and the appropriate

selective antibiotics. Triplicate plates were established for each time point to

64

 

determine the average population size. As with the Batch culture
competitions, the change in the ln ratio for competitive ﬁtness were
determined. The slopes for the lines are reported from the time following

nutrient depletion in the chemostats.

“1‘

1L - L's -

 

65

Acrodisk
0.22pm

   

I

     
   
  
     

  
     

 

 

 
  

 

  
 
  
 

 

 

 

 

 
 

 

   

 
 
 
   
   

 

 

     
 

    

 

   

   

   

 

     

 

 

 

 

 

 

o . \
:7- _
lllll , 5 $117.
5' \N'Ell I .
I l 1 5 ‘57:! ~ Acrodisk
“I” 1.] III" V} 0.22pm
‘ m. H" . Media
@ (1.9 Inﬂow
. g Glass Wool
u an ,~
Reserva'r . Alll
Sampling Port
Efﬂuent
Tube
«t. J)
O
((#3)) ‘— Aeration 311d
0 0 Mixing
0‘1
v —> Vacuun

A A A A A A l A A
«o-c0:><=0=>40> 0 <0><o><a>-o-

Vacutm
Manibld

  

 

Figure 2.2. Schematic representation of the chemostat used in the
competition experiments under low nutrient conditions. All
connections were made with luer-lock ﬁttings. A glass wool plug
surrounding the media inﬂow tube ﬁltered the air, which was drawn into the
chemostat using vacuum. Vacuum was also used for drawing media into
the culture vessel, and for aeration and mixing.

 

HF.

RESULTS

Growth of pure cultures under batch conditions. Plasmids pJP4 and
pEMTl beneﬁt their host Ralstonia eutropha by supplying a catabolic
pathway that can convert the herbicide 2,4-D to metabolic intermediates
which then enter the TCA cycle. Both plasmids, pJP4 and pEMTl, were
conjugally transferred into two isogenic host strains, Ralstonia eutropha
(formerly Alcaligenes eutrophus) JMP228R and strain JMP228S as
described in Table 2.4. The growth curves determined by optical density
measurements and 2,4-D depletion curves of the constructed strains are
shown in ﬁgure 2.3. The inset panel indicates the log-linear (exponential)
portion of the growth curve used to calculate the growth rate.

The growth rates for R. eutropha strains carrying pJP4 and pEMTl
did not differ signiﬁcantly when succinate served as the source of carbon
and energy. The data for the batch cultures grown in Erlenmeryer ﬂask
and/or the microtiter plate grown cells were comparable, and the growth
rates for the strains are compared in Table 2.5. When grown with succinate
(5 mM) as a limiting resource, both strains had pmax of 0.54 to 0.60 h"
(n=6, P =0.08). Neither of the R. eutropha strains appeared to be
signiﬁcantly different. Furthermore, for R. eutropha growing on 5 mM

succinate, chromosomal antibiotic resistance had no effect on the growth

67

 

rate (n =6, P = 0.96.) (Table 2.5). Growth rates of the strains were also
determined on 200 pg ml" 2,4-D as the carbon source to determine if there
was a difference between the plasmids (table 2.5). Plasmids pJP4 and
pEMTl both conferred a relatively fast growth rate of 0.23 (+/- 0.04) h'1
upon ' strains JMP228R/pJP4, JMP228S/pJP4, JMP228R/pEMT1 and
JMP228S/pEMT1 when 2,4-D was supplied as a limiting substrate. Under
batch culture conditions, both pJP4 and pEMTl harboring R. eutropha
strains maintained equivalent population densities of approximately 4 X
108 CFU ml" (Table 2.6) g

The Microtiter plate experiments showed that the growth curves
obtained on succinate and on 2,4-D (100 or 200 pg ml" ) were alike for
JMP228S/pJP4, JMP228R/pJP4, JMP228S/pEMT1 and JMP228R/pEMT1.
Replicate microtiter plate results were comparable, and while initial rates
were the same for both 100 and 200 ppm 2,4-D, the growth rates decreased
more rapidly in the 100 ppm 2,4-D samples as the substrate was consumed.
The comparison with ﬂask experiments use rates obtained with 200-250 pg
ml" 2,4-D to calculate the values in Table 2.5, since a longer growth phase
was observed. There was no signiﬁcant difference in the grth rates of
JMP228R/pJP4 and JMP228S/pEMT1 (n=12, p=0.96) as indicated by a

one-tailed t-test.

68

 

Table 2.5. Growth rates of parent strains and transconjugants on
succinate and 2,4-D. Growth rates (h") :1: standard deviation. Number of
replicates is shown in parentheses.

 

 

 

Strain Carbon and energy source a
Succinate 2,4-D
Flask (h") Microtiter (h") Flask (h'l) Microtiter (1F)
JMP228R 0.47 :1: 0.012 0.57 i 0.067 No growth No growth
(3) (6)
JMP228S 0.42 d: 0.003 0.54 :t 0.018 No growth No growth
(3) (6)
JMP228R/pJP4 0.51 i 0.048 0.58 :1: 0.030 0.20 :t 0.009 0.24 i 0.048
(4) (6) (3) (6)
JMP228S/pJP4 0.48 :t 0.007 0.56 i 0.037 0.24 d: 0.004 0.22 :1: 0.026
(3) (6) (3) (6)
JMP228R/pEMT1 0.50 d: 0.034 0.60 :t 0.037 0.23 i 0.011 0.24 :h 0.059
(4) (6) (3) (6)
JMP228S/pEMT1 0.50 d: 0.007 0.55 i 0.052 0.21 :t 0.004 0.23 :1: 0.043
(3) (6) (3) (6)
DB01R/pJP4 0.47 :1: 0.036 0.41 :t: 0.048 0.15 d: 0.012 0.20:1: 0.052
(4) (6) (3) (6)

 

69

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.20 .
0.15 . If ‘
l/
0.10 . 'j '
'l
OD600 i it
0.05 . /[ 6 l
h L 5 l
1’ ’. n 1"
" or) 4 . ..150 -
0.00 g --. 600 .
: m.“ "'\-'6 3 - ..
.g r: ‘ 17' L
‘ ,‘ . x 20 5 10 15 20
‘~ \|\“l‘l\. ‘|" Hours _lm
g "r 2,4-D
2211:; (ppm)
l“ 1"I1 I... d .50
'1': ‘1
l‘ \\ '\ I
\ ‘ I‘
Iii". l
t\:' 1‘ 1 n
0 10 ‘30 C 30 40 50 60'
Hours

Figure 2.3. Growth of pure cultures carrying plasmid pJP4 or
pEMTl on 2,4-D. JMP228R/pJP4 (squares), JMP228S/pJP4 (diamonds),
JMP228R/pEMTl (triangles) and JMP2288/pEMTl (circles) on 2,4-D
minimal media (solid symbols). 2,4-D depletion is shown in open symbols.
Each point is the average of measurements of three replicate cultures.
Error bars represent one standard deviation around the average. The inset
panel shows the natural log of the population size over time. From top to
bottom the lines are JMP228R/pEMT1, JMP2288/pJP4, JMP228S/pEMTl

and JMP228R/pJP4.

70

Colony morphology on agar plates. R. eutropha JMP228 carrying
plasmids pJP4 or pEMTl had the same colony morphology on both R2A
agar plates, and on 2,4-D minimal media plates. On R2A agar, colonies
were round, of uniform size and tan in color. The translucent colonies had
smooth edges and appeared after an incubation period of 48 hours at 20 °C.
On 2,4-D minimal media plates, the colonies maintained the round, smooth,
morphology, and translucent appearance. Colonies could be observed after
an incubation period of 72 hours, but were much easier to count after 96
hours.. The R. eutropha strains could readily be distinguished from the B.
cepacia strain on R2A agar plates. B. cepacia DB01r/pJP4 was
characterized by opaque cream colored colonies, with irregular size and
rough edges. Furthermore, the B. cepacia strain did not form colonies
before 60 hours of incubation. The B. cepacia strains also formed colonies
on 2,4-D agar plates after incubation for 72 hours, however, visual
distinction from R. eutropha was difﬁcult. Based on these results, plate
counts were carried out after 72 hours (3 days) with the R2A agar plates,
and after 120 hours (5 days) with the 2,4-D minimal medium plates.

Batch competition experiments. R. eutropha JMP228 carrying
plasmids pJP4 and pEMTl (JMP228R/pJP4 and JMP228S/pEMTl,
JMP228S/pJP4 and JMP228R/pEMT1) coexisted at a constant ratio in

intra-speciﬁc batch culture competitions. Panel A in Figure 2.4 indicates the

71

average population sizes for the competing strains. No effect was observed
for the strains carrying different antibiotic markers (n = 12, P = 0.21). In
24 independent, replicated batch culture competition experiments between
JMP228R/pJP4 and JMP228S/pEMT1 or the reciprocal set, both strains
coexisted at a relatively constant ratio. The slope of the line generated by
the ratios of the natural logarithms of the population sizes did not differ
signiﬁcantly from zero, as indicated by the very low r2 values (average
slope = -0.1025, r2 = 0.1794) (Figure 2.4 B). When the natural log of the
rifampicin resistant to streptomycin resistant strains was compared, the
slope of the line also did not deviate signiﬁcantly from zero (slope = 0.0108,
r2 = 0.11942). Similar results were obtained in competition experiments
with JMP228S/pJP4 versus JMP228R/pEMT1 (data not shown).

In another set of experiments, equal numbers of JMP228R/pJP4
orJMP228R/pEMT1 were competed against DB01R/pJP4. In four replicate
competition ﬂasks, the R. eutropha strains became dominant over the B.
cepacia strain when 200 pg ml" 2,4-D was supplied as the limiting
substrate. Furthermore, both JMP228R/pJP4 and JMP228R/pEMT1
followed the same pattern of dominance over DB01R (Fig. 2.5 A). The
average slope of the line positively deviated from zero, indicating dominance
by R. eutropha. JMP228R/pEMTl had an average slope of 2.809 and
expressed an r2=0.9680, whereas JMP228R/pJP4 had an average slope of

2.9637 and an r2=0.9842). Comparing the slopes of the lines generated by

72

the two R. eutropha strains indicated no signiﬁcant differences in the ability
of both strains to compete with B. cepacia (n=8, P=0.27). The plateau of
the ratio of competing populations represents the point at which no more B.
cepacia colonies could be detected on plates otherwise solely consisting of

R. eutropha colony morphotypes (Fig. 2.5, panel B).

73

 

1.E+09
1.E+O8 -

 

CFU/ml l.E+07 ~

 

 

 

 

 

 

 

1.E+06 -
1.E+05 -
Days
B 5
3 4
Ln ratio 1 ~ 4) L
PJP4/pEMTl ,1 4 ' ”"6 --.. a r s ”—78“?! ~——- % m
-3 - 8
-5
Days

Figure 2.4. Competition among R. eutropha strains JMP228/pJP4, and
JMP228/pEMT1 on 200 ppm 2,4-D. Panel A: Averages of the colony
forming units of 24 replicate competition ﬂasks. Diamonds are
JMP228/pJP4 and squares are JMP228/pEMTl. Error bars are the standard
deviation around the mean value for the population sizes of all competition
ﬂasks. Panel B: Natural logarithm of ratio of the pJP4 carrying JMMP228 to
the pEMTl carrying strains. Each of the 24 replicates is shown by
diamonds. Diamonds of different shades of gray represent different
experimental pairs.

74

A 1.E+09

 

 

 

 

 

 

 

 

1.E+08 ~
CFU/ml 1.E+07 -
1.E+06 .
1.E+05
0 l 2 3 4 5 6 .
Days
B 8
6 . 0
Ln (R.e.lB.c.)4 _ B
2 -
0 i
0 1 2 3 4 5 6
Days

Figure 2.5. Ralstonia eutropha hosting pJP4 or pEMTl are equal
competitors against Burkholderia cepacia hosting pJ P4. Batch culture
competitions in minimal medium amended with 250 ppm 2,4-D were
transferred to fresh media every 24 hours for ﬁve consecutive days.

A: Each point represents the average of four replicates. Error bars are the
standard deviation. Filled symbols represent competition between
JMPZZ8R/pJP4 (diamonds) and DB01R/pJP4 (circles). Open symbols
represent competition between JMP228R/pEMT1 (squares) and
DB01R/pJP4 (circles). Panel B: The natural logarithm of the ratio of R.
eutropha with pJP4 or pEMTl to B. cepacia hosting plasmid pJP4. Filled
diamonds are four separate competitions of JMP228R/pJP4 versus
DB01R/pJP4. Open squares are JMP228R/pEMTl versus DB01R/pJP4.
DB01 populations could not be detected below 1% of the total population.

75

Chemostat growth and competition experiments. Both pure culture
and mixed culture competition experiments were performed concurrently
in chemostats. Strains JMP228S/pJP4 and JMP228R/pEMT1 attained cell
densities of approximately 4-6 x 108 CFU ml". Strain JMP228S/pEMT1
grew to 2-4 x 108 CFU ml" (Table 2.5). In all culture vessels, the
concentration of 2,4-D was utilized to undetectable levels within the ﬁrst
24 hours. These results suggested that the pJP4 and pEMTl carrying R.
eutropha populations appeared to behave similarly under mixed and pure
culture growth conditions, and coexist in a chemostat operated at low
dilution rates. In a single chemostat experiment, strains JMP228R/pEMT1
and JMP228S/pJP4 were competed. The two strains coexisted, and
neither appeared to have a deﬁnitive ﬁtness advantage during the course of

6 days (Figure 2.6).

Table 2.6. Carrying capacity of R. eutropha strains in batch and
chemostat conditions

 

Strain Carrying Capacity

Batch a Chemostatb
JMP228R/pJP4 3.8 x 10% 2.3 x 107 nd
JMP228S/pJP4 3.7 x 108:1: 5.3 x 107 6 x 10%: 1 x 108
JMP228R/pEMT1 4.6x 108: 6.1 x 107 6x108:h1x108
JMP228S/pEMTl 4.3 x 108:1: 1.2 x 108 3 x 1083: 3 x 108

 

l: Pure culture growth on 200 ppm 2,4-D

Pure culture growth with 200 ppm 2,4-D in the reservoir, and a 0.33 h“l dilution rate.
2.4-D concentration in the chemostats was at undetectable concentrations. Each
eXperiment was run in triplicate.

76

A

30 A ~ 200
20 m1 [2,4-D]
ln cfulml 13 x l (1)00 (ppm)

 

 

 

 

 

 

0 50 100 150
hours
B
30 - 200
20 .
10 ‘ I 100 (ppm)
0 . Q ‘ - - 0
0 50 100 150
hours
C 30 - - 200
20 [2,4-D]
In cfulml 10 s‘ 100 (ppm)
0 m .- 0
0 50 100 150
hours
D
30 - 7 200
20 m I [2,4-D]
In cfulml 10 x 100 (ppm)
0 - . 0
0 50 days 100 150

Figure 2.6. R. eutropha with pJP4 or pEMTl in pure culture or in
competition in chemostats. Panels A—C: pure culture experiments. (A)

JMP228S/pJP4, (B) JMP228R/pEMT1, (C) JMP2288/pEMTl.

Panel D:

competition experiment. JMP228S/pJP4 (diamond) and JMP228R/pEMTl

(squares).

2,4-D concentration in chemostats is shown by the open

diamonds. Chemostats were run at a dilution rate of 0.033 h-l, and each

time point represents the mean of three replicate samples.
deviations fall within the presented symbols.

77

Standard

DISCUSSION

Many bacteria isolated from soils degrade 2,4-D, often by means of
catabolic genes carried on plasmids. The effect of different plasmids
carrying homologous degradative pathways with very similar genes for
2,4-D utilization on competitive ﬁtness was examined. The plasmids were
readily maintained within sub-populations of a single host strain. The goal
was to determine if two isogenic populations behaved in a similar manner
when 2,4-D was supplied as the sole source of carbon and energy. To
determine if the growth rates of an R. eutrophus host carrying different
2,4-D degradative plasmids (pJP4 and pEMTl) inﬂuenced competitive
outcome, strains JMP228R/pJP4, JMP228S/pJP4, JMP228R/pEMT1 and
JMP228S/pEMT1 were competed in batch competition. When the relative
ﬁtness of the strains were determined in head to head competition
experiments, coexistence was observed. The ratio of the two competing
populations remained constant. In this instance, it appears that the two
populations are nearly equally ﬁt to compete for 2,4-D, and therefore the
different plasmid backbones of pEMTl and pJP4 did not signiﬁcantly
inﬂuence the individual strains ﬁtness under the experimental conditions

chosen.

78

[tramwaAFﬁ

Both pJP4 and pEMTl containing cultures exhibited similar growth
rates on succinate, a chromosomally encoded pathway. The similarity of
growth rates on succinate indicated that the downstream portion of the
2,4-D degradative pathway was comparable in both strains. Plasmids p1 P4
and pEMTl beneﬁted their host strain by supplying a catabolic pathway
which converts the herbicide 2,4-D to metabolic intermediates, including
succinate, which then enter the TCA cycle. Therefore, following
degradation to 2-chloromaleylacetate, the genes needed to ﬁlrther
metabolize the intermediates are encoded by the host (Kukor et al., 1989).
Since the chromosomally-encoded genes for the’ degradation of 2-
chloromaleylacetate are the same for the two R. eutropha strains, any
differences in ﬁtness when grown on 2,4-D, could be presumed to be due
to the plasmid itself, or the genes carried on the plasmid. Therefore, the
similarities in both the growth characteristics and competitive ability of
strains JMP228R/pJP4, JMP228S/pJP4, JMP228R/pEMT1 and
JMP2288/pEMT1, were due to physiologically comparable genes for
degrading 2,4-D, carried in relatively neutral plasmid backbones.

Although plasmids pJP4 and pEMTl belong to different
incompatibility groups, they have a similar impact on the grth of their
host when competing for 2,4-D. Under the described experimental
conditions, no selective advantage was observed when the two populations

were grown on 2,4-D. The genes for 2,4-D utilization can be transferred

79

-—A£Ij

 

horizontally on two very different plasmid backbones, and confer
comparable physiological capabilities (2,4-D degradation) to the same host
species. In nature, however, there may be differences in the survival in
these strains due to more complex environmental conditions and different
host responses. The survival of the plasmids themselves in the environment
and the probability that they transfer to other populations in soil or water

ecosystems may be due to the host range and promiscuity of the plasmid.

80

 

REFERENCES

Bollmer, M.D., K. Stadler-Fritzche, and M. Schlémann. 1993. Conversion of 2-
cloromaleylacetate in Alcaligenes eutrophus JMP134. Arch. Microbiol.
159:182-188

Bouma, LE. and RE. Lenski. 1988. Evolution of a bacteria/plasmid association.
Nature. 335:351-352.

Chao, L., B.R. Levin and RM. Stewart (1977). Complex community in a simple
habitat: Experimental study with bacteria and phage. Ecology. 58:369-378.

Don, R. H. and J.M. Pemberton. 1985. Genetic and physical map of the 2,4-
dichlorophenoxyacetic acid-degradative plasmid pJP4. J. Bacteriol.
161 :466-468,

Don, R. H. and J.M. Pemberton. 1981. Properties of six pesticide degradation
plasmids isolated from Alcaligenes eutrophus and Alcaligenes paradoxus.
J. Bacteriol. 145:681-686

Don, R. H., A.J. Weightman, H.J. Knackmuss and K. N. Timmis. 1985.
Transposon mutagenesis and cloning analysis of the pathways for
degradation of 2,-4dichlorphenoxyacetic acid and 3-chlorobenzpate in
Alcaligenes eutrophus JMP134 (pJP4). J. Bacteriol. 161285-90.

DiGiovanni G.D., J.W. Neilson, I.L. Pepper and NA Sinclair. 1996. Gene
transfer of Alcaligenes eutrophus JMP134 plasmid pJP4 to indigenous soil
recipients. Appl. Enviom. Microbiol. 62:2521-2526

Dunbar, J.M. 1996. Effect of selection on the structure of a 2,4-
dichlorophenoxyacetic acid (2,4-D) degrading guild in soil. Ph.D
dissertation. Michigan State University, Department of Microbiology.

Friedrich, B. M. Meyer, and HG. Schlegel. 1983. Transfer and expression of the
herbicide degrading plasmid pJP4 in aerobic autotrophic bacteria. Arch .
Microbiol. 134:92-97

Fulthorpe, R.R., C. Mcgowan, O.V. Maltseva, W.E. Holben, and J.M. Tiedje.
1995 2,4-Dichlorophenoxyacetic acid-degrading bacteria contain mosaics of
catabolic genes. Appl. Environ. Microbiol. 61:3274-3281

Holben, W.E., B. M. Schroeter, B.G. Calabrese, R.H. Olsen, J.K. Kukor, B.O.
Biederbeck, A.E. Smith and J.M. Tiedje. 1992. Gene probe analysis of
soil microbial populations selected by amendment with 2,4-
dichlorophenoxyacetic acid. . Appl. Eviron. Microbiol. 58:3941-3948.

81

3.- ‘ﬂ

 

Kukor, J.J, R.H. Olsen and J.S. Sialc 1989. Recruitment of a chromosomally
encoded maleylacetate reductase for degradation of 2,4-
dichlorophenoxyacetic acid by plasmid pJP4. J. Bacteriol. 171:3385-3390.

Lenski, 1992, Relative Fitness: its estimation and its signiﬁcance for
environmental applications of microorganisms in Microbial Ecology (eds.
M. A. Levin, Seideler R.J. and Rogul, M.) McGraw-Hill, Inc. pp183-198

Levin, B. 1972. Coexistence of two asexual strains on a single resource. Science.
175:1272-1274

Mae A.A., O. M. Reet, N.R., Ausmees, V.M.Koiv and AL. Heinaru. 1993.
Characterization of a new 2,4-dichlrophenoxyacetic acid degrading plasmid
pEMT4011: physical map and localization of catabolic genes. J. Gen.
Microbiol. 139:3165-3170.

Murray, R.E., L.L. Parsons and M. S. Smith. 1992, Competition between two
isolates of denitrifying bacteria added to soil. Applied and Environmental
Microbiology. 58:3890-3895.

Neilson, J.W., K.L. Josephson, I.L. Pepper, R.B. Arlnold G.D. DiGiovanni and
NA. Sinclair. 1994. Frequency of horizontal gene tranfer of a large
catabolic plasmid (pJP4) in soil. Appl. Environ. Microbiol. 60;4053-4058.

Rutter, M. and DE. Nedwell, 1994. Inﬂuence of changing temperature on growth
rate and competition between two psychrotolerant antarctic bacteria:
competition and survival in non-steady state temperature environments

Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A
Laboratory Manual. Second Edition. Cold Spring Harbor Laboratory
Press. USA.

Top, E.M., P. Van Daele, N. De Saeyer, and L.J. F omey. 1998. Enhancement of
2,4-dichlorophenoxyacetic acid (2,4-D) degradation in soil by
dissemination of catabolic plasmids. Antonie Van Leeuwenhoek. 73:87-
94.

Top E.M., WE. Holben and L.J. Fomey. 1995. Characterization of diverse 2,4-
Dichlorophenoxyacetic acid-degrading plasmids isolated from soil by
complementation. Appl. Environ. Microbiol. 61 :3274-3281

Wilson, M. and SE. Lindow. 1994. Ecological similarity and coexistence of
epiphytic ice-nucleating (Ice+) Pseudomonas syringe strains and a non-ice
nucleating (Ice-) biological control agent. Appl. Environ. Microbiol.
60:3128—3137.

82

CHAPTER 3

RALSTONIA EUTROPHA AND BURKHOLDERIA CEPA CIA HOST POPULATIONS
AS RESERVOIRS FOR 2,4—DICHLOROPHENOYACETIC ACID DEGRADATIVE
PLASMIDS

SUMMARY

Ralstonia eutropha and Burkholderia cepacia are two of the many
naturally occurring bacterial populations that utilize 2,4-
dichlorophenoxyacetic acid (2,4-D) when they possess catabolic genes for
the upper degradation pathway for this aromatic compound. The work
presented in this chapter explores characteristics that were obtained from
pure culture experiments with R. eutropha JMP228 and B. cepacia DB01
as well as some of the previously documented characteristics of the two
strains. The ultimate goal of these experiments was to determine which
host-plasmid combinations offer the ﬁttest phenotype when competing for
2,4-D as a limiting resource. Before performing the competition
experiments, growth rates and Ks values were determined for pure cultures
of each host-plasmid combination. Experimental results were compared
with models generated for substrate limited bacterial growth, and will be
used to predict the outcome of competition experiments performed in

chapter 4.

83

R. eutropha and B. cepacia are representative of their genera based
upon BIOLOG proﬁles. Different substrate utilization proﬁles were
obtained for the two species on BIOLOG GN plates, with B. cepacia
utilizing a broader range of substrates than R. eutropha. Both strains were
able to reduce Tween 40 and Tween 80, fructose, two esters (methyl
pyruvate and mono-methyl succinate), 15 aliphatic and aromatic carboxylic
acids, 12 amino acids, bromosuccinic acid, urocanic acid, inosine and
phenyl ethylamine. Only two substrates were uniquely utilized by R.
eutropha: succinamic acid and L-leucine. B. cepacia utilized 30 substrates
that R. eutropha was unable to oxidize. These included 12 carbohydrates,
three carboxylic acids, four amino acids, two amines, two alcohols and
three phosphorylated sugars.

When two different plasmids (pJP4 and pEMT3) containing
different cassettes for the degradation of 2,4-D were introduced into these
two different hosts, the growth rates of the host-plasmid combinations
differed. When 2,4-D served as the sole source of carbon and energy,
carrying pJP4 was advantageous relative to hosting pEMT3. Pure culture
growth rate experiments, using 2,4-D as the limiting resource, indicated
that R. eutropha JMP228/pJP4 grew fastest (p: 0.23 h"). B. cepacia
DB01/pJP4 also had a relatively fast growth rate (p: 0.15 h"), while R.
eutropha JMP228/pEMT3 (p: 0.09 h"), and B. cepacia DB01/pEMT3

(p=0.08 h") displayed the slowest growth. Within a host population,
84

neither plasmid impacted growth rate when the populations utilized the
“neutral” carbon source, succinate. This compound was readily
metabolized by either host strain, independent of the presence of the
catabolic plasmid. R. eutropha grew slightly faster (p: 0.49 h") than B.
cepacia (p: 0.47 h") when succinate served as the sole source of carbon

and energy.

INTRODUCTION

The manner in which a population utilizes the resources and grows in
its environment may predict which population will become dominant under
different habitat conditions. This chapter is dedicated to the physiological
and genetic features of R. eutropha and B. cepacia, the two host strains
that will be used in upcoming competition experiments.

Ralstonia eutropha (formerly Hydrogenomonas eutropha,
Alcaligenes eutrophus) (Davis et al., 1969; Yabuuchi et al. 1996) and
Burkholderia cepacia (formerly Pseudomonas cepacia) (Palleroni, 1981;
Yabuuci et al. 1992) were been identiﬁed as two naturally occurring
bacterial populations that utilize 2,4-D when they possess genes encoding
for the degradative pathway. Both bacterial species are members of the B-
proteobacteria (Palleroni, 1992a). Some features are shared among the two
strains. For instance, both genera are gram negative, non-spore forming

rods and the cells tend to occur singly or in pairs, rather than forming long
85

chains or clumps. Cell aggregates are easily suspended in liquid media, and
cultures grow without forming ﬂocks when grown on a rotary shaker.
Neither Ralstonia nor Burkholderia possess sheaths or prosthecae. Both
populations are aerobic chemo-organotrophs, and are capable of growth on
a wide variety of organic substrates. Neither grows ferrnentatively. Both
are non-phototrophic and do not ﬁx nitrogen. In times of substrate excess
B. cepacia and R. eutropha are both capable of producing poly-B-
hydroxybutyrate (PHB) as a storage compound (Palleroli, 1992a; Palleroni,
1992b).

Burkholderia species are metabolically very versatile. Consequently,
these organisms can be found in many habitats. B. cepacia species are
nutritionally diverse, and at least 102 of 136 substrates were utilized by 14
of 18 strains tested by Ballard et al. (1970). B. cepacia maintains a wide
range of antibiotic resistances to penicillin-type antibiotics, chloramphenicol,
aminoglycosides, and polymyxin, largely due to the impermeability of the
cell wall and capsule (Burns et al. 1989). All B. cepacia strains are sensitive
to trimethoprim. For marking the genome of B. cepacia for molecular and
population studies, rifampicin has proved to be most successful, as the
strains can readily be isolated, and resistance is stable. However, there
appears to be a slight affect of rifampicin resistance on the competitive

ability of bacterial strains bearing this marker (Lewis et al., 1987).

86

Genetic diversity is tremendous among Burkholderia cepacia strains,
whose genomic complexity and plasticity are rivaled by few other genera.
With the advent of Pulsed Field Gel Electrophoresis, the visualization of the
large DNA molecules of the bacterial genome became possible, and sparked
the discovery that the B. cepacia genome contains multiple chromosomes
(Cheng et al., 1994; Hendrickson et al., 1996). On average, B. cepacia
isolates contain two to four replicons with total genome sizes spanning the
range of 5 to 9 Mb. B. cepacia strain DB01 (Keyser et al., 1976), exhibits
three large replicons of 3.4, 2.4 and 1.3 Mb, with a total genome size of 7.1
MB. Each large replicon has at least one set of rrn operons. In strain DB01,
a total of 6 rm operons were found among its 3 chromosomes.

Mobile genetic elements are commonplace in Burkholderia strains,
and together with the different chromosome numbers, genome sizes and
arrangements, support the idea that lateral gene transfer and genetic
rearrangement is a prominent feature of this genus. IS elements are
common, and have been implicated as a force driving diversiﬁcation of the
genome and the ability to acquire novel catabolic functions, such as the
ability to degrade the herbicide 2,4,5-T (Haugland et al., 1990; Daubaras et
al., 1995) The presence of plasmids is also common among populations of
B. cepacia. Some confer resistance to antibiotics and antibacterial agents.
(Chakrabarty et al., 1976; Jacoby, 1986). Some may provide resistance to

metals such as mercury, borate and terllurite. Other plasmids have been
87

shown to increase the catabolic range by providing pathways for the
utilization of aromatic compounds such as toluene, benzene, napthalene and
salicylate.

Based on rRNA similarity groups, Palleroni (1992b) categorized B.
cepacia as a Group II Pseudomonad. As such, B. cepacia clusters with the
B—subgroup of the Proteobacteria, and shares a common ancestry with
Alcaligenes species. Both these genera are genetically distinct from the
members of the y—subclass, that includes the type species of the
Pseudomonads, Pseudomonas aeruginosa. A comparison of the 16S rRNA
of the Ralstonia and Burkholderia groups indicates that members of
Burkholderia share 94% sequence similarity with each other, the genus
Ralstonia shares from 93-95% similarity and the two groups are joined at a
level of 88% sequence similarity (Y abuuchi et al., 1992). The members of
the B. cepacia group exhibit greater than a 98% similarity in the 16S rRNA
sequence (Tabacchioni et al., 1995).

Ralstonia eutropha, is a B-proteobacteria and shares many
physiological characteristics with B. cepacia. R. eutropha is also ubiquitous
in soil and freshwater environments, and has frequently been isolated from
habitats with high levels of aromatic or heavy metal contaminants. These
peritrichously ﬂagelated bacteria function as facultative chemolithotrophs,
capable of oxidizing hydrogen. They are oxidase positive, catalase positive,

and are capable of anaerobic growth with nitrate as the electron acceptor.
88

R. eutropha utilizes more carbohydrates than the type strain of Alcaligenes
from which it was reclassiﬁed in 1992. Among its distinguishing catabolic
traits are that hydroxybenzoate is metabolized via the gentisate pathway, and
protochatechuate via an ortho cleavage pathway (Davis et al., 1969). A
variety of aromatic compounds can be utilized, and the regulation of the B-
ketoadipate pathway was described by the classic work of Johnson and
Stanier (1971). R. eutropha makes frequent use of large catabolic plasmids
for increasing its metabolic potential. Those isolates obtained from sites
contaminated with chlorinated aromatic compounds, such as herbicides,
frequently possess large broad-host-range plasmids like pJP4 and pAC27,
which carry the genes for degradation of 2,4-dichlorophenol and 4-
chlorophenol.

While many of the typical characteristics of Reutropha and B.
cepacia are known from literature, this research focused on a few of the
phenotypic traits that could be used to distinguish between the two
population when they are used as hosts for two different catabolic plasmids,
pJP4 and pEMT3 (Ka et al., 1992; Top et al. 1995). The 2,4-D degradative
pathway encoded by plasmid pJP4 was discussed in Chapter 1. Another
plasmid, pEMT3, also encoding the 2,4-D degradative pathway, was
obtained by Top et al. (1994). The genetic and physiological
characterization of this plasmid is shown in Table 2.1. Plasmid pEMT3

belongs to the Inc-P incompatibility group, like pJP4. Consequently,
89

pEMT3 carries a similar set of genes for its own transfer and replication.
Although the gene encoding for the ﬁrst enzyme of the 2,4-D degradative
pathway (tfdA) is homologous among these two plasmids, pEMT3 carries a
different set of 2,4-D gene alleles for the tdeDFE operon and for the tde
gene. This plasmid’s phenotype is reﬂected in slower growth when 2,4-D

is supplied as a sole source of carbon and energy as compared to pJP4.

MA TERIALS AND METHODS

Antibiotic Resistance proﬁles. Before plasmids were introduced to
the host strains to construct the different host-plasmid combinations,
antibiotics were evaluated for their ability to serve as suitable selective
agents for the plasmid-free populations Ralstonia eutropha JMP228 and
Burkholderia cepacia DB01. Rifampicin (50, 100, and 150 pg ml"),
streptomycin (100, 250, and 500 pg ml"), trimethoprim (20, 40, and 80
pg/ml), tetracycline (15, 20, and 25 pg ml"), chloramphenicol (25, 50, and
75 pg ml"), kanamycin (25, 50, and 75 pg ml"), naladixic acid (10, 15,
and 20 pg m1") and ampicillin (50 and 100 pg ml") were tested at 22°C.
R. eutropha and B. cepacia were inoculated from freezer stock into 13
media (Sambrook et. a1, 1990), and allowed to grow for 12 hours. A 5%
(vol/vol) inoculum was then transferred to 50 ml fresh LB medium, and
grown to late log phase as determined by absorption readings at 600 nm.

Aliquots (100 pl) were directly plated onto LB plates containing each of the
90

antibiotics to determine the susceptibility and the frequency of appearance
of resistant mutants. A sample was diluted and plated to LB media to
determine the concentration of viable cells in the culture. The frequency of
appearance of the resistant phenotype was calculated as the number of
resistant colonies/ml divided by the total CFU ml".

Bacterial strains. Ralstonia eutropha JMP228 (referred to as
JMP228), the pJP4 cured strain of R. eutropha JMP134 (Don and
Pemberton, 1981) and Burkholderia cepacia DB01 (referred to as DB01)
(Ka and Holben 1994) were used as the recipient host strains for either of
two 2,4—D degradative plasmids. Spontaneous mutants of JMP228,
resistant to rifampicin or streptomycin were used to distinguish between
plasmid bearing populations, and these strains were called JMP228R and
JMP228S. The JMP228R was obtained from Top et. al.(1995) and the
streptomycin resistant strain was a spontaneous resistant mutant generated
for these competition experiments. JMP2288 was obtained by growing a
culture of JMP228 in minimal medium containing succinate (5 mM) to late
log phase. A one ml aliquot of the cell suspension was rinsed with sterile
phosphate buffered saline (PBS) and resuspended in 100 pl PBS. This
suspension was serially diluted in ten-fold increments and 100 pl was
spread over the surface of a minimal medium agar plate containing 2,4-D
(200 pgml") and streptomycin (100, 150, 200 and 250 pgml"). Three

colonies which grew on the 10'2 dilution plate with 150 pg ml"
91

streptomycin were clonally puriﬁed on the same medium three times. All
spontaneous mutants were equivalent in terms of growth rates and colony
morphology. Loss of the streptomycin resistance marker in the absence of
selection pressure was not observed. No difference was observed in the
growth rates of cultures of JMP288S grown in the presence or the
absence of 150 pg ml" streptomycin. The number of colonies isolated on
R2A agar containing or lacking streptomycin was equal for JMP228S with
or without selective pressure. B. cepacia strain DB01, served as the other
host strain Two antibiotic resistant strains were generated using rifampicin
(100, 150, 200, 250 pgml") and trimethoprim (2.5, 5 and 7.5 pgml").
Three colonies from the highest concentration were repeatedly transferred
for puriﬁcation. After verifying that there were no revertants, the resistant
Burkholderia strains were designated DB01R and DB01T. In addition to
the antibiotic resistance markers, Ralstonia and Burkholderia strains could
be distinguished based on the colony morphology on R2A medium. The

bacterial strains used in the experiments are listed in Table 3.1.

92

 

Table 3.1. Bacterial Strains used in the competition experiments

 

 

Species Strain Antibiotic Reference
resistance
Ralstonia eutropha J MP228 rifS strs Don and Pemberton 1981
Ralstonia eutropha JMP228R riP‘ strs Top et al. 1994
Ralstonia eutropha JMP228S rifis strR This study
Burkholderia cepacia DB01 rifs triS Holben et al. 1993
Burkholderia cepacia DB01 R rilR triS This study
Burkholderia cepacia DBOlT rifs triR This study

 

Plasmids and transconjugant construction. Plasmid pJ P4, isolated
from R. eutropha JMP134, belongs to the Inc-P incompatibility group, and
carries the most thoroughly characterized 2,4-D degradative pathway (Don
1985). Plasmid pEMT3 was isolated in a previous study as a result of an
exogenous plasmid trapping experiment using strain JMP228R as a
recipient (Top 1994). Plasmid pEMT3 is an Inc-P plasmid, and carries a
tfdA gene which hybridizes under high stringency conditions to a gene
probe generated from the internal fragment of the tfdA gene of plasmid
pJP4. In contrast, using internal ﬁ'agments of tde and tde genes of
plasmid pJP4 as probes, successful hybridization to pEMT3 only occurs
under intermediate and low stringency conditions, indicating differences in

the sequences of these genes (Top 1995) (Table 3.2).

93

 

Table 3.2: Plasmids used in the competition experiments

 

Hybridization a

 

tfd genes of pJP4 IncP replicon
Plasmid tfdA tde tde rep R Reference
pJP4 *" *** *** *** Don and Pemberton 1981
pEMT3 *** ** * *** Toget al. 1994

 

a ”* indicates hybridization under high stringency conditions and correlates to >90%
sequence similarity. ** hybridizes under medium stringency and 75-90 % sequence
similarity. "‘ hybridizes under low stringency conditions and represents 60 - 75 %
sequence similarity

Plasmids pJP4 and pEMT3 were transferred from JMP228R/pEMT3
or JMP228/pJP4 into strains R. eutropha JMP228R and JMP228S using
ﬁlter mating. Donor cells (1x107) were combined with recipient cells
(2x106) for 8 hours on a polycarbonate ﬁlter (Poretics 13 mm, 0.2 pm
pore size) placed atop LB agar. The cells from the ﬁlter were resuspended
in A+N minimal medium broth (composition below), without a carbon
source. A dilution series was plated to medium containing antibiotics to
prevent donor strain growth, and 2,4-D (200 ppm) as a sole source of

carbon to select for transconjugants.

94

 

Plasmid pJP4 was introduced to B. cepacia DBOlR and DB01T as
described for the JMP228 populations. Plasmid pEMT3, however, had to
be introduced to B. cepacia DB01R and DBOlT by combining donor cells
(1x107) and recipient cells (2x106) onto a single spot on R2A agar (Difco).
After 12 hours of incubation, colony material was transferred with an
inoculating loop to A+N minimal medium containing 2,4-D and antibiotics
to counterselect the donor strain. The presence of the plasmid was
veriﬁed by isolation of plasmid DNA using the QIAGEN midi-prep column
(QIAGEN). Plasmids from donor strains and transconjugants were
compared by RFLP analysis Eco-RI (Gibco) digest (Fig 3.1).
Transconjugants are listed in table 3.3. All cultures were archived in

selective media containing 15% glycerol, ﬁnal concentration, at -80 °C.

95

 

Table 3.3: Transconjugants used in the competition experiments

 

Transconjugant Donor Recipient 2,4-D Antibiotic

. . . a .
utilization resrstance

 

JMP228/pJP4 ++ rif5 s05 n15
JMP228R pEMT3 + rif" strs tris
JMP228R - iii" strS tris
JMPZZSS - rir‘s strR tris
DBOlR - n'l“ tn'S
DBOlT - rifs tn"
JMP228R/pJP4 JMP228/pJP4 JMP228R ++ n1“ strS tn'S
JMP228S/pJP4 JMP228/pJP4 JMP228S ++ sirs strR tris
JMP228R/pEMT3 indigenous JMP22 8R + rifR strs tn'S
JMP228S/pEMT3 JMP228R pEMT3 JMP228S + n15 snR tnS
DB01R/pJP4 JMP228/pJP4 DBOIR ++ n1“ tris
DB01T/pJP4 JMP228/pJP4 DBO 1 r ++ n15 triR
DB01R/pEMT3 DBOl/ pEMT3 DBOIR + n1“ n13
DB01T/pEMT3 JMP228R/pEMT3 DB01T + r115 n—iR

 

a A minus indicates that no growth with 2,4-D occurred, a plus indicates slow growth,
two plusses indicate fast growth..
The antibiotics rifampicin (rif) and streptomycin (str) were used at a ﬁnal
concentration of 250 pg ml’l and trimethoprim (tri) resistance was used at a ﬁnal
concentration of 75 pg ml".

96

 

‘I.-n-.. «a .

5 67 8 9101112131415161718192021

 

Figure 3.1: Restriction of plasmids isolated from transconjugants and
cut with enzyme EcoRl. Lanes 1, 5, 21 HindIII digested lambda DNA.;
Lane 2, plasmid pJP4 from JMP134 (Don and Pemberton, 1981); Lane 3,
JMP228r/pJP4; Lane 4, JMP228S/pJP4; Lane 6 plasmid pEMT3 from
K19 (JMP228/pEMT3) (Top et al. 1995); Lane 7 JMP228r/pEMT3; Lane
8, JMP228S/pEMT3; Lane 9, JMP228r; Lane 10, B. cepacia DB01; Lane
11, DB01r/pJP4; Lane 12, DB01t/pJP4; Lane 13, DB01t/pEMT3; Lane
14, DB01t/pEMT3-212a; Lane 15, DB01t/pEMT3-212b1; Lane 16,
DB01t/pEMT3-212b2; Lane 17, DB01t/pEMT3-213; Lane 18,
DB01r/pEMT3 Lane 19, DB01r/pEMT3-236; Lane 20, DB01t/pEMT3-
241. The DB01 transconjugants in Lanes 14, 15, 16, 17, 19 and 20 were
not used in this study. Lanes 14, 15, 16 and 20 suggest changes in the
plasmid backbone of pEMT3 could occur in strain DB01.

97

 

Cultivation of cells. Cells were grown in A+N mineral salts medium to

which carbon sources and antibiotics were added (Wyndham, et al., 1986).
A+N medium consisted of 1.71 g KZHP04-3HZO, 0.34 g N aHzPO4-H20,

0.33 g (NH4)2S04, 0.25 g MgS04-7H20, and 10 ml of a trace metals

solution. The trace metal solution contained, per liter: 12.0 g
NazEDTAa-ZHZO, 2.0 g NaOH, 0.4 g ZnSOa-7H20, 0.4 g MnSOa-4H20,
0.1 g CuSO4-5HZO, 3.0 g FeS04-7HZO, 5.2 g NaZSOa, 0.1 g

NaMoOa-ZHZO. (Wyndham 1986). The pH was adjusted with 10 N H2S04
to 7.2 prior to autoclaving (45 min, 121 psi). Prior to autoclaving,
amendments of 5 mM sodium succinate (Sigma), or 1.7 % agar (Difco,
Detroit, M1) were added if needed. Stock solutions of 2,4-D (20 mg ml"
in 0.1 M NaH2 P04, pH7.2) (Sigma), rifampicin (25 mg ml" in 100 %
methanol), streptomycin (25 mg ml" in water) and Trimethoprim (7.5 mg
ml" in 100% methanol) were ﬁlter sterilized prior to supplementing
autoclaved media. Solid LB agar medium consisted, per liter, of 10 g yeast
extract (Difco), 5 g tryptone (Difco), 5 g NaCl (Sigma) and 17g agar
(Difco). For each independent experiment, cultures were started from a
frozen culture stock. The cell cultures were grown to late log phase in 16
ml test tubes with 3 ml A+N minimal medium amended with 2,4-D (200
pgml") and the appropriate antibiotic (rifampicin, streptomycin or

trimethoprim). A 2.5% (vol/vol) inoculum to A+N medium with succinate
98

(SmM) or 2,4-D (200 ppm) was used to start any other culture used in the
growth and competition experiments. Population identiﬁcation in the
competition experiments was carried out using 2,4-D minimal medium
plates supplemented with the appropriate antibiotics (Table 3.3).
Differences in colony morphology among Ralstonia and Burkholderia
were determined on R2A agar plates.

Substrate utilization. Substrate utilization proﬁles were determined
using BIOLOG GN plates with 95 different carbon sources. JMP228r,
JMP2283, DB01r and DB01t were grown in A+N media supplemented with
5 mM succinate to an absorbance of 0.3 at 600 nm. The bacteria were
harvested by centrifugation at 5000 rpm (Sorvall SS34 rotor), rinsed twice
in sterile saline (0.85% NaCl), and suspended to an absorbance reading of
0.1 at 600 nm in sterile saline. Triplicate BIOLOG plates were inoculated
with the different strains by adding 300 p1 of culture ﬂuid to each well.
Development of purple color in the wells due to reduction of the reazurin
dye was used as the criterion for substrate use. Color formation was
determined visually after 8, 24 and 48 hour intervals. The 48 hour readings
were used as the endpoint, as no additional wells developed color when
incubated longer.

Growth Parameters. Culture Growth Rates. Growth rates of the
different strains were determined with succinate and 2,4-D, using two two

methods described in chapter 2. In the traditional method, three replicate
99

 

cultures were initiated in 100 ml of medium in a 500 ml Erlenmeyer ﬂask
with a l % (vol/vol) inoculum. The Flasks were incubated at 30 °C in a
rotary shaker (200 rpm, 30 °C). During grth with 2,4-D, 3 m1 samples
were periodically removed to determine the viable population size, optical
density, and concentration of 2,4-D. The optical density of a growing
culture was monitored spectrophotometrically with an HP 8452A diode
array spectrophotometer (Hewlett Packard, Waldbronn, Germany) at 600
nm, and the growth rates were determined by regression analysis. Bacterial
growth rates were calculated using the equation N1 = Noe“t where N is the
population size, and No is the population size at time 0. p is the growth rate
and t is the time elapsed from No to N]. The variable to solve for is p, and
it is the slope of the line from No to N1 during time t. During exponential
growth, an increase in population size is reﬂected by a linear increase in
absorbance. The growth rate p can be determined from the slope of the
straight line obtained on a log-linear plot by regression analysis during
exponential growth. The maximum growth rate for a substrate is the
intrinsic rate of increase in cell number above which additional substrate
will not allow for more rapid growth.

The second method determined growth rates in 96-well microtiter
plates. Optical densities at 590 nm were measured using a BIO-TEC
Microtiter EL312e microplate reader (BIO-TEK instruments Inc.,

Wenooski, VT) as described in Chapter 2. The data was saved to a
100

 

computer disk and analyzed using Microsoft Excel 5.0. Growth rates were
determined by regression analysis.

Disappearance of Substrate. Concentrations of 2,4-D were
determined with High Performance Liquid Chromatography (HPLC). The
samples were prepared as described previously in chapter 2. To calculate
the concentrations of 2,4-D in the samples, the peak areas were compared
with a 6-point calibration curve ranging from S to 250 pg ml" 2,4-D
standards. The samples were analyzed with a Hewlett Packard Series 1050
HPLC with a 250 mm x 4.6 mm Lichrosorb RP-l8, 10 pm particle size,
chromatography column (Alltech, Bannockbum, IL) at a ﬂow rate of 1.5
ml per minute and with 70% methanol and 30% phosphoric acid (0.1%)
(vol/vol) as the solvent.

The Half-Saturation Constant. The half saturation constants (Ks)
were determined for strains JMP228R/pJP4, JMP228R/pEMT3,
DB01R/pJP4 and DB01R/pEMT3, by measuring the rate of uptake and
incorporation of ring-labeled l4C- 2,4-D. 100 ml-cultures of each strain
were grown to late log phase with 200 pg ml" 2,4-D in A+N medium. The
optical densities of the cultures were measured and an appropriate volume
of culture ﬂuid that provided 1 x 1010 CFU ml" , were harvested by
centrifugation at 8000 rpm (Sorvall SS34 rotor) for 10 minutes. The cell
pellets were rinsed twice with phosphate buffered saline (PBS) solution

(Lbfﬂer, 1997), and resuspended in 1 ml PBS. CFU ml", were determined
10]

for each sample. Actual cell counts were approximately 2 x 109 CFU ml".
At time zero, a sample was removed from each vial before inoculation with
the test culture. A 10 pl aliquot of the dense cell suspension was added at 1
minute intervals to 100 ml serum bottles each containing 10 m1 of a
solution containing 14C ring-labeled 2,4-D at concentrations of 10, 1, 0.2,
0.1, 0.67 and 0.05 pg ml" 2,4-D. The bottles were capped with Teﬂon-
lined rubber stoppers, crimp-sealed, placed on a rocking platform after
inoculation, and incubated at room temperature. At each of the designated
intervals of 0, 6, 13, 20, 27, 34 and 41 minutes, 0.4 ml samples were
removed from each bottle using a syringe. The samples were transferred to
Centricon micro-concentrator ﬁlter units (Amicon) containing 40 pl 2N
HCl in the bottom of each unit. After centrifugation at 12,000 rpm in a
microcentrifuge for 1 minute, the ﬁlters containing the cells were placed
into scintillation vials containing 8 ml of scintillation ﬂuid. The micro-
concentration tubes containing the ﬂow through, without the caps, were
placed upside down into another scintillation vial. The DPMs for all
samples were counted with a scintillation counter. The resulting counts were
transformed into pg m1" 2,4-D using a standard curve generated using
samples with known 2,4-D concentrations and the DPMs obtained for the
different dilutions. 2,4-D uptake curves for each strain were ﬁtted to ﬁrst
order degradation kinetics using the Kaleidagraph software. An example of

the degradation curves is shown in Figure 3.2, for strain DB01R/pJP4.
102

‘_ "ha-Ii‘l‘u'u in tit—“1
o
1
\

Similar depletion curves were prepared for the four strains tested. The
equations generated by the graphs were used to calculate the initial rates of
2,4-D uptake at each of the substrate concentrations mentioned above. The
inverse of the substrate concentrations and rates, were used to create
Lineweaver-Burke graphs for each of the strains by plotting US over l/V
(Figure 3.3). The point at which the line intercepted the X-axis represented
-1/Ks, which was then used to ﬁnd the half saturation constant for

JMP228R/pJP4, JMP228R/pEMT3, DB01R/pJP4 and DB01R/pEMT3.

 

 

 

 

Time (min)

Figure 3.2. 2,4-D depletion curves used to calculate rates of 2,4—D
degradation for use in calculating Ks for strain DB01R/pJP4. This
graph illustrates curves for 10, 1 and 0.2 ppm. Curves for 0.067 and 0.05.
ppm cannot be resolved on this scale. Kaleidagraph software was used to
determine the initial degradation rate for each curve, which was then used
to generate a lineweaver-burke plot.

103

 

 

3500 7
3000
2500

1500 ~

1/[S]

500' 'f

 

-5 -5000 5 10 15 20 25 30 !

L_.___- __ __-

Figure 3.3. Lineweaver—Burke plot to determine the Ks for each of the
four host/plasmid combinations. Open diamonds are DBOl/pEMT3,
ﬁlled diamonds are DB01/pJP4, light grey circles are JMP228/pEMT3, dark
ﬁlled circles are JMP228/pJP4. The Ks values calculated for each of the
strains based on the interception of the line with the X axis are 1.2 ppm, 0.9
ppm, 0.8 ppm and 1.2 ppm respectively.

Chemostat Conditions: Chemostats of each pure culture were
operated as described in Chapter 2 for a duration of 12 days. Operating
parameters were as follows. A. Succinate-grown culture in mid-late log
phase (optical density of ~ 0.1 at 600 nm, about 2-3 x 107 CFU ml"), was
gently harvested (4,000 rpm for 20 minutes in a Sorvall SS34 rotor), rinsed
with A+N minimal medium, and suspended to a ﬁnal OD600 of 0.2 in the
same medium. This cell suspension (75 ml) was placed in a chemostat and
2,4-D was added to a ﬁnal concentration of 200 pg ml". Each chemostat
Was incubated for 8 hours at 30° C, before a constant ﬂow conditions were

104

initiated. The dilution rate (D) of a chemostat determines the growth rate
(p), and for the 75 ml volume chemostats, D= p = 0.033 h" which is well
below the maximum grth rate for the strains under investigation. To
sample the chemostats, the port was swabbed with 70% ethanol and 3 ml of
culture was removed with a 3 ml syringe. The withdrawn sample was used
to obtain measurements of the 2,4-D concentration (HPLC), optical density
and the CFU ml". Culture ﬂuids aliquots (100 p1) were plated onto A+N
agar plates containing 200 pg ml" 2,4-D and R2A agar plates, both
amended with the appropriate selective antibiotics to obtain viable cell
numbers. Three measurements of each chemostat were used for each time

point to determine the average population size.

Equation (a)

dN

=N S -

In order to evaluate the grth of the different strains in chemostats,
the values for pm“, Ks and yield were inserted into the Monod equation (a)
(Dykhuizen and Hartl, 1983). Then models were run using the Stella
Research Software 5.0 (High Performance Systems Inc. Hanover, NH). The
critical equation in this model is the Monod equation: dN/dt is the rate of
change in population size, N is the population size, pmax is the maximum

growth rate on the resource and S is the substrate concentration. The model

105

automatically calculates the population sizes over time and the remaining
concentration of 2,4-D, based on the substrate utilization equation by Monod.
Equation (b) represents the utilization of the substrate in the chemostats for a
single population. S0 and S are substrate concentration in the reservoir and
chemostat respectively. Y represents the growth yield, (CFU/mg substrate).
An illustration of the model is presented in Figure 3.4. The values to be
tracked are population size and 2,4-D concentration, shown by the black
boxes in Figure 3.4. Arrows connected to clouds indicate the direction of
energy ﬂow, and the equation associated with the arrow will calculate the rate
at which the population size or resource concentration changes depending on
the various inputs represented by open circles (in this case using the Monod
equation. The efﬁciency of utilization (mg/CFU) is the inverse of the growth

yield, CFU/mg 2,4-D, as calculated for strains grown on 2,4-D medium. The

dilution rate (D) determines the grth rates that the populations

experience.

 

Equation (b)
_di = D (50-3) - YN S
dt Ks + S

106

 

Bacterial
growth Washout

—-‘ i ": ___“
a w B ICICI l lgw.

 

   

 

Figure 3.4. Growth model developed on the Stella Research Program
to generate expected p0pulation growth curves and 2,4-D depletion
curves for pure cultures. The constants for KS, growth rate, and efﬁciency
Were chosen using an average value determined for each host/plasmid
Combination. The initial population size (N) and 2,4-D concentrations (200
Ppm) that were used for the model, were obtained using values from the

CXperimental chemostats.

107

RESULTS

Antibiotic resistance proﬁles. To determine which antibiotics to use

as selective markers, 1.7 x 108 to 2.7 x 108 cells of each bacterial strain

were plated to LB agar plates containing the different antibiotics.

Table 3.4: Testing the antibiotic resistance profiles of strains

 

Number of resistant colonies / total number of

 

Antibiotic pg/ml R. eutropha JMP228 B. cepacia DB01
Rifampicin 50 K1 x10W
100 9. x10")9 3.7 x10’09
150 6.2 x1009 4.0 x10’09
Streptomycin 100 3-9 X1009 TNTC‘
250 5.8 x1009 7.4 x10")7
500 3.9 x1009 9.8 x1009
Trimethoprim 20 TNTC 2.3 X1008
4o TNTC 5.8 1110‘08
30 7.5 x10‘07 0
Tetracycline 15 0" TNTC
20 0 TNTC
25 0 3.7 x10’06
Chlorarnphenicol 25 TNTC TNTC
50 5.1 1110-"7 TNTC
75 6.4 x1008 1.6 x1006
Kanamycin 25 0 O
50 0 0
75 0 0
Naladixic acid 10 3.4 x1007 TNTC
15 1.5 xlOm TNTC
20 6.1 x10’08 1.4 x1007
Ampicillin 50 TNTC TNTC
100 TNTC TNTC

8.8 x10”;

 

IN'I'C indicates a lawn of bacteria, or colonies were too numerous to be counted on the

plate.

0 indicates that no visible colonies were formed on the plate.

108

Table 3.4 displays the antibiotics that were tested. Chloramphenicol,
naladixic acid and ampicillin were not suitable, since the B. cepacia strains
were resistant, and the R. eutropha strains readily generated spontaneously
resistant mutants. Kanamycin produced no spontaneous mutants for all of
the populations tested. Results of the selection experiments and literature
review (Lewis 1987) indicated that rifampicin would be a good choice for
both R. eutropha and B. cepacia. Only a few spontaneous mutants arose,
and the mutants did not lose this phenotype even after repeated transfer
through media without selection in the presence“ of the antibiotic.
Rifampicin is bacteriostatic and inhibits RNA synthesis. Frequently
resistance is obtained through a mutation in the B—subunit of RNA
polymerase that provides resistance by blocking the rifampicin from
complexing with the RNA polymerase. Trimethoprim allowed for growth
of B. cepacia, and not R. eutropha, which would give a selective means of
distinguishing between the two different host strains, and to allow for
competition between B. cepacia populations resistant to trimethoprim or
rifampicin. Trimethoprim is a drug that inhibits dihydrofolate reductase.
Streptomycin, while allowing for the growth of B. cepacia, was appropriate
for use in distinguishing between two populations of R. eutropha.
Streptomycin is a bacteriocidal antibiotic that inhibits protein synthesis.

Substrate utilization proﬁles. Of the 95 carbon sources tested by the

BIOLOG-GN plate, R. eutropha utilized 38 of the 95 substrates. B. cepacia
109

 

 

rm ---- manna—n

DB01 utilized 63 of 95 carbon sources. Only two substrates were used by
the Ralstonia strains that were not used by Burkholderia: succinamic acid
and L-leucine. On the other hand, Burkholderia utilized 30 substrates that
were not used by Ralstonia. (Table 3.5). Utilization was based on
development of a deep purple color in the wells of the microtiter plate.
Inconclusive color formation was not scored.

Pure culture growth characteristics. R. eutropha colonies are
smooth, shiny and translucent on R2A, agar plates, but small (< 0.2 mm in
diameter) and translucent when grown on A+N plates without any carbon
source after 48 h. Growth without an added carbon source is not
completely unexpected, as the agar may contain impurities that could be
used for growth. Also, R. eutropha is a chemoautotroph, and can grow
slowly by ﬁxing carbon from the atmosphere (Palleroni, 1987). As the
colonies developed over a week, their edge became irregular. When grown
on A+N plates with 2,4-D the colonies carrying a degradative plasmid
(pJP4or pEMT3) did have a larger diameter (1 mm) than those on the no-
Carbon control plates, and they are much more robust (what is a robust

Colony?) In the presence of antibiotics no growth occurred on plates

lacking a carbon source.

110

Table 3.5:

JMP228 and Burkholderia cepacia DB01.

BIOLOG substrate utilization by Ralstonia eutropha

 

 

 

 

 

 

 

 

 

 

 

Used by all Used by none B. cepacia R. eutropha
only only
N-acetyl-D- N-acetyl-D-
Carbohydrates D-Fructose galactosamine glucosamine
Adonitol L-arabinose
Cellobiose D-arabitol
I-Erythritol L-fucose
Gentibiose D-galactose W-
Lactulose a-D-lactose .
Maltose D-mannitol
D-melibiose D-mannose
B-methyl-D- D-rafﬁnose
glucoside
L-rhamnose D-Sorbitol
Turanose Sucrose :—
Xylitol D-trehalose
D-psicose
Esters Methyl
pyruvate
Mono-methyl
succinate
Aromatic Urocanic acid Uridine
Inosine Thymidine
Amines Phenyl Putrecine
ehylamine
2-amino
ethanol
Amides Alaninamide Glucuronamide Succinarnic
acid
Alcohols 2,3—butanediol Glycerol
Phosphorylated Glucose-l- D,L, 0t-
Compounds phosphate glycerol
phosphate
Glucose-6-
phosphate
Brominated Bromosuccinic
acid
Polymers Tween 40 a-cyclodextrin
Tween 80 Dextrin
Glycogen

lll

 

 

 

Used by all Used by none B. cepacia R. eutropha
only only
Carboxylic Acetic acid D-galacturonic D-galacturonic
Acids acid lactone acid
Cis-aconitic acid or-hydroxy
butyric acid
Citric acid 'y-hydroxy
butyric acid
F ormic acid or-ketobutyric
acid
D-gluconic acid or-ketoglutaric
acid
B-hydroxy Ot-ketovaleric
butyric acid acid
p-hydroxy
phenylacetic
acid
Itaconic acid
D-L-lactic acid
Malonic acid
Propionic acid
Quinic acid
D-saccharic acid
Sebacic acid
Succinic acid
Amino Acids D-alanine Glycyl-L- Hydroxy-L- L-leucine
aspartic acid proline
L-alanine Glycyl-L- L-omithine
glutamic acid
L-alanyl glycine D-serine
L-asparagine D,L-Camitine
L-aspartic acid y—amino
butyric acid
L-glutamic acid
L-histidine
L-phenylalanine
L-proline
L-pyroglutamic
acid
L-serine
L-threonine

112

 

B. cepacia colonies were creamy white with an opaque appearance on
R2A agar. Older colonies tended to develop lumpy edges, whereas younger
ones were smooth. The trimethoprim resistant strain name-it exhibited two
distinguishable colony morphologies at a constant l-to-l-ratio: smooth large
mucoid colonies and ﬂatter, smaller, colonies with a dry appearance.
Neither morphology bred true. Mucoid colonies yielded both colony
morphologies, as did the drier ones. Purity of the culture was conﬁrmed
microscopically, by repeated palting and transfer of well-isolated colonies,
and by molecular analyses. Colonies grown in the absence of antibiotics
tended to be larger than those grown in the presence of antibiotics. This
was especially true for DBOlT grown with trimethoprim. Colonies were
consistently smaller when grown in the presence of the antibiotic, but still
larger than the colonies of strain DB01r when grown on R2A plates. The
colonies were transparent pinpoints on A+N plates without a carbon source.
They remained the same size for 3 weeks. On 2,4-D minimal media,
colonies appeared after a period of 72 hours, and developed a colony
morphology that was indistinguishable from the R. eutropha type. A
distinct difference occurred between the growth of JMP228 and DB01 on
R2A plates. R. eutropha colonies became visible after approximately 24
hours, whereas at least 36-48 were needed for B. cepacia colonies to

appear. However, colonies of the B. cepacia strains grew to be as large as

113

 

R. eutropha colonies within one week, in fact, the colonies continued to
become even larger when incubated longer.

Plasmids pJP4 and pEMT3 both beneﬁt their host strains Ralstonia
eutropha and Burkholderia cepacia by supplying a catabolic pathway which
can convert the herbicide 2,4-D to metabolic intermediates that can then be
channeled into the TCA cycle. Plasmids pJP4 and pEMT3 carrying the
2,4__D degradative pathway and different antibiotic resistance genes were
easily conjugated into strains JMP228R and JMP2288. Plasmid pJP4
readily conjugated into B. cepacia DB01, but pEMT3 could not be
transferred to the same strain when the matings were performed on LB
medium. Transconjugants were, however, obtained from matings on R2A
agar.

The growth rates of the two host populations differed only slightly
when succinate (5 mM) served as the source of carbon and energy. When
grown in batch cultures, strains JMP228 and DB01 had very similar growth
rates of 0.448 $0.047 h" and 0.416 $0.020 h", respectively (n = 18, P =
0.089). When pure cultures were grown on succinate under 24 hour batch
transfer conditions, the JMP228 strains attained densities of 1 x 109 (:1: 0.8 x
108) CFU ml". DB01 strains attained densities of 5 x 108 (:t 0.5 x 108)
CFU ml" (Table 3.6). The results of a one-tailed T-test for comparing the

strains using the batch culture data are shown in Table 3.7. Pairs 1 to 5

114

-1 Ana.” ‘

 

refer to the succinate-grown cultures, and pairs 6 to 10 are the 2,4-D-grown

cultures.

Table 3.6. Growth rates of strains supplied with succinate or 2,4-D
under batch culture conditions. Growth rates are in h", :1: SD.

 

 

 

l
113 n=6 n=if n=6
strain Succinate, Succinate, 2,4-D 2,4-D
microtiter Flask microtiter ﬂask
(h") (h") (h") (h")
JMP228 0.62 110.11 0.45 $10.02 3".
JMP228S nd 0.43 10.013
DB01 0.50 10.15 0.36 10.02
DBOlr nd 0.46 10.0531
DB01t nd 0.43 10.033
JMP228R/pJP4 0.52 110.11 0.54 10.09 0.24 1 0.07 0.22 10.06
JMP228S/pJP4 0.54 110.16 0.46 10.03 0.23 1 0.06 0.23 10.05
JMP228R/pEMT3 0.36 110.03 0.37 10.02 0.10 1 0.02 0.10 10.01
JMP228S/pEMT3 0.46 110.03 0.51 10.05 0.10 1 0.02 0.10 10.01
DB01R/pJP4 0.52 110.23 0.49 10.16 0.13 1 0.02 0.16 10.04
DB01T/pJP4 0.47 110.09 0.53 10.22 0.13 1 0.02 0.15 10.04
DB01R/pEM'1‘3 0.48 110.04 nd 0.09 1 0.02 0.09 1 0.03
DBOlT/pEMT3 0.44 110.04 nd 0.09 1 0.02 0.08 10.03

 

“ For these growth rate calculations, n= 3.

JMPZZS/pJP4 grew rapidly (n = 0.23 :1: 0.05 h") when 2,4-D was

supplied as a limiting resource. A lag period of 1 to 2 hours was observed

115

when succinate-grown cells were inoculated into medium amended with
2,4-D. In contrast, no lag time was observed when a culture exponentially
growing with 2,4—D served as the inoculum. When stationary phase was
reached, Ralstonia strains attained a population size of 2 x 108 CFU ml"
(Fig. 3.5). In the slow growth rate conditions in the chemostats with
dilution rates equal to 0.03 h", the populations attained cell densities of
about 8x108 CFU ml".

In contrast to JMP228/pJP4, the growth rate of strain
JMP228/pEMT3 was much slower (p= 0.10 :1: 0.01 h"). These rates
signiﬁcantly differed from those of JMP228/pJP4 (Pair 8, table 3.7 (n= 24,
P<0.0001). JMP228/pEMT3 grown with succinate displayed a lag time of
15 to 20 hours when transferred to 2,4-D medium. A stationary phase
population size of 3 x 106 CFU ml" could be maintained in culture vessels
sustaining a 10 % volume transfer to fresh 2,4-D minimal medium every 24
hours, a maximum density of 2 orders of magnitude less than that of
JMP228/pJP4. When a similar experiment was carried out using 10%
transfer every 72 h, JMP228/pEMT3 carrying strains obtained and
maintained a population density equal to that of JMP228/pJP4 (ﬁg. 3.5). In
the slow growth rate chemostats with dilution rates equal to 0.03 h-l, the

populations attained a size of 1x109 CFU/ml, similar to JMP228/ pJP4.

116

Table 3.7. One-tailed t-test on growth rates of strains determined
under batch culture conditions. Column 11 shows the number of culture
vessels used to calculate the P value. Boldface print indicates a signiﬁcant
difference between the growth. A P value <0. 001 is considered to represent
a signiﬁcant difference between growth rates.

 

 

 

 

succinate strain Marker plasmid n P
1. Antibiotic JMP228 rif 18 0.9852
J MP228 str
2. Plasmid JMP228 pJP4 24 0.0024
JMP228 pEMT3
3. Antibiotic DB01 rif 12 0.3455
DB01 tri
4. Plasmid DB01 pEMT3 12 0.6179
DB01 p1 P4
5. Strain JMP228 36 0.5415
DBOr
2,4.1) strain Marker plasmid n P
6. Antibiotic JMP228 rif p1 P4 12 0.4190
JMP228 str p1 P4
7. Antibiotic JMP 228 lif pEMT3 18 0.0647
JMP 228 str pEMT3
8. Plasmid JMP 228 rif or str pJP4 24 < 0.0001
JMP 228 str or rif pEMT3
9. Antibiotic DB01 rifR p1 P4 18 0.5376
DB01 triR pJ P4
10. Antibiotic DB01 rifR pEMT3 12 0.3559
DB01 triR pEMT3
ll. Plasmid DB01 rif or tri pJP4 36 < 0.0001
DB01 tri or rif pEMT3
12. Strain JMP 228 rif or str pJP4 36 < 0.0001
DBO] tri or rif pJP4
l3. Strain JMP 228 rif or str pEMT3 36 < 0.0001
DB01 tri or rif pEMT3
14. Strain] JMP 228 rif or str pJP4 24 < 0.0001
Plasmid
DB01 tri or rif pEMT3
15. Strain! JMP 228 rif or str pEMT3 36 < 0.0001
Plasmid
DB01 tri or rif pJP4 '

 

117

DB01/pJP4 had a fairly rapid growth rate (0.15 h" ) on 2,4-D that is
signiﬁcantly lower than that of JMP228/pJP4 (0.22 h" ) and higher than
that of JMP228/pEMT3 (0.10 h" ) (Table 3.7, pairs 12 and 15). Under
batch transfer conditions and 200 pg ml" 2,4-D, strain DB01/pJP4 attained
cell densities of only 1 x 106, similar to the population maintained for
JMP228/pEMT3. Under 72 hour transfer conditions, the strains grew to
5x107 CFU ml", similar to the maximum cell density of JMP228/pJP4
under these conditions (Fig. 3.5). In chemostats, DB01/pJP4 attained
densities of up to 5 x 109 CFU ml", higher than for the JMP228 strains.
DB01/pEMT3 was the slowest grower of the four host/plasmid
combinations (p= 0.08 h"). In the 24 hour batch transfer environment with
2,4-D as the limiting substrate, DB01/pEMT3 attained densities of only 1 x
10" CFU ml", two orders of magnitude less than the cell densities observed
with DB01/pJP4. Both of these B. cepacia strains reached and maintained
high densities similar to that of JMP228/pJP4 under 72 h transfer conditions

(Fig. 3.5). All four strains grew at different rates on 200 pg ml" 2,4-D.

118

 

 

 

 

 

 

A 1.E+09
1.E+08~
1.E+07~
E 1.E+06-
E 1.E+05-
U 1.E+04-
1.E+03— . . . m . .

5* 5* sr <r i- {-3 E m

95.. E: a: a 2 2 2 E

E“ 9' E LI-l LIJ m m

5 a R a: ‘3: 8* e e

g} g} S o 320 a E 1-1

E g 8 8 a a S 8

m m

E E Q o

B

1.E+091
1.E+081
g l.E+07-.
\ 1.E+0611
E 1.E+051
U 1.E+04

     

v
E:
E:
E:
o
a:
Q

 

M
[_.
2
LL]
S
O
m
D

JMP228R/pEMT3

Figure 3.3. Average population sizes for selected strains grown in batch
culture under 24 and 72 hours Transfer conditions. Panel A: Population
sizes attained in 24 hour batch transfer experiments. The antibiotic
resistance markers did not affect the maximum population sizes in 24 hour
transfer, batch culture conditions. Panel B: Population sizes with 24 and 72
hour transfer experiments. Black columns represent the 24 hour transfer
cultures, and the gray columns represent the 72 hour transfers. Colonies
were counted on R2A agar plates and 2,4-D minimal medium plates
amended with antibiotics (striped). Error bars indicate the standard
deviation among three replicates. Because DB01T/pEMT3 formed fewer

colonies on R2A plates, the population sizes in experiments were
determined on 2,4-D agar only.

119

 

lﬁ

From the utilization of 2,4—D it was calculated that JMP228/pJP4
strains produce 3.7x106 :1: 6 x 105 cells per mg of 2,4-D, JMP228/pEMT3
strains produce 3.7x106 1 6.5 x 105 cells per mg of 2,4-D, DB01/pJP4
strains produce 3.4x107 1 3.1 x 105 cells per mg of 2,4-D, and
DB01/pEMT3 strains produce 3.5x107 1 6.1 x 105 cells per mg of 2,4-D.
These were used as the efﬁciency values in the chemostat models.

The BIOLOG plate reader was also used to assess whether there were
differences in the growth rates of the strains on different concentrations of
2,4-D. While the 2,4-D was exhausted much more rapidly at lower
resource concentrations, the strains grew at the same rate at 2,4-D
concentrations between 10 and 500 ppm. Hence, a concentration of 200
ppm 2,4-D was chosen for physiological tests. This concentration allowed
the cells to attain sufﬁcient densities, without subjecting the bacteria to
toxic levels of 2,4-D.

Ks values for 2,4-D utilization were not signiﬁcantly different for
JMP228R/pJP4, JMP228R/pEMT3, DB01/pJP4 or DB01/pEMT3, and
ranged from 0.8 to 1.2 pg ml" 2,4-D. Additional information was gained
while measuring Ks. The sum of the l“C-2,4-D in the cell pellet and the
supernatant were subtracted from the initial amount of 2,4-D to give a
measurement of the quantity of 2,4-D presumably converted to CO; by each
strain during the course of the experiment. At a 2,4—D concentration of 10

ppm, all of the cultures showed similar distributions of 2,4-D.
120

Interestingly, as the 2,4-D concentration dropped to below 1 ppm, a greater
proportion of the 2,4-D was metabolized to CO; by B. cepacia strains
relative to R. eutropha strains at the ﬁnal timepoint (after 42 minutes).
There appeared to be no noticeable difference between the two plasmids,
pJP4 or pEMT3, in the R. eutropha host. B. cepacia with pJP4 mineralized
a greater proportion of 2,4-D than the same host with pEMT3 (Fig. 3.6).
Chemostat growth. Pure cultures of the different populations were
grown under chemostat conditions. The chemostats contained a total
volume of 75 m1, and a constant ﬂow rate of 0.33 h" was maintained.
Mathematical models to predict the outcome of these experiments were run
using the Stella Research Software. The graphs presented in Figure 3.7 and
3.8 shows the experimental data and the data obtained from the model.
Table 3.8 shows the values that were used in the model as described in the

Materials and Method section.

121

 

 

H"

JMP228/ pJ P4 JMP228/pEMT3

           

      
 
   

100% ‘ I 100% j. I I I I
80% -‘ g 80% ‘ g g
o .
%°f 60%* g a /°°f60%‘w
total 40V J 5 3 total 40‘V 1‘5 a 5 ”J $
24-D °. » « 24-D ° w$r$$s
’ i A ’ o 4' 5 5 $ $ V
20%, g y 2W" a :6 5 L4 5 g
0%. 1’“ 0% I I v V V
-0 0* W D -o N v-d 1""- W
D. C! ... ' ‘ O 0
1:1 0 O 0 1:5 0'
2,4-D concentration (ppm) 2,4-D concentration (ppm)
DBWPJ"4 DBollpEMT3
100% , I 100% a .
80% t -
80V ‘ ‘
total 40V . total o .x .
2,4-D ° . g 2.4-D 40% ’1 § §
20% T is 20% - a a
0%” _, - -\ 0%._ . .y
D -c N -g 0* V1 c) H or .—c [- In
"' o D Q Q "' ' ' D. D.
c1 1:) D O 1:) 1:1
2,4-D concentratlon (ppm) 2,4-D concentration (ppm)

Figure 3.6. Allocation of ring-labeled l‘C—2,4—D in growing cultures of
JMP228/pJP4, JMP228/pEMT3, DB01/pJP4 and DB01/pEMT3. The
values were obtained from the ﬁnal time point of the experiment taken after
41 minutes. 2,4-D concentration on the X axis is the initial amount of 2,4-
D supplied. Bars represent the percentage of labeled 2,4-D remaining in
each fraction. Wavy (slang) lines indicate the liquid fraction, the black
lines represent the cell fraction, and the white lines show the fraction
released as C02.

122

“‘71

Table 3.8. Parameters used to calculate expected growth curves based
on the Monod model, using Stella Research Software.

 

N0 11 KS Y So D SR
Strain (cﬁ1m|") (11") (11181-0 (80151") (cfuml‘) -(h") (cfum1")

 

JMP228R/pJP4 3 x 107 0.23 l 2.7 x 107 200 0.033 200
JMP228S/pJP4 8 x 107 0.23 1 2.7 x 107 200 0.033 200
JMP228R/pEMT3 7 x 107 0.10 1 2.7 x 107 200 0.033 200

JMP228S/pEMT3 211107 0.10 1 2.711107 200 0.033 200 1.
DBOlR/pJP4 411107 0.15 1 2.911107 200 0.033 200 7
g
DB01T/pJP4 41107 0.15 1 2.911107 200 0.033 200 5
1

DB01R/pEMT3 5,1107 0.08 1 2.911107 200 0.033 200
DB01T/pEMT3 4:1107 0.08 1 2.911107 200 0.033 200 '—

 

 

Abbreviations: N0 = initial population size, p = maximum growth rate (h "), Ks = half
saturation constant (mg/L), Y = efﬁciency (mg 2,4-D/CFU), So = initial 2,4-D (mg/L), D =
dilution rate ( h"), and SR = substrate concentration in the reservoir (mg/L)

123

 

 

   
 

 

 

 

 

 

  
 
 
 

 

 

 

 

 

 

 

 
 

 

 

 

 

 
 
 

 

 

   

 

 

1.E+10 JMP228R/pJP4 250
1.E+09 , 150
CFU/ml 1.E+08 ‘ 2 4-D
1'E+07 oW—w -<>-—-—---o— V, 50 ’
1.E+06 ‘ -50
0 3 6 days 9 12 15
1.E+10 JMP2288/pJP4 250
”“07 ‘b—o—w—O-o-o-"u-o-u-o- V50
1.E+06 - 1 -50
o 3 6 days 9 12 15 i
I
l
1.E+10 JMP228R/pEMT3 250
1.E+09 y 150 :—
CFU/ml 1.E+08 \‘ l 50 2,4-D
LE+O7 o—o—o—-—o-o--o——--o—-o-
1.E+06 , 1 -50
0 3 6 days 9 12 15
1.E+10 JMP2288/pEMT3 250
1.E+O9 1. 150
CFU/ml 1.E+08 ‘ 2,4-D
113+07 - 50
- A—o—o—o—o—w -<>-----<>-
1.E+06 , ‘ ' -50
0 3 6 days 9 12 15

Figure 3.7. Chemostat growth of JMP228R/pJP4, JMP228S/pJP4,
JMP228R/pEMT3 and JMP2288/pEMT3. The dilution rate was 0.03 h",
and the initial inoculum densities were 5 x 107 CFU m1". Grey lines
indicate modeled values, the black lines show the experimental data. Solid
lines represent culture growth and the dashed lines show the 2,4-D
concentrations in ppm.

124

 

 

   

 

 

 

 

 

 

   

 

 

 
   
 

 

 

 

 

 

    
   

 

 

 

 

 

 

 

 

 

 

1.E+10 DB01R/pJP4 250
CFU/ l 1.E+09 150
m 1.E+08 - 1A 2,4-D
1°E+07 it-b-o-o—w -<>-—-—-—-r-.-o. 1' 50
1.E+06 » ' -50
0 3 6 days 9 12 15
1.E+10 DB01T/pJP4 250
1.E+09 E : - 150
CFU/ml 1.E+08 ,‘ 2,4,1) T“
1.E+07 m+o___o__°_ 50 1‘
1.E+06 ~ * . -50 .
0 3 6 days 9 12 15
1.E+10 DBﬂlR/pEMT3 250 E.
1.E+O9 ' g 150 v
CFU/ml 1.E+08 2’44)
1.E+07 sw¢°,,,o_+ '50
1.E+06 -50
0 3 6 days 9 12 15
1.E+10 DB01T/pEMT3 250
CFU/ml 1.E+08 2,4-1)
1.E+07 5°
1.E+06 -50
0 3 6 days 9 12 15

Figure 3.8. Chemostat Growth of DB01R/pJP4, DB01T/pJP4,
DB01R/pEMT3 and DBOlT/pEMT3. The dilution rate was 0.03 h' , and
initial inoculum densities were 5 x 107 CFU ml". 2,4-D is in ppm. Grey
lines indicate modeled values, the black lines show the experimental data.
Solid lines represent culture growth and the dashed lines show the 2,4-D
concentrations in ppm.

125

JMP228S/pJP4 attained cell densities of 5 x 108 to 8 x 108 CPU ml",
and the concentration of 2,4-D in the chemostat decreased to undetectable
levels within the ﬁrst day. JMP228R/pEMT3 attained cell densities of 2 x
108 to 5 x 108 CFU ml", and the concentration of 2,4-D in the chemostat
also decreased to undetectable levels within the ﬁrst day. These results
implied that the pJP4 and pEMT3 containing R. eutropha populations did
not differ much in their growth in chemostats, as they did in batch culture.
All of the populations followed the trends predicted by the model closely.
The decline in population numbers for JMP228S/pEMT3 was due to
invasion by a fungal strain at the terminal portion of the experiment.
Actual degradation of 2,4-D by the pEMT3 carrying strains was slightly
faster than expected by the model (Fig. 3.7).

DB01T/pJP4 attained a density of 1 x 108 CFU ml", with the
concentration of 2,4-D in the chemostat at 20 pg ml" for the ﬁrst 5 days.
Then the population rose to 1 x 109 CPU ml" during days 6 to 12, and the
2,4-D concentration decreased to undetectable levels. Following a similar
growth proﬁle, DB01T/pEMT3 grew to 1 x 108 CFU ml", with the
concentration of 2,4—D in the chemostat at 20 to 50 pg ml" for the ﬁrst 5
days. Eventually, this culture also attained a ﬁnal density of 8 x 108
CFU/ml, and the concentration of 2,4-D in the chemostat decreased to about

5 ppm by day 7. 2,4-D levels gradually decreased to the limit of detection

126

—-A‘MTﬁ‘j

 

by days 8 to 12. The growth curves of the B. cepacia strains did not follow
the growth curves predicted by the model of Monod kinetics. By the time
the 2,4-D was depleted to the values predicted in the model, the population
sizes also reached the expected values. The reason for this oddity is as yet
unknown (Fig. 3.8), but may be a reﬂection of di-auxic growth where the
bacteria degrade 2,4-D to one of the intermediates, and then degrade the

intermidiate to carbon dioxide.

DISCUSSION

Ralstonia eutropha and Burkholderia cepacia are phylogenetically
related bacterial populations who are opportunists and abundant in a wide
variety of environments. Their physiology and their genetics are well
understood (Cheng et al., 1994; Chakrabarty et al., 1976; Palleroni, 1992;
Yabuuchi et al., 1992). Both are nutritionally diverse, Burkholderia
particularly so, as indicated by the wide range of substrates that can be
utilized as a resource. These bacteria are not only limited to the carbon
sources encoded within their chromosome(s), but can mutate or acquire
additional catabolic capabilities carried on plasmids or other
extrachromosomal genetic elements. The occurrence of catabolic genes on
insertion sequence (IS) elements and on catabolic plasmids give them the

ability to utilize a wide range of substrates, including aromatic compounds.

127

 

Many of the catabolic plasmids have a wide host range, and the genes they
carry are composed of cassettes and organized in operons that may be
independently assembled in different hosts, then rearranged in new hosts,
before continuing on to subsequent bacterial populations. The consequence
of this promiscuous genetic exchange is a shufﬂing of possible gene types
and plasmids, as well as the possible lateral transfer of chromosomal
segments (Fulthorpe et al., 1995; Top et al., 1995; McGowan et al., 1998;
Jacoby, 1986; Haugland et al. 1990).

When a bacterial population acquires a new set of catabolic genes,
the ability to fully mineralize an additional compound is determined by the
host population’s innate metabolic capabilities, as well as by the additional
catabolic capabilities of the plasmid carrying the degradative genes. Since
both of the investigated species are wealthy in terms of their metabolic
options, they have a higher probability of acquiring additional metabolic
pathways that will result in intermediates that can enter one of the existing
central metabolic cycles (e.g. the TCA cycle). My research took advantage
of two naturally occurring 2,4-D degradative plasmids, pJP4 and pEMT3,
which utilize different sets of degradative alleles. Pure cultures with either
of the two plasmids were examined for a range of characteristics. It was
hoped that this data, together with what was already known for both
species, would help to more accurately predict the outcome of competition

experiments among constructed host-plasmid combinations.
128

 

Plasmids pJP4 and pEMT3 both beneﬁt their host strains by
supplying a catabolic pathway which can convert the herbicide 2,4-D to
intermediates of the cell’s central metabolism. These conjugative plasmids
were both obtained from natural environments and can shuttle between a
wide range of bacterial hosts. Plasmids pJP4 and pEMT3 were conjugated
into R. eutropha JMP228. Plasmid pJP4 was also easily transferred into the
recipient strain B. cepacia DB01. All of the transconjugants carrying pJP4
exhibited a relatively fast growth rate with 2,4—D. When JMP228R/pEMT3
served as the donor strain for plasmid pEMT3, it was challenging to obtain
B. cepacia DB01/pEMT3 transconjugants. The low frequency of transfer
was surprising since two similar plasmids, pKA2 and pKA4, with nearly
identical restriction patterns and 2,4-D gene hybridization proﬁles, were
found in a bacterial strain identiﬁed as B. cepacia. (Ka et al. 1994). The
difﬁculty in introducing pEMT3 to B. cepacia DB01 was further reﬂected
in poorer (slower) growth of DB01/pEMT3 transconjugants, while utilizing
the degradative pathway ﬁ'om the plasmid. In addition, when pEMT3 was
isolated from the transconjugants, in more than in more than half of the
preparations additional restriction fragments were visible after gel
electrophoretic separation (Fig. 3.1, lanes, 14-16 and 20). The unusual
restriction proﬁles were in agreement with previous research supporting the

high levels of genomic plasticity in B. cepacia populations (Hendrickson et

129

 

al., 1996). Interestingly, this level of rearrangement was not observed
when inserting pJP4 into the Burkholderia strains.

The different plasmids confer different growth rates to a host
population when 2,4-D is supplied as the limiting substrate. These genes,
however, may also have other effects and inﬂuence grth and
reproductive success of these strains. A relevant factor affecting growth
rate, and therefore competitive outcome, is the substrate concentration.

Theoretically, a medium to which a substrate is added that allows the
population to grow at the maximum growth rate would favor a population
with the faster growth rate. This strategy is typically used to enrich for
populations exhibiting high growth rates, and has been observed in several
experiments, including those involving bacteria isolated from environmental
(Fulthorpe et al., 1994), domestic and heavily polluted sites such an alkaline
salt lake in Oregon (Maltseva and Oriel, 1997) or sewage sludge (Greer et
al., 1990). The diversity of 2,4-D degradative alleles that are recovered
from a site initially was thought to be low, however, this number increased
signiﬁcantly over the past decade. When soils were challenged with the
frequent application of 2,4-D, particular microbial populations and
degradative gene mosaics become enriched and dominant. Several different
investigators have added 2,4-D to untreated soils in proximity to those that
were treated regularly with 2,4-D. Enrichment studies using materials from

these sites yielded different dominant species despite enrichment in a similar
130

 

fashion. Experiments utilizing high concentrations of 2,4-D (>50 pg ml")
usually resulted in the enrichment of populations exhibiting high growth
rates and carrying plasmids of the pJP4 type. Typically, in enrichments
with high 2,4-D concentrations members of the genera Alcaligenes,
Ralstonia and Burkholderia will become dominant (Fulthorphe, 1995,
McGowan, 1998).

Ka and Holben found that a Sphingomonas species became dominant
in agricultural soils that were repeatedly treated with 2,4-D (Ka et al.,
1994). Asuming-Brempong (1998) showed that a B. cepacia strain became
dominant in agricultural soils with repeated 2,4-D application. In the cases
where 113-proteobacteria predominated, they tended to carry pJP4-like
plasmids and genes. In soil microcosms and ﬁeld studies, frequently
identiﬁed isolates often cluster within the B—proteobacteria, largely
belonging to the genera Burkholderia or Ralstonia (Fulthorpe et al.,
1995)

Enrichment with low concentrations of 2,4-D shows a very different
proﬁle of the 2,4-D degrading community. Interestingly, Dunbar et al.
found that the method of strain isolation correlated with the number of
different bacterial species and 2,4-D degradative alleles that were recovered
from a site. This study also presented evidence that enrichment-
independent, direct isolation methods resulted in a much greater diversity of

2,4-D degrading populations. Furthermore, these organisms belonged to a
131

""1

 

II
1

 

variety of taxonomic groups with a majority of their 2,4-D alleles being
different from those of the frequently studied plasmid pJP4 (Dunbar et al.,
1997)

It has also been recognized that soils without direct application of the
herbicide 2,4-D yield many populations capable of degrading 2,4-D. These
organisms frequently harbor genes involved in 2,4-D degradation that differ
from the commonly found pJP4-like genes (Dunbar et al., 1997; Kamagata
et al., 1997). Interestingly, pJP4-like and non-pJP4-like genes seem to
coexist in environments that have not been subject to enrichment of certain
populations due to the application of 2,4-D for agricultural purposes.
Supplying a nutrient, such as 2,4-D, skews the population toward
B—proteobacteria and genes of the pJP4 type.

Growth rate, however, is not the sole factor in determining how well
a population will perform under different growth conditions. Frequently,
environmental conditions are such that organisms spend much of their
existence deprived of resources, and must eke out a living either on very
low concentrations of a substrate, or have a means of avoiding starvation.
This study showed that strains of R. eutropha and B. cepacia performed
similarly in terms of growth rates when grown with succinate. The
presence of the catabolic plasmid did not affect the growth rate relative to
the plasmid free strain. In contrast, plasmid-harboring strains of the same

species showed a much larger difference in the growth rates when grown
132

 

 

with 2,4-D. Under high substrate concentration conditions, greater than 50
ppm, the R. eutropha strains always grew faster and B. cepacia. This high
growth rate at high substrate concentrations occurred at.a cost of lesser
efﬁciency at the lower concentrations. While this evidence is not
conclusive, the results of the 1"C distribution experiment suggest that at less
than 1 ppm 2,4-D, B. cepacia is more efﬁcient at low resource
concentrations. This observation is ﬁrrther supported by the chemostat
ﬁndings, which showed that the B. cepacia populations, carrying either
plasmid (pJP4 or pEMT3), attained population levels higher than those of

R. eutropha under resource-limited conditions

ACKNOWLEDGMENTS

This research was funded by a grant from the National Science
Foundation (BIR 912006)

133

 

REFERENCES

Ballard, R.W., NJ. Palleroni, M. Doudoroff, RY. Stanier, and M. Mandel. 1970.
Taxonoy of the aerobic Pseudomonads: Pseudomonas cepacia, P.
marginata, P. allicola and P. caryophylli. J. Gen. MicrObiol. 60:199-214.

Burkholder, W.H. . 1950. Sour skin, a bacterial rot of onion bulbs.
Phytopahtyology. 40:1 15-1 17.

Burns, J.L., LA. Hedin, and D.M. Lien. 1989. Chloramphenicol resistance in
Pseudomonas cepacia because of decreased permeability. Antimicrobial
agents and Chemotherapy 33:136-141.

Chakrabarty, AM. 1976. Plasmids in Pseudomonas. Annual Review of
Genetics 10:7-30.

Cheng, HP and T.G. Lessie. 1994. Multiple replicons constituting the genome of
Pseudomonas cepacia 17616. J. Bacteriaol. 176:4034-4042

Daubaras, D.L., C.D. Hershberger, K. Kitano, and AM. Chakrabarty. 1995.
Sequence analysis of a gene cluster involved in metabolism of 2,4-5-

trichlorophenoxyacetic acid by Burkholderia cepacia AC1100. Appl.
Envrion. Microbiol. 60:4100-4106

Davis, D. H., M. Doudoroff, R.Y. Stanier, and M. Mandel. 1969. Proposal to
reject the genus Hydrogenomonas: Taxonomidc implications. Int. J.
Bacteriol. 19:375-390.

Don, R. H. and J.M. Pemberton. 1985. Genetic and physical map of the 2,4-
dichlorophenoxyacetic acid-degradative plasmid pJP4. J. Bacteriol.
161:466-468,

Dunbar, J., S. White, and L. J. Fomey. 1997. Genetic Diversity through the
looking glass: effect of enrichment bias. Appl. Environ.
Microbiol.63: 1326-133 1 .

Dykhuizen, D. E. and D. L. Hartl, 1983. Selection in chemostats Microbiol. Rev.
47:150-168.

Fulthorpe, R.R., C. Mcgowan, O.V. Maltseva, WE. Holben, and J.M. Tiedje.
1995 2,4-Dichlorophenoxyacetic acid-degrading bacteria contain mosaics of
catabolic genes. Appl. Environ. Microbiol. 61:3274-3281

134

44—. _.____—

Greer C.W., J. Hawari and R. Samson. 1990, Inﬂuence of environmental factors
on 2,4-dichlorophenoxyacetic acid degradation by Pseudomonas cepacia
isolated from peat. Arch. Microbiol. 154:317-322.

Haugland, R.A.. U.M.X. Sangodkar, and AM. Chakrabarty. 1990. Repeated
sequences including RS1100 from Pseudomonas cepacia AC1100 function
as IS elements. Mol. Gen. Genet. 220:222-228

Hendrickson, W, A. Hilbner, and A. Kavanaugh-Black. 1996. Chromosome
multiplicity in Burkholderia cepacia . pp. 259-269. In: Nakazawa, T, K.
Furukawa, D. Haas, and S. Silver (eds). Molecular Biology of
Pseudomonads. Washington, DC. Proceedings of the ﬁfth international
symposium on Pseudomonads. American Society of Microbiology

Isles, A., I. Maclusky, M. Corey, R. Gold, C. Proper, P. Fleming and H. Levison.
1984. Pseudomonas cepacia infection in cystic ﬁbrosis: and emerging
problem. Journal of Pediatrics. 104: 206-210

Jacoby, GA. 1986. Resistance plasmids of Pseudomonas. P 265-293. In:
Sokatch .J.R. (ed.), The bacteria, volume 10, The biology of Pseudomonas.
Orlando: Academic Press.

Ka, J.O. ,W.E. Holben, and J.M. Tiedje. 1994. Analysis of Competition in Soil
among 2,4-dichlorophenoxyacetic acid-degrading bacteria. Appl. Environ.
Microbiol. 60:1121-1128

Kamagata, Y., R.R. Fulthorpe, K. Tamura, H. Takami, L. J. Fomey and J.M.
Tiedje. 1997. Pristine environments harbor a new group of oligotrophic
2,4-dichlorophenoxyacetic acid-degrading bacteria. Appl. Enviom.
Microbiol. 63: 2266-2272.

Keyser, P, B.F., Pujar, R. W. Eaton, and D.W. Ribbons. 1976. Biodegradation of
the phthalates and their esters by bacteria. Environ. Health Perspect.
18:159-166

Lewis, D.M., E.S.P. Bloomﬁeld, and L. R. Barran. 1987) Effect of rifampicin
resistance on nodulating competitiveness of Rhizobium meliloti. Can. J.
Microbiol. 33:343-345.

Lbfﬂer, F.E., JE. Champine, KM. Ritalahti, S.J. Sprague adn J .M. Tiedje. 1997.

Complete reductive dechlorination of 1,2-dichloropropane by anaerobic
bacteria. Appl. Environ. Microbiol. 63:2870-2875.

135

Maltseva, 0. And P. Oriel. 1997. Monitoring of an alkaline 2,4,6-tlichlorophenol-
degrading enrichment culture by DNA ﬁngerprinting methods and isolation

of the responsible organism, haloalkaliphilic Nocardioides sp. Strain M6.
Appl. Environ. Microbiol. 63 :4145-4 149.

McGowan, C., R. Fulthorpe, A. Wright, and J.M. Tiedje. 1998. Evidence for
interspecies gene transfer in the evolution of 2,4-dichlorophenoxyacetic
acid . Appl. Environ. Microbiol. 64:4089-4092.

Monod, J. 1949. The grth of bacterial cultures. Annu. Rev. Microbiol. 3:371-
394.

Palleroni, NJ. 1992a. Introduction to the Pseudomonads. pp 3071-3085. In: J.
Balows, A, H.G., Triiper, M. Dworkin, W. Harder, K-H Schleifer (eds.),
The Prokaryotes: A Handbook on the Biology of Bacteria: Ecophysiology,
Isolation, Identiﬁcation, Applications. 2nd ed. New York. Springer-Verlag,

Palleroni, N.J. 1992b. Present situation in the taxonomy of aerobic
pseudomonads. pp 105-115. In: Galli, E., S. Silver, and B. Withold (eds.),
Psudomonads: Molecular Bioloty and Biotechnology. Washington DC.
American Society for Microbiology,

Palleroni, NJ. and B. Holmes. 1981. Pseudomonas cepacia sp. nov.,nom. rev.
International Journal of Systematic Bacteriology. 31:479-481.

Top E.M., WE. Holben and L.J. Fomey (1995) Characterization of diverse 2,4-
Dichlorophenoxyacetic acid-degrading plasmids isolated ﬁ'om soil by
complementation. Appl. Environ. Microbiol. 61 :3274-3281

Stanier R.Y., N.J. Palleroni, and M. Doudoroff, 1966. The aerobic
pseudomonads: a taxonomic study J. Gen. Microbiol. 43: 159-271.

Tabacchioni, S., P. Visca, L. Chiarini, jA. Bevivino, C. Di Serio, S. Fancelli, and R.
Paul. 1995. Molecular characterization of rhizosphere and clinical isolates
of Burkholderia cepacia. Res. Microbial. 146:531-542

Wyndham R.C. 1986. Evolved aniline catabolism in Acinetobacter calcoaceticus
during continuous culture of river water. Appl. Environ Microbiol.
51 :781-789

Yabuuchi E., Y. Kosako, H. Oyaizy, I. Yano , H. Hotta, and Y. Nishiuchi. 1992.
Transfer of two Burkholderia and an Alcaligenes species to Ralstonia
gen. nov. : Proposal for Ralstonia picketii . (Ralston, Palleroni, Doudoroff
1973) comb. nov., Ralstonia solanacerum (Smith 1896) comb. nov. and
Ralstonia eutropha (Davis 1969). comb. nov. Microbiol. Immunol. 39:897-
904.

136

I‘LD‘LAL ’a .

 

CHAPTER 4

Two HOSTS AND Two PLASMIDS IN SIMPLE ENVIRONMENTS: THE INFLUENCE
OF 2,4-D DEGRADATIVE PLASMIDS ON THE COMPETITIVE FITNESS OF A HOST
STRAIN

ABSTRACT

The contribution of different plasmid-encoded degradative alleles to
the competitive ﬁtness of two host populations was investigated. Natural
variation in the catabolic gene cassettes for the degradation of aromatic
compounds has resulted in the evolution of different, but functionally
similar degradation pathways for the herbicide 2,4-dichlorophenoxyacetic
acid (2,4-D). Frequently, these degradation pathways are carried on broad
host range plasmids that are horizontally transferred among a variety Of
bacterial populations. In this study we utilized two such plasmids, pJP4 and
pEMT3, that confer the ability to utilize 2,4-D upon two bacterial host
species, Ralstonia eutropha JMP228 and Burkholderia cepacia DB01.
Each of the plasmids has a similar replication origin and belongs tO the Inc-
P incompatibility group. The plasmids have different alleles for the
catabolism of 2,4-D. Pure cultures were characterized by maximum growth

rates and KS(2,4-D)’ as described in the previous chapter. The values for these

137

 

two parameters for each host/plasmid combination were used to predict the
outcome of competition experiments performed under batch and chemostat
conditions with 2,4-D as the limiting resource. Pairwise combinations of
the two different bacterial species , R. eutropha JMP228 and B. cepacia
DB01, carrying the two different plasmids, pJP4 or pEMT3, were carried
out in the two culture conditions. In Chapter 3, it was determined that the
fastest growth and most rapid 2,4-D degradation were observed for the
JMP228/pJP4 combination, followed by DBOl/pJP4, and JMP228/pEMT3.
Population DB01/pEMT3 grew the slowest. There was no measured
difference in the Ks values as determined for any of the strain/plasmid
combinations. Hence, in batch culture competitions where resources were
replenished periodically at a high concentration, a faster growth rate should
lead to dominance, and thus the following order of dominance was
predicted: JMP228/pJP4 > DB01/pJP4 > JMP228/pEMT3 > DB01/pEMT3.
As expected, possessing plasmid pJP4 was advantageous, as it offered a
faster growth rate than pEMT3, and the expected order of dominance was
observed. In chemostat experiments conducted at a very low dilution rate
of 0.03 h", it was also expected that the faster growth rate should
predominate if the populations have the same Ks. The host/plasmid

combinations should reﬂect the same rank order as the batch competitions.

138

L mugs-n”

Surprisingly, the following order of dominance was observed: DB01/pJP4 Z
DBOl/pEMT3 > JMP228/pJP4 = JMP228/pEMT3. This result was
unexpected. When the substrate was supplied at constant but low
concentrations, Burkholderia, rather than Ralstonia predominated,
regardless of the plasmid. Possible causes for the different dominance
patterns under the two growth conditions are discussed, as are possible
ecological implications for maintaining different catabolic plasmids and

host species in the environmental gene pool.

INTRODUCTION

Among bacterial populations, most genes are transferred vertically,
from “mother” to “daughter”. The chromosomal genes encode for all of the
basic requirements of the cell, and determine much of the population's
physiological diversity. Additional characteristics to expand a population’s
substrate range or genetic diversity may reside on extrachromosomal
elements, such as conjugative plasmids, that are transferred horizontally and
vertically. Naturally occurring broad host range plasmids have been
identiﬁed, and can carry genes for metabolic functions such as the

degradation of aromatic compounds and resistance to heavy metals or

139

 

antibiotics. Plasmids that carry pathways for 2,4-D, 3-chlorobenzoate,
napthalene and toluene utilization have been isolated, and their horizontal
transfer by conjugation has been demonstrated (Di Giovanni et al., 1996;
Ravatin et al., 1998; Herrick et al., 1997; Wright et al., 1994). For a
degradation pathway involving any of these compounds, there exists some
amount of inherent variation in the gene sequences or pathways (Ghosal and
You, 1988; Koh et al., 1997).

For example, in order to mineralize phenol to carbon dioxide and to
use it as a source of carbon and energy, a bacterial population may utilize a
catabolic pathway that hydroxylates phenol to catechol followed by an ring-
cleavage event to form muconic acid (Rochind-Dubinsky et al., 1987;
Haggblom et al., 1992). There may be several alleles for each of the genes
involved in phenol degradation. Different enzymes may have slightly
different substrate speciﬁcities and different kinetic constants (Vmax and
Km). Thus, one set of enzymes may prefer chlorinated substrates, while
another prefers the non-substituted analogues, but both have the capacity to
degrade either, albeit with different kinetics. When required to grow by
utilizing one substrate as a sole carbon source, a population will do so by

means of whatever suitable pathway it may possess.

140

 

 

 

The products of those different degradation pathways enter into the
organism’s intermediary metabolic pathways, where they are used for
generating energy and biomass. The ability to utilize additional carbon
sources has obvious advantages. Hence, plasmid-encoded and
chromosomally-encoded genes contribute to the reproductive success of the
population capable of utilizing the new resource. Figure 4.1 illustrates the
integration of plasmid-encoded and chromosomally-encoded genes when
2,4-D is utilized as a carbon source. The portion of the pathway depicted in
gray is plasmid encoded, and determines the rate of metabolizing the
substrate (2,4-D) into intermediates, in this case chloromaleylacetate
(CMA), that can be channeled into the organism’s intermediary
metabolism. The host bacterium’s genome encodes for the transformation
of CMA to succinate, a common intermediate of central metabolism.
Consequently, what we measure is reproductive success: which population
leaves more offspring, and why? Is greater success due to the genes of the
chromosome, or those of the plasmid?

For pure cultures we can use growth rate and the half-saturation
constant (Ks) as indicators of reproductive success. In competition studies,
the more successful competitor is the one that achieves numerical

dominance.

141

It.“

 

we

Reproductive
Success

A

Plasmid Chromosome

 

 

Energy and
Meta? lism

      
   
 

 
 
  
  
 
   
  
  

 

Chromosomally encoded
Plasmid enzymes for utilization of
encoded succinate

enzymes
for degrad-

ing 2 4'1) ‘ Chromosomally
‘0 CMA encoded enzymes
for CMA to

(\I

Substrate
Substrate (2,4-D) .
aware]

Figure 4.1. An overview of the genetic contribution of different genes
in the competition experiments. Black regions are encoded by the
bacterial chromosome, the gray portions are controlled by the alleles for the
2,4-D degradative pathway that are encoded on the plasmid.

 
  
 
   
  

 

142

Bacteria in the environment may compete for a limiting resource and
the population that can most rapidly utilize the resource tends to become
enriched. The enrichment effect, in which the fastest growing phenotype
becomes dominant, is commonly observed during the enrichment and
isolation of bacteria with the ability to degrade recalcitrant substrates
(Dunbar et al., 1996; Dunbar et al., 1997; Fulthorpe et al., 1995). The
observations by Dunbar et al. (1997) correlated 2,4-D application with
enrichment for the pJP4—like genotype among several different bacterial
populations. As a result of enrichment and isolation techniques, it was
initially believed that the bacterial utilization of 2,4-D proceeded through a
very conserved and related set of catabolic genes. The genes encoded by
plasmid pJP4 were chosen to represent the model system to which newly
isolated degradative genes were compared, since they represented the most
frequently described set of 2,4-D utilization genes. Over time, it was
discovered that when alternative isolation methods were utilized, the
environment held a greater diversity of catabolic pathways that degraded this
herbicide (Ka. and Tiedje, 1994; Top et al., 1995). The emerging trend was
that of a high natural diversity in catabolic plasmids, 2,4-D degradation

genes, and suitable populations capable of metabolizing 2,4-D.

143

 

 

Plasmids pJP4 and pEMT3 were chosen for this investigation, because
both encode a pathway for 2,4-D degradation, however, they carry different
alleles. Residing on pJP4 are genes which encode for the most extensively
studied 2,4-D degradative pathway; they encode for six enzymes that
sequentially break down 2,4-D to chloromaleylacetate (Don and Pemberton,
1981; Don and Pemberton, 1985; Don et al., 1985). The canonical pathway
from plasmid pJP4 is described in Chapter 1. Although the ﬁrst enzyme of
the pathway (tfdA) is homologous for both plasmids, pEMT3 carries a
different set of 2,4-D gene alleles for the tdeDFE operon and for the tde
gene. Host strains carrying pEMT3 show slower growth rates when 2,4-D
is supplied as a sole source of carbon and energy. However, the genes of
pEMT3 encode for enzymes with a higher activity toward non-chlorinated
substrates than the enzymes encoded on pJP4 (Maltseva 1994).

Ralstonia eutropha JMP228 and Burkholderia cepacia DB01 served
as hosts for the two plasmids, pJP4 or pEMT3, described above (Don and
Pemberton, 1981; Top et al., 1995; Ka et al., 1994). Both bacterial species
are members of the B-proteobacteria, and have been shown to degrade 2,4-D
when they acquire the upper 2,4-D degradation pathway on a catabolic
plasmid (Ka et al., 1994; Top et al., 1995). This study evaluated the

contribution of the two different plasmids and their allelic genes encoding

144

 

for the enzymes involved in 2,4-D catabolism, to the ﬁtness of their host
strains in two simple habitats (batch culture versus chemostat). The
physiological characteristics of the different strain/plasmid combinations are
described in Chapter 3. In high nutrient environments, the presence of
plasmid pJP4 resulted in a higher growth rate in both host strains and
outcompeted the same strain carrying pEMT3. When both hosts had the
same plasmid, Ralstonia eutropha had a higher growth rate on 2,4-D than
Burkholderia cepacia. There was no measured difference in the Ks for any
of the four host/plasmid combinations.

Batch conditions represent a high nutrient environment selecting for
fast growth-rate, whereas chemostats simulate a constant, low nutrient
environment that selects for a lower Ks. Based on models of substrate
limited growth in the batch cultures, if abundant 2,4-D is supplied,
populations should grow at their maximal rate until 2,4-D is depleted. If a
portion of this culture is transferred to fresh medium, the population that had
grown to a larger size should again produce more offspring in a successive
spurt of growth, leading to an ever widening chasm in population sizes. A
stable coexistence could be maintained if the population that grew more
slowly at the high substrate concentration could grow more rapidly after the

substrate was nearly exhausted. While it may not become dominant, it

145

‘l

would not become extinct. From these theoretical considerations it was
predicted that the strain with the faster growth rate would dominate. In
competition experiments with the same bacterial species with either of the
two plasmids, pJP4 and pEMT3, it was expected that the strain with pJP4
would dominate, since this plasmid bestows a higher growth rate on its host.
Under continuous ﬂow chemostat conditions, a high maximum
growth rate is desirable, but in addition, the afﬁnity of the organisms to the
limiting resource becomes relevant. Populations with a lower half saturation
constant (Ks) should outcompete populations with lower afﬁnities to the
limiting resource. Using the substrate limited growth kinetics described by
Monod (1949), the following predictions can be made. First we assert that
the natural logarithm of the ratio of the competing populations (Ln (N A/NB))
plotted against time results in a line with the slope representing the rate of
difference at which the populations grow. NA and NB represent the
population sizes of population A and B respectively. A positive slope
indicates that population A became numerically dominant, and a negative
slope indicates dominance by population B. Next consideration is given to
the inﬂuence of substrate concentration on the Monod equation (Chapter 3).
If the substrate is present at a concentration much higher than the Ks of

either competitor then the population with the higher maximum growth rate

146

 

should predominate (slope = pm, A - umax B). If the substrate concentration in
the chemostat equals the K3 for the strains, then the slope of the line created
by (Ln (N A/NB)) = 1/2 (pm, A - pmax B), and again the population with the
faster growth rate should win. In the third case, when the substrate is
present at concentrations well below the Ks of the strains, then it is the ratio
of the Irma/Ks for each strain that determines ﬁtness, so the slope becomes
(pm, A / Ks A )- (pmax B/ KS 3 ), and therefore the strain with the lower value
for Ks should have the long term advantage. In this system, it was expected
that the pJP4 bearing populations predominate over the pEMT3 bearing
strains, since the measured KS for the four populations did not differ. To
predict the outcome in competition experiments, hypothetical curves for
each competitor using the Monod model were generated, and average
growth rate and Ks values for each pure culture host/plasmid combination
were determined. By comparing modeled growth curves to our observed
outcomes of competition, we describe the contribution of the two 2,4—D
degradative plasmids with different alleles to the ﬁtness of the two host

populations under two contrasting environmental conditions.

147

 

MA T ERIALS AND METHODS

Bacterial strains. Ralstonia eutropha JMP228 (JMP228), the pJP4
cured strain of R. eutropha JMP134 (Don and Pemberton, 1981) and
Burkholderia cepacia DB01 (Ka et al., 1994) were used as the recipient
host strains for the two 2,4-D degradative plasmids. Descriptions of the
plasmids and the construction of the transconjugants used in the
competition experiments are described in Chapter 3. Strain characteristics

are summarized in table 4.1

Table 4.1. Pure culture characteristics for strains grown on mineral
salts medium with succinate or 2,4—D under batch culture conditions.

 

 

strain Succinate, 2,4-D Ks (2,4-D) Yi e1 db
u 1 soa “ 1 SD (mg L") (CFU a")
IMP228R/pJP4 0.538 11.082 0.223 :I: 1.056 1.2 3.7x106
JMP228S/pJP4 0.456 14.031 0.238 :1: L050 nd nd
JMP228R/pEMT3 0.371 11.020 0.094 :1: L014 0.8 3.7x106
JMP2288/pEMT3 0.509 11.054 0.106 :I: I.014 nd nd
DBOlR/pJP4 0.489 :I:|.l60 0.159 :I: 1.038 1.0 3.4x10"
DB01T/pJP4 0.534 31.228 0.152 1; 1.039 nd nd
DB01R/pEMT3 nd 0.085 11.027 1.2 3.5x10‘5
DBOlT/pEMT3 nd 0.075 11.028 nd ' nd

 

“ so = Standard deviation

b Yield is the number of colony forming units per gram of 2,4—D.

148

 

Cell cultivation. Cells were grown in A+N mineral salts medium to
which carbon sources and antibiotics were added (Wyndham et al., 1986).
The preparation of A+N medium was described previously for the pure
culture competitions in Chapter 3. For each independent experiment,
cultures were started from a loopful of frozen culture stock. The cell
cultures were grown to late log phase in l6-ml test tubes with 3 ml A+N
minimal amended with 2,4-D (200 ppm) and the appropriate antibiotic
(rifampicin, streptomycin or trimethoprim). A 2.5% (vol/vol) inoculum to
ﬂesh A+N medium with succinate (5 mM) or 2,4-D (200 ppm) was used to
start any other culture used in the competition experiments. Population
identiﬁcation in the competition experiments was carried out using 2,4—D
minimal medium plates supplemented with antibiotics. Additionally,
morphological differences in colony types on R2A agar could distinguish
between Ralstonia and Burkholderia species.

2,4-D concentration. 2,4-D concentration was determined using High

Performance Liquid Chromatography (HPLC), as described in Chapter 3.

149

 

Batch culture competition experiments. Cultures were initially grown
to late log phase in minimal medium with succinate (5 mM) or 2,4-D (200
ppm), depending on whether the competition would take place in medium
amended with succinate or 2,4-D. Competitive pairs consisted of pair-wise
combinations of strains with different antibiotic susceptibilities carrying
either plasmid pJP4 or pEMT3. Competitions with succinate were used to
determine if plasmids pJP4 and pEMT3 affected the competitive ability of
JMP228 when the plasmid was not required for growth. Succinate was also
used to determine which host, JMP228 or DB01, would become dominant in
the batch environment. Competition experiments using 2,4-D as the limiting
resource were much more exhaustive. Reciprocal mating pairs between each
of the antibiotic resistant strains were used to account for the effect of the
chromosomal marker, rifampicin versus streptomycin and rifampicin versus
trimethoprim for JMP228 and DB01 respectively. Cells were harvested by
centrifugation, rinsed with sterile medium, and suspended to an optical
density of 0.1, in minimal medium without 2,4-D. To determine if ten-fold
differences in the initial ratio of competitors affected the outcome, 1:10, 1:1
and 10:1 ratios of the populations were combined. The CFU/ml was

determined on selective medium prior to inoculating (10 % [vol/vol])

150

 

triplicate cultures with the mixture in 2 ml minimal medium with 300 ppm
2,4-D. Alternating wells of a 24-well microtiter plate were used to perform
many replicates while avoiding cross-contamination. The cultures were
incubated at 30°C in a rotary shaker (200 rpm). Every 24 hours, 10% of the
culture volume was transferred to ﬂesh medium, as well as diluted in PBS,
and plated to selective medium for determining viable bacterial population
sizes. Additional inter-species competition experiments were carried out
with equal numbers of starting populations of the competing strains. Instead
of transferring the culture every 24 hours, 10% (vol/vol) of the culture was
transferred every 72 hours. With a longer time allotted for growth, slower
growing strains were able to complete their entire grth cycle and use all

the available 2,4-D. Competitive pairs and treatments are listed in table 4.2.

151

 

 

Table 4.2. Competition experiments .

“Fa... '~—- —

 

Batch competition '

 

 

 

 

 

 

Competition Strain 1 Strain 2 Succinate 2,4-D 2,4-D Chemostat
24 h 24 h 72 h

Same strain, JMP228R/ JMP228S/ nd ° 3a nd nd

Same PIP 4 pJP4

plasmid I":
JMP228R/ JMP2288/ nd 3 nd nd _l
pEMT3 pEMT3 ;
DBOlR/ DB01T/ nd 3 nd 1 ‘
PJP4 pJP4
DBOIR/ DBOlT/ nd 3 nd nd
pEMT3 pEMT3 5"

Same strain, JMP228/ JMP228/ 8 6" nd 3

Different. pJP4 pEMT3

plasmid b
DBOl/ DBOl/ nd 9 nd 2
pJP4 pEMT3

Different. JMP228/ DB01/ 8 6" 3 2

strain, pJP4 p JP4

Siamed JMP228/ DB01/ 8 9b 3 2

P 38“ pEMT3 pEMT3

Different JMP228/ DBOl/ nd 9b 3 2

strain, pJP4 pEMT3

Different. b

plasmid JMP228/ DB01/ nd 6 3 2
pEMT3 pJP4

’ number of replicates.

b the two antibiotic resistant populations bearing the same plasmids are being
considered as the same population, and the marker designation has been omitted.

c not done

152

 

 

The dominant strain can be detected using the natural log (Ln) of the
ratio of population sizes (Malthusian parameters) plotted over time. The
resulting curve represents the rate of difference at which the populations
grow. The slope of the line away from zero indicates dominance by one of

the populations (Lenski et al., 1992).

Chemostat Competition Experiments. Chemostats based upon the
design of Chao et al. ( 1977) were used, and were described in Chapter 2.
Cultures were initially grown as described for the batch competition
experiments in 200 ml of A+N minimal salts medium with 200 pg ml'l 2,4-
D as the sole carbon source. When the cultures attained late-log phase,
which correlated to an optical density of about 0.1 at 600 nm, the cells were
gently collected by centrifugation (4,000 rpm for 20 min in a Sorvall SS34
rotor). Aﬁer rinsing with A+N minimal medium, cultures were suspended to
an OD of 0.1 in A+N medium containing 200 pg ml’l 2,4-D. Initially, 37.5
ml of each competitor population was combined in the chemostat, to give
approximately the same total number of cells per vessel as in pure culture
controls (Chapter 3). 2,4-D was added to an initial concentration of 200 pg
ml". Each chemostat was sampled and incubated for 8 hours at 30° C before
a constant ﬂow of medium was begun. The dilution rate was set to 1 volume
exchange (75 ml) per day (3 ml h") with a reservoir 2,4-D concentration of

153

 

 

200 pg ml". Chemostats were operated for 10 to 12 days. Under these ﬂow
conditions, the populations should all grow at doubling times of 17.5 hours.
To sample the chemostats, the ports were rinsed with 70% ethanol. Samples
(1 ml) were removed with a plastic syringe, and the 2,4-D concentrations,
OD 600 and the CFU/ml of each competing strain were determined.
Additional l-ml samples were withdrawn from each chemostat to determine
the CFU/ml over time. A+N agar with 200 pg ml'l 2,4-D and the
appropriate selective antibiotics was used to count the population sizes of the
competing bacterial strains. In addition, the CFU/ml were also determined
on R2A agar amended with the appropriate antibiotics. R2A agar without
antibiotics was used to monitor for possible contaminants. As with the batch
competition experiments, the Malthusian parameter of the competing strains
was plotted over time to determine the relative ﬁtness as described by Lenski

et al., (1992). Competitive pairs and treatments are listed in Table 4.2.

Modeling the Monod equation. In order to evaluate the outcome of
the chemostat competitions, the expected population sizes and 2,4-D
concentrations for each combination of competitors were estimated using the
Monod equation. Models were run using the Stella Research Software v 5.0
(High Performance Systems Inc. Hanover, NH). Values determined for each
host/plasmid combination were inserted into the model. The resulting curves

154

 

 

‘1:

 

give the predicted population size for each competitor, given that the grth
rate, growth yield, and Ks are known for each strain. The Monod equation
for each strain is shown in equation (a). The Monod eqriation describing
competition in chemostats was previously described by Dykhuizen and Hartl
(1983). dN/dt is the rate of change in population size, N is the population
size, pmax is the maximum growth rate on the resource, and S is the substrate
concentration. Ks is the half—saturation constant, a value that describes the
concentration of resource at which the grth rate is one half the maximum.
D is the dilution rate of a chemostat. In a chemostat, where continuous
growth is the goal, the dilution rate (D) is the rate at which culture volume
and cells are removed with the equivalent input of fresh media. The dilution
rate D determines the grth rate of the bacterial population in a chemostat,

and equals the growth rate.

Equation (a)

dN_N -
‘31—” ”mﬂxssis DN

Considering there are two competitor populations, we can set up the
equation for each population A and B, and call their population sizes NA and
NB. The values for the pmax, Ks and efﬁciency are run through a model.

Equations (b) describe the growth of population A and population B.

155

 

 

Equations (b)

dNA_ N S

and

T = B I‘max(B) K—LSS(B)+ - DNB

For each competitor there is a value for S, the substrate concentration,
at which the cell obtains just enough energy to avoid being washed out of the
chemostat. The winner of competition, then, is the strain with a lower break—
even concentration of substrate. Populations that obtain enough energy to
replace their cell numbers are not washed out, and cells that starve (do not
make sufﬁcient replacement numbers) are washed out. The break-even
concentration (8*) is the same value as the resource threshold value (R*)
described in Tilman’s resource ratio theory discussed in Chapter 1 (Tilman
1981). If one of the populations is able to reduce the level of resource in the
chemostat to below the break-even concentration of its competitor, the
competitor will be diluted out. For each strain the break-even concentration

can be calculated using equation (c).

156

 

‘1

 

Equation (0)

DKS(A)
“max (A) - D

3*:

 

The lower 8* can be achieved in two main ways. If both strains have
a very similar Ks, then the population with the faster growth rate will have
the lower break-even concentration for the substrate, and the strain with the
lower growth rate will be diluted out. On the other hand, if both populations
have a very similar pmax, then the population with the lower Ks value should
attain the lower value for S*. In this way, a population with a lower grth
rate could become dominant under low substrate concentrations, and it
would become extinct in a high growth rate selective environment.

Similar to the batch culture competitions, the natural log of the ratio of
the population sizes was determined during the course of the experiment.
The natural logarithms of the ratio of the two populations (Malthusian
parameters) of the competing strains were plotted over time to determine
their relative ﬁtness (Lenski 1992). A signiﬁcant deviation in the slope of
this line away ﬁom zero indicated that one strain had become competitively
dominant to another. Using the equations in (b) the slope of the line can be

described by the equations in (d) and (e). Equation ((1) is the slope of the

157

 

 

line as represented by rearranging the equations in (b) as the rate of growth
of the population sizes.

Equation (d)

dNA dNB
1‘7th ' Ne—dt
The same equation can be represented in Equation (e) using the other
half of equation (b). By substituting values for Ks and pmax, for the two
different populations in equation (d), the slope of the line can be described

under three widely different substrate (S) concentrations.

Equation (e)

ﬂmaxm) S - #maxm) 'K—ST
KS(A)+S S(B)+

By making the value of S smaller than the K3, the half saturation
constant becomes the main determinant of the slope of the line. In the ﬁrst
case stated below in equations (f), the population with the lowest value of Ks
will be the one that determines the direction of the slope in its favor. For the
second case, we ﬁnd that if S is sufﬁciently larger than Ks, then the half-

saturation constant can be ignored, and the population with the higher

158

 

 

growth rate controls the direction of the slope of the line. Even when the
value of S is nearly equal to the values of Ks for the two populations, the

growth rate still determines the slope of the line, although it will be less

steep.
Equations (f)
If S<< Ks(A) and Ks(B) slope: (M , it‘ll—tug) ) S
KS(A) KS(B)
If S>> Ks(A) and Ks(B) slope = ”maxm) - “maX(B)
If S~ Ks(A) and KS(B) slope = 1/2(llmax(A) ' ”maX(B))

Toxicity of DB01 chemostat ﬁltrate. To determine if the reason that
DB01 became dominant in chemostats, the possibility of the production of
an inhibitory compound toward JMP228 was brieﬂy explored. At the
conclusion of one chemostat experiment, the vessels containing pure cultures
of DB01/pJP4 and DB01/pEMT3 were each ﬁltered through a 0.22 pm ﬁlter
to obtain two cell-free spent media stocks. A dilution series of the spent
media and phosphate buffered saline were prepared, so that ﬁnal

concentrations of the spent media were 100 %, 75 %, 50 %, 25 % and 0% by
159

 

lf~_‘

 

volume. 2,4-D (100 ppm) was added to each dilution. Three ml of each
dilution was put into three l6-ml test tubes per susceptibility test
(JMP228/pJP4 and JMP228/pEMT3). 200 pl of chemostat grown cultures
of the two strains of R. eutropha were inoculated into each tube, and the
tubes were incubated in a rotary shaker (30 °C, 200 rpm, 24hours). The ﬁnal
optical density of each tube was recorded spectrophotometrically, and a
decrease in turbidity with an increase in the concentration of ﬁltrate would
warrant further investigation into the presence of growth inhibiting
compounds produced by DBOl. No difference in the turbidity would

indicate that there was little possibility of inhibitor production by DB01.

RESULTS

Succinate batch competitions. Competitions in minimal medium
supplemented with succinate, indicated that for host strain JMP228, there
was a greater difference in the effect of the antibiotic resistance marker,
rifampicin or streptomycin, than the catabolic plasmid (pJP4 or pEMT3) in
24 hour batch transfer environments. In either reciprocal combination of
JMP228R/pJP4 vs JMP2288/pEMT3 (four replicates), or JMP2288/pJP4 vs
JMP228R/pEMT3 (four replicates), the two host populations coexisted at a

summed population size of about 1 x 109 CFU ml". The total population

160

 

ll—
5

 

size was equivalent to the pure culture’s population sizes when grown alone
under these conditions. Figure 4.2 shows the average population sizes of
the two reciprocal sets of competitors

The slope of the regression line (-0.0042) drawn through the points
derived from the Malthusian parameter, did not deviate signiﬁcantly from
zero, as indicated by a very poor r2 value when all data were considered
together (r2 = 0.0155) (Fig. 4.3). On the other hand, there was always a
slight, but signiﬁcant difference in slope of the lines in favor of the
rifampicin resistant strain over the streptomycin resistant one. For each of
the four replicate reciprocal competitions, all of the points were plotted on
the same graph, to generate a slope for all the points. The best ﬁt line
yielded slopes of 0.33, and 0.50 with r2 values of 0.7896 and 0.9252,
respectively, indicating a slight advantage of the rifampicin resistant

population.

161

 

 

 

 

 

 

 

1.E+10
1°E+09€ _’ Q-~— —- ~-
CFU/ml 1.E+08 A M
E
1.E+07
—I—JMP228$(pJ P4) — -n- -JMP228r( EMT3)
___-___JMP228r(pJP4) — —Cl- -JMP2288 pEMT3)
1.E+06
0 50 100 150 200
hours

Figure 4.2. Growth of JMP228/pEMT3 and JMP228/pJP4 competing
for 5 mM succinate as a limiting resource. Grey and black lines indicate
two separate experiments and each symbol represents the average of four
replicates with error bars indicating the standard deviation around the mean.

 

 

 

 

 

 

 

10
5“ a
ln(pJP4/ JR 0 9 o o 5
PEMT3) 0 8 ﬁ‘ ‘ 6 U 9‘ : ~~~~-—-~i
-5
-10
0 50 100 150 200

hours

Figure 4.3. Plasmids pEMT3 and pJP4 do not govern fitness of R.
eutropha competing for succinate. With succinate-grown cells the slope of
the line generated by the 1n of the ratio of JMP228/pJP4 / JMP228/pEMT3
is not different from zero, as indicated by the low rzvalue (0.0155). The
mirror image appearance is due to the replicate competitions performed for
each antibiotic resistance, and the sloping away from zero is due to a slight
advantage for the rifampicin resistant population.

162

 

EV

 

In contrast, the four replicate competition experiments between
JMP228R/pJP4 vs DB01R/pJP4 and JMP228R/pEMT3 vs DB01R/pEMT3,
showed that the R. eutropha populations always dominated over the B.
cepacia competitors when both rival strains had the same plasmid (Fig.
4.4). The slope of the regression line of Malthusian parameters was
positive in favor of JMP228 (0.055), and the r2 value was high (>0.9761).

The DBOlR/pJP4 population was decreased to values lower than
those of the pure cultures, while the JMP228R/pJP4 and JMP228R/pEMT3

populations attained sizes equal to those of their non-competing controls

(Figure 4.5).

163

m1

 

 

+JMP2285(pJP4) +DB01r(pJP4)
—o—-JMP228s(pEMT3) —I:l—DB01r(pJP4)~

1.E+10

1.E+09

1.E+08
CFU/ml 1.E+07
1.E+06
1.E+OSe
1.E+04»

0 20 40 60 80 100 120 140 160

hours

 

 

 

Figure 4.4. JMP228 is a superior competitor to DBOI when succinate is
supplied as the limiting resource.

 

 

 

 

10
/l
8 « 5 // 8
In (JMP228/ 6 “ /°/
2 /
P a
0
0 50

100 150 200
hours

Figure 4.5. In of the population sizes of JMP228/DB01 when succinate
is supplied as the limiting resource. The positive slope of the line (0.055)
and the r2 value of 0.9761, indicate dominance by JMP228. Filled
diamonds are competitions of DB01/pJP4 with JMP228/pJP4, and circles
are competitions of DB01/pJP4 with JMP228/pEMT3.

164

 

 

Plasmid pJP4 conferred dominance in batch culture intra-speciﬁc
competitions. All competitors were able to maintain a viable population
size in minimal medium with 200 pg ml'I 2,4-D and a daily transfer of 10%
culture volume, although only JMP228/pJP4 was capable of (completely
utilizing the provided 2,4-D in that time period. No signiﬁcant differences
were noticed in pair-wise combinations of the same host population with
different antibiotic resistance markers that were carried on the same
plasmid. For example, in 12 independent, replicate, batch culture
competitions among JMP228R/pJP4 and JMP2288/pJP4, the effects of host
strain antibiotic resistance determinants did not have a large effect on the

dominant population (Fig. 4.6).

 

 

 

 

 

10
5
lnrifR/strR 0e .5 . r . ‘L . 9-1
-5
—10
O 20 4O 6O 80 100
hours

Figure 4.6. Effect of antibiotic resistance markers to fitness of
JMP228/pJP4 and JMP228/pEMT3. In of the population sizes of
JMP228R/pJP4 to JMP2288/pJP4 (solid circles), and JMP228R/pEMT3 to
JMP228S/pEMT3 (open circles) when 2,4—D (200 ppm), was supplied as
the limiting resource. The slope of the line is 0.002 with an r2 value of
—O.5465, indicating that the population coexisted.

165

 

 

The results of 12 independent replicates also indicated that
JMP228/pJP4 was competitively dominant over JMP228/pEMT3 when
grown on 200 pg ml'l 2,4-D. These ﬁndings conﬁrm the predictions made
using the pure culture growth kinetics, in which the strain with the faster
growth rate predominated. There was a signiﬁcant increase in the
proportion of JMP228S/pJP4 relative to JMP228R/pEMT3 (Figure 4.7).
The same trend was evident even if the pEMT3 carrying strain was given a

ten-fold greater initial population size (data not shown).

166

T - . til—1

 

 

 

1.E+10 ’
1.E+09

 

1.E+08
CFU/ml 1.E+07
1.E+06

1.E+05 ' —I—-JMP228R/pJP4 —O—JMP2288/ JP4
-ﬂ--JMP2288/pEMT3 --0--JMP228 pEMT3

 

 

\ ‘\

 

 

 

 

 

 

1.E+04
O 20 40 6o 80 100 120
hollrs
10
5 ,- -,,,_,,/g//._-,,_ a
In (JMP228/PIP4/ 0 ,, -6 ,, r r—
JMP228/ PEMT3) */
-5
-10
O 20 40 6O 80 100
hours

Figure 4.7. Competition between JMP228/pJP4 and JMP228/pEMT3
for 2,4-D in batch culture. Top panel: the CFU/m1 of each competitor
JMP228/pJP4 to JMP228/pEMT3 in 24 hour batch transfer experiments.
Squares are one experiment, and circles are another, each with three
replicates. Solid symbols represent pJP4 carrying strains, and the open
symbols are pEMT3 carrying JMP228. Bottom panel: the In of the
population sizes of JMP228R/pJP4 to JMP22SS/pEMT3 (open circles), and
JMP228S/pJP4 to JMP228R/pEMT3 (closed circles) when 2,4-D (200
ppm) was supplied as the limiting resource. The slope of the line (0.07)
with an r2 of 0.9232, suggested dominance by the pJP4-bearing strain of R.

eutropha.

167

'“—_:l

Although the two antibiotic resistant populations of DB01/pJP4
exhibited similar growth characteristics when in pure culture, an advantage
was noted for the rifampicin resistant populations relative to the
trimethoprim resistant populations when they were grown together (Figure
4.8). While the effect of antibiotic resistance cannot entirely be ignored, it
did not have a profound effect on competition when the strains carried
different plasmids or when DB01 strains competed with JMP228/pJP4 or
JMP228/pEMT3. As with the JMP228 strains, intra-species competition
experiments between DB01R/pJP4 and DB01T/pEMT3 or DB01T/pJP4 and

DBOlR/pEMT3) lead to dominance by the pJP4 carrying population (Fig.

 

 

 

4.9).
10 " A"
Ln(DBOlrirR/ 0 ___, ”a [:Bf’ma .
DB01 triR) A ' g
-10
0 20 40 60 80 100
hours

Figure 4.8. Effect of antibiotic resistance markers on fitness of
DB01/pJP2 and DBOl/pEMT3. In of the population sizes of DB01R/pJP4
to DB01T/pJP4 (triangles), and DB01R/pEMT3 to DB01T/pEMT3
(squares). The slope of the line (0.034) and the r2 value (0.7360) indicate
that the rifampicin resistant population may have an advantage over the
trimethoprim resistant population.

168

 

1.E+10 -
—I—DBOIR/pJP4 —-D--DB01T/pEMT3

1.E+08 .. .. a... DBOlR/pJP4 {E - DB01T/pEMT3

  

CFU/ml 1.E+06

 

 

 

 

 

 

1.E+04 ‘
1.E+02 _ f V F
0 2O 4O 6O 80 100 120 :
hours l‘
1
10 * I .5
5 ,-../x ,...}! E
1“(DBOl/pJP4/ ,......mx .--» E
DBOl/pEMT3) 0 I - - _1
-5
-10
0 20 40 60 80 100 120
hours

Figure 4.9. Competition between DB01/pJ P4 and DB01/pEMT3 for 2,4-
D in batch culture. Top panel: the CFU/ml of each competitor DBOl/pJP4
to DB01/EMT3 in 24 hour batch transfer. Black and gray lines indicate two
independent experiments, each run in triplicates. Closed symbols represent
pJP4 carrying strains and open ones are pEMT3 carrying DB01 strains.
Bottom panel: the In of the population sizes of DB01R/pJP4 to
DB01T/EMT3 (solid symbols), and DB01T/pJP4 to DB01R/EMT3 (open
symbols). The slope of the line for DB01T/pJP4 to DB01R/EMT3 had r2
values of 0.72, and DB01R/pJP4 to DB01T/EMT3 had r2 values equal to
0.94.

 

169

Inter-species batch c0mpetition between R. eutropha JMP228 and B.
cepacia DB01, resulted in dominance by the JMP228 str'ainsjwhen both
competitors had the same plasmid, whether transferred at 24 hour or 72 hour
intervals. In the 24-hour competitions, three replicate ﬂasks for each
antibiotic resistance combination were performed. In the 72-hour
competitions, only one of the two possible antibiotic resistance combinations
was combined, but three replicates at three different starting ratios (1 :10, 1:1
and 10:1) of each competitor strain was utilized. Regardless of the ratio of
competitors, the same strain from the competitive set became dominant.
Results of the 1:1 ratio are presented, because they are representative of all
three starting ratios. The outcome of competition for both 24 and 72 hour
transfer experiments were similar. DB01/pJP4 strains were outcompeted by
JMP228/pJP4 and similarly DB01/pEMT3 strains were outcompeted by
IMP228/pEMT3 (Figures 4.10 and 4.11).

When alternate host/plasmid pairs were subjected to interspecies
competition, again, the population with the fastest growth rate became
dominant. Intra-species competition among the pairs DB01/pJP4 vs

JMP228/pEMT3 and DB01/pEMT3 vs JMP228/pJP4 were carried out

170

(Figures 4.12 and 4.13). The dominant population in each case was the

strain with the faster growth rate; the one that carried pJP4.

 

 

 

 

 

 

 

 

 

1.E+08
1.E+07 F
l
1.E+06t l
CFU/ml i
1.E+05
+JMP228R/pJP4
1°E+04 —A—DB01T/pJP4 L-
1.E+03 "
0 50 100 150 200 250 300 350
hours
15
10
In (JMP228/pJP4)/
(DBOl/pJP4)) 5 [If 9
t/,,_
-5

0 50 100 150 200 250 300 350
hours

 

Figure 4.10. Competition between JMP228/pJP4 and DB01/pJP4 for
2,4-D in batch culture. Top panel: the CFU/m1 of competitors
JMP228R/pJP4 (solid symbols) and DB01R/pJP4 (open symbols). Bottom
panel: the In of the population sizes of JMP228R/pJP4 to DB01T/pJP4. The
slope of the line, 0.02 had an r2 of 0.9366, indicating that R. eutropha
populations increased signiﬁcantly relative to B. cepacia.

171

 

1.E+08 ~
1.E+07
1.E+O6

CFU/ml 1.E+05

    

 

 

 

 

 

 

1.E+04-
' +Dl301r/pEMT3
1.E+02 : A .
0 50 100 150 200 250 300 350 F
hours :~
15 L
10 o ///,/"'//6
1n (JMP228/ pEMT3 / /
DB01/pEMT3) 5 e 3/ﬂ/f0’”
0 -m

 

 

0 50 100 150 200 250 300 350

hours

Figure 4.11. Competition between JMP228/pEMT3 and DB01/pEMT3
for 2,4-D in batch culture. Top panel: the CFU/ml of competitors
JMP228R/pEMT3 (solid symbols) and DB01T/pEMT3 (open symbols).
Bottom panel: the In of the population sizes of JMP228R/pEMT3 to
DB01T/pEMT3. The slope of the line, 0.036 had an 12 of 0.963, indicating
that R. eutropha again was dominant to B. cepacia.

 

172

 

 

 

 

 

 

 

 

1.E+08
1.13+07 4
1.E+06
1.E+O4
—I— J MP22 8R/pJP4
1'E+03
—A— DBO 1 T/pEMT3
l-E+02
0 100 200 300 400
hours
10
8 , W/B
[11 (JMP228R/P1P4 / 4 (ll/aw,“
DB01/pEMT3)
2 &/a/'
0? ._ ‘73-‘- A #74
2

 

0 50 100 150 200 250 300 350

hours

Figure 4.12. Competition between JMP228/pJP4 and DB01/pEMT3 for
2,4-D in batch culture. Top panel: the CFU/ml of competitors
JMP228R/pJP4 (solid symbols) and DB01T/pEMT3 (open symbols).
Bottom panel: the ln of the population sizes of JMP228R/pJP4 to
DBOlT/pEMT3. The slope of the line, 0.023 had an r2 of 0.926, indicating
the faster growth conferred by pJP4 results in dominance.

173

 

 

1.E+08

 

 

 

1.E+07*
1.E+0t5I
CFU/ml 1.E+05
1.E+04
+JMP228R/pEMT3
1'15“” -A—DBOlT/pJP4
1.E+02
0 100 200 300 400
hours

 

3i

'2 a“\
1n (JMP228/pEMT3 /_4 \
DB01/pJP4) _6 . ' ““\-.
0 \\__Q
-3 v- o
-10

 

 

 

O 50 100 150 200 250 300 350

hours

Figure 4.13. Competition between JMP228/pEMT3 and DB01/pJP4 for
2,4-D in batch culture. Top panel: the CFU/ml of competitors
JMP228R/pEMT3 (solid symbols) and DB01T/pJP4 (open symbols).
Bottom panel: ln of the population sizes of JMP228R/pEMT3 to
DB01T/pJP4. The slope of the line, -0.025 had an r2 of 0.950, again
supporting that the faster growth conferred by pJP4 resulted in dominance.

174

 

 

Modeling the chemostat competitions. Chemostat competitions were
modeled using the Stella Research Software Package (v 5.0). The software
is designed to run mathematical models, when the appropriate constants and
equations are introduced to the various parts of the model. A drawing of the
model for a two species competition is shown in ﬁgure 4.14. In this ﬁgure,
the large black rectangles represent the population sizes of each competing
bacterial strains (1 and 2), and the concentration of 2,4-D remaining in the
chemostats. Initially this value was equal to each original population size,
and 200 ppm 2,4-D, as in the experimental populations. The thick arrows
connected to clouds with a gray circle describe rates, and all of them are
represented by an equation that inﬂuences the item in the black box. In the
case of bacterial growth, there is one input and one output. Both Ks and pImam
are needed as constants for the input, and the Monod (1954) equation can
solve for the population size over time. For example, “growth Bacteria l” is
equal to the change in population size over time described by equation (g).
The output is equal to the p0pulation size (N) times the dilution rate (D). So,
the population size of “Bacteria 1” is described by equation (3).

Equation (g)

dN_N S
Tr " “m” lags

175

 

1g-

 

  

growth Bacteria l
@———+

3:13.74

i o . V" .

f e.. 7,745“
.-

Ksl

washout Bacteria l

   
   

Bacterial

growth rate 1

inﬂow 2.4-D

/-**\ ___} 2,4-1) '________§ “ dilution
\ . a , )3: e1- " . a rate
airmen
rate '.
reservoir 24D
growth Bacteria 2 washout Bacteria 2

 

Bacterial _—

“”0

growth rate 2

Figure 4.14. The model that was used by the Stella Research Program
to generate expected population growth curves for two competing
populations. The constants for KS, growth rate, and efﬁciency were chosen
using an average value determined for each host/plasmid combination (Table
4.3).

176

 

 

Substrate concentration (2,4-D) has one input and three outputs. Input
is the reservoir concentratiOn of 2,4-D (200 ppm) times the dilution rate.
Outputs are the utilization of 2,4-D by the two bacterial strains (a product of
the population size and the efficiency of 2,4-D utilization by each strain)
and dilution (dilution rate times substrate remaining). The efﬁciency of
utilization (g / CFU) is the inverse of the growth yield CFU/g. The curved

arrows indicate that there is a relationship between one component and

another.

Table 4.3: Values for constants in the Stella model.

 

 

JMP228/ JMP228/ Dnor/ DB01]
Variable pJP4 pEMT3 pJP4 pEMT3
pm, (h") 0.22 0.10 0.15 0.08
Ks 1 (ppm) 1 1 1 1
K3 2 (ppm) 1 1 0.25 0.25
Starting Emulation 2 x 107 2 x 107 2 x 107 2 x 107

1x108 1x108 1x108 1x108

°m°iency (g’CFU) 2.7 x 10'7 2.7 x 10'7 2.9 x 10"7 2.9 x 10'7
dilution rate (11") 0.033 0.033 0.033 0.033
reservoir (mg/ L) 200 200 200 200
original 2,4-D 200 200 200 200

 

177

 

 

For each host/plasmid pair that was competed in a chemostat, a model
was run. These are graphically shown as dashed lines superimposed on each
graph of the measured values in the chemostat. Two different values of Ks
were used to generate models of competition with DB01 strains, in order to
illustrate how a lower value of Ks could allow a population with a slower
growth rate to predominate at these low nutrient concentrations. For the
point to be made, only one graph of the DB0] strains with Ks == 1 ppm is
shown. The reasons for using a Ks value of 0.25 ppm for the DB01 host,
will become evident in the next section.

Chemostat competitions. Chemostat competitions were carried out
simultaneously with pure cultures. The chemostats were fed very slowly
(0.033 b"), so that nutrients (2,4-D) would become limiting. In the
computer simulated models run at the above conditions, 2,4-D levels in the
chemostats attained ﬁnal concentrations of between 0.15 and 0.7 ppm,
values that lie below the 2 ppm limit of resolution of the HPLC. The results
of the competitions were compared with predicted population sizes that were
determined using the Stella Research Software Package. Plasmid pJP4
plays a lesser role in chemostat competition than in batch culture, and also
supports the theory that if the Ks of the two populations is the same, then the

strain with the maximum growth rate would become dominant. First,

178

P...

 

"E.—

 

consider the case of the same host population with a different plasmid. In
reciprocal combinations within chemostats between JMP228S/pJP4 and
JMP228R/pEMT3 or between JMP228R/pJP4 and JMP228S/pEMT3 both
competitors coexisted, but the JMP228/pJP4 combination was numerically
dominant. (Figure 4.15 and 4.16) A plot of the difference in the Malthusian
parameters tracked over 300 h for each pair indicated a slope of 0.004 and
0.190 in favor of the pJP4 carrying strain. Since this is a small slope with an
r2 value of 0.566 and 0.621, there was a marginal deviation from a slope of
zero. Carried out over a longer period of time, this may have lead to the
exclusion of JMP228/pEMT3. Qualitatively, the strain that was expected to
dominate, did. However, the magnitude of the difference was not nearly as
great as expected by the model. The pEMT3 carrying strain was not
excluded nearly as rapidly as one would have predicted.

In competition studies between DB01/pJP4 and DB01/pEMT3 it was
observed that the experimental results were consistent with the modeled
results, except that the ﬁnal population sizes of the two competitors were
less than predicted. (Figure 4.17). When the slope of the In of the ratio of
the pJP4 carrying strain to the pEMT3 carrying strain was graphed, it could
be seen that the DB01/pJP4, with a faster grth rate, also became dominant

over DB01/pEMT3.

179

 

 

 

 

 

 

 

1.E+10 -
.. 225
1.E+09
L 175
1.E+08
CFU/ml ~ “* ‘25
1.E+07 «~ g 75 2.4-D (ppm)
1.E+06 ~~ -_ 25
3 b
1.E+05 : ' ‘ 1 t -25
0 50 100 150 200 250 300

 

10

5 h

In (JMP228S/pJP4/ I ,- ,. ”ﬁrm“
1‘
JMP228R/pEMT3) . r ' '. . . an

-5-

 

 

 

-10
0 50 100 150 200 250 300
hours

Figure 4.15. Chemostat competition between JMP2288/pJP4 and
JMP228R/pEMT3. Top panel: the colony forming units of JMP2288/pJP4
(black squares) and JMP228R/pEMT3 (gray open squares) in chemostat
competition. The expected growth curves for each strain are indicated with
black and gray lines, respectively. Solid diamonds indicate measured 2,4-D
concentration, open ones are predicted by the model. Bottom panel: the
natural log of the ratio pJP4 carrying to pEMT3 carrying strains of JMP228.
The slope of the line (0.004) and r2 (0.567) indicate a marginal advantage by
JMP228S/pJP4.

180

 

 

 

 

 

 

 

 

l225
1.E+09 ,_ ___.
0.] ‘“ 175
1.E+08 /. .. #125
CFU/ml ll. .1.- . ~ ' u. ‘_ 2,4-D ppm
1.E+07 ~ 3 \-\\ - _ . 75
1.E+06 3 \-\\ _25
4‘6}
1.E+05 - a a -25
0 50 100 150 200 250 300
hours
10-
5 ._
1 (JMP228/pJP4/ ' ’
n L... M-,...ﬂ/j”
JMP228/pEMT3) 0"! 'e ' " ' '
-5 J
-10

 

 

 

0 50 100 150 200 250 300

hours

Figure 4.16. Chemostat competition between JMP228R/pJP4 and
JMP2288/pEMT3. Top panel: the colony forming units of JMP228R/pJP4
(black squares) vs. JMP2288/pEMT3 (gray open squares) in chemostat
competition. The expected growth curves for each strain are indicated with
black and gray lines, respectively. Solid diamonds indicate measured 2,4-D
concentration, open ones are expected. Bottom panel: the natural log of the
ratio pJP4 carrying to pEMT 3 carrying strains of JMP228. The slope of the
line (0.190) and r2 (0.621) indicate a marginal trend in favor of
IMP228R/pJP4.

181

 

 

 

 

 

 

L 225
/’_‘\/ " 125
“x- \x‘ . 75 2,4-D ppm
3 T \“e 25
M‘- "l’
* i -25

 

 

0 50 100 150 200 250 300

 

 

 

hours
10
5 .. .
_,2_-_M
ln (DB01T/pJP4/ .. .--f
DB01R/pEMT3) (lift/v P . . . ,
-5 i
-10

 

 

0 50 100 150 200 250 300
hours

Figure 4.17. Chemostat competition between DB01T/pJP4 and
DB01R/pEMT3. Top panel: the colony forming units of DB01T/pJP4
(black triangle) and DB01R/pEMT3 (gray open triangle) in chemostat
competition. The expected growth of each strain is indicated with black and
gray lines, respectively. Solid diamonds indicate measured 2,4-D
concentration, open ones are obtained from the model. Bottom panel: the
natural log of the ratio pJP4 carrying to pEMT3 carrying strains of DBOl.
The slope of the line (0.02) and r2 (0.868) indicate a marginally signiﬁcant
trend in favor of DB01/pJP4.

182

 

 

In two independent replicate chemostat experiments with reciprocal
antibiotic resistance markers on the populations, JMP228/pJP4 and
DB01/pJP4 were competed. A surprising outcome was observed. While
JMP228/pJP4 dominated over DB01/pJP4 in batch transfer competition, the
DB01/pJP4 strain outcompeted JMP228/pJP4 in these slow growth rate
chemostats (ﬁgure 4.18). Notice that the expected values for these strains
did not correlate well with the observed values, when p=0.22 for
JMP228/pJP4 and p= 0.15 for DB01/pJP4, and each has a Ks of 1 ppm. The
prediction would be that the faster growing strain (JMP228/pJP4) should
dominate. On separate occasions these two host/plasmid combinations were
competed. The same outcome was observed: DB01/pJP4 became dominant
over JMP228/pJP4. One representative graph of the competition between
JMP228S/pJP4 and DB01R/pJP4 is shown (Figures 4.18 and 4.20). The
SIOpe and r2 of the ratio of the population is reported here for comparison of
the three experiments. The slope was calculated after the initially supplied
2,4-D was depleted, to eliminate the rapid growth period by R. eutropha.
The slope of the difference in the Malthusian parameter of In
(JMP228S/pJP4)/ (DB01R/pJP4) was -0.072 and correlated to a r2 value of
0.908 in the graphed example. The other two replicates resulted in a 1n ratio

or JMP228/DB01 with slopes = —0.045 and -0.01 with r2 = 0.8924, and

183

""7

 

 

0.8043 respectively. The negative slope to the line indicated dominance by
DB01. The increase in DBOlR/pJP4 occurred after 120 hours of incubation,
and by 150 hours, the D801 strain had increased to numbers higher than that

of JMP22SS/pJP4.

 

250
5* 200
r 150

CFU/ml -. 100

294-1) (ppm)

 

~~0

+ ~50
0 50 100 150 200 250 300 350

hours

 

 

 

Figure 4.18. Chemostat competition between JMP2288/pJP4 and
DB01R/pJP4. JMP228S/pJP4 (black square) and. DB01R/pJP4 (gray open
triangle) in chemostat competition. The expected growth curves for each
strain are indicated with black and gray lines, respectively. Solid diamonds
indicate measured 2,4-D concentration, and open ones are expected. The
modeled lines (black and gray for each strain as identiﬁed), were for
JMP228S/pJP4, p=0.22 h“, KS=1 ppm and DBOlR/pJP4, p=0.15 h", Ks=1
ppm. Notice that it is predicted that JMP228 should become dominant.

184

 

VF. .1? .r'T‘R 1—._ -4»

 

Since the observed population dynamics did not match the predicted
curve, alternative explanations for why DB01 became dominant over
JMP228 were explored. The ﬁrst of these ties back into 'the pure culture
experiments for determining Kg. The method used may not have been
sensitive enough to accurately calculate the Kg for these strains at these low
concentrations of substrate. Evidence from ﬁgure 3.6 suggests that the Kg
being at least as low as 1 ppm, if not signiﬁcantly lower. It is also possible
that DB01 may be more efﬁcient at utilizing very low concentrations of as
described by the 14C studies. DB01 appeared to completely mineralize a
greater fraction of the total 2,4-D at concentrations below 1 ppm (Chapter 3).

In a set of hypothetical curves, also run on the Stella v 5.0 software,
we can see that as little as a 50% decrease in the Kg of a strain with a lower
growth rate, can be sufﬁcient to change the outcome of competition. Figure
4.19, illustrates the consequence of a sequential decrease in the Kg of
DB01/pJP4, which has p = 0.15 h", when it competes with JMP228/pJP4,
which always has a Kg=1, and p = 0.22 h". Hypothetical case A, both
strains have Kg =1. In case B, DB01/pJP4, has Kg= 0.1, in case C, Kg= 0.25

and in case D, Kg= 0.01 ppm.

185

 

T

 

;_

 

1.E+10 *
DBOl/pJP4, Ks= 0.1 (D)

1.E+09 st . nannunnauunnnnﬂnuu JMP228/pJP4, KS=1 (A)

' DBOl/ JP4, Ks: 0.25 (C)
DBOl pJP4, Ks: 0. 5 (B)

 

 

 

1.E+08 ' ' - — . JMP228/pJP4, Ks=1 (B)
CFU/ml “n. - . . ..
1.E+07 Ig“.‘ DBOl/pJP4, Ks—l (A)
JMP228/pJP4,Ks=1 (C)
1.E+06
JMP228/pJP4, Ks=l (D)
1.E+05 , ' ' '

 

hours

Figure 4.19. different KS values for DB01 inﬂuence competition between
JMP2288/pJP4 and DB01R/pJP4 in chemostats. Four hypothetical
competitions between JMP2288/pJP4 (black ﬁlled square) and DB01R/pJP4
(gray open square). With a decrease of KS to 0.5 ppm 2,4-D the slower
growing DB01 is able to become dominant.

For a combination of strains, if we assume that the K5 is the same, (1
ppm), the populations should diverge at a rate proportional to the difference
in their maximum growth rates. This is not observed. Burkholderia
outcompeted Ralstonia. To support the possibility that the KS for strain
DB01 is lower than that of strain JMP228, the upcoming graphs will show
the expected ratios for DBOl calculated with a KS of 0.25 ppm. Using
values generated for the hypothetical population curves, when DB01/pJP4

was assigned a Ks=0.25 expected values did ﬁt the observed data more

186

0 50 100 150 200 250 300 350 400 450 500

. use”
E

 

 

closely. Figure 4.20 plots the same empirical data as in ﬁgure 4.18, but the
hypothetical curves were generated for DB01/pJP4 with a KS = 0.25 ppm
2,4-D. While KS differences can explain one way in which the DB01/pJP4
strain could become dominant, the fact that the shift appears much later than
expected, and more rapidly may indicate the inﬂuence of another parameter.
Evolution of the DB01/p] P4 strain to grow better on the low concentrations
of 2,4-D or perhaps some other physiological change or pathway induction is
responsible for the rapid reversal of the dominant strain in Figure 4.20.
Chemostat competitions between JMP228/pEMT3 and DB01/pEMT 3
indicated that DB01/pEMT3 became dominant. The slope of the ln ratio of
(JMP228/pMET3) / (DB01/pEMT3) was -0.013 and correlated with an r2
value of 0.815. (Figure 4.21). Again a value of 0.25 was used for the
hypothetical Ks of DB01/pEMT3. The emerging trend is that DB01 strains
outcompete JMP228 under the low growth rate, nutrient poor conditions

simulated in these chemostat conditions.

187

l“»ﬂas’.uv I.“ _..-. _7

 

 

 

1.E+10 +- 250
1.E+09 1- " 200

-‘ ﬂ snafﬂe-vii 3 . '4: s" - ﬁﬁ—ﬁ-ﬁ-kﬁﬁyb

1.E+08 3“ ' _ “ - 1' 150
CFU/ml " _ -- 100
1.E+07 4» ‘ 2.4-D (ppm)
‘ .I “ 50
l - ‘ “rﬁ‘?
1.E+06 - . 0

 

 

 

 

 

 

1.E+05 t -50
0 50 100 150 200 250 300 350
hours
10
5 0..
ln (JMP228S/pJP4/ ’ \\
DBOlR/nIP4) 0% . . r. \__ , . .
_5 ‘3‘\_o
-10

 

 

0 50 100 150 200 250 300 350
hours

Figure 4.20. Chemostat competition between JMP2288/pJP4 and
DB01R/pJP4. T0p panel: population size of JMP2288/pJP4 (black solid
square) vs. DB01R/pEMT3 (gray open triangle) in chemostat competition.
The expected growth curves for each strain are indicated with black and gray
lines, respectively. Solid diamonds indicate measured 2,4-D concentration,
open ones are expected from the model. Bottom panel: the natural log of
the ratio JMP228 to DB01. The slope of the line (-0.072) and r2 (0.908)
indicate a signiﬁcant trend in favor of DB01/pJP4. This ﬁnding suggests
that a lower KS may be a factor in allowing the strain with the slower growth
rate to become dominant.

188

 

 

 

 

 

1.E+10 250
.. 200
1.E+09
-_ 150
CFU/ml 1-E+08 100
» O 2,4-D (ppm)
1.E+O7
r 0
1.E+06 ~ -50

 

0 50 100 150 200 250 300 350
hours

 

10

5

1n (JMP228R/pEMT3 0
/DB01T/pEMT3) f “irks“;ox;
5 _. O MR“

 

 

 

 

-10
0 50 100 150 200 250 300 350
hours

Figure 4.21. Chemostat competition between JMP228R/pEMT3 and
DB01T/pEMT3. Top panel: the population size of JMP228R/pEMT3
(black square) vs. DB01T/pEMT3 (gray open triangle) in chemostat
competition. The expected growth curves for each strain are indicated with
black and gray lines, respectively. Solid diamonds indicate measured 2,4-D
concentration, open ones are expected from the model. Bottom panel: the
natural log of the ratio JMP228 to DB01. The negative slope of the line
(-0.013) and r2 (0.819) indicate a signiﬁcant trend in favor of DB01/pEMT3.

189

 

 

The observation that strain DBOl became dominant held true for the
observed population sizes for all of the chemostats of interspecies
competition. Figures 4.22 and 4.23 show competition betWeen populations
of different species with different degradative plasmids: JMP228R/pEMT3
vs. DB01T/pJP4 and for JMP228R/pJP4 vs. DB01T/pEMT3. In both pairs,
DB01 dominated at the end of the experiment. When JMP228R/pEMT3
competed with DB01T/pJP4, it was expected that DB01 should dominate.
First, DB01/pJP4 had a faster growth rate, and additionally, it may have a

lower value for KS. The regression line generated by the natural log of the

ratio of population sizes was negative (048), and the r2 of 0.923, indicating ‘

superior ﬁtness of strain DB01.
On the other hand, the results were more surprising for the

competition between J MP228R/pJ P4 and DB01T/pEMT3. These strains
have a very large difference in their pmax. In fact, when the model is run with

the values of p=0.22 and KS = 1 ppm for JMP228/pJP4 and p: 0.08 and KS
= 0.25 for DBOl/pEMT3, the predicted population dynamics would support
coexistence of the two populations, with a greater proportion of the
community being composed of strain JMP228R/pJP4. In theory the two
populations could coexist. In reality, JMP228/pJP4 was adversely affected

by competition with DB01/pEMT3 (ﬁgure 4.23).

190

 

(EL... , ..

 

 

  

 

 

1.E+10 ~ 250
1.E+09 . 200
1.E+08 ~~ 150
CFU/ml 1.E+07 it 100 2,4-D ppm
1.E+06 50
1.E+05 ~ -» 0
1.E+04 ~ -50
0 50 100 150 200 250 300 350

hours

 

10
5
Ln JMP228R/pEMT3 / at-
DBOlT/pJP4 Oi'ﬁ' \ ' '
O
-5 \i‘\

\\’

 

 

 

 

-10
0 50 100 150 200 250 300 350
hours

Figure 4.22. Chemostat competition between JMP228R/pEMT3 and
DB01T/pJP4. Top panel: population size of JMP228R/pEMT3 (black
square) vs. DB01T/pJP4 (gray open triangle) in chemostat competition. The
expected growth curves for each strain are indicated with black and gray
lines, respectively. Solid diamonds indicate measured 2,4-D concentration,
open ones are expected from the model. Bottom panel: the natural log of
the ratio JMP228 to DB01. The slope of the line (-0.48) and r2 (0.923)
indicate a signiﬁcant trend in favor of DB01/pEMT3.

191

 

 

 

 

 

 

 

 

 

 

250
1.E+09 -~ 200
“150
CFU/ml 1'5”” ‘ .. 100 2,4-D ppm
1.E+05 ““ 50
-. 0
1.E+03 , ‘ I -50
0 50 100 150 200 250 300 350
hours
250
200
- 150
-- 100 2,4—D ppm
-* 50
. 0
1.E+03 -50

 

 

 

0 50 100 150 200 250 300 350

hours

 

10
5 .
ln (JMP228R/pJP4/ 0 W \ . .

DBOlT/ pEMT3) \
-5 - X

-10 ‘

0 l 00 200 300
hours

 

 

 

 

Figure 4.23. Chemostat competition between JMP228R/pEMT3 and
DBOlT/pEMT3. Top panel: the population size of JMP228R/pJP4 (black
square) and DB01T/pEMT3 (gray open triangle) in chemostat competition
when DBOl/pEMT3 has a KS value of 0.25 ppm. Middle panel: the same
observed values as top panel, but with the expected outcome of competition
calculated with the KS of DBOl/pEMT3 assigned to 0.1 ppm. The expected
growth curves for each strain are indicated with black and gray lines,
respectively. Solid diamonds indicate measured 2,4-D concentration, open
ones are expected values. Bottom panel: the natural log of the ratio
JMP228/pJP4 to DB01. The slope of the line (-0.039) and r2 (0.9062)
indicate a trend in favor of DBOl/pEMT3.

192

 

TI“ 3.2mm. spa-z ...an -
3 .

 

Inhibitor production by DB01. Of three scenarios that could have
resulted in the dominance of DB01 over JMP228, two were investigated. In
the above examples, B. cepacia DB01 may indeed posses a lower Kg than
R. eutropha JMP228, although it was not detected by empirical means, and
additional factors may be involved as well. Another plausible cause for the
decline in JMP228, would be if DBOl was producing an inhibitory
compound. Using cell-free ﬁltrates of the chemostat ﬂuid of DB01 strains
with supplemental 2,4-D as the substrate for the growth of JMP228
indicated two things. All of the different dilutions of the culture ﬂuid
supported the grth of the JMP228/pJP4 and JMP228/pEMT3 strains to
the same extent, as measured by turbidity. All cultures reached comparable
densities (0.1 - 0.14), at 600 nm. Less turbidity was seen in phosphate
buffered saline with 2,4-D (0.09). Consequently, there was no indication of
an inhibitory compound secreted to the medium by DB01 during chemostat
growth that adversely affected the growth of JMP228. The experiment also
suggested that there were sufﬁcient micro-nutrients present in the “spent”
media to support growth, especially if more of the primary nutrient (2,4-D)
was supplied. These experiments did not address the possibility of

evolution of ﬁtness among the D801 strains.

193

' ” ""7

 

 

DISCUSSION

When populations compete for limited resources, whether for habitat,
shelter, food or breeding grounds, the interaction may have different
outcomes. The ﬁrst may relate to Gause’s Principle of Competitive
Exclusion. Two competing species cannot coexist indeﬁnitely in a single
niche (Gause, 1934). A mutual requirement for the resource will result in
competition for survival and the population that can utilize the resource
more effectively will dominate. Consequently, populations of bacteria such
as R. eutropha or B. cepacia, competing for one limiting resource (2,4-D),

should have one winning population.

A competitive mechanism that could be used by bacteria is to interfere
directly with a competitor through the production of antibiotics or toxins
(Schroth 1992). This is analogous to “ﬁghting it out” for territory or mates
among land mammals or birds. Results from these experiments did not
indicate the production of inhibitory compounds toward R. eutropha by B
cepacia growing on 2,4-D in chemostats. However, the potential exists that
the analyses performed were insufﬁcient to detect toxins, particularly if they
were lipid soluble, or did not pass through the membrane during ﬁltration.

Lipid soluble toxins, like cepalysins with hemolytic or antifungal properties,
194

 

 

are made by B. cepacia strains (Abe and Nakazawa, 1994). Biocidal activity
toward bacteria and fungi has been reported to involve lipid associated
compounds that are secreted through vesicles. (Hu and Young, 1998). Most
of the toxic activity acts against eukaryotic cells, such as in the colonization
of lung tissue of Cystic Fibrosis patients by B. cepacia. So far the activity of
the cepalycins have not been shown to be active against members of R.

eutropha.

In the absence of inhibitory compounds, the interaction is probably
one of exploitative competition. Among bacterial populations, a competitive
interaction that occurs by more efﬁcient exploitation of a resource has been
demonstrated in numerous experiments using different bacterial populations
and different substrates (Duetz et al., 1994; Gisi et al., 1988; Hansen and

Hubbell, 1980; Lenski et al. 1994).

Once it has been observed that a population becomes dominant over
another, the exact mechanism by which a population becomes a more
successful competitor must be deﬁned. From ecological studies of
succession and community evolution, it has been suggested that populations
fall on a gradient of two basic life style groupings. Populations that are r-

selected maximize their growth rates when resources are plentiful.

195

 

tili'ﬁ‘h r-l --t.’\_~.d‘..". -r.;.‘.

 

Pioneering species that have the capability to grow quickly utilize resources
rapidly in order to fuel a rapidly increasing population size. Weedy plants
(that are the targets of the 2,4-D used in this study) are just such an example.
In bacteria, like in plants, rapid growth may have a price, and as the
populations expand in size resources are exhausted and the environmental
pressures change. Nutrient poor conditions arise, and competition for the
resource becomes ferocious. Levels of utilizable substrate may drop below
the detection capabilities of the r-selected bacteria, and they begin to starve.
Under nutrient-poor conditions, a populations that can detect the lowest
quantities of a resource will be more successful. These are the K-selected
populations, and they are adapted to survival under conditions of crowding
or resource limitation. Populations thought of as r-selected should have high

growth rates, and those that are K-selected, should have low Kg values.

The two different growth conditions used in these experiments, batch
and continuous culture, represent two environments that could favor growth
by bacterial populations utilizing different life history strategies (Andrews,
1986). The r- and K- selection strategies were originally used to describe
sexually reproducing populations, but also have some application to
microbial ecology (Roughgarden, 1971, Kasahara,l99l, Gorlach, 1994). In

essence, bacterial populations that maximize their growth rates should be

196

 

 

good pioneering species and invaders. Both host strains, R. eutropha
JMP228 and B. cepacia DB01 fall into this category of the continuum of
survival strategies. Differences in how the strains compete with one another
in batch and continuous culture suggest that B. cepacia DB01 is more K-

selected and R. eutropha JMP228 is more r-selected.

Obviously, r- and K- selection theories can be tied directly to bacterial
growth kinetics, the properties that have traditionally been used to predict
the outcome of resource based competition among two populations (Monod,
1949; Hansen and Hubbell, 1980). In several experiments, populations were
characterized based on measurable traits such as maximum growth rate
(pm), and the half saturation constant (K,) for a compound supplied as a
sole source of carbon and energy. (Duetz et al. 1994, Greer et al. 1992,
Graham, 1993). Previously, a kinetic comparison was performed among
seven 2,4-D degrading strains of bacteria but this study did not determine
how the different strains would interact when in competition (Greer et al.,
1992). Growth parameters were estimated using the Monod model ﬁtted to
2,4-D depletion data. This data indicated that there was generally very little
difference among the examined strains. Since at least ﬁve of the isolates
were obtained by enrichment from sewage sludge, the likelihood of

comparing only fast growing phenotypes was quite high. The experiments

197

presented in this Chapter were designed to test not only the fastest growing
2,4-D degraders, but rather to examine combinations of alleles that would
confer either a high growth rate or a low growth rate to the host population.
In order to predict dominance among two populations carrying either of two
plasmids in terms of their 2,4-D degradative characteristics, the maximum
growth rate and Kg of the bacterial strains were determined. Based on
resource competition theory, under high nutrient concentrations and batch
transfer conditions, the dominant strain would be the one with a faster
growth rate. In contrast, the dominant strain under low nutrient continuous
culture conditions would be the one to reduce the substrate to the lowest
concentration. When resource levels are depleted to a level at which a
competing population would be unable to maintain a positive grth rate or
viable population size, the population with the lower threshold value should

become dominant. (Tilman, 1981; Smith, 1993).

Differences in alleles encoding enzymes for the degradation of a
compound can inﬂuence the ﬁtness of the bacterial population that utilizes
that compound as a growth substrate. The ability to fully mineralize a
compound is determined by a combination of the effects of the host
population and the plasmid carrying the degradative alleles. The research
presented here, took advantage of the naturally occurring diversity in the

198

degradative plasmids, alleles and bacterial populations capable of the
complete catabolism of the herbicide, 2,4-D. To test whether our
observations matched our pure culture predictions, populations were
combined in equal ratios and the relative ﬁtness’ of the strains were
assessed using the slope of the line generated by the ratio of the natural log
of the population sizes (Lenski et al. 1992). By competing isogenic host
populations carrying different 2,4-D alleles, as well as different species
with the same alleles, the contribution of the different plasmids, with
different degradative genes to the ﬁtness of the host populations were
determined.

It was expected ﬁom pure culture growth kinetics and prior research,
that supplying high concentrations of 2,4-D would result in the enrichment
of the pJP4-like alleles (Dunbar et al., 1996). In all batch competition
experiments the successful competitor was the one with the faster growth
rate, regardless of whether it was in members of the same host strain or
different host strains. Plasmid pJP4 conferred dominance over plasmid
pEMT3 in isogenic host populations, in both JMP228 and DB01 strains.
These results supported a correlation between plasmid borne genes and

ﬁtness of the host strain.

199

 

In order to evaluate the outcome of competition in batch culture, it
was necessary to allow enough time for all of the 2,4-D to be degraded by
the competing populations. In 24 hours, only IMP228/pJP4' could utilize all
of the substrate, while all strains degraded the substrate in 72 hours. Of
what signiﬁcance is stationary phase dynamics to the competitive interaction
among the populations? Once the substrate is depleted, the cultures entering
into stationary phase may begin to die off. Mortality could be signiﬁcant if
death rates during stationary phase are high. Declines in the R. eutropha and
B. cepacia populations used in these experiments were not seen during the
ﬁrst three days of stationary phase. In the 72 hour transfer experiments,
population sizes were determined daily during each of the ﬁrst 3 days
(Figures 4.10, 4.11, 4.12 and 4.13). JMP228/pJP4 populations did not
decline despite spending 48 hours in stationary phase, and the slower
growing strains were not in stationary phase for more than 24 h.

Interestingly, in these experiments, the 2,4-D degradative plasmid-
allele combination was most inﬂuential in intra-species interactions, and was
most pronounced in the batch competition environments. These results
contrasted those observed for dichloromethane (DCM) dehalogenases ﬁom
two different species of methylotrophic bacteria. In the experiments by Gisi,

(1998), two isofunctional dehalogenases supplied to Methylobacterium DM4

200

 

(without dehalogenase) using a plasmid vector, had different activities under
batch and chemostat conditions. The dehalogenase of DMll lead to
dominance of its host strain in conditions of substrate exCess, and that of
DM4 lead to dominance under substrate-limited conditions. For different
catabolic functions, such as the degradation of 2,4-D and DCM
dehalogenases, it is conceivable that there is a different relative contribution
of the donated and chromosomally encoded genes

While the batch competition experiments indicated that the plasmid
and its corresponding genes did have an effect on intra-species competition,
the properties of the host strain were also critical to interspecies competition
in the two different growth environments. Batch culture, discussed above,
was selective for populations with a fast grth rate, and would represent
the r-selected environments. In fact, while very little difference was
observed for the growth rate of Ralstonia and Burkholderia with succinate as
a carbon source, JMP228 with either plasmid, out competed DB01/pJP4.
The host strain R. eutropha had a distinct advantage over B. cepacia. The
dominance of strain JMP228 over DB01 might have been due to a longer lag
time for the B. cepacia strain, as well as to the slightly faster growth of R.
eutropha. When grown on 2,4-D in batch conditions, R.eutropha also

outcompeted B. cepacia, when both utilized the same plasmid.

201

The half-saturation constant (Kg) refers to the concentration of

resource that allows for growth at one-half the maximum growth rate, and
reﬂects an increased afﬁnity of an organism for a substrate. Low KS values

and high population densities under nutrient limited conditions, are
indicative of K-selected populations. It is most practical to examine low
substrate conditions in continuous culture, where byproducts and cells are
removed as new resource is supplied. The organism that displays a lower
substrate threshold should be able to reduce the resource to a lower
concentration than its competitor can utilize. The magnitude of the
difference would be related to the amount of extra available resource the
successﬁll competitor will have access to after the other strain is no longer
capable of responding to the presence of the nutrient.

Kg values obtained experimentally, were very similar among the four
strain/plasmid combinations. With the same Kg, the same dominant
population should have been observed in both batch and chemostat
competitions, but surprisingly, different results were obtained. Intra-speciﬁc
competition between JMP228/pJP4 and JMP228/pEMT3 strains resulted in
coexistence of the two different plasmid-bearing populations, and the same
was seen in competition between DB01/pJP4 and DB01/pEMT3. The
pEMT3 carrying strain initially decrease relative to the pJP4 carrying strain,

202

 

due to the initial 2,4-D that was utilized by the faster growing strain.
Eventually the two populatiOns carrying different plasmids equilibrated, with
slightly larger numbers of the p] P4 bearing population.

The most surprising result was obtained in the inter-speciﬁc
chemostat competitions. During the ﬁrst 8 hours of equilibration, R.
eutropha outcompeted B. cepacia, as would be expected by its faster grth
rate on the initial 2,4-D. After the original resource was depleted and both
populations sought a new equilibrium, DB01/pJP4 and DB01/pEMT3 strains
began to increase in numbers and slowly replace the JMP228 competitor
populations. How did the slower growing DBOl dominate?

Another explanation for dominance of DB01 populations may lie in
the differences in the resource uptake mechanisms or afﬁnity for the
substrate, despite the fact that there was little observed difference in the
measured value for Kg. Perhaps the difference was in the manner in which
the 2,4-D was utilized in the slow growth rate environments. In pure culture
experiments (Chapter 3), the DB01 populations appeared to convert more of
the 2,4-D to C02, which may have allowed them to more completely utilize
more of the substrate than the JMP228 strains. Perhaps the Kg for DB01
was lower than that of JMP228, but the experiment to determine Kg was not

sensitive enough to detect a difference between the strains. When models

203

 

were run with lower Kg values assigned to DBO], expected and observed
population dynamics resulted in similar outcomes, supporting the hypothesis
that DB01 does, indeed, possess greater afﬁnity for the 2,4-D at low
resource concentrations.

It is also possible that there are differences between JMP228 and
DB01 in the portion of the degradation pathway involved with the
conversion of CMA to succinate. The enzymes responsible for this
transformation are chromosomally encoded, and may differ between the two
species of bacteria. To date, it is not known how different the mar genes of
Ralstonia and of Burkholderia are, or if variations in this portion of the
pathway could be sufﬁcient for inﬂuencing the outcome of the competitive
interaction between the two strains.

Since antibiotic resistance markers were used to selectively identify
the pJP4 and pEMT3 carrying strains, two scenarios could explain the
coexistence of the same host with different plasmids. The ﬁrst was that a
pJP4 carrying strain, resistant to both antibiotics was being detected, rather
than the pEMT3 carrying strain. Alternatively, pJP4 had transferred from
JMP228/pJP4 to the JMP228/pEMT3, or ﬁ'om DB01/pJP4 to DB01/pEMT3.
Initial experiments conducted on antibiotic resistant populations indicated

that there was a very low frequency of occurrence of a doubly resistant

204

 

antibiotic phenotype (<10 CFU per 108 CPU plated), and colonies resistant
to both antibiotics were obtained from the chemostats at < 102 CPU ml".
Resistance to both antibiotics could not account for the high numbers of the
pEMT3-host type detected in the competition experiments. Could plasmid
tranfer promote the replacement of pEMT3 due to incompatibility of the two
Inc-P plasmids? Although the possibility for plasmid transfer exists, there
was little evidence to suggest that this was the driving force behind the
coexistence. Plasmid transfer was investigated by using a technique to
determine which 2,4-D alleles for tfdA and tde were found in of 10
colonies of JMP228R/pEMT3 from a the 10'7 selective plate, as well as 10
colonies for each of the JMP229S/pJP4 strains from the 10"7 dilution. The
colonies were analyzed using the Polymerase Chain Reaction (PCR) with
primers speciﬁc for the tfdA and tde genes of pJP4 as described by
(Vallaeys et a1, 1993). Results indicated that since the JMP228/pJP4 and
DB01/pJP4 yielded a product of the appropriate size with both primer sets,
all those isolates maintained genes like those of plasmid pJP4. The
JMP228/pEMT3 and DB01/pEMT3 colonies all yielded product with the
tfdA primer set, but did not amplify with the tde set, suggesting that pJP4

had not been acquired by the population.

205

In conclusion, host encoded characteristics are crucial to inter-speciﬁc
competition, particularly under different growth conditions. If a plasmid
provides a faster growth rate to its host, that population will gain a
competitive advantage when resources are plentiful, regardless of the strain
hosting that plasmid. However, at the lower limits of substrate acquisition,
small differences in resource utilization, such as the Kg for 2,4-D, may

swing the pendulum in favor of a different, slower growing population.

ACKNOWLEDGMENT

This research was funded by a grant from the National Science
Foundation (BIR 912006)

206

 

REFERENCES

Abe, M, and T. Nakazawa. 1994. Characterization of hemolytic and antifungal
substance, cepalycin, from Pseudomonas cepacia. Microbiol. Immunol.
38:1-9

Andrews H. H. and R. F. Harris. 1986. r- and K- Selection and Microbial
Ecology. In: ED. K.C. Marshall. Advances in Microbial Ecology. Vol.9.
Plenum Publishing. pp 99-147.

Chao, L., B.R. Levin and FM. Stewart (1977). Complex community in a simple
habitat: Experimental study with bacteria and phage. Ecology. 58:369-
378.

DiGiovanni G.D., J.W. Neilson, I.L. Pepper and NA. Sinclair. 1996. Gene
transfer of Alcaligenes eutrophus JMP134 plasmid pJP4 to indigenous soil
recipients. Appl. Enviom. Microbiol. 62:2521-2526

Don, R. H. and J .M. Pemberton. 1981. Properties of six pesticide degradation
plasmids isolated from A Icaligenes eutrophus and Alcaligenes paradoxus.
J. Bacteriol. 145:681-686

Don, R. H., A.J. Weightman, H.J. Knackmuss and K. N. Timmis. 1985.
Transposon mutagenesis and cloning analysis of the pathways for

degradation of 2,-4dichlorphenoxyacetic acid and 3-chlorobenzpate in
Alcaligenes eutrophus JMP134 (pJP4). J. Bacteriol. 161:85-90.,

Duetz W. A., C. DeJong, P.A. Williams, and J. G. Van Andel. 1994. Competition
in chemostat culture between Pseudomonas strains that use different
pathways for the degradation of toluene. Appl. Environ. Microbiol.
60:2858-2863.

Dunbar, J., D.C. L. Wong, M.J. Yarus, and L. J. Forney. 1996.
Autoradiographic mothod for isolation of diverse microbial species with
unique catabolic traits. Appl. Envrion. Microbiol. 62:4180-4185.

Dunbar, J., S. White, and L. Fomey. 1997. Genetic diversity through the
looking Glass: effect of enrichment bias. Appl. Envrion. Microbiol.
63:1326-1331.

Dykhuizen, D. E. and D. L. Hartl, 1983. Selection in chemostats Microbiol.
Rev. 47: 150-168.

207

 

Fukumori F. and RP. Hausinger. 1993. Alcaligenes eutrophus JMP134 “2,4-D
dichlorophenoxyacetate monooxygenase” is an a-ketoglutarate-dependent
dioxygenase. J. Bacteriol. 175:2083-2086.

Fulthorpe, R.R., C. Mcgowan, O.V. Maltseva, W.E. Holben, and J.M. Tiedje.
1995 2,4-Dichlorophenoxyacetic acid-degrading bacteria contain mosaics
of catabolic genes. Appl. Environ. Microbiol. 61:3274-3281

Gause, G. F. 1934. The struggle for existence. Hafner, New York. 1964. Pp163-

Ghosal D. and I.-S.You. 1988. Nucleotide homology and organization of
chlorocatechol oxidation genes of plasmids pJP4 and pAC27. Mol. Gen.
Genet. 211:113-120.

Gisi, D., L. Willi, H. Traber, T. Leisinger and S. Vuilleumier. 1998. Effercts of
bacterial host and dichlromethane dehalogenase on the competitiveness of
methylotrophic bacteria growing with dichloromethane. Appl. Environ.
Microbiol. 64:1 194-1202

Gorlach, K, R. Shingaki, H. Morisaki and T. Hattori. 1994. Construction of eco-
collection of paddy ﬁeld soil bacteria for population analysis. J. Gen.
Appl. Microbiol. 40:509-517.Greer C.W., J. Hawari and R. Samson.
1990, Inﬂuence of environmental factors on 2,4-dichlorophenoxyacetic

acid degradation by Pseudomonas cepacia isolated from peat. Arch.
Microbiol. 154:317-322.

Graham, D. W. , J. A. Chaudhary, R. S. Hanson and R. G. Arnold. 1993. Factors
affecting competition between Type I and Type II methanotrophs in two-
organisms, continuous-ﬂow reactors. Microbial Ecology, 25:1-17

Haggblom, M.M. 1992. Microbial breakdown of halogenated aromatic pesticides
and related compounds. FEMS Microbiology Reviews 103:29-72

Hansen S. R. and S. P. Hubbell. 1980. Single-nutrient microbial competition:
qualitative agreement between experimental and theoretically forecast
outcomes. Science. 207: 1491 -l493.

Herrick J.B., K.G. Stuart-Keil, W.C. Ghiorse and E. L. Madsen. 1997. Natural
horizontal transfer of a naphthalene dioxygenase gene between bacteria
native to a coal tar-contaminated ﬁeld site. Appl. Eviron. Microbiol.

63:2330-2337.
Hu, F. P., and J. M. Young. 1998. Biocidal activity in plant pathogenic

Acidovorax, Burkholderia, Herbaspirilum, Ralstonia and Xanthomonas
spp. J. Applied Microbiol. 84:263-271

208

 

Ka, J.O. ,W.E. Holben, and J .M. Tiedje. 1994a. Analysis of Competition in Soil
among 2,4-dichlorophenoxyacetic acid-degrading bacteria. Appl.
Environ. Microbiol. 60: 1 121-1 128

Ka, JD. and J.M. Tiedje 1994b. Integration and Excision of a 2,4-
dichlorophenoxyacetioc acid-degradative plasmid in Alcaligenes

paradoxus and Evidence of its natural intergeneric transfer. J. BActeriol.
176:5284-5289,

Kasahara, Y. and T. Hattori. 1991. Analysis of bacterial populations in a
grassland soil according to rates of development on solid media. FEMS
Micobiol. Ecol. 86:95-102.

Koh,, S., M.V. McCullar, and DD. Focht. 1997. Biodegradation of 2,4-
dichlorophenol through a distal meta-ﬁssion pathway. Appl Environ.
Microbiol. 63:2054-2057.

Kukor, J.J, R.H. Olsen and J.S. Siak. 1989. Recruitment of a chromosomally
encoded maleylacetate reductase for degradation of 2,4-

dichlorophenoxyacetic acid by plasmid pJP4. J. Bacteriol. 171:3385-
3390.

Lenski, 1992, Relative Fitness: its estimation and its signiﬁcance for
environmental applications of microorganisms in Microbial Ecology (eds.
M. A. Levin, Seideler R.J. and Rogul, M.) McGraw-Hill, Inc. pp] 83-198

Lenski, R.E., S.C. Simpson, and T.T. Nguyen. 1994. Genetic analysis of a
plasmid-encoded host genotype-speciﬁc enhancement of bacterial ﬁtness

Maltseva, O. And P. Oriel. 1997. Monitoring of an alkaline 2,4,6-
trichlorophenol-degrading enrichment culture by DNA ﬁngerprinting

methods and isolation of the responsible organism, haloalkaliphilic
Nocardioides sp. Strain M6. Appl. Environ. Microbiol. 63:4145-4149.

McGowan, C., R. Fulthorpe, A. Wright, and J.M. Tiedje. 1998. Evidence for
interspecies gene transfer in the evolution of 2,4-dichlorophenoxyacetic
acid . Appl. Environ. Microbiol. 64:4089-4092.

Monod, J. 1949. The growth of bacterial cultures. Annu. Rev. Microbiol. 3:371 -
394.

Ravatin R., A.J.B. Zehnder and JR. Van Der Meer. 1998. Low-frequency
horizontaltransfer of an element containing the chlorocatechol degradation
genes from Pseudomonas sp. strain 813 to Pseudomonas putida F1 and to
indigenous bacteria in laboratory-scale activated-sludge microcosms.

Appl. Eviron. Microbiol. 64:2126-2132.

209

 

Rochkind-Dubinsky, M.L., G.S. Sayler, and J. W. Blackburn, Microbiological
Decompostition of Chlorinated Aromatic Compounds. New Yorek,
Marcel Dekker Inc. pp 108-123

Roughgarden, J. 1971. Density-dependent natural selection; Ecology. 52:453-
468.

Schroth, M. N., D. C. Hildebrand and N. Panopoulos. 1992. Phytopathogenic
pseudomonads and related plant-associated pseudomonads. Pp 3104-
3131. In: J. Balows, A, H.G., Triiper, M. Dworkin, W. Harder, K-H
Schleifer (eds.), The Prokaryotes: A Handbook on the Biology of
Bacteria: Ecophysiology, Isolation, Identiﬁcation, Applications. 2n ed.
New York. Springer-Verlag,

Smith, V. 1993, Implications of Resource-Ratio Theory for microbial ecology.
Advances in Microbial Ecolgy. 13: 1-37.

Tilman, D. 1981. Tests of resource competition theory using four species of
Lake Michigan algae. Ecology. 62:802-815.

Top E.M., W.E. Holben and L.J. Fomey (1995) Characterization of diverse 2,4-
Dichlorophenoxyacetic acid-degrading plasmids isolated from soil by
complementation. Appl. Environ. Microbiol. 61:3274-3281

Vallaeys T., F. PerselloCartieaux, N. Rouard, C. Lors, G. Laguerre, and G.
Soulas. 1993. PCR-RFLP analysis of 168 rRNA, tfdA and tde genes
reveals a divresity of 2,4-D degraders in soil aggregates. FEMS
Microbiology Ecology. 24:269-278

Vollmer M. D. K. Stadler-Fritzche and M. Schlbmann. 1993. Conversion of 2-
chloromaleylacetate in Alcaligenes Eutrophus JMP134. Microbiology.

Wright, A. and RH. Olsen. 1994. Self-mobilization and organization of the genes
encoling the toluene metabolic pathway of Pseudomonas mendocina KRl.
Appl. and Environ. Microbiol. 60:235-242.

Wyndham R.C. 1986. Evolved aniline catabolism in Acinetobacter calcoaceticus
during continuous culture of river water. Appl. Envrion. Microbiol.
51:781-789

210

 

CONCLUDING STATEMENT

WHY ARE 2,4-D DEGRADING MICROBES SO DIVERSE?

Microbes with the capacity to utilize 2,4-D as a substrate, are
abundant in many environments, particularly in agricultural and urban areas
where broad-leaf weeds are controlled. Table 1.3 illustrates the presence of
great diversity in 2,4-D degrading microbes. Natural variation also exists in
pathways for degrading 2,4-D, and in the vectors that move them among
microbial populations. What is the advantage of such diversity in bacterial
genes? Why do they persist and continue to diversify, when a particular set
of alleles may offer a ﬁtness advantage to its host? Why do we consistently
ﬁnd many populations carrying out similar activities in the same habitat?

At the root of this bacterial diversity, is environmental variation. The
fact that most soils are composed of numerous abiotic and biotic
components: moisture, rock, sand, silt, decaying organic matter, predators,
which lead to numerous microhabitats. To a microbe with a size of only 1
pm length, a grain of sand is a great boulder. Decaying leaves provide a
source of nutrients to many different microbes, each potentially suited to
one very speciﬁc task in the degradation process. Among plant components
that can be degraded by microbial populations, are the products of lignin
degradation by fungi. The lignin products are many, and include

compounds like chlorinated phenols. To degrade chlorinated phenols, is

211

 

only a step away from the ability to utilize 2,4-D. If the microbe possesses
a means of breaking down 2,4-D to a chlorinated phenol, and already can
degrade lignin-like compounds, it has acquired the ability to utilize a new
source of carbon and energy. In environments where the entire 2,4-D
degradative pathway is favorable, such as in maintained gardens and
locations frequently treated with 2,4-D, the components may become
assembled on mobile genetic elements that can be shared among microbial
populations. In this way, familiar elements and patterns become
recognizable in our analysis of 2,4-D degrading genes and plasmids.

Why is this diversity maintained in nature? Perhaps it is because of
differences in the microhabitats that occur in close proximity to one
another. At different times, under different resource limitations, a greater
diversity of pathways in the environment can lead to either a more complete
degradation of compounds or the maintenance of variants capable of
degrading slightly different compounds with different afﬁnities for each.
Even the plasmids used in this study, pJP4 and pEMT3 differed in the
speciﬁcity of their enzymes toward chlorinated or non-chlorinated
substrates, respectively. Plasmid pJP4 provided greater efficiency toward
chlorinated substrates, while pEMT3 provided more rapid degradation of
non-chlorinated substrates. Depending on the natural ﬂuctuation of the
environment, the variant pathways may both be advantageous at different

times.

212

 

Microhabitats provided by differences in resource concentration, may
be as important as ones created in physical space. The research conducted
in this dissertation examined two very simple habitats; one in which the
substrate concentration was plentiful (batch transfer), and one in which the
substrate was limiting (low ﬂow rate chemostats). In each of the different
habitats, there was a propensity for a different bacterial population to
become dominant. Ralstonia eutropha tended to dominate under conditions
of high 2,4-D concentrations, and the presence of pJP4 was important, as a
faster growth rate was favored. To the contrary, the low ﬂow rate
chemostats, in which the concentration of the resource was maintained at
levels below 1 ppm 2,4-D, Burkholderia cepacia was able to dominate over
the R. eutropha strains. The dominance of B. cepacia strains appeared to
be the result of slightly lower half-saturation constants relative to the R.
eutropha strains. In environments where 2,4-D application is occasional,
but at high concentrations, the pJP4 carrying strains may become more
abundant shortly after application, but after the initial substrate is
diminished, a different physiological parameter becomes more important.
The ability to utilize lower concentrations of 2,4-D (or any other resource)
will lead to an increase in the size of the populations with lower thresholds
for a substrate. Variation in substrate concentrations over time, can
therefore support a greater diversity of bacteria than any constant

environment. Since natural habitats are much more complex than the

213

 

environmental parameters that were tested, the potential niches provided to
microbial populations in nature provide sufﬁcient variation to support a
diverse range of genes, continuously adapting and evolving to suit its
current habitat. Spatially separated micropopulations can begin to
diversify, along their own evolutionary path, thus continuously increasing

the diversity of microbial populations on earth.

214