Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Physiological roles and evolutionary implications of alternative DNA polymerases in Escherichia coli
(USC Thesis Other)
Physiological roles and evolutionary implications of alternative DNA polymerases in Escherichia coli
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
PHYSIOLOGICAL ROLES AND EVOLUTIONARY IMPLICATIONS OF
ALTERNATIVE DNA POLYMERASES IN ESCHERICHIA COLI
by
Christopher Hale Corzett
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MOLECULAR BIOLOGY)
December 2012
Copyright 2012 Christopher Hale Corzett
ii
Epigraph
Thus, from the war of nature, from famine and death, the most exalted object which we
are capable of conceiving, namely, the production of the higher animals, directly follows.
There is grandeur in this view of life, with its several powers, having been originally
breathed into a few forms or into one; and that, whilst this planet has gone cycling on
according to the fixed law of gravity, from so simple a beginning endless forms most
beautiful and most wonderful have been, and are being, evolved.
- Charles Darwin, 1859
Darwin, C. (1859). On the origin of species by means of natural selection, or the
preservation of favoured races in the struggle for life. London. J. Murray. 1
st
ed. p. 490
iii
Dedication
The entirety of this work is dedicated to my beautiful and loving wife Stephanie, without
whom I would have likely never embarked on this wonderfully fulfilling journey.
Throughout the duration of this process you have willfully made innumerable sacrifices
and provided unwavering support. It has been your confidence in my potential that has
enabled and fueled my pursuits. Simply put, you are at once my motivation and my
inspiration.
iv
Acknowledgements
First and foremost, I would like to express my heartfelt gratitude to my mentor and
advisor, Steven Finkel. In deference to Virginia Woolf, Steve has literally and
figuratively provided a bench of one’s own. Throughout my graduate education I have
consistently had the resources, space, and freedom to envision and implement the whims
of my curiosity. While often providing sage and practical advice to reign-in my
imagination, he has never once impeded the pursuit of any inquiries or experiments I
might have conceived. With an uncanny ability to provide inspiration and enthusiasm
when needed most while simultaneously raising expectations for productivity, Steve has
enabled me to achieve far more than I might have previously thought possible. Practically
speaking, he has fulfilled my greatest hopes for a scientific advisor.
On a more personal level, Steve has been an ideal mentor that has fostered my ongoing
development as a scientist and in life. Since our first interactions, his infectious
enthusiasm for science and his students has cemented his standing as a personal role
model. While a single anecdote does not serve justice, the fact that Steve allowed and
encouraged me, as a first-year student in his lab, to present the work I initiated during my
rotation at a national conference speaks volumes about his character. While routine for
Steve, acts like these illustrate his unwavering support and confidence in his students and
their capabilities. While instilling these values in me, he has conveyed valuable insights
and perspectives that will not soon be forgotten.
v
I would also like to express my sincerest gratitude to the remaining members of my
dissertation committee: Myron Goodman, Steve Goodman, and Ken Nealson. Each of
these individuals has contributed to my education in a profound manner. Be it through
donated reagents, shared lab space, probing questions, constructive feedback, sage
wisdom or sound advice, I am indebted to each of you for your assistance throughout my
graduate career. I sincerely hope we can continue to maintain ongoing and fruitful
relationships in the future.
Next, I would like to acknowledge my parents Doug and Michele. It goes without saying
that the person I am today intimately reflects their influences, and words fail when
attempting to convey all that they have given to me. They have both instilled in me the
value of hard work, dedication and personal conviction, and this work is but one
manifestation of their shared sacrifices. Despite learning early how some view biology as
merely a bunch of “hand waving” and warnings that it certainly doesn’t pay well, the
moonlit strolls, camping expeditions and tide-pooling adventures have engendered my
fascination with nature. Furthermore, our occasionally heated discussions and, at times,
endless allusions to the devil’s advocate have shown me the importance of considering
alternative hypotheses and to substantiate my beliefs; lessons that continue to serve me in
the laboratory and in life.
Of course, I would be remiss if I did not acknowledge the contributions of my fellow lab
mates, and offer some insight into the wonderful experiences I have had with each of
vi
you. Perhaps my gratitude for each of these individuals can be best summarized and
expressed within the context of the lab culture. Throughout the long-term duration, the
overall population of the lab has remained relatively stable to the untrained eye.
However, we have actually remained extremely dynamic as novel individuals appear,
each with their own interesting capabilities and beneficial traits. While admittedly
stressful at times, as we sometimes compete for space and resources, there is good reason
to believe that those who survive and thrive do so by expressing and capitalizing on what
makes them unique and better suited for a particular role within this culture. While these
selective pressures can seem suffocating, you have each been like a gasp of fresh air as
you progressively swept into my life. With the profound influence you each have had on
my experience, I sincerely hope that I too might have influenced you all in at least some
microscopic way. Furthermore, I am eager to see how you all perform as you take over
and influence the trajectory of the lab culture as it continues to evolve.
Finally, I would also like to acknowledge the role of the greater Molecular and
Computational Biology department and University at-large. I can honestly attest to the
fact that I have been a happy and fulfilled graduate student throughout my time at the
University of Southern California, and this feat is due in no small part to the people
surrounding me. A variation of a common phrase might suggest it takes a department to
raise a graduate student, and the faculty, postdocs, fellow graduate students,
undergraduates and administrators have all had a lasting and positive impact on the
amazing experience I have had at USC.
vii
Table of Contents
Epigraph ii
Dedication iii
Acknowledgements iv
List of Tables ix
List of Figures x
Abstract xiv
Chapter 1 Introduction 1
Chapter 2 Fitness during feast and famine: How alternative DNA 19
polymerases II, IV and V each influence physiology and
evolution in Escherichia coli
2.1 Abstract 19
2.2 Introduction 20
2.3 Materials and methods 23
2.4 Results 30
2.5 Discussion 52
Chapter 3 DNA polymerase II confers a physiological fitness advantage 59
within continuous culture chemostat competitions
3.1 Abstract 59
3.2 Introduction 60
3.3 Materials and methods 63
3.4 Results 65
3.5 Discussion 77
Chapter 4 Stationary phase conditions induce expression of alternative 84
DNA polymerases in the absence of exogenous stressors
4.1 Abstract 84
4.2 Introduction 85
4.3 Materials and methods 88
4.4 Results 89
4.5 Discussion 101
viii
Chapter 5 Conclusions and discussion 107
References 117
Appendix A The effect of serial passage and batch culture aging on fitness 127
during long-term stationary phase and the GASP phenotype
A.1 Abstract 127
A.2 Introduction 128
A.3 Materials and methods 129
A.4 Results 132
A.5 Discussion 148
Appendix B Characterization of the Escherichia coli death phase 153
B.1 Abstract 153
B.2 Introduction 153
B.3 Materials and methods 156
B.4 Results 156
B.5 Discussion 164
Appendix C Applications of whole-genome re-sequencing for the 168
characterization of genetic diversity generated during
long-term stationary phase
C.1 Materials and methods 169
C.2 Identification of genomic changes following prolonged 172
incubation in long-term stationary phase using whole-genome
re-sequencing
C.3 The characterization of genetic diversity within aged alternative 184
DNA polymerase deficient strains using whole-genome
re-sequencing
ix
List of Tables
Table 2.1 Strains used in chapter 2. 24
Table 2.2 qRT-PCR primers used in chapter 2. 29
Table 2.3 Rif
R
single-nucleotide polymorphisms identified. 45
Table 2.4 Rif
R
deletions and amplifications identified. 47
Table 2.5 Strain-specific mutation spectrum. 47
Table 3.1 Strains use in chapter 3. 63
Table A.1 Summary of co-outgrowth competitions. 144
Table A.2 Summary of stationary phase competitions. 144
Table C.1 Characterization of large-scale amplifications and deletions 180
observed in aged colony morphotypes.
x
List of Figures
Figure 1.1 Y-family DNA polymerase homology. 4
Figure 1.2 The process of translesion synthesis. 7
Figure 1.3 The SOS response induces alternative DNA polymerase 9
expression.
Figure 1.4 The five phases of the bacterial life cycle. 14
Figure 1.5 The GASP phenotype. 16
Figure 1.6 Population dynamics during long-term stationary phase. 17
Figure 2.1 Outgrowth of polymerase-deficient strains. 31
Figure 2.2 Co-outgrowth and stationary phase competitions between 32
double-mutant strains.
Figure 2.3 GASP competitions against wildtype. 34
Figure 2.4 GASP competitions against double-mutant strains. 35
Figure 2.5 Increased relative fitness of Pol II
+
cells following 37
serial passage.
Figure 2.6 Pol II
+
serial passage relative fitness summary. 38
Figure 2.7 Stationary phase competitions following serial passage. 39
Figure 2.8 Pol II
+
relative fitness during chemostat competitions. 40
Figure 2.9 Fluctuations in Pol II
+
relative fitness during chemostat 41
competitions.
Figure 2.10 Strain-specific Rif
R
mutation frequency. 43
Figure 2.11 Strain-specific Rif
R
mutation spectrum. 48
Figure 2.12 Alternative polymerase transcript abundance increases 50
over the cell cycle.
xi
Figure 2.13 Alternative polymerase transcript proportion increases 51
over the cell cycle.
Figure 3.1 Initial Pol II
+
advantage during serial passage competitions. 66
Figure 3.2 Pol II
+
relative fitness during chemostat competitions. 68
Figure 3.3 Pol II
+
relative fitness according to overall population 69
density during chemostat competitions.
Figure 3.4 Pol II
+
relative fitness according to dissolved oxygen 70
content during chemostat competitions.
Figure 3.5 Fluctuations in Pol II
+
relative fitness during chemostat 72
competition #1.
Figure 3.6 Fluctuations in Pol II
+
relative fitness during chemostat 73
competition #2.
Figure 3.7 Fluctuations in Pol II
+
relative fitness during chemostat 74
competition #3.
Figure 3.8 Fluctuations in Pol II
+
relative fitness during chemostat 75
competition #4.
Figure 3.9 Fluctuations in Pol II
+
relative fitness during chemostat 76
competition #5.
Figure 4.1 Alternative DNA polymerases are induced upon entering 90
stationary phase.
Figure 4.2 Entry into stationary phase signaled by dps and fis. 92
Figure 4.3 Alternative polymerase transcript levels remain elevated 93
throughout long-term stationary phase.
Figure 4.4 Alternative polymerase transcript proportion remains 95
elevated throughout long-term stationary phase.
Figure 4.5 Pol II transcript increases proportionally greater and faster 96
than other alternative DNA polymerases.
Figure 4.6 Mitomycin C dose response curves. 97
xii
Figure 4.7 Effect of stationary phase and mitomycin C treatment on 99
gene expression.
Figure 4.8 Gene induction following mitomycin C treatment versus 100
stationary phase.
Figure 5.1 Fitness during feast and famine 108
Figure 5.2 The modulation of mutation 113
Figure A.1 Co-outgrowth and stationary phase competitions between 133
unaged and batch (+5) populations.
Figure A.2 Co-outgrowth and stationary phase competitions between 134
unaged and batch (+10) populations.
Figure A.3 Co-outgrowth and stationary phase competitions between 135
unaged and serial passage (+5) populations.
Figure A.4 Co-outgrowth and stationary phase competitions between 136
unaged and serial passage (+10) populations.
Figure A.5 Co-outgrowth and stationary phase competitions between 137
batch (+5) and batch (+10) populations.
Figure A.6 Co-outgrowth and stationary phase competitions between 138
batch (+5) and serial passage (+5) populations.
Figure A.7 Co-outgrowth and stationary phase competitions between 139
batch (+5) and serial passage (+10) populations.
Figure A.8 Co-outgrowth and stationary phase competitions between 140
batch (+10) and serial passage (+5) populations.
Figure A.9 Co-outgrowth and stationary phase competitions between 141
batch (+10) and serial passage (+10) populations.
Figure A.10 Co-outgrowth and stationary phase competitions between 142
serial passage (+5) and serial passage (+10) populations.
Figure A.11 Summary of co-outgrowth and stationary phase competitions 143
following aging treatment against unaged populations.
Figure A.12 Effect of aging regimen on GASP phenotype . 145
xiii
Figure B.1 Escherichia coli long-term growth curve (Trial #1). 158
Figure B.2 Escherichia coli long-term growth curve in 159
dilute LB (Trial #2).
Figure B.3 Timing of the Escherichia coli death phase. 161
Figure B.4 Modeling the Escherichia coli death phase. 163
Figure C.1 Analysis of chromosomal copy number variation in 175
Wildtype 1 and Wildtype 2.
Figure C.2 Analysis of chromosomal copy number variation in 176
colonies 150-c4 and 150-c5.
Figure C.3 Analysis of chromosomal copy number variation in 177
colonies 150-O and 180-E.
Figure C.4 Analysis of chromosomal copy number variation in 178
colonies 240-M and 750-MO.
Figure C.5 Analysis of chromosomal copy number variation in 179
colonies 840-Sm-O and 900-Sm-2.
Figure C.6 Analysis of chromosomal copy number variation in aged 186
populations of Wildtype and Pol II
-
.
Figure C.7 Analysis of chromosomal copy number variation in aged 187
populations of Pol IV
-
and Pol V
-
.
Figure C.8 Analysis of chromosomal copy number variation in aged 188
populations of Pol II
+
and Pol IV
+
.
Figure C.9 Analysis of chromosomal copy number variation in aged 189
populations of Pol V
+
and Pol -/-/-.
xiv
Abstract
Escherichia coli DNA polymerases II, IV and V serve dual roles within cells by
facilitating efficient replication past potentially lethal DNA damage while simultaneously
introducing genetic variation that can promote adaptation and evolution within stressful
environments. While long recognized to be important for these physiological and
evolutionary roles, the specific molecular mechanisms and relative contributions
attributable to each of these alternative DNA polymerases within natural environmental
conditions has remained elusive. Using a series of alternative polymerase-deficient
strains analyzed during conditions of both feast and famine, we establish distinct
hierarchies of polymerase activity. Pol II confers a significant physiological advantage by
facilitating efficient replication and creating genetic diversity during periods of rapid
growth, whereas Pol IV and Pol V make the largest contributions to evolutionary fitness
during long-term stationary phase. Pol V is responsible for maximizing allelic diversity,
yet Pol IV is the single greatest determinant of mutation frequency. Furthermore, we
demonstrate that these alternative polymerases, along with additional members of the
SOS regulon, are induced as cells transition from exponential to stationary phase growth
in the absence of exogenous stress-stimulated SOS induction, and that they remain
elevated throughout long-term stationary phase. These findings reveal each alternative
DNA polymerase is vital to physiological and evolutionary fitness under dynamic and
unpredictable conditions akin to those experienced in nature, and indicate their
contributions to replication and adaptation within microbial communities are greater than
previously appreciated.
1
Chapter 1: Introduction
Biological success over extended timescales requires an exquisite balance between
maintaining physiological fitness (the capacity to survive and produce offspring) and
evolutionary fitness (the capacity to adapt to and thrive within changing environments).
Accordingly, cellular processes facilitating growth and the formation of genetic diversity
directly influence both physiological and evolutionary fitness, respectively. DNA
polymerases are uniquely situated at the intersection of both of these selective forces:
required for efficient replication, and responsible for generating mutations. Accordingly,
elucidating the mechanism and extent to which DNA polymerases introduce genetic
variation during replication remains critical to understanding the long-term survival and
evolution of bacterial populations.
1.1 The DNA polymerases of Escherichia coli
Escherichia coli encodes five DNA polymerases (Friedberg, 2006; Johnson and
O'Donnell, 2005). The “housekeeping” high-fidelity DNA polymerase III (Pol III)
performs the majority of leading-strand DNA synthesis during replication under
vegetative conditions, while DNA Polymerase I (Pol I) contributes principally to the
maturation of Okazaki fragments during lagging-strand synthesis (Friedberg, 2006;
Kornberg and Baker, 1992). Three alternative DNA polymerases (Pol II, Pol IV and Pol
V) can be induced under a variety of environmental stresses (Layton and Foster, 2003,
2005; Stumpf and Foster, 2005; Yeiser et al., 2002) and are characterized most
2
extensively following induction of the SOS regulon in response to DNA damage, leading
them to be commonly referred to as SOS-induced polymerases (Courcelle et al., 2001;
Friedberg, 2006; Nohmi, 2006; Yang and Woodgate, 2007).
First characterized in 1970 (De Lucia and Cairns, 1969; Knippers, 1970), DNA
polymerase II (encoded by polB) was subsequently rediscovered and found to play an
important role during replication following DNA damage (Bonner et al., 1990; Bonner et
al., 1988; Iwasaki et al., 1990). A B-family polymerase (Bonner et al., 1990; Braithwaite
and Ito, 1993), Pol II is capable of 3’-exonuclease proofreading (Cai et al., 1995a) and
replicating undamaged DNA with considerable accuracy (Cai et al., 1995b). Pol II has
also been implicated in the process of replication during stress, particularly for its role
initiating DNA synthesis following UV exposure in a process known as replication restart
(Banach-Orlowska et al., 2005; Pham et al., 2001; Rangarajan et al., 1999, 2002).
DNA polymerases IV (dinB) and V (umuDC) are both Y-family polymerases
(Goodman,
2002; Nohmi, 2006) that replicate DNA with significantly lower fidelity (Fuchs et al.,
2004; Jarosz et al., 2007; Tang et al., 2000). Pol IV was first characterized in 1999
(Wagner et al., 1999) and has since been implicated in the formation of small frameshift
mutations and the generation of “adaptive mutations” using plate-based mutation assays
(Foster, 2000, 2007; Galhardo et al., 2009; McKenzie et al., 2001; Roth et al., 2006).
Around the same time, Pol V was also recognized as an error-prone DNA polymerase
important for stress-induced mutagenesis (Reuven et al., 1999; Tang et al., 1998; Tang et
3
al., 1999), though first characterized for its role in replication following DNA damage in
1992 (Rajagopalan et al., 1992). Homologs of these alternative polymerases are found in
all three domains of life (Figure 1.1) (Ohmori et al., 2001). Humans possess no fewer
than ten of these specialized alternative polymerases (Hubscher et al., 2002), and their
actions have been implicated in a variety of diseases and human health (Lange et al.,
2011; Robbins et al., 1975; Robbins et al., 1974; Stallons and McGregor, 2010; Xie et al.,
2010).
In E. coli these three alternative DNA polymerases are distinguished by having a more
open conformation within their active site that enables them to replicate past bulky
template lesions, albeit at a significantly lower fidelity. Accordingly, these enzymes are
capable of performing a vital physiological role by mediating translesion synthesis (TLS),
enabling DNA synthesis past damage that would otherwise halt replication (Bichara et
al., 2011; Fuchs et al., 2004; Goodman, 2002; Nohmi, 2006; Tippin et al., 2004).
4
Figure 1.1. Y-family DNA polymerase homology.
An unrooted phylogenetic tree of Y-family polymerases based upon protein homology.
Different branches are shaded in different colors. The abbreviations used in the figure are
as follows: Bacteria: Ban, Bacillus anthracis; Bha, Bacillus halodurans; Bsu, Bacillus
subtilis; Cfr, Citrobacter freundii; Eco, Escherichia coli; Efa, Enterococcus faecalis; Lla,
Lactococcus lactis; Mge, Mycoplasma genitalium; Mmo, Morganella morganii; Mpn,
Mycoplasma pneumoniae; Mtu, Mycobacterium tuberculosis; Nme, Neisseria
meningitides; Pae, Pseudomonas aeruginosa; Pmu, Pasteurella multocida; Psy,
Pseudomonas syringae; Sco, Streptomyces coelicolor; Sen, Salmonella enterica; Sfl,
Shigella flexneri; Sma, Serratia marcescens; Spn, Streptococcus pneumoniae; Sth,
Salmonella typhi; Sty, Salmonella typhimurium; Uur, Ureaplasma urealyticum; Vco,
Vibrio cholerae. Archaea: Hal, Halobacterium sp. NRC-11; Sso, Sulfolobus solfataricus.
Eukaryotes: At, Arabidopsis thaliana; Ce, Caenorhabditis elegans; Dm, Drosophila
melanogoster; Hs, Homo sapiens; Lm, Leishmania major; Mm, Mus musculus; Sc,
Saccharomyces cerevisiae; Sp, Schizosaccharomyces pombe. (From: Ohmori et al., 2001)
5
1.2 Translesion synthesis
During DNA synthesis, as the replication machinery progresses, the high-fidelity DNA
polymerase III will occasionally encounter sites of DNA damage with altered chemical
composition that its relatively constricted and highly-specific active site cannot replicate
past. This polymerase stalling triggers a cascade of events known as translesion synthesis,
where an alternative DNA polymerase can take over and enable cells to copy past DNA
damage to continue replication (Figure 1.2).
First, the arrested replication machinery creates regions of single-stranded DNA as the
helicase DnaB continues to unwind DNA ahead of the replication fork. RecA protein then
binds this single-stranded DNA forming nucleoprotein filaments that trigger the SOS
response and induce the expression of alternative DNA polymerases, detailed below.
Eventually, the stalled Pol III holoenzyme falls off the template DNA, allowing access
for one of the alternative DNA polymerases to take its place on the template primer and
resume synthesis past the site of DNA damage. This polymerase switch is facilitated by
the sliding β-clamp, a processivity factor that encircles DNA (mediated by the clamp
loader γ-complex) and remains stalled at sites of DNA damage. Comprised of a
homodimer that can simultaneously bind at least two polymerases, an increasing body of
evidence suggests this β-clamp effectively serves as a “toolbelt”, capable of tethering
alternative polymerases to the replication fork so they can efficiently switch places to
resolve template lesions when necessary (Heltzel et al., 2009a; Heltzel et al., 2009b;
Indiani et al., 2005). Due to their comparatively low processivity, the alternative
6
polymerase quickly falls off from the replication machinery, facilitating another
polymerase switch so Pol III can resume high-fidelity synthesis. This alternative
polymerase-mediated translesion synthesis provides a vital function by facilitating the
resumption of replication, which could otherwise prevent cell division.
While mediating the vital physiological role of translesion synthesis, these comparatively
error-prone alternative DNA polymerases frequently introduce mutations and can
generate a significant amount of genetic variation that has the potential to drastically
affect the evolutionary fitness of bacterial populations. In fact, the mutagenic
contributions of each alternative DNA polymerase have been implicated in the formation
of antibiotic resistance within clinically relevant infections of animal models (Cirz et al.,
2005). However, given their capacity to introduce potentially deleterious mutations, it is
perhaps not surprising that these alternative polymerases are also highly regulated. While
all three alternative polymerases are induced under a variety of environmental stresses,
including long-term stationary phase (Yeiser et al., 2002), they have been primarily
characterized following their induction by the SOS response.
7
Figure 1.2. The process of translesion synthesis.
Escherichia coli translesion synthesis is proposed to operate in a sequential step-wise
process, based upon in vitro experiments. (1) When the replication machinery (shown
here as only the dimeric β-clamp, γ-complex, and DNA polymerase (Pol) III
holoenzyme) perform DNA synthesis on leading-strand DNA it will occasionally (2)
encounter and stall at sites of DNA damage, and ssDNA accumulates as the DnaE
helicase continues to unwind dsDNA. (3) RecA protein then coats downstream ssDNA as
a nucleoprotein filament that can induce the SOS response and expression of alternative
DNA polymerases. While stalled, Pol III is eventually released from the replication
machinery and an alternative DNA polymerase, here Pol V, can gain access to the primer
template. (4) This alternative polymerase can facilitate replication past the lesion, albeit
with reduced fidelity and the incorporation of mutations in newly synthesized DNA. (5)
Once past the lesion, the alternative polymerase is released and Pol III can again gain
access, and (6) continue high-fidelity replication.
8
1.3 The SOS response
First characterized following UV-induced DNA damage (Radman, 1974; Radman, 2007),
the E. coli SOS response is regulated by a two-component repressor and activator system
of LexA and RecA proteins (Figure 1.3). The LexA repressor binds to a 20 base pair
consensus sequence in the operator region of SOS genes, normally suppressing their
expression. As alluded to earlier, the SOS response is initiated by the presence of single-
stranded DNA (ssDNA), which can arise when housekeeping DNA polymerases stall at
template lesions. This ssDNA is then bound by RecA protein, forming active RecA*
nucleoprotein filaments. These RecA nucleoprotein filaments serve as coproteases that
facilitate the autocatalytic cleavage of LexA repressor, leading to the derepression of
approximately 40 genes as part of the SOS response (Courcelle et al., 2001). These genes
include all three alternative DNA polymerases, as well as both recA and lexA. This
induction serves as a feedback loop increasing both regulators, enabling LexA pools to
rise following the resolution of DNA damage and repress further induction.
9
Figure 1.3. The SOS response induces alternative DNA polymerase expression.
LexA protein serves as a repressor of over 40 genes within the SOS regulon, binding to
operator regions and preventing transcription. As DNA damage causes regions of single-
stranded DNA to form ahead of stalled replication forks, the ensuing formation of RecA*
nucleoprotein filament can promote the autocatalytic cleavage of LexA protein. This
derepression induces the expression of SOS response genes, including polB, dinB, and
umuDC which encode alternative DNA polymerases. recA and polB are induced quickly,
whereas umuDC expression increases later. Furthermore, UmuD protein must undergo
RecA*-mediated cleavage to produce UmuD’ before combining with UmuC to form the
UmuD’
2
C-complex known as DNA Polymerase V. DNA Polymerase V is then activated
when a molecule of RecA and ATP (red triangle) is transferred from the DNA 3’-end of
RecA* to form the complex UmuD’
2
C-RecA-ATP. (Modified from Tippin et al., 2004
and Jiang et al., 2009)
RecA*
RecA
recA
LexA
polB dinB umuDC
Pol II
Pol IV
UmuC UmuD
UmuD’
Pol V
LexA
< 1 min < 1 min ~ 45 min
DNA
Damage
Active Pol V
RecA*-ATP
10
The basal expression and induction of genes responsive to the SOS response is contingent
on the sequence specificity of the consensus LexA binding sequence (5’-
TACTG(TA)
5
CAGTA-3’) in the operator of each gene (Fernandez De Henestrosa et al.,
2000; Walker, 1984). Depending on the affinity of the LexA repressor to the consensus
recognition sequence, specific SOS genes can experience varying extents of repression
that can produce a temporal cascade of gene induction. Following induction of the SOS
response by UV irradiation during logarithmic growth, polB (Pol II) is among of the
earliest genes induced (<1 minute after SOS induction), wheras umuD and umuC (Pol V,
~45 minutes) are among the last (Courcelle et al., 2001; Kenyon and Walker, 1980; Kim
et al., 2001; Qiu and Goodman, 1997; Woodgate and Ennis, 1991). DNA polymerase V is
also subject to an additional level of regulation as UmuD protein must undergo RecA*-
mediated autocleavage (similar to LexA autocleavage) before two shorter, mutagenically
active UmuD’ molecules can combine with one UmuC molecule to form a heterotrimeric
UmuD’
2
C complex, known as error-prone DNA polymerase V (Tippin et al., 2004).
While Pol V (UmuD’2C) is barely active in the absence of RecA*, a molecule of RecA
and ATP can be transferred from the DNA 3’-end of RecA* and form an activated Pol V
complex (UmuD’
2
C-RecA-ATP) that can catalyze TLS in the absence of RecA* (Jiang et
al., 2009; Patel et al., 2010).
In this manner, the induction of alternative DNA polymerases serves to resolve template
lesions and allow replication to continue, thus preventing the formation of single-stranded
DNA and relieving the trigger of SOS induction. However, while serving this vital
11
physiological role, these enzymes are also capable of introducing significant genetic
diversity that can have important ramifications for the evolution of microbial populations
in nature.
1.4 Alternative DNA polymerases in nature
Although it has been known since the 1970s that UV irradiation and various exogenous
chemical agents interfering with replication or inducing DNA damage could elicit the E.
coli SOS response (Radman, 1975; Radman, 2007), comparatively little attention has
focused on the role of this stress response within natural populations in the absence of
exogenous stressors. Instead, most in vivo studies have focused on the roles of Pols II, IV
and V either in the survival of rapidly dividing cells in the presence of DNA damage, or
in a plate-based assay system studying the process of “adaptive mutation” (Foster, 2005,
2007; Galhardo et al., 2007; Roth et al., 2006). This is despite the fact that patterns of
sequence variability within genes coding for alternative polymerases from natural isolates
suggest they exhibit levels of variation similar to those observed in housekeeping Pol I
and Pol III (Bjedov et al., 2003a). These findings are indicative of a strong selective
pressure maintaining their function and specificity within natural environments, and
suggest they play vital roles within natural populations.
It has been proposed that in nature these alternative polymerases play an important role in
generating genetic diversity that facilitates adaptation and survival during environmental
and nutritional stress. Although most mutations are detrimental or neutral, within large
12
microbial populations these deleterious effects can be outweighed by the generation of
rare beneficial alleles. Therefore, within dynamic and unpredictable habitats higher
mutation frequencies can be advantageous, and adopting a strategy of “mutate or die” can
become evolutionarily favorable. In this manner, the error-prone replication of alternative
DNA polymerases is often referred to as stress-induced mutagenesis, and offers a
molecular mechanism of temporarily introducing genetic variation during unfavorable
conditions that might serve to ameliorate strong selective forces.
1.5 Long-term stationary phase
To investigate the roles and relative contributions of alternative DNA polymerases under
conditions more akin to those experienced by microbial communities in nature, we
sought to characterize their physiological roles and evolutionary implications throughout
the five phases of the bacterial life cycle (Finkel, 2006), as microbial populations
encounter nutritional and environmental stress during fierce competition for limited
resources. During long-term incubation in rich media, E. coli can exhibit five distinct
phases of growth (Figure 1.4). When bacteria are first inoculated in rich medium at low
population density, cells exhibit a lag phase before growing as they retool their
physiology and initiate synthesis of cellular components required for division. Soon
thereafter, cells enter what is called exponential or logarithmic phase growth,
characterized by rapid cell growth with populations doubling at a constant rate. This rapid
growth will continue until nutrients are depleted, toxic byproducts of metabolism
accumulate in sufficient concentrations to inhibit further growth, or cell-to-cell signaling
13
induces physiologic changes. Once growth stops and the overall population density no
longer increases, cells enter stationary phase. Eventually, typically on the order of 3 days
in batch culture, cells begin to die and population density exhibits a significant decline
known as death phase, often with 99% or more of cells losing viability. Following death
phase, E. coli can enter what is known as long-term stationary phase where relatively
stable population densities can be maintained for extended periods (years) without
providing additional nutrients (Finkel, 2006; Finkel and Kolter, 1999). While the overall
population density remains nearly flat as the rate of cell growth approximates the rate of
cell death, this stability belies the highly dynamic and evolving population structure of
these cultures as demonstrated by the Growth Advantage in Stationary Phase (GASP)
phenotype.
14
Figure 1.4. The five phases of the bacterial life cycle.
Following the introduction of bacteria at low density into fresh LB medium there is an
initial (1) lag phase as cells retool their physiology and prepare for growth, followed by
rapid (2) exponential-phase growth. Soon thereafter, as nutrients are depleted and toxic
byproducts accumulate, growth is inhibited and cells enter (3) stationary phase. After 2-3
days, cultures enter (4) death phase where approximately 99% of cells lose viability
before entering (5) long-term stationary phase as relatively stable population density is
maintained for extensive periods of time. By day 10, populations of bacteria isolated
from these cultures are capable of expressing the Growth Advantage in Stationary Phase
(GASP) phenotype. (Modified from Finkel, 2006)
2
3
4
5
1
15
1.6 The GASP phenotype
The Growth Advantage in Stationary Phase (GASP) phenotype can be defined as the
ability of cells aged in long-term batch culture to outcompete cells from younger cultures.
This phenomenon can be demonstrated through competitions initiated by mixing a
subpopulation of cells obtained from a 10-day culture as a 1:1000 minority against a
dense overnight culture of an unaged parental strain (Figure 1.5). By monitoring the
population density of each competing strain using chromosomally encoded antibiotic
resistance markers, within a matter of days one can observe the “aged” 10-day-old cells
outcompeting the unaged 1-day-old cells, often driving them below the limit of detection.
This competitive advantage is genetically encoded, and not due to a physiological
adaptation as evidenced by the capacity to serial passage aged cells through repeated
outgrowth and still observe the GASP phenomenon. To date, previously characterized
GASP alleles have either altered the activity of the general stress response sigma factor
RpoS, or improved the capacity to scavenge and catabolize amino acids. Additional
findings indicate the formation of novel beneficial alleles continues throughout long-term
stationary phase, such that multiple waves of GASP occur as different subpopulations
compete for limited nutrients during this dynamic growth phase (Figure 1.6).
16
Figure 1.5. The GASP phenotype.
When “aged” wildtype cells obtained from 10-day-old, long-term stationary phase
cultures (Black lines, filled markers) are inoculated as a 1:1000 (vol:vol) minority
population into a dense overnight culture of unaged wildtype cells (Gray lines, open
marker), the 10-day-old population can quickly increase in frequency and often drive the
1-day-old population to extinction. Three representative pair-wise competitions are
presented, denoted by marker shape.
3
5
7
9
11
0
2
4
6
8
10
12
14
log
10
CFU/mL
Day
*
*
*
17
Figure 1.6. Population dynamics during long-term stationary phase.
Although the overall population density remains relatively stable as cells incubate in
long-term stationary phase, the underlying population structure is highly dynamic. This
stable overall density is the result of a balanced birth rate and death rate, creating a
dynamic equilibrium of newly created GASP mutants growing while less competitive
cells die. The black like reflects overall population density, while each colored line
represents a different GASP mutant that appears during long-term incubation. (From
Finkel, 2006)
Ultimately, long-term culture and the GASP phenotype can serve as experimental tools
capable of demonstrating the formation and selection of novel beneficial alleles from
populations thriving within dynamic and stressful conditions. Accordingly, understanding
the molecular mechanisms and dynamics of survival and the generation of genetic
diversity within long-term stationary phase can inform our understanding of evolutionary
processes at play within natural environments.
18
1.7 Physiological roles and evolutionary implications of alternative
DNA polymerases
Preliminary findings indicated each alternative DNA polymerase was important for
survival and evolutionary fitness within long-term competitions (Yeiser et al., 2002),
likely through their contributions to genetic diversity; however, the mechanisms and
relative contributions of DNA Polymerases II, IV and V remained ambiguous. Here,
upon initiating a series of experiments designed to differentiate and elucidate the cellular
functions and genetic byproducts attributable to each alternative polymerase, we have
discovered specific physiological roles and evolutionary implications attributable to each
alternative DNA polymerase. These findings not only serve to inform our mechanistic
understanding of their relative contributions to cellular fitness throughout the bacterial
lifecycle, but fundamentally influence our understanding of the molecular processes
affecting the adaptation and evolution of microbial communities.
19
Chapter 2: Fitness during feast and famine: How alternative DNA
polymerases II, IV and V each influence physiology and evolution in
Escherichia coli
The content of this chapter appears as submitted in Corzett, C.H., Goodman, M.F.,
Finkel S.E., “Fitness during feast and famine: How alternative DNA polymerases
influence physiology and evolution in Escherichia coli”, manuscript submitted.
2.1 Abstract
Escherichia coli DNA polymerases II, IV and V generate genetic diversity during periods
of feast and famine, but it is not known when or how each enzyme acts. Here we show
that these alternative polymerases, with additional members of the SOS regulon, are
induced as cells transition from exponential to stationary phase growth in the absence of
exogenous stress-stimulated SOS induction. We establish distinct hierarchies of
polymerase activity. Pol II confers a significant physiological advantage by facilitating
efficient replication and creating genetic diversity during periods of rapid growth (II >V
>IV). Pol IV and Pol V make the largest contributions to evolutionary fitness during
long-term stationary phase. Pol V is responsible for maximizing allelic diversity
(V>IV>II) and Pol IV is the greatest determinant of mutation frequency (IV>V>II).
Therefore, to attain optimal fitness and undergo adaptive evolution bacterial populations
require robust expression of all three alternative DNA polymerases during exponential
and stationary phase.
20
2.2 Introduction
The generation of genetic diversity directly impacts the evolutionary fitness of a
population, yet little is known about the molecular mechanisms generating allelic
variation in complex bacterial communities, including batch cultures (Bjedov et al.,
2003b; Chao and Cox, 1983; Pigliucci, 2008; Saint-Ruf and Matic, 2006; Woods et al.,
2011). Given that the majority of mutations are introduced during replication,
perturbations in the fidelity of replication can have dramatic consequences on the
evolutionary success of populations
(Foster, 2007; Galhardo et al., 2007; Kunkel, 2004;
Yeiser et al., 2002). Characterization of the mechanisms and extent to which DNA
polymerases introduce genetic variation is critical to understanding the physiology and
evolution of bacteria.
Escherichia coli encodes five DNA polymerases (Friedberg, 2006; Johnson and
O'Donnell, 2005). High-fidelity DNA polymerase III performs the majority of DNA
replication under vegetative conditions, with Pol I contributing principally to maturation
of Okazaki fragments (Friedberg, 2006; Kornberg and Baker, 1992). Three alternative
DNA polymerases (Pol II, Pol IV and Pol V) can be induced under a variety of
environmental stresses (Layton and Foster, 2003, 2005; Stumpf and Foster, 2005; Yeiser
et al., 2002) and are characterized most extensively following induction of the SOS
regulon in response to DNA damage, leading them to be referred to as SOS-induced
polymerases (Courcelle et al., 2001; Friedberg, 2006; Nohmi, 2006; Yang and Woodgate,
2007). They perform a vital physiological role by mediating translesion synthesis (TLS),
21
enabling efficient replication past DNA damage that would otherwise halt replication
(Bichara et al., 2011; Fuchs et al., 2004; Goodman, 2002; Nohmi, 2006; Tippin et al.,
2004).
DNA polymerase II (encoded by polB) is a B-family polymerase (Banach-Orlowska et
al., 2005; Pham et al., 2001; Rangarajan et al., 1999, 2002) capable of 3’-exonuclease
proofreading enabling it to replicate undamaged DNA with considerable accuracy (Cai et
al., 1995b). DNA polymerases IV (dinB) and V (umuDC) are Y-family polymerases
(Goodman, 2002; Nohmi, 2006) that replicate DNA with relatively lower fidelity (Fuchs
et al., 2004; Jarosz et al., 2007; Tang et al., 2000). Homologs of these alternative
polymerases are found in all three domains of life (Ohmori et al., 2001), and have been
implicated in a variety of human diseases (Lange et al., 2011; Robbins et al., 1975;
Robbins et al., 1974; Stallons and McGregor, 2010; Xie et al., 2010).
Most in vivo studies have focused on the roles of Pols II, IV and V either in the survival
of rapidly dividing cells in the presence of DNA damage sufficient to activate the E. coli
SOS response, or in a plate-based assay system studying the process of “adaptive
mutation” (Foster, 2005, 2007; Galhardo et al., 2007; Roth et al., 2006). We have
previously found (Yeiser et al., 2002) that in the absence of exogenous DNA damage,
Pols II, IV and V provide a substantial fitness advantage to wild type cells during long-
term stationary phase competition, including expression of the GASP (Growth Advantage
in Stationary Phase) phenotype where mutants of increased fitness are isolated after long-
22
term incubation in stationary phase (Finkel, 2006; Zambrano et al., 1993). However,
Yeiser et al. used strains mutant for only a single alternative DNA polymerase, making it
impossible to assign specific activities to each enzyme.
Here we investigate the roles and quantitative contributions of Pols II, IV, and V in cell
growth and evolutionary fitness using a series of isogenic mutant strains lacking each
alternative polymerase in all possible combinations: as single null mutants, as
combinations of double null mutants that express only one alternative polymerase, and as
a triple null mutant. This study provides a comprehensive analysis of the roles of Pols II,
IV and V in the absence of exogenous DNA damage by determining each polymerase’s
contribution toward cell survival and evolutionary fitness during periods of feast, in batch
culture growth or in a chemostat where nutrients are continually being replenished, and
famine, as cells enter stationary phase and transition into long-term stationary phase
where nutrients are being depleted. These data establish individual roles for each
polymerase under conditions where different strains must compete to survive. We show
that while any one of the polymerases can randomly generate a mutation that enables a
cell to survive during either rapid or slow growth, there are nevertheless conditions
during batch culture where the activity of each individual enzyme is either dominant or
co-dominant. Previous studies on the regulation of the alternative polymerases have
focused on increased transcription in response to DNA damage (Courcelle et al., 2001;
Kenyon and Walker, 1980). Here we provide new data showing the transcription pattern
23
of each alternative polymerase throughout the five phases of the bacterial life cycle
(Finkel, 2006), in the absence of “standard” SOS induction by exogenous stressors.
2.3 Materials and Methods
2.3.1 Strains used and mutant construction
All strains (Table 2.1) are derived from E. coli K-12 strain ZK126 (W3110 ∆lacU169
tna-2), including nalidixic acid-resistant wildtype strain ZK1142 (Zambrano et al., 1993).
DNA polymerase single, double, and triple mutants were constructed by bacteriophage
P1 transduction into ZK126 using the following donor strains: for Pol II
−
, SH2101
(polB::Spc) (Bonner et al., 1992); for Pol IV
−
, RW626 (dinB::Kan); and for Pol V
−
,
RW82 (umuDC::Cam)(both RW626 and RW82 were generous gifts from Roger
Woodgate [National Institutes of Health, Bethesda, MD]). Genetic elements conferring
antibiotic resistance are effectively neutral in the absence of drug selection (Kraigsley
and Finkel, 2009; Yeiser et al., 2002). Strains lacking a single alternative polymerase are
designated with a superscript “minus” sign (Pol II
-
, Pol IV
-
, and Pol V
-
), whereas double-
mutant strains capable of expressing only one alternative polymerase are designated with
a superscript “plus” sign (Pol II
+
, Pol IV
+
, and Pol V
+
).
24
Table 2.1. Strains used in chapter 2.
Strain
Relevant genotype
or phenotype
Nomenclature
Pol
II
Pol
IV
Pol
V
Reference
ZK126 W3110 ΔlacU169 tna-2 WT + + + Zambrano et al. (1993)
ZK1142
ZK126 Nal
R
WT + + + Zambrano et al. (1993)
SF2003
ZK126 polB::Spc
R
Pol II
-
- + + Yeiser et al. (2002)
SF2006
ZK126 dinB::Kan
R
Pol IV
-
+ - + Yeiser et al. (2002)
SF2009
ZK126 umuDC::Cam
R
Pol V
-
+ + - Yeiser et al. (2002)
SF2012
ZK126 polB::Spc
R
dinB::Kan
R
Pol V
+
- - + This study
SF2014
ZK126 polB::Spc
R
umuDC::Cam
R
Pol IV
+
- + - This study
SF2016
ZK126 dinB::Kan
R
umuDC::Cam
R
Pol II
+
+ - - This study
SF2018
ZK126 polB::Spc
R
dinB::Kan
R
umuDC::Cam
R
- / - / - - - - Yamanaka et al. (2011)
2.3.2 Culture conditions, media, and titering assays
Strains were cultured in 5.0mL LB Broth, Lennox (Difco-BD) and incubated at 37ºC with
aeration in a TC-7 test tube roller (New Brunswick Scientific), unless otherwise
specified. Experiments were initiated from overnight cultures inoculated from frozen LB-
glycerol stocks. Viable counts were determined by serial dilution of cells periodically
removed from cultures, and plating on selective medium containing the appropriate
antibiotics: Nal, nalidixic acid (20µg/mL); Str, streptomycin (25µg/mL); Spc,
spectinomycin (100µg/mL); Kan, kanamycin (50µg/mL); Cam, chloramphenicol
(30µg/mL); and Rif, rifampicin (100µg/mL). This method of titering is accurate within
+/- 3-fold (Kraigsley and Finkel, 2009), and has a limit of detection of 1000 colony
forming units (CFU)/mL in this study.
25
2.3.3 Batch culture long-term competition assays
Two types of batch culture competitions, distinguished by their initial cell densities, were
performed: Type 1, where both strains are mixed at a low initial density (~10
6
CFU/mL)
and co-outgrown, and Type 2, where high-density stationary phase cultures are mixed
(~10
9
CFU/mL). In Type 1 experiments, competitions were initiated by inoculating 5µL
of each competing strain (1:1000 dilution, vol:vol) into the same 5.0mL LB culture. In
Type 2 experiments, competitions were initiated by combining 2.5mL of overnight
stationary phase cultures of each strain (1:1 mix, vol:vol). Viable counts were determined
as described above using the appropriate combinations of antibiotics (Kan/Cam for Pol
II
+
, Spc/Cam for Pol IV
+
, and Spc/Kan for Pol V
+
). At the conclusion of each 14-day
competition, strains showing a 10-fold greater relative population density were scored as
winners.
2.3.4 Serial passage aging regimen
To observe changes in relative fitness following repeated outgrowth, strains were serially
passaged in 5.0mL LB cultures. Independent cultures of each strain were incubated for 24
hours, sampled and diluted 1:1000 (vol/vol) into fresh medium. Cultures were passaged
daily for 5 days and frozen LB-glycerol stocks were prepared following each passage.
Passaged strains are identified with the number of additional passages in parentheses
prior to its phenotype [e.g. (+5) Pol II
+
denotes Pol II
+
cells following five serial
passages.]
26
2.3.5 Selection of GASP mutants and GASP competitions
To select for GASP mutants, strains were inoculated into independent 5.0mL LB
cultures, incubated for 10 days, and 150µL was frozen as LB glycerol stocks. To initiate
GASP competitions, an overnight culture of each aged population was grown and
introduced as a minority at a 1:1000 (vol/vol) dilution into a 5.0mL culture of an unaged
population of wildtype cells (ZK1142) or polymerase mutant strains (Zambrano et al.,
1993). Population densities for each strain were determined by titering, as described
above.
2.3.6 Chemostat competitions
To assess the relative fitness of polymerase-deficient strains under conditions promoting
rapid growth, competitions were performed under continuous culture conditions in
chemostats (Chao and Cox, 1983; Harder and Kuenen, 1977). For each competition,
250µL (1:3000, vol:vol) of each double-mutant strain (Pol II
+
, IV
+
and V
+
) was
inoculated into 750mL LB in a BioFlo 110 bioreactor (New Brunswick Scientific). The
chemostat was run for up to 2 hours under batch conditions to obtain the desired density
before initiating flow of fresh medium into the growth chamber. Average dilution rates
varied from 1 to 4 volumes per hour and chemostats were run between ~6 to 10 hours.
The chemostat culture was regularly sampled to monitor optical density and determine
viable counts.
27
2.3.7 Mutation frequency assay
The frequency of spontaneous rifampicin resistance was determined in wildtype and all
seven polymerase-deficient strains. For each strain, 159 independent 5.0mL overnight
cultures were grown and 100µL of each was spread onto plates containing rifampicin.
Total cell counts were determined by plating an appropriate dilution of each culture on
LB agar. The number of Rif
R
colonies was determined after 48h of incubation at 37ºC.
The frequency of spontaneous rifampicin resistance was calculated by dividing the
number of Rif
R
CFU/mL by the total CFU/mL. The distributions of observed Rif
R
frequencies were compared using the two-sample K-S Test (p <
0.05)(http://www.physics.csbsju.edu/stats/KS-test.html).
2.3.8 Rif
R
mutant sequencing
The sequences of Rif
R
mutants were determined using cells obtained from the mutation
frequency assay described above. The Rif
R
colony nearest the center of each plate was
resuspended in 20µL LB. 2µL was used as template for PCR amplification of the rpoB
gene using the primers 5’-AATGTCAAATCCGTGGCGTG-3’ and 5’-
TTCACCCGGATACATCTCGTC-3’, with the remaining sample frozen in LB-glycerol.
Amplified products were sequenced at High-Throughput Sequencing Solutions (Seattle,
WA) using the primers 5’-AATGTCAAATCCGTGGCGTG-3’ and 5’-
CGTCGTATCCGTTCCGTTGG-3’ specific for Cluster I and Cluster II, respectively
(Garibyan et al., 2003).
28
2.3.9 Quantitative RT-PCR
Real-time PCR was used to determine the expression patterns of each alternative
polymerase gene, as well as induction of the SOS response. LB cultures inoculated with a
1:1000 dilution of an overnight wildtype population were incubated at 37ºC and
periodically sampled for RNA extraction. Samples were treated with RNAlater (Qiagen)
and total RNA was isolated using the RNeasy Mini Kit (Qiagen). qRT-PCR reactions
with 100ng template RNA were performed using the One-step RT-PCR Kit (Qiagen)
with SYBR Green (Molecular Probes) and amplified on an Opticon-2 Real-time PCR
Cycler (MJ Research). Primers used for amplification are provided in Table 2.2. Relative
transcript abundance and changes in gene expression were determined using the 2
-ΔCt
method (Livak and Schmittgen, 2001). In control experiments with artificial SOS
induction, 1 µg/ml mitomycin C (Sigma) was added after 2 hours of incubation, at an
OD
600
of ~0.1, and RNA was sampled hourly.
29
Table 2.2. qRT-PCR primers used in chapter 2.
Primer Sequence (5'-3')
polB Forward agttcgaaatcaagcgagga
polB Reverse tgcacaatggcactatcgtt
dinB Forward gacctgcggtgatgtacaaa
dinB Reverse tcgcttcacattcagaccag
umuC Forward ttatctgttcccgctcgttt
umuC Reverse cagttttaccgacgcgctat
umuD Forward tcgtcaaagcaagtggtgat
umuD Reverse gatggtaatgggcgagtacg
polA Forward ctttttccagcttcgcaatc
polA Reverse agttcaaacgctggactgct
dnaE Forward taccatttccacgtcaacga
dnaE Reverse cccggacatgatcagttttt
recA Forward cagcatcgataaacgcacag
recA Reverse gaagaccgttccatggatgt
sulA Forward tgcggtgttaaccagagttg
sulA Reverse tcaggctatgcacatcgttc
gapA Forward gtgatccggctaacctgaaa
gapA Reverse gtcctggccagcatatttgt
sbmC Forward tcgggaagcgtaaagtaacc
sbmC Reverse gttgcaggtttccatctcgt
fis Forward cgaacaacgcgtaaattctg
fis Reverse attgcatcaccatgtccaac
lexA Forward ttgcaggaagaggaagaagg
lexA Reverse ttatgcactgccagcaagtc
dps Forward ctgccatggtatccagatga
dps Reverse tgctttatacccgcaacgat
30
2.4 Results
2.4.1 Pol IV and Pol V confer greater relative fitness during long-term
stationary phase
When cultured individually in rich medium, log phase growth and stationary phase
survival of all seven polymerase-deficient strains are indistinguishable from wildtype
(Figure 2.1). To determine whether strains capable of expressing only one alternative
polymerase display altered fitness, each double mutant was competed against each of the
other two mutants in Type 1 co-outgrowth (Figure 2.2A-C) and Type 2 stationary phase
(Figure 2.2E-G) long-term competition experiments.
In Type 1 competitions, initiated at low population density, most competitions (59%)
ended as ties, with both strains reaching final densities within ten-fold of each other
(Figure 2.2D). Among those competitions with a clear winner, Pol V
+
outperformed both
Pol IV
+
and Pol II
+
strains. However, in Type 2 competitions, initiated with stationary
phase populations at high cell density, there was a decisive winner in nearly all (93%)
competitions (Figure 2.2H). Strains expressing only Pol II were significantly less fit than
strains expressing only Pol IV (Figure 2.2E) or Pol V (Figure 2.2F), losing in 78% of
competitions. In competitions between Pol IV
+
and Pol V
+
both strains performed equally
well, with Pol IV
+
winning as many competitions as Pol V
+
(Figure 2.2G).
31
Figure 2.1. Outgrowth of polymerase-deficient strains.
Cell densities of wildtype and all combinations of mutant strains are plotted during 24
hours of growth in LB. Timepoints were taken every 45 minutes for the first 12 hours of
growth and every 90 minutes during the second 12 hours of growth. All strains exhibited
indistinguishable growth curves; representative data are shown. Black, wildtype; light
red, Pol II
-
, light green, Pol IV
-
; light blue, Pol V
-
; dark blue, Pol V
+
; dark green, Pol IV
+
;
dark red, Pol II
+
; gray, triple mutant.
6
7
8
9
10
11
0
4
8
12
16
20
24
log
10
CFU/mL
Hour
32
Figure 2.2. Co-outgrowth and stationary phase competitions between double-
mutant strains.
Representative Type 1 co-outgrowth (A-C) and Type 2 stationary phase (E-F)
competitions between unaged polymerase-deficient strains: red lines, Pol II
+
; green lines,
Pol IV
+
; blue lines, Pol V
+
. Three representative competitions are shown where squares,
circles, and triangles indicate competition pairs. Competition data are summarized in (D)
& (H). Asterisks indicate that titers were below the limit of detection (<1000 CFU/ml.)
33
2.4.2 Pol II enables cells to express the GASP phenotype faster
To assess the capacity of each alternative polymerase to generate beneficial alleles, each
double-mutant strain was subjected to GASP competition assays. In a typical GASP
assay cultures are aged for 10 days to allow spontaneous random mutants to take over the
culture. Cells from these cultures are then incubated as a minority with unaged wildtype
cells (Zambrano et al., 1993). After incubating in monoculture for ten days, all three aged
mutant strains were able to outcompete the unaged wildtype population when introduced
as a minority (Figure 2.3A-C), indicating the presence of a beneficial GASP allele.
However, the aged Pol II
+
(Figure 2.3A) and Pol V
+
(Figure 2.3C) populations drive
unaged wildtype populations to extinction faster than aged Pol IV
+
(Figure 2B).
We also determined the GASP phenotype of each polymerase-deficient strain with
respect to one another, rather than the wildtype. Again, every aged population expressed
the GASP phenotype over unaged populations (Figure 2.4A-F), however the strength of
the GASP phenotype differed among strains as determined by the time it took for the
minority population to take over the culture. For each competition, the day the minority
became the majority was determined and the average day of takeover was calculated. The
Pol II
+
strain consistently exhibited the GASP phenotype fastest, with an average time to
takeover of 3.5 days, compared to 4.5 days or 5.8 days for the strains capable of
expressing only Pol V or Pol IV, respectively.
34
Figure 2.3. GASP competitions against wildtype.
Polymerase double-mutant strains were aged for 10 days and competed to determine their
GASP phenotypes against unaged wildtype strains. Filled symbols correspond to strains
aged for 10 days; open symbols correspond to unaged strains. Strains are indicated by
line color: wild type, black; Pol II
+
, red; Pol IV
+
, green; Pol V
+
. blue. Unaged wildtype
cells were competed against 10-day-old (A) Pol II
+
, (B) Pol IV
+
, or (C) Pol V
+
. Three
representative competitions are shown for each pair where squares, circles, and triangles
indicate competition pairs. Asterisks indicate that titers were below the limit of detection
(<1000 CFU/ml.)
35
Figure 2.4. GASP competitions against double-mutant strains.
Polymerase double-mutant strains were aged for 10 days and competed to determine their
GASP phenotypes against each unaged polymerase mutant strain. Filled symbols
correspond to strains aged for 10 days; open symbols correspond to unaged strains.
Strains are indicated by line color: Pol II
+
, red; Pol IV
+
, green; Pol V
+
, blue. Aged Pol II
+
strains were competed against unaged (A) Pol IV
+
or (D) Pol V
+
; aged Pol IV
+
strains
were competed against unaged (B) Pol II
+
or (E) Pol V
+
; aged Pol V
+
strains were
competed against unaged (C) Pol II
+
or (F) Pol IV
+
. Three representative competitions are
shown for each pair where squares, circles, and triangles indicate competition pairs.
Asterisks indicate that titers were below the limit of detection (<1000 CFU/ml.)
A
C
B
D
E
F
0 2 4 6 8 10 14 12 0 2 4 6 8 10 14 12
Day
*
*
*
*
*
*
*
*
*
*
*
*
11
9
7
5
3
11
9
7
5
3
11
9
7
5
3
log
10
CFU/mL
36
2.4.3 Pol II contributes to relative fitness during serial passage
Although the Pol II
+
strain was less fit than Pol IV
+
and Pol V
+
in Type 2 stationary phase
competitions (Figure 2.2E-G), in Type 1 competitions, initiated at low cell densities, the
Pol II
+
strain performed significantly better (Figure 2.2A-C). The critical difference
between these experiments is that in Type 1 competitions cells experience an additional
outgrowth, providing more opportunity to replicate and generate genotypic diversity
under nutrient-rich conditions. Since this additional outgrowth on the first day of
competition profoundly influenced the fitness of the Pol II
+
strain weeks later, we
determined whether subjecting each strain to five serial passages prior to competition
would amplify this effect.
Serially passaged Pol II
+
populations were then competed against unpassaged Pol IV
+
or
Pol V
+
to assess changes in fitness (Figure 2.5). With each additional passage, the Pol II
+
populations displayed an increase in relative fitness, to the extent that Pol II
+
populations
passaged five times [(+5) Pol II
+
] consistently outcompeted both Pol IV
+
and Pol V
+
strains (Figure 2.5E&J). Prior to the passaging regimen, the (+0)Pol II
+
population never
won a stationary phase competition (Figure 2.5A&F), yet after two or more additional
passages it never lost to either unpassaged Pol IV
+
(Figure 2.5E) or Pol V
+
(Figure 2.5J),
outcompeting Pol IV
+
by a wider margin than Pol V
+
(Figure 2.6). Serially passaged
(+5)Pol II
+
also performed significantly better against passaged (+5)Pol IV
+
and (+5)Pol
V
+
(Figure 2.7).
37
Figure 2.5. Increased relative fitness of Pol II
+
cells following serial passage.
Populations of unaged Pol IV
+
(green lines, A-E) or unaged Pol V
+
(blue lines, F-J) were
each competed against the Pol II
+
strain (red lines) after one to five serial passages. (A &
F) The control Pol II
+
strain with no additional passage, (+0) Pol
+
; (B & G) one additional
passage, (+1) Pol II
+
; (C & H) two additional passages, (+2) Pol II
+
; (D & I) three
additional passages, (+3) Pol II
+
; or (E & J) five additional passages, (+5) Pol II
+
. The
limit of detection is (<1000 CFU/ml).
38
Figure 2.6. Pol II
+
serial passage relative fitness summary.
The average log
10
ratio of final cell densities between serially passaged Pol II
+
versus Pol
IV
+
(green circles) or Pol V
+
(blue squares) for each of the five competitions is plotted.
-2
-1
0
1
2
3
0
1
2
3
4
5
log
10
Relative Density of Pol II
+
Number of Additional Passages
39
Figure 2.7. Stationary phase competitions following serial passage.
Representative Type 2 stationary phase competitions between polymerase-deficient
strains following five serial passages: red lines, (+5) Pol II
+
; green lines, (+5) Pol IV
+
;
blue lines, (+5) Pol V
+
. Three representative competitions are shown where squares,
circles, and triangles indicate competition pairs. Competition data are summarized in (D)
& (H). Asterisks indicate that titers were below the limit of detection (<1000 CFU/ml.)
A
C
B
11
9
7
5
3
Day
log
10
CFU/mL
0 2 4 6 8 10 14 12
11
9
7
5
3
11
9
7
5
3
*
40
2.4.4 Pol II confers greater relative fitness during continuous culture
To determine the relative physiological contributions of each polymerase under
conditions promoting high rates of cell division, a series of competitions were performed
under continuous culture conditions in a chemostat. The three strains encoding a single
alternative polymerase were co-cultured in chemostats at increasing flow rates,
corresponding to faster growth rates. In every competition Pol II
+
outperformed both Pol
IV
+
and Pol V
+
(Figure 2.8), with average cell yields >10% more at lower flow rates and
as much as 250% more at higher flow rates.
Figure 2.8. Pol II
+
relative fitness during chemostat competitions.
The average relative fitness of Pol II
+
compared to Pol IV
+
(in green) and Pol V
+
(in blue)
during continuous culture competitions is shown for different dilution rates (volumes per
hour).
0
1
2
3
1.6
2.0
2.3
2.5
4.0
Average Pol II
+
Ratio
Average Dilution Rate (Volumes/Hour)
41
While the presence of Pol II provided a significant advantage during continuous culture,
the cell density of the Pol II
+
mutant exhibited periodic fluctuations with respect to the
other two strains (Figure 2.9). At a dilution rate of 4 volumes per hour, Pol II
+
cell
density fluctuated ~3.5-fold with respect to the Pol V
+
strain and ~2.5-fold with respect to
Pol IV
+
; similar fluctuations were observed in all chemostat competitions (see Chapter 3).
While the frequency and magnitude of the fluctuations varied between experiments, Pol
II
+
always remained the dominant strain.
Figure 2.9. Fluctuations in Pol II
+
relative fitness during chemostat competitions.
The relative fitness of Pol II
+
compared to Pol IV
+
(in green) and Pol V
+
(in blue) is
shown over time for a chemostat competition with a constant dilution rate of 4.0
volumes/hour.
0
1
2
3
4
0
1
2
3
4
5
6
Pol II
+
Ratio
Time (Hours)
42
2.4.5 Alternative polymerases affect mutation frequency and spectrum
To elucidate the molecular basis of the differences in adaptive potential observed during
long-term competition, we characterized the frequency and spectrum of spontaneous
mutations in rpoB, encoding the β-subunit of RNA polymerase, conferring rifampicin
resistance (Rif
R
). Mutations known to confer drug resistance include all six classes of
missense mutations, as well as in-frame amplifications and deletions (Garibyan et al.,
2003; Jin and Gross, 1988; Makiela-Dzbenska et al., 2011; Reynolds, 2000; Severinov et
al., 1994; Singer et al., 1993; Wolff et al., 2004; Wrande et al., 2008). 1272 independent
cultures were assayed; 159 cultures each for the wildtype, all three single-mutant, all
three double-mutant, and the triple-mutant strains. Although the Rif
R
mutation
frequencies varied among independent replicates, significant strain-specific differences in
the distribution of mutation frequencies were observed (Figure 2.10A-C). While Pol IV
-
was similar to wildtype, both Pol II
-
and Pol V
-
showed significantly lower mutation
frequencies (Figure 2.10A). Among double-mutant strains, Pol V
+
had the greatest
mutation frequency, significantly higher than Pol IV
+
, which in turn was significantly
higher than that of Pol II
+
(Figure 2.10B). Median Rif
R
values reflect these shifts in
mutation frequency (Figure 2.10C).
43
Figure 2.10. Strain-specific Rif
R
mutation frequency.
The frequency of spontaneous rifampacin resistance was determined for 159 independent
replicates of the wildtype and each polymerase mutant strain. Mutation frequencies are
presented in ascending order. (A) Mutation frequencies for the wildtype (black) and all
three double-mutant strains (Pol II
+
, red; Pol IV
+
, green; Pol V
+
, blue). (B) Mutation
frequencies for the wildtype (black) and all three single-mutant strains (Pol II
-
, magenta;
Pol IV
-
, light green; Pol V
-
, cyan). (C) Median Rif
R
frequencies for each strain, including
the triple null mutant.
44
We also assessed the spectrum of mutations generated in each strain by sequencing rpoB
in individual Rif
R
clones. Out of 1009 sequenced isolates we identified 86 alleles across
53 nucleotide positions (Tables 2.3 and 2.4), with several alleles that appear to be
previously unreported (indicated in Table 2.3). Strain-specific mutation spectrum
differences were observed. The wildtype and Pol V
+
strains produced the greatest
numbers of different Rif
R
alleles across the greatest number of nucleotide positions, while
the strain deficient in all three alternative polymerases generated the fewest alleles at the
fewest positions. Strain-specific mutation spectra data presented in Table 2.5, and
summarized in Figure 2.11. Wildtype cells generated about half as many GCAT
mutations compared to any other strain, but had more than twice as many GCCG
mutations than all other strains except Pol V
+
. Pol II
+
generated more deletions than all
other strains combined.
45
Table 2.3. Rif
R
single-nucleotide polymorphisms identified.
Nucleotide
Previous
Base
Mutation
Previous
Codon
New
Codon
AA
Change
Wildtype Pol II
-
Pol IV
-
Pol V
-
Pol V
+
Pol IV
+
Pol II
+
- / - / -
428 G C* CGT CCT R143P* 0 1 0 1 1 0 0 0
436 G T GTT TTT V146F 1 2 0 1 0 0 1 0
437 T A* GTT GAT V146D* 1 0 0 0 0 0 1 0
442 C A CAG AAG Q148K 0 0 1 0 0 0 0 0
C CAG CCG Q148P 6 6 1 6 2 8 1 4
G CAG CGG Q148R 2 0 2 0 0 0 0 1
T CAG CTG Q148L 15 4 24 10 7 11 6 31
C CAG CAC Q148H 0 0 1 0 0 0 0 0
T CAG CAT Q148H 0 0 0 1 1 0 1 1
446 T C* CTG CCG L149P* 0 0 0 2 0 0 0 0
448 C G* CAC GAC H150D* 0 0 0 0 1 0 0 0
1525 A C AGC CGC S509R 0 0 0 0 1 1 0 1
1526 G T* AGC ATC S509I* 1 0 0 0 0 0 0 0
1527 C A AGC AGA S509R 2 0 1 1 1 1 1 0
A CTG CAG L511Q 1 1 4 0 3 0 1 2
C CTG CCG L511P 1 2 0 2 1 0 0 1
G CTG CGG L511R 2 1 1 3 2 1 1 1
1534 T C TCT CCT S512P 3 4 6 5 7 5 1 4
A TCT TAT S512Y 0 4 1 3 5 1 2 3
T TCT TTT S512F 0 0 2 2 2 2 2 1
A CAG AAG Q513K 1 0 1 0 0 0 0 0
G* CAG GAG Q513E* 0 0 0 0 0 1 0 0
C CAG CCG Q513P 5 1 3 5 2 1 4 1
G CAG CGG Q513R 1 0 1 0 1 2 0 0
T CAG CTG Q513L 5 2 2 1 1 2 2 2
1539 G T CAG CAT Q513H 0 0 0 0 1 0 0 0
A GAC AAC D516N 0 1 0 1 0 2 1 0
T GAC TAC D516Y 1 0 1 1 1 0 0 0
C GAC GCC D516A 0 0 0 0 0 0 1 0
G GAC GGC D516G 11 9 7 16 11 7 17 8
T GAC GTC D516V 0 1 1 3 0 2 1 0
1552 A G AAC GAC N518D 1 0 0 1 0 0 1 0
1565 C T TCT TTT S522F 0 0 2 0 0 0 1 1
1574 C G ACG AGG T525R 0 0 0 0 0 2 0 0
A CAC AAC H526N 1 4 0 0 0 1 0 0
G CAC GAC H526D 0 0 1 2 1 1 1 0
T CAC TAC H526Y 6 7 11 8 12 9 6 10
1577 A T CAC CTC H526L 0 2 1 0 1 0 1 0
A CAC CAA H526Q 0 0 0 0 0 0 0 1
G CAC CAG H526Q 0 0 0 0 1 0 0 0
Missense Mutation Observed in Strain
443 A
444 G
1532 T
1535 C
1537 C
1538 A
1546 G
1547 A
1576 C
1578 C
46
Table 2.3 (Continued)
All missense mutations in rpoB conferring rifampicin resistance (Rif
R
) identified in this
study are listed, along with the number observed within each strain background. Asterisks
(*) denote missense mutations conferring rifampicin resistance that to our knowledge
have not previously been reported.
Nucleotide
Previous
Base
Mutation
Previous
Codon
New
Codon
AA
Change
Wildtype Pol II
-
Pol IV
-
Pol V
-
Pol V
+
Pol IV
+
Pol II
+
- / - / -
1585 C T CGT TGT R529C 1 1 1 1 0 0 0 0
A CGT CAT R529H 0 4 1 3 3 1 0 0
T CGT CTT R529L 0 2 0 0 0 0 0 1
A TCC TAC S531Y 1 1 0 0 1 0 0 0
G* TCC TGC S531C* 0 0 0 0 1 0 0 0
T TCC TTC S531F 1 2 0 2 2 6 6 3
1594 G C* GCA CCA A532P* 0 1 0 0 0 0 1 0
1595 C A GCA GAA A532E 1 0 0 0 2 2 0 1
1597 C G* CTC GTC L533V* 1 0 0 0 0 0 0 0
A CTC CAC L533H 0 0 1 1 0 1 0 1
C CTC CCC L533P 1 6 2 4 1 2 1 0
G CTC CGC L533R 2 0 0 0 0 0 0 1
A GGC AGC G534S 0 1 0 0 3 1 0 3
T GGC TGC G534C 2 2 4 1 0 2 0 1
A GGC GAC G534D 0 1 0 0 1 0 0 1
C GGC GCC G534A 4 1 2 2 3 0 0 1
T GGC GTC G534V 2 0 2 0 2 0 0 0
1607 G T* GGC GTC G536V* 3 0 1 1 0 0 0 0
1610 G A GGT GAT G537D 0 1 1 0 2 0 0 0
1687 A C ACC CCC T563P 13 31 21 25 18 28 33 18
A* CCT CAT P564H* 1 0 0 0 0 0 0 0
T CCT CTT P564L 1 3 1 2 2 0 4 2
1702 A G* AAC GAC N568D* 0 1 0 0 0 0 0 0
1708 G T GGT TGT G570C 3 2 0 1 0 1 2 1
1709 G C GGT GCT G570A 2 1 0 0 4 0 0 0
1712 T A* CTG CAG L571Q* 0 1 1 0 1 1 0 0
C ATC CTC I572L 3 2 2 3 4 0 1 4
T ATC TTC I572F 2 0 1 0 2 1 1 1
A ATC AAC I572N 3 0 3 0 2 0 0 2
C ATC ACC I572T 1 0 0 0 0 1 0 0
G ATC AGC I572S 3 5 6 3 6 9 5 6
1716 C G ATC ATG I572M 1 0 0 0 0 0 0 0
C A TCT TAT S574Y 2 0 0 0 1 5 2 2
C T TCT TTT S574F 2 0 2 0 2 4 3 4
2060 G A CGT CAT R687H 0 0 1 0 0 0 0 0
1691 C
1714 A
1715 T
1721
Missense Mutation Observed in Strain
1586 G
1592 C
1598 T
1600 G
1601 G
47
Table 2.4. Rif
R
deletions and amplifications identified.
All unique deletions and amplifications in rpoB conferring rifampicin resistance (Rif
R
)
identified in this study are listed. Some deletions were identified more than once within a
given strain. The codons affected are provided.
Table 2.5. Strain-specific mutation spectrum.
The absolute number of mutations identified in each class of mutation, as well as the
percent of overall mutations they represent, within each strain is provided.
Strains Mutation Nucleotide Previous Codon AA Position
Previous
AA
New Codon New AA
Pol II
+
Deletion 1516 TTC GGT TCC 506/8 FGS - -
Pol II
+
Deletion 1519 GGT TCC 507/8 GS - -
Pol II
+
Deletion 1587 TAT 530 I - -
Pol II
+
Deletion 1594 GCA 532 A - -
Pol II
+
,
Pol V
+
Deletion 1594 GCA CTC 532/3 AL - -
Pol IV
+
Deletion 1594 GCA CTC GGC 532/4 ALG - -
Pol II
+
Deletion 1594 GCA CTC GGC CCA 532/5 ALGP - -
Pol V
-
,
Pol IV
+
Deletion 1604 CCA GGC GGT 535/7 PGG - -
WT Insertion 1528-1557 - - -
CAG CTG TCT CAG TTT ATG
GAC CAG AAC AAC
QLSQFMDQNN
- / - / - Insertion 1585-1590 - - - CGT ATC RI
# % # % # % # % # % # % # % # %
GC ! AT 11 9% 21 17% 22 17% 19 15% 29 22% 25 20% 23 18% 25 20%
GC ! TA 22 18% 17 14% 12 9% 10 8% 15 12% 13 10% 9 7% 12 9%
GC ! CG 9 7% 4 3% 4 3% 5 4% 12 9% 5 4% 2 2% 1 1%
AT ! GC 21 17% 22 18% 18 14% 30 24% 21 16% 17 13% 20 16% 14 11%
AT ! TA 27 22% 11 9% 38 30% 15 12% 17 13% 18 14% 13 10% 39 30%
AT ! CG 34 27% 46 38% 34 27% 45 36% 35 27% 48 38% 46 36% 36 28%
Deletions 0 0% 0 0% 0 0% 1 1% 1 1% 2 2% 14 11% 0 0%
Insertions 1 1% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0% 1 1%
Total Identified 125 121 128 125 130 128 127 128
Pol IV
+
Pol II
+
- / - / - WT Pol II
-
Pol IV
-
Pol V
-
Pol V
+
48
Figure 2.11. Strain-specific Rif
R
mutation spectrum.
The mutation spectrum of alleles conferring spontaneous rifampacin resistance was
determined for the wildtype and each polymerase mutant strain. Each class of mutation is
presented as the percent of each type of mutation observed in each strain background.
Detailed mutation data is presented in Tables 2.3 and 2.4, with source data in Table 2.5.
49
2.4.6 Stationary phase induces transcription of alternative polymerases and
additional SOS genes
Given their observed physiological and evolutionary impacts during periods of feast and
famine in the absence of endogenous stressors, we determined the expression patterns for
each alternative DNA polymerase during long-term incubation. Wildtype cultures were
sampled hourly for the first twelve hours and then every 24 to 48 hours over 14 days of
long-term stationary phase incubation. Quantitative RT-PCR on total cellular RNA
throughout long-term incubation (Figure 2.12) revealed an induction of transcripts
encoding all three alternative polymerases (polB, dinB, umuC), as well as the SOS
response gene sulA. Relative to their abundance after 2 hours of exponential phase
growth, by five hours, as the cells transition into stationary phase, the average abundance
of transcripts encoding Pol II and Pol IV more than doubled (dinB, +2.15-fold; polB,
+2.22-fold) and the catalytic subunit of Pol V was induced 50% by hour 7 (umuC, +1.51-
fold). Meanwhile, transcript levels for the dnaE-encoded catalytic subunit of Pol III
decreased by more than 40% within 4 hours (0.54-fold). For comparison, expression
levels of the early-exponential phase gene fis (Ball et al., 1992) peak at ~2 hours and
decrease significantly from mid-logarithmic phase through stationary phase and
expression of the stationary phase-specific dps gene (Martinez and Kolter, 1997)
increases from late log through stationary phase (see Figure 4.2).
50
Figure 2.12. Alternative polymerase transcript abundance increases over the cell
cycle.
mRNA transcript abundance, in the absence of exogenous SOS inducers, was determined
by qRT-PCR. Transcript abundance is shown for each gene relative to its concentration
after 2 hours of incubation. Genes are identified as: polB, red, open circles; dinB, green
open triangles; umuC, blue, open squares; umuD, blue, closed squares; dnaE, black
diamonds; sulA, purple, filled circles; lexA, orange, filled triangles.
51
When comparing the relative abundance of dnaE, polB, dinB, and umuC transcripts
(Figure 2.13) there is a dramatic decline in the average proportion of dnaE transcript
from ~69% during logarithmic growth to 34% upon entry into stationary phase. By 24
hours the dinB transcript is most abundant (56%), significantly exceeding that of dnaE
(31%). For all time points, there was a consistent hierarchy of transcript abundance
among alternative polymerases, with dinB > polB > umuC. The induction of alternative
polymerases corresponded with induction of sulA, an SOS gene, indicating that stationary
phase conditions induce the SOS response in the absence of exogenous stressors; these
genes are induced at ~25% of the levels observed during mitomycin C induction (see
Figure 4.7).
Figure 2.13. Alternative polymerase transcript proportion increases over the cell
cycle.
mRNA transcript abundance, in the absence of exogenous SOS inducers, was determined
by qRT-PCR. The proportion of polymerase transcripts over time, expressed as a
percentage of total transcript abundance, is shown for four representative transcripts: Pol
III, dnaE, black diamonds; Pol IV, polB, green triangles; Pol II, polB, red circles; Pol V,
umuC, blue squares.
52
2.5 Discussion
The long-term viability and evolutionary success of bacterial populations requires a
balance between maintaining the capacity to replicate efficiently with high fidelity while
simultaneously generating sufficient genetic diversity to facilitate adaptation and survival
in changing environments. Our previous study, which showed that alternative DNA
polymerases play a role in the evolutionary fitness of bacterial populations (Yeiser et al.,
2002), left open the key issue of each enzyme’s physiological and mutagenic
contribution. Here we show fundamental roles for each of the alternative DNA
polymerases. During periods of rapid cell division, Pol II contributes significantly to
relative fitness by facilitating faster growth and the generation of genetic diversity,
whereas Pol IV and Pol V introduce greater genetic variation, conferring increased
relative fitness, under the more stressful conditions of long-term stationary phase.
Furthermore, differences in the frequency and spectrum of allelic variation attributable to
each polymerase suggest a competitive hierarchy for access to replicate DNA in vivo.
Given that success during long-term stationary phase is influenced by the formation of
advantageous mutations, our competition experiments illustrate a race to generate
beneficial alleles quickly (Figure 2.2). The observation that the double-mutant Pol IV
+
and Pol V
+
strains perform better than Pol II
+
in both Type 1 and Type 2 competitions
indicates that these polymerases play a greater role in creating mutations, including
beneficial alleles, during long-term stationary phase. Therefore, Pols IV and V, which
53
have inherently low deoxynucleotide insertion fidelity and lack proofreading, are more
likely than Pol II to produce genetic diversity quickly in slowly dividing cells.
However, following a 10-day aging regimen where beneficial alleles are generated and
selected, Pol II
+
expressed the strongest GASP phenotype (Figure 2.3 and 2.4). This
finding indicates that, after providing ample time for beneficial mutations to appear, Pol
II provides a greater physiological contribution to overall fitness than either Pol IV or Pol
V via increased replicative efficiency during growth, enabling cells to capitalize on
beneficial GASP alleles more quickly. This role for Pol II is bolstered by our observation
that the Pol II
+
strain performed significantly better when experiments were initiated at
low density (allowing additional outgrowth) than at high density. The fact that serially
passaged (+5)Pol II
+
populations exhibit greater competitive fitness against passaged
(+5)Pol IV
+
or (+5)Pol V
+
leads to the conclusion that Pol II not only plays an important
physiological role with respect to the efficiency of DNA replication, but also contributes
considerable genetic diversity during periods of rapid growth.
This conclusion is reinforced by the chemostat competition experiments where the Pol II
+
strain dominated (Figure 2.8). The relative abundance of the Pol II
+
strain peaked at
regular intervals, potentially reflecting modulation of alternative polymerase expression.
Recent single-cell studies investigating the SOS and other stress responses have shown
transient bursts of activity occurring in discrete and repetitive peaks (Friedman et al.,
2005; Krishna et al., 2007; Locke et al., 2011). We speculate that Pol V may play a role
54
in maintaining this regulatory pattern, acting as a prokaryotic DNA-damage checkpoint to
produce a pause in replication (Murli et al., 2000; Opperman et al., 1999; Sutton et al.,
2001). These iterative bursts of expression could provide a method of limiting the
potentially harmful mutagenic effects of error-prone polymerases by enabling cells to
assess whether template lesions have been resolved before inducing further expression.
We analyzed the frequency and mutation spectrum of spontaneous Rif
R
alelles in each of
the seven polymerase-deficient strains and wildtype (Figures 2.10 and 2.11). Previous
work identified differences in the frequency and spectrum of mutations generated by
alternative polymerases, however many of these studies analyzed either a limited subset
of mutation classes, evaluated reporter genes on extra-chromosomal elements, looked at
mutations generated under artificial expression conditions, following treatment with
exogenous stressors, or used strains with altered repair systems (Curti et al., 2009; Foster,
2005, 2007; Galhardo et al., 2007; Hersh et al., 2004; Wolff et al., 2004). Here we looked
at the distribution of mutation frequencies under standard laboratory conditions with
endogenous expression in the absence of exogenous stressors, in wildtype and isogenic
mutant strains. The distribution of mutation frequencies reveals significant differences in
the in vivo fidelity of each polymerase. Use of the double-mutant strains enables the
analysis of each polymerase’s contribution to overall mutation frequency. The Pol V
+
strain was responsible for generating the greatest allelic diversity, followed by Pol IV
+
,
and Pol II
+
. This hierarchy of fidelity is consistent with previously published in vitro data
and several in vivo studies (Cai et al., 1995b; Kobayashi et al., 2002; Maor-Shoshani et
55
al., 2000; Tang et al., 2000). It is also important to compare the mutation frequency of the
single-mutant strains where two polymerases can compete for access to replicate DNA,
enabling additional inferences regarding their relative contributions when acting in
concert. Here the Pol IV
-
strain exhibited a greater mutation frequency than either Pol II
-
or Pol V
-
, which were approximately equal.
Three factors are often implicated in the competition for access to replicate and mediate
TLS: (1) the relative expression levels of each polymerase, (2) their affinity for the β-
clamp and (3) their kinetic affinity for a lesion (Al Mamun and Humayun, 2006; Delmas
and Matic, 2006; Hastings et al., 2010; Heltzel et al., 2009a; Heltzel et al., 2009b; Indiani
et al., 2005; Sutton, 2004, 2010). Any polymerase present at a higher concentration, with
a greater affinity for the β-clamp, or with more favorable kinetics for resolving particular
lesions would be expected to replicate more and perform a greater fraction of TLS, and
would be expected to mask or supersede the effects of competing polymerases.
Therefore, Pol IV, the most abundant alternative DNA polymerase (Nohmi, 2006), may
have a numerical advantage and dominate the majority of TLS. This numerical
domination should presumably persist in the absence of exogenous DNA damage,
provided that Pols II, IV, and V are abundantly transcribed during in long-term stationary
phase, as appears to be true (Figure 2.12). Consistent with this model, we show that the
presence or absence of Pol IV is the single greatest determinant of mutation frequency.
Among strains capable of expressing only one alternative polymerase, Pol IV confers an
intermediate mutation frequency. Strikingly, this intermediate frequency is maintained in
56
the presence of competing polymerases Pol II or Pol V. When Pol II is absent, Pol IV
may be outcompeting and masking the greater mutagenic potential of Pol V. When Pol V
is missing, Pol IV appears to outcompete the higher fidelity polymerase Pol II, hence the
similar mutation frequencies in these three strain backgrounds.
In the strain lacking only Pol IV, leaving Pol II and Pol V to compete, the mutation
frequency was significantly higher and remarkably similar to the mutation frequency of
the double-mutant strain expressing only Pol V. Therefore, Pol V outcompetes Pol II to
replicate during long-term batch culture. Together, these findings are consistent with a
competitive hierarchy for access to replicate DNA in vivo of Pol IV > Pol V > Pol II,
when cells are incubating in long-term stationary phase. Therefore, even though Pol V is
inherently more mutagenic, its activity is modulated by Pol IV. Pol IV’s higher levels of
expression enable it to outcompete the other polymerases and give it more opportunity to
introduce genetic variation, consistent with its significant role in generating mutations in
the lac system (Foster, 2000; McKenzie et al., 2001).
Among over 1000 spontaneous mutants resistant to rifampicin, we identified 86 alleles
encompassing all six classes of missense mutations, as well as in-frame insertions and
deletions. We found that the wildtype and Pol V
+
strains generated the greatest numbers
of different alleles across the greatest number of nucleotide positions in rpoB. The
triple-mutant strain completely deficient in alternative polymerase expression introduced
57
the least variety of alleles at the fewest positions, demonstrating the importance of these
enzymes in creating diverse alleles.
Previous reports have identified ‘fingerprints’ or mutational ‘hotspots’ left by specific
polymerases in various genetic backgrounds
(Curti et al., 2009; Wolff et al., 2004). We
also identified strain-specific fluctuations in the frequency of specific alleles and
mutation classes. For example, nearly half as many GCAT mutations were identified in
the wildtype strain compared to any other polymerase-deficient strain. Twice as many
GCCG mutations were observed in wildtype relative to all strains except Pol V
+
,
suggesting Pol V plays an important role in generating these transversions. The number
of deletions detected in the Pol II
+
strain was greater than all other strains combined.
Since this strain is incapable of expressing Pol IV or Pol V, it suggests Pol II plays a
significant role in generating deletions, as has recently been reported (Koskiniemi and
Andersson, 2009; Wang and Yang, 2009). However, the wildtype and Pol IV
-
strains,
capable of expressing Pol II, did not produce any deletions, while the Pol IV
+
and PolV
+
strains did. Although the Y-family polymerases can form deletions, they do so at a much
lower frequency than Pol II. If both Pol IV and Pol V outcompete Pol II for access to
mediate translesion synthesis, their presence will mask the likelihood of Pol II
introducing a deletion. While the alterations in mutation spectrum observed differ from
polymerase fingerprints reported elsewhere, these discrepancies may reflect strain and
environmental differences.
58
Our findings reveal important roles of alternative DNA polymerases under conditions of
feast and famine. The fact that the alternative polymerases, members of the SOS regulon,
are expressed in the absence of endogenous stressors during long-term stationary phase
further illustrates their impact on the survival and evolution of bacterial populations.
When replacing Pol III, these polymerases each play a central role in providing a balance
that achieves sufficiently rapid growth while generating sufficient mutations to ensure
genetic diversity. The observation that each alternative polymerase contributes differently
to the variety and quantity of mutations demonstrates the importance of competitive
interactions among the polymerases. One explanation for the apparent increase in
alternative polymerase expression during long-term stationary phase may be a positive
selection on bacterial populations that increase genetic diversity during times of stress.
Given the ubiquitous nature of alternative DNA polymerases and their impact on cell
fitness and survival, a deeper understanding of factors affecting their expression,
competitive interactions, mutation preferences, and other cellular functions will yield
valuable insights toward understanding the physiological responses and evolutionary
trajectories of bacterial populations.
59
Chapter 3: DNA polymerase II confers a physiological fitness advantage
within continuous culture chemostat competitions
3.1 Abstract
Microbial communities often cope with dynamic and unpredictable changes in nutrient
availability and environmental stress. The capacity to quickly adjust and thrive during
intermittent cycles of feast and famine is crucial to evolutionary fitness. Accordingly,
mechanisms facilitating rapid growth can provide a significant physiological advantage.
Although alternative DNA polymerases have been extensively characterized for their
roles during stressful environmental conditions, by enabling replication past sites of DNA
damage that might otherwise halt growth, their physiological roles might also be of vital
importance during periods of rapid growth. Using a series of isogenic mutant strains
capable of expressing a single alternative DNA polymerase we investigate the effects and
relative contributions of Pol II, IV, and V on cell growth within nutrient rich
environments. Competitions conducted during serial passage and within chemostats
reveal Pol II confers a physiological advantage by enabling faster growth during periods
of feast as cells are rapidly dividing, and that this fitness advantage increases with growth
rate. These results offer compelling insights into the roles of each alternative DNA
polymerase upon transitioning from stressful conditions to those conducive to growth, as
well as the dynamics of polymerase regulation, DNA synthesis and growth.
60
3.2 Introduction
Within their natural habitats, microbial communities face dynamic and unpredictable
changes in nutrient availability and environmental stress, so the ability to quickly adjust
and thrive during intermittent periods of feast and famine is crucial to evolutionary
fitness. Accordingly, bacteria have evolved to utilize numerous strategies to ensure both
physiological and evolutionary fitness within changing environments.
Among many factors critical for success within unpredictable environments is the
capacity to “strike while the iron’s hot” and reproduce quickly when environments are
conducive to growth. Within a complex microbial consortium with competition for finite
resources, any molecular mechanisms improving the capacity of an organism to sense
and capitalize within nutrient-rich conditions can confer a significant evolutionary
advantage (Ferenci, 2008; Harder and Kuenen, 1977). Accordingly, individuals capable
of quickly responding to an influx of nutrients and replicating faster can increase their
relative density compared to slower growing competitors.
While alternative DNA polymerases have been widely characterized for their roles in
facilitating replication past DNA damage under stressful environmental conditions
(Kenyon and Walker, 1980; Layton and Foster, 2003, 2005; Stumpf and Foster, 2005;
Yeiser et al., 2002), there are also indications that they could be playing a previously
unexpected role during nutrient rich conditions as well. DNA polymerase II and IV are
both constitutively expressed at relatively low but appreciable abundance throughout
61
logarithmic growth (Nohmi, 2006), making them available even during rapid growth.
Given their well-characterized roles in translesion sythesis, alternative DNA polymerase
likely also serve to facilitate efficient replication during times of growth. This is because
even during periods of rapid growth any lesions in template DNA can act as a roadblock
for the progression of replication forks and can halt DNA synthesis until the lesion can be
resolved. Within nutrient rich environments, excessive stalling would prevent organisms
from replicating as fast as possible and make them less efficient at capitalizing with
growth during periods of feast.
Previous results indicated the capacity to express DNA polymerase II conferred a
significant advantage following a serial passage aging regimen, after populations were
repeatedly introduced into new cultures with fresh medium (Figure 2.5). While the
observed advantage in long-term stationary phase competitions was genetically encoded,
this finding also suggested Pol II might play a greater relative role in replication during
outgrowth possibly by enabling faster replication and growth during serial passage. Here
we sought to amplify, identify and characterize any possible physiological growth
advantage attributable to each alternative polymerase through the use of competitions
during serial passage and within a chemostat.
Chemostats (coined from “chemical environment is static”) can be defined as bioreactors
in which fresh culture medium is continuously added while spent culture fluid is
continuously removed (Ferenci, 2008; Harder and Kuenen, 1977; Novick and Szilard,
62
1950). In this manner, by changing the rate at which fresh medium is added to the growth
chamber, the growth rate of the microorganisms can be held constant. This controlled
growth is capable of creating conditions wherein growth and all culture parameters
(volume, density, nutrient concentration, pH, dissolved oxygen etc.) can theoretically
remain constant, creating “steady-state” conditions ideal for analyzing cell physiology.
Typically, chemostats utilize a defined growth medium in which a single growth factor
becomes limiting, thus inhibiting further growth. However, to recapitulate conditions
used during the initial observation during serial passage, the complex growth medium LB
was used here. Accordingly, by providing a steady supply of fresh nutrients, growth rates
remained high and enabled us to assess the relative capacity of each strain to reproduce as
quickly as possible under nutrient-rich continuous growth.
Using a series of isogenic mutant strains capable of expressing a single alternative DNA
polymerase we investigated the effects and relative contributions of Pol II, IV, and V on
cell growth within nutrient rich environments. Competitions conducted during serial
passage and within chemostats reveal Pol II confers a physiological advantage by
enabling faster growth during periods of feast as cells are rapidly dividing, and that this
fitness advantage increases with growth rate. These results offer compelling insights into
the roles of each alternative DNA polymerase upon transitioning from stressful
conditions to those conducive to growth, as well as the dynamics of polymerase
regulation, DNA synthesis and growth.
63
3.3 Materials and Methods
3.3.1 Strains used and mutant construction
All strains (Table 3.1) are derived from E. coli K-12 strain ZK126 (W3110 ∆lacU169
tna-2) (Zambrano et al., 1993). DNA polymerase double mutants were constructed by
bacteriophage P1 transduction into ZK126 using the following donor strains: for Pol II
−
,
SH2101 (polB::Spc) (Bonner et al., 1992); for Pol IV
−
, RW626 (dinB::Kan); and for Pol
V
−
, RW82 (umuDC::Cam)(both RW626 and RW82 were generous gifts from Roger
Woodgate [National Institutes of Health, Bethesda, MD]). Genetic elements conferring
antibiotic resistance are effectively neutral in the absence of drug selection (Kraigsley
and Finkel, 2009; Yeiser et al., 2002). Double-mutant strains capable of expressing only
one alternative polymerase are designated with a superscript “plus” sign (Pol II
+
, Pol IV
+
,
and Pol V
+
).
Table 3.1. Strains used in chapter 3
Strain Relevent Genotype/Phenotype Nomenclature Pol II Pol IV Pol V
SF2012
ZK126 polB::Spc
R
dinB::Kan
R
Pol V
+
- - +
SF2014
ZK126 polB::Spc
R
umuDC::Cam
R
Pol IV
+
- + -
SF2016
ZK126 dinB::Kan
R
umuDC::Cam
R
Pol II
+
+ - -
3.3.2 Culture conditions, media, and titering Assays
Strains were initially cultured in 5.0mL LB Broth, Lennox (Difco-BD) and incubated at
37ºC with aeration in a TC-7 test tube roller (New Brunswick Scientific). Competitions
were initiated using overnight cultures inoculated from frozen LB-glycerol stocks. Viable
counts were determined by serial dilution of cells periodically sampled from each culture,
64
and plating on selective medium containing the appropriate combination of antibiotics:
Spc, spectinomycin (100µg/mL); Kan, kanamycin (50µg/mL); and Cam,
chloramphenicol (30µg/mL). This method of titering is accurate within +/- 3-fold
(Kraigsley and Finkel, 2009), and has a limit of detection of 1000 colony forming units
(CFU)/mL in this study.
3.3.3 Serial passage competitions
To identify any strain-specific difference in relative fitness during serial passage
conditions, all possible pair-wise competitions were initiated between each double-
mutant strain (Pol II
+
, IV
+
and V
+
). For each competition, 5µL (1:1000, vol:vol) of each
competing strain was inoculated into a fresh 5.0mL culture. Competitions were allowed
to grow 24 hours before diluting dense cultures 1:1000 into another fresh 5.0mL culture.
This serial passage regimen was performed every 24 hours for two weeks and viable
counts were determined at the end of every 24-hour period as described above using
appropriate combinations of antibiotics (Kan/Cam for Pol II
+
, Spc/Cam for Pol IV
+
, and
Spc/Kan for Pol V
+
).
3.3.4 Chemostat competitions
To assess the relative fitness of polymerase-deficient strains under conditions promoting
rapid growth, competitions were performed under continuous culture conditions in
chemostats (Chao and Cox, 1983; Harder and Kuenen, 1977). For each competition,
250µL (1:3000, vol:vol) of each double-mutant strain (Pol II
+
, IV
+
and V
+
) was grown
65
overnight in batch culture and inoculated into 750mL LB within a BioFlo 110 bioreactor
(New Brunswick Scientific). The chemostat was run for up to 2 hours under batch
conditions to obtain the desired population density before initiating flow of fresh LB
medium into the growth chamber. The chemostat culture was regularly sampled to
monitor optical density (A
600
) and titered on appropriate plates to determine the viable
counts of each strain. Average dilution rates varied from 1 to 4 volumes per hour and
chemostats were run between ~6 to 10 hours.
3.4 Results
3.4.1 Pol II confers an initial advantage during serial passage competitions
Previous results (Figure 2.5) indicated Pol II populations obtained following serial
passage had increased in relative fitness, and suggested Pol II might play a greater role
facilitating replication during outgrowth. To further assess whether Pol II might confer a
physiological advantage during rapid growth a series of pairwise competitions were
initiated within cultures passaged daily over twelve days (Figure 3.1). Within all serial
passage competitions between Pol IV
+
and Pol V
+
both strains remained present at similar
densities throughout the duration of every competition (Figure 3.1A). However, the Pol
II
+
strain exhibited titer counts greater than both Pol IV
+
(Figure 3.1B) and Pol V
+
(Figure
3.1C) throughout the initial days of each competition. Under these conditions, Pol II
+
was
present at an average density greater than Pol IV
+
until day 8 (Figure 3.1E), and greater
than Pol V
+
until day 4 (Figure 3.1F). After Pol V
+
and Pol IV
+
became the majority
populations, the Pol II
+
strain remained in the minority throughout the duration of the
competition.
66
Figure 3.1. Initial Pol II
+
advantage during serial passage competitions.
Polymerase double-mutant strains were mixed and competed to during serial passage to
assess their relative fitness during periods of rapid growth. Eight competitions were
initiated for each pair of double-mutant strains: (A) Pol IV
+
vs. Pol V
+
, (B) Pol II
+
vs. Pol
IV
+
, and (C) Pol II
+
vs. Pol V
+
. Strains are indicated by line color: Pol II
+
, red; Pol IV
+
,
green; Pol V
+
. blue. (B) Average cell densities for each strain throughout serial passage
competition: red lines, Pol II
+
; green lines, Pol IV
+
; blue lines, Pol V
+
.
67
3.4.2 Pol II confers greater relative fitness during continuous culture
To determine the relative physiological contributions of each polymerase under
conditions promoting high rates of cell division, a series of competitions were performed
under continuous culture conditions within a chemostat. The three strains encoding a
single alternative polymerase were co-cultured in chemostats at increasing flow rates,
corresponding to faster growth rates. Optical density was monitored to ensure culture
turbidity remained at appropriate levels and samples were regularly titered to determine
the population density of each competing strain. In every competition Pol II
+
outperformed both Pol IV
+
and Pol V
+
(Figure 3.2), with average cell yields >10% more
at lower flow rates and as much as 250% more at higher flow rates. The relative
performance of Pol II
+
did not appear to be strictly dependent on overall population
density or dissolved oxygen content (Figures 3.3 and 3.4, respectively). Although Pol II
+
exhibited the greatest advantage when cultures were at lower densities and greater
oxygen content, there was not a statistically significant correlation.
68
Figure 3.2. Pol II
+
relative fitness during chemostat competitions.
The average relative fitness of Pol II
+
compared to Pol IV
+
(in green) and Pol V
+
(in blue)
during continuous culture competitions is shown for different dilution rates (volumes per
hour).
0
1
2
3
1.6
2.0
2.3
2.5
4.0
Average Pol II
+
Ratio
Average Dilution Rate (Volumes/Hour)
69
Figure 3.3. Pol II
+
relative fitness according to overall population density during
chemostat competitions.
The relative fitness of Pol II
+
compared to Pol IV
+
(green triangles) and Pol V
+
(blue
circles) is plotted according to the overall population density observed at that timepoint.
All timepoints throughout all chemostat competitions are shown.
0
1
2
3
4
5
6
4
5
6
7
8
9
Pol II
+
Ratio
log
10
CFU/mL
II+/IV+
II+/V+
70
Figure 3.4. Pol II
+
relative fitness according to dissolved oxygen content during
chemostat competitions.
The relative fitness of Pol II
+
compared to Pol IV
+
(green triangles) and Pol V
+
(blue
circles) is plotted according to the dissolved oxygen content observed at that timepoint.
All timepoints throughout all chemostat competitions are shown.
1
2
3
4
5
6
0
20
40
60
80
100
Pol II
+
Ratio
Disolved Oxygen (%)
II+/IV+
II+/V+
71
3.4.3 Periodic fluctuations in Pol II competitive advantage
While the presence of Pol II provided a significant advantage during continuous culture
competitions, the cell density of the Pol II
+
mutant exhibited periodic fluctuations with
respect to the other two strains throughout five independent chemostat competitions
conducted with different average flow rates (Figures 3.5-3.9). For example, at a constant
dilution rate of 4 volumes per hour, Pol II
+
cell density fluctuated ~3.5-fold with respect
to the Pol V
+
strain and ~2.5-fold with respect to Pol IV
+
(Figure 3.9). Although the
frequency of peaks in relative abundance varied among experiments, Pol IV
+
was
observed to peak on average more frequently (84 minutes) than Pol V
+
(108 minutes).
While the magnitude of these fluctuations varied between experiments, Pol II
+
consistently remained the dominant strain.
72
Figure 3.5. Fluctuations in Pol II
+
relative fitness during a chemostat competition
with an average flow rate of 1.6 volumes/hour.
All three polymerase double-mutant strains were mixed and competed during continuous
culture within a chemostat to assess relative fitness. (A) The population density of each
strain is shown throughout the competition, along with the flow rate of fresh media
(mL/min). Strains are indicated by line color: Pol II
+
, red; Pol IV
+
, green; Pol V
+
. blue.
Flow rate is indicated in black. (B) The relative fitness of Pol II
+
compared to Pol IV
+
(in
green) and Pol V
+
(in blue) is shown over time for this competition, which had an
average dilution rate of 1.6 volumes/hour upon initiating flow of nutrients.
0.0
0.5
1.0
1.5
2.0
2.5
0
2
4
6
8
Pol II Ratio
Hour
II+/IV+
II+/V+
0
10
20
30
40
50
4
5
6
7
8
9
0
2
4
6
8
Flow Rate (mL/min)
log
10
CFU/mL
Hour
Pol II+
Pol IV+
Pol V+
Flow Rate
0
10
20
30
40
50
4
5
6
7
8
9
Flow Rate (mL/min)
log
10
CFU/mL
0.0
0.5
1.0
1.5
2.0
2.5
0
2
4
6
8
Pol II
+
Ratio
Hour
A
B
73
Figure 3.6. Fluctuations in Pol II
+
relative fitness during a chemostat competition
with an average flow rate of 2.0 volumes/hour.
All three polymerase double-mutant strains were mixed and competed during continuous
culture within a chemostat to assess relative fitness. (A) The population density of each
strain is shown throughout the competition, along with the flow rate of fresh media
(mL/min). Strains are indicated by line color: Pol II
+
, red; Pol IV
+
, green; Pol V
+
. blue.
Flow rate is indicated in black. (B) The relative fitness of Pol II
+
compared to Pol IV
+
(in
green) and Pol V
+
(in blue) is shown over time for this competition, which had an
average dilution rate of 2.0 volumes/hour upon initiating flow of nutrients.
0.0
0.5
1.0
1.5
2.0
2.5
0
2
4
6
8
Pol II Ratio
Hour
II+/IV+
II+/V+
0
10
20
30
40
50
4
5
6
7
8
9
0
2
4
6
8
Flow Rate (mL/min)
log
10
CFU/mL
Hour
Pol II+
Pol IV+
Pol V+
Flow Rate
0
10
20
30
40
50
4
5
6
7
8
9
Flow Rate (mL/min)
log
10
CFU/mL
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0
2
4
6
8
10
Pol II
+
Ratio
Hour
A
B
74
Figure 3.7. Fluctuations in Pol II
+
relative fitness during a chemostat competition
with an average flow rate of 2.3 volumes/hour.
All three polymerase double-mutant strains were mixed and competed during continuous
culture within a chemostat to assess relative fitness. (A) The population density of each
strain is shown throughout the competition, along with the flow rate of fresh media
(mL/min). Strains are indicated by line color: Pol II
+
, red; Pol IV
+
, green; Pol V
+
. blue.
Flow rate is indicated in black. (B) The relative fitness of Pol II
+
compared to Pol IV
+
(in
green) and Pol V
+
(in blue) is shown over time for this competition, which had an
average dilution rate of 2.3 volumes/hour upon initiating flow of nutrients.
0.0
0.5
1.0
1.5
2.0
2.5
0
2
4
6
8
Pol II Ratio
Hour
II+/IV+
II+/V+
0
10
20
30
40
50
4
5
6
7
8
9
0
2
4
6
8
Flow Rate (mL/min)
log
10
CFU/mL
Hour
Pol II+
Pol IV+
Pol V+
Flow Rate
0
10
20
30
40
50
4
5
6
7
8
9
Flow Rate (mL/min)
log
10
CFU/mL
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0
2
4
6
8
10
12
Pol II
+
Ratio
Hour
A
B
75
Figure 3.8. Fluctuations in Pol II
+
relative fitness during a chemostat competition
with an average flow rate of 2.5 volumes/hour.
All three polymerase double-mutant strains were mixed and competed during continuous
culture within a chemostat to assess relative fitness. (A) The population density of each
strain is shown throughout the competition, along with the flow rate of fresh media
(mL/min). Strains are indicated by line color: Pol II
+
, red; Pol IV
+
, green; Pol V
+
. blue.
Flow rate is indicated in black. (B) The relative fitness of Pol II
+
compared to Pol IV
+
(in
green) and Pol V
+
(in blue) is shown over time for this competition, which had an
average dilution rate of 2.5 volumes/hour upon initiating flow of nutrients.
0.0
0.5
1.0
1.5
2.0
2.5
0
2
4
6
8
Pol II Ratio
Hour
II+/IV+
II+/V+
0
10
20
30
40
50
4
5
6
7
8
9
0
2
4
6
8
Flow Rate (mL/min)
log
10
CFU/mL
Hour
Pol II+
Pol IV+
Pol V+
Flow Rate
0
10
20
30
40
50
4
5
6
7
8
9
Flow Rate (mL/min)
log
10
CFU/mL
0
2
4
6
8
10
12
14
0
2
4
6
8
10
Pol II
+
Ratio
Hour
A
B
76
Figure 3.9. Fluctuations in Pol II
+
relative fitness during a chemostat competition
with a flow rate of 4.0 volumes/hour.
All three polymerase double-mutant strains were mixed and competed during continuous
culture within a chemostat to assess relative fitness. (A) The population density of each
strain is shown throughout the competition, along with the flow rate of fresh media
(mL/min). Strains are indicated by line color: Pol II
+
, red; Pol IV
+
, green; Pol V
+
. blue.
Flow rate is indicated in black. (B) The relative fitness of Pol II
+
compared to Pol IV
+
(in
green) and Pol V
+
(in blue) is shown over time for this competition, which had a constant
dilution rate of 4.0 volumes/hour upon initiating flow of nutrients.
0.0
0.5
1.0
1.5
2.0
2.5
0
2
4
6
8
Pol II Ratio
Hour
II+/IV+
II+/V+
0
10
20
30
40
50
4
5
6
7
8
9
0
2
4
6
8
Flow Rate (mL/min)
log
10
CFU/mL
Hour
Pol II+
Pol IV+
Pol V+
Flow Rate
0
10
20
30
40
50
4
5
6
7
8
9
Flow Rate (mL/min)
log
10
CFU/mL
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0
1
2
3
4
5
6
7
Pol II
+
Ratio
Hour
A
B
77
3.5 Discussion
Microbial communities must often cope with dynamic and unpredictable changes in
nutrient availability and environmental stress. The capacity to quickly adjust and thrive
during intermittent cycles of feast and famine is crucial to evolutionary fitness.
Previously, we have demonstrated the importance of alternative DNA polymerases in
generating beneficial alleles that facilitate adaptation and long-term evolution. Here we
show these enzymes also have important physiological roles that help enable cells to
quickly retool and capitalize on environments conducive to growth by facilitating DNA
synthesis and rapid growth.
Within serial passage competitions, as cells experience a daily nutritional upshift, the
capacity to resume growth and facilitate efficient replication faster than competing strains
can confer a significant selective advantage. As observed throughout the first days of
serial passage competition, the Pol II
+
strain demonstrated an advantage compared to
both Pol IV
+
and Pol V
+
(Figure 3.1B&C), suggesting the capacity to express Pol II can
enable cells to adjust and grow faster during serial passage. While the effects of this Pol
II physiological advantage dissipated over time during serial passage, this is likely the
effect of beneficial alleles conferring a novel advantage arising faster within the Pol V
+
and Pol IV
+
populations, respectively. The relative timing of Pol II
+
populations
becoming the minority against Pol V
+
and Pol IV
+
is consistent with this hypothesis given
that their previously characterized in vivo mutation frequencies (Figure 2.9) would
indicate Pol V
+
could generate a beneficial allele sooner than Pol IV
+
.
78
The initial physiological advantage of Pol II during serial passage similarly manifested
itself within chemostat competitions. We demonstrate for the first time that the Pol II
+
strain dominated both Pol IV
+
and Pol V
+
within all chemostat competitions (Figure 3.2).
As average dilution rates increased to necessitate faster growth rates, the average Pol II
advantage also increased. Accordingly, these findings demonstrate the Pol II
+
strain was
capable of growing faster than the Pol IV
+
and Pol V
+
strains. While this physiological
advantage could be the manifestation of many non-mutually exclusive possibilities, we
speculate that this physiological advantage is the result of Pol II acting to enable cells to
replicate more efficiently and facilitate faster growth within nutrient rich conditions. This
conclusion is bolstered by a previously characterized role for Pol II during replication
restart following DNA damage and the cessation of stress (Rangarajan et al., 2002).
Although initially discovered in 1970 (Knippers, 1970), the role of Pol II during DNA
repair and replication remained enigmatic without any characterized in vivo phenotypes
until nearly three decades later when Pol II was first implicated in catalyzing “replication
restart” following UV irradiation (Rangarajan et al., 1999, 2002). Following the
introduction of DNA damage via UV irradiation, wildtype cells normally exhibit a brief
~10 minute pause in replication before resuming growth. However, deletion of Pol II
results in an extended delay (~50 minutes) before DNA synthesis resumes. Since then it
has also been discovered that cells experiencing DNA damage during stationary phase
79
exhibit a delay in replication restart, however this delay is only observed following a
nutritional upshift and the resumption of growth (Murli et al., 2000).
Similar to replication restart following UV-induced DNA damage, Pol II also appears to
confer a physiological advantage during serial passage competitions as cells resume
growing after experiencing a nutritional upshift following the comparatively stressful
conditions of stationary phase. If, indeed, the capacity to express Pol II enables cells to
facilitate replication restart, resume DNA synthesis faster, and replicate sooner following
stationary phase stress, it would confer a significant advantage during serial passage
competitions. Thus, replication restart offers a compelling mechanistic explanation for
the observed physiological advantage of Pol II.
Surprisingly, the relative abundance of the Pol II
+
strain appeared to peak at regular
intervals over the course of each chemostat competition (Figures 3.5-3.9). This finding
appears to reflect modulation of alternative polymerase expression and regulation.
Recent single-cell studies investigating the SOS and other stress responses have shown
transient bursts of activity occurring in discrete and repetitive peaks of induction
(Friedman et al., 2005; Krishna et al., 2007; Locke et al., 2011). Evolutionarily, these
iterative bursts of expression could provide a method of limiting the potentially harmful
mutagenic effects of error-prone polymerases by enabling cells to attempt error-free
repair and assess whether template lesions have been resolved before inducing further
expression. However, within the context of our chemostat competitions we speculate that
80
the observed peaks in relative Pol II
+
performance might be the result of discrete peaks in
SOS induction that can temporarily inhibit growth.
Similar to the aforementioned role of Pol II during replication restart, uncleaved UmuD
and UmuC have also been implicated in the process of resuming DNA synthesis
following UV irradiation; however, unlike Pol II, these proteins act to cause a delay in
the resumption of DNA replication (Opperman et al., 1999). Following UV exposure,
uncleaved UmuD remains the predominant form of the protein for approximately 20
minutes before RecA-mediated processing into UmuD’ and the formation of active Pol V
(UmuD’
2
C) which can resume DNA replication via translesion synthesis. Importantly,
this delay in replication restart also occurs within cells that experienced UV damage
during stationary phase by interacting with Fis to delay replication following a nutritional
upshift (Murli et al., 2000). After making this finding, the authors proposed a model
wherein prior to UmuD cleavage these proteins act to create a delay before resuming
replication following UV exposure. This pause would serve to provide additional time to
allow nucleotide excision repair to repair any DNA damage before resuming replication.
In this manner, cells could minimize the formation of excessive mutations by repairing
UV-induced DNA damage, then following this delay any remaining lesions could be
resolved with the error-prone DNA polymerase.
Within an ecological context over an evolutionary timescale this cell-cycle control
function could serve as a multi-level regulatory mechanism to maintain genomic integrity
81
by first preventing DNA damage from being lethal by halting replication, then allowing
the chance for error-free repair, before finally using error-prone replication to resolve any
residual lesions (Opperman et al., 1999). However, within the context of serial passage
and chemostat competitions, this delay in the resumption of replication could serve to
impede rapid growth and manifest as a decrease in relative fitness. Therefore, this
pausing mechanism also offers a potential explanation of the periodic peaks in population
density relative to Pol II since periodic bursts of SOS induction could create periodic
pauses in replication within the Pol V
+
background.
Another factor contributing to the formation of peaks might be transient bursts of Pol II-
mediated replication. While initially characterizing the role of Pol II during replication
restart, Rangarajan et al. (1999) also observed a transient “burst” in DNA synthesis
within a temperature-sensitive background following a transition to non-permissive
temperatures. They found that after an initial burst of DNA synthesis following
replication restart the rate of dNTP incorporation drops again after 10-15 minutes,
suggesting under normal conditions another polymerase switch occurs with Pol III
replacing Pol II. This rapid burst in DNA synthesis followed by another delay in
replication could also contribute to periodic shifts in relative Pol II
+
performance within
chemostat competitions. Here, within the Pol II
+
strain, transient peaks in SOS induction
could create transient bursts of DNA synthesis that enable brief periods of rapid growth
followed by slower growth, which would enable Pol IV
+
and Pol V
+
strains to catch-up in
relative density.
82
In this manner, the roles of both Pol II and Pol V during replication restart act in
coordination as opposing forces that can contribute to peaks in relative abundance.
During peaks of SOS induction, the Pol II
+
strain would have an initial advantage by
quickly initiating replication restart, however this burst in DNA synthesis might be
transient and allow competing strains to catch-up. Meanwhile, the Pol V
+
strain would
have an initial disadvantage due to a UmuD-mediated pause in replication that the Pol II
+
strain wouldn’t experience. Similarly, although the Pol IV
+
strain would also not
experience this UmuD-mediated delay, replication restart experiments suggest the Pol
IV
+
stain might be at an even greater disadvantage. This is because following DNA
damage the delay of DNA synthesis was further exacerbated (~90 minutes) in double-
mutant ΔpolBΔumuDC strains (Rangarajan et al., 1999), which would put the Pol IV
+
strain at an even greater disadvantage. Interestingly, this temporal cascade in resuming
DNA synthesis during replication restart corresponds with our previous GASP
competition hierarchy of fitness (Pol II
+
> Pol V
+
> Pol IV
+
) (Figures 2.3&2.4), providing
a compelling causal mechanism for our previous observation.
Regardless of the molecular mechanism, our serial passage and chemostat competitions
clearly demonstrate a physiological advantage conferred by Pol II. Upon experiencing a
nutritional upshift within serial passage and during rapid growth within chemostats Pol II
enables faster growth, presumably by enabling rapid DNA synthesis and replication.
These findings further emphasize the importance and relative contributions of each
83
alternative DNA polymerase during periods of rapid growth, and demonstrate their
important physiological roles enabling microbial populations to quickly adjust and
capitalize during conditions suitable for growth within dynamic environments.
84
Chapter 4: Stationary phase conditions induce expression of alternative
DNA polymerases in the absence of exogenous stressors
4.1 Abstract
Alternative DNA polymerases have long been recognized for their physiological roles in
facilitating replication past DNA damage introduced by potentially lethal UV radiation,
chemical mutagens, and other forms of exogeneous stress. However, these error-prone
polymerases are also implicated in the formation of genetic diversity that can profoundly
influence adaptation and the evolutionary fitness of microbial communities. To further
assess their significance within environmental conditions experienced by natural
populations, the relative abundance and expression pattern of each alternative polymerase
was determined throughout long-term incubation in the absence of artificial SOS-
induction. We find the abundance of transcripts encoding all three E. coli alternative
DNA polymerases (polB, dinB, umuC, and umuD) dramatically increases early during the
transition from exponential growth to stationary phase, while the catalytic domain of the
high-fidelity Pol III (dnaE) exhibits a corresponding decrease in abundance. Furthermore,
transcription levels of each alternative polymerase remain elevated throughout long-term
incubation, with Pol IV transcript consistently exceeding that of dnaE. Surprisingly, early
induction of alternative polymerase expression appears to occur in an SOS-independent
manner, whereas their abundance during long-term stationary phase coincides with peaks
of SOS induction. These findings reveal that alternative DNA polymerases are induced in
the absence of exogenous stressors under conditions akin to those experienced in nature,
85
suggesting their contributions to comparatively error-prone replication may reflect the
status quo within microbial communities.
4.2 Introduction
Molecular mechanisms influencing the formation of genetic diversity can have profound
effects on the evolutionary fitness of microbial communities (Bjedov et al., 2003b; Chao
and Cox, 1983; Pigliucci, 2008; Saint-Ruf and Matic, 2006; Woods et al., 2011). Widely
recognized for their roles during translesion synthesis, error-prone DNA polymerases can
also serve to introduce beneficial alleles within bacterial population that promote
adaptation and evolution (Foster, 2007; Galhardo et al., 2007; Kunkel, 2004; Yeiser et al.,
2002). However, given their capacity to introduce potentially deleterious mutations, it is
perhaps not surprising that these enzymes are also highly regulated. While all three
alternative polymerases can be expressed under a variety of environmental stresses
(Layton and Foster, 2003, 2005; Stumpf and Foster, 2005; Yeiser et al., 2002), they have
been primarily characterized following their induction by the SOS response, and are
widely referred to as SOS DNA polymerases (Courcelle et al., 2001; Friedberg, 2006;
Nohmi, 2006; Yang and Woodgate, 2007).
The effects of a two-component repressor and activator system of LexA and RecA
proteins regulate the E. coli SOS response (Bichara et al., 2011; Fuchs et al., 2004;
Goodman, 2002; Nohmi, 2006; Tippin et al., 2004). While LexA normally suppresses
expression by binding to a consensus sequence in the operator region of SOS genes,
86
DNA damage can create regions of single-stranded DNA during replication that are
quickly bound by RecA which can become activated and induce the autocatalytic
cleavage of the LexA repressor. Cleavage of LexA then derepresses approximately 40
genes within the SOS regulon, including all three alternative DNA polymerases (polB,
dinB, and umuDC), as well as both recA and lexA. Accordingly, this induction serves as a
feedback loop increasing both regulators, enabling LexA pools to rise following the
resolution of DNA damage and repress further induction.
The basal expression and timing of induction for genes within the SOS regulon is
contingent on the sequence of their consensus LexA binding sequence (5’-
TACTG(TA)
5
CAGTA-3’) (Fernandez De Henestrosa et al., 2000; Walker, 1984).
Depending on the affinity of the LexA repressor to the LexA recognition sequence,
specific SOS genes can experience varying levels of repression that can result in a
temporal cascade of gene induction. Following induction of the SOS response by UV
irradiation, polB (Pol II) is among of the earliest genes induced (<1 minute), wheras
umuD and umuC (Pol V, ~45 minutes) are among the last along with the replication and
cell division inhibitors sbmC and sulA, respectively (Courcelle et al., 2001; Kenyon and
Walker, 1980; Kim et al., 2001; Qiu and Goodman, 1997; Woodgate and Ennis, 1991).
While we have known since the 1970s that UV irradiation and various exogenous
chemical agents interfering with replication or introducing DNA damage can elicit the
SOS response (Radman, 1975), relatively little attention has focused on the role of this
87
stress response within natural populations in the absence of exogenous stressors. Instead,
most in vivo studies have focused on the roles of Pols II, IV and V either in the survival
of rapidly dividing cells in the presence of DNA damage, or in a plate-based assay system
studying the process of “adaptive mutation” (Foster, 2005, 2007; Galhardo et al., 2007;
Roth et al., 2006). This is despite the fact that an increasing body of evidence suggests
this stress response and alternative DNA polymerases are important in nature. For
example, patterns of sequence variability within genes coding for alternative polymerases
from natural isolates indicated they exhibit levels of variation similar to those observed in
Pol I and Pol III, which suggests there is a strong selective pressure for maintaining their
function and specificity within natural environments (Bjedov et al., 2003a).
To investigate the extent and dynamics of alternative DNA polymerase expression under
conditions more akin to those experienced by microbial communities in nature, we have
monitored their relative abundance and induction throughout the five phases of the
bacterial life cycle (Finkel, 2006) as microbial populations encounter nutritional stress
and the build-up of potentially toxic metabolic byproducts during fierce competition for
limited resources. Accordingly, understanding the extent and dynamics of alternative
DNA polymerase expression during the transition to stationary phase and long-term
incubation, in the absence of exogenous stressors or mutagens, more accurately reflects
conditions experienced in nature, and can inform our understanding of their relative
contributions to genetic diversity and evolution.
88
We demonstrate that alternative DNA polymerases and the SOS response are induced
early and remain elevated throughout long-term stationary phase in the absence of
exogenous stressors. These findings emphasize the importance of this stress response
within the dynamic conditions experienced by environmental microbial communities, and
indicate error-prone DNA polymerases may facilitate a much greater proportion of
replication than previously recognized, reinforcing the significance of alternative DNA
polymerases in shaping microbial evolution.
4.3 Materials and Methods
4.3.1 Quantitative RT-PCR
Real-time PCR was used to determine the expression patterns of each alternative
polymerase gene, as well as induction of the SOS response. LB cultures inoculated with a
1:1000 dilution of an overnight population of wildtype E. coli K-12 strain ZK126
(W3110 ∆lacU169 tna-2) were incubated at 37ºC and periodically sampled for RNA
extraction. Samples were treated with RNAlater (Qiagen) and total RNA was isolated
using the RNeasy Mini Kit (Qiagen). qRT-PCR reactions with 100ng template RNA were
performed using the One-step RT-PCR Kit (Qiagen) with SYBR Green (Molecular
Probes) and amplified on an Opticon-2 Real-time PCR Cycler (MJ Research). Primers
used for amplification are shown in Table 2.2. Relative transcript abundance and changes
in gene expression were determined using the 2
-ΔCt
method (Livak and Schmittgen,
2001). Relative changes in gene expression were normalized to the metabolic
housekeeping gene gapA using the 2
-ΔΔCt
method (Livak and Schmittgen, 2001).
89
4.3.2 Mitomycin C dose response curve
Dose response curves were used to determine an appropriate concentration of mitomycin
C to induce artificial SOS induction without causing cell death. From dense overnight
cultures 5µL was used to inoculate 5.0mL LB cultures (1:1000) and incubated at 37ºC.
Optical density (OD
600
) and titer counts were monitored every 30 minutes. Following 2
hours of incubation, mitomycin C (Sigma) was introduced at increasing final
concentrations (0.1, 0.5, 1, 2, 5 and 10 µg/mL) within treated samples.
4.3.3 Artificial induction of the SOS response
In control experiments using artificial SOS induction, a 1.0 mg/ml mitomycin C stock
solution was added after 2 hours of incubation, at an OD
600
of ~0.1, to a final
concentration of 1.0 µg/mL. RNA was sampled hourly thereafter for quantitative RT-
PCR analysis, as described above.
4.4 Results
4.4.1 Stationary phase induces transcription of alternative polymerases and
additional SOS genes
Given their observed physiological and evolutionary impacts during periods of feast and
famine in the absence of endogenous stressors, we determined the expression patterns for
each alternative DNA polymerase during long-term incubation. Wildtype cultures were
sampled hourly for the first twelve hours and then every 24 to 48 hours over 14 days of
long-term stationary phase incubation. Quantitative RT-PCR on total cellular RNA
90
throughout long-term incubation revealed an induction of transcripts encoding all three
alternative polymerases (polB, dinB, umuC), as well as the SOS response gene sulA,
within the first hours of outgrowth (Figure 4.1).
Figure 4.1. Alternative DNA polymerases are induced upon entering stationary
phase.
mRNA transcript abundance was determined by qRT-PCR during the transition to
stationary phase in the absence of exogenous SOS inducers. Transcript abundance is
shown for each gene relative to its concentration after 2 hours of incubation.
0
1
2
3
4
2
3
4
5
6
7
8
Transcript abundance relative to Hour 2
Hour
gapA
polB
dinB
umuC
umuD
dnaE
sulA
recA
lexA
91
Relative to their abundance after 2 hours of incubation and entry into exponential phase
growth, by five hours, as the cells transition into stationary phase, the average abundance
of transcripts encoding Pol II and Pol IV more than doubled (dinB, +2.15-fold; polB,
+2.22-fold) and the catalytic subunit of Pol V was induced 50% by hour 7 (umuC, +1.51-
fold). Meanwhile, transcript levels for the dnaE-encoded catalytic subunit of Pol III
decreased by more than 40% within 4 hours (0.54-fold). Transcripts for the SOS
response regulators LexA and RecA increased slightly (lexA, +1.43-fold; recA +1.53-
fold) by hour 3, but quickly decreased and remained relatively flat during the first 8 hours
of incubation. Unexpectedly, abundance of the SOS-regulated cell division inhibitor sulA
also rose until hour 4 (+3.4-fold) before falling through hour 7.
This early induction of alternative DNA polymerase transcripts corresponded with the
transition between exponential growth and early stationary phase. For comparison,
expression levels of the early-exponential phase gene fis (Ball et al., 1992) peak at ~2
hours and decrease significantly from mid-logarithmic phase through stationary phase
while expression of the stationary phase-specific dps gene (Martinez and Kolter, 1997)
increases from late log through stationary phase (Figure 4.2).
92
Figure 4.2. Entry into stationary phase signaled by dps and fis.
Population density of wildtype cells inoculated 1:1000 (vol/vol) into LB is plotted during
the first 6 hours of growth in LB as cells enter stationary phase (circles). Transcript
abundance of dps (squares) and fis (triangles) was determined by qRT-PCR in the
absence of exogenous SOS inducers and transcript abundance relative to gapA is plotted.
4.4.2 Alternative DNA polymerases remain induced throughout long-term culture
In addition to early stationary phase, the abundance and expression patterns of each
alternative DNA polymerase was monitored throughout long-term stationary phase, along
with other SOS response genes. When the change in transcript abundance relative to hour
2 was normalized to the change in abundance of the metabolic housekeeping gene gapA,
part of the glycolysis pathway, the abundance and induction pattern of each gene could
be standardized and monitored relative to the nutritional status of the culture. Following
their initial induction within the first hours of incubation, all alternative DNA polymerase
genes remained elevated throughout long-term culture incubation (Figure 4.3).
dps
fis
Wildtype
0.001
0.01
0.1
1
10
6
7
8
9
10
0
1
2
3
4
5
6
7
Abundance relative to gapA
log
10
CFU/mL
Hour
93
Figure 4.3. Alternative polymerase transcript levels remain elevated throughout
long-term stationary phase.
mRNA transcript abundance, in the absence of exogenous SOS inducers, was determined
by qRT-PCR throughout long-term stationary phase. Transcript abundance is shown for
each gene normalized the metabolic housekeeping gene gapA. (A) Normalized
abundance of polymerase genes. (B) Normalized abundance of SOS response genes.
94
When normalized to changes in gapA transcript abundance, the relative abundance of
each alternative DNA polymerase continued to rise during stationary phase (Figure
4.3A). During the first day of growth, the normalized change in abundance of each
alternative DNA polymerase abundance peaked at hour 10 (polB, +8.30-fold; dinB,
+5.46-fold; umuC, +4.88-fold; umuD, +7.17-fold). Unlike the alternative DNA
polymerases, the normalized abundance of dnaE remained comparatively flat. This day 1
peak in normalized abundance corresponded with those of both lexA (+3.41-fold) and
recA (+3.41-fold), as well as the SOS response gene sulA (+11.69-fold) which exhibited
the greatest relative induction of all genes analyzed (Figure 4.3B).
Following the first day of incubation, alternative polymerase abundance remained
elevated throughout long-term stationary phase. However, the normalized abundance of
each polymerase exhibited discrete peaks over time. The greatest peaks in relative
polymerase expression were observed on days 3 and 5, which corresponded with elevated
levels of lexA, recA, and sulA.
4.4.3 Alternative DNA polymerase transcripts exceed Pol III transcript during
stationary phase
When comparing the relative abundance of dnaE, polB, dinB, and umuC transcripts
(Figure 4.4) upon entry into early stationary phase we observed a dramatic decline in the
average proportion of dnaE. The relative abundance of dnaE transcript quickly dropped
from ~69% during logarithmic growth (hour 2) to 34% upon entry into stationary phase
(hour 5). This decline was accompanied by a corresponding increase in the relative
95
proportion of dinB. Within 24 hours the dinB transcript became most abundant (56%),
significantly exceeding that of dnaE (31%), and this dominance continued throughout
long-term culture. Throughout all timepoints there was a consistent trend in the hierarchy
of transcript abundance among alternative polymerases, with dinB > polB > umuC. While
umuC typically remained scarce, there was a dramatic peak in abundance on days 3 and 5
where it exceeded both dinB and dnaE.
Figure 4.4. Alternative polymerase transcript proportion remains elevated
throughout long-term stationary phase.
mRNA transcript abundance, in the absence of exogenous SOS inducers, was determined
by qRT-PCR throughout long-term stationary phase. The average proportion of
polymerase transcripts over time, expressed as a percentage of total transcript abundance,
is shown for four representative transcripts: Pol III, dnaE, black diamonds; Pol IV, polB,
green triangles; Pol II, polB, red circles; Pol V, umuC, blue squares.
96
While considering the relative proportion of each alternative DNA polymerase, it is also
informative to compare the proportional change in transcript abundance. When the
change in proportion of each polymerase is normalized during the initial entry into
stationary phase, the relative proportion of polB increases sooner and to the greatest
extent, followed by umuC, and finally dinB (Figure 4.5). Accordingly, although dinB
remains the most prevalent alternative polymerase transcript, polB is induced
proportionally faster and to a greater extent.
Figure 4.5. Pol II transcript increases proportionally greater and faster than other
alternative DNA polymerases.
The change in average polymerase proportion was determined during the transition to
stationary phase, and is plotted over time for each gene: Pol III, dnaE, black diamonds;
Pol IV, polB, green triangles; Pol II, polB, red circles; Pol V, umuC, blue squares.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
2
3
4
5
6
Change in polymerase proportion
Hour
dnaE
polB
dinB
umuC
97
4.4.4 Stationary phase induces moderate levels of the SOS response compared to
mitomycin C
To assess the relative amount of induction attributable to the conditions experienced
during stationary phase compared to those observed during artificial SOS induction, a
series of control experiments were conducted using mitomycin C (MMC). Initial dose-
response experiments were conducted to determine an appropriate concentration of MMC
to elicit a sufficient SOS response to observe an effect without excessive inhibition of cell
growth or lethality. Upon introducing MMC following 2 hours of growth, a concentration
of 1.0 µg/mL was found to slow cell growth and reduce overall population density
without resulting in cell death (Figure 4.6).
Figure 4.6. Mitomycin C dose response curves.
Wildtype cells were treated with increasing concentrations of mitomycin C (MMC) to
assess the affect on growth and survival. (A) Optical density of cultures treated with
MMC. (B) Cell survival following MMC treatment. Control populations without MMC
treatment are shown in black, and increasing concentrations of MMC are denoted by
darker shades of blue. Red arrows signify the addition of MMC at hour 2.
98
This method of chemical SOS induction was then utilized on cultures entering stationary
phase, and qRT-PCR was employed to assess the relative extent of induction attributable
to stationary phase in the absence of exogenous stressors (Figure 4.7). As controls, the
abundance of genes not under SOS regulation were also assessed (gapA, fis, and dps).
SOS induction had no effect on gapA transcript abundance, and the early exponential
gene fis and early stationary phase gene dps remained unaffected by MMC (Figure 4.7A).
However, each gene encoding an alternative DNA polymerase was induced to a greater
extent following MMC treatment compared to the effect of entering stationary phase
(Figure 4.7D-F). This trend was also observed for the well-characterized SOS response
genes sulA and sbmC (Figure 4.7B). As expected, MMC treatment also elicited a strong
induction of both lexA and recA (Figure 4.7C). However, unlike the other genes known to
be responsive to SOS induction, neither lexA nor recA appeared to be induced within the
untreated stationary phase samples over this timeframe. On average, entry into stationary
phase (hour 5) induced the expression of each alternative DNA polymerase gene to levels
approximately 25% of those observed during MMC induction (Figure 4.8).
99
Figure 4.7. Effect of stationary phase and mitomycin C treatment on gene
expression.
mRNA transcript abundance was determined using qRT-PCR on cells entering stationary
phase with or without the addition of mitomycin C (MMC). Transcript abundance
relative to untreated gapA abundance is shown over time. Transcripts from cells treated
with MMC are denoted with circles, while transcripts from untreated cell are triangles.
Treated and untreated gapA is shown compared to (A) fis and dps. Change relative to
untreated gapA is shown for (B) sulA and sbmC, (C) recA and lexA, (D) polB, (E) dinB,
and (F) umuC and umuD.
100
Figure 4.8. Gene induction following mitomycin C treatment versus stationary
phase.
mRNA transcript abundance was determined using qRT-PCR on cells entering stationary
phase with or without the addition of mitomycin C (MMC). The amount of transcript
abundance relative to untreated expression at hour 2 was calculated, and the ratio of
MMC-treated transcript abundance relative to untreated transcript abundance is shown
over time. (A) Alternative DNA polymerase (polB, dinB, umuC, umuD) induction is
compared to gapA. (B) Induction of SOS response genes (sbmC, sulA, recA, lexA) is
compared to non-SOS response genes (gapA, dps, fis).
101
4.5 Discussion
It has been known since the 1970s that UV irradiation and various exogenous chemical
agents interfering with replication or inducing DNA damage could elicit the E. coli SOS
response (Radman, 1974; Radman, 2007). A resurgence of interest and investigation has
revealed this SOS-mediated induction of alternative DNA polymerases can have
important ramifications for cell survival and evolution, yet the extent and implications of
this stress response within natural environmental populations has remained elusive. Here,
we show that the SOS response and alternative DNA polymerases are induced early and
remain elevated throughout long-term stationary phase in the absence of exogenous
stressors. These findings emphasize the importance of this stress response within the
dynamic conditions experienced by environmental microbial communities, and reinforce
the significance of alternative DNA polymerases in shaping microbial evolution.
Using qRT-PCR to assess the relative abundance and patterns of gene expression upon
entry into stationary phase, we demonstrate the rapid induction of each alternative DNA
polymerase early within the first hours of growth. Corresponding with the rapid increase
in transcription of dps signifying entry into early stationary phase, we observe a dramatic
increase in the overall abundance of transcripts encoding all three alternative polymerases
(Figure 4.1), as well as their induction normalized to a metabolic housekeeping gene
(Figure 4.3). This initial induction occurred early during outgrowth, commencing even
before the end of exponential growth, such that transcript abundance levels first peaked at
102
the same time population density plateaued. Across the same timespan, the abundance of
the well-characterized SOS response gene sulA also increased.
Surprisingly, this initial induction occurred with comparatively modest levels of lexA and
recA transcript. While mature LexA and RecA proteins mediate the SOS response, we
observed the overall transcript abundance of these SOS regulators initially increased only
slightly (Figure 4.1). Once normalized to changes in the abundance of gapA, both
regulators appeared to remain uninduced during late exponential growth and entry into
stationary phase (Figure 4.3). Only later, beginning at hour 7, when population growth
had plateaued were both lexA and recA induced. This increase in relative abundance
peaked at hour 10, corresponding with similar secondary peaks in the abundance of all
other SOS response genes (polB, dinB, umuC, umuD, and sulA). Accordingly, it appears
the initial increase in the alternative polymerases and sulA might occur in the absence of
a full-fledged SOS response. The initial increase in transcription of these SOS response
genes without an accompanying induction of lexA and recA was also observed within the
control mitomycin C experiments, where expression of both regulators remained flat in
the absence of MMC-mediated SOS induction (Figure 4.7C). These findings indicate that
these genes, typically associated as part of the SOS response, might also become induced
via an SOS-independent mechanism during entry into stationary phase before population
density plateaus.
103
Recently, an analysis of Pol IV expression observed regulation of dinB attributable to the
effects of the general stress-response sigma factor RpoS in addition to SOS-mediated
induction (Layton and Foster, 2003). Strains with a disruption in rpoS produced a 3-fold
decrease in the amount of dinB transcript. While this phenomenon was observed during
growth in minimal media, the effect of RpoS-mediated regulation on dinB transcript was
observed during late stationary phase after RpoS levels reach their maximum in early
stationary phase. Here, using strains with functional RpoS, we observed an approximately
3-fold increase in dinB as well as the other alternative DNA polymerases. It is tempting
to speculate that RpoS, rather than an induction in the SOS response, might mediate this
initial peak of induction.
Similarly, the DNA gyrase inhibitor sbmC is also known to be co-regulated by both LexA
and RpoS (Baquero et al., 1995; Oh et al., 2001). Oh et al. observed an increase in sbmC
expression between exponential and stationary phase in wildtype cells that was reduced
4.8-fold within an rpoS mutant. Within our MMC-induced control experiments we also
monitored sbmC and observed an induction during the first 5 hours in untreated cells that
was approximately 5-fold less than those observed with artificial SOS induction.
Together, these observations suggest the initial induction of genes observed during entry
into stationary phase might be lexA- and recA-independent. Instead, the initial induction
of alternative polymerases and other stress response proteins might be mediated by RpoS
rather than the SOS response.
104
Following the initial induction of alternative DNA polymerases and other SOS genes
during the transition to stationary phase, we continued to observe elevated levels of
normalized abundance throughout long-term stationary phase (Figure 4.3). Interestingly,
while the level of each SOS gene remained elevated relative to levels during exponential
growth, we observed periodic fluctuations in their relative abundance. Unlike the first
peak in expression, these secondary stationary phase peaks coincided with elevated levels
of both lexA and recA, consistent with full-fledged SOS-mediated induction.
When comparing the abundance and induction of each alternative DNA polymerase to
the housekeeping high-fidelity Pol III we found a dramatic shift in relative polymerase
transcription throughout long-term incubation. During exponential growth dnaE
transcripts dominated transcript abundance, but the average proportion of transcripts
dropped quickly by hour 5 as cells transitioned to stationary phase. This decrease
coincided with a dramatic increase of dinB, with Pol IV transcript remaining the most
abundant throughout long-term stationary phase. While less dramatic with regards to their
overall proportion, the proportion of polB and umuC also rose quickly during entry into
stationary phase. In fact, in terms of proportional increase, polB and umuD were induced
faster and to a greater extent than dinB (Figure 4.5).
This proportional increase in alternative DNA polymerase transcription has far-reaching
implications with regards to our understanding of replication during stationary phase and
long-term incubation. With greater relative abundance, the alternative DNA polymerases
105
have a better chance at binding the β-clamp, becoming associated with the replication
machinery. Therefore, there is an increased likelihood of alternative polymerases
performing a greater proportional role in replication, and suggests the conditions
encountered during long-term stationary phase promote comparatively error-prone DNA
synthesis. While the importance of alternative DNA polymerases in generating diversity
during stress continues to garner greater appreciation, our results further substantiate their
significance during conditions without the addition of exogenous stressors of the artificial
introduction of DNA damage, more akin to conditions experienced within natural
environments.
As microbial communities encounter nutritional stress and build-up of potentially toxic
metabolic byproducts during competition for limited resources, they experience
conditions similar to those found during entry into stationary phase and long-term
incubation. The rapid induction of alternative DNA polymerases observed here not only
serves to facilitate translesion synthesis past damage that might otherwise inhibit efficient
replication, but also increases the likelihood of mutations forming. Given the abundance
of alternative DNA polymerases remains elevated throughout long-term stationary phase,
it appears this comparatively error-prone replication more accurately reflects the status
quo within natural occurring populations. Accordingly, our findings implicate these
stress-induced error-prone DNA polymerases in facilitating a much greater proportion of
replication than previously appreciated, significantly influencing our understanding of the
106
molecular mechanisms influencing the formation and overall abundance of genetic
diversity within nature.
107
Chapter 5: Conclusions and Discussion
Mutation is the ultimate source of all genetic variation: providing the raw material for
natural selection, and serving as a catalyst of evolution. Therefore, understanding the
molecular mechanisms by which allelic variation is generated can offer fundamental
insights into the process of adaptation and evolution. The work presented here illustrates
how alternative DNA polymerases serve a vital physiological role in cell survival while
simultaneously generating genetic diversity that shapes microbial evolution.
Cumulatively, the results presented here illustrate and elucidate the balance between the
dual roles of physiological success and evolutionary fitness mediated by alternative DNA
polymerases. We have shown that DNA polymerase II is most vital to physiological
fitness and efficient replication during periods of rapid growth within batch culture and
chemostat competitions. Meanwhile, DNA polymerase IV and V are comparatively more
important to evolutionary fitness throughout long-term culture by introducing significant
levels of genetic diversity. Together, each alternative DNA polymerase serves a specific
role balancing the physiological and evolutionary demands throughout changing
environmental stresses, with all three contributing to evolutionary fitness.
5.1 Fitness during fluctuating feast and famine
Natural environments are highly dynamic and unpredictable, with nutrient availability
fluctuating between periods of feast and famine. Over evolutionary timescales, successful
108
organisms must 1) survive and endure during periods of stress, 2) take advantage of
conditions suitable for growth, and 3) remain adaptable to changing environmental
conditions. Together, our results illustrate how alternative DNA polymerases mediate
each of these demands within microbial communities, as summarized in Figure 5.1.
Figure 5.1. Fitness during feast and famine.
The alternative DNA polymerases mediate important physiological and evolutionary
roles as nutritional conditions and environmental stresses change throughout the bacterial
lifecycle. During periods of both feast and famine, the expression of each alternative is
crucial to long-term competitive success. When nutritional resources are readily
available, Pol II confers a significant physiological advantage and enables cells to “strike
while the iron is hot” by facilitating efficient and rapid growth during outgrowth. As
nutrients become depleted and environments become increasingly stressful during long-
term stationary phase, populations adopt a strategy of “mutate or die” and the more error-
prone Pol IV and Pol V promote the generation of genetic diversity and the formation of
adaptive alleles.
Hours!
0! 6! 12!
Days!
2! 6! 10! 14!
4!
6!
8!
10!
5!
7!
9!
log
10
CFU/mL!
Pol 2! Pol 4 & Pol 5 !
“Mutate or die”
“Strike while the
iron is hot”
“Express for success”
109
Microbial communities in nature face innumerable forms of both endogenous and
exogenous stresses capable of introducing potentially lethal DNA damage. While the
genomic integrity of every bacterium is under perpetual assault, alternative DNA
polymerases offer a molecular mechanism to tolerate this DNA damage. By enabling
translesion synthesis, the expression of Pol II, IV, and V is crucial for long-term success
by enabling cells to mitigate the effects of DNA damage and survive within stressful
environmental conditions.
When these stresses are alleviated and nutrients become readily available, alternative
DNA polymerases serve to facilitate efficient DNA synthesis enabling rapid growth.
During these brief periods of feast, bacteria capable of replicating fastest can best
capitalize on favorable conditions and “strike while the iron it hot” to increase in
frequency within microbial communities. By preventing excessive stalling of replication
forks, each of the alternative DNA polymerases can ensure the capacity to resume
replication. Here, the roles of DNA polymerase II appeared most vital to physiological
fitness. We discovered the capacity to express Pol II conferred a significant advantage
during periods of rapid growth within outgrowth, serial passage, and chemostat
competitions. Presumably this physiological advantage stems from the capacity of Pol II
to facilitate rapid replication restart and catalyze efficient DNA synthesis, giving cells the
ability to jumpstart and maintain faster growth.
110
As nutrients inherently become limiting with any population, environmental conditions
inevitably return to those of famine. Here again, alternative DNA polymerases confer a
significant advantage through the generation of genetic diversity. As alternative DNA
polymerases became induced and performed a greater proportion of replication during the
stressful conditions of stationary phase, they were shown to be capable of introducing
significant quantities of genetic variation. While often deleterious, the occasional
beneficial allele can enable individuals within a population to perform better, and confer
a selective advantage relative to competitors. As evidenced throughout long-term
stationary phase and GASP competitions, the more error-prone Pol IV and Pol V
conferred a significant advantage as populations adopt a strategy of “mutate or die”
during the ever-changing conditions of long-term cultures. Accordingly, we demonstrate
the alternative DNA polymerases also enable microbial populations to adapt and thrive
within dynamic and unpredictable conditions like those experienced in nature.
5.2 Stress-induced mutagenesis: The modulation of mutation
As demonstrated, each alternative DNA polymerase serves specific and vital
physiological and evolutionary roles throughout the bacterial lifecycle; however, our
results also attest to the exquisite regulation of these enzymes and contribute to our
understanding of the rate and tempo of bacterial evolution. Mutations introduced during
replication account for the majority of overall genetic diversity within bacterial
populations, and perturbations in the fidelity of replication directly influence the
adaptation and evolutionary fitness of microbial communities as they encounter dynamic
111
and unpredictable changes in environmental conditions. Accordingly, the mechanisms
and extent to which DNA polymerases introduce genetic variation during replication
remains critical to understanding the long-term survival and evolution of bacterial
populations.
Given that many molecular mechanisms influencing mutation rates, including alternative
DNA polymerases, are genetically heritable, it should come as no surprise that rates of
mutagenesis are themselves under strong selective pressures. Indeed, E. coli has evolved
an elaborate system of regulation controlling the potentially deleterious effects of
mutagenic error-prone polymerases. As illustrated by our observations throughout the
bacterial lifecycle, we have demonstrated the induction of all three alternative DNA
polymerases upon the transition to stationary phase, and the presence of periodic peaks in
relative abundance. These dynamics of alternative polymerase expression are consistent
with our understanding of selective pressures influencing optimal mutation frequencies,
and expand our understanding of bacterial evolution in nature.
Optimal rates of mutagenesis, dictating overall levels of genetic diversity, can vary with
population structure and environmental conditions. Elevated rates of mutation can be
advantageous within highly dynamic conditions, as it can be critical to maintain sufficient
diversity within a population to survive strong selective pressures. However, it is also
vital to prevent excessive mutagenesis given that the vast majority of mutations are
deleterious. This is particularly true within relatively stable and nutrient-rich
112
environments, since deleterious mutations will often create a relative burden that will
slow or inhibit further growth. Therefore, the rate of mutation can have enormous
implications on overall evolutionary fitness, and optimal rates of mutation can vary in a
condition-specific manner.
The evolutionary pressures influencing optimal rates of mutation are also at play within
the context of our long-term culture experiments, and illustrate the exquisite regulation of
alternative DNA polymerases. As summarized in Figure 5.2, the roles and relative
importance of each alternative DNA polymerase throughout the bacterial lifecycle reflect
these evolutionary pressures. When first inoculated into fresh medium, bacterial cultures
experience feast conditions that promote rapid growth. Cells are experiencing a relatively
stable environment with comparatively little stress. Accordingly, under these growth
conditions one can understand why it can be beneficial to have low levels of mutation: as
the saying goes, “don’t fix what isn’t broken.” Therefore, it is perhaps unsurprising that
the use of DNA polymerases more prone to making potentially deleterious mutations
would be avoided, and expression of the alternative DNA polymerases would remain
low. In the event that lesions in template DNA are encountered that require translesion
synthesis within these conditions of feast, it appears cells continue to avoid excessive
mutagenesis and utilize Pol II to facilitate rapid growth. Given that Pol II has the greatest
fidelity among alternative polymerases, our findings suggest microbial communities may
have evolved to maintain comparatively low mutation rates during conditions of rapid
growth.
113
Figure 5.2. The modulation of mutation.
As nutritional and environmental stresses fluctuate throughout the bacterial lifecycle, the
relative importance and utilization of each alternative DNA polymerase appears to shift
in a mechanism that modulates overall mutation rates. During the stable and nutrient rich
conditions of outgrowth, utilization of the comparatively high fidelity of Pol II maintains
relatively low mutation rates as cells thrive and grow rapidly. As nutrients are depleted
and environmental conditions become more stressful, the comparatively error-prone Pol
IV and Pol V play a greater role during replication and promote the formation of genetic
diversity with higher mutation rates.
However, as cells transition into the comparatively dynamic and stressful conditions of
long-term stationary phase, a greater mutation frequency can become more advantageous
by potentially enabling the formation of a beneficial allele. Here, as competition for
nutrients becomes increasingly fierce and cells become increasingly desperate, the
potential cost of a deleterious mutation decreases relative to the price of cell death.
Hours!
0! 6! 12!
Days!
2! 6! 10! 14!
4!
6!
8!
10!
5!
7!
9!
log
10
CFU/mL!
Pol 2! Pol 4 & Pol 5 !
Natural!
Selection!
Fidelity! Diversity!
Nutrient Availability! Stressful Conditions!
114
Accordingly, nature appears to favor a “mutate or die” strategy embracing the mantra
“nothing to lose and everything to gain”, and has selected for a mechanism of inducing
error-prone alternative DNA polymerase expression during periods of stress. Given that
the comparatively error-prone Pol IV and Pol V confer a significant advantage by
introducing genetic diversity during long-term incubation, our findings indicate bacterial
populations may have evolved mechanisms that facilitate replication with a
comparatively high mutation rate during periods of stress.
Furthermore, we demonstrate an additional level of regulatory complexity influencing
mutation rates as this stress-induced mutagenesis and alternative DNA polymerase
expression can occur within discrete peaks. We observed the alternative DNA
polymerases being induced as part of the SOS response during temporary bursts of
expression, providing a mechanism of transient expression that might serve to limit or
mitigate the potentially deleterious effect of elevated mutation rates by limiting the
timing of error-prone replication. Accordingly, nature appears to have developed and
selected an additional layer of risk management wherein a temporary peak in error-prone
replication can be coordinated with a pause in cell division, thereby limiting mutagenesis
and enabling cells to potentially sense whether stress has been alleviated before initiating
another round of induction.
Additionally, there is an increasing body of evidence indicating that SOS induction and
stress-induced mutagenesis might be limited to a sub-population of cells within microbial
115
communities, providing yet another system of checks and balances to limit potentially
harmful effects of excessive mutation. Single-cell analysis has shown that while
individual cells exhibit pulses of SOS induction, the effect is not homogenous across the
entire community and that only a subpopulation of cells exhibits this pulsing at any given
time (Friedman et al., 2005). These findings are consistent with a proposed model of
stress-induced mutation occurring within only a small minority of cells at any given time
(Galhardo et al., 2007). If this proves to be the case, it appears microbial communities are
capable of utilizing a mechanism to maintain the genetic stability of the majority of cells,
while still exploring genetic space by inducing mutation in a fraction of cells. This
proposed bet-hedging strategy would thereby limit the risk of excessive mutagenesis by
isolating their deleterious effects to a hypermutable subpopulation. While the results
presented in this work reflect patterns observed across entire populations of cells, our
findings are consistent with this model of stress-induced mutation.
Together, our results illustrate a mechanism of stress-induced regulation controlling
alternative DNA polymerase expression that elucidates selective pressures at play within
dynamic conditions akin to natural environments. Given that we found alternative DNA
polymerase expression to remain consistently elevated throughout long-term stationary
phase in the absence of exogenous stressors, it appears this more error-prone replication
might reflect the status quo within natural populations experiencing dynamic changes.
Accordingly, our results suggest previous studies conducted under standard feast
116
conditions of laboratory growth likely underestimated the importance of these enzymes in
nature.
5.3 Physiological roles and evolutionary implications of alternative
DNA polymerases
Microbial communities must often cope with dynamic and unpredictable changes in
nutrient availability and environmental stress. The capacity to survive, adapt, and thrive
during intermittent cycles of feast and famine is crucial to the evolutionary success of an
organism. Ultimately, the findings presented here are among the first in vivo phenotypes
associated with each alternative DNA polymerase in E. coli and demonstrate their
importance to overall fitness. We have defined discrete roles for each polymerase
throughout each phase of the bacterial lifecycle, revealing their contributions to both
physiological and evolutionary fitness during conditions of fitness during feast and
famine. These results illustrate their complex regulation, and emphasize the selective
pressures facing environmental microbial communities. While serving vital physiological
roles, the alternative DNA polymerases influence the rate and patterns of evolution
within natural environments, significantly contributing to the survival and evolution of
microbial communities.
117
References
Al Mamun, A.A., and Humayun, M.Z. (2006). Escherichia coli DNA polymerase II
can efficiently bypass 3,N(4)-ethenocytosine lesions in vitro and in vivo. Mutation
Research 593, 164-176.
Aldsworth, T.G., Sharman, R.L., and Dodd, C.E. (1999). Bacterial suicide through
stress. Cellular and Molecular Life Sciences 56, 378-383.
Atwood, K.C., Schneider, L.K., and Ryan, F.J. (1951a). Periodic selection in
Escherichia coli. Proceedings of the National Academy of Sciences of the United States
of America 37, 146-155.
Atwood, K.C., Schneider, L.K., and Ryan, F.J. (1951b). Selective mechanisms in
bacteria. Cold Spring Harbor Symposia on Quantitative Biology 16, 345-355.
Ball, C.A., Osuna, R., Ferguson, K.C., and Johnson, R.C. (1992). Dramatic changes in
Fis levels upon nutrient upshift in Escherichia coli. Journal of Bacteriology 174, 8043-
8056.
Banach-Orlowska, M., Fijalkowska, I.J., Schaaper, R.M., and Jonczyk, P. (2005).
DNA polymerase II as a fidelity factor in chromosomal DNA synthesis in Escherichia
coli. Molecular Microbiology 58, 61-70.
Baquero, M.R., Bouzon, M., Varea, J., and Moreno, F. (1995). sbmC, a stationary-
phase induced SOS Escherichia coli gene, whose product protects cells from the DNA
replication inhibitor microcin B17. Molecular Microbiology 18, 301-311.
Benjamini, Y., and Hochberg, Y. (1995). Controlling the False Discovery Rate - a
Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical
Society, Series B 57, 289-300.
Bichara, M., Meier, M., Wagner, J., Cordonnier, A., and Lambert, I.B. (2011).
Postreplication repair mechanisms in the presence of DNA adducts in Escherichia coli.
Mutation Research 727, 104-122.
Bjedov, I., Lecointre, G., Tenaillon, O., Vaury, C., Radman, M., Taddei, F.,
Denamur, E., and Matic, I. (2003a). Polymorphism of genes encoding SOS
polymerases in natural populations of Escherichia coli. DNA Repair 2, 417-426.
Bjedov, I., Tenaillon, O., Gerard, B., Souza, V., Denamur, E., Radman, M., Taddei,
F., and Matic, I. (2003b). Stress-induced mutagenesis in bacteria. Science 300, 1404-
1409.
118
Bonner, C.A., Hays, S., McEntee, K., and Goodman, M.F. (1990). DNA polymerase II
is encoded by the DNA damage-inducible dinA gene of Escherichia coli. Proceedings of
the National Academy of Sciences of the United States of America 87, 7663-7667.
Bonner, C.A., Randall, S.K., Rayssiguier, C., Radman, M., Eritja, R., Kaplan, B.E.,
McEntee, K., and Goodman, M.F. (1988). Purification and characterization of an
inducible Escherichia coli DNA polymerase capable of insertion and bypass at abasic
lesions in DNA. Journal of Biological Chemistry 263, 18946-18952.
Bonner, C.A., Stukenberg, P.T., Rajagopalan, M., Eritja, R., O'Donnell, M.,
McEntee, K., Echols, H., and Goodman, M.F. (1992). Processive DNA synthesis by
DNA polymerase II mediated by DNA polymerase III accessory proteins. Journal of
Biological Chemistry 267, 11431-11438.
Braithwaite, D.K., and Ito, J. (1993). Compilation, alignment, and phylogenetic
relationships of DNA polymerases. Nucleic Acids Research 21, 787-802.
Buchanan, R.E. (1918). Life phases in a bacterial culture. Journal of Infectious Diseases
23, 109-125.
Cai, H., Yu, H., McEntee, K., and Goodman, M.F. (1995a). Purification and properties
of DNA polymerase II from Escherichia coli. Methods in Enzymology 262, 13-21.
Cai, H., Yu, H., McEntee, K., Kunkel, T.A., and Goodman, M.F. (1995b).
Purification and properties of wild-type and exonuclease-deficient DNA polymerase II
from Escherichia coli. Journal of Biological Chemistry 270, 15327-15335.
Chao, L., and Cox, E.C. (1983). Competition between high and low mutating strains of
Escherichia coli. Evolution 37, 125-134.
Cirz, R.T., Chin, J.K., Andes, D.R., de Crecy-Lagard, V., Craig, W.A., and
Romesberg, F.E. (2005). Inhibition of mutation and combating the evolution of
antibiotic resistance. PLoS Biology 3, e176.
Courcelle, J., Khodursky, A., Peter, B., Brown, P.O., and Hanawalt, P.C. (2001).
Comparative gene expression profiles following UV exposure in wild-type and SOS-
deficient Escherichia coli. Genetics 158, 41-64.
Curti, E., McDonald, J.P., Mead, S., and Woodgate, R. (2009). DNA polymerase
switching: effects on spontaneous mutagenesis in Escherichia coli. Molecular
Microbiology 71, 315-331.
De Lucia, P., and Cairns, J. (1969). Isolation of an E. coli strain with a mutation
affecting DNA polymerase. Nature 224, 1164-1166.
119
Delmas, S., and Matic, I. (2006). Interplay between replication and recombination in
Escherichia coli: impact of the alternative DNA polymerases. Proceedings of the
National Academy of Sciences of the Unites States of America 103, 4564-4569.
Engelberg-Kulka, H., and Glaser, G. (1999). Addiction modules and programmed cell
death and antideath in bacterial cultures. Annual Review of Microbiology 53, 43-70.
Ferenci, T. (2008). Bacterial physiology, regulation and mutational adaptation in a
chemostat environment. Advances in Microbial Physiology 53, 169-229.
Fernandez De Henestrosa, A.R., Ogi, T., Aoyagi, S., Chafin, D., Hayes, J.J., Ohmori,
H., and Woodgate, R. (2000). Identification of additional genes belonging to the LexA
regulon in Escherichia coli. Molecular Microbiology 35, 1560-1572.
Finkel, S.E. (2006). Long-term survival during stationary phase: evolution and the GASP
phenotype. Nature Reviews Microbiology 4, 113-120.
Finkel, S.E., and Kolter, R. (1999). Evolution of microbial diversity during prolonged
starvation. Proceedings of the National Academy of Sciences of the Unites States of
America 96, 4023-4027.
Foster, P.L. (2000). Adaptive mutation in Escherichia coli. Cold Spring Harbor
Symposia on Quantitative Biology 65, 21-29.
Foster, P.L. (2005). Stress responses and genetic variation in bacteria. Mutation
Research 569, 3-11.
Foster, P.L. (2007). Stress-induced mutagenesis in bacteria. Critical Reviews in
Biochemistry and Molecular Biology 42, 373-397.
Friedberg, E.C., Walker G.C., Siede W., Wood, R.D., Schultz, R.A., et al. (2006).
DNA Repair and Mutagenesis. 2
nd
Edition. (Washington, D.C., ASM Press).
Friedman, N., Vardi, S., Ronen, M., Alon, U., and Stavans, J. (2005). Precise
temporal modulation in the response of the SOS DNA repair network in individual
bacteria. PLoS Biology 3, e238.
Fuchs, R.P., Fujii, S., and Wagner, J. (2004). Properties and functions of Escherichia
coli: Pol IV and Pol V. Advances in Protein Chemistry 69, 229-264.
Galhardo, R.S., Do, R., Yamada, M., Friedberg, E.C., Hastings, P.J., Nohmi, T., and
Rosenberg, S.M. (2009). DinB upregulation is the sole role of the SOS response in
stress-induced mutagenesis in Escherichia coli. Genetics 182, 55-68.
120
Galhardo, R.S., Hastings, P.J., and Rosenberg, S.M. (2007). Mutation as a stress
response and the regulation of evolvability. Critical Reviews in Biochemistry and
Molecular Biology 42, 399-435.
Garibyan, L., Huang, T., Kim, M., Wolff, E., Nguyen, A., Nguyen, T., Diep, A., Hu,
K., Iverson, A., Yang, H., et al. (2003). Use of the rpoB gene to determine the
specificity of base substitution mutations on the Escherichia coli chromosome. DNA
Repair (Amsterdam) 2, 593-608.
Goodman, M.F. (2002). Error-prone repair DNA polymerases in prokaryotes and
eukaryotes. Annual Reviews of Biochemistry 71, 17-50.
Harder, W., and Kuenen, J.G. (1977). A review. Microbial selection in continuous
culture. Journal of Applied Bacteriology 43, 1-24.
Hastings, P.J., Hersh, M.N., Thornton, P.C., Fonville, N.C., Slack, A., Frisch, R.L.,
Ray, M.P., Harris, R.S., Leal, S.M., and Rosenberg, S.M. (2010). Competition of
Escherichia coli DNA polymerases I, II and III with DNA Pol IV in stressed cells. PLoS
One 5, e10862.
Heltzel, J.M., Maul, R.W., Scouten Ponticelli, S.K., and Sutton, M.D. (2009a). A
model for DNA polymerase switching involving a single cleft and the rim of the sliding
clamp. Proceedings of the National Academy of Sciences of the Unites States of America
106, 12664-12669.
Heltzel, J.M., Scouten Ponticelli, S.K., Sanders, L.H., Duzen, J.M., Cody, V., Pace,
J., Snell, E.H., and Sutton, M.D. (2009b). Sliding clamp-DNA interactions are required
for viability and contribute to DNA polymerase management in Escherichia coli. Journal
of Molecular Biology 387, 74-91.
Hersh, M.N., Ponder, R.G., Hastings, P.J., and Rosenberg, S.M. (2004). Adaptive
mutation and amplification in Escherichia coli: two pathways of genome adaptation
under stress. Research in Microbiology 155, 352-359.
Hubscher, U., Maga, G., and Spadari, S. (2002). Eukaryotic DNA polymerases.
Annual Review of Biochemistry 71, 133-163.
Indiani, C., McInerney, P., Georgescu, R., Goodman, M.F., and O'Donnell, M.
(2005). A sliding-clamp toolbelt binds high- and low-fidelity DNA polymerases
simultaneously. Molecular Cell 19, 805-815.
Iwasaki, H., Nakata, A., Walker, G.C., and Shinagawa, H. (1990). The Escherichia
coli polB gene, which encodes DNA polymerase II, is regulated by the SOS system.
Journal of Bacteriology 172, 6268-6273.
121
Jarosz, D.F., Beuning, P.J., Cohen, S.E., and Walker, G.C. (2007). Y-family DNA
polymerases in Escherichia coli. Trends Microbiology 15, 70-77.
Jiang, Q., Karata, K., Woodgate, R., Cox, M.M., and Goodman, M.F. (2009). The
active form of DNA polymerase V is UmuD'(2)C-RecA-ATP. Nature 460, 359-363.
Jin, D.J., and Gross, C.A. (1988). Mapping and sequencing of mutations in the
Escherichia coli rpoB gene that lead to rifampicin resistance. Journal of Molecular
Bioliology 202, 45-58.
Johnson, A., and O'Donnell, M. (2005). Cellular DNA replicases: components and
dynamics at the replication fork. Annual Reviews in Biochemistry 74, 283-315.
Kenyon, C.J., and Walker, G.C. (1980). DNA-damaging agents stimulate gene
expression at specific loci in Escherichia coli. Proceedings of the National Academy of
Sciences of the United States of America 77, 2819-2823.
Kim, S.R., Matsui, K., Yamada, M., Gruz, P., and Nohmi, T. (2001). Roles of
chromosomal and episomal dinB genes encoding DNA pol IV in targeted and untargeted
mutagenesis in Escherichia coli. Molecular Genetics and Genomics 266, 207-215.
Knippers, R. (1970). DNA polymerase II. Nature 228, 1050-1053.
Kobayashi, S., Valentine, M.R., Pham, P., O'Donnell, M., and Goodman, M.F.
(2002). Fidelity of Escherichia coli DNA polymerase IV. Preferential generation of small
deletion mutations by dNTP-stabilized misalignment. Journal of Biological Chemistry
277, 34198-34207.
Kornberg, A., and Baker, T.A. (1992). DNA Replication, 2
nd
Edition (New York, W.H.
Freeman).
Koskiniemi, S., and Andersson, D.I. (2009). Translesion DNA polymerases are
required for spontaneous deletion formation in Salmonella typhimurium. Proceedings of
the National Academy of Sciences of the Unites States of America 106, 10248-10253.
Kraigsley, A.M., and Finkel, S.E. (2009). Adaptive evolution in single species bacterial
biofilms. FEMS Mirobiology Letters 293, 135-140.
Krishna, S., Maslov, S., and Sneppen, K. (2007). UV-induced mutagenesis in
Escherichia coli SOS response: a quantitative model. PLoS Computational Biology 3,
e41.
Kunkel, T.A. (2004). DNA replication fidelity. Journal of Biological Chemistry 279,
16895-16898.
122
Lange, S.S., Takata, K., and Wood, R.D. (2011). DNA polymerases and cancer. Nature
Reviews Cancer 11, 96-110.
Layton, J.C., and Foster, P.L. (2003). Error-prone DNA polymerase IV is controlled by
the stress-response sigma factor, RpoS, in Escherichia coli. Molecular Microbiology 50,
549-561.
Layton, J.C., and Foster, P.L. (2005). Error-prone DNA polymerase IV is regulated by
the heat shock chaperone GroE in Escherichia coli. Journal of Bacteriology 187, 449-
457.
Li, H., and Durbin, R. (2009). Fast and accurate short read alignment with Burrows-
Wheeler transform. Bioinformatics 25, 1754-1760.
Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative gene expression data
using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402-
408.
Locke, J.C., Young, J.W., Fontes, M., Hernandez Jimenez, M.J., and Elowitz, M.B.
(2011). Stochastic pulse regulation in bacterial stress response. Science 334, 366-369.
Makiela-Dzbenska, K., Jonczyk, P., Schaaper, R.M., and Fijalkowska, I.J. (2011).
Proofreading deficiency of Pol I increases the levels of spontaneous rpoB mutations in E.
coli. Mutation Research 712, 28-32.
Maor-Shoshani, A., Reuven, N.B., Tomer, G., and Livneh, Z. (2000). Highly
mutagenic replication by DNA polymerase V (UmuC) provides a mechanistic basis for
SOS untargeted mutagenesis. Proceedings of the National Academy of Sciences of the
Unites States of America 97, 565-570.
Martinez, A., and Kolter, R. (1997). Protection of DNA during oxidative stress by the
nonspecific DNA-binding protein Dps. Journal of Bacteriology 179, 5188-5194.
McKenzie, G.J., Lee, P.L., Lombardo, M.J., Hastings, P.J., and Rosenberg, S.M.
(2001). SOS mutator DNA polymerase IV functions in adaptive mutation and not
adaptive amplification. Molecular Cell 7, 571-579.
Murli, S., Opperman, T., Smith, B.T., and Walker, G.C. (2000). A role for the
umuDC gene products of Escherichia coli in increasing resistance to DNA damage in
stationary phase by inhibiting the transition to exponential growth. Journal of
Bacteriology 182, 1127-1135.
Nohmi, T. (2006). Environmental stress and lesion-bypass DNA polymerases. Annual
Reviews in Microbiology 60, 231-253.
123
Novick, A., and Szilard, L. (1950). Description of the chemostat. Science 112, 715-716.
Oh, T.J., Jung, I.L., and Kim, I.G. (2001). The Escherichia coli SOS gene sbmC is
regulated by H-NS and RpoS during the SOS induction and stationary growth phase.
Biochemical and Biophysical Research Communications 288, 1052-1058.
Ohmori, H., Friedberg, E.C., Fuchs, R.P., Goodman, M.F., Hanaoka, F., Hinkle, D.,
Kunkel, T.A., Lawrence, C.W., Livneh, Z., Nohmi, T., et al. (2001). The Y-family of
DNA polymerases. Molecular Cell 8, 7-8.
Opperman, T., Murli, S., Smith, B.T., and Walker, G.C. (1999). A model for a
umuDC-dependent prokaryotic DNA damage checkpoint. Proceedings of the National
Academy of Sciences of the Unites States of America 96, 9218-9223.
Patel, M., Jiang, Q., Woodgate, R., Cox, M.M., and Goodman, M.F. (2010). A new
model for SOS-induced mutagenesis: how RecA protein activates DNA polymerase V.
Critical Reviews in Biochemistry and Molecular Biology 45, 171-184.
Pham, P., Rangarajan, S., Woodgate, R., and Goodman, M.F. (2001). Roles of DNA
polymerases V and II in SOS-induced error-prone and error-free repair in Escherichia
coli. Proceedings of the National Academy of Sciences of the Unites States of America
98, 8350-8354.
Pigliucci, M. (2008). Is evolvability evolvable? Nature Reviews Genetics 9, 75-82.
Qiu, Z., and Goodman, M.F. (1997). The Escherichia coli polB locus is identical to
dinA, the structural gene for DNA polymerase II. Characterization of Pol II purified from
a polB mutant. Journal of Biological Chemistry 272, 8611-8617.
Radman, M. (1974). Phenomenology of an inducible mutagenic DNA repair pathway in
E. coli: SOS repair hypothesis. Molecular and Environmental Aspects of Mutagenesis
(6th Rochester Conference on Environmental Toxicity, 1973), S.F. Prakash L, Miller
MW, Lawrence CW, Tabor WH, ed. (Springfield, IL, Charles C Thomas Publisher).
Radman, M. (1975). SOS repair hypothesis: phenomenology of an inducible DNA repair
which is accompanied by mutagenesis. Basic Life Sciences 5A, 355-367.
Radman, M. (2007). SOS Hypothesis and the Emergence of Integrative Biology
Supramolecular Structure and Function 9. In, G. Pifat-Mrzljak, ed. (Springer
Netherlands), pp. 307-313.
124
Rajagopalan, M., Lu, C., Woodgate, R., O'Donnell, M., Goodman, M.F., and Echols,
H. (1992). Activity of the purified mutagenesis proteins UmuC, UmuD', and RecA in
replicative bypass of an abasic DNA lesion by DNA polymerase III. Proceedings of the
National Academy of Sciences of the United States of America 89, 10777-10781.
Rangarajan, S., Woodgate, R., and Goodman, M.F. (1999). A phenotype for
enigmatic DNA polymerase II: a pivotal role for pol II in replication restart in UV-
irradiated Escherichia coli. Proceedings of the National Academy of Sciences of the
Unites States of America 96, 9224-9229.
Rangarajan, S., Woodgate, R., and Goodman, M.F. (2002). Replication restart in UV-
irradiated Escherichia coli involving pols II, III, V, PriA, RecA and RecFOR proteins.
Molecular Microbiology 43, 617-628.
Reuven, N.B., Arad, G., Maor-Shoshani, A., and Livneh, Z. (1999). The mutagenesis
protein UmuC is a DNA polymerase activated by UmuD', RecA, and SSB and is
specialized for translesion replication. The Journal of Biological Chemistry 274, 31763-
31766.
Reynolds, M.G. (2000). Compensatory evolution in rifampin-resistant Escherichia coli.
Genetics 156, 1471-1481.
Robbins, J.H., Kraemer, K.H., and Flaxman, B.A. (1975). DNA repair in tumor cells
from the variant form of xeroderma pigmentosum. Journal of Investigative Dermatology
64, 150-155.
Robbins, J.H., Kraemer, K.H., Lutzner, M.A., Festoff, B.W., and Coon, H.G. (1974).
Xeroderma pigmentosum. An inherited diseases with sun sensitivity, multiple cutaneous
neoplasms, and abnormal DNA repair. Annals of Internal Medicine 80, 221-248.
Roth, J.R., Kugelberg, E., Reams, A.B., Kofoid, E., and Andersson, D.I. (2006).
Origin of mutations under selection: the adaptive mutation controversy. Annual Review
of Microbiology 60, 477-501.
Saint-Ruf, C., and Matic, I. (2006). Environmental tuning of mutation rates.
Environmental Microbiolgy 8, 193-199.
Severinov, K., Soushko, M., Goldfarb, A., and Nikiforov, V. (1994). RifR mutations
in the beginning of the Escherichia coli rpoB gene. Molecular and General Genetics 244,
120-126.
Singer, M., Jin, D.J., Walter, W.A., and Gross, C.A. (1993). Genetic evidence for the
interaction between cluster I and cluster III rifampicin resistant mutations. Journal of
Molecular Biology 231, 1-5.
125
Skvortsov, D., Abdueva, D., Stitzer, M.E., Finkel, S.E., and Tavare, S. (2007). Using
expression arrays for copy number detection: an example from E. coli. BMC
Bioinformatics 8, 203.
Stallons, L.J., and McGregor, W.G. (2010). Translesion synthesis polymerases in the
prevention and promotion of carcinogenesis. Journal of Nucleic Acids 2010.
Stumpf, J.D., and Foster, P.L. (2005). Polyphosphate kinase regulates error-prone
replication by DNA polymerase IV in Escherichia coli. Molecular Microbiology 57, 751-
761.
Sutton, M.D. (2004). The Escherichia coli dnaN159 mutant displays altered DNA
polymerase usage and chronic SOS induction. Journal of Bacteriology 186, 6738-6748.
Sutton, M.D. (2010). Coordinating DNA polymerase traffic during high and low fidelity
synthesis. Biochimica Biophysica Acta 1804, 1167-1179.
Sutton, M.D., Murli, S., Opperman, T., Klein, C., and Walker, G.C. (2001). umuDC-
dnaQ Interaction and its implications for cell cycle regulation and SOS mutagenesis in
Escherichia coli. Journal of Bacteriology 183, 1085-1089.
Tang, M., Bruck, I., Eritja, R., Turner, J., Frank, E.G., Woodgate, R., O'Donnell,
M., and Goodman, M.F. (1998). Biochemical basis of SOS-induced mutagenesis in
Escherichia coli: reconstitution of in vitro lesion bypass dependent on the UmuD'2C
mutagenic complex and RecA protein. Proceedings of the National Academy of Sciences
of the United States of America 95, 9755-9760.
Tang, M., Pham, P., Shen, X., Taylor, J.S., O'Donnell, M., Woodgate, R., and
Goodman, M.F. (2000). Roles of E. coli DNA polymerases IV and V in lesion-targeted
and untargeted SOS mutagenesis. Nature 404, 1014-1018.
Tang, M., Shen, X., Frank, E.G., O'Donnell, M., Woodgate, R., and Goodman, M.F.
(1999). UmuD'(2)C is an error-prone DNA polymerase, Escherichia coli pol V.
Proceedings of the National Academy of Sciences of the United States of America 96,
8919-8924.
Tippin, B., Pham, P., and Goodman, M.F. (2004). Error-prone replication for better or
worse. Trends in Microbiology 12, 288-295.
Wagner, J., Gruz, P., Kim, S.R., Yamada, M., Matsui, K., Fuchs, R.P., and Nohmi,
T. (1999). The dinB gene encodes a novel E. coli DNA polymerase, DNA pol IV,
involved in mutagenesis. Molecular Cell 4, 281-286.
126
Walker, G.C. (1984). Mutagenesis and inducible responses to deoxyribonucleic acid
damage in Escherichia coli. Microbiological Reviews 48, 60-93.
Wang, F., and Yang, W. (2009). Structural insight into translesion synthesis by DNA
Pol II. Cell 139, 1279-1289.
Wolff, E., Kim, M., Hu, K., Yang, H., and Miller, J.H. (2004). Polymerases leave
fingerprints: analysis of the mutational spectrum in Escherichia coli rpoB to assess the
role of polymerase IV in spontaneous mutation. Journal of Bacteriology 186, 2900-2905.
Woodgate, R., and Ennis, D.G. (1991). Levels of chromosomally encoded Umu
proteins and requirements for in vivo UmuD cleavage. Molecular and General Genetics
229, 10-16.
Woods, R.J., Barrick, J.E., Cooper, T.F., Shrestha, U., Kauth, M.R., and Lenski,
R.E. (2011). Second-order selection for evolvability in a large Escherichia coli
population. Science 331, 1433-1436.
Wrande, M., Roth, J.R., and Hughes, D. (2008). Accumulation of mutants in "aging"
bacterial colonies is due to growth under selection, not stress-induced mutagenesis.
Proceedings of the National Academy of Sciences of the Unites States of America 105,
11863-11868.
Xie, K., Doles, J., Hemann, M.T., and Walker, G.C. (2010). Error-prone translesion
synthesis mediates acquired chemoresistance. Proceedings of the National Academy of
Sciences of the Unites States of America 107, 20792-20797.
Yang, W., and Woodgate, R. (2007). What a difference a decade makes: insights into
translesion DNA synthesis. Proceedings of the National Academy of Sciences of the
Unites States of America 104, 15591-15598.
Yeiser, B., Pepper, E.D., Goodman, M.F., and Finkel, S.E. (2002). SOS-induced DNA
polymerases enhance long-term survival and evolutionary fitness. Proceedings of the
National Academy of Sciences of the Unites States of America 99, 8737-8741.
Zambrano, M.M., Siegele, D.A., Almiron, M., Tormo, A., and Kolter, R. (1993).
Microbial competition: Escherichia coli mutants that take over stationary phase cultures.
Science 259, 1757-1760.
127
Appendix A: The effect of serial passage and batch culture aging on
fitness during long-term stationary phase and the GASP phenotype
A.1 Abstract
Environmental microbial communities can quickly adapt and thrive within ever-changing
and unpredictable environments. Accordingly, characterization of the mechanisms, rates,
and dynamics in which bacteria adapt to diverse selective pressures is critical to
understanding their evolution. To assess the effects of prolonged aging of microbial
communities under diverse conditions on the formation of genetic diversity and selection,
we systematically aged populations of bacteria under long-term batch culture as well as
daily serial passage before initiating a series of competitions to determine the effects of
aging on relative fitness. Our findings indicate mutations that improve performance
during long-term stationary phase competitions are generated quickly within both aging
regimens, but that aging within batch culture is more beneficial than serial passage. We
show that serial passage creates significant genetic diversity that can generate alleles
beneficial within stationary phase, but that prolonged adaptation to outgrowth conditions
eventually precludes further success within long-term batch culture. Therefore, we
demonstrate both serial passage and batch culture incubation aging regimens create
significant genetic diversity, but serve to select and amplify alleles beneficial under
different conditions.
128
A.2 Introduction
Microbial communities are capable of quickly adapting to survive and grow within to
diverse environmental habitats. Characterizing the mechanisms, rate, and dynamics of
bacterial evolution to different conditions remains vital to understanding how these
organisms, as well as populations in general, change over time.
Even within a standard laboratory culture of Escherichia coli in a rich medium, such as
LB, bacteria experience at least five distinct phases of growth that reflect changing
nutritional availability and environmental stresses (Figure 1.4). While it has been known
that prolonged incubation of bacterial populations within batch culture can generate
beneficial alleles that improve relative fitness compared to unaged cells (Atwood et al.,
1951a, b), and can demonstrate the Growth Advantage in Stationary Phase (GASP)
phenotype (Finkel, 2006), the timing and impact of different aging regimens on the
generation of genetic diversity and selection of beneficial alleles remained poorly
characterized.
While previous data has indicated beneficial alleles continue to form during prolonged
batch culture incubation (20-day old cells virtually always GASP over 10-day old cells),
the timing and rate of allele formation has remained ambiguous. Similarly, the impact of
adaptation and selection under conditions different from those experienced during long-
term stationary phase (LTSP) has not previously been explored. For example, it remains
unknown whether incubation and aging in conditions other than LTSP might also
129
increase performance within these conditions, or serve to preclude success under
starvation conditions.
To assess the effects of prolonged aging of microbial communities under diverse
conditions on the formation of genetic diversity and selection, we systematically aged
populations of bacteria under long-term batch culture as well as daily serial passage
before initiating a series of competitions to determine the effects of aging on relative
fitness. Our results further elucidate the impact of serial passage and batch culture
incubation on the formation of genetic diversity and relative fitness within long-term
competitions.
A.3 Materials and Methods
A.3.1 Strains used in this study
The two parental strains used in this study are derived from E. coli K-12 strain ZK126
(W3110 ∆lacU169 tna-2), and consist of a nalidixic acid-resistant (ZK1142) and a
streptomycin-resistant (ZK1143) wildtype strain. Genetic elements conferring antibiotic
resistance are effectively neutral in the absence of drug selection (Kraigsley and Finkel,
2009; Yeiser et al., 2002).
A.3.2 Culture conditions, media, and titering assays
Strains were cultured in 5.0mL LB Broth, Lennox (Difco-BD) and incubated at 37ºC with
aeration in a TC-7 test tube roller (New Brunswick Scientific), unless otherwise
130
specified. Experiments were initiated from overnight cultures inoculated from frozen LB-
glycerol stocks. Viable counts were determined by serial dilution of cells periodically
removed from cultures, and plating on selective medium containing the appropriate
antibiotics: Nal, nalidixic acid (20µg/mL); Str, streptomycin (25µg/mL). This method of
titering is accurate within +/- 3-fold (Kraigsley and Finkel, 2009), and has a limit of
detection of 1000 colony forming units (CFU)/mL in this study.
A.3.3 Serial passage aging regimen
To observe changes in relative fitness following repeated outgrowth, strains were serially
passaged in 5.0mL LB cultures. Independent cultures of each strain were incubated for 24
hours, sampled and diluted 1:1000 (vol/vol) into fresh medium. Three independent
cultures were passaged daily for 10 days and frozen LB-glycerol stocks were prepared
following each passage. Competitions were initiated using populations experiencing
either 5 or 10 passages, as indicated in parentheses [e.g. serial (+5) and serial (+10),
respectively].
A.3.4 Batch culture aging regimen
The same parental strains used to initiate serial passage aging were also used to initiate
cultures for aging within batch cultures. Three independent cultures were allowed to
incubate for 10 days, and frozen LB-glycerol stocks were prepared from 5-day-old and
10-day-old populations. Competitions were initiated from these batch culture aged
131
populations, and are indicated with their age in parentheses [e.g. batch (+5) and batch
(+10), respectively].
A.3.5 Batch culture long-term competition assays
Two types of batch culture competitions, distinguished by their initial cell densities, were
performed: Type 1, where both strains are mixed at a low initial density (~10
6
CFU/mL)
and co-outgrown, and Type 2, where high-density stationary phase cultures are mixed
(~10
9
CFU/mL). In Type 1 experiments, competitions were initiated by inoculating 5µL
of each competing strain (1:1000 dilution, vol:vol) into the same 5.0mL LB culture. In
Type 2 experiments, competitions were initiated by combining 2.5mL of overnight
stationary phase cultures of each strain (1:1 mix, vol:vol). Viable counts were determined
as described above using the appropriate antibiotics (Nal for ZK1142, Str for ZK1143).
At the conclusion of each 14-day competition, strains showing a 10-fold greater relative
population density were scored as winners.
A.3.6 GASP competitions
To initiate GASP competitions, an overnight culture of each aged population (derived
from parental ZK1142) was grown and introduced as a minority at a 1:1000 (vol/vol)
dilution into a 5.0mL culture of an unaged population of wildtype cells (ZK1143).
Population densities for each strain were determined by titering, as described above.
132
A.4 Results
All possible combinations of Type 1 (co-outgrowth) and Type 2 (stationary phase)
competitions were conducted between unaged and aged populations within both genetic
backgrounds used in this study. Populations derived from cells aged during serial passage
and batch culture incubation for either 5- or 10-days were competed against one another,
as well as against unaged parental strains. The results of each competition are presented
in Figures A.1-A.10, and summarized in Tables A.1-A.2.
133
Figure A.1 Co-outgrowth and stationary phase competitions between unaged and
batch (+5) populations.
Cell densities of unaged and batch (+5) populations are plotted during Type 1 co-
outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging treatments
were performed in each strain background (ZK1142 and ZK1143), and competitions were
initiated using swapped markers for each type of competition. Strain backgrounds for
each population are displayed above each panel. Aging treatments are indicated by line
color: unaged, black; batch (+5), cyan. Three competitions are shown in each panel where
circles, squares, and triangles indicate competition pairs.
1142 Unaged vs. 1143 Batch (+5) 1142 Unaged vs. 1143 Batch (+5)
1143 Unaged vs. 1142 Batch (+5) 1143 Unaged vs. 1142 Batch (+5)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
134
Figure A.2 Co-outgrowth and stationary phase competitions between unaged and
batch (+10) populations.
Cell densities of unaged and batch (+10) populations are plotted during Type 1 co-
outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging treatments
were performed in each strain background (ZK1142 and ZK1143), and competitions were
initiated using swapped markers for each type of competition. Strain backgrounds for
each population are displayed above each panel. Aging treatments are indicated by line
color: unaged, black; batch (+10), blue. Three competitions are shown in each panel
where circles, squares, and triangles indicate competition pairs.
1142 Unaged vs. 1143 Batch (+10) 1142 Unaged vs. 1143 Batch (+10)
1143 Unaged vs. 1142 Batch (+10) 1143 Unaged vs. 1142 Batch (+10)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
135
Figure A.3 Co-outgrowth and stationary phase competitions between unaged and
serial passage (+5) populations.
Cell densities of unaged and serial passage (+5) populations are plotted during Type 1 co-
outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging treatments
were performed in each strain background (ZK1142 and ZK1143), and competitions were
initiated using swapped markers for each type of competition. Strain backgrounds for
each population are displayed above each panel. Aging treatments are indicated by line
color: unaged, black; serial passage (+5), pink. Three competitions are shown in each
panel where circles, squares, and triangles indicate competition pairs.
1142 Unaged vs. 1143 Serial (+5) 1142 Unaged vs. 1143 Serial (+5)
1143 Unaged vs. 1142 Serial (+5) 1143 Unaged vs. 1142 Serial (+5)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
136
Figure A.4 Co-outgrowth and stationary phase competitions between unaged and
serial passage (+10) populations.
Cell densities of unaged and serial passage (+10) populations are plotted during Type 1
co-outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging treatments
were performed in each strain background (ZK1142 and ZK1143), and competitions were
initiated using swapped markers for each type of competition. Strain backgrounds for
each population are displayed above each panel. Aging treatments are indicated by line
color: unaged, black; serial passage (+10), red. Three competitions are shown in each
panel where circles, squares, and triangles indicate competition pairs.
1142 Unaged vs. 1143 Serial (+10) 1142 Unaged vs. 1143 Serial (+10)
1143 Unaged vs. 1142 Serial (+10) 1143 Unaged vs. 1142 Serial (+10)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
137
Figure A.5 Co-outgrowth and stationary phase competitions between batch (+5)
and batch (+10) populations.
Cell densities of batch (+5) and batch (+10) populations are plotted during Type 1 co-
outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging treatments
were performed in each strain background (ZK1142 and ZK1143), and competitions were
initiated using swapped markers for each type of competition. Strain backgrounds for
each population are displayed above each panel. Aging treatments are indicated by line
color: batch (+5), cyan; batch (+10), blue. Three competitions are shown in each panel
where circles, squares, and triangles indicate competition pairs.
1142 Batch (+5) vs. 1143 Batch (+10) 1142 Batch (+5) vs. 1143 Batch
(+10)
1143 Batch (+5) vs. 1142 Batch (+10) 1143 Batch (+5) vs. 1142 Batch
(+10)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
138
Figure A.6 Co-outgrowth and stationary phase competitions between batch (+5)
and serial passage (+5) populations.
Cell densities of batch (+5) and serial passage (+5) populations are plotted during Type 1
co-outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging treatments
were performed in each strain background (ZK1142 and ZK1143), and competitions were
initiated using swapped markers for each type of competition. Strain backgrounds for
each population are displayed above each panel. Aging treatments are indicated by line
color: batch (+5), cyan; serial passage (+5), pink. Three competitions are shown in each
panel where circles, squares, and triangles indicate competition pairs.
1142 Batch (+5) vs. 1143 Serial (+5) 1142 Batch (+5) vs. 1143 Serial (+5)
1143 Batch (+5) vs. 1142 Serial (+5) 1143 Batch (+5) vs. 1142 Serial (+5)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
139
Figure A.7 Co-outgrowth and stationary phase competitions between batch (+5)
and serial passage (+10) populations.
Cell densities of batch (+5) and serial passage (+10) populations are plotted during Type
1 co-outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging
treatments were performed in each strain background (ZK1142 and ZK1143), and
competitions were initiated using swapped markers for each type of competition. Strain
backgrounds for each population are displayed above each panel. Aging treatments are
indicated by line color: batch (+5), cyan; serial passage (+10), red. Three competitions
are shown in each panel where circles, squares, and triangles indicate competition pairs.
1142 Batch (+5) vs. 1143 Serial (+10) 1142 Batch (+5) vs. 1143 Serial (+10)
1143 Batch (+5) vs. 1142 Serial (+10) 1143 Batch (+5) vs. 1142 Serial (+10)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
140
Figure A.8 Co-outgrowth and stationary phase competitions between batch (+10)
and serial passage (+5) populations.
Cell densities of batch (+10) and serial passage (+5) populations are plotted during Type
1 co-outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging
treatments were performed in each strain background (ZK1142 and ZK1143), and
competitions were initiated using swapped markers for each type of competition. Strain
backgrounds for each population are displayed above each panel. Aging treatments are
indicated by line color: batch (+10), blue; serial passage (+5), pink. Three competitions
are shown in each panel where circles, squares, and triangles indicate competition pairs.
1142 Batch (+10) vs. 1143 Serial (+5) 1142 Batch (+10) vs. 1143 Serial (+5)
1143 Batch (+10) vs. 1142 Serial (+5) 1143 Batch (+10) vs. 1142 Serial (+5)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
141
Figure A.9 Co-outgrowth and stationary phase competitions between batch (+10)
and serial passage (+10) populations.
Cell densities of batch (+10) and serial passage (+10) populations are plotted during Type
1 co-outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Aging
treatments were performed in each strain background (ZK1142 and ZK1143), and
competitions were initiated using swapped markers for each type of competition. Strain
backgrounds for each population are displayed above each panel. Aging treatments are
indicated by line color: batch (+10), blue; serial passage (+10), red. Three competitions
are shown in each panel where circles, squares, and triangles indicate competition pairs.
1142 Batch (+10) vs. 1143 Serial (+10) 1142 Batch (+10) vs. 1143 Serial (+10)
1143 Batch (+10) vs. 1142 Serial (+10) 1143 Batch (+10) vs. 1142 Serial (+10)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
142
Figure A.10 Co-outgrowth and stationary phase competitions between serial
passage (+5) and serial passage (+10) populations.
Cell densities of serial passage (+5) and serial passage (+10) populations are plotted
during Type 1 co-outgrowth (A&B) and Type 2 stationary phase (C&D) competitions.
Aging treatments were performed in each strain background (ZK1142 and ZK1143), and
competitions were initiated using swapped markers for each type of competition. Strain
backgrounds for each population are displayed above each panel. Aging treatments are
indicated by line color: serial passage (+5), pink; serial passage (+10), red. Three
competitions are shown in each panel where circles, squares, and triangles indicate
competition pairs.
1142 Serial (+5) vs. 1143 Serial (+10) 1142 Serial (+5) vs. 1143 Serial (+10)
1143 Serial (+5) vs. 1142 Serial (+10) 1143 Serial (+5) vs. 1142 Serial (+10)
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
143
Figure A.11 Summary of co-outgrowth and stationary phase competitions
following aging treatment against unaged populations.
A composite of competitions between each aged and unaged populations for Type 1 co-
outgrowth (A&B) and Type 2 stationary phase (C&D) competitions. Competitions
between unaged 1142 populations against aged 1143 populations are shown in A and C,
while those between unaged 1143 and aged 1142 are shown in B and D. Line colors
reflect the aged population in each competition: serial (+10), red; serial (+5); batch (+5),
cyan; batch (+10), blue. Solid lines reflect aged populations, unaged populations are
denoted with dashed lines.
Serial (+10) a! Serial (+5) a! Batch (+5) a! Batch (+10) a!
O/N Growth a! O/N Growth a! O/N Growth a! O/N Growth a!
Serial (+10) a! Serial (+5) a! Batch (+5) a! Batch (+10) a!
O/N Growth a! O/N Growth a! O/N Growth a! O/N Growth a!
Serial (+10) a! Serial (+5) a! Batch (+5) a! Batch (+10) a!
O/N Growth a! O/N Growth a! O/N Growth a! O/N Growth a!
Serial (+10) a! Serial (+5) a! Batch (+5) a! Batch (+10) a!
O/N Growth a! O/N Growth a! O/N Growth a! O/N Growth a!
Serial (+10) Batch (+10) Serial (+5) Batch (+5)
vs. S(+10) vs. S(+5) vs. B(+5) vs. B(+10)
1142 Unaged vs. 1143 Aged 1142 Unaged vs. 1143 Aged
1143 Unaged vs. 1142 Aged 1143 Unaged vs. 1142 Aged
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Outgrowth Stationary Phase
Day Day
A
B
C
D
144
Table A.1 Summary of co-outgrowth competitions.
Final outcomes of all head-to-head co-outgrowth competitions are summarized,
regardless of parental strains. Populations with a 10-fold greater density at the end of 14
days were considered winners. Values are color-coded, with lowest values in yellow and
highest values in red.
Table A.2 Summary of stationary phase competitions.
Final outcomes of all head-to-head stationary phase competitions are summarized,
regardless of parental strains. Populations with a 10-fold greater density at the end of 14
days were considered winners. Values are color-coded, with lowest values in yellow and
highest values in red.
> = <
Unaged 0 0 6 Batch (+5)
Unaged 0 0 6 Batch (+10)
Unaged 0 2 4 Serial (+5)
Unaged 0 3 3 Serial (+10)
Batch (+5) 0 5 1 Batch (+10)
Batch (+5) 5 0 1 Serial (+5)
Batch (+5) 5 1 0 Serial (+10)
Batch (+10) 5 1 0 Serial (+5)
Batch (+10) 6 0 0 Serial (+10)
Serial (+5) 1 4 1 Serial (+10)
> = <
Unaged 0 0 6 Batch (+5)
Unaged 0 0 6 Batch (+10)
Unaged 0 4 2 Serial (+5)
Unaged 0 5 1 Serial (+10)
Batch (+5) 0 1 5 Batch (+10)
Batch (+5) 4 2 0 Serial (+5)
Batch (+5) 5 1 0 Serial (+10)
Batch (+10) 6 0 0 Serial (+5)
Batch (+10) 6 0 0 Serial (+10)
Serial (+5) 2 3 1 Serial (+10)
145
Figure A.12 Effect of aging regimen on GASP phenotype.
Wildtype strains were aged during serial passage (A&B) or batch culture (C&D) for
either 5 (A&C) or 10 (B&D) days, then competed against unaged wiltype populations to
determine their GASP phenotypes. Filled symbols correspond to aged populations; open
symbols correspond to unaged strains. Aging treatments are indicated by line color:
unaged, black; serial passage (+5), pink; serial passage (+10), red; batch (+5), cyan; batch
(+10), blue. Unaged wildtype populations were competed against (A) serial passage
(+5), (B) serial passage (+10), (C) batch (+5), or batch (+10). Three competitions are
shown for each pair where squares, circles, and triangles indicate competition pairs.
Serial Passage Batch Culture
11
9
7
5
3
11
9
7
5
3
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
log
10
CFU/mL
Day Day
A
B
C
D
Serial (+5)
Serial (+10) Batch (+10)
Batch (+5)
146
A.4.1 Unaged populations perform worse against cells aged in batch culture than
those aged during serial passage
When assessing the relative performance of each aged population against unaged cells,
cells aged in batch culture (Figures A.1 and A.2) performed better against unaged cells
than those aged during serial passage (Figures A.3 and A.4). Populations aged for either 5
or 10 days always outcompeted unaged populations. Batch (+10) populations drove
unaged cells to extinction faster and more often than batch (+5) cells. This trend held
within both Type 1 and Type 2 competitions, and can be clearly observed when each
competition of aged versus unaged is plotted together (Figure A.11)
Among cells aged during serial passage, neither serial (+5) nor serial (+10) ever lost to
unaged populations, however the majority of competitions ended as ties. Quantitatively,
serial (+5) won one more competition than serial (+10) against unaged cells during Type
1 and Type 2 competitions (Tables A.1 and A.2, respectively).
A.4.2 10 days of batch culture incubation is more beneficial than 5 days of
incubation
Among competitions between batch (+5) versus batch (+10), the batch (+10) cells never
lost to batch (+5) cells (Figure A.5). However, batch (+5) populations performed
comparatively better within co-outgrowth competitions (Figure A.5a and A.5b) than they
did in stationary phase competitions (Figure A.5c and A.5d). Batch (+5) tied with batch
(+10) in most Type 1 competitions, while it lost most Type 2 competitions.
147
A.4.3 Cells aged in batch culture defeat cells aged during serial passage
Within competitions between cells aged in batch culture versus those aged during serial
passage (Figures A.6-A.9), batch cells cumulatively defeated serial passage cells in each
variety of competition. In both Type 1 and Type 2 competitions, batch cells won 88% of
all competitions against serially passaged cells. Quantitatively, serial (+5) cells won or
tied more competitions against batch cells than serial (+10) cells (Tables A.1 and A.2).
A.4.4 Repeated serial passage does not improve performance during long-term
stationary phase competitions
Co-outgrowth and stationary phase competitions between serial (+5) and serial (+10)
demonstrated approximately equal relative fitness (FigureA.10). Qualitatively and
quantitatively (Tables A.1 and A.2, respectively) both populations performed similarly.
While most competitions ended as ties, the serial (+5) populations won one more
stationary phase competition than serial (+10).
A.4.5 Cells aged in batch culture exhibit stronger GASP phenotypes.
In addition to head-to-head long-term stationary phase competitions, each aged strain was
also subjected to GASP competitions. Typical GASP competitions are initiated with a 10-
day-old population from batch culture incubation inoculated as a 1:1000 (vol:vol)
minority population into a dense overnight of unaged cells. Here, we initiated similar
competitions with each aged population [serial (+5), serial (+10), batch (+5), and batch
(+10)] inoculated as a minority against a dense overnight of unaged cells (Figure A.12).
148
Serial (+5) minority populations began to increase cell density, but never exceeded
unaged populations (Figure A.12a). Serial (+10) cells increased in density faster than
serial (+5), but also failed to take over the unaged population (Figure A.12b). Similarly,
batch (+5) cells quickly increased in population density, but never defeated the unaged
population (Figure A.12c). This is in contrast to the batch (+10) minority population,
which quickly took over the culture and exhibited greater final population densities than
the unaged cells (Figure A.12d). While the GASP competitions illustrated show each
aged ZK1142 strain against unaged ZK1143, similar trends were also observed when
genetic markers were swapped (data not shown).
A.5 Discussion
We found that the longer cells were incubated in batch culture, the better they performed
within all long-term stationary phase competitions. Given that the long-term competitions
utilized in this study were akin to those experienced in cells aging in long-term batch
culture, it is perhaps unsurprising that these populations would perform better. After all,
enough time was provided for genetic diversity to be generated, and for beneficial alleles
under these growth conditions to be selected. However, it is important to note that enough
diversity and beneficial alleles were already present within only 5 days of incubation to
enable these cells to defeat unaged cells, emphasizing the rate at which beneficial alleles
are generated and selected with batch culture.
149
Furthermore, batch (+10) cells consistently performed better than batch (+5) cells. While
this observation was not unexpected, it serves to demonstrate beneficial alleles continue
to be generated and selected within long-term incubation. It is often presumed that upon
entry into stationary phase cell growth arrests and remains relatively stagnant until death
phase, where the majority of cells die. This death phase selects for cells most capable of
scavenging nutrients and surviving within the comparatively harsh conditions of long-
term stationary phase. Given that death phase occurs prior to day 5, both batch (+5) and
batch (+10) populations consisted of cells derived from the survivors of death phase.
While death phase is considered the initial selective pressure enabling the isolation of
GASP alleles, the fact that batch (+10) cells outcompete batch (+5) cells suggests these
populations continued to accrue additional beneficial alleles during the additional 5 days
of batch culture incubation. Since batch (+5) cells initially drop in density within most
competitions against batch (+10), it appears unlikely the additional 5 days of incubation
merely served to provide additional time to select beneficial alleles initially generated
early during growth, but rather that new beneficial alleles are generated between days 5
and 10.
In addition to characterizing the rate and dynamics of beneficial allele formation and
selection during batch culture, we demonstrate that alleles beneficial for long-term
stationary phase also occur during serial passage. Populations aged during serial passage
performed qualitatively and quantitatively better than unaged cells within long-term
stationary phase competitions. Unsurprisingly, these serially passaged strains performed
150
better in Type 1 competitions where they experienced another outgrowth under
conditions similar to those previously experienced during the aging regimen. However,
the fact that they often won (and never lost) to unaged cells indicated aging during serial
passage can generate alleles better-suited for survival during long-term stationary phase,
conditions these cells were not accustomed to following their daily passage. This finding
serves to demonstrate that significant levels of genetic diversity continue to be generated
within outgrowth and serial passage, but that this diversity remains underappreciated
given that conditions experienced during outgrowth and entry into stationary phase do not
provide a strong selective force for the amplification and observation of these diverse
alleles.
The importance and influence of different selective forces acting during serial passage
and batch culture aging are further emphasized by the results of competitions between
serial (+5) and serial (+10) cells. Here, these populations experienced an aging regimen
that would presumably select for alleles that might improve the capacity to quickly
initiate growth, provide faster growth rates, or enable populations to reach greater
densities. Alleles selected under these conditions might provide little or no advantage
during conditions experienced in long-term stationary phase, which were used throughout
the competitions shown here. As predicted, the serial (+10) cells had more time to adapt
to conditions very different from those experienced in long-term stationary phase, and
performed worse relative to serial (+5) cells within each competition. Therefore, it
appears the genetic diversity created and selected during prolonged adaptation to serial
151
passage came with a corresponding loss of fitness within the environmental conditions
experienced in head-to-head long-term stationary phase competitions.
Our GASP competition results reflect similar findings within a different form of
competition. Here, we again observed batch culture competitions performing better
compared to serially passaged cells when competing against an unaged population. The
batch (+10) cells exhibited a standard GASP phenotype, but it appears batch (+5) cells
were not capable of fully outcompeting the unaged population. This finding again
suggests important beneficial GASP alleles continue to form between day 5 and day 10 of
batch culture incubation.
Interestingly, our GASP competitions initiated using cells aged during serial passage
revealed slightly different phenotypes. Again, serial passage appeared to generate enough
beneficial alleles to allow these populations to remain prevalent and grow as minority
populations even though they never performed as well as cells aged in batch culture.
However, under these conditions the serial (+10) cells were better competitors than serial
(+5) cells. Since each population was able to grow under these conditions, it indicates
they both generated a beneficial allele during the aging regimen that let them survive and
continue to grow. However, it appears the serial (+10) population was able to better
capitalize on the beneficial GASP allele and grow faster than serial (+5). Given that the
serial (+10) population had more time to adapt to the selective pressures during serial
passage conditions, characterized by repeated outgrowth, it appears this population had
152
also evolved the capacity to grow faster under these conditions. Accordingly, this
unexpected finding appears to illustrate a confluence between the formation of GASP
alleles even during serial passage and the adaptation to conditions of rapid growth that
likely enabled the serial (+10) cells to increase in initial frequency faster during GASP
competitions.
Together, our findings indicate alleles that improve performance during long-term
stationary phase competitions are generated quickly within diverse growth conditions.
These beneficial mutations are not only generated and selected throughout batch culture
incubation, but also during serial passage conditions. Therefore, the respective aging
regimens both serve to create significant genetic diversity but serve to select and amplify
alleles beneficial under their respective conditions.
153
Appendix B: Characterization of the Escherichia coli death phase
B.1 Abstract
As nutrient availability and environmental stresses wax and wane, populations of
microbial communities ebb and flow. Despite the importance of cell death in
understanding the populations dynamics of bacteria in nature, few investigations have
characterized the death phase of the bacterial lifecycle. Using readily available plate-
based assays for cell viability, we characterized the Escherichia coli death phase within
batch culture conditions in the absence of exogenous stressor agents and demonstrate that
the timing, extent, and rate of cell death is remarkably consistent and predictable. We
found populations of bacteria exhibited exponential declines in cell density, and suggest
further characterization of these dynamics might eventually serve to elucidate the cellular
and environmental factors influencing bacterial cell death.
B.2 Introduction
Natural environments are inherently limited for nutrients, and population growth is
limited accordingly. As nutrient availability and environmental stresses wax and wane, so
do population densities. Though integral to understanding population dynamics in nature,
nutrient and environmental factors influencing cellular death within microbial
populations have remained ambiguous despite the availability of techniques amenable to
investigation for more than a century. While the importance of understanding how and
why microbial death occurs within natural environments is widely acknowledged, the
154
timing, rate, and extent of population decline within controlled conditions without
exogenous chemical stressors has rarely been described in the scientific literature. Here,
we demonstrate that the Escherichia coli death phase in rich medium represents an
exponential decline in population density where the onset, extent, and the rate of death all
appear to be remarkably predictable.
During long-term incubation in rich media, E. coli can exhibit five distinct phases of
growth (Figure 1.4). When bacteria are first inoculated in rich medium at low population
density, cells exhibit a lag phase before growing as they retool their physiology and
initiate synthesis of cellular components required for division. Soon thereafter, cells enter
what is called exponential or logarithmic phase growth, characterized by rapid cell
growth with populations doubling at a constant rate. This rapid growth will continue until
nutrients are depleted or toxic byproducts of metabolism accumulate in sufficient
concentrations to inhibit further growth. Once growth stops and the overall population
density no longer increases, cells enter stationary phase. Eventually, typically on the
order of 3 days during aerobic batch culture in rich medium, cells begin to die and
population density exhibits a significant decline known as death phase, often with 99% or
more of cells losing viability. Following death phase, E. coli can enter what is known as
long-term stationary phase where relatively stable population densities can be maintained
for extended periods (years) without providing additional nutrients (Finkel, 2006).
155
While it has long been know that cells incubated in batch culture will eventually lose
viability and enter death phase (Buchanan, 1918), the causal mechanisms and dynamics
of death phase remain unresolved. Furthermore, the initial triggers of death phase remain
enigmatic. It is often suggested that cell death is a stochastic byproduct of environmental
conditions, as a particular set of conditions can only support a limited number of cells for
a given period of time, after which cells begin to los the capacity to perform maintenance
and repair functions that can lead to the formation and accumulation of cellular damage
leading to cell death. Alternatively, the buildup of metabolic byproducts within the
extracellular environment might result in the loss of cellular integrity and cell death.
Intriguingly, it has also been proposed that cell death might be the byproduct of cell-
mediated apoptosis, or programmed cell death (Aldsworth et al., 1999; Engelberg-Kulka
and Glaser, 1999). In this model it has been proposed that over evolutionary time
populations of bacteria might have ‘learned’ that during nutrient limited conditions the
sacrifice of a portion of the overall population can facilitate the survival and eventual
reproduction of the remaining minority cells.
At present, there is little evidence to distinguish between these alternative models of cell
death, but preliminary investigations have begun to characterize the dynamics of cell
death. While very few reports have been published characterizing the bacterial death
phase, many authors have alluded to cultures exhibiting an exponential decline in
population density during death phase. However, the onset, extent, and dynamics of cell
death have not been systematically characterized. Here, we report on the dynamics and
156
reproducibility of the E. coli death phase in rich medium using standard plate-based
viability assays over time.
B.3 Materials and Methods
B.2.1 Strain, culture condition, media and titering assay
The E. coli K-12 strain ZK126 (W3110 ∆lacU169 tna-2) was grown in 5.0mL cultures
using various dilutions of LB Broth, Lennox (Difco-BD) and incubated at 37ºC with
aeration in a TC-7 test tube roller (New Brunswick Scientific). Cultures were either
grown in LB broth, or in a 1:10 or 1:100 dilution of LB broth in saline (0.5% NaCl).
Experiments were initiated from overnight cultures inoculated from frozen LB-glycerol
stocks. Viable counts were determined by plating serial dilutions of cells periodically
removed from cultures on LB agar. This method of titering is accurate within +/- 3-fold,
and has a limit of detection of 1000 colony forming units (CFU)/mL.
B.4 Results
B.3.1 Characterizing the five phases of E. coli growth in LB broth
In an attempt to characterize the dynamics of death phase in E. coli, two independent
time course experiments were conducted. During the first experiment, a dense overnight
culture of ZK126 was used to inoculate 8 technical replicates at a 1:1000(vol:vol)
dilution. Preliminary data suggested death phase commenced around the start of day 3
and ended before day 5 under these conditions, but it was not known precisely when it
would begin or end. Accordingly, titer counts were taken approximately every 24 hours
157
until hour 64, when data was taken every two hours for the following 36 hours, and
sporadically thereafter. The growth curves for all eight replicates obtained from this time
course are presented in Figure B.1a, with average data plotted in Figure B.1b.
From this data we observed a descending population density followed by a stable density
for the remaining data points, signifying the end of death phase and entry into long-term
stationary phase. However, the beginning of death phase was not observed in high
resolution. Additionally, the media used to dilute samples for titering appeared to become
contaminated around hour 85, creating a spike in the titer counts until the solution was
changed before taking the time point at hour 101.
With the information obtained from this initial experiment, a similar experiment was
conducted with a greater frequency of data collection over the period of interest. In
addition, four cultures were also inoculated in dilute LB (1:10 and 1:100) to determine
whether nutrient concentrations influenced the onset or rate of death phase. To clearly
observe the outgrowth and final densities achievable within dilute LB, populations were
inoculated at 1:1,000,000 (vol:vol) within this second experiment. Data points were taken
every two hours throughout the first 24 hours of growth, six hours later at hour 30, then
again (in what might best be described as a feat of masochism) every two hours between
hours 36 and 120 to characterize all phases of growth at high resolution (Figure B.2).
158
Figure B.1 Escherichia coli long-term growth curve (Trial #1).
Population density was determined for wildtype E. coli over time. (A) The growth curve
for eight independent cultures inoculated at 1:1000 (vol:vol) from a dense overnight
culture. (B) Average cell density over time. Asterisks (*) denote the period in which the
solution used to titer samples became contaminated.
7
9
11
8
9
10
0 20 40 60 80 100 120 140
log
10
CFU/mL
Hour
A
B
*
*
159
Figure B.2 Escherichia coli long-term growth curve in dilute LB (Trial #2).
Population density was determined for wildtype E. coli over time. (A) The growth curve
for sixteen independent cultures inoculated at 1:1,000,000 (vol:vol) from a dense
overnight culture. Eight were inoculated into 1xLB (Blue), four into a 1:10 dilution of LB
(green), and four into a 1:100 dilution of LB (red). (B) Average cell densities over time.
7!
9!
11!
7!
9!
11!
0! 20! 40! 60! 80! 100! 120! 140!
log
10
CFU/mL!
Hour!
A!
B!
160
In this experiment, all five phases of the bacterial growth curve can be observed in high
resolution within 1x LB. As might be predicted, each 10-fold dilution resulted in an
approximate 10-fold decrease in population density following outgrowth. The 1:100x and
1:10x LB cultures grew to maximal density following outgrowth and remained at stable
population densities, without death phase, throughout the duration of the experiment. The
culture grown in 1x LB exhibited all five phases of the bacterial growth curve, with death
phase occurring beginning at hour 56 and ending with entry into long-term stationary
phase at hour 82.
B.3.2 The timing of death phase
When assessing the bacterial growth curve, we sought to characterize the timing of death
phase and determining whether it commenced at the same time in different cultures. In
the first trial, using a 1:1,000 inoculum, death phase started at hour 47 while it
commenced at hour 56 in the second trial inoculated at a 1:1,000,000 dilution. This delay
in the second data set is likely the effect of the smaller inoculum density, as it would be
expected that this population would take longer to reach maximal density. As seen in
Figure B.3, when the average data from Trial #1 is advanced 7 hours to make the death
phases of each trial overlap, the initial density was very similar to the population density
of Trial #2 at that time.
161
Figure B.3 Timing of the Escherichia coli death phase.
Wildtype populations of E. coli were grown in LB from different inoculation densities
and the timing and dynamics of death phase were characterized. Population density of a
culture inoculated from an overnight culture at 1:1000 (vol/vol)(Trial #2, black)
compared to one inoculated at 1:1,000,000 (vol/vol)(Trial #1, blue). For comparison, the
data from Trial #1 is also presented with data points shifted forward by 7 hours [Trial #1
(+7), red].
5!
7!
9!
11!
0! 30! 60! 90!
Trial #2!
Trial #1!
Trial #1(+7)!
log
10
CFU/mL!
Hour!
162
B.3.3 Death phase is exponential
We also sought to characterize the dynamics of death phase and model the rate of decline
in population density. For both Trial #1 and Trial #2, the onset and duration of cell death
was determined, and used to formulate a best-fit trend line within Excel. Although the
regularity of data points at the onset of death phase measurements within Trial #1 was
low, the available data (hour 47) and remaining points exhibiting a decline in density
(hours 65-73) fit well with an exponential decline in cell density (Figure B.4a).
Similarly, the death phase within Trial #2 also exhibited an exponential decline. Due to
small variations in the average population density, it was unclear whether death phase
commenced at hour 56 or hour 60. It was determined that data from hour 56 and 68 fit
remarkably well with a modeled exponential decline based upon the remaining data
points observed during death phase (hours 64-78). When the two data points from hours
60 and 62 were eliminated from consideration, the modeled rate of exponential decline fit
the remaining data points extremely well.
163
Figure B.4 Modeling the Escherichia coli death phase.
The decline in average population density (black lines) over time was monitored from
cultures inoculated at a (A) 1:1,000 dilution and (B) 1:1,000,000. For each data set an
exponential trend line (dotted lines; red) was fit to the highlighted death phase (solid
lines; pink). Trend line equations and R
2
values are presented, along with calculated
halving times.
Trial #2
Halving time:
6.5 Hours
Trial #1
Halving time:
6.8 Hours
A
B
log
10
CFU/mL
y = 2E+12e
-0.103x
R² = 0.9981
50
60
70
80
90
8
9
10
y = 7E+11e
-0.099x
R² = 0.9964
40
50
60
70
80
8
9
10
Hour
164
Furthermore, using the modeled best-fit lined of exponential decline, the death phases
observed in both trials appeared very similar. Using the start points and end points of the
characterized death phases, we calculated the rate of population decline. Using the
formula N
t
= N
0
(1/2)
t/t
1/2
, (N
o
, pre-death phase density; N
t
, post-death phase density; t,
time; t
1/2
, half-life), the halving time of E. coli was determined to be 6.8 hours in Trial #1
and 6.5 hours in Trial #2. Therefore, the observed death rates were consistent in both
experiments.
B.5 Discussion
In this study we sought to characterize the bacterial death phase by monitoring population
density over time. In two separate trials, we observed the onset of death phase to occur at
approximately the same time once normalized to the cessation of exponential growth.
Similarly, the extent of cell death was approximately the same, with final populations
leveling off into long-term stationary phase at the same overall density. This decline in
population density fit well with a model of exponential decline in both trials, and
appeared to be consistent under these conditions. Compared to a doubling time of
approximately 20 minutes during exponential growth phase in LB, Escherichia coli
exhibited a halving time of approximately 6.5-6.8 hours during death phase under these
conditions.
Characterization of the dynamics of cell death within microbial population remains
important to not only understanding the lifecycles of environmental microbial
165
communities, but also the environmental and cellular factors that influence cell death.
Despite the importance of understanding how and why microbial death occurs within
natural environments, the timing, rate, and extent of population decline within controlled
conditions without exogenous chemical stressors has rarely been described within the
literature. Here, we have demonstrated the Escherichia coli death phase in LB represents
an exponential decline in population density where the onset, extent, and the rate of death
all appear to be remarkably predictable.
This onset of cell death has often been assumed to be the result of either cell starvation,
the buildup of toxic byproducts in the cell or environment, or a combination of both.
Relatively recently, it has also been suggested that death phase might be the result of a
form of prokaryotic programmed cell death or apoptosis. The exponential decline in
population density observed in this study is not inconsistent with any of these proposed
mechanisms of cell death, but serves to parameterize the dynamics of this period.
The fact that cell death is exponential indicates the same proportion of cells die per unit
time throughout this phase, and that the population decline is proportional to the overall
population density. With regards to a causal mechanism of cell death, our data is
consistent with a completely stochastic model of cell death. Much like the decay of
radioactive isotopes, the death of an individual cell might be entirely random and
impossible to predict. However, the probability of any individual cell losing viability at a
166
given time might be equally likely at any time, making cell death proportional to overall
population density.
Alternatively, an exponential decline in population density is also consistent with causal
mechanisms contingent on the concentration of one or more cellular or environmental
factors. For example, a variety of chemical and toxicological reactions are known to
exhibit exponential decay. The rate of many chemical reactions is depends on the
concentration of a single reactant, known as first-order reactions, and consequently
follow exponential decay. Interestingly, many enzyme-catalyzed reactions behave in this
manner, and would be consistent with a form of enzyme- or toxin-mediated cell death.
Similarly, many pharmacological or toxicological substances are distributed or
metabolized according to exponential decay patterns. Interestingly, the exponential decay
of population density could be consistent with the formation of an exogenous substance
that induces cell death, and the well-characterized pharmacokinetics of toxin clearance or
metabolism could explain how steady-state population densities are maintained
throughout long-term stationary phase.
While the environmental or cellular triggers of cell death were not specifically
investigated here and remain ambiguous, the procedures employed enable the
characterization of the mechanisms and factors influencing the onset, rate, and extent of
microbial death. Although the current investigation did not assess many of these factors
specifically, it does serve to indicate that the dynamics of death phase can be remarkably
167
consistent under controlled environments. Accordingly, this ubiquitous phenomenon is
also amenable to rigorous scientific investigation, and the results presented here perhaps
best serve to open the door for further questioning.
Despite the easily accessible and rudimentary techniques employed within this study, few
investigations have ever ventured to characterize the ubiquitous phenomenon of death
phase within microbial populations. What cellular events immediately precede the onset
of cell death? How does the growth environment change once cells begin to lose
viability? What cellular or environmental factors might induce, prevent, or mitigate cell
death? Is the rate of cell death consistent regardless of organism? Do the dynamics of cell
death show any correlation with other cellular characteristics such as growth or metabolic
rates? The answers to questions like these remain ambiguous despite initial observations
of death phase at least nearly a century ago, and can offer valuable insights into the
mechanisms and processes influencing the death of microbial populations.
168
Appendix C: Applications of whole-genome re-sequencing for the
characterization of genetic diversity generated during long-term
stationary phase.
Nearly two decades ago it was discovered that populations of bacteria incubated during
long-term stationary phase rapidly and consistently obtain beneficial alleles that confer a
selective advantage (Zambrano et al., 1993). Since these initial observations, an
increasing body of evidence continues to demonstrate that significant quantities of
genetic diversity are generated within microbial communities under these nutrient limited
conditions (Finkel, 2006; Finkel and Kolter, 1999). However, identification of these
novel alleles and characterization of adaptive and evolutionary dynamics have remained
largely impractical. However, next-generation sequencing technology offers the
compelling possibility of a method that could facilitate the efficient characterization of
adaptive changes and population structures in complex microbial communities.
To further characterize the evolutionary dynamics observed within long-term culture, we
have performed whole-genome re-sequencing on representative individuals and
populations following incubation during long-term batch culture within two distinct
experiments. In the first, clonal isolates obtained over the course of more than three years
of incubation can offer insight into the kinds of adaptive changes selected throughout
long-term culture. In the second, populations of cells derived from alternative DNA
polymerase-deficient strains were aged in batch culture to assess their impact on mutation
frequency and spectrum. While data analysis remains ongoing, preliminary observations
169
demonstrate the applicability and potential of whole-genome re-sequencing for the
investigation of experimental evolution.
C.1 Materials and Methods
Bacterial strains, culture media, and cell growth conditions. E. coli strains ZK1142
(Nal
R
) and ZK1143 (Str
R
) were incubated in 5.0mL of LB broth in 18 x 150mm glass test
tubes with constant aeration on a New Brunswick TC-7 roller at 37
o
C.
Long-term growth of bacterial cultures. Cultures of ZK1142 and ZK1143 were
incubated for more 1350 days without providing additional nutrients with titers remaining
at approximately 10
6
CFU/mL (data not shown). Sterile water was added to each culture
monthly to adjust for the loss of volume due to evaporation.
Isolation of cells with altered colony morphotypes. Every 30 days, cells from long-
term incubated batch cultures were plated onto LB agar and incubated for 24-48 hours at
37
o
C. Colonies showing distinct morphotypes were picked and restreaked onto LB agar
plates. LB-glycerol stocks were prepared from colonies with distinct morphologies at
various time points. Colonies were named according to the date at which they were
isolated, and their appearance (c, normal; O, opaque; E, “fried egg”; M, mini; Sm; small).
Batch culture aging of alternative DNA polymerase-deficient strains. Wildtype
(ZK126) and every combination of alternative DNA polymerase-deficient strains (see
170
Table 2.1 for nomenclature: Pol II
-
, Pol IV
-
, Pol V
-
, Pol II
+
, Pol IV
+
, Pol V
+
, and -/-/-)
were grown in LB at 37
o
C for 10 days, and populations were frozen in LB-glycerol
stocks. These stocks were used to inoculate fresh cultures and DNA was harvested
following overnight growth.
DNA isolation and genomic library preparation. Isolates with distinct colony
morphologies were outgrown from LB-glycerol stocks in 5.0mL LB broth overnight prior
to harvesting DNA. Genomic DNA was prepared from each overnight culture using the
Gentra Puregene Yeast/Bacteria Kit (QIAGEN) as specified by the protocol for DNA
purification from Gram-negative bacteria. Genomic libraries were constructed as
described in Preparing Samples for Sequencing Genomic DNA (Illumina). Libraries were
constructed as specified according to manufacturer’s recommendations except for the
fragmentation of DNA, which was performed by sonication instead of nebulization.
Sequencing. Genomic DNA from 18 samples was sequenced as either single-ended 44bp
reads or paired-ended 41bp or 36bp reads at the Cambridge Research Institute (University
of Cambridge, United Kingdom). Each sample was sequenced across one flow cell lane
on an Illumina GAII, generating a hundreds of millions of reads and billions of bases.
Sequenced reads were aligned using BWA 0.5.7 (Li and Durbin, 2009) to the complete E.
coli K-12 MG1655 genome sequence available from NCBI. Reads were allowed up to 3
mismatches per end and unique reads were mapped to the genome. All other BWA
alignment parameters were set as default. Approximately 97% of the reads mapped
171
uniquely to the genome, resulting in more than 4600X coverage of the 4.6Mb genome
across all 18 samples. All sequence data will be deposited in the Sequence Read Archive
of NCBI before publication. Peter Chang (University of Southern California, Los
Angeles) performed all data processing, including single-nucleotide polymorphism (SNP)
calling and copy number variation (CNV) analysis, in conjunction with the author.
SNP calling and allele frequency differences. Distributions of read depth across the
genome show that the average depth was 100 reads in the single-ended samples and 400
reads in the paired-ended samples. Based on these distributions, genomic positions that
were covered by less than 10 reads in the single-ended samples and 100 reads in the
paired-ended samples were not analyzed for allelic differences. To identify genomic
positions with differences in allele frequencies, a Fisher Exact test was used to compare
the counts of all base calls at each position between any two samples. We applied an
FDR cutoff of 0.05 to call allele differences significant (Benjamini and Hochberg, 1995).
Copy number variation and structural rearrangements. To search for CNVs, the
average read depth was calculated for all nonoverlapping 100bp windows of the genome
for each sample. CNVs were identified using the segmentational program CNV-seq.
172
C.2 Identification of genomic changes following prolonged incubation
in long-term stationary phase using whole-genome re-sequencing.
Whole-genome re-sequencing was first applied to clonal isolates displaying altered
colony morphology following prolonged incubation in long-term stationary phase. Here,
four initially isogenic cultures were maintained and regularly sampled for more than 1000
days. When samples from each of these populations were plated on nutrient rich agar
plates, it was discovered that numerous distinct colony morphologies could be identified
within a single time point. Representative colony morphotypes obtained from one of
these independent lineages were selected for further analysis.
Within this culture, colony morphotypes remained beige in color and matte in appearance
throughout the first 120 days of incubation. On day 150, three colony variants were
discovered: parental, opaque, and mucoid. By day 180 colonies had appeared that
exhibited ruffled edges like a fried egg. Still later, miniature colonies that were mucoid
were identified on day 240, while opaque miniature colonies formed by day 750.
Eventually, novel small colonies were seen on days 840 and 900.
Given these morphologies remained stable when colonies were picked and restreaked,
these readily observable phenotypes reflect the presence of several distinct genotypes
present within these aged populations at any given time. Furthermore, the proportion of
each morphology was found to rise and fall over the course of the experiment, reflecting
dynamic population structures within these cultures. Genomic analysis using complete
173
genomic hybridization onto Affymetrix DNA arrays (Skvortsov et al., 2007) confirmed
that all 20 colony variants assessed were descendants from the same parent, and not the
result of contamination. While large-scale amplifications and deletions were uncovered,
we sought to further assess the genetic composition of eight selected colony variants
using whole-genome re-sequencing to gather higher resolution information concerning
the genetic changes within each isolate. Assessment of these data remains ongoing, and
consists primarily of two forms of analysis: (1) single-nucleotide polymorphism (SNP)
identification and (2) copy number variation (CNV) analysis.
Preliminary analysis has already identified numerous SNPs within the genomes of each
evolved strains that have changed relative to the original parental strain. Once completed,
a comprehensive assessment of this data will enable the identification base changes in the
genomes of evolved clones that are likely either neutral or beneficial within the
conditions experienced in long-term stationary phase. The overall abundance of
mutations could address long-standing questions and shed new light on the rate of
mutations and extent of genetic diversity under stressful conditions. Furthermore, the
location (coding or non-coding), and nature (synonymous or non-synonymous) of each
change, as well as their conservation over time, could provide valuable information
concerning regulatory or protein sequences under strong selection within these
environments.
174
Initial analysis of copy number variation has already revealed large-scale differences
among isolates from evolved populations. In this analysis, changes in the overall
abundance of sequence reads covering a region of the genome are used to identify
deletions or amplifications: the absence of reads indicates a deletion, while a rapid and
significant increase in read abundance signals an amplification. Base coverage was
averaged across nonoverlapping 100 base pair windows, enabling the detection of
amplifications or deletions on the order of hundreds of bases or more within this analysis.
As readily observed in Figures C.1-C.5, large-scale deletions and amplifications can be
observed in each clone (summarized in Table C.1).
175
Figure C.1 Analysis of chromosomal copy number variation in Wildtype 1 and
Wildtype 2.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for unaged populations of parent strains (A) Wildtype 1 and (B) Wildtype
2. Triangles indicate previously characterized deletions within parental strains.
Wildype 1!
Wildtype 2!
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
A!
B!
176
Figure C.2 Analysis of chromosomal copy number variation in colonies 150-c4
and 150-c5.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for cells from colonies (A) 150-c4 and (B) 150-c5. Amplifications not
observed in the parental wildtype are indicated with red brackets.
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
150-c4!
150-c5!
A!
B!
177
Figure C.3 Analysis of chromosomal copy number variation in colonies 150-O
and 180-E.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for unaged populations of parent strains (A) 150-E and (B) 180-E.
Amplifications (red brackets) and deletions (red triangles) not observed in the parental
wildtype are indicated.
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
150-O!
150-E!
A!
B!
178
Figure C.4 Analysis of chromosomal copy number variation in colonies 240-M
and 750-MO.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for cells from colonies (A) 240-M and (B) 750-MO. Amplifications (red
brackets) and deletions (red triangles) not observed in the parental wildtype are indicated.
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
240-M!
750-MO!
A!
B!
179
Figure C.5 Analysis of chromosomal copy number variation in colonies 840-Sm-
O and 900-Sm-2.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for cells from colonies (A) 840-Sm-O and (B) 900-Sm-2. Amplifications
(red brackets) and deletions (red triangles) not observed in the parental wildtype are
indicated.
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
840-Sm-O!
900-Sm-2!
A!
B!
180
Table C.1 Characteristics of large-scale amplifications and deletions observed in
aged colony morphotypes.
Colony Change Start End Coverage Size (kb)
Parental
Deletion 273300 372000 - 98.7
Deletion 1410100 1433000 - 22.9
Deletion 3128400 3146800 - 18.4
150-c4 Amplification 1588900 1645100 +5-6x 56.2
150-c5 Amplification 1588900 1645100 +2-3x 56.2
150-O Deletion 3629400 3636500 - 7.1
180-E Amplification 2995800 3363600 +2x 367.8
240-M Amplification 2995800 3363600 +1x 367.8
750-MO
Deletion 1195600 1210600 - 15.0
Deletion 4348500 4364000 - 15.5
840-Sm-
O
Amplification 700 931800 +1x 931.1
Deletion 570500 585000 - 14.5
900-Sm-2
Deletion 1195600 1210600 - 15.0
Amplification 1342400 1529400 +4-5x 187.0
Amplification 2995800 3363600 +2x 367.8
181
When first assessing the unaged wildtype ZK126 read coverage depth aligned to the
reference MG1655 sequence we identify three previously characterized deletions
(marked with triangles) inherent to our parental strain. Accordingly, these deletions are
present in all colonies assays. While both wildtype control populations appear identical
with regards to amplifications or deletions, we see changes quickly accrue within evolved
populations.
As seen in Figures C.2a and C.2b, by day 150 an ~56.2kb amplification was present in
morphotypes 150-c4 and 150-c5. Interestingly, the read depth of this region was
approximately 5-6x greater than the average coverage across the remaining genome
within 150-c4, while it was approximately 2x greater in the 150-c5. This observation
suggests this region was amplified to different extents in each clone, and is consistent
with observations made by Skortzov et al. (2007) using Affymetric arrays. Furthermore,
the 150-O obtained from the same population (Figure C.3a) did not have this
amplification, but did have an ~7.1kb deletion not observed in the other two strains.
Accordingly, these preliminary findings verify that at least three large-scale differences
were already present by day 150.
Novel large-scale alterations were also observed at later time points as the culture
continued to age. As seen in Figure C3b, there was a massive 367.8kb amplification
identified 30 days later in the 180-E clone that was covered approximately 2x greater
than the rest of the chromosome. In the clone isolated at the next representative time
182
point (Figure C.4a), the same region was amplified, but with an average 1x greater depth,
indicating one of the additional copies of this region was not present within this
genotype.
Within colony morphotypes isolated at significantly later time points, we continued to
observe novel changes. The 750-MO colony revealed two deletions of approximately
15.0kb and 15.5kb (Figure C.4b). Neither of these deletions were detected in the 840-Sm-
O colony (Figure C.5a), but rather a 0.9Mb amplification was detected that represents the
duplication of an astounding ~1/5 of the E. coli genome. However, prior to this massive
amplification, a 14.5kb deletion must have occurred given the absence of coverage within
the amplified region. Finally, within the 900-Sm-2 colony two distinct and amplifications
were identified (Figure C.5b). One appeared to be a novel 187kb amplification an
approximately 4-5x increase in coverage, along with the identical 367.8kb amplification
observed in 180-E. Similarly, same 15.0kb deletion detected in the 750-MO was also
found in the 900-Sm-2 clone.
When these amplifications and deletions are considered in the context of each other, it
reveals a remarkable number of large-scale rearrangements are created and maintained
within long-term cultures. The fact that same 2x amplification could be observed in 180-
E and 900-Sm-2 indicates at least a subpopulation of cells with this amplification
remained present for at least 720 days. Similarly, the identical deletion detected in 750-
MO and 900-Sm-2 suggests these clones are also related. With these genomic
183
rearrangements taken together, it implies the amplification likely occurred prior to day
180, but was removed in a subpopulation that gave rise to 750-MO.
While only large-scale genomic abnormalities can be readily observed with CNV data at
this time, future SNP analysis might inform our understanding of cell lineages within the
few clonal isolates identified in the current study. While this comprehensive analysis
remains incomplete, our CNV data demonstrates a remarkable number of genomic
rearrangements occurring within populations of E. coli maintained in long-term culture.
Furthermore, we have shown how these alterations can be stably maintained within
subpopulations of different cell lineages competing for survival within these stressful
environments. As informative as these data are at the moment, they provide only a small
glimpse into the abundant genetic diversity and population dynamics within long-term
stationary phase.
184
C.3 The characterization of genetic diversity within aged alternative
DNA polymerase deficient strains using whole-genome
re-sequencing.
With improving sequencing capabilities, the ease of generating massive base coverage
continues to steadily rise. Accordingly, as read depths increase, we improve the capacity
to distinguish between background levels of inherent sequencing errors from actual SNP
variants present at low frequencies in the sample. Functionally applied to the analysis of
microbial populations, this technology can theoretically enable the identification and
tracking of bacterial subpopulations present at low frequency.
Practically speaking, within an organism like E. coli with a comparatively small genome,
we have already achieved average sequencing coverage of ~100x coverage across the
entire genome. Therefore, even after considering background sequencing errors and
conservative limits of detection, it is already possible to observe low frequency SNPs,
indicative of minority subpopulations, when present as low as approximately 10% of the
overall population. With sequencing capacity continuing to improve, this technology
opens the door for numerous applications for studying the formation of genetic diversity,
population dynamics, and experimental evolution.
Similar to the analysis of clonal isolates from cultures experiencing prolonged stationary
phase, we are also attempting to utilize whole-genome re-sequencing technology to
identify and characterize the formation of genetic diversity within populations of bacteria.
185
Within this experiment, alternative DNA polymerase-deficient strains were incubated in
batch culture for a comparatively brief 10 days before populations were harvested for
sequencing. The aim of this study was to quantify the abundance and characterize the
spectrum of mutations generated within each polymerase background to determine the
mutagenic effects of each alternative DNA polymerase.
Given previous findings demonstrating the impact of each alternative DNA polymerase
on the evolutionary fitness and the formation of rifampicin resistance, we sought to
analyze the number and characteristics of mutations observed across the entire genome
rather than one well-characterized gene. Accordingly, we allowed each strain to age for
10 days in batch culture before harvesting populations of each culture for whole-genome
resequencing. CNV analysis on these aged polymerase-deficient strains has revealed few
large-scale abnormalities (Figures C.6 – C.9).
186
Figure C.6 Analysis of chromosomal copy number variation in aged populations
of Wildtype and Pol II
-
.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for populations of aged cells from the (A) wildtype strain and (B) Pol II
-
.
Wildtype (+10)!
Pol II
-
(+10)!
A!
B!
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
187
Figure C.7 Analysis of chromosomal copy number variation in aged populations
of Pol IV
-
and Pol V
-
.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for populations of aged cells from strains (A) Pol IV
-
and (B) Pol V
-
.
Pol IV
-
(+10)!
Pol V
-
(+10)!
A!
B!
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
188
Figure C.8 Analysis of chromosomal copy number variation in aged populations
of Pol II
+
and Pol IV
+
.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for populations of aged cells from strains (A) Pol II
+
and (B) Pol IV
+
.
Amplifications (red brackets) not observed in the parental wildtype are indicated.
Pol II
+
(+10)!
Pol IV
+
(+10)!
A!
B!
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
189
Figure C.9 Analysis of chromosomal copy number variation in aged populations
of Pol V
+
and Pol -/-/-.
Average read depth was calculated for nonoverlapping 100bp windows of each genome,
and presented for populations of aged cells from strains (A) Pol V
+
and (B) Pol -/-/-.
Pol V
+
(+10)!
-/-/- (+10)!
A!
B!
Average Read Depth!
1!
10!
100!
1000!
0! 1M! 2M! 3M! 4M!
Chromosome Position!
1!
10!
100!
1000!
190
Neither of the wildtype aged populations showed any unique amplifications or deletions,
and nearly all 10-day-old polymerase-deficient populations appeared similar to wildtype.
Given this CNV analysis is performed on populations of cells rather than clonal isolates,
this finding was expected since nearly all individuals in the population would need to
possess the same amplification or deletion for the effect to be detected. The only
exception was the Pol IV
+
strain, which displayed a large 112kb amplification (between
575,100 - 687,100). Accordingly, this would indicate that either this amplification was
present at the time of inoculum, or that by day 10 of incubation a subpopulation with a
large and beneficial amplification of this region had already begun to sweep through the
culture to detectable levels. Overall, while the low frequency of large-scale aberrations
seen in the CNV data is consistent with expectations, SNP analysis would be expected to
more appropriately address our aim of identifying overall mutation frequency and
spectrum.
While SNP analysis within these aged polymerase-deficient strains remains a work in
progress, we can report the identification loci with statistically significant differences in
allele frequencies relative to a representative unaged wildtype population. However, the
definitive identification and analysis of mutations generated over the course of these
experiments remains ongoing.
Given that fixed SNP calls likely represent base changes present from the time of
inoculum, these loci will likely be disregarded from further analysis since they do not
191
reflect novel mutations originating during the course of aging. Among the remaining
positions of interest, it is necessary to identify loci with SNP frequencies above
background sequencing error and significantly different from any background variation
present in the frozen LB-glycerol stocks used to inoculate each culture. While a
comprehensive and reliable analysis accounting for these concerns has not been achieved
at this time, there are instances where noticeable shifts in allele frequency appear to
indicate the presence of a genuine mutation within sequenced populations.
While these findings are promising, they also highlight yet another challenge when
interpreting the results of this particular experiment. While we allowed populations to age
for 10 days to allow sufficient levels of genetic diversity to arise to detectable levels, this
also provided enough time for beneficial alleles to be selected within the population.
Therefore, the increase in frequency of individuals with advantageous mutations would
be expected to confound analysis of overall mutation frequency and suggest an artificially
high mutation frequency. However, the growth and selection of novel beneficial
mutations would also be expected to occur with a concurrent decrease in the abundance
of other genotypes within the population, and this purifying selection could prevent the
detection of rare mutations resulting in an underestimation of overall genetic diversity.
While the effects of these selective forces must be considered when assessing overall
mutation frequency, they also illustrate the capacity to monitor population dynamics and
selection within these microbial communities.
192
Despite the challenges of interpreting whole-genome re-sequencing data for the purposes
described here, the potential promise of such techniques remains compelling. Upon
establishing baseline expectations for inherent sequencing errors and background
variation in initial populations, the development of an automated analysis system for the
identification and comparison of mutations generated within bacterial populations could
offer fundamental insights concerning the abundance and formation of genetic diversity.
Similarly, SNP calling for the purpose of identifying low frequency subpopulations
within dynamic environmental or laboratory communities of microorganisms could
potentially revolutionize our capacity to characterize population dynamics and adaptive
evolution in real-time at high resolution.
Abstract (if available)
Abstract
Escherichia coli DNA polymerases II, IV and V serve dual roles within cells by facilitating efficient replication past potentially lethal DNA damage while simultaneously introducing genetic variation that can promote adaptation and evolution within stressful environments. While long recognized to be important for these physiological and evolutionary roles, the specific molecular mechanisms and relative contributions attributable to each of these alternative DNA polymerases within natural environmental conditions has remained elusive. Using a series of alternative polymerase-deficient strains analyzed during conditions of both feast and famine, we establish distinct hierarchies of polymerase activity. Pol II confers a significant physiological advantage by facilitating efficient replication and creating genetic diversity during periods of rapid growth, whereas Pol IV and Pol V make the largest contributions to evolutionary fitness during long-term stationary phase. Pol V is responsible for maximizing allelic diversity, yet Pol IV is the single greatest determinant of mutation frequency. Furthermore, we demonstrate that these alternative polymerases, along with additional members of the SOS regulon, are induced as cells transition from exponential to stationary phase growth in the absence of exogenous stress-stimulated SOS induction, and that they remain elevated throughout long-term stationary phase. These findings reveal each alternative DNA polymerase is vital to physiological and evolutionary fitness under dynamic and unpredictable conditions akin to those experienced in nature, and indicate their contributions to replication and adaptation within microbial communities are greater than previously appreciated.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Investigating the E. Coli DNA polymerase V mutasome
PDF
Genetic diversity and bacterial death in the context of adaptive evolution
PDF
Activation mechanism of damaged-induced DNA polymerase V in Escherichia coli
PDF
The role of ATP in the regulation of Escherichia coli DNA polymerase V activity
PDF
Decrypting Escherichia coli DNA polymerase V mutasome ATP regulation
PDF
Investigating highly mutagenic polymerases in pathogenic bacteria
PDF
The mechanism of damage-induced mutations in Escherichia coli
PDF
Survival and evolution of Shewanella oneidensis MR-1: applications for microbial fuel cells
PDF
Generation of monoclonal antibodies via phage display and in vitro affinity maturation using activation induced deoxycytidine deaminase and DNA polymerase eta
PDF
Evolution of E. coli in unstructured and structured environments
PDF
Mechanisms of long-term survival in Escherichia coli
PDF
Nutritional competence in Escherichia coli: the use of double-stranded DNA as a nutrient
PDF
New tools for whole-genome analysis of DNA replication timing and fork elongation in saccharomyces cerevisiae
PDF
The intra-S phase checkpoint and its effect on replication fork dynamics in saccharomyces cerevisiae
PDF
Genomic and physiological characterization of Escherichia coli evolving in long-term batch culture
PDF
Mechanism of human nonhomologous DNA end joining
PDF
Molecular genetics of longevity in Escherichia coli
PDF
Mechanisms of long-term survival of Vibrio harveyi and Escherichia coli: linking damage and senescence
PDF
Genome-wide GATC methylation in long-term cultures of Escherichia coli K-12
PDF
Structural and biochemical studies of large T antigen: the SV40 replicative helicase
Asset Metadata
Creator
Corzett, Christopher Hale
(author)
Core Title
Physiological roles and evolutionary implications of alternative DNA polymerases in Escherichia coli
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Molecular Biology
Publication Date
11/12/2012
Defense Date
05/03/2012
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
alternative DNA polymerase,error-prone DNA polymerase,microbial evolution,OAI-PMH Harvest,SOS response,stationary phase,translesion synthesis
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Finkel, Steven E. (
committee chair
), Goodman, Myron F. (
committee member
), Goodman, Steven D. (
committee member
), Nealson, Kenneth H. (
committee member
)
Creator Email
corzett@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-110357
Unique identifier
UC11288740
Identifier
usctheses-c3-110357 (legacy record id)
Legacy Identifier
etd-CorzettChr-1285.pdf
Dmrecord
110357
Document Type
Dissertation
Rights
Corzett, Christopher Hale
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
alternative DNA polymerase
error-prone DNA polymerase
microbial evolution
SOS response
stationary phase
translesion synthesis