University of Southern California Dissertations and Theses

Genetic diversity and bacterial death in the context of adaptive evolution

(USC Thesis Other)

Genetic diversity and bacterial death in the context of adaptive evolution

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

GENETIC DIVERSITY AND BACTERIAL DEATH IN THE
CONTEXT OF ADAPTIVE EVOLUTION

By

Christina M Ferraro

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MOLECULAR BIOLOGY)

MAY 2019

ii
Acknowledgments
First, I would like to thank my Ph.D. advisor, Dr. Steven Finkel, for his guidance and support over
the past six years. His support has been invaluable, and I will forever be thankful to him for that.
Thank you to my committee as well for their support and guidance over the years. I would also
like to thank my lab mates and friends that I have gained during this time, particularly Nicole
Ratib, Lacey Westphal, Namita Shroff, Karin Kram, and Chris Corzett.

Also, thank you to my family. Without them, I do not think that it is an exaggeration to say that
none of this would be possible. I feel immensely lucky to have such a great support system. Finally,
and most importantly, thank you to my partner Jonathan Lee for his never-ending support over the
years. We met during our USC graduate school interviews, and he has been a pillar of support ever
since.

iii
Table of Contents
Acknowledgments………………………….………………………………………………... ii
List of Tables……………………………………………………………………………......... v
List of Figures………………………………………………………………………………… vi
Abstract……………………………………………………………………………………... . 1
Chapter 1 Dynamics of Stationary Phase and Bacterial Death
in Batch Culture……………………………………………………………………………... 2
1.1 Abstract………………………………………………………………………….. 2
1.2 Introduction……………………………………………………………………… 3
1.3 References……………………………………………………………………….. 17

Chapter 2 Environment-Specific Signatures of Alternative DNA Polymerase
Activity and Mutational Spectra in Escherichia coli………………………………………… 27
2.1 Abstract…………………………………………………………………………... 27
2.2 Introduction………………………………………………………………………. 29
2.3 Materials and Methods…………………………………………………………… 31
2.4 Results……………………………………………………………………………. 33
2.5 Discussion……………………………………………………………………….... 39
2.6 References………………………………………………………………………… 44

Chapter 3 Physiological, Genetic, and Transcriptomic Analysis of Alcohol-
Induced Delay of Bacterial Death in Escherichia coli………………………………………... 49
3.1 Abstract…………………………………………………………………………… 49
3.2 Introduction……………………………………………………………………….. 50
3.3 Materials and Methods……………………………………………………………. 52
3.4 Results…………………………………………………………………………….. 55
3.5 Discussion………………………………………………………………………… 66
3.6 References………………………………………………………………………… 72
3.7 Supporting Material………………………………………………………………. 79

Chapter 4 Comprehensive Physiological and Genetic Profiling of Alcohol-
Induced Delay of Bacterial Death of Escherichia coli………………………………………... 85
4.1 Abstract……………………………………………………………………………. 85
4.2 Introduction………………………………………………………………………... 86
4.3 Materials and Methods…………………………………………………………….. 87
4.4 Results……………………………………………………………………………... 89
4.5 Discussion…………………………………………………………………………. 102
4.6 References…………………………………………………………………………. 108

iv
Chapter 5 The Far Reach of Alcohol: Alcohol-Induced Bacterial Death Delay
in Other (Non-Escherichia coli) Microbes…………………………………………………….. 111
5.1 Abstract……………………………………………………………………………. 111
5.2 Introduction………………………………………………………………………... 112
5.3 Materials and Methods…………………………………………………………….. 113
5.4 Results……………………………………………………………………………... 116
5.5 Discussion…………………………………………………………………………. 122
5.6 References…………………………………………………………………………. 126

Chapter 6 Concluding Remarks……………………………………………………………….. 129
6.1 Conclusion and broader impact……………………………………………………. 129
6.2 Future direction……………………………………………………………………. 131
6.3 References…………………………………………………………………………. 133

Appendix A: RNA-sequencing and Whole Transcriptome Analysis of Alcohol-
Induced Delay of Escherichia coli Death………………………………………………………134
A.1 Brief Comments and Experimental Design……………………………………….. 134

v
List of Tables

Table 1.1 Brief list of bacteria with alternative sigma factors (Adapted
from Jaishankar & Srivastava, 2017)…………………………………………………… 7
Table 2.1 Strains used in this study (Adapted from Corzett et al., 2013)……………… 33
Table 2.2 List of Rif
R
mutations identified in this study and number of
occurrences………………………………………………………………………………41
Table 3.1 E. coli strains used in this study………………………………………………52
Table 3.2 Chemicals tested for delayed death effect……………………………………60
Table 3.3 Most significantly upregulated genes one hour after ethanol
addition…………………………………………………………………………………. 63
Table S3.1 Significantly upregulated genes one hour after ethanol addition…………...79
Table S3.2 Significantly downregulated genes one hour after ethanol
addition………………………………………………………………………………......80
Table 4.1 E. coli strains used in this study………………………………………………87
Table 4.2 Potential alcohol effect gene candidate descriptions…………………………97
Table 5.1 List of strains used in this study……………………………………………..114
Table 5.2 List of species and overall alcohol effect…………………………………....121

vi
List of Figures

Figure 1.1 Brief summary of the changes that occur between stationary
phase and LTSP……………………………………………………………………… 6
Figure 1.2 Schematic of physiological portrayal of delay of bacterial death
phase…………………………………………………………………………………. 14
Figure 2.1 Wild-type strain grown in different media show altered growth
phenotypes…………………………………………………………………………… 34
Figure 2.2 Distribution of media-specific Rif
R
mutation frequency………………… 35
Figure 2.3 Media-specific mutation spectra………………………………………… 36
Figure 2.4 Transversion:transition ratios of bacteria grown in each medium……….. 37
Figure 2.5 Alternative DNA polymerase mutant strains grown in each medium
look indistinguishable from wild-type………………………………………………... 38
Figure 3.1 Wild-type strain treated with different concentrations of ethanol on
Day 1 shows a dose-dependent alcohol effect………………………………………… 55
Figure 3.2 Depletion of ethanol over time in active cultures…………………………. 58
Figure 3.3 Daily ethanol addition results in a prolonged stationary phase effect…….. 59
Figure 3.4 Other n-alcohols can produce the alcohol effect………………………….. 60
Figure 3.5 An rpoS null mutant strain exhibits the alcohol effect……………………. 61
Figure 3.6 An alcohol dehydrogenase double mutant strain exhibits the alcohol
effect…………………………………………………………………………………… 62
Figure 3.7 Single gene null mutations of glycolate degradation pathway genes
glxR, glcD, and gcl, have altered delayed death phenotypes………………………….. 65
Figure S3.1 The majority of natural isolate and laboratory strains show the
delayed death effect induced by the addition of ethanol………………………………. 83
Figure S3.2 Single gene null strains corresponding to mutations in glycolate
degradation pathway genes have altered delayed death phenotype……………………. 84
Figure 4.1 The “alcohol effect” is a stationary phase-specific phenomenon………….. 89
Figure 4.2 Multiple small ethanol doses cause a stronger effect than a one-time
larger dose……………………………………………………………………………… 90
Figure 4.3 Daily ethanol additions delay death phase for ~20 days…………………… 91
Figure 4.4 Depletion of ethanol over time in active cultures…………………………... 91
Figure 4.5 Change in pH is not the main cause of the alcohol effect………………….. 93
Figure 4.6 Alcohol effect is not caused by cell division……………………………….. 94
Figure 4.7 Distribution of Rif
R
mutation frequency due to ethanol addition…………... 95
Figure 4.8 Human and canine natural isolate strains exhibit the alcohol effect………... 96
Figure 4.9 RpoS activity affects entry into death phase in treated and untreated
cultures…………………………………………………………………………………... 97
Figure 4.10 Growth physiologies of frmA, frmR, and frmAR mutant strains…………… 98
Figure 4.11 RNA-seq candidate genes do not cause the alcohol effect………………….. 99
Figure 4.12 Supplementing cultures with glycolate degradation pathway substrates
do not affect cell populations…………………………………………………………….. 100
Figure 5.1 Many strains exhibit the alcohol effect……………………………………… 117
Figure 5.2 There are species that do not exhibit the alcohol effect……………………… 118
Figure 5.3 Some species do not show an alcohol effect but exhibit other phenotypes
in response to alcohol addition…………………………………………………………… 120
vii
Figure A.1 Significantly upregulated genes 1 hour after ethanol addition……………… 135
Figure A.2 Significantly downregulated genes 1 hour after ethanol addition…………... 136
Figure A.3 Significantly upregulated genes 3 hours after ethanol addition…………….. 137
Figure A.4 Significantly downregulated genes 3 hours after alcohol addition…………. 138
Figure A.5 Significantly upregulated genes 24 hours after alcohol addition…………… 139
Figure A.6 Significantly downregulated genes 24 hours after alcohol addition………… 140

1
Abstract
Both genetic diversity and bacterial death play large roles in growth and survival of bacterial
populations. In batch culture conditions, where bacteria undergo lag phase, exponential phase,
stationary phase, death phase, and long-term stationary phase (LTSP), expression of alternative
DNA polymerases (polB, dinB, and umuDC) is extremely important to creating that genetic
diversity and causing mutations that may aide in population survival in LTSP. Many of these
beneficial mutations appear prior to death phase and often do not take over the population until
after death phase. This thesis characterizes how genetic diversity, through mutation frequency and
mutation spectrum analysis via a rifampicin-resistance (Rif
R
) reporter system, is affected by
environmental changes. Further, death phase occurrence helps allow for beneficial mutations to
overtake the population upon entry into LTSP. Given the inherent difficulty in studying dying
populations, my work utilizes a phenomenon known as the “alcohol effect,” where addition of
small, sublethal doses of alcohols between two and six carbons in length can delay the onset of
death phase in a dose-dependent manner. I characterize that the alcohol effect is not strain or
species-specific and potentially allows for a better understanding regarding how bacterial
populations die.

2
Chapter 1: Dynamics of Stationary Phase and Bacterial Death in Batch Culture

The content of this chapter is a draft of an invited minireview manuscript in preparation for Applied
& Environmental Microbiology: Ferraro CM, Finkel SE., “Death after life: Dynamics of
stationary phase and bacterial death in batch culture.”

1.1 Abstract
Death phase remains a topic that is little understood in the life cycle of bacteria in the laboratory.
Dogma stipulates that the onset of death phase in Escherichia coli is a primarily stochastic event,
with a few notable exceptions, in which ~99% of the bacterial population lose viability caused by
a loss of nutrients and buildup of metabolic waste products. However, the death phase transition
from stationary phase to long-term stationary phase (LTSP) is essentially a “black box,” having
implications for many other actively studied phenomena such as the viable but nonculturable
(VBNC) state and bacterial persistence. This review introduces and analyzes current thinking on
the topic of bacterial death, as well as explores the requirement for novel approaches to better
understand the mechanisms encountered during this important phase of laboratory incubation.

3
1.2 Introduction
According to the “feast or famine” model of survival (Koch, 1971), bacteria outside of laboratory
conditions mainly live in extremely nutrient-poor conditions with the occasional ‘feast,’ a usually
short period of time where microbes have plentiful carbon, nitrogen, phosphate, and/or energy
sources. The famine condition is prevalent in both terrestrial and marine environments (Grimm et
al., 2003). For example, in marine environments, microbes are thought to survive for thousands of
years or more while buried under progressively deeper layers of sediment, remaining viable under
the most minimal conditions of nutrient availablity (Jorgensen & Marshall, 2016; Lever et al.,
2015; Starnawski et al., 2017; Walsh et al., 2016). These bacteria are estimated to replicate
extremely slowly, with potential generation times of hundreds of years for a single cell division
event to occur (Lever et al., 2015). Terrestrial conditions also have extreme variation in nutrient
availability, often affected by temperature, water flow, and wind (Grimm et al., 2003; Hartman &
Richardson, 2013). Even in the mammalian gut, bacterial populations undergo numerous batch
culture-like stresses, including periods of feast or famine, pH stress, and oxygen stress (Pereira &
Berry, 2017).

Batch culture incubation can emulate these famine conditions (Finkel, 2006). When incubated in
a nutrient-rich medium such as Luria-Bertani (LB) broth, Escherichia coli experience five phases
in the laboratory: Lag phase, log or exponential phase, stationary phase, death phase, and long-
term stationary phase (LTSP). While exponential and stationary phase are mechanistically fairly
well-defined (Finkel, 2006), the other three phases are considerably less well understood. LTSP is
a famine-like period of cell survival in which E. coli populations have long since exhausted the
readily metabolizable nutrients initially found in the medium and are constantly evolving,
4
apparently surviving by metabolizing the detritus of the cells in the culture that have died (Finkel
& Kolter, 2001). A key feature of LTSP is the emergence of cells expressing the growth advantage
in stationary phase (GASP) phenotype (Zambrano et al. 1993; Finkel, 2006). The GASP phenotype
is defined by an aged population’s ability to outcompete its unaged parent population when
cocultured (Zambrano et al. 1993; Zambrano & Kolter, 1996; Finkel & Kolter, 1999; Finkel,
2006).

Essential to understanding a population’s entry into LTSP is, first, understanding the bacterial
death phase. When death phase occurs in rich media in the laboratory, >99% of the population
frequently loses cell viability while the survivors enter LTSP (Finkel & Kolter, 2001; Finkel,
2006). Understanding which cells in the population survive death phase, to provide the starter
population for LTSP, is essential to understanding the dynamics of long-term survival and
evolution as a whole. There is currently debate about whether a random subset of the bacterial
population enters LTSP or if there is a “programmed” mechanism (Finkel, 2006), akin to a form
of bacterial apoptosis or programmed cell death (PCD). Programmed mechanisms have been
proposed (Sutterlin et al., 2016; Yamaguchi & Inouye, 2011), but the identification of molecular
mechanisms for this remain elusive (Finkel, 2006; Navarro-Llorens, 2010).

Understanding how death occurs to bring about LTSP also has implications for other growth-
arrested states, including, but not limited to, the ‘viable but nonculturable’ (VBNC) and persistence
states (Xu et al., 1982; Bergkessel et al., 2016; Navarro Llorens et al., 2010). Most studied in E.
coli and Vibrio cholerae, the VBNC state occurs across a wide variety of genera (Oliver, 2005;
Zhao et al., 2017) and is induced by stresses that also occur in the lead-up to death in the batch
5
culture system, including osmotic shock, pH stress, and starvation (Ramamurthy, Ghosh, &
Gururaja, 2014). Bacterial persistence, though apparently a stochastic and reversible mechanism,
is primarily controlled by stress signaling response pathways (i.e. the SOS response) (Harms et
al., 2016). As with pre-death phase batch culture populations and VBNC cells, persisters may arise
in a similar manner to those populations experiencing nutrient limitation (Maisonneuve et al.,
2013; Harms et al., 2016; Fisher et al., 2017). It is possible that there may be shared features
between these three non-growth conditions (Bergkessel et al., 2016).

In this Minireview, we will describe what is known about the transition at the population-level cell
death in E. coli and other bacteria and juxtapose that knowledge with potential laboratory,
biotechnology, and industrial applications. A better understanding of bacterial stress response will
provide further insight into research interest areas such as antibiotic resistance, evolution, and
bacterial survival under extreme conditions and nutrient limitation.

THE STRESSES OF STATIONARY PHASE
Upon entry into stationary phase, cells no longer divide but remain metabolically active (Navarro-
Llorens, 2010) (Fig. 1.1). Also, as bacterial populations transition out of exponential growth and
into stationary phase, cells shift from a state of frequent cell divisions to the multitude of stress
involved in the maintenance of a high-density population size, including nutrient, osmotic, and pH
stress (Finkel, 2006; Navarro-Llorens, 2010). In particular, upon entry into stationary phase, the
excessive nutrient availability corresponding to exponential growth in rich media is absent, leaving
bacteria to compete for resources in an already crowded environment. How bacteria maintain cell
densities while responding to so many environmental stresses remains an active area of study.
6

Figure 1.1 | Brief summary of the changes that occur between stationary phase and LTSP.
Once bacteria are inoculated in fresh rich media, such as Luria-Bertani (LB) broth, cell populations
grow rapidly and enter stationary phase. Populations remain at high density for ~2 days before
entering death phase, where ~99% of cells die. The remaining populations enter long-term
stationary phase (LTSP). During stationary phase, cells are morphologically and metabolically
distinct from when they are in stationary phase. Figure adapted from Finkel, 2006.

Characteristics of stationary phase. E. coli populations in stationary phase have a very different
morphology and experience a multitude of stresses compared to populations in exponential phase.
Morphologically, stationary phase cells appear smaller and more spherical than logarithmic phase
cells (Lange & Hengge-Aronis, 1991). Expression of transcriptional regulator BolA, regulated by
RpoS, helps control this phenotype and is essential for ‘normal’ coccoid stationary phase cell
morphology (Santos et al., 2002). Other common changes upon entry into stationary phase include
condensation of the nucleoid, formation of 100S ribosome dimers, and alterations in expression of
Log10 CFU/ml
Day
1 2 3 4
6
7
8
9
10
Stationary Phase
-Small, spherical cells
-Increased RpoS expression
-Dps DNA compaction
-100S ribosome dimers
-Increased resistance to stresses
Death Phase
-~99% of population dies
-Stochastic onset?
-PCD cause?
-Delay via buffering or alcohol
LTSP
-GASP
7
gene regulators like Lrp and oxyR Navarro Llorens et al., 2010; Pletnev et al., 2015; other
citations). Further, many of the proteins that control these changes, like with bolA, are controlled
by alternative sigma factors.

Stationary phase stress responses and adaptations
Bacteria possess a plethora of methods to combat the multitude of stresses associated with
stationary phase. In E. coli, the primary global gene regulator for stationary phase stress is rpoS,
encoding the sigma factor σ
S
/σ
38
, also known as RpoS (Lange & Hengge-Aronis, 1991), an
alternative sigma factor that is upregulated upon entry into stationary phase and directly or
indirectly regulates expression of ~23% of E. coli’s genome (Wong et al., 2017). This involves
RpoS in almost every aspect of the stationary phase stress response. RpoS also plays a role in
biofilm formation (Adams & McLean, 1999; Corona-Izquierdo & Membrillo-Hernández, 2002;
Collet et al., 2008) and many GASP mutations are associated with rpoS in LTSP cultures
(Bohannon et al., 1991; Zambrano et al., 1993; Zinser & Kolter, 1999; Finkel, 2006). Stationary
phase sigma factors are found in many other genera of bacteria, both Gram-positive and Gram-
negative, including Bacillus subtilis, Pseudomonas aeruginosa, and Streptomyces coelicolor
(Jaishankar & Srivastava, 2017) (Table 1.1).
Table 1.1 | Brief list of bacteria with alternative sigma factors (Adapted from Jaishankar &
Srivastava, 2017)
Name of organism G+/G- Sigma Factors Reference
Escherichia coli G- 7; ECF sigma factors Navarro Llorens et al., 2010
Bacillus subtilis G+ 18 Haldenwang, 1995; Gruber &
Gross, 2003
Streptomyces
coelicolor
G+ 65 Kim et al., 2008; Tripathi et al.,
2014
Pseudomonas
aeruginosa
G- 24 Potvin et al., 2008
Corynebacterium
glutamicum
G+ 7

Pátek & Nešvera, 2013
8

While RpoS is the primary global stress response and stationary phase sigma factor, other
alternative sigma factors function as part of various stress responses. For example, RpoE, or σ
E

(σ
24
), is associated with envelope stress (Hayden & Ades, 2008). Osmotic and envelope stress are
common during stationary phase (Jenkins et al., 1990), making it necessary for cells to adapt more
rigid and resistant cell envelopes and cause the periplasm to accumulate osmoprotectants such as
trehalose (Huisman et al., 1996). Both RpoE and RpoS play active roles in cell envelope protection
during stationary phase (Santos et al., 1999; Charoenwong et al., 2011; Kabir et al., 2005). Two
other alternative sigma factors in E. coli that can be associated with stationary phase are RpoH (σ
H

or σ
32
) and RpoN (σ
N
or σ
54
) (Jaishankar & Srivastava, 2017). RpoH works primarily to combat
heat shock stress (Wagner et al., 2008), while RpoN is associated with nitrogen stress (Hunt &
Magasanik, 1985). Notably, ~60% of the RpoN

regulon is also under RpoS regulation (Dong et
al., 2011).

In addition to stationary phase-specific changes in the cell envelope and gene expression via
alternative sigma factors, E. coli experience an alternative nucleoid condensation and protein
composition modulated by Dps (Antipov et al., 2017). The dps gene (which stands for DNA-
binding protein from starved cells) is the most highly expressed gene in stationary phase cells
(Almiron et al., 1992) and forms dodecameric structures which bind DNA, condensing it into a
highly ordered structure called the biocrystal (Grant et al., 1998; Wolf et al., 1999). Dps is
essential for cells to survive various stresses such as oxidative and UV stress (Martinez & Kolter,
1997; Nair & Finkel, 2004; Gundlach & Winter, 2014).

9
LABORATORY STUDY OF PHASES AND EVOLUTION
There are three common experimental platforms that researchers use to study long-term cell
physiology and evolution in the laboratory: chemostats, serial passages, and batch culture. The
chemostat method uses a constant volume to grow cell populations (Gresham & Hong, 2015).
Cultures receive a constant influx of nutrients, but in return, artificially lose some of the population
to account for the volume difference. One drawback to this method is bottlenecking in which part
of the viable culture is constantly lost, causing populations also lose diversity. With serial
passaging techniques, a small fraction of the community is transferred into fresh media, usually
daily. As with a chemostat, this method provides additional selective pressure given that there is
constant bottlenecking in the population. Natural bacterial death does not occur in chemostats due
to the constant influx of new media and purging of part of the population. Meanwhile, death phase
can occur in serial passage systems depending on when the culture fraction is transferred to new
media. The majority of serial passage experiments, however, are passaged after 24 hours, prior to
most bacterial death phases (Barrick & Lenski, 2013).

Batch culture systems differ from chemostats and serial passaging because unlike the two systems
described above, populations do not artificially lose biodiversity at any point in their evolution
(Finkel, 2006). Cultures are inoculated at low concentrations and allowed to undergo all five
phases: lag, exponential, stationary, death, and long-term stationary phase (LTSP); no additional
nutrients are added at any point. By allowing populations to undergo all phases, the batch culture
system arguably provides a robust method for studying whether death phase is stochastic or
“programmed.” Another benefit of this laboratory system is that in many ways, it likely
10
approximates the natural environments in which there is often a ‘feast or famine’ cycle (Finkel,
2006).

Another method to study non-exponentially growing populations is the retentostat. Retentostats
were developed as a modification of a chemostat to account for the dilemma imposed by the loss
of biomass in chemostats (Chesbro et al., 1979; Overkamp et al., 2015). While functionally very
similar to a chemostat, in retentostats, fresh medium is constantly fed into the system and a filter
prevents any biomass from leaving via the effluent stream (Ercan et al., 2015). This allows the
microbes to metabolize potential detrital nutrients supplied by dying cells.

BACTERIAL DEATH: THE DEBATE
Currently, there is no clear consensus regarding the mechanism, whether stochastic or
programmed, through which death occurs in E. coli. Aside from several specialized examples
(discussed below), no single programmed cell death (PCD) mechanism seems ubiquitous
throughout E. coli (Finkel 2006; Ramisetty et al., 2016). Other Gram-negative and Gram-positive
bacteria do have PCD systems (Bayles, 2014). Holin-like CidA and antiholin-like LrgA serve as
an example of a toxin-antitoxin (TA) system in which CidA, which can permeabilize the cell wall,
is inhibited by LrgA (Bayles, 2007). If excessive damage is done to the cells, LrgA no longer
inhibits CidA, and the population enters PCD (Bayles, 2007). E. coli differs from other bacteria in
that while there are defined examples of PCD mechanisms, such as with the mazEF toxin-antitoxin
system (Yamaguchi & Inouye, 2011), there is no ubiquitous PCD system across all strains E. coli.
Further, unpublished data from the Finkel laboratory indicate that deletion of up to 10 toxin-
antitoxin systems in a single strain of E. coli result in no discernable effect on the timing of entry
11
into death phase or the corresponding extent of loss of viability, further indicating that, at the very
least, TA systems are not the sole factor in the regulation of PCD mechanisms across bacterial
species. It should also be noted that if there is such a genetic basis for bacterial death, a death-
resistant mutant strain would likely have been identified by this point. Mechanistically, this is a
complex event that cannot be explain by disruption of a single gene.

Even with growing interest in the topic of bacterial PCD over the last few years (Bayles 2014),
studying the mechanism of bacterial death in a population remains practically difficult. A number
of groups have tried to circumvent this difficulty by using synthetic biological or computer
simulation modeling approaches (Avrani et al., 2017; Tanouchi et al., 2012). Usually, the exact
timing of death depends on a multitude of factors, including the bacterial strain and the medium in
which the strain is grown in (Finkel, 2006; Kram & Finkel, 2014; Kram & Finkel, 2015), so it can
be difficult to capture populations as they are dying. Because of such complexities, the question
remains extremely understudied.

The lead-up to death phase and potential stochastic causes
By the end of stationary phase, bacteria have encountered a multitude of stresses, including pH
and nutrient stress. The presence of such stresses activates multiple global stress response
pathways, including those involving alternative sigma factors RpoS and RpoE (Hengge-Aronis,
2002; Hengge 2009).

What mechanisms might allow bacteria to enter death phase in a seemingly random manner? The
first and simplest explanation is that cell death is truly random. In this model, a particular
12
environment can only support a given population size for a given amount of time. When the
population is no longer supported and all cells can no longer perform maintenance and repair
functions, such as repair of oxidative to proteins, lipids, or nucleic acids that occur as a result of
normal metabolism (Nystrom, 2003), the vast majority of the population dies. It has been
postulated that the ability to ‘sense’ the high density of cells and lack of nutrients, leading to the
eventual onset of death phase, is mediated by quorum sensing mechanisms (Finkel, 2006). Once
the bulk of cells begin dying and lose cellular integrity, their still-living siblings are able to utilize
the detritus of the dead cells. From these deceased siblings, the surviving population acquires
nucleic acids, carbohydrates, amino acids, and lipids (Finkel & Kolter, 2001), allowing the group
to sense the change and exit of death phase to begin replicating and surviving again. This model
doesn’t, however, explain why all of the cells don’t die and a small subset of the population is able
to exit of death phase and enter long-term stationary phase (LTSP).

Programmed cell death and toxin-antitoxin gene pairs
Over the past two decades, there has been increased interest in the topic of bacterial ‘apoptosis,’
or ‘programmed cell death’ (PCD) in bacteria (Navarro-Llorens, 2010; Bayles, 2014). The
definition of programmed cell death in this context is a genetically-programmed process(es) that
leads to bacterial cell suicide, often contextualized at the population-level. Examples of PCD in
bacteria come in many forms, from mother cell lysis in Bacillus subtilis during sporulation (Lewis,
2000) to the cell death associated with structures formed in fruiting bodies of Myxococcus xanthus
(Shimkets, 1999). Further, PCD is suggested to occur in response to environmental stresses such
as starvation (Hazan et al., 2004).

13
TA systems are among the best studied PCD mechanisms. First identified in the 1980s (Gerdes et
al., 1986; Hiraga et al., 1986; Bravo et al., 1987), TA pairs are common across all genera of
bacteria and archaea (Yamaguchi et al., 2011; Makarova et al., 2013) and are comprised of a toxin
gene and a corresponding antitoxin gene (Unterholzner et al., 2013). When there is too much
‘toxin’ in a cell with no ‘antitoxin’ to counteract it, the cell in question will die. Lehnherr and
Yarmolinsky were among the first to describe TA pairs when they characterized ‘addiction
molecules’ phd (antitoxin) and doc (toxin), isolated from prophage P1 (Lehnherr & Yarmolinsky,
1995), though there are many TA pairs in Gram positive as well as Gram negative bacteria
(Claverys & Havarstein, 2007).

The exact role of TA systems and PCD as a whole across bacterial systems remains open given
the difficulty in replicating a number of the experiments across different strains (Ramisetty et al.,
2016). It is possible that TA modules play a key role in generating a persister state, a phenomenon
actively studied for the part it plays in antibiotic resistance (Page & Peti, 2016). In regard to PCD,
however, it is unlikely that TA systems are the common mode of programmed death across species
(Ramisetty et al., 2016). As all bacteria enter death phase at some point, identification of a common
mechanism, or at least similar mechanisms, would be equivalent to ‘the holy grail’ in this field of
study.

Modes of delaying death
One potential avenue for investigating mechanisms of bacterial death is to study ways in which
death is delayed (Fig. 1.2). There have been multiple ways in which groups have explored this
phenomenon. Farrell and Finkel have shown that buffering the culture media at neutral pH delays
14
death phase (Farrell & Finkel, 2003). By maintaining the pH at 7.0 in Luria-Bertani (LB) broth,
the authors observed a delay of entry into death phase, maintaining high cell densities, in E. coli
for multiple days (Farrell & Finkel, 2003). The addition of acid directly to basic cultures prolongs
the acidity of the cultures, causing the cells to remain in stationary phase.

Figure 1.2 | Schematic of physiological portrayal of delay of bacterial death phase. One of the
methods that researchers use to study the mechanism of death phase is to see how bacterial death
is delayed. Both buffering and sublethal low dose alcohol additions to rich media, like LB broth,
have been shown to delay death phase in Escherichia coli (Farrell & Finkel, 2003; Ferraro &
Finkel, 2019). The actual length of death delay is dose dependent and can be shorter or longer than
represented here. Colony forming units (CFU)/ml of wild-type untreated populations are
represented as a solid line while the buffered populations are represented with a dashed line. Figure
adapted from Finkel, 2006.

The delay of bacterial death through a phenomenon called the “alcohol effect” (Chapter 3; Chapter
4; Chapter 5; Vulic & Kolter, 2002; Ferraro & Finkel, 2019). When sublethal low doses of straight-
chain alcohols between 2 and 6 carbons in length are added to bacterial cultures, populations
exhibit a dose-dependent death delay effect. It was recently found in E. coli that the
Log10 CFU/ml
Day
1 2 3 4
6
7
8
9
10
15
glycolate/glyoxylate degradation pathway may contribute to the alcohol effect (Chapter 3; Ferraro
& Finkel, 2019). Particularly, given the structural similarity between alcohol and glycolate, this
model suggests that alcohol may be allosterically mimicking glycolate to bind GlcC to derepress
the glycolate degradation pathway and cause an increase in gluconeogenesis, which would then
allow for energy for increased scavenging of detrital nutrients and energy (Chapter 3). The alcohol-
induced delay of death effect is present in environmental isolate E. coli strains (Chapter 3; Chapter
4) as well as in non-E. coli genera (Chapter 5), including Pseudomonas and Klebsiella (Ferraro &
Finkel, 2019).

CONCLUSIONS AND FUTURE DIRECTIONS
Progress is made all the time that betters our understanding of complex biological systems and the
interplay between each cell in those populations. Though ever changing, both stationary phase and
LTSP are becoming better understood (Finkel, 2006; Navarro-Llorens, 2010; cite Nicole’s paper;
more?). In spite of all this, death phase remains under-studied (Finkel, 2006; Navarro-Llorens,
2010), understandably so given the inherent difficulty of studying population-level cell death.
Assumption remains that a “programmed” sort of bacterial death is more of the exception than the
norm. Characterization of how entire populations enter wide-scale death where ~99% of a
population dies (Finkel, 2006; Kram & Finkel, 2014; Kram & Finkel, 2015) remains elusive
(Finkel, 2006; Navarro-Llorens, 2010). Possibly, this is a stochastic affair. Another possibility is
that quorum sensing in the population plays a role in the distribution of a genetic signal not just to
die, but to ensure that at least ~1% of the remaining population survives to enter LTSP.

16
Extensive characterization of bacterial death and survival also has large scale implications for
medicine and technology. Given that the natural environments of the majority of bacterial
populations is more of a “feast or famine” model (Finkel, 2006) of survival, this signifies a large
amount of stress (and stress resistance), large-scale nutrient deprivation and intense competition
for the remainder, and relying on genetic diversity, often through SOS DNA polymerases (Chapter
4; Corzett et al., 2013), for a better chance at survival. An understanding of causes and
mechanisms, if any, of bacterial death phase has the potential to allow researchers a better overall
understanding of long-term population survival. In particular, the work done in this Thesis explores
how genetic diversity is affected by medium environment (Chapter 2) and how addition of small,
sublethal doses of alcohols causes a delay in the onset of bacterial death phase, thus possibly
allowing researchers a method to studying the inherently difficult topic of death phase (Chapter 3;
Chapter 4; Chapter 5; Appendix A).
17
1.3 References
Adams JL, McLean RJ. 1999. Impact of rpoS deletion on Escherichia coli biofilms. Appl Environ
Microbiol 65:4285-4287.
Adnan M, Sousa AM, Machado I, Pereira MO, Khan S, Morton G, Hadi S. 2017. Role of bolA and
rpoS genes in biofilm formation and adherence pattern by Escherichia coli K-12 MG1655
on polypropylene, stainless steel, and silicone surfaces. Acta Microbiol Immunol Hung
64:179-189.
Aizenman E, Engelberg-Kulka H, Glaser G. 1996. An Escherichia coli chromosomal “addiction
module” regulated by 3’,5’-bispyrophosphate: a model for programmed bacterial cell
death. Proc Natl Acad Sci 93:6059-6063.
Almiron M, Link AJ, Furlong D, Kolter R. 1992. A novel DNA-binding protein with regulatory
and protective roles in starved Escherichia coli. Genes Dev 6:2646-2654.
Antipov SS, Tutukina MN, Preobrazhenskaya EV, Kondashov FA, Patrushev MV, Toshchakov
SV, Dominova I, Shvyreva US, Vrublevskaya VV, Morenkov OS, Sukharicheva NA,
Panyukov VV, Ozoline ON. 2017. The nucleoid protein binds genomic DNA of
Escherichia coli in a non-random manner. PLoS One 12:e0182800.
Avrani S, Bolotin E, Katz S, Hershberg R. 2017. Rapid genetic adaptation during the first
four months of survival under resource exhaustion. Mol Biol Evol 34:1758-1769.
Barrick JE, Lenski RE. 2013. Genome dynamics during experimental evolution. Nat Rev Genet
14:827-839.
Bayles KW. 2007. The biological role of death and lysis in biofilm development. Nat Rev
Microbiol 5:721-726.
18
Bayles KW. 2014. Bacterial programmed cell death: making sense of a paradox. Nat Rev
Microbiol 12:63-69.
Belitsky M, Avshalom H, Erental A, Yelin I, Kumar S, London N, Sperber M, Schueler-Furman
O, Engelberg-Kulka H. 2011. The Escherichia coli extracellular death factor EDF induces
the endoribonucleolytic activities of the toxins MazF and ChpBK. Cell 41:625-635.
Bergkessel M, Basta DW, Newman DK. 2016. The physiology of growth arrest: uniting molecular
and environmental microbiology. Nat Rev Microbiol 14:549-562.
Bohannon DE, Connell N, Keener J, Tormo A, Espinosa-Urgel M, Zambrano MM, Kolter
R. 1991. Stationary-phase-inducible “gearbox” promoters: differential effects of katF
mutations and role of s
70
. J Bacteriol 173:4482-4492.
Bravo A, de Torrontegui G, Diaz R. 1987. Identification of components of a new stability
system of plasmid R1, ParD, that is close to the origin of replication of this plasmid. Mol
Gen Genet 210:101-110.
Cao H, Wei D, Yang Y, Shang Y, Li G, Zhou Y, Ma Q, Xu Y. 2017. Systems-level
understanding of ethanol-induced stresses and adaptation in E. coli. Sci Rep 7:44150.
Charoenwong D, Andrews S, Mackey B. 2011. Role of rpoS in the Development of Cell Envelope
Resilience and Pressure Resistance in Stationary-Phase Escherichia coli. Appl Environ
Microbiol 77:5220-5229.
Chaudhari SS, Thomas VC, Sadykov MR, Bose JL, Ahn DJ, Zimmerman MC, Bayles KW. 2016.
The LysR-type transcriptional regulator, CidR, regulates stationary phase cell death in
Staphylococcus aureus. Mol Microbiol 101:942-953.
Chesbro W, Evans T, Eifert R. 1979. Very slow growth of Escherichia coli. J Bacteriol 139:625-
638.
19
Claverys JP, Havarstein LS. 2007. Cannibalism and fratricide: mechanisms and riasons d’etre. Nat
Rev Microbiol 5:219-229.
Collet A, Cosette P, Beloin C, Ghigo JM, Rihouey C, Lerouge P, Junter GA, Jouenne T. 2008.
Impact of rpoS deletion on the proteome of Escherichia coli grown planktonically and as
biofilm. J Proteome Res 7:4659-4669.
Corona-Izquierdo FP, Membrillo-Hernández J. 2002. A mutation in rpoS enhances biofilm
formation in Escherichia coli during exponential phase of growth. FEMS Microbiol Lett
211:105-110.
Corzett CH, Goodman MF, Finkel SE. 2013. Competitive fitness during feast and famine: how
SOS DNA polymerases influence physiology and evolution in Escherichia coli. Genetics
194:409-420.
Dong T, Yu R, Schellhorn H. 2011. Antagonistic regulation of motility and transcriptome
expression by RpoN and RpoS in Escherichia coli. Mol Microbiol 79:375-386.
Ercan O, Bisschops MM, Overkamp W, Jorgensen TR, Ram AF, Smid EJ, Pronk JT, Kuipers OP,
Daran-Lapujade P, Kleerebezem M. 2015. Physiological and transcriptional responses of
different industrial microbes at near-zero specific growth rates. Appl Environ Microbiol
81:5662-5670.
Farrell MJ, Finkel SE. 2003. The growth advantage in stationary-phase phenotype conferred by
rpoS mutations is dependent on the pH and nutrient environment. J Bacteriol 185:7044-
7052.
Farrell MJ, Finkel SE. 2003. The growth advantage in stationary-phase phenotype conferred by
rpoS mutations is dependent on the pH and nutrient environment. J Bacteriol 185:7044-
7052.
20
Ferraro CM, Finkel SE. 2019. Physiological, Genetic, and Transcriptomic Analysis of Alcohol-
Induced Delay of Escherichia coli Death. Appl Environ Microbiol 85:e02113-18.
Finkel SE, Kolter R. 2001. DNA as a nutrient: novel role for bacterial competence gene homologs.
J Bacteriol 183:6288-6293.
Finkel SE. 2006. Long-term survival during stationary phase: evolution and the GASP
phenotype. Nat Rev Microbiol 4:113-120.
Gerdes K, Rasmussen PB, Molin S. 1986. Unique type of plasmid maintenance function:
postsegregational killing of plasmid-free cells. Proc Natl Acad Sci 83:3116-3120.
Grant RA, Filman DJ, Finkel SE, Kolter R, Hogle JM. 1998. The crystal structure of Dps,
a ferritin homolog that binds and protects DNA. Nat Struct Biol 5:294-303.
Gresham D, Hong J. 2015. The functional basis of adaptive evolution in chemostats. FEMS
Microbiol Rev 39:2-16.
Grimm NB, Gergel SE, McDowell WH, Boyer EW, Dent CL, Groffman P, Hart SC, Harvey J,
Johnston C, Mayorga E, McClain ME, Pinay G. 2003. Merging aquatic and terrestrial
perspectives of nutrient biogeochemistry. Oecologia 137: 485-501.
Gruber TM, Gross CA. 2003. Multiple sigma subunits and the partitioning of bacterial
transcription space. Annu Rev Microbiol 57:441-466.
Gundlach J, Winter J. 2014. Evolution of Escherichia coli for maximum HOCl resistance through
constitutive expression of the OxyR regulon. Microbiol 160:1690-1704.
Haldenwang WG. 1995. The Sigma Factors of Bacillus subtilis. Microbiol Rev 59:1-30.
Harms A, Maisonneuve E, Gerdes K. 2016. Mechanisms of bacterial persistence during stress and
antibiotic exposure. Science 16:354.
21
Hartman WH, Richardson CJ. 2013. Differential nutrient limitation of soil microbial biomass and
metabolic quotients (qCO2): is there a biological stoichiometry of soil microbes? PLoS
One 8:e57127.
Hayden JD, Ades SE. 2008. The extracytoplasmic stress factor, s
E
, is required to maintain cell
envelope integreity in Escherichia coli. PLoS ONE 3:e1573.
Hayden JD, Ades SE. 2008. The extracytoplasmic stress factor, s
E
, is required to maintain cell
envelope integreity in Escherichia coli. PLoS ONE 3:e1573.
Hazan R, Sat B, Engelberg-Kulka H. 2004. Escherichia coli mazEF-mediated cell death is
triggered by various conditions. J Bacteriol 186:3663-3669.
Hengge R. 2009. Proteolysis of s
S
(RpoS) and the general stress response in Escherichia coli. Res
Microbiol 160:667-676.
Hengge R. 2011. Stationary-phase gene regulation in Escherichia coli. EcoSal Plus 4(2).
Hengge-Aronis R. 2002. Signal transduction and regulatory mechanisms involved in control of the
s
S
(RpoS) subunit of RNA polymerase. Microbiol Mol Biol Rev 66:373–395.
Hiraga S, Jaffe A, Ogura T, Mori H, Takahashi H. 1986. F plasmid ccd mechanism in
Escherichia coli. J Bacteriol 166:100-104.
Hunt TP, Magasanik B. 1985. Transcription of glnA by purified Escherichia coli components: core
RNA polymerase and the products of glnF, glnG, and glnL. Proc Natl Acad Sci 82:8453-
8457.
Jaishankar J, Srivastava P. 2017. Molecular basis of stationary phase survival and applications.
Front Microbiol doi:10.3389/fmicb.2017.02000.
Jenkins DE, Chaisson SA, Matin A. 1990. Starvation-Induced Cross Protection against Osmotic
Challenge in Escherichia coli. J Bacteriol 172:2779-2781.
22
Jorgensen BB, Marshall IPG. 2015. Slow microbial life in the seabed. Annu Rev Mar Sci 8:311-
332.
Kabir MS, Yamashita D, Koyama S, Oshima T, Kurokawa K, Maeda M, Tsunedomi R, Murata
M, Wada C, Mori H, Yamada M. 2005. Cell lysis directed by s
E
in early staitonary phase
and effect of induction of the rpoE gene on global gene expression in Escherichia coli.
Microbiology 151:2721-2735.
Kim ES, Song JY, Kim DW, Chater KF, Lee KJ. 2008. A possible extended family of regulators
of sigma factor activity in Streptomyces coelicolor. J Bacteriol 190:7559-7566.
Koch AL. 1971. The adaptive response of Escherichia coli to feast and famine existence. Adv
Microb Physiol 6:147-217.
Kolodkin-Gal I, Hazan R, Gaathon A, Carmeli S, Engelberg-Kulka H. 2007. A linear pentapeptide
is a quorum-sensing factor required for mazEF-mediated cell death in Escherichia coli.
Science 318:652-655.
Kolodkin-Gal I, Hazan R, Gaathon A, Carmeli S, Engelberg-Kulka H. 2007. A linear pentapeptide
is a quorum-sensing factor required for mazEF-mediated cell death in Escherichia coli.
Science 318:652-655.
Kram KE, Finkel SE. 2014. Culture volume and vessel affect long-term survival, mutation
frequency, and oxidative stress of Escherichia coli. Appl Environ Microbiol 80:1732-1738.
Kram KE, Finkel SE. 2015. Rich medium composition affects Escherichia coli survival,
glycation, and mutation frequency during long-term batch culture. Appl Environ
Microbiol 81:4442-4450.
Lange R, Hengge-Aronis R. 1991. Identification of a central regulator of stationary-phase gene
expression in Escherichia coli. Mol Microbiol 5:49-59.
23
Lehnherr H, Yarmolinsky MB. 1995. Addiction protein Phd of plasmid prophage P1 is a substrate
of the ClpXP serine protease of Escherichia coli. Proc Natl Acad Sci 92:3274-3277.
Lever MA, Rogers KL, Lloyd KG, Overmann J, Schink B, Thauer RK, Hoehler TM, Jorgensen
BB. 2015. Life under extreme energy limitation: a synthesis of laboratory – and field-based
investigations. FEMS Microbiol Rev 5:688-728.
Lewis K. 2000. Programmed death in bacteria. Microbiol Mol Biol Rev 64:503-514.
Makarova KS, Wolf YI, Koonin EV. 2013. Comparative genomics of defense systems in archaea
and bacteria. Nucleic Acids Res 41:4360-4377.
Martinez A, Kolter R. 1997. Protection of DNA during oxidative stress by the nonspecific DNA-
binding protein Dps. J Bacteriol 179:5188-5194.
Munoz-Gomez AJ, Santos-Sierra S, Berzal-Herranz A, Lemonnier M, Diaz-Orejas R. 2004.
Insights into the specificity of RNA cleavage by the Escherichia coli MazF toxin. FEBS
Lett 567:316-320.
Nair S, Finkel SE. 2004. Dps Protects Cells against Multiple Stresses during Stationary Phase. J
Bacteriol 186:4192-4198.
Navarro Llorens JM, Tormo A, Martínez-García E. 2010. Stationary phase in Gram-
negative bacteria. FEMS Microbiol Rev 34:476-495.
Nystrom T. 2003. Nonculturable bacteria: programmed survival forms or cells at death’s door?
Bioessays 25:204-211.
Overkamp W, Ercan O, Herber M, van Maris AJ, Kleerebezem M, Kuipers OP. 2015.
Physiological and cell morphology adaptation of Bacillus subtilis at near-zero specific
growth rates: a transcriptome analysis. Environ Microbiol 17:346-363.
24
Page R, Peti W. 2016. Toxin-antitoxin systems in bacterial growth arrest and persistence. Nat
Chem Biol 12:208-214.
Pátek M, Nešvera J. 2013. “Promoters and plasmid vectors of Corynebacterium glutamicum” in
Corynebacterium glutamicum. Microbiology Mongraphs, Vol 23, eds H. Yukawa and M.
Inui (Berlin: Springer), 51-88.
Pereira FC, Berry D. 2017. Microbial nutrient niches in the gut. Environ Microbiol 19: 1366-1378.
Pletnev P, Osterman I, Sergiev P, Bogdanov A, Dontsova O. 2015. Survival guide: Escherichia
coli in the stationary phase. Acta Naturae 7:22-33.
Potvin E, Sanschagrin F, Levesque RC. 2008. Sigma factors in Psuedomonas aeruginosa. FEMS
Microbiol Rev 32:38-55.
Ramamurthy T, Ghosh A, Pazhani GP, Sinoda S. 2014. Current perspectives on viable but non-
culturable (VBNC) pathogenic bacteria. Front Public Health 2:103.
Ramisetty BC, Natarajan B, Santhosh RS. 2015. mazEF-mediated programmed cell death in
bacteria: “What is this?”. Crit Rev Microbiol 41:89-100.
Ramisetty BC, Raj S, Ghosh D. 2016. Escherichia coli MazEF toxin-antitoxin system does not
mediate programmed cell death. J Basic Microbiol 56:1398-1402.
Ranjit DK, Endres JL, Bayles KW. 2011. Staphyloccus aureus CidA and LrgA proteins exhibit
holing-like properties. J Bacteriol 193:2468-2476.
Romantsov T, Guan Z, Wood JM. 2009. Cariolipin and the osmotic stress responses of bacteria.
Biochim Biophys Acta 1788:2092-2100.
Santos JM, Lobo M, Matos APA, de Pedro MA, Arraiano CM. 2002. The gene bola
regulates dacA (PBP5), dacC (PBP6) and ampC (AmpC), promoting normal morphology
in Escherichia coli. Mol Microbiol 45:1729-1740.
25
Shimkets LJ. 1999. Intercellular signaling during fruiting-body development of Myxococcus
xanthus. Annu Rev Microbiol 53:525-549.
Stewart EJ, Madden R, Paul G, Taddei F. 2005. Aging and death in an organism that
reproduces by morphologically symetric division. PLoS Biol 3:e45.
Sutterlin HA, Shi H, May KL, Miguel A, Khare S, Huang KC, Silhavy TJ. 2016. Disruption of
lipid homeostasis in the Gram-negative cell envelope activates a novel cell death pathway.
Proc Natl Acad Sci 113:e1565-1574.
Tanouchi Y, Pai A, Buchler NE, You L. 2012. Programmed stress-induced altruistic death in
engineered bacteria. Mol Syst Biol 8:626.
Tripathi I, Zhang Y, Lin Z. 2014. Bacterial sigma factors as targets for engineered or synthetic
transcriptional control. Front Bioeng Biotechnol 2:33.
Unterholzner SJ, Poppenberger B, Rozhon W. 2013. Toxin-antitoxin systems: biology,
identification, and application. Mob Genet Elements 3:e26219.
Wagner MA, Zahrl D, Rieser G, Koraimann G. 2008. Growth phase- and cell division-dependent
activation and inactivation of the σ
32
regulon in Escherichia coli. J Bacteriol 191:1695-
1702.
Walsh EA, Kirkpatrick JB, Rutherford SD, Smith DC, Sogin M, D’Hondt S. 2016. Bacterial
diversity and community composition from seasurface to subseafloor. ISME J 10:979-989.
Weber H, Polen T, Heuveling J, Wendisch VF, Hengge R. 2005. Genome-wide analysis of the
general stress response network in Eschericha coli: σ
S
-dependent genes, promoters, and
sigma factor selectivity. J Bacteriol 187:1591-1603.
Wolf SG, Frenkiel D, Arad T, Finkel SE, Kolter R, Minsky A. 1999. DNA protection by stress-
induced biocrystallization. Nature 400:83-85.
26
Wong GT, Bonocora RP, Schep AN, Beeler SM, Lee Fong AJ, Shull LM, Batachari LE, Dillon
M, Evans C, Becker CJ, Bush EC, Hardin J, Wade JT, Stoebel DM. 2017. Genome-wide
transcriptional response to varying RpoS levels in Escherichia coli K-12. J Bacteriol 199:1-
17.
Yamaguchi Y, Park JH, Inouye M. 2011. Toxin-antitoxin systems in bacteria and archaea. Annu
Rev Genet 45:61-79.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zhang Y, Zhang J, Hoeflich KP, Ikura M, Qing G, Inouye M. 2003. MazF cleaves cellular
mRNAs specifically at ACA to block protein synthesis in Escherichia coli. Mol Cell
12:913-923.

27
Chapter 2: Environment-Specific Signatures of Alternative DNA Polymerase Activity and
Mutational Spectra in Escherichia coli

The content of this chapter includes draft text of a manuscript in preparation: Corzett CH, Kram
KE Ferraro CM, Goodman, MF, Finkel SE., “Environment-specific signatures of mutational
spectra and alternative DNA polymerase expression in Escherichia coli.”

2.1 Abstract
Varying environmental conditions can cause changes in physiological responses within microbial
populations. The long-term stationary phase (LTSP) of cell growth is a particularly stressful time
caused by factors including a depletion of nutrients and buildup of reactive oxidative species and
toxins, and we know that alternative DNA polymerases (polB, dinB, and umuDC) are induced in
during phase. To determine whether differences in certain environments may affect polymerase
expression levels, and potentially mutation spectra, we monitored the survival of Escherichia coli
grown in a variety of laboratory media [LB, 2xYT, Super Broth (SB), and Terrific Broth (TB)]
while determining the patterns of polymerase gene expression and the resulting mutation
frequencies and mutation spectra. We found that the genes encoding all three alternative DNA
polymerases were induced within 8 hours in E. coli grown in each of the four media tested. In fact,
alternative polymerase transcripts were more abundant than those of high fidelity PolIII (dnaE)
within 24 hours in all four media. Mutation frequencies and mutation spectra of spontaneous
mutations were measured for the gene rpoB, which confer resistance to the antibiotic rifampicin
in all four environments. E. coli grown in LB and YT have a higher mutation frequency than more
nutrient-rich SB and TB. Additionally, cells grown in LB and YT have a higher number of
transversions whereas cells grown in SB and TB have a higher number of transitions. We found
correlations between mutation spectra in the different environments and the alternative polymerase
28
expression patterns. Together, these findings indicate that alternative DNA polymerase expression
induced during LTSP may mediate changes in mutation frequency and spectrum. Further, this data
indicates that these activities are medium-specific.

29
2.2 Introduction
Escherichia coli has long been the laboratory “work horse” for studying fundamental biological
and metabolic processes. For the most part, E. coli is incubated for short amounts of time though
the actual life cycle of the bacterium can span much longer without more nutrients being added to
the growth medium (Finkel, 2006). The first three phases of the bacterial life cycle in batch culture
systems, lag phase, exponential phase, and stationary phases, are fairly well defined while the last
two phases, death phase and long-term stationary phase (LTSP) remain somewhat less well-
understood (Finkel, 2006; Navarro Llorens et al., 2010). After stationary phase, cells enter death
phase, where at least ~99% of the cells in the population die, though this largely depends on the
growth medium, strain type, and other conditions (Finkel, 2006; Kram & Finkel, 2014; Kram &
Finkel, 2015). The surviving cells then enter LTSP, though any potential genotypie differences
between the surviving subpopulation versus the dead majority population remain unknown
(Ferraro & Finkel, manuscript in preparation). LTSP is potentially the most dynamic of any of the
phases (Finkel, 2006). Here, the overall population size remains fairly constant, although there are
rapid changes in the external medium environment and often cellular turnover of subpopulations
(Finkel, 2006). Cells that are better suited to a given LTSP culture condition causes selection of
mutations with beneficial alleles (Zambrano et al., 1993; Zinser & Kolter, 2004; Zinser et al.,
2003; Zinser & Kolter, 2000), including the growth advantage in stationary phase (GASP)
phenotype (Finkel, 2006; Finkel & Kolter, 1999; Zambrano et al., 1993; Zinser & Kolter, 2004;
Zinser et al., 2003; Zinser & Kolter, 2000).

Although, as previously mentioned, the exact nature of the transition from stationary phase into
LTSP remains unknown, many physiological markers serve as an indicator of a changing
30
environment. For example, pH, mutation frequency, and even glycation, a covalent modification
on a proteins and nucleic acid that affect activity (Kram & Finkel, 2014; Pepper et al., 2010;
Mironova et al., 2001; Mironova et al., 2005), all change in relation to the life cycle of a bacterial
population.

Previous work in the Finkel lab has demonstrated that when grown in Luria-Bertani (LB) broth
under batch culture conditions, three alternative DNA polymerases are naturally (without
exogenous SOS induction) differentially expressed, leading to differences in mutation frequency
and spectrum in E. coli (Corzett et al., 2013). These three alternative error-prone DNA
polymerases, often referred to as SOS DNA polymerases due to their extensive characterization
following SOS regulon induction (Courcelle et al., 2001; Goodman, 2002), consist of Pol II
(encoded by polB), Pol IV (dinB), and Pol V (umuDC). Pol II is a B-family polymerase
(Rangarajan et al., 1999; Pham et al., 2001; Banach-Orlowska et al., 2005) and is the least error-
prone as it contains 3’-exonuclease proofreading activity (Cai et al., 1995). Pol IV and Pol V are
part of the Y-family of DNA polymerases (Ohmori et al., 2001; Goodman, 2002; Nohmi, 2006)
and do not have exonuclease activity, resulting lower replication fidelity (Tang et al., 2000; Fuchs
et al., 2004; Jarosz et al., 2007). Further, the Y-family DNA polymerases have homologs in
eukaryotes in addition to various prokaryotes (Ohmori et al., 2001) and potentially play a role in
several human diseases (Robbins et al., 1974; Stallons & McGregor, 2010; Lange et al., 2011;
Sale et al., 2012). Corzett et al., 2013 show that Pol II, Pol IV, and Pol V can be expressed without
exogenous induction and that the polymerases contribute towards overall fitness and long-term
survival in batch culture.

31
The work described in this chapter combines the groundwork laid by Corzett et al., 2013 and Kram
& Finkel, 2015, where E. coli was grown in four different rich media conditions, including LB, 2x
yeast extract-tryptone (YT), Terrific Broth (TB), and Super Broth (SB), and characterized for
different genetic and physiological changes expressed by cells (Kram & Finkel, 2015). Here, I
grew wild-type E. coli in each medium in batch culture conditions and quantified the mutation
frequency and mutation spectrum for each sample set using a rifampicin-resistance (Rif
R
) reporter
assay. From there, I correlated differences in mutation spectrum with differential DNA polymerase
gene expression. This work contributes additional evidence that medium composition, something
not often considered in relation to its effect on the bacterium, can affect the genetics and
physiological responses of bacteria. Such results could prove impactful not only in the day-to-day
workings of a laboratory but also in fields such as evolution and biodefense. For example, being
able to identify different mutation spectra caused by certain medium nutrients or the culture vessel
potentially allows researchers to identify conditions in which a bioweapon is grown, simply by
measuring the mutation frequency and spectrum.

2.3 Materials & Methods
Bacterial strains, culture media, and growth conditions. E. coli strains used in this chapter are
listed in Table 2.1, with most experiments performed using E. coli K-12 strain ZK126, derived
from W3110 lineage (Zambrano et al., 1993). Unless otherwise stated, cultures were inoculated
from frozen 20% glycerol stocks into 5 ml of LB (Lennox), 2xYT, TB, or SB medium (Difco)
(Kram & Finkel, 2015) in 18- by 150-mm borosilicate test tubes (Thermo Fisher) and incubated at
37
°
C with aeration using TC-7 rolling drums (New Brunswick Scientific, Edison, NJ). Cells from
32
overnight cultures were then inoculated 1:1000 (vol:vol) into 5 ml of either LB, 2xYT, TB, or SB
rich medium (same medium as original inoculum).

Monitoring cell growth and survival. Viable cell counts were determined by serial dilution at
indicated time points and plating on LB agar (Kraigsley & Finkel, 2009). The limit of detection
for this method of titering is ≥1,000 CFU/ml (Kraigsley & Finkel, 2009).

Mutation frequency and spectrum assay and analysis. Mutation frequency was determined by
measuring spontaneous rifampicin resistance (Rif
R
) of E. coli ZK126 wild-type strain after 24
hours growth (Corzett et al., 2013) in either LB, 2xYT, TB, or SB medium. For each condition,
48 independent 5 ml overnight cultures were grown in their given medium. Total cell counts were
measured on LB agar (described above), and 100 µl of each overnight culture was plated onto
plates containing rifampicin (100 µg/ml). LB rifampicin plates were protected from light and
incubated overnight at 37
°
C. Rif
R
frequency was calculated by dividing the total number of Rif
R

colony forming units (CFU)/ml by the total CFU/ml of cells plated. The mutation frequency
distributions were compared using a two-sample Kolmogorov-Smirnov (K-S) test (p < 0.05)
(http://www.physics.csbsju.edu/stats/KS-test.html).

Two colonies from each LB rifampicin agar plate were picked to be sequenced. Each colony was
stored in a 20% glycerol solution in individual wells in a 96-well plate (Corning). Prior to freezing,
2 µl of the resuspended colony was used as the template for colony PCR of the rpoB gene. Cluster
I and cluster II (Garibyan et al., 2003) were sequenced using the primers described in Corzett et
al., 2013.
33
2.4 Results
Table 2.1 | Strains used in this study (Adapted from Corzett et al., 2013)
Strain Relevant
genotype/phenotype
Nomenclature Pol II Pol IV Pol V Reference
ZK126 W3110 ∆lacU169 tna-2 Wild-type + + + 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+ only - - + Corzett et al.,
2013
SF2014 ZK126 polB::Spc
R

umuDC::Cam
R

Pol IV+ only - + - Corzett et al.,
2013
SF2016 ZK126 dinB::Kan
R

umuDC::Cam
R

Pol II+ only + - - Corzett et al.,
2013
SF2018 ZK126 polB::Spc
R

dinB::Kan
R

umuDC::Cam
R

Triple mutant - - - Corzett et al.,
2013

E. coli shows differential growth patterns in batch culture. From Kram et al., 2015, it was
shown that E. coli M2 (MG1655 lineage) (Lee et al., 2012) populations exhibit different growth
phenotypes depending whether the cells are grown in Luria-Bertani (LB) broth, 2x yeast extract-
tryptone (YT), Terrific Broth (TB), or Super Broth (SB). All media described here are considered
undefined rich media though they vary in chemical composition (Kram & Finkel, 2015). Upon
inoculation from overnight cultures, the main wild-type strain used in the laboratory, ZK126
(Table 2.1), shows distinct growth patterns when grown in each medium (Fig. 2.1) (data obtained
from K. E. Kram in the Finkel lab). Whereas cells grown in standard rich media LB or YT reach
peak densities on Day 1 before slowly declining in viability on Days 2 and 3, entering LTSP by
Day 4, populations grown in super rich media TB and SB remain at about ~10x higher densities
on Days 2 and 3 compared to LB and YT populations before plummeting in viability by Day 4. In
34
all four media, population coutts level out in stationary phase at about 10
8
colony forming units
(CFU)/ml.

Figure 2.1 | Wild-type strain grown in different media show altered growth phenotypes. To
determine the effect of media contents on bacterial growth, our wild-type E. coli strain, ZK126,
was grown in four different common rich media environments. Cultures grown in LB are indicated
in black, YT cultures are shown in blue, TB in red, and SB in green. Error bars represent standard
deviation; n=3.

Populations exhibit variation in mutation frequency and mutation spectrum in response to
different media. It is well-established that alternative, error-prone DNA polymerases introduce
more chromosomal mutations to the system than the “housekeeping” DNA polymerase III
(Goodman, 2002), and we also know from previous work in the lab that gene expression of the
alternative polymerases is naturally induced (without exogenous chemical inducers) in LB (Corzett
et al., 2013) and differentially expressed between the four media (data now shown). Here, I wanted
to identify whether changes in alternative DNA polymerase expression correlated with differences
in mutation frequency and spectrum. Using a spontaneous rifampicin-resistance (Rif
R
) reporter
assay, we cultured 48 biological replicates for 24 hours and measured Rif
R
mutation frequency
35
(Fig. 2.2). We found that TB and SB cultures had the lowest average mutation frequencies with
~1.1x10
-8
mutations/CFU and ~4.4x10
-8
mutations/CFU, respectively. YT and LB cultures
exhibited the highest Rif
R
mutation frequencies with YT cultures at ~1.9x10
-7
mutations/CFU and
LB cultures at ~2.4x10
-6
mutations/CFU.

Figure 2.2 | Distribution of media-specific Rif
R
mutation frequency. The frequency of
spontaneous rifampicin resistance (Rif
R
) for 48 independent replicates for the wild-type strain
grown in each medium. In the large panel, mutation frequencies are indicated in ascending order.
The average Rif
R
mutation frequencies are depicted in the top left corner. LB (black lines), YT
(blue lines), TB (red lines), SB (green lines). Error bars represent standard deviation.

To examine the mutation spectrum for each data set, we isolated two spontaneous Rif
R
colonies
per antibiotic plate and sequenced the two regions of the rpoB gene, encoding the β-subunit of
RNA polymerase, known to contain rifampicin resistance mutations (Garibyan et al., 2003) (Fig.
2.3). To date, there are >100 unique mutant alleles known to confer rifampicin resistance (Corzett
et al., 2013) (Table 2.2) spanning all categories of transition and transversion mutations, making
Rif
R
Frequency
36
this reporter system a quick, easy, and informative assay. Across the different data sets, we found
54 different mutations, two of which are, to the best of our knowledge, novel to this study. The LB
cultures exhibit a large number of CG-GC (C->G or G->C) mutations, while in YT cultures, there
were a higher amount of AT-CG mutations compared to the other culture sets (Fig. 2.3). Notably,
both TB and SB cultures, with the lowest mutation frequencies, show a large quantity of CG-TA
mutations. LB and YT cultures, having the highest mutation frequencies, possess increased
numbers of AT-TA mutations.

Figure 2.3 | Media-specific mutation spectra. The mutation spectra of wild-type E. coli grown
in each of the four media were determined by sequencing the two clusters of the rpoB gene that
correspond to rifampicin resistance (Rif
R
). Ninety-six colonies were sequenced for each group. LB
(black bars), YT (blue bars), TB (red bars), and SB (green bars). X-axis represents the mutation
type, and y-axis denotes the number of mutations identified in the screen per culture medium.

To elucidate whether each culture set favors a certain mutation type, we recalculated the mutation
spectra data to determine transversion/transition ratios (Fig. 2.4). Once again, LB and YT cultures
0
5
10
15
20
25
30
35
40
45
AT-CG AT-GC AT-TA CG-AT CG-TA CG-GC Indel
Number of Mutations
Mutation Class
LB
YT
TB2
SB
37
can be grouped together because they largely favor transversion mutations while TB and SB
cultures slightly favor transitions.

Figure 2.4 | Transversion:transition ratios of bacteria grown in each medium. To identify if
any mutation type was being favored, the mutation spectra data from Fig. 2.3 was recalculated to
determine transversion:transition ratios. Rich media LB and YT cultures favor transversions while
super rich media TB and SB favor transitions. LB (black bar), YT (blue bar), TB (red bar), SB
(green bar).

DNA polymerase mutant strains exhibit no difference in growth phenotype when cultured
in different rich media. Because we see differences in mutation frequency and spectrum due to
growth media and its ties to expression of alternative DNA polymerases, we wanted to evaluate
whether culturing the DNA polymerase mutant strains in the four different media would cause
alterations in growth phenotype compared to wild-type. The single, double, and triple alternative
DNA polymerase mutant strains (Table 2.1) were grown in LB, YT, TB, or SB and their growth
physiologies were compared to that of the wild-type ZK126 strain (Fig. 2.5). Overall, in each
medium, the mutant strains appear largely indistinguishable from wild-type, though polymerase
0.1
1
10
Transversion:Transition
LB
YT
TB
SB
38
mutant strains grown in LB do appear to enter a more drastic death phase than wild-type. LB and
YT cultures exhibit ~2 days in stationary phase before losing viability (Fig 2.5A-B) while TB and
SB cultures all possess extended stationary phase, dying around Day 4 (Fig 2.5C-D).

Figure 2.5 | Alternative DNA polymerase mutant strains grown in each medium look
indistinguishable from wild-type. To determine if growth in different media causes a change in
phenotype in polymerase mutants compared to wild-type (WT), we grew wild-type and each of
the polymerase mutant strains (Table 2.1) in either LB (A), YT (B), TB (C), or SB (D). WT (open
squares), Pol II- (open diamonds), Pol IV- (open circle), Pol V- (open triangle), Pol V+ only
(closed squares), Pol IV+ only (closed diamonds), Pol II+ only (closed circles), triple mutant
(closed triangles). The error bars indicate standard deviation. n=2.

A. B.
C. D.
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10
LB
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10
YT
6
7
8
9
10
11
0 1 2 3 4 5 6 7 8 9 10
TB
6
7
8
9
10
11
0 1 2 3 4 5 6 7 8 9 10
SB
Log
10
CFU/ml
Day
WT
II-
IV-
V-
V+ only
IV+ only
II+ only
Triple mutant
39
2.5 Discussion
Researchers have long wished to understand environmental factors that can affect fundamental
aspects of microbial activity, including the generation of genetic diversity and long-term viability
of bacterial cultures (Corzett et al., 2013; Kram & Finkel, 2015). Previous work from our lab has
not only shown that while previously little considered, factors such as medium composition and
incubation vessel type can alter the physiologies and genetics of the bacteria (Kram & Finkel,
2014; Kram & Finkel, 2015), but the lab has also shown that alternative DNA polymerases
influence mutation frequency and that these polymerases are naturally induced after exponential
phase, potentially affecting long-term evolution (Corzett et al., 2013). In this work, we elucidate
the effect medium content have on mutation frequency and mutation spectrum. Further, we also
tie that into expression and known mutation patterns of the polymerases (Corzett et al., Manuscript
in preparation).

In this work, we note that wild-type E. coli populations exhibit altered growth phenotypes (Fig.
2.1) that are likely the result of the differences in medium composition (Kram & Finkel, 2015).
The four media may be categorized into two different groups: rich and super-rich media; LB and
YT broths fall into the former group. Both have a basic recipe of tryptone, yeast extract, and NaCl,
though YT broth contains twice the concentration of yeast extract and slightly more tryptone. TB
and SB media largely differ from LB and YT because they both have significantly more yeast
extract compared to the first two media (2.4% versus 0.5 and 1.0%, respectively), and they contain
glycerol as well as phosphate buffer (Kram & Finkel, 2015). Here, we observe that a predominant
physiological difference between media is in the length of stationary phase. While LB and YT
cultures are in stationary phase for only about two days, TB and SB cultures remain in stationary
40
phase for two days longer (Fig. 2.1). This observation further validates the concept developed by
Kram & Finkel (2015) that medium content significantly affects the growth patterns bacteria under
laboratory conditions during long-term stationary phase.

Taking this concept a step further, we sought to determine if there were observable differences in
mutation frequency and mutation spectrum that could be tied back to E. coli’s alternative DNA
polymerases. Using the rifampicin resistance (Rif
R
) reporter system, we found that the same E.
coli strain (ZK126) exhibits different mutation frequencies. LB and YT cultures show the highest
Rif
R
mutation frequencies while TB and SB cultures have the lowest. This could potentially be
explained by the observation that at 24 hours, Pol II (polB), with its lower mutation frequency, is
more highly expressed in TB and SB cultures than LB and YT (Corzett et al., Manuscript in
preparation).

Further, Pol II is likely the cause of the increased CG-TA mutations in the mutation spectra data
(Fig. 2.3) (Rangarajan et al., 1999). There are novel trends for each medium culture set, such as
the high frequency of AT-CG mutations in the YT culture. Given that aside from the observation
that Pol II preferentially generates CG-TA mutations, currently, there is little known regarding if
Pol IV or Pol V favors a certain mutation type. In regards to Pol II, a likely follow-up experiment
would be to evaluate the mutation spectrum of the Pol II- single mutant strain (SF2003) and
determine if the C-T mutation spike is diminished in TB and SB cultures.

The additional evaluation that rich media cultures (LB and YT) largely favor transversions while
super-rich media cultures (TB and SB) favor transition mutations (Fig 2.4) provides further
41
evidence of the effect culture medium has on bacterial populations and how this could further link
to alternative DNA polymerase expression. That is not to say, however, that there would be
alterations in growth phenotype of the alternative polymerase mutant strains compared to wild-
type (Fig. 2.5). Especially given the complexity of populations during stationary phase (Navarro
Llorens et al., 2010), it is not surprising that there would be no discernable phenotypic change in
mutant strains grown in the four media. No change in phenotype could simply mean that alternative
DNA polymerases do not play a central role in survival in stationary phase in batch culture
environments.

This work provides some of the first evidence that links alternative DNA polymerase expression
to mutation frequency and spectrum. And as previously stated, alternative DNA polymerases
possess a lower fidelity compared to replicative DNA polymerase III (Tang et al., 2000; Fuchs et
al., 2004; Jarosz et al., 2007), In stressful environments where these alternative polymerases are
naturally induced (Corzett et al., 2013), an increase in mutation frequency seems a likely side
effect of the lower fidelity enzymes. Further, there are many additional experiments to be
performed and extrapolations, such as potential biodefense applications, that could be gleaned
from this research. For example, extrapolation of this work will potentially allow for identification
of the growth conditions of released bioweapons through analysis of mutation frequency and
mutation spectrum.
Table 2.2 | List of Rif
R
mutations identified in this study and number of occurrences
Nucleotide Amino Acid Change Mutation Type LB YT TB SB
436

V146F G-T 0 2 1 6
V146L* G-C* 1 0 0 0
443

Q148P A-C 2 4 0 0
Q148R A-G 2 0 0 0
Q148L A-T 10 5 0 0
444 Q148H G-C 0 1 0 0
42
1525 S509R A-C 0 2 0 0
1526 S509I G-T 0 1 0 0
1527 S509R C-A 0 1 0 0
1532

L511R T-G 1 3 0 0
L511Q T-A 2 2 0 0
1534 S512P T-C 0 2 5 1
1535

S512Y C-A 0 0 3 5
S512F C-T 0 0 4 9
1537 Q513K C-A 0 0 2 2
1538

Q513P A-C 0 3 0 1
Q513R A-G 0 0 0 1
Q513L A-T 1 2 2 2
1546

D516Y G-T 1 1 4 3
D516N G-A 0 2 2 3
1547

D516G A-G 4 1 7 5
D516V A-T 0 0 0 1
1565 S522F C-T 0 0 1 1
1574 T525R C-G 1 0 0 0
1576

H526N C-A 0 2 1 1
H526Y C-T 3 4 16 11
H526D C-G 0 0 3 1
1577 H526L A-T 1 2 0 1
1578

H526Q C-A 0 1 0 0
H526Q C-G 32 0 0 0
1586 R529H G-A 0 0 2 1
1592

S531Y C-A 0 0 0 1
S531F C-T 0 1 7 4
S531C C-G 1 0 0 0
1594 A532P G-C 0 0 0 1
1597 L533V C-G 1 0 0 0
1598

L533R T-G 1 4 0 0
L533P T-C 0 0 2 0
L533H T-A 0 0 1 0
1600

G534C G-T 3 2 0 2
G534S G-A 0 1 0 0
1601

G534V G-T 1 0 0 1
G534D G-A 0 0 0 1
G534A G-C 3 2 0 0
1687 T563P A-C 4 10 0 1
1691 P564L C-T 0 2 1 1
1714

I572L A-C 2 5 2 0
I572F A-T 1 3 5 0
1715

I572S T-G 1 5 7 4
I572T T-C 1 0 0 0
I572N T-A 1 3 0 0
43
1721

S574Y C-A 1 1 0 0
S574F C-T 1 1 0 0
2441-2443 Indel* Indel* 0 0 0 1
* denotes a novel mutation, to the best of our knowledge
44
2.6 References
Banach-Orlowska M, Fijalkowska IJ, Schaaper RM, Jonczyk P. 2005. DNA polymerase II
as a fidelity factor in chromosomal DNA synthesis in Escherichia coli. Mol Microbiol
58:61-70.
Bohannon DE, Connell N, Keener J, Tormo A, Espinosa-Urgel M, Zambrano MM, Kolter
R. 1991. Stationary-phase-inducible “gearbox” promoters: differential effects of katF
mutations and role of s
70
. J Bacteriol 173:4482-4492.
Cai H, Yu H, McEntee K, Kunkel TA, Goodman MF. 1995. Purification and properties of
wild-type and exonuclease-deficient DNA polymerase II from Escherichia coli. J Biol
Chem 270:15327-15335.
Corzett CH, Goodman MF, Finkel SE. 2013. Competitive fitness during feast and famine: how
SOS DNA polymerases influence physiology and evolution in Escherichia coli. Genetics
194:409-420.
Courcelle J, Khodursky A, Peter B, Brown PO, Hanawalt PC. 2001. Comparative gene
expression profiles following UV exposure in wild-type and SOS-deficient Escherichia
coli. Genetics 158:41-64.
Finkel SE, Kolter R. 2001. DNA as a nutrient: novel role for bacterial competence gene homologs.
J Bacteriol 183:6288-6293.
Finkel SE. 2006. Long-term survival during stationary phase: evolution and the GASP
phenotype. Nat Rev Microbiol 4:113-120.
Fuchs R, Fujii S, Wagner J. 2004. Properties and functions of Escherichia coli: Pol IV and
Pol V. Adv Portein Chem 69:229-264.
45
Garibyan L, Huang T, Kim M, Wolff E, Nguyen et al. 2003. Use of the rpoB gene to
determine the specificity of base substitution mutation on the Escherichia coli
chromosome. DNA Repair (Amst.) 2:593-608.
Goodman MF. 2002. Error-prone repair DNA polymerases in prokaryotes and eukaryotes.
Annu Rev Biochem 71:17-50.
Jarosz DF, Beuning PJ, Cohen SE, Walker GC. 2007. Y-family DNA polymerases in
Escherichia coli. Trends Microbiol 15:70-77.
Kraigsley AM, Finkel SE. 2009. Adaptive evolution in single species bacterial biofilms. FEMS
Microbiol Lett 293:135-140.
Kram KE, Finkel SE. 2014. Culture volume and vessel affect long-term survival, mutation
frequency, and oxidative stress of Escherichia coli. Appl Environ Microbiol 80:1732-1738.
Kram KE, Finkel SE. 2015. Rich medium composition affects Escherichia coli survival,
glycation, and mutation frequency during long-term batch culture. Appl Environ
Microbiol 81:4442-4450.
Lange SS, Takata K, Wood RD. 2011. DNA polymerases and cancer. Nat Rev Cancer 11:96-110.
Lee H, Popodi E, Tang H, Foster PL. 2012. Rate and molecular spectrum of spontaneous
mutations in the bacterium Escherichia coli as determined by whole-genome
sequencing. Proc Natl Acad Sci 109:e2774-e2783.
Lewis K. 2000. Programmed death in bacteria. Microbiol Mol Biol Rev 64:503-514.
Makarova KS, Wolf YI, Koonin EV. 2013. Comparative genomics of defense systems in archaea
and bacteria. Nucleic Acids Res 41:4360-4377.
46
Mironova R, Niwa T, Handzhiyski Y, Sredovska A, Ivanov I. 2005. Evidence for non-
enzymatic glycosylation of Escherichia coli chromosomal DNA. Mol Microbiol 55:1801-
1811.
Mironova R, Niwa T, Hayashi H, Dimitrova R, Ivanov I. 2001. Evidence for non- enzymatic
glycosaylation in Escherichia coli. Mol Microbiol 39:1061-1068.
Munoz-Gomez AJ, Santos-Sierra S, Berzal-Herranz A, Lemonnier M, Diaz-Orejas R. 2004.
Insights into the specificity of RNA cleavage by the Escherichia coli MazF toxin. FEBS
Lett 567:316-320.
Navarro Llorens JM, Tormo A, Martínez-García E. 2010. Stationary phase in Gram-
negative bacteria. FEMS Microbiol Rev 34:476-495.
Nohmi T. 2006. Environmental stress and lesion-bypass DNA polymerases. Annu Rev
Microbiol 60:231-253.
Ohmori H, Friedberg EC, Fuch RP, Goodman MF, Hanaoka F et al. 2001. The Y-family of
DNA polymerases Mol Cell 8:7-8.
Pepper ED, Farrell MJ, Nord G, Finkel SE. 2010. Antiglycation effects of carnosine and other
compounds on the long-term survival of Escherichia coli. Appl Environ Microbiol
76:7925-7930.
Pham P, Rangarajan S, Woodgate R, Goodman MF. 2001. Roles of DNA polymerases V and II
in SOS-induced error-prone and error-free repair in Escherichia coli. Proc Natl Acad Sci.
98:8350-8354.
Rangarajan S, Gudmundsson G, Qiu Z, Foster PL, Goodman MF. 1997. Escherichia coli DNA
polymerase II catalyzes chromosomal and episomal DNA synthesis in vivo. Proc Natl Acad
Sci 94:946-951.
47
Rangarajan S, Woodgate R, Goodman MF. 1999. A phenotype for enigmatic DNA
polymerase II: a pivotal role for pol II in replication restart in UV-irradiated Escherichia
coli. Proc Natl Acad Sci 96:9224-9229.
Robbins JH, Kraemer KH, Lutzner MA, Festoff BW, Coon HG. 1974. Xeroderma
pigmentosum. An inherited disease with sun sensitivity, multiple cutaneous neoplasms,
and abnormal DNA repair Ann Intern Med 80:221-248.
Sale JE, Lehmann AR, Woodgate R. 2012. Y-family DNA polymerases and their role in
tolerance of cellular DNA damage. Nat Rev Mol Cell Biol 13:141-152.
Stallons LJ, McGregor WG. 2010. Translesion synthesis polymerases in the prevention and
promotion of carcinogenesis. J Nucleic Acids pii:643857.
Tang M, Pham P, Shen X, Taylor JS, O’Donnell M et al. 2000. Roles of E. coli DNA
polymerases IV and V in lesion-targeted and untargeted SOS mutagenesis. Nature
404:1014-1018.
Yeiser B, Pepper ED, Goodman MF, Finkel SE. 2002. SOS-induced DNA polymerases
enhance long-term survival and evolutionary fitness. Proc Natl Acad Sci 99:8737-8741.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zinser EER, Kolter R. 2000. Prolonged stationary-phase incubation selects for lrp
mutations in Escherichia coli K-12. J Bacteriol 182:4361-4365.
Zinser ER, Kolter R. 2004. Escherichia coli evolution during stationary phase. Res
Microbiol 155:328-336.
48
Zinser ER, Schneider D, Blot M, Kolter R. 2003. Bacterial evolution through the selective
loss of beneficial genes. Trade-offs in expression involving two loci. Genetics 164:1271-
1277.

49
Chapter 3: Physiological, Genetic, and Transcriptomic Analysis of Alcohol-Induced Delay
of Bacterial Death in Escherichia coli

The content of this chapter appears essentially as published in January 2019 in Applied &
Environmental Microbiology, 85:e02113-18. Publication title: Ferraro CM, Finkel SE.,
“Physiological, Genetic, and Transcriptomic Analysis of Alcohol-Induced Delay of Bacterial
Death in Escherichia coli.”

3.1 Abstract
When Escherichia coli K-12 is inoculated into rich medium in batch culture, cells experience five
phases. While the lag and logarithmic phases are mechanistically fairly well-defined, stationary
phase, death phase, and long-term stationary phase are less well understood. Here, we characterize
a mechanism of delaying death, a phenomenon we call the “alcohol effect,” where the addition of
small amounts of certain alcohols prolongs stationary phase for at least 10 days longer than in
untreated conditions. We show that stationary phase is extended when ethanol is added above a
minimum threshold concentration. Once ethanol levels fall below a threshold concentration, cells
enter death phase. We also show that the effect is conferred by the addition of straight-chain
alcohols 1-propanol, 1-butanol, 1-pentanol, and, to a lesser degree, 1-hexanol. However, methanol,
isopropanol, 1-heptanol and 1-octanol do not delay entry into death phase. Though modulated by
RpoS, the alcohol effect does not require RpoS activity or the activities of the AdhE or AdhP
alcohol dehydrogenases. Further, we show that ethanol is capable of extending the lifespan of
stationary phase cultures for non-K-12 E. coli strains and that this effect is caused in part by genes
of the glycolate degradation pathway. These data suggest a model where ethanol and other shorter
1-alcohols can serve as signaling molecules, perhaps by modulating patterns of gene expression
that normally regulate the transition from stationary phase to death phase.
50

3.2 Introduction
The mechanism(s) modulating bacterial death during batch culture are poorly understood (Finkel,
2006; Navarro Llorens et al., 2010) and debate continues regarding whether the transition from
stationary phase to death phase in Escherichia coli is a stochastic or a “programmed” process akin
to a form of bacterial apoptosis (Azienman et al., 1996; Finkel, 2006; Navarro Llorens et al., 2010;
Bayles, 2014; Allocati et al., 2015; Chandra et al., 2016). In typical batch culture in rich media, E.
coli K-12 strains will begin to die after one to two days in stationary phase, resulting in the loss of
viability of 99% of cells with the surviving ~1% of the population transitioning into long-term
stationary phase (LTSP) (Finkel, 2006; Kram & Finkel, 2014; Kram & Finkel, 2015). Not
surprisingly, there is an inherent difficulty in studying bacterial cell death mechanisms since most
experimental approaches focus on the surviving subpopulations. While previous studies have
utilized computer simulation models and synthetic biological techniques to study population-level
bacterial death (Tanouchi et al., 2012; Avrani et al., 2017), an alternative method is to characterize
modes of delaying the transition to death phase (Lazar, 1996; Vulić & Kolter, 2002; Farrell &
Finkel, 2003). Farrell and Finkel characterized one such method of temporary death delay through
buffering of the growth medium (Farrell & Finkel, 2003). These authors showed that by keeping
a constant pH of 7.0 in Luria-Bertani (LB) complex growth medium, stationary phase could be
extended for multiple days. This is the case in both wild-type and a rpoS-attenuated mutant strain
with severely diminished functionality. Both this work and that of Farrell and Finkel describe two
of the main modes of delaying death.

The “alcohol effect,” or “alcohol-induced delay of death,” was first observed in the mid-1990s by
members of Roberto Kolter’s laboratory (Lazar, 1996; Vulić & Kolter, 2002) when they sought to
51
determine if there was a protein synthesis-dependent mechanism for entry into death phase, using
chloramphenicol as a translation inhibitor (Rendi & Ochoa, 1962; Kohanski et al., 2010). This
experiment meant to elucidate a role for active protein synthesis being required for the transition
into death phase (Azienman et al., 1996; Finkel, 2006; Navarro Llorens et al., 2010; Bayles, 2014;
Kram & Finkel, 2014; Kram & Finkel, 2015). While it was initially observed that the addition of
sub-lethal amounts of the antibiotic could cause a delay in the initiation of death phase, control
experiments revealed that chloramphenicol was not the causal agent. Instead, it was the ethanol
solvent that lead to the maintenance of stationary phase (Lazar, 1996). Traditionally, studies have
focused on adding nearly lethal concentrations of alcohol to better understand ethanol tolerance
for biofuel production studies (Horinouchi et al., 2010; Haft et al., 2014). Some of the common
physiological effects resulting from addition of high doses of alcohol include increased membrane
permeability and slowed outgrowth of E. coli populations compared to wild-type (Ingram, 1976;
Ingram & Vreeland, 1980; Ly & Longo, 2004; Horinouchi et al., 2010). The phenomenon, where
sublethal alcohol additions actually confer a potentially positive benefit, was subsequently dubbed
the “alcohol effect.”

Vulić and Kolter (Vulić & Kolter, 2002) previously demonstrated that the onset of death phase can
be delayed when ethanol, 1-propanol, or 1-butanol were added after 24 hours of batch culture
incubation. They also showed that an rpoS null mutant strain “lost” the effect and strains without
the AdhE alcohol dehydrogenase still exhibited the delay. These data suggested that the effect does
not require the catabolism of ethanol as a carbon source (Clark & Cronan, 2005), but instead that
some regulated change is required. Here, we more extensively characterize the physiological,
52
genetic, and transcriptomic factors associated with alcohol-mediated delay of death phase,
including a further exploration of the requirement of a functional rpoS gene.

3.3 Materials and Methods
Table 3.1 | E. coli strains used in this study
Strain Relevant Genotype/Origin Nomenclature Reference
ZK126 W3110 ∆lacU169 tna-2 Wild-type Zambrano et al., 1993
PFM2 MG1655 ∆pyrE748 rph
+
IS186 Wild-type Lee et al., 2012
BW25113 BD792 rrnB3 ∆lacZ4787 hsdR514
∆(araBAD)567 ∆(rhaBAD)568 rph-1
Wild-type Datsenko & Wanner,
2000; Baba et al., 2006
ZK1000 ZK126 rpoS::Kan RpoS null Bohannon et al., 1991
SF2602 ZK126 adhE adhP::Kan Double mutant This study
SF2603 ZK126 gcl::Kan Gcl null This study
SF2604 ZK126 glcD::Kan GlcD null This study
SF2605 ZK126 glxR::Kan GlxR null This study
ECOR-04 Human; Iowa Natural isolate Ochman & Selander, 1984
ECOR-13 Human; Sweden Natural isolate Ochman & Selander, 1984
ECOR-14 Human; Sweden Natural isolate Ochman & Selander, 1984
ECOR-15 Human; Sweden Natural isolate Ochman & Selander, 1984
ECOR-28 Human; Iowa Natural isolate Ochman & Selander, 1984
ECOR-29 Kangaroo rat; Nevada Natural isolate Ochman & Selander, 1984
ECOR-37 Marmoset; Washington (zoo) Natural isolate Ochman & Selander, 1984
ECOR-38 Human; Iowa Natural isolate Ochman & Selander, 1984
ECOR-40 Human; Sweden Natural isolate Ochman & Selander, 1984
ECOR-51 Human infant; Massachusetts Natural isolate Ochman & Selander, 1984
ECOR-62 Human; Sweden Natural isolate Ochman & Selander, 1984
ECOR-63 Human; Sweden Natural isolate Ochman & Selander, 1984
ECOR-68 Giraffe; Washington (zoo) Natural isolate Ochman & Selander, 1984
ECOR-71 Human; Sweden Natural isolate Ochman & Selander, 1984

Bacterial strains, culture media, and growth conditions. E. coli strains used in the study are
listed in Table 3.1, with most experiments performed using the E. coli K-12 strain ZK126, derived
from the W3110 lineage (Zambrano et al., 1993). Other strains discussed included Pseudomonas
aeruginosa PA-14, Shewanella oneidensis MR-1, Vibrio harveyi B392, Streptococcus and
Klebsiella (laboratory isolates from human fecal samples). Unless stated otherwise, cultures were
53
inoculated from frozen 20% glycerol stocks into 5 ml of Luria-Bertani (Lennox) medium (LB)
(Difco) in 18- x 150-mm borosilicate test tubes (Thermo Fisher) and incubated at 37
°
C with
aeration using TC-7 rolling drums (New Brunswick Scientific, Edison, NJ). Cells from overnight
cultures were then inoculated 1:1000 (vol:vol) into 5 ml LB culture. After 24 hours of incubation,
alcohols (Koptec; Sigma) were added to different specified concentrations. For experiments
requiring many replicate cultures, one large volume of LB was inoculated from the overnight
culture, and then aliquoted into test tubes. Other strains tested include rpoS null mutant strain
ZK1000 of the ZK126 lineage (Bohannon et al., 1991), strain PFM2 (Lee et al., 2012) of the
MG1655 lineage, strain BW25113 (the parent of the Keio Collection of gene knockouts; Datsenko
& Banner, 2000; Baba et al., 2006), and the E. coli Reference Collection (ECOR) strains (26)
listed in Table 3.1. The isogenic adhE adhP::Kan (SF2602) double mutant strain and gcl::Kan
(SF2603), glcD::Kan (SF2604), and glxR::Kan (SF2605) single mutant strains were constructed
by a combination of P1 transduction and FLP-FRT recombination to remove the kanamycin gene
cassette interrupting the genes of interest (adhE), as described (Datsenko & Wanner, 2000).

Monitoring cell growth, cell survival, and culture pH. Viable cell counts were determined by
serial dilution at indicated time points and plating on LB-agar (Kraigsley & Finkel, 2009); the limit
of detection for this method of titering is ≥1,000 CFU/ml (Kraigsley & Finkel, 2009). Where
appropriate, the pH was measured using 6.0 to 10.0 range pH paper with ~0.3 pH unit increments
(EMD Chemicals, La Jolla, CA).

Ethanol colorimetric concentration assay. Ethanol concentration in cultures was measured using
a colorimetric assay (BioVision Inc., Milpitas, CA), according to the manufacturer’s instructions,
54
using a standard curve generated with known concentrations of ethanol ranging from 2-20 mM.
Briefly, in a sterile biological hood, samples of culture medium were obtained and resuspended in
ethanol assay buffer. 50 µl samples were then transferred to a 96-well plate with lid (Corning). 46
µl enzyme assay buffer, 2 µl enzyme mix, and 2 µl of enzyme probe were then added and the plates
were incubated at room temperature for one hour with no light exposure. After incubation, OD
570 nm was determined for each sample and compared to the standard curve. As appropriate,
several different sample volumes were obtained to ensure that the ethanol concentrations were
within the linear range of the assay.

RNA-sequencing preparation, sequencing, and analysis. E. coli K-12 strain ZK126 was
inoculated in 5 ml of LB medium from frozen a 20% glycerol stock and grown at 37
°
C as described
above. After 24 hours of incubation, 10 µl (~35.2 mM) of 95% ethanol (Koptec; Sigma) was added
to duplicate cultures, an untreated pair of cultures serving as the negative control. Treated and
untreated cultures were incubated for one additional hour. The mRNA was then purified from 0.5
ml of each bacterial culture using the RNeasy Mini Kit (Qiagen). Samples were sent to BGI
Americas Corporation (Cambridge, MA) for library preparation and sequencing using the Illumina
HiSeq 4000 platform. The 100 bp paired-end reads were aligned to the E. coli K-12 W3110
genome. Normalized counts (transcripts per million, TPM), accounting for total number of reads
and gene size, were calculated using SAMtools, Bowtie2, TopHat2, and HTSeq (Trapnell et al.,
2009; Trapnell et al., 2012; Anders et al., 2015). EdgeR (Bioconductor) software was used to
analyze differential expression between treatments (Anders et al., 2013).
55
3.4 Results
Ethanol addition prolongs stationary phase in a dose-dependent manner. To define the
minimum concentration of ethanol that can produce the delayed death effect, 1 to 6 µl of 95%
ethanol was added to 5 ml batch cultures of E. coli on Day 1 of incubation (Fig. 3.1A), which
corresponds to stationary phase populations. 5 µl (~17.6 mM) was found to be the minimum
amount necessary to cause a one-day delay in death (Fig. 3.1A) as the addition of 1 to 4 µl (~3.5-
14.1 mM) shows no effect.

Figure 3.1 | Wild-type strain treated with different concentrations of ethanol on Day 1 shows
a dose-dependent alcohol effect. To determine the minimum and maximum concentrations
necessary to manifest a delay of death effect, various concentrations of 95% ethanol were added
6
7
8
9
10
0 1 2 3 4 5
6 µl
5 µl
4 µl
3 µl
2 µl
1 µl
0 µl
50 µl
15 µl
10 µl
5 µl
0 µl
A.
B.
C.
6
7
8
9
10
0 1 2 3 4 5 6 7
Log
10
CFU/ml
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10
Day
450 µl
350 µl
250 µl
150 µl
50 µl
0 µl
❋
56
to 5 ml LB cultures. (A) Wild-type cultures were checked for the minimum concentration that
generates an effect. Untreated (closed squares), 1 µl (closed diamonds; ~3.5 mM), 2 µl (closed
circles; ~7.0 mM), 3 µl (closed triangles; ~10.6 mM), 4 µl (open diamonds; ~14.1 mM), 5 µl (open
squares; ~17.6 mM), and 6 µl (open circles; ~21.1 mM). (B) E. coli exhibit a dose-dependent
effect. Cultures were either treated with 0 µl (closed squares), 5 µl (open squares), 10 µl (open
diamonds; ~35.2 mM), 15 µl (open circles; ~52.8 mM), or 50 µl (open triangles; ~176 mM). (C)
Different concentrations were added to cultures to identify the maximum concentration able to
generate an effect. Cultures were treated with 0 µl (closed squares), 50 µl (closed diamonds), 150
µl (open squares; ~528 mM), 250 µl (open diamonds; ~880 mM), 350 µl (open circles; ~1.2 M),
or 450 µl (open triangles; ~1.6 M). Error bars represent standard error of replicates; n=3. Asterisks
indicate viable cell counts below the limit of detection (<1000 CFU/ml).

Incrementally increasing the amount of added ethanol further extends the length of stationary
phase (Fig. 3.1B). Adding 10 µl (~35.2 mM) delays death by two days, while a 15 µl (~52.8 mM)
addition delays death by three days. When 50 µl (~176 mM) of ethanol is added on Day 1, cells
remain in stationary phase until at least Day 7.

This effect is not strain-specific. In addition to the W3110-lineage strain ZK126 (Zambrano et al.,
1993), we tested: another W3110 strain, the Keio Collection parental strain BW25113 from the
BD792 background (Datsenko & Wanner, 2000; Baba et al., 2006); PFM2 from the MG1655
lineage (Lee et al., 2012); and 14 strains from the ECOR collection of E. coli natural isolate
“reference” strains (Ochman & Selander, 1984) (Fig. S3.1). An additional six E. coli strains
isolated from human and canine fecal samples (obtained from the laboratory of I. Ehrenreich) were
also tested (data not shown). Every strain tested exhibited the effect except for two strains from
the ECOR collection (ECOR-37 and ECOR-40); one showed no effect (Fig. S3.1G), and one
showed extreme sensitivity to the presence of ethanol (Fig. S3.1I).

To determine the maximum concentration for which the ethanol effect is observed, doses of 50 µl
to 450 µl (~1.6 M) were added on Day 1 of incubation (Fig. 3.1C). Dosages of 50 µl and 150 µl
57
(~528 mM) delay death for 7 days, while a dose of 250 µl (~880 mM) prolongs stationary phase
until after Day 10 in batch culture. Cultures receiving 350 µl (~1.2 M) or 450 µl of ethanol do not
exhibit the delay of death phenotype, with cells receiving the highest dose showing a significant
loss of viability starting on Day 2. This shift from beneficial to harmful concentrations of ethanol
likely reflects the balance between the alcohol potentially serving as a beneficial signaling
molecule, rather than a toxic denaturant (Yomano et al., 1998).

Further, this effect appears to be phase-specific phenomenon. When alcohol is added to cultures
on Day 0 (lag phase), Day 1 (stationary phase), or Day 2 (stationary phase), E. coli populations
exhibit the delayed death phenotype. However, if alcohol is added after the transition into death
phase (Day 3 or later) there is no phenotypic difference compared to untreated cultures (data not
shown).

Since increasing concentrations of ethanol cause dose-dependent delays in death, we next sought
to verify that the cells present in the medium were responsible for the depletion of ethanol. To test
this, we compared the extracellular concentrations of ethanol in the presence or absence of active
cultures over time (Fig. 3.2). 5 µl of ethanol was added on Day 1 and the ethanol concentration
was determined over 4 days. By Day 3, ethanol is almost entirely depleted from the culture medium
in the presence of cells (Fig. 3.2), correlating with the transition of the population into death phase.
In the control culture lacking E. coli cells, the decrease of ethanol in the culture medium is much
slower, with concentrations of ~10 mM still present after 4 days of incubation.
58

Figure 3.2 | Depletion of ethanol over time in active cultures. Viable cell counts of wild-type
cells over four days in LB medium is shown (closed squares). On Day 1, 5 µl of 95% ethanol
(~17.6 mM) was added to 5 ml cultures, or no cell controls were prepared. Ethanol concentration
measurements (mM) are indicated for ZK126 cultures (open squares) and no cell controls (open
circles). Ethanol is depleted from cultures immediately prior to the onset of death phase and at a
faster rate than with the no cell control. Error bars represent standard error of replicates; n=3.

Maintenance of a minimum ethanol concentration is required to prolong stationary phase.
Given that greater doses of ethanol lead to a longer delay of entry into death phase, we next
determined whether small daily doses of ethanol, ensuring that ethanol is always present in the
medium, would similarly increase the length of stationary phase. Starting on Day 1, daily additions
of 5 µl ethanol were added to wild-type cultures (Fig. 3.3), which prolonged stationary phase for
greater than 9 days. In contrast, a single addition of 5 µl ethanol resulted in only a single day of
delayed death (Fig. 3.1A; Fig. 3.3). And conversely, cells treated with a one-time dose of 50 µl
ethanol, corresponding to more than the total amount added over the nine-day time course, die
sooner than those treated with smaller daily amounts (Fig. 3.1C). Further, cells treated with later
administrations of ethanol, such as on Day 2, do show an “alcohol effect” as long as populations
are still in stationary phase when the alcohol is added (data not shown). Populations are not
affected by a dose of ethanol after the onset of death phase.
0
5
10
15
20
6
7
8
9
10
0 1 2 3 4
mM Ethanol
Log
10
CFU/ml
Day
59

Figure 3.3 | Daily ethanol addition results in a prolonged stationary phase effect. 5 µl ethanol
was added daily (open circles) to ZK126 cultures in LB and viable cell counts (CFU/ml) were
measured. Untreated cultures are indicated with closed squares, and cultures treated with a single
dose on Day 1 are represented as open diamonds. Error bars represent standard error of replicates;
n=3.

Death is delayed in the presence of other short, straight-chain alcohols between 2 and 6
carbons in length. Cultures treated with equimolar concentrations of either 1-propanol, 1-butanol,
or 1-pentanol are able to delay death phase of E. coli for one to two days longer compared to the
addition of ethanol (Fig. 3.4; Table 3.2). While the addition of an equimolar amount of 1-hexanol
causes loss of viability, the addition of half the dose prolongs stationary phase, although not to the
same extent as other short-chain alcohols. The addition of 1-heptanol and 1-octanol proved lethal
at the concentrations tested. Surprisingly, neither the addition of equimolar amounts of methanol
nor 2-propanol induces the effect (Fig. 3.4; Table 3.2). Similarly, the addition of other diols and
amines that have structural similarity to ethanol, such as ethanolamine, 3-amino-1-propanol, and
ethylene glycol, do not cause the effect (Table 3.2).
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9
Log
10
CFU/ml
Day
60

Figure 3.4 | Other n-alcohols can produce the alcohol effect. Viable counts of cultures with
either 17.6 mM 1-propanol (open diamonds), 1-butanol (open circles), 1-pentanol (open triangles),
or 5 µl (8.0 mM) 1-hexanol (open squares) added to one-day-old cultures is shown. 2-Propanol
(closed diamonds) shows no effect compared to untreated cultures (closed squares). Error bars
represent standard error of replicates; n=3.

Table 3.2 | Chemicals tested for delayed death effect
Compound Chemical Formula Concentration
(mM)
Phenotype
a
Methanol CH3OH 17.6-24.7 -
Ethanol CH3CH2OH 17.6 +
1-Propanol CH3(CH2)2OH 17.6 +
2-Propanol CH3CHOHCH3 13.1-17.6 -
1-Butanol CH3(CH2)3OH 17.6 +
1-Pentanol CH3(CH2)4OH 17.6 +
1-Hexanol

CH3(CH2)5OH 8.0-17.6
b
+/-

1-Heptanol

CH3(CH2)6OH 4.8-17.6
c
-

1-Octanol

CH3(CH2)6CH2OH 3.8-17.6
d
-

Ethanolamine

H2NCH2CH2OH 6.6-17.6
c
-

3-Amino-1-propanol

H2NCH2CH2CH2OH 5.2-17.6
c
-

4-Amino-1-butanol

H2N(CH2)4OH 4.3-17.6
c
-

Ethylene glycol HOCH2CH2OH 7.2-17.6 -
1,3-Propanediol HOCH2CH2CH2OH 5.5-17.6 -
a
Plus = has delayed death; Minus = no delay; Plus/minus = has delayed death at lower but not
equimolar (~17.6 mM) concentrations.
b
Lethal at equimolar (~17.6 mM) concentrations; has effect at lower concentrations.
c
Lethal at equimolar concentrations; no effect at lower concentrations.
d
Lethal at all concentrations tested.
6
7
8
9
10
0 1 2 3 4 5 6 7
Log
10
CFU/ml
Day
Untreated
Ethanol
2-Propanol
1-Propanol
1-Butanol
1-Pentanol
1-Hexanol
61

The “alcohol effect” is modulated by RpoS activity. Previously, Vulić and Kolter (Vulić &
Kolter, 2002) reported that the ethanol effect is a stationary phase-specific phenomenon and
depends on the activity of RpoS, which either directly or indirectly regulates the expression of
approximately 23% of the genome (Wong et al., 2017). They showed that a rpoS null mutant strain
shows no effect when treated with ethanol. We sought to verify and expand upon these findings
by testing for the effect using a more frequently sampled time course with the isogenic rpoS null
mutant strain ZK1000 (Bohannon et al., 1991) where a kanamycin-resistance gene cassette has
replaced the rpoS gene, completely disrupting its function.

Figure 3.5 | An rpoS null mutant strain exhibits the alcohol effect. The rpoS null strain ZK1000
shows a small effect when 5 µl ethanol (open circles) is added on Day 1 of growth compared to
untreated cultures (closed circles). Wild-type treated (5 µl ethanol; open circles) and untreated
(closed squares) cultures serve as controls. Error bars represent standard error of replicates; n=3.

The rpoS null strain was treated with ethanol, and viable cell counts were determined several times
a day over three days of incubation. The rpoS null mutant still shows a prolonged stationary phase
effect, although shorter than for wild-type strains (Fig. 3.5), suggesting that while not essential,
RpoS modulates the effect. This was likely previously unreported (Vulić & Kolter, 2002) because
Wild-type 0 µl
Wild-type 5 µl
RpoS Null 0 µl
RpoS Null 5 µl
6
7
8
9
10
0 1 2 3 4
Log
10
CFU/ml
Day
62
the effect’s induction occurs at an earlier time for the rpoS mutant, and the timing for this event
would have been missed due to less frequent sampling compared to this study. Further, a rpoS
mutant with reduced activity exhibits a similarly robust phenotype to the wild-type strain (data not
shown).

Alcohol dehydrogenase activity is not essential to prolong stationary phase. One possible
mechanism of the “alcohol effect” is the metabolism of ethanol as a carbon source (Clark &
Cronan, 2005). To address this, survival patterns were determined for mutants lacking both of E.
coli’s alcohol dehydrogenase genes, adhE and adhP (Rudolph et al., 1968; Hoog et al., 1999).
Cultures of mutant cells were treated with varying amounts of either ethanol or 1-butanol, an
alcohol that shows an effect in wild-type cells but is not metabolized (Clark & Rod, 1987; Zhang
et al., 2008; Tseng et al., 2012), and viable cell counts were determined over 4 days (Fig. 3.6).

Figure 3.6 | An alcohol dehydrogenase double mutant strain exhibits the alcohol effect. To
determine if ethanol is being metabolized as a carbon source, an alcohol dehydrogenase adhE adhP
double mutant strain (open symbols) or wild type cells (closed symbols) were tested. Double
mutant cultures were treated with either no ethanol (open squares), 5 µl (17.6 mM; open
diamonds), or 50 µl (open circles) ethanol additions on Day 1 to LB cultures and viable cell counts
WT 0 µl
WT 5 µl
WT 50 µl
WT Butanol
Double Mutant 0 µl
Double Mutant 5 µl
Double Mutant 50 µl
Double Mutant Butanol
6
7
8
9
10
0 1 2 3 4
Log
10
CFU/ml
Day
63
(CFU/ml) were determined. In addition, 17.6 mM 1-butanol (open triangles) was added on Day 1.
Error bars represent standard error of replicates; n=2.

Without treatment (Fig. 3.6), the alcohol dehydrogenase double mutant behaves like the wild-type
strain, experiencing two days in stationary phase. While the addition of 5 µl ethanol on Day 1
causes only a slight effect, the addition of 50 µl ethanol causes a prolonged extension of stationary
phase (Fig. 3.6). Cultures treated with equimolar amounts of 1-butanol also show an effect, further
indicating that alcohol catabolism is not required to cause a delay in the onset of death phase.

Enzymes involved in the glycolate degradation pathway may help modulate the alcohol
effect.
Given that essential causal gene(s) involved in the alcohol effect remain unknown after testing
several hypotheses, we next chose to analyze the transcriptome of the E. coli populations in the
presence or absence of alcohol treatment by RNA-seq. Wild-type cultures were either left
untreated or treated with 10 µl (~35.2 mM) ethanol after 24 hours. After one hour of additional
incubation, mRNA was prepared from culture samples and submitted for sequencing.
Transcriptome data were normalized and pairwise comparisons were made between duplicate
treated and untreated populations (Anders et al., 2013).
Table 3.3 | Most significantly upregulated genes one hour after ethanol addition
Gene Synonym(s) Description Fold
Change
p-Value
glxR glxB1, ybbQ Tartronate semialdehyde reductase 2 36.25 7.10e-57
hyi ybbG, gip Hydroxypyruvate isomerase 19.83 1.28e-36
gcl Glyoxylate carboligase 12.05 2.88e-75
glcD yghM, gox Glycolate dehydrogenase, putative
FAD-linked subunit
11.46 2.14e-94

64
An analysis of the most highly induced genes showed that three of the four genes identified are
involved in the glycolate/glyoxylate degradation pathway: glxR, glcD, and gcl (Clark & Cronan,
2005) (Table 3.3; Table S3.1; Table S3.2). To determine whether any of these genes modulate the
alcohol effect, single gene knockout mutations were constructed by bacteriophage P1 transduction.
When treated with 5 µl ethanol, all three single mutants have longer delayed death phenotypes
compared to wild-type cells (Fig. 3.7B; Fig. S3.2), though to varying degrees. Both the untreated
glxR and glcD mutant strains only stay in stationary phase for one day before entering death phase
(Fig. 3.7A; Fig. S3.2A-B); the untreated gcl mutant strain survives two days of stationary phase
before dying, similar to wild-type (Fig. 3.7A; Fig. S3.2C). Interestingly, the gcl null mutant also
shows a severely reduced butanol effect compared to wild-type while the glxR and glcD null
populations are unaffected with respect to the addition of butanol (Fig. 3.7C).
65

Figure 3.7 | Single gene null mutations of glycolate degradation pathway genes glxR, glcD,
and gcl, have altered delayed death phenotypes. The RNA-sequencing analysis (Table 3; Table
S3.1; Table S3.2) showed significant upregulation of three genes involved in the
glycolate/glyoxylate degradation pathway: glxR, glcD, and glxR. To test if these play a role in the
causing the “alcohol effect,” single mutant knockouts were made of each (19). Wild-type
populations (closed squares), GlxR null populations (open diamonds), GlcD null populations
(open circles), and Gcl null populations (open triangles). (A) Untreated cultures; (B) Treated
cultures with 5 µl (~17.6 mM) ethanol; (C) Treated cultures with ~17.6 mM 1-butanol. Error bars
represent standard error of replicates; n=3.

6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
Log
10
CFU/ml
6
7
8
9
10
0 1 2 3 4 5
Day
Wild-type
GlxRNull
GlcDNull
GclNull
A.
B.
C.
66
3.5 Discussion
We show that the addition of ethanol, and several other straight-chain alcohols, causes a delay in
the onset of death phase leading to prolonged stationary phase during batch culture incubation in
a rich medium. Increasing concentrations of ethanol delay death in a dose-dependent manner, until
toxic levels are reached (Fig. 3.1B-C). This dose-dependency is likely caused by the presence of
ethanol above a minimum threshold (Fig. 3.1A). We also show that ethanol is depleted at a faster
rate in the presence of cells (Fig. 3.2) and that once ethanol is no longer detected in the culture
medium, populations enter death phase. We posit that the faster depletion of ethanol is due to the
cells themselves acting as a sink for the alcohols since they can readily pass through the membrane
(Cooper, 2000), and apparently this happens at a rate faster than simple evaporation. We initially
proposed several models to explain the “alcohol effect”: (i) alcohol is being metabolized as a
carbon source (Clark & Cronan, 2005); (ii) the presence of ethanol or other alcohols in cultures
triggers a stress response; or (iii) alcohols are serving as signaling molecules.

E. coli possesses a natural ethanol degradation pathway (Clark & Cronan, 2005), so while it is
plausible that ethanol could be utilized as a carbon source whose metabolism might lead to a delay
of entry into death phase, data from several experiments argue against this model. We show that
the alcohol dehydrogenase double mutant strain displays the alcohol-induced delayed death
phenotype (Fig. 3.6). Also, treatment with 1-butanol (Fig. 3.6), which cannot be catabolized as a
nutrient by the E. coli strains used in this study (Table 3.1) (Clark & Rod, 1987; Zhang et al., 2008;
Tseng et al., 2012), still induces the effect. Together, these data suggest that metabolism of ethanol,
or any alcohol, as a nutrient is not responsible for the delayed entry into death phase.

67
We next speculated that the presence of alcohol in the culture medium may trigger a protective
stress response through one or more alternative sigma factors. The “alcohol effect” is a stationary
phase-specific phenomenon (Vulić & Kolter, 2002), and RpoS is a global regulator of E. coli’s
stress response (Lange & Hengge-Aronis, 1991; Membrillo-Hernández & Lin, 1999; Hengge-
Aronis, 2002; Hengge, 2009; Battesti et al., 2011; Wong et al., 2017). In the absence of added
alcohol, the rpoS null mutant strain behaves differently than the wild-type ZK126 strain. The rpoS
mutant strain has a shorter stationary phase of only one day in untreated medium, compared to two
days for the wild-type (Fig. 3.5), and enters death phase before Day 2. However, the rpoS null
strain still exhibits the effect (Fig. 3.5), suggesting that while RpoS may modulate stationary phase
activities, it is not essential to prolong stationary phase in the presense of added alcohol.

Previous reports have shown that ethanol stresses the E. coli cell envelope by increasing its
permeability, affecting growth rate (Ingram, 1976; Ingram & Vreeland, 1980; Ly & Longo, 2004;
Horinouchi et al., 2010). A key regulator involved in response to envelope stress, including that
caused by ethanol, is the alternative sigma factor RpoE (Hayden & Ades, 2008; Cao et al., 2017).
While other work has described a role for RpoE in responding to the increased membrane stress
caused by ethanol, the vast majority of the studies describing RpoE’s role in alcohol stress involve
doses of ethanol that are significantly higher (~2.5-5%) than those used in this study (~0.1-0.2%)
(Horinouchi et al., 2010; Haft et al., 2014; Cao et al., 2017). Further, no significant change in gene
expression of members of the RpoE regulon directly regulated by the sigma factor were detected
in the RNA-seq experiment.

68
In addition to 1-butanol, other non-catabolizable straight-chain alcohols also cause the delayed
death phase effect (Fig. 3.4). Straight-chain n-alcohols containing between 2 and 6 carbons lead to
the alcohol effect. Equimolar concentrations of straight-chain alcohols containing between 3 and
5 carbons not only induce the effect but also result in the cells staying an additional day in
stationary phase. The increased amphiphilicity of these longer alcohols could result in greater
membrane permeability (Ly & Longo, 2004), allowing more alcohol to enter the cells, causing a
prolonged effect. 1-hexanol, while lethal at higher concentrations, causes an effect when added at
a half dose. The intermediate dose response of 1-hexanol likely represents a balance between the
positive life-extension alcohol effect versus the alcohol’s toxicity. Further, the 1-carbon alcohol,
methanol, and branched alcohol, 2-propanol, do not induce the effect, while 1-heptanol causes
either no effect or proves lethal to cells. We conclude that to cause the death delay phenotype,
alcohols require straight-chain structures between 2 and 6 carbons in length. The fact that 2-
propanol does not induce the effect, while 1-propanol does, further supports a model where the
alcohol may be directly interacting with some protein as a signaling molecule and the presence of
the hydroxyl group at the 2-position interferes with the interaction.

Other groups have previously performed transcriptomic analyses of ethanol treatment to E. coli
(Horinouchi et al., 2010; Haft et al., 2014), but those studies were done while performing directed
evolution to yield strains with increased ethanol tolerance. Therefore, significantly higher
concentrations of ethanol were added to cultures compared to this study. Here, using lower
concentrations, we identify three ethanol-induced genes involved in the glycolate/glyoxylate
degradation pathway (Clark & Cronan, 2005) that alter the delayed death phenotype when knocked
out: glxR (tartronate semialdehyde reductase), glcD (glycolate dehydrogenase), and gcl
69
(glyoxylate carboligase) (Table 3.3; Fig. 3.7; Table S3.1; Table S3.2). These three genes serve as
enzymes in the first three steps of the glycolate degradation pathway (Clark & Cronan, 2005). This
same pathway appears to be essential for E. coli to grow on either glycolate or glyoxylate as a sole
carbon source and also feeds into gluconeogenesis. These genes were not noted in the previous
ethanol transcriptome studies as responding to the addition of excess ethanol (Horinouchi et al.,
2010; Haft et al., 2014). However, a recent study showed that genes involved in the glycolate
degradation pathway, including all three of the genes identified here, are upregulated in the
presence of butanol (Si et al., 2016). Here, all three null strains show varying degrees of increased
delayed death compared to wild-type, while only the gcl null populations show a reduced, but still
present, alcohol effect with equimolar amounts of butanol (Fig. 3.7C). Given that butanol cannot
be catabolized by our parental E. coli strain (only by non-engineered E. coli strains (Clark & Rod,
1987; Zhang et al., 2008; Tseng et al., 2012)), these data support a model where Gcl plays an
important role in modulating the alcohol effect. Both the glxR and glcD mutants, like the rpoS null
strain, also exhibit shortened stationary phase lengths in the absence of added ethanol (Fig. 3.7A),
suggesting that both of these genes contribute to the fitness of untreated stationary phase
populations.

A null mutant strain of the second most highly upregulated gene in the transcriptomic analysis, hyi
(Table 3.3), showed no phenotypic difference when compared to wild-type cultures (data not
shown). Likely, this is due to the fact that, while co-expressed with the glxR and gcl genes in the
same operon, hyi does not act in the glycolate/glyoxylate degradation pathway (Clark & Cronan,
2005).

70
The “alcohol effect” is not a strain-specific phenomenon. It occurs in the vast majority of
laboratory and natural isolate E. coli strains that we have tested (Fig. S3.1) as well as in other
genera associated with humans, including strains of Pseudomonas, Streptococcus, and Klebsiella
(data not shown); interestingly, non-human-associated Vibrio and Shewanella strains show no
effect (data not shown). Ethanol is present in many different natural environments, including the
human gastrointestinal tract (Halsted et al., 1973; Cederbaum, 2012), and multiple different natural
isolates of E. coli strains have ethanol oxidation pathways (Salaspuro et al., 1999). Recent studies
looking at the change in microbial diversity due to alcohol consumption in mouse and human
models show dramatic shifts in microbial diversity profiles (Hartmann et al., 2015; Dubinkina et
al., 2017; Kosnicki et al, 2018). However, little information is availabile distinguishing how
microbes in the gut are affected by alcohols present from bacterial fermentation versus from excess
human host consumption.

In our experimental system, it appears likely that ethanol and other straight-chain alcohols
containing between 2 and 6 carbons may be serving primarily as signaling molecules, ultimately
delaying the onset of death phase in cell populations. Given the structual similarity between
alcohols and glycolate, it is possible that the alcohol is allosterically mimicking glycolate to bind
to the GlcC regulator (Pellicer et al., 1999), causing a deprepression of the glycolate degradation
pathway. This could, in turn, result in an increase in gluconeogenesis, providing additional carbon
and energy sources available through increased scavanging of detrital nutrients during stationary
phase and, thus, delaying death.

71
Despite being one of the best-studied organisms, we still do not understand what triggers
populations of E. coli to die. This work has the potential to shed light on the mechanisms by which
signaling molecules can impact community dynamics. A better grasp of the mechanisms
underlying the “alcohol effect” may also improve our understanding of the interplay between
ethanol and bacteria, whether that be in the laboratory or natural environments like the poorly
understood host-microbe dynamics of the human gastrointestinal tract.
72
3.6 References
Aizenman E, Engelberg-Kulka H, Glaser G. 1996. An Escherichia coli chromosomal
“addiction module” regulated by 3’,5’-bispyrophosphate: a model for programmed
bacterial cell death. Proc Natl Acad Sci 93:6059-6063.
Allocati N, Masulli M, Di Ilio C, De Laurenzi V. 2015. Die for the community: an overview of
programmed cell death in bacteria. Cell Death Dis 6:e1609.
Anders S, McCarthy DJ, Chen Y, Okoniewski M, Smyth GK, Huber W, Robinson MD. 2013.
Count-based differential expression analysis of RNA sequencing data using R and
Bioconductor. Nat Protoc 8:1765-1786.
Anders S, Pyl PT, Huber W. 2015. HTSeq – a Python framework to work with high-
throughput sequencing data. Bioinformatics 31:166-169.
Avrani S, Bolotin E, Katz S, Hershberg R. 2017. Rapid genetic adaptation during the first
four months of survival under resource exhaustion. Mol Biol Evol 34:1758-1769.
Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M,
Wanner BL, Mori H. 2006. Construction of Escherichia coli K-12 in-frame, single-gene
knockout mutants: the Keio collection. Mol Syst Biol 2:2006.0008.
Battesti A, Majdalani N, Gottesman S. 2011. The RpoS-mediated general stress response in
Escherichia coli. Annu Rev Microbiol 65:189–213.
Bayles KW. 2014. Bacterial programmed cell death: making sense of a paradox. Nat Rev
Microbiol 12:63–69.
Bohannon DE, Connell N, Keener J, Tormo A, Espinosa-Urgel M, Zambrano MM, Kolter R. 1991.
Stationary-phase-inducible “gearbox” promoters: differential effects of katF mutations and
role of s
70
. J Bacteriol 173:4482-4492.
73
Cao H, Wei D, Yang Y, Shang Y, Li G, Zhou Y, Ma Q, Xu Y. 2017. Systems-level
understanding of ethanol-induced stresses and adaptation in E. coli. Sci Rep 7:44150.
Cederbaum AI. 2012. Alcohol metabolism. Clin Liver Dis 16:667-685.
Chandra B, Ramisetty M, Raj S, Ghosh D. 2016. Escherichia coli MazEF toxin-antitoxin system
does not mediate programmed cell death. J Basic Microbiol 56:1398-1402.
Clark D, Cronan J. 2005. Two-carbon compounds and fatty acids as carbon sources. EcoSal Plus
2005 doi:10.1128/ecosalplus.3.4.4.
Clark DP, Rod ML. 1987. Regulatory mutations that allow the growth of Escherichia coli
on butanol as carbon source. J Mol Evol 25:151-158.
Cooper GM. 2000. The cell: a molecular approach, 2
nd
ed. Sinauer Associates, Sunderland, MA.
Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in Escherichia
coli K-12 using PCR products. Proc Natl Acad Sci 97:6640-6645.
Dubinkina VB, Tyakht AV, Odintsova VY, Yarygin KS, Kovarsky BA, Pavlenko AV, Ischenko
DS, Popenko AS, Alexeev DG, Taraskina AY, Nasyrova RF, Krupitsky EM, Shalikiani
NV, Bakulin IG, Shcherbakov PL, Skorodumova LO, Larin AK, Kostryukova ES,
Abdulkhakov RA, Abdulkhakov SR, Malanin SY, Ismagilova RK, Grigoryeva TV, Ilina
EN, Govorun VM. 2017. Links of gut microbiota composition with alcohol dependence
syndrome and alcohol liver disease. Microbiome 5:141.
Farrell MJ, Finkel SE. 2003. The growth advantage in stationary-phase phenotype
conferred by rpoS mutations is dependent on the pH and nutrient environment. J
Bacteriol 185:7044-7052.
Finkel SE. 2006. Long-term survival during stationary phase: evolution and the GASP
phenotype. Nat Rev Microbiol 4:113-120.
74
Haft RJF, Keating DH, Schwaegler T, Schwalbach MS, Vinokur J, Tremaine M, Peters JM,
Kotlajich MV, Pohlmann EL, Ong IM, Grass JA, Kiley PJ, Landick R. 2014. Correcting
direct effects of ethanol on translation and transcription machinery confers ethanol
tolerance in bacteria. Proc Natl Acad Sci 111:E2576-E2585.
Halsted CH, Robles EA, Mezey E. 1973. Distribution of ethanol in the human
gastrointestinal tract. Am J Clin Nut 26:831-834.
Hartmann P, Seebauer CT, Schnabl B. 2015. Alcohol liver disease: The gut microbiome and
liver cross talk. Alcohol Clin Exp Res 39:763-775.
Hayden JD, Ades SE. 2008. The extracytoplasmic stress factor, s
E
, is required to maintain cell
envelope integreity in Escherichia coli. PLoS ONE 3:e1573.
Hengge R. 2009. Proteolysis of s
S
(RpoS) and the general stress response in Escherichia coli. Res
Microbiol 160:667–676.
Hengge-Aronis R. 2002. Signal transduction and regulatory mechanisms involved in control of the
s
S
(RpoS) subunit of RNA polymerase. Microbiol Mol Biol Rev 66:373– 395.
Horinouchi T, Tamaoka K, Furusawa C, Ono N, Suzuki S, Hirasawa T, Yomo T, Shimizu H. 2010.
Transcriptome analysis of parallel-evolved Escherichia coli strains under ethanol stress.
BMC Genomics 11:579.
Ingram LO, Vreeland NS. 1980. Differential effects of ethanol and hexanol on the
Escherichia coli cell envelope. J Bacteriol 144:481-488.
Ingram LO. 1976. Adaption of membrane lipids to alcohols. J Bacteriol 125:670-678.
Kohanski MA, Dwyer DJ, Collins JJ. 2010. How antibiotics kill bacteria: from targets to
networks. Nat Rev Microbiol 8:423-435.
75
Kosnicki KL, Penprase JC, Cintora P, Torres PJ, Harris GL, Brasser SM, Kelley ST. 11 May 2018,
posting date. Effects of moderate, voluntary ethanol consumption on the rat and human gut
microbiome. Addict Biol doi:10.1111/adb.12626.
Kraigsley AM, Finkel SE. 2009 Adaptive evolution in single species bacterial biofilms. FEMS
Microbiol Lett 293:135-140.
Kram KE, Finkel SE. 2014. Culture volume and vessel affect long-term survival, mutation
frequency, and oxidative stress of Escherichia coli. Appl Environ Microbiol 80:1732-1738.
Kram KE, Finkel SE. 2015. Rich medium composition affects Escherichia coli survival,
glycation, and mutation frequency during long-term batch culture. Appl Environ
Microbiol 81:4442-4450.
Lange R, Hengge-Aronis R. 1991. Growth phase-regulated expression of bolA and
morphology of stationary-phase Escherichia coli cells are controlled by the novel sigma
factor s
S
. J Bacteriol 173:4474–4481.
Lazar S. 1996. PhD thesis. Harvard University, Cambridge, MA. Survival of Escherichia coli in
stationary phase.
Lee H, Popodi E, Tang H, Foster PL. 2012. Rate and molecular spectrum of spontaneous
mutations in the bacterium Escherichia coli as determined by whole-genome
sequencing. Proc Natl Acad Sci 109:e2774-e2783.
Ly HV, Longo ML. 2004. The influence of short-chain alcohols on interfacial tension,
mechanical properties, area/molecule, and permeability of fluid lipid bilayers. Biophys J
87:1013-1033.
76
Membrillo-Hernández J, Lin EC. 1999. Regulation of expression of the adhE gene,
encoding ethanol oxidoreductase in Escherichia coli: transcription from a downstream
promoter and regulation by fnr and RpoS. J Bacteriol 181:7571-7579.
Navarro Llorens JM, Tormo A, Martínez-García E. Stationary phase in Gram-negative
bacteria. FEMS Microbiol Rev 34:476-495.
Ochman H, Selander RK. 1983. Standard reference strains of Escherichia coli from natural
populations. J Bacteriol 157:690-693.
Pellicer MT, Fernandez C, Badía J, Aguilar J, Lin ECC, Baldomà. 1999. Cross-induction of glc
and ace operons of Escherichia coli attribute to pathway intersection. J Biol Chem
274:1745-1752.
Rendi R, Ochoa S. 1962. Effect of chloramphenicol in cell-free preparations on protein
synthesis of Escherichia coli. J Biol Chem 237:3711-3713.
Rudolph FB, Purich DL., Fromm HJ. 1968. Coenzyme a-linked aldehyde dehydrogenase from
Escherichia coli. J Biol Chem 243:5539-5545.
Salaspuro V, Nyfors S, Heine R, Siitonen A, Salaspuro M, Jousimies-Somer H. 1999.
Ethanol oxidation and acetaldehyde production in vitro by human intestinal strains of
Escherichia coli under aerobic, microaerobic, and anaerobic conditions. Scand J
Gastroenterol 34:967-973.
Shafqat J, Höög J, Hjelmqvist L, Oppermann UCT, Ibáñez C, Jörnvall H. 1999. An ethanol-
inducible MDR ethanol dehydrogenase/acetaldehyde reductase in Escherichia coli:
structural and enzymatic relationships to the eukaryotic protein forms. Eur J Biochem
263:305-311.
77
Si HM, Zhang F, Wu AN, Han RZ, Xu GC, Ni Y. 2016. DNA microarray of global
transcription factor mutant reveals membrane-related proteins involved in n-butanol
tolerance in Escherichia coli. Biotechnol Biofuels 8:114.
Tanouchi Y, Pai A, Buchler NE, You L. 2012. Programmed stress-induced altruistic death in
engineered bacteria. Mol Syst Biol 8:626.
Trapnell C, Pachter L, Salzberg S. 2009. TopHat: discovering splice junctions with RNA-seq.
Bioinformatics 25:1105-1111.
Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, Pimentel H, Salzberg SL, Rinn JL,
Pachter L. 2012. Differential gene and transcript expression analysis of RNA-seq
experiments with TopHat and Cufflinks. Nat Protoc 7:562-578.
Tseng H, Prather KLJ. 2012. Controlled biosynthesis of odd-chain fuels and chemicals via
engineered modular metabolic pathways. Proc Natl Acad Sci 109:17925-17930.
Vulić M, Kolter R. 2002. Alcohol-induced delay of viability loss in stationary phase
cultures of Escherichia coli. J Bacteriol 184:2898-2905.
Wong GT, Bonocora RP, Schep AN, Beeler SM, Lee Fong AJ, Shull LM, Batachari LE, Dillon
M, Evans C, Becker CJ, Bush EC, Hardin J, Wade JT, Stoebel DM. 2017. Genome-wide
transcriptional response to varying RpoS levels in Escherichia coli K-12. J Bacteriol 199:1-
17.
Yomano LP, York SW, Ingram LO. 1998. Isolation and characterization of ethanol-
tolerant mutants of Escherichia coli KO11 for fuel ethanol production. J Ind Microbiol
Biotechnol 20:132–138.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
78
Zhang K, Sawaya MR, Eisenberg DS, Liao JC. 2008. Expanding metabolism for biosynthesis of
nonnatural alcohols. Proc Natl Acad Sci 105:20653-20658.

79
3.7 Supporting Material

Table S3.1 | Significantly upregulated genes one hour after ethanol addition
Gene Synonym(s) Description Fold
Change
p-Value
glxR glxB1, ybbQ Tartronate semialdehyde reductase 2 36.25 7.10e-57
hyi ybbG, gip Hydroxypyruvate isomerase 19.83 1.28e-36
gcl Glyoxylate carboligase 12.05 2.88e-75
glcD yghM, gox Glycolate dehydrogenase, putative FAD-
linked subunit
11.46 2.14e-94
glcE yghL, gox Glycolate dehydrogenase, putative FAD-
linked subunit
9.17 6.83e-50
cysA Sulfate/thiosulfate ABC transporter ATP
binding subunit
4.78 1.65e-34
cysI Sulfite reductase, hemoprotein subunit 4.76 1.30e-20
frmR yaiN DNA-binding transcriptional repressor 4.63 7.96e-49
cysJ Sulfite reductase, flavoprotein subunit
complex
4.30 2.57e-30
asnA Asparagine synthetase A 3.73 3.83e-28
cysD Sulfate adenylyltransferase subunit 2 3.71 9.94e-32
cysH Phosphoadenosine phosphosulfate
reductase
3.69 3.82e-25
frmA adhC S-(hydroxymethyl) glutathione
dehydrogenase, formaldehyde
dehydrogenase
3.67 4.40e-23
allA glxA2, ybbT Ureidoglycolate lyase 3.47 2.67e-07
yciW Putative oxidoreductase 3.42 9.62e-35
yeeD Putative sulfurtransferase 3.11 3.65e-25
cysN Sulfate adenyltransferase subunit 1 3.10 3.22e-27
cysW Sulfate/thiosulfate ABC transporter inner
membrane subunit
3.08 5.32e-18
mgtA atmA, corB,
mgt, mtg
Mg
2+
-importer P-type ATPase 2.96 2.51e-24
cysU Sulfate/thiosulfate ABC transporter inner
membrane subunit
2.95 7.23e-13

80
Table S3.2 | Significantly downregulated genes one hour after ethanol addition
Gene Synonym(s) Description Fold
Change
p-Value
kdpF K
+
-transporting P-type ATPase subunit -51.91 0.000183
elbA iraM, elb1,
ycgW
Anti-adaptor protein, inhibitor of s
S

proteolysis
-4.42 0.00126
ygeL Uncharacterized protein -4.30 0.0121
yadK Putative fimbrial protein -4.18 0.000641
yehC Putative fimbrial chaperone -4.04 0.00234
yhhH PF15631 family protein -3.92 0.00434
ydeM Putative anaerobic sulfatase maturation
enzyme
-3.89 0.000370
ybcQ DLP12 prophage, putative antitermination
protein
-3.86 0.00208
ygiZ Conserved inner membrane protein -3.62 0.00438
ypjC DUF5507 domain-containing protein -3.54 0.00779
ynfO Qin prophage -3.53 0.00105
ydeQ Putative fimbrial adhesin protein -3.48 0.00419
ydcC H repeat-associated putative transposase -3.39 0.000994
bglH yieC Carbohydrate-specific outer membrane
porin, cryptic
-3.36 0.000515
yddL Putative uncharacterized protein -3.33 0.0185
yhaC Uncharacterized protein -3.33 0.00305
yohH mdtQ, yohG Putative multidrug resistance outer
membrane protein
-3.31 0.00779
tdcR DNA-binding transcriptional activator -3.20 0.0191
ydeS Putative fimbrial protein -3.20 0.0181
yiaB Conserved inner membrane protein -3.17 0.0113
ycaK Putative NAD(P)H-dependent
oxidoreductase
-3.16 0.00487
yddK Leucine-rich repeat domain-containing
protein
-3.15 0.0180
agaB yraD Galactosamine-specific PTS enzyme IIB
component
-3.14 0.0156
ygcW Putative deoxygluconate dehydrogenase -3.14 0.0112
yaiS Putative deacetylase -3.11 0.00508
ycjM ggaP Glucosylglycerate phosphorylase -3.10 0.00389
ychS Putative uncharacterized protein -3.09 0.00479
essQ ydfS Qin prophage, putative S lysis protein -3.08 0.00721
81
fixX yaaT Putative ferredoxin -3.08 0.0171
yddJ Uncharacterized protein -3.07 0.00495
yqeK Uncharacterized protein -3.05 0.0105
yafT Lipoprotein -3.03 0.00859
yjfM DUF1190 domain-containing protein -3.00 0.00819
ygeK Putative DNA-binding transcriptional
regulator
-2.98 0.0200
nrfB yjcI Periplasmic nitrite reductase penta-heme
c-type cytochrome
-2.97 0.00347
yafU Putative inner membrane protein -2.97 0.0108
ydfR Qin prophage -2.94 0.0183
ybcO DLP12 prophage, putative nuclease -2.93 0.0362
yhcA Putative fimbrial chaperone -2.92 0.0142
yhhZ Putative endonuclease -2.91 0.00417
sfmF ybcG Putative fimbrial protein -2.89 0.0225
ybbD Putative uncharacterized protein -2.89 0.00623
mokC gefL Regulatory protein -2.88 0.00399
pin argU, dnaY tRNA-Arg(UCU) -2.88 0.00784
yigE DUF2233 domain-containing protein -2.85 0.00122
ynbB Putative CDP-diglyceride synthase -2.84 0.00514
ydeT fmlC Fimbrial usher domain-containing protein -2.83 0.00298

82

6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
Log
10
CFU/ml
Day
ECOR-04 ECOR-13
ECOR-14
ECOR-15
ECOR-28
6
7
8
9
10
0 1 2 3 4 5
ECOR-29
ECOR-37 ECOR-38
A. B.
C. D.
E. F.
G. H.
83

Figure S3.1 | The majority of natural isolate and laboratory strains show the delayed death
effect induced by the addition of ethanol. The effect was identified across multiple different
ECOR (Ochman & Selander, 1984) strains and laboratory strains. Cultures were either left
untreated (squares) or treated (diamonds) on Day 1. (A) ECOR-04; (B) ECOR-13; (C) ECOR-14;
(D) ECOR-15; (E) ECOR-28; (F) ECOR-29; (G) ECOR-37; (H) ECOR-38; (I) ECOR-40; (J)
ECOR-51; (K) ECOR-62; (L) ECOR-63; (M) ECOR-68; (N) ECOR-71; (O) M2; and (P)
BW25113 strains.
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
ECOR-40 ECOR-51
ECOR-62
ECOR-68
ECOR-63
ECOR-71
M2
BW25113
I.
P.
L.
O.
M. N.
J.
K.
Log
10
CFU/ml
Day
84

Figure S3.2 | Single gene null strains corresponding to mutations in glycolate degradation
pathway genes have altered delayed death phenotype. The RNA-sequencing analysis (Table
3.3; Table S3.1; Table S3.2) showed significant upregulation of the glxR, glcD, and gcl genes.
Here, we restructured the data from Figure 7 to make each panel correspond to the growth
physiology of one of the null strains made. Untreated cultures (closed squares), 5 µl (~17.6 mM)
ethanol (open diamonds), and ~17.6 mM 1-butanol (open squares). (A) GlxR null populations; (B)
GlcD null populations; (C) Gcl null populations.

6
7
8
9
10
0 1 2 3 4 5
6
7
8
9
10
0 1 2 3 4 5
Log
10
CFU/ml
6
7
8
9
10
0 1 2 3 4 5
Day
A.
B.
C.
85
Chapter 4: Comprehensive Physiological and Genetic Profiling of Alcohol-Induced Delay of
Bacterial Death of Escherichia coli

4.1 Abstract
Though arguably one of the most well-studied organisms in the laboratory, there are many
components of the life cycle of Escherichia coli that remain poorly understood, particularly how
populations react to complex and stressful environments. Here, we further explore a phenomenon
known as the “alcohol effect,” where the addition of small, sublethal alcohol concentrations to
bacterial cultures causes a dose-dependent delay in entry into death phase. An extension of Chapter
3, the work presented here lays the groundwork for the physiological and genetic parameters
associated with the alcohol effect. We evaluate a variety of phenotypes, including a comparison of
the long-term survival of wild-type and rpoS mutant strains in the presence or absence of ethanol,
as well as the physiological effects of gene candidates associated with the presence or absence of
alcohol. These data, along with that presented in Chapter 3, further support a model in which the
alcohol effect is a widespread, yet complex, phenotype.
86
4.2 Introduction
The concept of alcohol-induced delay of death phase in bacteria remains a somewhat new as well
as seemingly counterintuitive topic. The phenomenon was discovered, as described by Ferraro &
Finkel (2019; Chapter 3) in the 1990s when the Kolter group attempted to determine if there was
a requirement for active protein synthesis (i.e. apoptosis) in moderating the onset and timing of
death phase (Ferraro & Finkel, 2019). Instead, they discovered what is now known as the “alcohol
effect,” a process in which the presence of sublethal, low-doses of short straight-chain alcohols
leads to a delay of entry into death phase is delayed in a dose-dependent manner. Given that this
effect alters the timing of the transition from stationary phase, where populations are still at a high
cell density, to death phase, where the population loses ~99% of viability (1, 7-8), the alcohol
effect has the potential to provide insight the mechanisms modulating the transition to death phase
(Ferraro & Finkel, 2019; Ferraro & Finkel, manuscript in preparation).

To date, there are only two publications that discuss this effect in any detail, Vulić & Kolter (2002)
and Ferraro & Finkel (2019). Vulić and Kolter previously demonstrated that stationary phase can
be prolonged upon the addition of ethanol, 1-propanol, or 1-butanol was added after one day of
incubation under rich media batch culture conditions (Vulić & Kolter, 2002). In our paper, we
extensively expanded on what is known about the alcohol effect. We showed that short, straight-
chain alcohols between two and six carbons in length delay the onset of death phase and that this
effect does not require the activity of the stationary phase-specific sigma factor RpoS or the AdhE
or AdhP alcohol dehydrogenases. Further, for the first time, our work provided evidence that genes
in the glycolate degradation pathway may play a role in causing this effect (Ferraro & Finkel,
2019).
87

This chapter expands on the groundwork resulting in our first publication on this topic (Ferraro &
Finkel, 2019). These experiments provide further insight into the alcohol effect that help define
the physiological and genetic limits of the phenomenon.

4.3 Materials & Methods
Bacterial strains, culture media, and growth conditions. E. coli strains used in this chapter are
listed in Table 4.1, with most experiments performed using E. coli K-12 strain ZK126, derived
from W3110 lineage (Connell et al., 1987; Zambrano et al., 1993). Unless otherwise stated,
cultures were inoculated from frozen 20% glycerol stocks into 5 ml of Luria-Bertani (Lennox)
medium (LB) into 18- by 150-mm borosilicate test tubes (Thermo Fisher) and incubated at 37
°
C
with aeration using TC-7 rolling drums (New Brunswick Scientific, Edison, NJ). Cells from
overnight cultures were then inoculated 1:1000 (vol:vol) into 5 ml of LB and treated with alcohol
(or other component) as described in Chapter 3. Mutant strains were constructed via P1
transduction, as described in Chapter 3. Laboratory isolate strains (Fig. 4.8) were isolated in the
Ehrenreich lab.

Table 4.1 | E. coli strains used in this study
Strain Relevent Genotype Nomenclature Reference
ZK126 W3110 ∆lacU169 tna-2 Wild-type Connell et al., 1987
ZK819 ZK126 rpoS819 RpoS attenuated Zambrano et al., 1993
ZK1000 ZK126 rpoS::Kan RpoS null Bohannon et al., 1991
SF2606 ZK126 frmA::Kan FrmA null This study
SF2607 ZK126 frmR::Kan FrmR null This study
SF2608 ZK126 allR::Kan AllR null This study
SF2609 ZK126 allA::Kan AllA null This study
SF2610 ZK126 hyi::Kan Hyi null This study
SF2611 ZK126 yeeD::Kan YeeD null This study

88
Monitoring cell growth and survival. Viable cell counts were determined by serial dilution at
indicated time points and plating on LB agar (Kraigsley & Finkel, 2009). The limit of detection
for this method of titering is ≥1,000 CFU/ml (Kraigsley & Finkel, 2009). Where indicated, the pH
was measured using 6.0 to 10.0 range pH paper with ~0.3 pH unit increments (EMD Chemicals,
La Jolla, CA).

Mutation frequency analysis. Mutation frequency was determined by measuring spontaneous
rifampicin resistance (Rif
R
) of E. coli ZK126 wild-type strain after 24 hours growth (Corzett et al.,
2013) in LB. Independent 5 ml overnight cultures were grown in their given medium. Total cell
counts were measured on LB agar (described above), and 100 µl of each overnight culture was
plated onto plates containing rifampicin (100 µg/ml). LB rifampicin plates were protected from
light and incubated overnight at 37
°
C. Rif
R
mutation frequency was calculated by dividing the total
number of Rif
R
colony forming units (CFU)/ml by the total CFU/ml. The mutation frequency
distributions were compared using a two-sample Kolmogorov-Smirnov (K-S) test (p < 0.05)
(http://www.physics.csbsju.edu/stats/KS-test.html) and plotted.

Ethanol colorimetric concentration assay. Ethanol concentration in cultures was measured using
a colorimetric assay (BioVision Inc., Milpitas, CA), as described in Chapter 3 (Ferraro & Finkel,
2019), following the manufacturer’s instructions.
89
4.4 Results
Alcohol must be present during stationary phase to exhibit the alcohol effect. To better
understand the timing of when the alcohol-induced death delay (or alcohol effect) occurs, we
treated wild-type cultures with 5 µl of 95% ethanol (~17.6 mM) on either Day 0, Day 1, Day 2, or
Day 3 (Fig. 4.1). Only cultures treated on Day 1 or 2, when the populations are still in stationary
phase, exhibit the alcohol effect. Populations that are pre-stationary phase (Day 0) or post-death
phase (Day 3) show long-term growth and survival patters no different than untreated cells.
However, when greater concentrations of alcohol are added to pre-stationary phase cultures, the
ethanol effect starts to be observed, suggesting that there appears to be enough alcohol present by
the time entry into stationary phase occurs to delay death phase (data not shown).

Figure 4.1 | The “alcohol effect” is a stationary phase-specific phenomenon. 5 µl 95% ethanol
was added on either Day 0 (open diamonds), Day 1 (open circles), Day 2 (open triangles), or Day
3 (closed diamonds). Untreated cultures are indicated by the open squares. Lines signify averages
of triplicate cultures.

Further, the magnitude of the effect appears to be stronger when smaller doses of alcohol are
administered to cultures versus a large one-time dose early in stationary phase (Fig 4.2). A single
5 µl dose on Day 1 delays death for one day. A two-time dose of 5 µl on Day 1 and 2, prolongs
stationary phase for three days versus the one-time 10 µl two-day delay in the transition into death
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5 6 7
CFU/ml
Day
Untreated
Day0
Day1
Day2
Day3
90
phase (Chapter 3). Strikingly, a 15 µl total dose, distributed equally over three days, has an equal
magnitude effect to a single Day 1 dose of ten times that amount.

Figure 4.2 | Multiple small ethanol doses cause a stronger effect than a one-time larger dose.
5 µl 95% ethanol was added either one time (open diamonds) after 24 hours of incubation or added
for one (open circles) or two (open triangles) additional days at the same low dose. A one-time (on
Day 1) 150 µl 95% ethanol dose is indicated by the closed diamonds while the untreated ZK126
wild-type strain is indicated with open squares. Lines signify averages of triplicate cultures.

Extending this concept further, we treated wild-type ZK126 populations with daily small doses (5
µl) of ethanol to examine just how long death phase could be delayed in our system (Fig. 4.3). We
found that death phase could be delayed for upwards of ~20 days with spread out doses of ~100
µl ethanol, much longer than typical ~7 day delay of a single 150 µl dose on Day 1. Much of the
discrepancy could be due to ethanol presence in the medium. Data from Chapter 3 and Figure 4.4
indicate that ethanol depletion from the medium directly factors into the exact timing of death
phase. Surprisingly, these aged treated cultures, after undergoing death phase starting on Day 21,
increase in viability again by the end of the time course to pre-death levels.
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5 6 7 8 9
CFU/ml
Day
WT
5 µl 1x
5 µl 2x
5 µl 3x
150 µl 1x
91

Figure 4.3 | Daily ethanol additions delay death phase for ~20 days. Wild-type cultures were
treated with daily 5 µl 95% ethanol additions (diamonds) starting on Day 1. Untreated cultures are
represented with square markers. Lines denote average of triplicate culture, and error bars
represent standard deviation.

Figure 4.4 | Depletion of ethanol over time in active cultures. Viable cultures left untreated
(squares) or treated with 50 µl (~176 mM) 95% ethanol (diamonds) on Day 1 of incubation. Viable
counts (CFU/ml) are indicated on the primary y-axis, and concentration (mM) of ethanol is
indicated on the secondary y-axis. Lines denote average of triplicate culture, and error bars
represent standard deviation.

1E+06
1E+07
1E+08
1E+09
1E+10
0 5 10 15 20 25 30
CFU/ml
Day
0
20
40
60
80
100
120
140
160
180
200
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5 6 7
CFU/ml
Day
mM
92
The alcohol effect is not caused by pH, acetate, or dividing cells. Given that alcohol is added
to cultures to generate the effect, we sought to identify if a secondary factor, pH, could be the
actual cause of the effect (Fig. 4.5). Here, we added increasing amounts of either ethanol (Fig.
4.5A) or acetate/acetic acid (Fig. 4.5B) and measured long-term survival. Previous studies show
that buffering a culture medium at neutral pH can delay entry into death phase (Farrell & Finkel,
2003). We found that although alcohol treatment can cause a lower pH of the cultures, this is not
the root cause of the effect. This is best exemplified by the 50 µl treated sample (Fig. 4.5A). In this
condition, there is an alcohol effect, but there is not a lowering of pH.
93

Figure 4.5 | Change in pH is not the main cause of the alcohol effect. Monoculture growth of
wild-type where either 95% ethanol (A) or acetate (B) was added and the pH checked. Lines
indicate averages of triplicate cultures. Closed squares designate CFU/ml (primary axis), and open
squares designate pH (secondary axis) of the culture. (A) 95% ethanol was added to Day 1 cultures
and the pH checked daily. Untreated (black lines), 5 µl (~17.6 mM) (blue lines), 50 µl (purple
lines), 150 µl (orange lines), 250 µl (green lines), and 350 µl (red lines). (B) Acetate was added to
wild-type cultures on either Day 1 or Day 2 and the pH checked. Untreated (black lines), Day 1
17.6 mM acetate addition (blue lines), Day 2 17.6 mM acetate addition (purple lines), Day 1 35.2
mM acetate (orange lines), and Day 1 88.0 mM acetate addition (green lines).

Ethanol is present in the medium for longer periods of time when added at higher initial
doses. Another possible cause of the alcohol effect is the idea that alcohol causes the cells to divide
5
6
7
8
9
10
11
12
1E+00
1E+01
1E+02
1E+03
1E+04
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5 6 7 8 9 10 11
4
5
6
7
8
9
10
11
12
1E+03
1E+04
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
CFU/ml
Day
pH
A.
B.
94
while others die, thus maintaining the high viability counts (CFU/ml). To test this, we performed
single treatments with either ethanol or ampicillin or a double treatment with both, and then
determined growth and survival yields (Fig. 4.6). If cell division is the cause of the alcohol effect,
then adding ampicillin to an ethanol-treated culture should prevent the effect from occurring by
killing actively dividing cells. The addition of ampicillin to treated cultures was tested and cultures
still display the alcohol effect, indicating that replacement cell division is not the cause of the
phenotype.

Figure 4.6 | Alcohol effect is not caused by cell division. Wild-type cultures were left untreated
(squares) or treated with ethanol only (diamonds), ampicillin only (circles), or ethanol and
ampicillin (triangles). Lines denote average of triplicate culture, and error bars represent standard
deviation.

The alcohol effect is not correlated with a change in mutation frequency. Further investigating
potential causes of the alcohol effect, we sought to determine if there was a change in mutation
frequency that could be contributing to the phenotype. Using a rifampicin-resistance (Rif
R
)
reporter system, we found that the mutation frequency distributions between untreated and ethanol-
treated cells were indistinguishable (Fig. 4.7).
Untreated
Ethanol only
Ampicillin only
Ethanol +Ampicillin
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5 6
CFU/ml
Day
95

Figure 4.7 | Distribution of Rif
R
mutation frequency due to ethanol addition. The frequency
of spontaneous rifampicin resistance (Rif
R
) for 40 independent cultures of untreated (black lines)
or treated with 50 µl 95% ethanol added on Day 0. Y-axis shows the distribution of mutation
frequencies from lowest to highest.

This effect is present in natural lab isolate strains (Ehrenreich strains). One of the biggest
questions regarding the alcohol effect is how widespread the phenomenon is. Chapter 3 delves into
the topic a bit by examining natural isolate ECOR strains and determining that the majority of E.
coli strains tested exhibit the effect. We took that evaluation further here by testing six different E.
coli strains that were isolated in the Ehrenreich lab (Fig. 4.8). We determined that all six freshly
isolated strains show the alcohol effect, though the exact magnitude and timing of entry into death
phase, even in untreated cultures, varied.
1.00E-09
1.00E-08
1.00E-07
1.00E-06
Rif
R
Frequency
96

Figure 4.8 | Human and canine natural isolate strains exhibit the alcohol effect. One sample
of each natural isolate collection in the Finkel lab (from well 1A) was tested on Day 1 for the
alcohol effect. Untreated samples (squares), 5 µl 95% ethanol (diamonds), 10 µl (circles), and 50
µl (triangles). First letter of strain name represents initial strain donor. (A) I, (B) F, (C) S, (D) J,
(E) O (infant), and (F) D (dog). Lines denote average of three replicates, and error bars indicate
standard deviation.

rpoS mutant strains exhibit altered patterns of entry into death phase, but still exhibit the
alcohol effect. Previous work from this lab (Chapter 3) and Vulić and Kolter demonstrated that
while not essential (Chapter 3), RpoS function plays a role in the magnitude of the alcohol effect
magnitude. Both wild-type and an RpoS-null strain were previously examined where the RpoS-
null strain has a severely diminished, but present effect (Chapter 3). Here, we tested an RpoS
attenuated strain, which has severely diminished RpoS function to determine whether it has a
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5 6 7 8
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5 6 7 8
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5 6 7 8
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5 6 7 8
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5 6 7 8
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5 6 7 8
CFU/ml
I-1A strain
A. B.
C. D.
E. F.
F-1A strain
S-1A strain J-1A strain
O-1A strain D-1A strain
97
moderate effect between that of wild-type and the null strain (Fig. 4.9). In untreated cultures, the
RpoS-attenuated strain enters death phase prior to Day 2 of incubation. Surprisingly, treated RpoS-
attenuated cultures exhibit an alcohol effect of much greater magnuitude than wild-type, entering
death phase one day later than treated wild-type cultures.

Figure 4.9 | RpoS activity affects entry into death phase in treated and untreated cultures.
Wild-type (WT) and RpoS mutant strains (Table 4.1) were either left untreated (solid markers) or
treated with 5 µl 95% ethanol (open markers) after 24 hours incubation. Wild-type (squares), RpoS
attenuated (diamonds), and RpoS null (circles). Lines denote average of three replicates, and error
bars indicate standard deviation.

Table 4.2 | Potential alcohol effect gene candidate descriptions
Gene Synonym(s) Description
frmA adhC S-(hydroxymethyl)glutathione dehydrogenase
frmR yaiN DNA-binding transcriptional repressor
allR glxA3, ybbU DNA-binding transcriptional repressor
allA glxA2, ybbT Ureidoglycolate lyase
hyi ybbG, gip Hydroxypyruvate isomerase
yeeD Putative sulfurtransferase

FrmR may play a role in the ethanol aspect of the alcohol effect. Given the difficulty in
determining genes directly involved in generating the alcohol effect, we performed RNA-
sequencing (Chapter 3 and Appendix A) to help narrow down potential gene candidates. Of the
gene candidates from the transcriptomic analyses (Chapter 3), we found that only two genes are
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
CFU/ml
Day
WT untreated
RpoS attenuated untreated
RpoS null untreated
WT + 5 µl ethanol
RpoS attenuated + 5 µl ethanol
RpoS null + 5 µl ethanol
98
upregulated both one and three hours after addition of ethanol: frmA and frmR. Therefore, we next
created single and double mutants of these two genes in the wild-type background and tested for
the effect (Fig. 4.10). Overall, there seem to be two notable observations from this experiment: 1)
The FrmR mutant, encoding for a DNA-binding transcriptional repressor (Table 4.2), strain (Fig.
4.10B) shows a severely diminished alcohol effect at low doses but not at higher ethanol or
equimolar butanol doses, and 2) In the untreated FrmA mutant (Fig. 4.10A), encoding a glutathione
dehydrogenase, and the double mutant strain (Fig. 4.10C), populations exhibit a day shorter
stationary phase than wild-type. This may indicate that the genes in this pathway are important for
regulating the amount of time cells remain in stationary phase in batch culture.

Figure 4.10 | Growth physiologies of frmA, frmR, and frmAR mutant strains. To determine if
either frmA or frmR plays a role in the alcohol effect, single and double mutant strains were made
via P1 transduction. Untreated (open squares), 5 µl 95% ethanol Day 1 treatment (open diamonds),
50 µl (open circles), 150 µl (open triangles), and 17.6 mM butanol (closed diamonds). (A) FrmA,
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
CFU/ml
Day
0 µl
5 µl
50 µl
150 µl
17.6 mM
Butanol
A.
B.
C.
99
(B) FrmR, (C) FrmA FrmR double mutant. Lines denote average of three replicates, and error bars
indicate standard deviation.

Other gene candidates identified by RNA-seq analysis did not show an altered alcohol effect.
We tested mutants in three additional “+1 hour upregulated genes,” including allA, hyi, and yeeD
and one transcriptional repressor, allR, that regulates the expression of many of the most
upregulated genes in the screen (Fig. 4.11, Table 4.2). The genes AllR regulates include allA, hyi,
and members of the glycolate degradation pathway including gcl, glcD, and glxR (Chapter 3).

Figure 4.11 | RNA-seq candidate genes do not cause the alcohol effect. RNA-sequencing gene
candidate mutant strains were made by transducing via P1 transduction into the wild-type strain.
Mutant strains were either left untreated (squares) or treated on Day 1 with 5 µl 95% ethanol
(diamonds), 50 µl (circles), or 17.6 mM butanol (triangles). (A) AllR, (B) AllA, (C) Hyi, (D) YeeD.
Lines denote average of three replicates, and error bars indicate standard deviation.

Similar to the frmA and frmR mutants, mutants of the genes here have an effect in the presence of
butanol as well as most ethanol concentrations. The only possible minor exception is the Hyi
mutant strain, encoding a hydroxypyruvate isomerase, which exhibits a diminished effect in the
Day
CFU/ml
AllR
A. B.
C. D.
AllA
Hyi YeeD
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
100
presence of 5 µl 95% ethanol treatment. This, however may be misleading as sometimes even in
wild-type, the 5 µl dose can enter death phase slightly before the Day 3 sample is taken.

Addition of glycolate degradation pathway does not trigger the effect. The work done in
Chapter 3 introduced that the glycolate/glyoxylate degradation pathway, whose gene members
include gcl, glxR, glcD, and glcE, may play a role in mediating the alcohol effect. We tested this
further by supplementing wild-type cultures with intermediate chemical substrates of the pathway:
glycolate, glyoxylate, and D-glyceric acid (Fig. 4.12). Addition of none of the compounds yield a
delay of entry into death phase.

101

Figure 4.12 | Supplementing cultures with glycolate degradation pathway substrates do not
affect cell populations. Wild-type cultures were left untreated (squares) or treated on Day 1 with
glycolate (A), glyoxylate (B), or D-glyceric acid (C). 17.6 mM (diamonds), 35.2 mM (circles).
Lines denote average of three replicates, and error bars indicate standard deviation.

Day
0 mM
17.6 mM
35.2 mM
CFU/ml
A.
B.
C.
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4
102
4.5 Discussion
The work presented in this chapter builds on the data presented in Chapter 3 (Ferraro & Finkel,
2019). While the work done in Chapter 3 serves to lay the foundation to the mechanism of the
“alcohol effect,” the data from this chapter more extensively define the genetic and physiological
parameters to yield the effect, providing greater insight to all facets of the alcohol effect.

Here, we show that the alcohol effect is a stationary phase-specific phenomenon (Fig 4.1). If
ethanol is not present in the medium prior to death phase during stationary phase, then the small,
sublethal alcohol doses will have no effect on the viability of the population. The importance of
alcohol presence in the medium appears to be key. This statement is supported by the data
presented in Figures 4.2, 4.3, and 4.4. We show that small, daily doses of alcohol create an effect
of equal magnitude (Fig 4.2, open triangles) to that of a one-time dose of 10x the other amount
(Fig. 4.2, closed diamonds). Extrapolating this though, we see that when maintaining a constant,
but moderate, concentration of alcohol in a culture via daily dosage, we can prolong stationary
phase for at least 20 days (Fig 4.3). Plus, the ethanol concentration measurement experiments from
Chapter 3 and Fig 4.4 exemplify that when ethanol is depleted from the medium, populations soon
thereafter enter death phase, whether it be from smaller doses (Chapter 3) or a larger dose (Fig
4.4). These data combined further reinforce the importance of alcohol presence in the medium to
generate the effect.

We next sought to evaluate potential causes of the alcohol effect. Upon initial examination, there
appear to be two easily testable causes: 1) the addition of alcohol to the cultures creates a change
in culture pH, thus affecting the timing of death phase (Farrell & Finkel, 2003), or 2) alcohol serves
103
as a nutrient causing cell division to occur while another sub-portion of the population dies,
creating the illusion of stationary phase maintenance. The former theory is easily testable by
determining the culture pH in the presence or absence of alcohol. In Figure 4.5, the pH of wild-
type cultures was determined when either ethanol or acetic acid was added to the medium.
Although we do notice that the pH lowers after ethanol addition in the 150 µl and 250 µl treatments
while the viability remains high (Fig 4.5A), it is also notable that this is not the case with lower
dosages where there is still a delayed death effect, signifying that pH is not the cause of the effect.
Similarly, noted by Farrell and Finkel (2003), we identify that addition of small amounts of acid
to the culture medium can temporarily affect viability (Farrell & Finkel, 2003). However, the effect
of acid and alcohol on cultures appear to be different in physiology and mechanism. Altering the
pH can provide a short-term reprieve from the alkalization of the medium and does not appear be
as complex or function to the same magnitude as with alcohol’s effect on population viability.

We tested the second possibility, that there is some level of death/cell division turn around in the
culture, by adding ampicillin to alcohol-treated cultures (Fig. 4.6). Ampicillin, a beta-lactam
antibiotic, prevents cell division by interfering with a bacterium’s ability to form its cell wall
(Kohanski et al., 2010). If cell division was the cause of the effect, then the ampicillin/ethanol-
treated cultures would appear similar to untreated wild-type. Both the ethanol-only and dual-
treated cultures show the effect, signifying that cell division does not cause the alcohol effect.

To determine if cells that were not entering death phase might be novel mutants, we tested the
mutation frequency of the alcohol-treated cultures using a rifampicin resistance (Rif
R
) reporter
system (Corzett et al., 2013). Checking the Rif
R
mutation frequencies of 40 replicate treated and
104
untreated cultures (80 tubes total), we found that there is no difference in mutation frequency
between the cultures, an expected, but still important find.

Of particular concern in our analysis was whether or not the alcohol effect was specific to our
laboratory strain of E. coli or more widespread throughout different strains on E. coli (this chapter
and Chapter 3) and different bacteria genera (Chapter 5). In this Chapter, we tested six different
E. coli strains freshly isolated in the Ehrenreich lab: 4 adult humans, 1 human infant, and 1 adult
canine (Fig. 4.8). All six strains show a delayed death effect, though the timing and magnitude of
the effects are different between each other and from the laboratory wild-type strain, ZK126. What
factors affect a strain’s magnitude and timing of effect could prove a key avenue of study in the
future.

On a related note, another potentially interesting future avenue of study could be further analysis
of the RpoS-attenuated mutant strain’s increased magnitude effect when treated with 5 µl 95%
ethanol. The attenuated strain, ZK819, carries a 46-base pair duplication at the 3’ end of rpoS
(Zambrano et al., 1993). Given ZK819’s severely diminished RpoS function, which serves as a
global regulator of E. coli’s stress response (Wong et al., 2017; Hengge, 2009; Battesti et al., 2011;
Hengge-Aronis, 2002; Lange & Hengge-Aronis, 1991; Membrillo-Hernández & Lin, 1999) as well
as directly or indirectly regulating ~23% of E. coli’s genome (Wong et al., 2017), it initially seems
counterintuitive that the attenuated strain would have such a strong alcohol effect when the treated
RpoS-null strain shows a severely diminished effect (Fig 4.9, Chapter 3). Potentially, there is a
gene in the RpoS regulon, but not the altered RpoS attenuated regulon, that plays a role in the
105
effect. Additional transcriptomic studies will need to be performed to further evaluate this
possibility.

The last avenue of study into the mechanism of the alcohol effect was to focus on potential genetic
causes of the effect. After performing RNA-sequencing in the presence or absence of
ethanol(Chapter 3, Appendix A), we analyzed several candidate genes, constructed mutants via P1
transduction (Chapter 3), and tested for the effect. The first set of genes that we tested were the
frmA and frmR single and double mutants (Fig 4.10, Table 4.2). We focused on these two genes
because of all of the genes upregulated due to ethanol treatment, frmA, a dehydrogenase (Table
4.2), and frmR, a transcriptional regulator of frmA, are the only genes that are positively
differentially expressed at more than one time point (Appendix A). Upon examination of the
physiologies of the single and double frm mutants, we notice that all three strains show an effect
due to ethanol, though there does seem to be a lessened effect in the FrmR mutant strain when 5
µl ethanol is added (Fig. 4.10B). This could mean that FrmR plays a role in ethanol’s role in the
alcohol effect but not other alcohols since there’s no loss of phenotype compared to wild-type
when butanol is added. There is also a possibility that the 5 µl treatment did have an effect on the
FrmR strain, but that this effect was not observed because of the timing of the sampling of the
culture to determine viable counts, which we know can cause misleading results (Chapter 3). Of
note, both the FrmA single mutant (Fig. 4.10A) and the double mutant strains (Fig. 4.10C) exhibit
a shorter length of time in stationary phase, by one full day, compared to wild-type untreated
populations (Fig. 4.1). This potentially suggests the potential importance of FrmA in survival of
populations during stationary phase.

106
Given that the frm mutant strains do not show the alcohol effect when treated with butanol, an
alcohol that cannot be naturally metabolized by E. coli (Clark & Rod, 1987; Zhang et al., 2008;
Tseng & Prather, 2012), we sought to identify other potential gene candidates from the RNA-seq
screen that show a lessened effect with butanol addition. In Figure 4.11, we tested allR (Fig.
4.11A), allA (Fig. 4.11B), hyi (Fig 4.11C), and yeeD (Fig. 4.11D) due to their strong upregulation
one hour post-treatment (Chapter 3, Appendix A). Or, in allR’s case, though the gene didn’t appear
in the RNA-seq screen, it was chosen to test because many of the genes that did show up in the
screen, including glxR, glcD, hyi, and allA, are regulated by allR (Rintoul et al., 2002; Walker et
al., 2006). All four single mutant strains were tested, and we identified that none show an altered
butanol effect, though like frmA, hyi shows a lessened effect with the addition of 5 µl ethanol (Fig.
4.11C).

The data presented in Chapter 3 showed that genes involved in the glycolate/glyoxylate
degradation pathway potentially play a role in the alcohol effect since the Gcl mutant strain
exhibits an altered effect with butanol addition to cultures (Ferraro & Finkel, 2019). We took this
work a step further in this Chapter by evaluating whether supplementation of substrates in the
glycolate degradation pathway could cause an “alcohol effect” themselves. However, this does not
appear to be the case since cultures supplemented with these substrates results in no distinguishable
phenotype compared to wild-type cells (Fig. 4.12).

As shown above, the work performed in this chapter provides an extensive evaluation into the
physiological limits and potential genetic causes of the alcohol effect. Here, we provide
fundamental and novel work that gives researchers a better overall understanding of the nuances
107
of this effect in E. coli. Future work would benefit by delving into whether the alcohol effect is
present in non-E. coli strains (Chapter 5).

108
4.6 References
Battesti A, Majdalani N, Gottesman S. 2011. The RpoS-mediated general stress response in
Escherichia coli. Annu Rev Microbiol 65:189–213.
Bohannon DE, Connell N, Keener J, Tormo A, Espinosa-Urgel M, Zambrano MM, Kolter
R. 1991. Stationary-phase-inducible “gearbox” promoters: differential effects of katF
mutations and role of s
70
. J Bacteriol 173:4482-4492.
Clark DP, Rod ML. 1987. Regulatory mutations that allow the growth of Escherichia coli
on butanol as carbon source. J Mol Evol 25:151-158.
Connell IN, Han Z, Moreno F, Kolter R. 1987. An E. coli promoter induced by the cessation
of growth. Mol Microbiol 1:195-201.
Farrell MJ, Finkel SE. 2003. The growth advantage in stationary-phase phenotype
conferred by rpoS mutations is dependent on the pH and nutrient environment. J
Bacteriol 185:7044-7052.
Finkel SE, Kolter R. 2001. DNA as a nutrient: novel role for bacterial competence gene homologs.
J Bacteriol 183:6288-6293.
Finkel SE. 2006. Long-term survival during stationary phase: evolution and the GASP
phenotype. Nat Rev Microbiol 4:113-120.
Garibyan L, Huang T, Kim M, Wolff E, Nguyen et al. 2003. Use of the rpoB gene to
determine the specificity of base substitution mutation on the Escherichia coli
chromosome. DNA Repair (Amst.) 2:593-608.
Hengge-Aronis R. 2002. Signal transduction and regulatory mechanisms involved in control of the
s
S
(RpoS) subunit of RNA polymerase. Microbiol Mol Biol Rev 66:373–395.
109
Hengge R. 2009. Proteolysis of s
S
(RpoS) and the general stress response in Escherichia coli.
Res Microbiol 160:667–676.
Kohanski MA, Dwyer DJ, Collins JJ. 2010. How antibiotics kill bacteria: from targets to
networks. Nat Rev Microbiol 8:423-435.
Kraigsley AM, Finkel SE. 2009. Adaptive evolution in single species bacterial biofilms. FEMS
Microbiol Lett 293:135-140.
Kram KE, Finkel SE. 2014. Culture volume and vessel affect long-term survival, mutation
frequency, and oxidative stress of Escherichia coli. Appl Environ Microbiol 80:1732-1738.
Kram KE, Finkel SE. 2015. Rich medium composition affects Escherichia coli survival,
glycation, and mutation frequency during long-term batch culture. Appl Environ
Microbiol 81:4442-4450.
Lange R, Hengge-Aronis R. 1991. Growth phase-regulated expression of bolA and
morphology of stationary-phase Escherichia coli cells are controlled by the novel sigma
factor s
S
. J Bacteriol 173:4474–4481.
Lewis K. 2000. Programmed death in bacteria. Microbiol Mol Biol Rev 64:503-514.
Makarova KS, Wolf YI, Koonin EV. 2013. Comparative genomics of defense systems in archaea
and bacteria. Nucleic Acids Res 41:4360-4377.
Navarro Llorens JM, Tormo A, Martínez-García E. 2010. Stationary phase in Gram-
negative bacteria. FEMS Microbiol Rev 34:476-495.
Rintoul MR, Cusa E, Baldomá L, Badia J, Reitzer L, Aguilar J. 2002. Regulation of the
Escherichia coli allantoin regulon: coordinated function of the repressor AllR and the
activator AllS. J Mol Biol 324:599-610.
110
Tseng H, Prather KLJ. 2012. Controlled biosynthesis of odd-chain fuels and chemicals via
engineered modular metabolic pathways. Proc Natl Acad Sci 109:17925-17930.
Walker JR, Altamentova S, Ezersky A, Lorca G, Skarina T, Kudritska M, Ball LJ, Bochkarev A,
Savchenko A. 2006. Structural and biochemical study of effector molecule recognition by
the E. coli glyoxylate and allantoin utilization regulatory protein AllR. J Mol Biol 358:810-
828.
Wong GT, Bonocora RP, Schep AN, Beeler SM, Lee Fong AJ, Shull LM, Batachari LE, Dillon
M, Evans C, Becker CJ, Bush EC, Hardin J, Wade JT, Stoebel DM. 2017. Genome-wide
transcriptional response to varying RpoS levels in Escherichia coli K-12. J Bacteriol 199:1-
17.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zhang K, Sawaya MR, Eisenberg DS, Liao JC. 2008. Expanding metabolism for biosynthesis of
nonnatural alcohols. Proc Natl Acad Sci 105:20653-20658.

111
Chapter 5: The Far Reach of Alcohol: Alcohol-Induced Bacterial Death Delay in Other (Non-
Escherichia coli) Microbes

5.1 Abstract
In previous work, we extensively characterized a mode of delaying death in Escherichia coli, a
phenomenon known as the “alcohol effect,” where the addition of small, sublethal doses of certain
alcohols delay the transition from stationary phase into death phase. Here, we elucidate the growth
physiologies of twenty-two microbes from different genera, comprising both Gram-positive (G+)
and Gram-negative (G-) bacteria from many different isolation origins. Though not all of the
microbes exhibited the alcohol effect, ~59% (13/22), both G+ and G-, of the species tested show
obvious death delay phenotypes when exposed to alcohol. These data identify that the alcohol
effect is not just limited to E. coli, but is present in an extensive variety of genera, further
suggesting the importance of this effect’s presence throughout evolution and providing an exciting
avenue for future study.
112
5.2 Introduction
Using Escherichia coli as an example, after initial outgrowth in rich media (exponential phase),
bacteria enter stationary phase where viability remains high. In typical batch culture media,
populations will then enter death phase, where the majority of the population dies, before the
surviving subpopulation enters long-term stationary phase (LTSP) (1, 7-8). The exact cause of
entry into death phase in bacteria remains a poorly understood (1-2) and understandably difficult
topic to study.

One manner in which research groups are studying how bacterial populations die is by studying a
phenomenon known as the “alcohol effect,” (Vulić & Kolter, 2002; Ferraro & Finkel, 2019) where
when treated with small, sublethal doses of certain alcohols, a population’s entry into death phase
is delayed. To the best of our knowledge, this work has only been studied to any degree in E. coli
(Vulić & Kolter, 2002; Ferraro & Finkel, 2019). In these studies, it was found that when a straight-
chain alcohol between two (ethanol) and six carbons (hexanol) in length is added to E. coli, death
phase is delayed (Ferraro & Finkel, 2019; Chapter 3). This effect is a dose-dependent response and
can cause death delay for as many as ~20 days, as long as ethanol is constantly present in the
medium (Chapter 4). Further, the glycolate/glyoxylate degradation pathway appears to be involved
in generating the effect (Ferraro & Finkel, 2019). Overall, it appears that short-chain alcohols may
be acting as signalling molecules to delay the onset of death phase (Ferraro & Finkel, 2019).

One of the biggest curiosities about this effect is whether the alcohol effect is E. coli-specific or if
other microbes from different genera also display the phenotype. The effect, if localized to E. coli
only, would still be an interesting point of study. If, however, the phenotype appears in different
113
Gram-positive and Gram-negative genera, this potentially opens a fascinating avenue of study
regarding why this phenomenon is present long after the bacteria branched from their last common
ancestor, implying that there may be an evolutionary cause for keeping the genes that cause the
effect in the genome.

In this Chapter, we test twenty-two different species from different Gram classifications and from
a wide variety of genera for the alcohol effect phenotype. Of the twenty-two, thirteen species show
obvious positive phenotypes (has alcohol effect, like Pseudomonas aeruginosa PA-14), while the
rest show no distinguishable effect or are highly sensitive to alcohol toxicity (i.e. Shewanella
oneidensis MR-1 (Myers & Nealson, 1988) and Vibrio harveyi B392 (Reichelt & Baumann,
1973)). This work helps open up avenues of future research regarding the purpose of the alcohol
effect and why it is present in so many species.

5.3 Materials & Methods
Bacterial strains, culture media, and growth conditions. Bacterial strains used in this chapter
are listed in Table 5.1. Unless otherwise stated, cultures were inoculated from frozen 20% glycerol
stocks into 5 ml of LB (Lennox), LB + 100 mM HEPES, or SWC medium (CITE) in the case of
Vibrio harveyi B392 (Reichelt & Baumann, 1973), in 18- by 150-mm borosilicate test tubes
(Thermo Fisher) and incubated at 30
°
C or 37
°
C, depending on standard growth procedures for each
microbe, with aeration using TC-7 rolling drums (New Brunswick Scientific, Edison, NJ). Cells
from overnight cultures were then inoculated 1:1000 (vol:vol) into 5 ml of the aforementioned
media and treated with ethanol (Koptec) as described in Chapter 3. If a microbe light-sensitive,
114
like Shewanella oneidensis MR-1, the test tubes were wrapped in foil and incubated. Strains
Streptococcus R1C4 and Klebsiella R1C5 were isolated by Dr. Namita Shroff of the Finkel lab.

Monitoring cell growth and survival. Viable cell counts were determined by serial dilution at
indicated time points and plating on LB agar (Kraigsley & Finkel, 2009). If light-sensitive, like
Shewanella oneidensis MR-1, the agar plates were covered in foil and incubated. The limit of
detection for this method of titering is ≥1,000 CFU/ml (Kraigsley & Finkel, 2009).
Table 5.1 | List of strains used in this study
Species ATCC
#
Gram
Stain
Incubation
Temp (
°
C)
Isolation
Origin
Reference
Bacillus cereus + 30 Human isolate
- blood
Hoffmaster et al., 2006
Bacillus subtilis 23059 + 37 Soil Kuzma et al., 1995
Enterococcus faecalis + 37 Human isolate Lebreton et al., 2014
Halomonas C2 - 30 Crustal fluids
– North Pond

Halomonas D2 - 30 Crustal fluids
– North Pond

Klebsiella (Enterobacter)
aerogenes
13048 - 37 Intestinal tract
of animals -
sewage
Skerman et al., 1980
Klebsiella pneumoniae - 37
Klebsiella R1C5 - 37 Finkel lab
isolate- Open
Biome

Micrococcus luteus +/- 30 Soil, dust,
water, air,
flora on human
skin
Fleming, 1922
Mycobacterium
smegmatis
+ 37 Water, soil,
food

Proteus mirabilis - 37 Human Warren et al., 1982
Proteus vulgaris - 37 Soil, water,
feces

Pseudomonas aeruginosa
PA-14
- 37 Plant Schroth et al., 1977;
Rahme et al., 1995
Pseudomonas fluorescens - 30 Soil, water Palleroni, 1984
Salmonella enterica
serovar Typhimurium
- 37
115
Serratia marcescens - 37 Pond water
Shewanella oneidensis
MR-1
- 37 Lake Oneida,
NY
Myers & Nealson, 1988
Shigella sonnei - 37
Staphylococcus aureus 25923 + 37 Clinical isolate Boyle et al., 1973
Staphylococcus
epidermidis
14990 + 37 Human - nose Jones et al., 1963
Streptococcus R1C4 + 37 Finkel lab
isolate- Open
Biome

Vibrio harveyi B392 - 30 Marine Reichelt & Baumann,
1973

116
5.4 Results
Both Gram-positive and Gram-negative microbes exhibit the “alcohol effect.” To determine
whether different bacterial species possess the “alcohol effect,” defined as a phenotype where the
onset of death phase in bacteria is delayed in response to the addition of alcohol to cultures
(Chapter 3; Ferraro & Finkel, 2019), we added varying concentrations, ranging from 5 µl (~17.6
mM) to 50 µl (~176 mM) of 95% ethanol, to cultures. Alcohol was added on either Day 1 or Day
2 of incubation, depending on the timing of death phase for that species.

Of the twenty-two strains tested, thirteen of those (~59%) exhibit an obvious alcohol effect that
fall into one of two phenotypic categories (Fig. 5.1; Table 5.2). A number of strains, like
Pseudomonas aeruginosa PA-14 (Fig. 5.1A) (Schroth et al., 1977; Rahme et al., 1995) or non-
laboratory clinical isolate Streptococcus R1C4 (isolated courtesy of Namita Shroff of the Finkel
lab) fall into the first category of positive alcohol effect in that, like E. coli (Ferraro & Finkel,
2019), they show a level of dose-dependency in the magnitude of response to alcohol. Other
strains, like Proteus vulgaris (Fig 5.1C) or Staphylococcus aureus (data not shown) (Boyle et al.,
1973), do show an alcohol effect, but this phenotype is not dose-dependent. With P. vulgaris, both
the 10 µl and 25 µl ethanol treatments exhibit a nearly identical-magnitude 1-day delay in the onset
of death phase compared to untreated cultures.
117

Figure 5.1 | Many strains exhibit the alcohol effect. Of the twenty-two non-Escherichia coli
strains tested, thirteen exhibited noticeable delays in the onset of death phase. Of those thirteen,
however, strains showed differing magnitudes of effect. Populations were grown in 5 ml LB media
and treated with various concentrations of 95% ethanol on the day indicated. (A) Pseudomonas
aeruginosa PA-14 treated with ethanol addition on Day 2, (B) Streptococcus R1C4 (isolated
courtesy of Namita Shroff of the Finkel lab) treated with ethanol addition on Day 1, (C) Proteus
vulgaris treated on Day 1. Untreated (open squares), 5 µl (open diamonds), 10 µl (open circles),
25 µl (open triangles), and 50 µl (closed diamonds). Lines represent averages of duplicate cultures,
and error bars denote standard deviation.

Some microbes show no or negative effect with alcohol addition. Of the nine bacterial species
that do not exhibit the alcohol effect, there were three distinguishable growth phenotypes observed
(Fig. 5.2; Table 5.2). The first phenotype is when the alcohol-treated populations enter death phase
CFU/ml
Day
0 µl
5 µl
10 µl
25 µl
50 µl
A.
B.
C.
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5
P. aeruginosaPA14
StreptococcusR1C4
P . vulgaris
118
almost identically to untreated cultures, like Bacillus subtilis (Kuzma et al., 1995). In these
populations, their growth phenotypes don’t indicate any level of stress due to sublethal, small
alcohol doses, varying largely from the second category (Fig. 5.2B). This group consists solely of
Vibrio harveyi B392 (Fig. 5.2B) (Reichelt & Baumann, 1973) and Shewanella oneidensis MR-1
(data not shown) (Myers & Nealson, 1988) where these populations not only show no alcohol
effect, but they are actually acutely sensitive to any addition of alcohol. In fact, V. harveyi appears
to show a dose-dependent sensitivity to ethanol.

CFU/ml
Day
0 µl
5 µl
10 µl
25 µl
50 µl
A.
B.
C.
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+04
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+04
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
B. subtilis
V . harveyiB392
E. faecalis
119
Figure 5.2 | There are species that do not exhibit the alcohol effect. Populations were grown in
5 ml of the medium indicated and treated with various concentrations of 95% ethanol on Day 1.
(A) Bacillus subtilis grown in LB + 100 mM HEPES, (B) Vibrio harveyi B392 grown in SWC
medium, (C) Enterococcus faecalis grown in LB. Untreated (open squares), 5 µl (open diamonds),
10 µl (open circles), 25 µl (open triangles), and 50 µl (closed diamonds). Lines represent averages
of duplicate cultures, and error bars denote standard deviation.

A small subgroup of species, comprised of Bacillus cereus (Hoffmaster et al., 2006), Enterococcus
faecalis (Fig. 5.2C) (Lebreton et al., 2014), and Micrococcus luteus (Fig. 5.3B) (Fleming, 1922),
did not enter death phase during the duration of the experiment (Fig. 5.2C), making it impossible
to identify if there is an alcohol effect.

CFU/ml
Day
0 µl
5 µl
10 µl
25 µl
50 µl
A.
B.
C.
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
0 1 2 3 4 5
1E+05
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5
1E+06
1E+07
1E+08
1E+09
1E+10
1E+11
0 1 2 3 4 5
KlebsiellaR1C5
M. luteus
S. marcescens
120
Figure 5.3 | Some species do not show an alcohol effect but exhibit other phenotypes in
response to alcohol addition. Populations were grown in 5 ml LB media and treated with various
concentrations of 95% ethanol on the day indicated. (A) Klebsiella R1C5 (isolated courtesy of
Namita Shroff of the Finkel lab) treated with ethanol addition on Day 2, (B) Micrococcus luteus
treated with ethanol addition on Day 1, (C) Serratia marcescens treated on Day 1. Untreated (open
squares), 5 µl (open diamonds), 10 µl (open circles), 25 µl (open triangles), and 50 µl (closed
diamonds). Lines represent averages of duplicate cultures, and error bars denote standard
deviation.

Certain species do not exhibit the alcohol effect, but they do show phenotypic difference when
alcohol is added. Working with bacteria from different genera, we not only identify those with an
alcohol effect from those without one. Here, we also identify species that do not possess the alcohol
effect in the classic definition, but these strains do exhibit clear, differing phenotypes due to
alcohol presence in the media (Fig. 5.3). In some cases, like Klebsiella (Fig. 5.3A), alcohol addition
on Day 2, when populations are beginning to enter death phase, both stops cells from dying further
and at higher doses, causes an increase in viable cell count. This increase in colony forming units
(CFU) per ml is also observed in M. luteus (Fig. 5.3B) (Fleming, 1922), In this case, however,
there is no death phase with these populations, only additional growth caused by alcohol addition.
Further, like the first two species mentioned, Serratia marcescens (Fig. 5.3C) shows population
growth with the 10 µl (~35.2 mM) dose of alcohol in a 5 ml culture. Though here, we observe
multi log-fold amounts of growth in the span of two days, bringing these treated replicates to
stationary phase levels of cell viability counts.

Table 5.2 | List of species and overall alcohol effect
Species Gram
Stain
LB LB + 100
mM HEPES
Other
Medium
Overall
Effect
Bacillus cereus + - ?
Bacillus subtilis + - - -
Enterococcus faecalis + ? ? ?
Halomonas C2 - + +
Halomonas D2 - + +
121
Klebsiella (Enterobacter)
aerogenes
- + + +
Klebsiella pneumoniae - + + +
Klebsiella R1C5 - + +
Micrococcus luteus +/- ? ?
Mycobacterium smegmatis + - -
Proteus mirabilis - + + +
Proteus vulgaris - + - +
Pseudomonas aeruginosa PA-14 - + +
Pseudomonas fluorescens - + +
Salmonella enterica serovar
Typhimurium
- + +
Serratia marcescens - - -
Shewanella oneidensis MR-1* - - -*
Shigella sonnei - - - -
Staphylococcus aureus + - + +
Staphylococcus epidermidis + - + +
Streptococcus R1C4 + + +
Vibrio harveyi B392* - (SWC) - -*
+ Denotes that species exhibits an alcohol effect.
- Denotes that species does not exhibit an alcohol effect.
* Denotes that species has not alcohol effect and is sensitive to alcohol.
? Denotes that species did not die during study, so unable to identify if exhibit alcohol effect.

122
5.5 Discussion
Prior to this study, only two published studies exist, to the best of our knowledge, that describe the
“alcohol effect,” a phenomenon defined as a population-level delay in the onset of death phase
when treated with sublethal, small doses of certain short, straight-chain n-alcohols (Vulić & Kolter,
2002; Ferraro & Finkel, 2019); both studies only characterize the effect in Escherichia coli in any
detail. Here, we show that the alcohol effect is not an E. coli-specific phenomenon. In fact, the
majority (thirteen) of the twenty-two species tested, from a variety of genera, exhibit the alcohol
effect to some magnitude. The bacterial species tested in this study are both Gram-negative or
Gram-positive and come from many different isolation points of origin, suggesting that the growth
and survival requirements of the strains themselves may vary significantly. Of those species tested,
we were able to categorize each into one of three phenotypic categories: (i) exhibits effect in a
similar manner to E. coli, (ii) no effect distinguishable, or (iii) may or may not show classic
definition of alcohol effect but also has some altered population response to alcohol.

Pseudomonas aeruginosa PA-14 (Fig 5.1A) (Schroth et al., 1977; Rahme et al., 1995),
Staphylococcus epidermidis (Jones et al., 1963), and non-laboratory isolate Streptococcus R1C4
(Fig. 5.1B) represent the bacteria that comprise the first category (Fig. 5.1; Table 5.1).
Pseudomonas aeruginosa PA-14 (Fig 5.1A) and Streptococcus R1C4 (Fig. 5.1B), like E. coli,
exhibit a dose-dependent-like effect due to alcohol. Notably, though the effect of each increases
in magnitude with increased small doses of alcohol, the length of delay in death phase in each of
these strains differs between the laboratory E. coli strain (Chapter 3; Ferraro & Finkel, 2019). For
example, when dosed with 50 µl 95% ethanol, Streptococcus R1C4 only experiences a ~2 day
death delay whereas wild-type E. coli experiences a ~7 day delay (Chapter 3; Ferraro & Finkel,
123
2019). P. aeruginosa exhibits a greater magnitude effect (>3 days) compared to Streptococcus
R1C4 (Fig 5.1A-B). Meanwhile, species like Proteus vulgaris do show an alcohol effect, though
there seems to be no dose-dependency (Fig 5.1C). With these species, 10 µl (~35.2 mM) and 25
µl (~88 mM) 95% ethanol doses cause a population effect of the same magnitude, very different
from E. coli. The fact that ~59% of bacteria tested possess a delayed death phenotype, though to
varying degrees, begs the question regarding what these microbes genetically have in common
even each other that allow for the effect.

The second major phenotypic class identified in this study is comprised of the nine other microbes
(Fig. 5.2; Table 5.2). The alcohol-treated microbes in this group either: (i) have no distinguishable
phenotype from wild-type (Fig. 5.2A) or (ii) are negatively affected by alcohol (Fig 5.2B). Vibrio
harveyi B392 (Fig. 5.2B) (Reichelt & Baumann, 1973) and Shewanella oneidensis MR-1 (data not
shown) (Myers & Nealson, 1988) are notable because they not only do not have an effect, but they
are sensitive to any concentration of alcohol added to cultures. Both microbes are from aquatic
environments where there is likely little-to-no alcohol ever present (Beale et al., 2010), potentially
signifying that marine microbes are less likely to possess the alcohol effect phenotype and are
more likely to be sensitive to alcohol additions. However, this is not the case with all marine-
isolated microbes. Halomonas C2 and Halomonas D2, both of which were isolated from crustal
fluids in the North Pond drilling station in the Atlantic Ocean (Russel et al., 2016) do exhibit a
mild positive alcohol effect (Table 5.2). More aquatic species may need to be tested for the
phenotype before any definitive correlations can be made.

124
Comprising the last phenotypic class, a number of microbes may or may not exhibit the alcohol
effect in the classic definition (delay of the onset of death phase), but show some manner of altered
growth phenotype due to ethanol treatment. Death phase of Klebsiella R1C5 (Fig 5.3A) is stopped
when alcohol is added. In fact, each species in Figure 5.3 shows some degree of growth due to
alcohol treatment. In these cases, it appears that small, sublethal alcohol doses enable cell division,
though to varying degrees and at different times in the population “life” cycle, signaling that
ethanol may be acting as food for these microbes.

The microbes in each class are not all from the same points of origins. For example, of the microbes
that have a positive effect, P. aeruginosa PA-14 was isolated as a plant pathogen (Mathee, 2017).
Streptococcus R1C4, meanwhile, was isolated from human fecal samples (OpenBiome,
Cambridge, MA). Both strains exhibit a dose-dependent effect. Still, if possible, future work for
each class of phenotype (Table 5.2) should be performed in the future to first sequence the species
to identify why some have the effect and others don’t and then elucidate whether any point of
origin commonalities exist between the classes. This is currently difficult as many of the current
species sample set do not have indicated ATCC catalog number or other indication regarding point
of origin or potential genome size. This confusion potentially causes trouble categorizing groups
because strains like M. luteus (Fleming, 1922) or P. aeruginosa can be isolated from a variety of
location types (Mathee et al., 2007).

Our work here opens up an exciting line of questioning that warrant further examination. In
particular, future work may provide answers regarding why this seemingly unusual phenotype
exists in so many microbes from different genera. Knowledge of the species’ genome will also
125
help identify if genes involved in glycolate degradation are present or absent, thus potentially
indicating the presence or absence of the alcohol effect. Future studies may also be able to
distinguish an evolutionary advantage for keeping this reaction, or, potentially, a reason to
eliminate genes that cause the ability to generate the phenotype from their crowded genomes.
Further, increased sample sets with knowledge of origins of isolation may give statistical power to
distinguish potential correlations between origin and presence (or not) of the alcohol effect
phenotype.

126
5.6 References
Beale R, Liss PS, Nightingale PD. 2010. First oceanic measurements of ethanol and
propanol. Geophy Res Lett 37:L24607.
Boyle VJ, Fancher ME, Ross RW Jr. 1973. Rapid, modified Kirby-Bauer susceptibility test
with single, high-concentration antimicrobial disks. Antimicrob Agents Chemother 3:418-
424.
Farrell MJ, Finkel SE. 2003. The growth advantage in stationary-phase phenotype
conferred by rpoS mutations is dependent on the pH and nutrient environment. J
Bacteriol 185:7044-7052.
Ferraro CM, Finkel SE. 2019. Physiological, Genetic, and Transcriptomic Analysis of
Alcohol-Induced Delay of Escherichia coli Death. Appl Environ Microbiol 85:e02113-18.
Finkel SE. 2006. Long-term survival during stationary phase: evolution and the GASP
phenotype. Nat Rev Microbiol 4:113-120.
Fleming A. 1922. On a remarkable bacteriolytic substance found in secretions and tissues.
Proc R Soc Lond Ser B 93:306-317.
Hoffmaster A, Hill K, Gee J, Marston C, De B, Popovic T, Sue D, Wilkins P, Avashia S,
Drumgoole R, Helma C, Ticknor L, Okinaka R, Jackson J. 2006. Characterization of
Bacillus cereus isolates associated with fatal pneumonias: strains are closely related to
Bacillus anthracis and harbor B. anthracis virulence. J Clin Microbiol 44:3352-3360.
Jones D, Deibel RH, Niven CF Jr. 1963. Identity of Staphylococcus epidermidis. J
Bacteriol 85:62-67.
Kraigsley AM, Finkel SE. 2009. Adaptive evolution in single species bacterial biofilms. FEMS
Microbiol Lett 293:135-140.
127
Kram KE, Finkel SE. 2014. Culture volume and vessel affect long-term survival, mutation
frequency, and oxidative stress of Escherichia coli. Appl Environ Microbiol 80:1732-1738.
Kram KE, Finkel SE. 2015. Rich medium composition affects Escherichia coli survival,
glycation, and mutation frequency during long-term batch culture. Appl Environ
Microbiol 81:4442-4450.
Kuzma J, Nemecek-Marshall M, Pollock WH, Fall R. 1995. Bacteria produce the volatile
hydrocarbon isoprene. Curr Microbiol 30:97-103.
Lebreton F, Willems RJL, Gilmore MS, Lebreton F, Willems RJL. 2014. Enterococcus
diversity, origins in nature, and gut colonization. In Gilmore MS, Clewell DB, Ike Y,
Shankar N (ed), Enterococci: from commensals to leading causes of drug resistant
infection. Massachusetts Eye and Ear Infirmary, Boston, MA.
Lewis K. 2000. Programmed death in bacteria. Microbiol Mol Biol Rev 64:503-514. Makarova
KS, Wolf YI, Koonin EV. 2013. Comparative genomics of defense systems in archaea and
bacteria. Nucleic Acids Res 41:4360-4377.
Mathee K, Narasimhan G, Valdes C, Qiu X, Matewish JM, Koehrsen M, Rokas A, Yandava CN,
Engels R, Zeng E, Olavarietta R, Doud M, Smith RS, Montomery P, White JR, Godfrey
PA, Kodira C, Birren B, Galagan JE, Lory S. 2007. Dynamic of Pseudomonas aeruginosa
genome evolution. Proc Natl Acad Sci 105:3100-3105.
Mathee K. 2017. Forensic investigation into the origin of Pseudomonas aeruginosa PA14 – old
but not lost. J Med Microbiol 67:1019-1021.
Myers CR, Nealson KH. 1988 Bacterial manganese reduction and growth with manganese oxide
as the sole electron acceptor. Science 240:1319-1321.
128
Navarro Llorens JM, Tormo A, Martínez-García E. 2010. Stationary phase in Gram-
negative bacteria. FEMS Microbiol Rev 34:476-495.
Palleroni NJ. 1984. Bergey’s Manual of Systematic Bacteriology. In Krieg NR, Holt JG (ed),
Baltimore: The Williams and Wilkens Co. pp. 141-199.
Rahme LG, Stevens EJ, Wolfort SF, Shao J, Tompkins RG, Ausubel FM. 1995. Common
virulence factors for bacterial pathogenicity in plants and animals. Science 268:1899-1902.
Reichelt JL, Baumann P. 1973. Taxonomy of the marine, luminous bacteria. P Archiv
Mikrobiol 94:283.
Russell JA, León-Zayas R, Wrighton K, Biddle JF. 2016. Deep subsurface life from North
Pond: Enrichement, isolation, characterization and genomes of heterotrophic bacteria.
Front Microbiol 6:678.
Schroth MN, Cho JJ, Green SK, Kominos SD. 1977. Epidemiology of Pseudomonas
aeruginosa in agricultural areas. In Young VM (editor) Pseudomonas aeruginosa:
Ecological Aspects and Patient Colonization. New York: Raven Press. pp. 1-29.
Skerman VB et al., 1980. Approved lists of bacterial names. Int J Syst Bacteriol 30:225-420.
Warren JW, Tenney JH, Hoopes JM, Kass EH. 1982. A prospective microbiologic study of
bacteriuria in patients with chronic indwelling urethral catheters. J Infect Dis 146:719-723.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R. 1993. Microbial competition:
Escherichia coli mutants that take over stationary phase cultures. Science 259:1757-1760.

129
Chapter 6: Concluding Remarks

6.1 Conclusion and broader impact
How genetic diversity occurs to enable evolution in response to environmental changes is an active
area of scientific study. Similarly, elucidating the mechanism(s) of how bacterial death occurs,
which allows for the takeover of beneficial mutants in a population, is crucial to better
understanding bacterial survival mechanism. Advancing our knowledge in these areas has the
potential to help better understanding of topics such as antibiotic resistance and long-term bacterial
survival in extreme conditions (Grimm et al., 2003; Hartman & Richardson, 2013; Lever et al.,
2015). This work has sought to address these questions:

1. How does environment affect genetic diversity in Escherichia coli?
Chapter 2 addresses this question by measuring the mutation frequencies and spectra of E. coli
using a well-established rifampicin-resistance (Rif
R
) reporter system (Corzett et al., 2013). Using
this assay, this work found that when grown in four different types of rich media: LB, YT, TB,
and SB, E. coli have different mutation frequencies, with LB and YT cultures having the highest
mutation frequencies (Fig 2.2). A mutation spectrum analysis reveals novel trends for each
environment (Fig 2.3), and, therefore, we are potentially able to relate some of these trends to
known mutation signatures of alternative DNA polymerases. Overall, the work performed in this
chapter provides novel data regarding different environment’s effect on the generation of genetic
diversity in microbial populations.

2. What is the potential mechanism for alcohol-induced death delay in E. coli?
130
Chapters 3 and 4 both elucidate the parameters and mechanism behind the “alcohol effect,” where
alcohol addition to E. coli cultures results in a delay in death phase phenotype. For the first time,
we found that addition of small, sublethal doses of straight-chain n-alcohols between two and six
carbons in length (Fig 3.4; Table 3.2) delay the onset of death phase in a dose-dependent manner
(Fig 3.1). The alcohol effect is stationary phase-specific (Fig 4.2), and alcohol depletion from the
medium precedes the onset of death phase (Fig 3.2). Similarly, when an above-threshold
concentration of alcohol is maintained in the cultures, populations can survive for approximately
~20 days longer in stationary phase than in under untreated conditions (Fig 4.3). Further, neither
an exacerbated stress response nor alcohol catabolism cause the delayed death phenotype since
both an RpoS null mutant strain (Fig 3.5) and a mutant strain that does not express either alcohol
dehydrogenase (Fig 3.6) both still exhibit alcohol effect.

After performing transcriptomic (RNA-sequencing) analyses to identify potential genes associated
with causing the alcohol effect, we found that three of the most highly upregulated genes are part
of the glycolate degradation pathway. Further, a Gcl mutant strain shows a reduced alcohol effect
compared to wild-type when butanol is added to cultures (Fig. 3.7C). This is particularly important
since butanol cannot be metabolized by E. coli as a source of carbon or energy. From this work,
we propose a model in which alcohol, which is structurally similar to glycolate, may be
allosterically mimicking glycolate to bind the GlcC regulator (Pellicer et al., 1999), causing a
derepression of the glycolate degradation pathway. In turn, this may result in an increase in
gluconeogenesis and provide additional energy and nutrient sources available through scavenging
of detrital nutrients during stationary phase to then delay death phase.

131
3. Is the alcohol effect present in non-E. coli microbes?
While the work performed in Chapter 3 (Ferraro & Finkel, 2019) showed that the majority of other
E. coli strains tested, both laboratory and non-laboratory isolates, exhibit the alcohol effect (Fig.
S3.1), no previous publications revealed whether or not microbes from different genera also exhibit
the effect. The work done in Chapter 5 evaluated twenty-two different microbes from different
genera and isolation locations (Table 5.1) and found that ~59% of those tested show a positive
alcohol effect (Fig. 5.1). Of those that not only show no alcohol effect but are sensitive to the effect
(Table 5.2), both species (S. oneidensis MR-1 and V. harveyi B392) are from aquatic environments
(Table 5.1) where ethanol is typically never present (Beale et al., 2010).

6.2 Future directions
The work performed here provides a gateway to studying bacterial death through thorough
examination of the alcohol effect phenotype in E. coli. Even so, there still remain a number of
questions regarding the complexity of the alcohol effect. Specifically, a number of experiments
are still required to support the model posited by us in Chapter 3 in which alcohol may be
allosterically mimicking glycolate to bind the GlcC regulator (Pellicer et al., 1999) to derepress
the glycolate degradation pathway, thus causing an increase in gluconeogenesis and enabling
additional scavenging of detrital energy and nutrients to cause the alcohol effect. Testing a GlcC
null mutant for the alcohol effect would be one means to effectively test this model. Further, the
transcriptomic experiments performed here have provide a tremendous amount of data from
several different time points post-treatment. A surprisingly high number of the differentially
expressed genes, particularly at the later time points, are genes of unknown and uncharacterized
function. Potentially, given the apparent genotypic complexity of the alcohol effect, there is a
132
possibility that some of these uncharacterized genes may contribute to generation of this
phenotype.

Another exciting avenue for future study regards examination of a sample set of other microbes to
determine which exhibit the alcohol effect. For a truly effective sample set, the microbes tested
would include those: (1) of known species or strain, (2) with sequenced genomes, and (3) that have
documented isolation origin locations. These three factors combined would enable us, assuming
the sample size was sufficiently high for statistical significance calculations, to elucidate potential
trends regarding which bacteria, and from which origins, show the effect and why.

Finally, some of the earliest work in this thesis evaluates the effect environment has on mutation
frequency and spectrum and how alternative DNA polymerases II, IV, and V may play a role in
generating this diversity. An obvious and essential follow-up experiment in the future would be to
subject polymerase mutant strains, single, double, and triple, to the four different media conditions
used here and determine if any of the mutation type peaks are diminished when a given polymerase
is no longer present in the cell population.
133
6.3 References
Beale R, Liss PS, Nightingale PD. 2010. First oceanic measurements of ethanol and
propanol. Geophy Res Lett 37:L24607.
Corzett CH, Goodman MF, Finkel SE. 2013. Competitive fitness during feast and famine: how
SOS DNA polymerases influence physiology and evolution in Escherichia coli. Genetics
194:409-420.
Grimm NB, Gergel SE, McDowell WH, Boyer EW, Dent CL, Groffman P, Hart SC, Harvey J,
Johnston C, Mayorga E, McClain ME, Pinay G. 2003. Merging aquatic and terrestrial
perspectives of nutrient biogeochemistry. Oecologia 137: 485-501.
Hartman WH, Richardson CJ. 2013. Differential nutrient limitation of soil microbial
biomass and metabolic quotients (qCO2): is there a biological stoichiometry of soil
microbes? PLoS One 8:e57127.
Lever MA, Rogers KL, Lloyd KG, Overmann J, Schink B, Thauer RK, Hoehler TM,
Jorgensen BB. 2015. Life under extreme energy limitation: a synthesis of laboratory – and
field-based investigations. FEMS Microbiol Rev 5:688-728.
Pellicer MT, Fernandez C, Badía J, Aguilar J, Lin ECC, Baldomà. 1999. Cross-induction of glc
and ace operons of Escherichia coli attribute to pathway intersection. J Biol Chem
274:1745-1752.

134
Appendix A: RNA-sequencing and Whole Transcriptome Analysis of Alcohol-Induced Delay
of Escherichia coli Death

A.1 Brief Comments and Experimental Design
To identify potential gene candidates involved in generating the alcohol effect, a transcriptomic
experiment (RNA-sequencing) was performed. Here, duplicate wild-type ZK126 cultures were
treated after 24 hours growth in Luria-Bertani (LB) medium and total RNA was collected 1 (25
hour timepoint), 3 (28 hour timepoint), and 24 hours (48 hour timepoint) after treatment. Sample
collection and data analysis is described in Chapter 3.

Below, I have included a cursory analysis of differentially expressed genes, through pairwise
comparison, from each time point with a p-value of less than or equal to 0.05 and a fold change of
at least three-fold. Red lines indicate treated replicates, and black/gray lines indicate untreated
replicates. Genes are represented on the x-axis, and transcripts per million (TPM; calculations
described in Chapter 3) are represented on the y-axis.
135

Figure A.1 | Significantly upregulated genes 1 hour after ethanol addition.
0
100
200
300
400
500
allA asnA cysA cysD cysH cysI cysJ cysN cysU cysW frmA frmR gcl glcD glcE glxR hyi mgtA yciW yeeD
genes
tpm
cond
tpm_25Ta
tpm_25Tb
tpm_25Ua
tpm_25Ub
25hr Treated vs. Untreated LogFC > 1.5
136

Figure A.2 | Significantly downregulated genes 1 hour after ethanol addition.
0
10
20
agaB bglH elbA essQ fixX kdpF mokC nrfB pin sfmF tdcR yadK yafT yafU yaiS ybbD ybcO ybcQ ycaK ychS ycjM ydcC yddJ yddK yddL ydeM ydeQ ydeS ydeT ydfR yehC ygcW ygeK ygeL ygiZ yhaC yhcA yhhH yhhZ yiaB yigE yjfM ykiB ynbB ynfO yohH ypjC yqeK
genes
tpm
cond
tpm_25Ta
tpm_25Tb
tpm_25Ua
tpm_25Ub
25hr Treated vs. Untreated LogFC < -1.5
137

Figure A.3 | Significantly upregulated genes 3 hours after ethanol addition.

0
200
400
600
argA argB argC argD argG argH argI artJ cysA fimA frmA frmR gadE hdeA hdeB lsrG ybdR ydiH
genes
tpm
cond
tpm_28Ta
tpm_28Tb
tpm_28Ua
tpm_28Ub
28hr Treated vs. Untreated Log2FC > 1.5
138

Figure A.4 | Significantly downregulated genes 3 hours after alcohol addition.

0
10000
20000
astA astC cpxP cutC fxsA htpX ibpB mmuP napD phnG yceP ychH yebE yfdM yliH yqgD
genes
tpm
cond
tpm_28Ta
tpm_28Tb
tpm_28Ua
tpm_28Ub
28hr Treated vs. Untreated Log2FC < -1.5
139

Figure A.5 | Significantly upregulated genes 24 hours after alcohol addition.
0
20000
40000
60000
asd dapB dksA gcl glxR lysC otsB pheLphnB potF rplB rplD rplP rplV rpmArpmC rpsC rpsQ rpsS sfsA ybgT yccJ yciG yeeDygaMyhbEyhdV yiaG yibT yjbJ yliH ymdFymgBymgEyodCyohC ytjA
genes
tpm
cond
tpm_48Ta
tpm_48Tb
tpm_48Ua
tpm_48Ub
Novel 48hr Treated vs. Untreated Log2FC > 2.5
140

Figure A.6 | Significantly downregulated genes 24 hours after alcohol addition.
0
2500
5000
7500
10000
12500
asd dapB dksA gcl glxR lysC otsB pheL phnB potF rplB rplD rplP rplV rpmArpmC rpsC rpsQ rpsS sfsA ybgT yccJ yciG yeeDygaMyhbE yhdV yiaG yibT yjbJ yliH ymgBymgEyodCyohC ytjA
genes
tpm
cond
tpm_48Ta
tpm_48Tb
tpm_48Ua
tpm_48Ub
Novel 48hr Treated vs. Untreated Log2FC > 2.5

Abstract (if available)

Abstract Both genetic diversity and bacterial death play large roles in growth and survival of bacterial populations. In batch culture conditions, where bacteria undergo lag phase, exponential phase, stationary phase, death phase, and long-term stationary phase (LTSP), expression of alternative DNA polymerases (polB, dinB, and umuDC) is extremely important to creating that genetic diversity and causing mutations that may aide in population survival in LTSP. Many of these beneficial mutations appear prior to death phase and often do not take over the population until after death phase. This thesis characterizes how genetic diversity, through mutation frequency and mutation spectrum analysis via a rifampicin-resistance (Rif®) reporter system, is affected by environmental changes. Further, death phase occurrence helps allow for beneficial mutations to overtake the population upon entry into LTSP. Given the inherent difficulty in studying dying populations, my work utilizes a phenomenon known as the “alcohol effect,” where addition of small, sublethal doses of alcohols between two and six carbons in length can delay the onset of death phase in a dose-dependent manner. I characterize that the alcohol effect is not strain or species-specific and potentially allows for a better understanding regarding how bacterial populations die.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Physiological roles and evolutionary implications of alternative DNA polymerases in Escherichia coli

PDF

Genomic and physiological characterization of Escherichia coli evolving in long-term batch culture

PDF

Mechanisms of long-term survival of Vibrio harveyi and Escherichia coli: linking damage and senescence

PDF

Bacterial survival under low nutrient conditions in the laboratory and the deep biosphere

PDF

Investigating the E. Coli DNA polymerase V mutasome

PDF

Dps contributes to typical growth, survival, and genome organization in E. coli

PDF

Evolution of E. coli in unstructured and structured environments

PDF

Generation of monoclonal antibodies via phage display and in vitro affinity maturation using activation induced deoxycytidine deaminase and DNA polymerase eta

PDF

Molecular genetics of longevity in Escherichia coli

PDF

Survival and evolution of Shewanella oneidensis MR-1: applications for microbial fuel cells

PDF

Genome-wide GATC methylation in long-term cultures of Escherichia coli K-12

PDF

Nutritional competence in Escherichia coli: the use of double-stranded DNA as a nutrient

PDF

Quorum sensing interaction networks in bacterial communities

PDF

Genome sequencing and transcriptome analysis of the phenotypically plastic spadefoot toads

PDF

Understanding the genetics, evolutionary history, and biomechanics of the mammalian penis bone

PDF

Packaging and delivery of plasmids by bacterial extracellular vesicles

PDF

From single molecules to bacterial nanowires: functional and dynamic imaging of the extracellular electron transfer network in Shewanella oneidensis MR-1

PDF

Genetic and molecular insights into the genotype-phenotype relationship

PDF

Molecular signature of aggressive disease and clonal diversity revealed by single-cell copy number analysis of prostate cancer cells across multiple disease states

PDF

The function of Rpd3 in balancing the replicaton initiation of different genomic regions

Asset Metadata

Creator Ferraro, Christina Mary (author)

Core Title Genetic diversity and bacterial death in the context of adaptive evolution

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Molecular Biology

Publication Date 04/26/2019

Defense Date 02/27/2019

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Alcohol,bacteria,death phase,E. coli,error-prone DNA polymerases,Evolution,LTSP,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Finkel, Steven (committee chair), Ehrenreich, Ian (committee member), El-Naggar, Moh (committee member), Goodman, Myron (committee member), Nealson, Ken (committee member)

Creator Email cferraro@usc.edu,cmferraro29@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-149739

Unique identifier UC11660175

Identifier etd-FerraroChr-7275.pdf (filename),usctheses-c89-149739 (legacy record id)

Legacy Identifier etd-FerraroChr-7275.pdf

Dmrecord 149739

Document Type Dissertation

Format application/pdf (imt)

Rights Ferraro, Christina Mary

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