B .R
r...

.r

. Egg ,
. I. ‘L.

I

.x .7
.315
..

.
a".
.55"... Farm

 

 

 

S:
.r
ut

.
..
é

 

 

s; x) L...‘

. :5». . 2
:rzwfiwu a.» .
anvihfiw . “a...

. . A 3%."

r
3

aarmmnmfi

_ a war

.19. .

(gt

a LI. ‘ n. :
swmfiufiww

. 2.: L».
h

A !
«lawn? a...”
n .. 1‘ up.
and . flux
. aux»;

mm.
Vw

.n
flaunt

. I).

I<
..

fir.
‘2 . :wfim
-

ah‘t . .35
‘ ‘ .5...

 

 

THESlS

25ml

 

LIBRARY
Michigan State
University

 

 

 

-.- .00 ‘

This is to certify that the

thesis entitled

ESCHERICHIA COLI RIBOSOMAL RNA OPERON MULTIPLICITY:
GENETIC INTERACTIONS AND FITNESS COSTS

presented by

Joel Ginj i Hashimoto

has been accepted towards fulfillment
of the requirements for

 

M.S. degree in Microbiology and
Molecular Genetics

\ (g (Cg/i

Major professor

 

Date MAUI/(>0

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

JUL 611 4
131204

 

 

 

 

 

 

 

 

 

 

 

 

 

11m chlRCIDateDuepGS—p.“

ESCHERICHIA COLI RIBOSOMAL RNA OPERON MULTIPLICITY:
GENETIC INTERACTIONS AND FITNESS COSTS

By

Joel Ginji Hashimoto

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE

Department of Microbiology and Molecular Genetics

2000

ABSTRACT

ESCHERICHIA COLI RIBOSOMAL RNA OPERON MULTIPLICTY:
GENETIC INTERACTIONS AND FITNESS COSTS

By
Joel Ginji Hashimoto
Recombination between 11778 and the other six rRNA operons of
Escherichia coli resulted in the removal of a sacB-neo insertion in rrsB at a rate
of 5x10'9 per cell division. The loss of the insertion was dependent on flanking
regions of sequence homology to the other rRNA operons. The relative fitness of
the recombinant isolates, measured in head-to-head competition experiments
with a wild-type strain, was ca. 1% higher than the strain carrying the insertion.
Two of 14 recombinant isolates from apparent gene conversion events resulted
in rTnB operons with alternate 168-238 internally transcribed spacer transfer
RNAs. These findings are discussed in the context of a conceptual model
describing the maintenance of sequence heterogeneity in co-evolving genes.
Previously constructed chromosomal deletions of rrnA were
complemented by reinserting the nnA operon in its native site using two different
approaches. Wild-type or near wild-type fitness was not restored in either of the
complemented strains, indicating the large fitness loss seen in strains with rrnA
deleted is not specifically attributable to the deletion of rrnA but rather to
differences in adjacent genes transferred by P1 transduction during the strain
construction. An alternative approach to generate rRNA operon deletions is
proposed that should circumvent the problems with the earlier strain

constructions.

ACKNOWLEDGEMENTS

I would first like to thank my parents Merle and Andrew Hashimoto for
their assistance and support in the many avenues I have pursued. To Merl,
Noelle, and Riki: Thank you for the belief you have shown in my abilities even
when l was doubtful. I would like to thank my advisor Dr. Thomas Schmidt for all
of the excellent advice, for showing the elegance of scientific inquiry, and the
passion with which it should be conducted. Thank you to my committee
members, Dr. Patterson and Dr. Jackson, for their guidance. To the Schmidt and
Breznak labs: Dr. Brad Stevenson, Dr. Dan Buckley, Joel Klappenbach, Kristen
Huizinga, Les Dethlefsen, Joe Graber, and Keri Byzek. Thank you all for the fun
and interesting times in and out of the lab. Thank you Kate Delaney and Ann
Stafford for the countless number of plates poured, miscellaneous support, and
for keeping the project going. And finally to Bryan, Shalette, Ryan, and Sean, for
helping me remember what i set out to do and the importance of the time we are

given.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vi
LIST OF FIGURES ............................................................................... vii
CHAPTER 1 ......................................................................................... 1
CHAPTER 2 ......................................................................................... 6
2.1 Introduction ................................................................................. 6
2.2 Materials and Methods .................................................................. 9
Strains and genotype verification ...................................................... 9
Growth conditions and media ......................................................... 1O
Recombination rate measurements ................................................. 12
Relative fitness assay .................................................................. 13
rRNA operon sequence alignment .................................................. 14
2.3 Results ..................................................................................... 15
Rate of recombination involving rrnB ............................................... 15
Fitness effects of recombination between rrn operons ......................... 21
Correlation analysis of pair-wise rrs and n1 sequence difference
and distance .............................................................................. 23
2.4 Discussion ................................................................................. 25
rRNA operon recombination results in the loss of the
sacB-neo insertion ....................................................................... 26
Fitness effects of rRNA recombination and rRNA operon interruption.....27
Factors affecting the co-evolution of rRNA operons in E. coli ................ 28
Conclusions ............................................................................... 33
CHAPTER 3 ....................................................................................... 35
3.1 Introduction ................................................................................ 35
3.2 Materials and Methods ................................................................. 37
Cloning, XL-PCR and general methods ............................................ 37
Version 1 plasmid constructions ..................................................... 38
Version 1 strain construction ......................................................... 43
Version 2 strain construction ......................................................... 47
Chemostat competition experiments ................................................ 47
Batch culture competition experiments .......................................... 48
3.3 Results ..................................................................................... 48
Version 1 competition experiments .................................................. 48
Version 2 batch culture competition experiments ................................ 50
3.4 Discussion ................................................................................. 53
Lack of complementation in rrnA replacement strains ......................... 53
Rational for constructing rrnA deletions and replacements from
the same transductant .................................................................. 54
Future directions ......................................................................... 55
Conclusions ............................................................................... 56

CHAPTER 4 ....................................................................................... 61

LITERATURE CITED ............................................................................ 64

LIST OF TABLES

Table 2.1. Recombination and sacB mutation rates

Table 2.2. Values used in equations 1(A) and 2(8) to determine recombination

and 8308 mutation rates

Table 3.1. Plasmids used for strain constructions

Table 3.2. E. coli strain constructions

vi

LIST OF FIGURES

Figure 2.1. Location and organization of E. coli rRNA operons. A. E. coli
chromosome with location (mbp) and orientation of rRNA operons. B. General
organization of rrn operons. P1 and P2 are the tandem promoters, rrs encodes
168 rRNA, ITS represents the internally transcribed spacer, m encodes 23S
rRNA, rrf encodes 58 rRNA, and ter is the rho-independent termination region.

C. Table of rRNA operons and their associated ITS and distal tRNA species.

Figure 2.2. BSV1 and BSV2 rmB operon constructions containing the sacB-neo
cassette. A. BSV1 has the sacB-neo cassette inserted into the Bglll site of rrsB,
upstream sequence similarity to the other rm operons, and a transcriptionally
active operon. B. BSV2 was constructed through the removal of sequences
upstream of rrnB, rrsB, the ITS region, and the 5’ end of n1 for the sacB-neo
cassette leaving no upstream sequence similarity to the other rm operons and a
transcriptionally inactive operon. The murl gene is unique to the m8 region and
represents the end of the upstream sequence similarity to the other rRNA

operons.

Figure 2.3. Southern hybridization of BSV1 and BSV2 isolate genomic DNA
digested with Pvull and probed for a conserved region of rrs. In BSV1, the band
corresponding to rrnB is shifted when containing the sacB-neo cassette inserted

in rrs (lane 1), after recombination with a rm operon with a tRNA"°-tRNAA'a ITS

vii

(lane 3), and back to wild-type after recombination with a rm operon with a
tRNAGlu ITS (lanes 4-8). BSV1 sacB isolates are sucrose resistant and
kanamycin resistant and still harbor the sacB-neo cassette in rrsB (lanes 9-12).
BSV2 120 and BSV2 126 are sucrose resistant and kanamycin sensitive
(recombinant phenotype) but have not regained the deleted region of rrsB (lanes
13-14). BSV2 118 and BSV2 142 are sucrose resistant and kanamycin resistant
(sacB phenotype) and have not regained the deleted region of 1738 (lanes 15-

16).

Figure 2.4. PCR amplification of the m3 region of the isolates used in the
relative fitness assay. Amplification of the rrnB operon with the sacB-neo
cassette produces a 10.4 kb fragment as in the BSV1 sacB isolates.
Amplification of the wild-type or a recombinant rmB region produces a 6.6 kb
fragment. The BSV1 recombinant isolates (rec) all have PCR amplification
products of 6.6 kb, indicating the removal of the sacB-neo cassette. The
BSV1rec#31 isolate was excluded due to the strange amplification size and

unexpected banding pattern in the Southern hybridization (Figure 2.5).

Figure 2.5. Southern hybridization of Pvull digsted genomic DNA from BSV1
isolates used in the relative fitness assay. The banding pattern confirms the
recombinant isolates (rec) have lost the sacB-neo cassette and sacB mutants

(sacB) still have the cassette in MB. Due to the unusual banding pattern of

viii

BSV1reo#31, it was excluded from further analysis. BSV1 rec#5 has a banding

pattern indicative of an alternate ITS in rrnB.

Figure 2.6. Relative fitness of BSV1, $303, and recombinant isolates. The
recombinant isolates are significantly different from both BSV1 and sacB groups
but BSV1 and sacB are not statistically distinguishable. Error bars represent +/-

1 standard deviation for each group.

Figure 2.7. Correlation of pair-wise sequence differences of rrs and n1 with
distance between operons of all available bacterial and archeal genome
sequences with multiple rRNA operons. Pair-wise sequence differences were
tallied and distances between operons were calculated based on their published
locations. E. coli comparisons are represented as open diamonds and the other

organism comparisons are represented as closed diamonds.

Figure 3.1. Xbal digestions of pUCA21 plasmid constructions. rrnA was cloned
as an Xbal fragment into pUC19 giving pUCA21 L1 and L2. pUCA21
constructions were further digested with restriction endonucleases to yield the
resulting deletion plasmids. A. Xbal digestions of pUCA21 plasmids. B.

Restriction sites used in the creation of the pUCA21 deletion plasmids.

Figure 3.2. Diagram of steps involved in the complementation of 88301 and

88300 creating A+KOBA* and A-KOBA*. The creation of JHB and JHAb strains

follow these steps starting at step 3 using plasmids pMAKA14 and pKTS102,

respectively.

Figure 3.3. PCR amplification of the rrnA region of the complementation and
deletion strains. The rrnA region amplifies as a 7.3 kb fragment as indicated in
the BS4548 lane. The deleted rmA region amplifies as a 2.6 kb fragment. All
band mobilities for the deletion and complementation strains show the expected

size.

Figure 3.4. Southern hybridization of Pvull digested genomic DNA from rmA
deletion and complementation strains probed for a conserved region of rrs. AII
strains have rrnB deleted. A+KOBA* and A-KOBA* are the complementation
strains starting from B8301 and 88300 with the rrnA operon reinserted as
outlined in Figure 3.2. A+JHB122, A-JHBIZZ, A+JHA8621, and A-JHAB311
were constructed using the intermediate strains B8311 and B8310 with the Ara+
and Ara- marker, respectively. JHB strains have rrnA reinserted into the

chromosome. JHAB strains have rrnA deleted from the chromosome.

Figure 3.5. Chemostat competitions of complementation strain A+KOBA*. The
A+KOBA* vs. A-KOAB competition suggests the presence of rrnA provides a
small fitness advantage. The A+KOBA* vs. A-KOB competition shows the fitness
loss seen in the KOAB strains (88301, 88300) is not complemented by the

reinsertion of MA. The KOAB vs. KOB competition represents data from

reference (49) and is included to show the similarity to A+KOBA* vs. KOB
competition. All chemostat competitions were run in four parallel chemostat
vessels with each time point representing the mean ratio of Ara+ to Ara- with
error bar representing plus and minus one standard deviation. The chemostats

were run at an average dilution rate of 0.11 hr".

Figure 3.6. Relative fitness of A-KOBA* and A-KOAB strains. A+Control is the
parental strain (BS4548). Error bars represent +/- one standard deviation. The
A-KOBA* competition clearly illustrates that the reinsertion of rrnA does not

improve fitness.

Figure 3.7. Relative fitness in batch culture of JHAB and JHB strains. Error bars
represent +/- one standard deviation. The large difference between the control
competitions (A+JHAB vs. A-JHAB and A+JHB vs. A-JHB) indicate that a large

amount of variability is introduced during the creation of the strains.

Figure 3.8. Ammons-Fox protocol for generating chromosomal deletions.
Plasmid and strain information is summarized in Tables 3.1 and 3.2. The first
two steps generate the required plasmids with the FRT-kan’-F RT cassette
inserted between the up/down sequence allowing site-specific integration of the
pasmid into the E. coli chromosome. The plasmid pKO3 has a temperature
sensitive origin of replication (ts-ori) and the sacB gene from Bacillus subtilis

conferring sucrose sensitivity. E. coli B strains REL4548 and BS4548 are then

xi

transformed with pKO3-deIA (or -delB) and cointegrates are isolated by growing
transformants at 42°C in the presence of kanamycin and chloramphenicol.
Growth of cointegrants at 30°C in the presence of 6% sucrose and kanamycin
selects cells with the plasmid excised and lost from the cell due to the lethal
effects of the 3308 gene in the presence of sucrose. Kanamycin resistant cells
are then transformed with pCP20 carrying the F lp recombinase resulting in the
site-specific excision of the kanr gene flanked by FRT sites (F lp recognition
sequence). After F Ip removal of the kanr gene, a partial F RT sequence remains

at the site of the deletion.

xii

Chapter 1
INTRODUCTION
rRNA operons and Bacterial physiology

In most bacterial species, ribosomal RNA (rRNA) is transcribed from an
operon with a rrs-rrI-nf (168—238-58) gene organization, in some instances with
transfer RNA (tRNA) genes located between rrs and rrl (Internally Transcribed
Spacer; ITS) and distal to nf (6). The organization of all three species of rRNA
into an operon allows for their coordinated regulation preventing the production of
an excess of one type of rRNA. In Escherichia coli, all seven rRNA operons
have two promoters (P1, P2) with transcription from the P1 promoter subject to
upstream activation, regulation by the stringent response, and growth rate
dependent regulation, while transcription from the P2 promoter occurs at a lower,
constitutive level (14).

In a growing cell, ribosome biosynthesis requires a large proportion of the
cellular energy expenditures. Ribosomal RNA accounts for over half of the total
RNA transcribed during rapid exponential growth illustrating the need for lots of
ribosomes to meet the translational requirements of the cell (6, 14). Thus, the
high cellular requirement for rRNA, especially during transitions to favorable
growth conditions, can only be met by the presence of strong promoters and, in

the case of many organisms, by having multiple rRNA operons.

rRNA operon copy number and ecological strategy

Ribosome biosynthesis is an energy intensive process that is tightly
regulated to match the cellular requirements for translation. However, the
cellular concentration of RNA is not proportional to the translational requirements
of the cell at slow growth rates. Excess ribosomes were detected in E. coli,
which has seven rRNA operons, at a growth rate of 0.1 doubling per hour
suggesting the increased transcriptional capacity gained by having multiple rRNA
operons may be a cost to the cell under slow growth conditions (31).

Recent observations have strengthened the apparent correlation of rRNA
operon copy number and general ecological strategies. In E. coli strains, an
increase in response time, or the interval of time required to enter exponential
growth upon the shift to more favorable growth conditions, was seen with
increasing numbers of rRNA operons disrupted by antibiotic cassettes
suggesting a link between high rRNA operon copy number and rapid response
times (12). These results, however, are complicated by the presence of several
antibiotic cassettes in the cell and active promoters producing truncated or
disrupted rRNA transcripts. The potential metabolic expense incurred by the
presence of multiple rRNA operons under slow growth conditions was also
studied in strains of E. coli that harbored additional plasmid-born copies of MB.
The additional rRNA operons increased doubling times by 33% and was
attributed to the diversion of energy to the synthesis of excess ribosomes (50).
The general observation of rRNA operon copy number as a predictor for

ecological strategy also holds true for environmental microbial isolates. Bacteria

isolated from the soil that respond quickly to the influx of nutrients and form
colonies early on agar plates had an average of 5.5 rrs genes while late
appearing colonies had an average of 1.4 rrs gene copies (30).

The pattern of species with few rRNA copies employing an ecological
strategy characterized by efficient growth under low, sustained nutrient
conditions and species with high numbers of rRNA operons characterized by
quick response times to favorable conditions suggests a key predictive role of
rRNA operon copy number for different ecological strategies. Both strategies
have benefits and trade-offs. For the low copy number species, while they are
efficient at growing at slow growth rates under constant nutrient conditions, they
are unable to respond quickly to an abundance of nutrients and may be out
competed by high copy number organisms when the environment is
characterized by periodic fluctuations in favorable growth conditions. For high
copy number organisms, while they may be able to respond quickly to an influx of
nutrients due to their higher capacity to produce rRNA, the constitutive
production of rRNA under slow growth conditions would favor low copy number
organisms able to grow more efficiently. While several observations suggest an
important association between rRNA operon copy number and ecological
strategy, a direct experimental link has yet to be established.

Recombination and rRNA operons

When present in multiple copies, rRNA operons are a conspicuous

deviation from the norm in Bacteria where almost all chromosomal genes are

present in a single copy (6, 39). While potential ecological explanations for the

presence of multiple rRNA operons are discussed above, an additional
implication of multiple rRNA operons is the potential for recombination between
the operons. Such recombination events between rRNA operons have been
shown to lead to inversions, deletions, duplications, and the exchange of
sequence between rRNA operons in E. coli (22-27). Thus, multiple rRNA
operons may also play an important role in the organization and reorganization of
the chromosome of an organism.

Homologous recombination between multiple copies of rRNA operons can
also result in the concerted evolution of the sequences and repair of operons that
have been disrupted or may be non-functional (2-4, 7, 10, 22). The co-evolution
of rRNA sequence is evident in the similarity of the multiple sequences within a
species as compared to closely related species. For example, the seven rRNA
operons of E. coli are more similar to one another (~99% identity) than to the
seven rRNA operons of Salmonella typhimurium (~97% identity) (3, 4, 7, 10). In
addition, several of the rRNA operons of E. coli have identical sequences but the
number and identity of the operons with identical sequence can vary in different
E. coli strains (10). Thus, despite decades of inquiry into the interactions
between the rRNA operons of E. coli, a coherent view of the underlying dynamics
which govern such interactions has yet to be addressed beyond the frequency

with which the phenomena occur.

Overview of Thesis

In Chapter 2 of this thesis I report the measurement of recombination
rates between rrnB and the remaining six rRNA operons of E. coli. Isolates that
had undergone recombination events to remove an insertion in the rrsB gene
were further examined to determine the fitness effects of recombination in batch
culture competitions with a wild-type strain. The results from the competition
experiments and recombination rate measurements are discussed in the context
of developing a conceptual model for the maintenance of sequence
heterogeneity in the rRNA operons of E. coli.

Chapter 3 summarizes two attempts to establish a growth rate specific
role for the rrnA operon by reinserting mrA into its native position in strains of E.
coli that have rrnA alone or rmA and rrnB deleted from the chromosome. The
resulting strain constructions were tested for rescue of the wild-type fitness levels
in batch culture and chemostat competitions. The shortcomings of the
complementing strain constructions will be discussed and a potential new

method for generating rRNA operon deletions will be proposed.

Chapter 2
RATE AND FITNESS EFFECTS OF RECOMBINATION BETWEEN rRNA
OPERONS OF ESCHERICHIA COLI
2.1 Introduction

Escherichia coli has seven ribosomal RNA (rRNA) operons located
asymmetrically on the chromosome that are highly similar in gene organization
and nucleotide sequence (Figure 2.1) (6). The large regions of sequence
similarity allow for inversions, duplications, and sequence exchange between the
rRNA operons at frequencies of 10’3-10'5 in overnight liquid cultures (6, 22, 24-
27). High recombination frequencies between rRNA operons suggests that the
E. coli chromosome is dynamic, yet the overall arrangement of rRNA operons in
the chromosome is very stable. For example, the number, chromosomal
location, and orientation of rRNA operons has been preserved in E. coli and
Salmonella typhimurium (S. enterica serovar Typhimurium) since their
divergence (6, 28). The observation of high recombination frequencies and the
conservation of chromosome and operon organization suggests strong selection
for the maintenance of rm organization and location over evolutionary time
scales.

While most genes in Bacteria and Archea are present in single copies, the
genes encoding the rRNAs are often present in multiple copies allowing for
concerted evolution of the sequences through sequence exchange (6, 34).
Recombination between genes present in multiple copies on a chromosome can

homogenize sequences by gene conversion (non-reciprocal exchange) or by

homologous recombination (reciprocal exchange) with sister chromosomes,
resulting in the substitution of one gene sequence for another (36). The
exchange of rm sequences within the E. coli lineage has resulted in the seven rm
operons of E. coli appearing more similar to one another than to any of the seven
rm operons of S. lyphimurium. If recombination between the rRNA operons that
resulted in homogenization and co-evolution of the sequences did not occur,
multiple rm operons within a species would be expected to accumulate neutral
mutations at a rate comparable to changes observed between species, which is
clearly not the case. Apparent exceptions to this phenomenon exist in
Thermomonospora chromogena, Thermobispora bispora, and Haloarcula
marismortui where rrs or rrl sequences differ by up to 10% with the high degree
of heterogeneity potentially explained by the lateral transfer of rDNA (17, 42, 53,
54).

This chapter focuses on the factors that affect the concerted evolution of
rRNA sequences. The rate of recombination involving rrnB and the six other
rRNA operons was measured to estimate the homogenizing force of
recombination. The resulting fitness effects of inter-operon recombination were
then measured to determine the dynamics affecting the eventual fixation or loss
of the homogenized operons. The results were then used to develop a
conceptual model of rRNA operon evolution, focusing on a mechanism for the

maintenance of sequence heterogeneity between rRNA operons.

 

 

 

 

 

 

 

rmG
B.
P1 P2 rrs fer
rrs H rrl rrf
IE1 H11 H}—
C.
rrn Operon ITS tRNA(s) Distal tRNA(s)
B: E: G GIU2 ..
C Glu Asp1, Trp
A "91, A3318 -
D lle1, Ala1B Thr1
H "61, AIa1B A8p1

 

Figure 2.1. Location and organization of E. coli rRNA operons. A. E. coli
chromosome with location (mbp) and orientation of rRNA operons. B. General
organization of rm operons. P1 P2 are the tandem promoters, rrs encodes168
rRNA, ITS represents the internally transcribed spacer, rrl encodes 238 rRNA, rrf
encodes 5S rRNA, and ter is the termination region. C. Table of rRNA operons
and their associated ITS and distal tRNA species.

2.2 Materials and Methods
Strains and genotype verification

Insertional inactivations of the rrnB operon were produced by linear
transformation of a recD' strain of E. coli (D308) with a DNA fragment containing
a rrnB operon and a sacB-neo cassette (8) inserted in the Bglll site of rrsB.
Generalized transduction (P1vir) was used to move the rrnB operon disrupted by
the sacB-neo cassette from D308 into a strain of E. coli B that had been
propogated in the lab for 10,000 generations (33). The resulting strain, BSV1,
contains a rrnB operon with intact promoters and homology upstream and
downstream of the sacB-neo cassette to the other rRNA operons (Figure 2.2). A
sacB-neo rrnB allele with no upstream homology to the other rRNA operons was
used to construct strain BSV2 with the promoters (P1 and P2), rrsB, and the 5'
end of n18 deleted, leaving a transcriptionally inactive rmB region with no
homology upstream of the sacB-neo cassette to any of the other rRNA operons
(Figure 2.2) (49).

BSV1 and BSV2 genotypes were verified by both PCR and Southern
hybridization. PCR amplification of rrnB was carried out using the upB fonlvard
(5’-CCGAATTACATATGACCGTGCTGGTGTI’TGAC-3’) and dnB reverse (5’-
CTACTCTAGACCTGATGCAAAAACGAGGCTAG'ITI'A-3’) primers and thh
DNA Polymerase XL (PE Applied Biosystems; Foster City, CA) under the
following cycle conditions: 94°C 3 min, 94° C 30”; 60° C 1 min; 72° C 10 min; 30
cycles, 72° C 10 min, hold at 4° C. The rRNA operons were visualized using

Southern hybridization on Pvull digested genomic DNA probed with a digoxignin-

dUTP-Iabeled DNA probe for a conserved region of rrs (positions 8-536 of rrsB)
(30).
QQWII'I corflionsarflmedig

Cultures were typically grown in Luria-Bertani liquid medium (LB), or with
1.5% Bacto Agar (Difco, Laboratories; Detroit, MI) added for solid medium.
When selecting for sucrose resistance, sucrose was added at a final
concentration of 6% to LB with NaCl omitted (LBsuc) (8). Where noted,
kanamycin was added to LBsuc at a final concentration of 50ug/ml (LBsuc-kan).
Overnight cultures consisted of 2 ml of LB in 16 x 125 mm culture tubes,
inoculated from freezer stocks. Competition experiments, conditioning cultures,
and cultures for rate measurements were grown in either 10 ml of media in a 50
ml Erlenmeyer flask or 50 ml of media in a 250 ml Erlenmeyer flask shaking at
225 rpm at 37° C. The conditioning cultures and competition experiments used
to measure relative fitness were grown in Davis-Mingoli broth (DM)
supplemented with 2.0 mg/L thiamine hydrochloride and 25 mg/I of glucose
(DM+25) (33).

10

A. BSV1

 

 

 

 

 

 

F sacB-neo l
(er
——i iii rm H:—
murl rrs rrl rrf
B. BSV2
ter
———l H sacB-neo l Jfl——
mur/ rrl rrf

Figure 2.2. BSV1 and BSV2 rrnB operon constructions containing the sacB-neo
cassette. A. BSV1 has the sacB-nee cassette inserted into the Bglll site of rrsB,
upstream sequence similarity to the other rrn operons, and a transcriptionally
active operon. B. BSV2 was constructed through the removal of sequences
upstream of rrnB, rrsB, the ITS region, and the 5' end of rrl for the sacB-neo
cassette leaving no upstream sequence similarity to the other rm operons and a
transcriptionally inactive operon. The gene murl is unique to the rrnB region and
represents the end of upstream sequence similarity to the other rRNA operons.

11

Recombination rate measurements

The rate measurements for recombination and $308 mutation were carried
out as outlined by Crane et. al (15), and is a modification of the fluctuation
analysis of Luria and Delbruck (38). Eleven flasks with 10 ml of LB were
inoculated with ~100 cells from an overnight LB culture. The cultures were
incubated at 37 °C shaking at 225 rpm for 15 hours. The stationary phase
cultures were sampled by spreading 75ul of the culture on 4 LBsuc plates, 4
LBsuc-kan plates, and diluted and spread on 4 LB plates. After overnight
incubation at 30° C, the number of colonies on each plate was counted. For
each replicate culture, the size of the recombinant population was calculated as
the number of colonies on LBsuc minus the number of colonies on LBsuc—kan.
The 3308 mutant population was represented as the number of colonies on
LBsuc-kan.

To determine the rate of recombination and 3308 mutation, the median
value for each population was determined from the 11 replicate cultures. The
following median estimator of mutation rates was used to determine rates of
mutation and recombination (15, 29):

Amed = ((rm/s) - In 2)/(ln(rmls) — ln(ln 2)) (equation 1)
where rm is the median number of recombinants or mutants per culture, 3 is the
dilution factor, and Amed is the median number of recombination events or
mutation events in the culture. The recombination or mutation rate was then

determined by dividing Amed by the median of the total number of cells per culture.

12

For the rate of double mutation in sacB and neo in BSV2 (recombinant
phenotype; sucrose resistant, kanamycin sensitive), a different equation was
used because many of the replicate cultures produced no colonies with the
recombinant phenotype. The following equation is more suited for results with
many ‘mutant free’ cultures (15, 29):

mg= ((1-p) In (le))/(p In (p)) (equation 2)
where p is the dilution factor, 2. is the number of cultures devoid of mutants, C is
the total number of cultures, and mg is the number of mutants per culture.
Relative fitness as_sa_y

Overnight LB cultures of each competing population were inoculated from
freezer stocks. The following day, conditioning cultures, consisting of 10 ml of
DM+25 in 50 ml Erlenmeyer flasks were inoculated with a final dilution of
1:10,000 from the overnight cultures. Competition cultures were initiated 24
hours after inoculating the conditioning cultures by adding 100ul of a 1:1 mix of
the two competitors to 5, 50ml flasks containing 10 ml of DM+25. BSV1, BSV2,
and the resulting mutants and recombinants carry a neutral arabinose utilization
marker (Ara”) which enables the enumeration of the competitor versus the
control strain (Ara‘), because on TA plates Ara+ colonies appear white and Ara'
colonies appear red (33). Cultures were transferred at 24 hr intervals for 4 days
and monitored daily by plating a 5x10“6 dilution of the each replicate culture on 3

TA plates.

13

Relative fitness measurements were calculated using the following

equafions:
m1 = In [N1(1)/N1(0)]/(1 d) (equation 3)
Wk; = m1/m2 (equation 4)

where N1(0) and N1(1) are the population size of Ara+ strains at 0 hours and 24
hours, respectively, and m1 is the Malthusian parameter for the Ara+ strain. The
relative fitness of the Ara+ strain to the control Ara' strain was calculated using
equation 4 where m is the Malthusian parameter of the Ara+ strain and m; is the
Malthusian parameter of the Ara' control strain (33, 49). Statistical analysis of
the relative fitness of BSV1, the BSV1 recombinant isolates, and the BSV1 sacB
isolates included ANOVA and pair-wise comparisons using the Bonferroni
multiple comparisons procedure.
rRNiA operon sequence alignments

Sequences of rrs (168) and rrl (23S) rRNA genes were obtained from
GenBank, The Institute for Genomic Research Microbial Genome Database
(www.tigir.org/t_db/mgb/md_bcomplete.html), and the National Center for
Biotechnology Information Microbial Genome Database
(www.ncbi.nlm.nih.gov/PMGifs/Genomes/bact.html) and analyzed with the
VWsconsin Package Version 10.0, GCG software package (Madison, WI). The
following microbial genomes containing multiple rRNA operons were used in the
analysis and represent all available sequenced genomes with multiple rRNA
operons at the time of preparation: Aquifex aeolicus VF5, Bacillus subtilis ATCC

23857, Campy/obacterjejuni ATCC 700819, E. coli ATCC 10798, Haemophilus

14

influenzae ATCC 51907, Helicobacter pylori 26695, Methanococcus jannaschii
DSMZ 2661, Methanobacterium thermoautotrophicum ATCC 29096, Neisseria
menigitidis M058, Treponema pallidum ATCC 25870, Ureaplasma urea/yticum
serovar 3, Vibrio cholerae ATCC 39315, and X yella fastidiosa 9a50. Pair-wise
differences were calculated for the rrs and n1 genes of each organism. Pair-wise
sequence differences were scored and distances between rRNA operons were
calculated based on the published locations of the operons (9, 16, 19, 20, 32, 43,

44, 46, 48, 51, 52).

2.3 Results
Rate of recombination involving rrnB

Fluctuation analysis of BSV1 showed the average rate of recombination
removing the sacB-neo cassette in rrnB of BSV1 to be 4.75x10'9 per cell division
(Table 2.1) in two independent measurements using equation 1 and the values in
Table 2.2. Of 14 independent BSV1 recombinant isolates screened by Southern
hybridization, two (14%) had recombined with a rRNA operon containing the
tRNA”‘°1-tRNAAlana ITS spacer (rrnA, rrnD, or rmH), resulting in a hybrid rrnB
operon with an alternate ITS region. In BSV2, the recombinant phenotype
occurred at a rate of 1.4x10‘10 per cell division (equation 2, Table 2.2), and is
within the range of values measured for the spontaneous inactivation of both
phenotypes of the sacB-neo cassette (8). All BSV2 isolates with the recombinant
phenotype (sucrose resistant, kanamycin sensitive) that were screened by PCR

and Southern hybridization maintained the 3.8 kb insertion in rrsB, indicating the

15

loss of both phenotypes of the sacB-neo cassette was a result of mutations
rather than the removal of the cassette by recombination. The average sacB
mutation rate was 5.7x10’8 per cell division in BSV1 and BSV2 (Table 2.1). In all
sacB mutants (sucrose resistant, kanamycin resistant) examined by PCR and
Southern hybridization, the sacB-neo cassette was present in the rrnB operon

(Figures 2.3, 2.4, 2.5).

16

Table 2.1. Recombination and sacB mutation rates.

 

 

Strain Recombination Rate sacB mutation rate
(per cell division) (per cell division)
BSV1 rep. 1 4.8x10'9 4.6x10'8
BSV1 rep. 2 4.7x10'9 7.4x10'8
BSV2 0* 5.1x10’8

 

*The rate of the recombinant phenotype (sucrose resistant, kanamycin sensitive)
in BSV2 cultures was 1.4x10'10 per cell division. However, all BSV2 isolates with
the recombinant phenotype were found to have maintained the insertion in the

rrnB gene.

Table 2.2. Values used in equations 1(A) and 2(B) to determine recombination
and sacB mutation rates

 

 

 

 

 

A. Strain rm 3 (dilution factor) lmed Population
BSV1rec (rep#1) 41 0.03 180 3.76x1010
BSV1rec (rep#2) 20 0.004 562 1.20x1011

BSV1sacB (rep#1) 523 0.03 1720 3.76x101°

BSV1sacB (rep#2) 424 0.004 8879 1.20x1011

BSV2 sacB 447 0.03 1493 2.91x10‘°

B. Strain j 2 C rng Population

BSV2rec 0.03 7 11 4.17 2.91x1010

 

17

  

11

>

BSV1 rec 320

, BSV1 rec 640

. BSV1 rec 460
S BSV1sacBZO

     

BSV1

          

i BSV1 rec1 2
‘. BSV1 rec 100
BSV1 sacB72
Sf BSV2 120
BSV2 126

BSV1 r

   

rrnE—» ., ‘ - "
rrnB—r “
alt. ITS rrnB_.

sacB-neo rrnB ——> g I . ' “'“ a—— 6.5 kb
rrnG _. _ f‘ -
rrnC —» .
rrnA _.
rrnD —> ._ 4.3 kb

”DH—r '_ ..

Figure 2.3. Southern hybridization of BSV1 and BSV2 isolate genomic DNA
digested with Pvull and probed for a conserved region of rrs. In BSV1, the band
corresponding to rrnB is shifted when containing the sacB-neo cassette inserted
in rrs (lane 1), after recombination with a rrn operon with a tRNAI'e-tRNAA'a ITS
(lane 3), and back to wild-type after recombination with a rrn operon with a
tRNAGlu ITS (lanes 4-8). BSV1 sacB isolates are sucrose resistant and
kanamycin resistant and still harbor the sacB-neo cassette in rrsB (lanes 9-12).
BSV2 120 and BSV2 126 are sucrose resistant and kanamycin sensitive
(recombinant phenotype) but have not regained the deleted region of rrsB (lanes
13-14). BSV2 118 and BSV 142 are sucrose resistant and kanamycin resistant
(sacB phenotype) and have not regained the deleted region of rrsB (lanes 15-
16).

     

' 1 kb ladder

10.4 kb—» -._ 10 kb
- e—8kb
6.6kb——> ' -._6kb

Figure 2.4. PCR amplification of the rrnB region of the isolates used in the
relative fitness assay. Amplification of the rrnB operon with the sacB-neo
cassette produces a 10.4 kb fragment as in the BSVISacB isolates. Amplification
of the wild-type or recombinant rrnB region produces a 6.6 kb fragment. The
BSV1recombinant isolates (rec) all have PCR amplification products of 6.6 kb
indicating the removal of the sacB-neo cassette. The BSV1rec#31 isolate was
excluded due to the strange amplification size and unexpected banding pattern in
the Southern hybridization (Figure 2.5).

BSV1 rec#61
5

3 V1 r
V

BSV2

BSV1

I!
if
C
R
t
I

u *— rrnB
.. . . _ ‘—\sac3-neo ””3

n n
Ii III

II I,
II In

I

Figure 2.5. Southern hybridization of Pvull digested genomic DNA from BSV1
isolates used in the relative fitness assay. The banding pattern confirms the
recombinant isolates (rec) have lost the sacB-neo cassette and $808 mutants
(sacB) still have the cassette in rrnB. Due to the unusual banding pattern of
BSV1rec#31, it was excluded from futher analysis. BSV1rec#5 has a banding
pattern indicative of an alternate ITS in rrnB.

20

Fitness effects of recombination between rRNA operons

Independent isolates of BSV1 cultures with the recombinant phenotype
(sucrose resistant, kanamycin sensitive) were confirmed by PCR and Southern
hybridization to have lost the sacB-neo cassette from rrsB (Figure 2.4,
Figure2.5). The amplification product from the rrnB operon with the sacB-neo
cassette in BSV1 is 10.4 kb, while the amplification product from the rrnB operon
without the cassette (either wild type or recombinant) is 6.6 kb (Figure 2.4).
Southern hybridization of genomic DNA digested with Pvull and probed for rrs
produces a 8.4 kb fragment in wild-type and recombinant isolates of E. coli and a
6.1 kb fragment in BSV1 (Figure 2.3, Figure 2.5).

The BSV1 recombinant populations had a higher relative fitness than both
the parental strain BSV1 and the BSV1 sacB isolates (p = 0.0089, p = 0.0034;
respectively). The relative fitness difference between the BSV1 recombinant
isolates and either BSV1 or the BSV1 sacB isolates was about 1% (Figure 2.6).
There was no significant difference between BSV1 and theB8V1 sacB isolates (p
= 0.2823). One of the BSV1 recombinant isolates had an abnormal Southern
hybridization pattern and a PCR amplification product at an unexpected size.
Due to this discrepancy, the data from BSV1rec#31 was excluded from our
analysis. One of the recombinant isolates used in the relative fitness
measurement (BSV1rec#5) was found to have recombined with a rRNA operon
with a tRNA”°-tRNAA'° ITS by Southern hybridization. The relative fitness of the
alternate ITS recombinant was indistinguishable from the other recombinants and

is included in the BSV1 recombinant group.

21

 

1.005
1.000 .. _.. __ _._.._ .. __________ ..__ __.__._ __ _ _ f _ _._. _.

 

0.995 “ 7"“ f e i i i — 4 ~ W-______ We l WWW
0.990 -. ~ ~ A a e A e Jw—m— _.__ W W -W _

 

0.985 '*” W‘ A mr-ile-w T— ‘ # ——_. ..__
0.980 ~~-—-~ l _l__. e e i W _ W _- ..... .

 

 

relative fitness

0.975 -~ .- e e W - WW ______
0.970 — —v 4 — —— 44 W _ __ -.__W

 

 

 

 

 

 

 

 

 

0.965
BSV1 sacB recombinant

Figure 2.6. Relative fitness of BSV1, sacB, and recombinant isolates. The
recombinant isolates are significantly different from both BSV1 and sacB groups
but BSV1 and sacB are not statistically distinguishable. Error bars represent +/- 1
standard deviation for each group.

22

Correlation analvsis of pair-wise rrs and n1 sequence difference and distance
The results of the pair-wise sequence comparisons of rrs and rrl from
sequenced microbial genomes with multiple rRNA operons are summarized in
Figure 2.7. There is a significant positive correlation between the total pair-wise
differences present in rrs and rrl and the distance between the operons
compared, suggesting a potential link between chromosomal distance between

operons and recombination rate.

23

r=053
50 p = < 0.0001

pair-wise differences

 

 

 

0 500 1000 1500 2000 2500
distance between operons (kb)

Figure 2.7. Correlation of pair-wise sequence differences of rrs and rrl with
distance between operons of all available bacterial and archeal genome
sequences with multiple rRNA operons. Pair-wise sequence differences were
tallied and distances between operons were calculated based on their published
locations. E. coli comparisons are represented as open diamonds and the other
organism comparisons are represented as closed diamonds.

24

2.4 Discussion

The objective of this investigation was to address and quantitatively
measure some of the factors that affect the co-evolution of rRNA sequences in E.
coli. The presence of reiterated sequences in a genome allow for various types
of recombination with a variety of outcomes. One result of recombination
between repeated sequences is the potential for sequence exchange resulting in
the co-evolution of the sequences. Ribosomal RNA operons are often present in
multiple copies in bacterial and archeal genomes setting them apart from most
other genes which are normally present in single copies. Prokaryotic species
with multiple copies of the rRNA operon have been shown to have rDNA
sequences that are more similar within a species than between species which
clearly indicates that the rRNA genes are co-evolving (34).

Gene conversion is typically invoked to explain the concerted evolution of
gene families. The clearest cases for gene conversion are from Ascomycetes
fungi where all the products of meiosis are easily examined (40, 41). Our
system, however, does not allow the easy isolation and identification of all the
products of a given recombination event and therefore any claim as to the nature
of the specific recombination mechanism would be tentative, at best. The type of
recombination events described here, however, clearly would lead to the
concerted evolution of rRNA sequences as a heterogeneity (sacB-neo cassette)
has been removed from the m8 operon through recombination with one of the

other rRNA operons.

25

rRNA operon @comgiflgtion results in the loss of the sacB-neo insertion

The involvement of recombination in genetic alterations is typically shown
by the requirement of one or a combination of recombination genes (i.e. recA,
recB, recF). Here we have not shown a requirement for any of the recombination
genes but two lines of evidence clearly show the involvement of recombination in
our system. First, the loss of the 3.8kb sacB-neo insertion was observed only
when flanking homology to the other rRNA operons was present. Second, the
loss of the sacB-neo cassette from rrnB was observed to coincide with the
exchange of internally transcribed spacer (ITS) regions resulting in an altered
mobility of the m8 fragment in Southern hybridizations (Figure 4). In 14% of the
BSV1 recombinant isolates (2 of 14), Southern hybridization indicated that the
rrnB operon had recombined with an operon containing the tRNA"°‘-tRNAA'°”.I3
ITS (rmA, MD, or rmH) resulting in an rmB operon with the tRNA'm-tRNAA'a1B
ITS. The presence of the tRNA”e1-tRNAA|a1l3 ITS region in rrnB could only occur
through recombination between 77718 and a rRNA operon with the tRNA"°1-
tRNAAI°1E3 ITS and, therefore, clearly implicates recombination in the loss of the
sacB-neo cassette.

The recombination rate measured for BSV1 is strikingly similar to other
rates reported for the exchange of sequence via recombination. The sacB-neo
cassette was lost from BSV1 at a rate of 5x10“9 per cell division which is within
the range of recombination rates (3x10‘9 to 2x108) observed between the tuf
genes of S. typhimurium (1). Mutation and recombination events are commonly

reported as a frequency of a population, which is a measure of the rate and the

26

subsequent growth of the recombinant or mutant sub-populations. The reporting
of recombination and mutation events as frequencies and not rates makes
comparisons to other studies difficult. In a strain of E. coli with only one tRNAG'“2
spacer, the exchange of ITS regions resulting in an increase in the number of
tRNAGluz spacers occurred at a frequency of 5x10'5 (22). The apparent
discrepancy between our measured rate of recombination and previously
measured frequencies can be attributed in part to the large differences in the
growth rate of the recombinant and parental populations in the experiments by
Harvey and Hill (22) caused by the large disparity of ITS tRNA species and their
lack of control for “jack-pot” cultures.
Fitness effects of rRN_A recombination MNAgLaon interruption

The loss of the sacB-neo cassette from ME through homologous
recombination with one of the other six intact rRNA operons resulted in a 1-2%
increase in relative fitness (Figure 2.6). Thus, the cumulative effects of rrnB
interruption, carrying the sacB-neo cassette, and recombination involving rrnB is
approximately 1%. It has been previously reported that antibiotic cassette
interruptions of either rrnA, rrnB, or miD resulted in no significant change in
growth rate in various types of media and temperatures (11, 12). Our results do
not contradict these earlier studies because the comparisons of relative fitness
made here were measured in head-to-head competition experiments which
detects small differences in lag, exponential growth, and stationary phase,

enabling us to resolve the differences summarized in Figure 2.6.

27

The BSV1 sacB isolates had a relative fitness indistinguishable from BSV1
and significantly lower than the recombinant isolates indicating that a functional
sacB gene has little impact on relative fitness. The relative fitness of the
recombinant isolate with an alternate ITS region was found to be within the range
of the other recombinant isolates, suggesting the exchange of spacer ITS has
little effect on relative fitness. Because of the paucity of alternate ITS
recombinants, statistically supported comparisons could not be made for this
group.

.F_actorsgffecting the co-evolution of rRNA operons in E. coli

The similarity of the seven rRNA operons of E. coli and preservation of
chromosome and rRNA operon organization since the divergence from S.
typhimurium (6, 28) suggests that selection against recombination between rRNA
operons is negligible when resulting in the exchange of short stretches of
sequence with no alteration of the organization of the chromosome or the rRNA
operons. Furthermore, apart from short insertions and deletions from the ITS
regions of the seven rRNA operons of E. coli K-12, the degree of heterogeneity in
the ITS is similar to the degree of heterogeneity seen in rrs and rrl (3, 4),
indicating that sequence exchange and homogenization affects the entire rRNA
operon and is not limited to the structural genes.

If the heterogeneity in the rRNA operons is not selectively maintained, the
presence of heterogeneity is surprising, due to the high rate and frequency of
recombination between rRNA operons observed in this and other studies. In light

of the observations presented here, the maintenance of sequence heterogeneity

28

in the rRNA operons of E. coli can be explained by a model in which the balance
of the antagonistic phenomenon of recombination and mutation results in similar
yet heterogeneous rRNA operons.

If we assume the 1% fitness increase seen in the recombinant isolates is
specifically attributable to the repair and reestablishment of a functional rRNA
operon, the fixation of recombination events between rRNA operons would be
expected to be a process that occurs very rarely. In an environment similar to the
conditions of the relative fitness assay, the probability of fixation of a recombinant
with a fitness increase of 1% over the rest of the population is 4.4x10’3 and would
occur only after ~2500 generations (33). However, the 1% increase in the relative
fitness of the recombinant isolates may not be solely attributable to the
reestablishment of the rrnB operon, but may also be due to the loss of the
cassette carrying and the kanamycin resistance phenotype. Thus,
homogenization may be occurring at a significant rate but the fixation of the
homogenized operons would rarely occur or would occur on the rare occasion
that the homogenized operons hitch-hike to fixation with a beneficial mutation
located elsewhere in the chromosome.

Comparisons of rRNA operon sequences of available Bacteria and Archea
complete genome sequences adds an additional factor to this model. When the
data from all of the prokaryotic organisms with sequenced genomes and two or
more rRNA operons are analyzed, there is a positive correlation between the
number of pair-wise sequence differences (of 168 and 238 genes within an

individual organism) and the distance between the operons on the chromosome

29

(Figure 2.7). In other words, the further apart two rRNA operons are, the more
likely they are to have sequence heterogeneity, suggesting that the proximity to
other operons may be an important factor in the rate with which operons
recombine. Thus, the outcome of the opposing forces of the recombination rate
resulting in homogenization and the de novo mutation rate resulting in
heterogeneity becomes more complex. Organisms with multiple rRNA operons
would have different rates of recombination for each operon depending on the
proximity to the other operons. Thus, by being localized distantly from other
rRNA operons, an operon may be able to escape some of the homogenization by
having a lower rate of recombination resulting in a divergent operon.

The dependence of recombination rate on the distance between the
regions participating in the exchange has been shown for plasmid and
chromosomal based systems in various types of organisms (18, 37, 45, 47). In
E. coli, recombination rates were reduced up to two orders of magnitude when
homologous regions were separated by 7kb (37). In S. typhimurium,
recombination between 5 kb regions is reduced by over one order of magnitude
when the distance separating the homologies is increased from 2 to 17 minutes
(45). In Eukarya, the pattern of decreasing recombination rate with increasing
distance between regions has been observed in the nontrascribed spacers (NTS)
of ribosomal RNA genes (18). The pattern of heterogeneity in NTS regions of
wheat, rye, maize, and mice was better fit by a model that assumes an increased
probability of sequence exchange for closely located regions (18). In the

bacteriophage T4, the frequency of recombination decreased over one order of

30

magnitude when the distance between recombining regions was increased from
79 to 1698 bp (47). Thus, the observed positive correlation of sequence
difference and distance has empirical support based on recombination rates
between homologous sequences separated by varying distances, and may be a
widespread phenomenon impacting the concerted evolution of gene families in
all domains of life.

The positive correlation of sequence heterogeneity between rRNA
operons and distance between the operons raises the question of how proximity
would affect recombination rates. The correlation may be a result of local
unwinding of the DNA during replication or transcription resulting in easier access
to close regions for strand invasion and sequence exchange. Additionally, the
proximity of closely located operons may result in an increased local
concentration of operons during replication. As a replication fork passes through
a region, the operons in that region would be at a higher relative concentration
than other operons resulting in an increased probability of sequence exchange.

A subtle yet interesting implication of the observed correlation of
increasing sequence heterogeneity with distance is the implied mechanistic
nature of the interactions leading to homogenization. While not a direct inference
from the data, it seems more likely that the observed correlation would arise
primarily through intrachromosomal interactions where replication is not required
in order to explain an increased frequency of interaction between closely located
operons. Only under near maximal growth rates, growth conditions assumed to

be rare and transitory, does gene copy number increase appreciably as a result

31

of the off-set timing of chromosomal replication and cell division in E. coli. Thus,
the potential for increased interaction based on the proximity of rRNA operons on
the chromosome seems to favor intrachromosomal interactions but in no way
excludes interchromosomal interactions from contributing to the observed
correlation.

The presence of sequence heterogeneity between the rRNA operons of E.
coli is probably a result of several factors. First, the lack of a large fitness benefit
associated with recombination events between rRNA operons provides a steep
barrier to the fixation of the recombinant operons in a population. Thus, very few
homogenizing events would be expected to rise to fixation in the population, but
the few that do become fixed result in the overall similarity seen in the rRNA
operons. In the case of E. coli, the rate of fixation of recombinants would have to
be lower than the rate of fixation of point mutations within the sequences
resulting in highly similar sequences but with some differences. Second, the
increased chance of recombination between operons located close to one
another would result in some very similar or identical sequences and some
operons that contain more sequence differences.

An additional mechanism for the appearance of heterogeneity between
rRNA operons other than point mutation is the inter-species lateral transfer of
rDNA (54). While not specifically considered by the model presented here,
lateral transfer has been suggested as the most likely mechanism for the
presence of highly heterogeneous (~10% pair-wise difference) rRNA genes in

several species and remains a potential source of introducing sequence

32

heterogeneity (17, 42, 53, 54). However, in our model, the origin of
heterogeneity within the rRNA operons is for the most part irrelevant because
once heterogeneity is present, it will be governed by the homogenizing force of
recombination in the same way as heterogeneity arising via point mutations.
Conclusions

We have provided direct measurements of recombination between the
rRNA operons of E. coli. A rate of ~5x10'9 per cell division for recombination
events involving rrnB was measured in two independent fluctuation experiments
and is in agreement with similar recombination events in related systems. The
fitness effects of recombination between rRNA operons resulting in the
reestablishment of an intact rRNA operon was found to be ~1%. These results
may be interpreted in an analogous situation where a mutation within a structural
rRNA gene arises resulting in a non—functional transcript. In this case, the
recombinational repair of such a lesion resulting in the reestablishment of a
functional rRNA transcript would be associated with an increase in fitness of
~1%.

The presence of highly similar, but not identical, rDNA sequences can be
explained by the rare chance of fixation of genotypes with a small fitness
advantage allowing for some homogenization of sequences through
recombination and heterogeneity arising by de novo mutation. In the case of E.
coli, the rate of mutation and fixation of such mutations giving rise to
heterogeneity must be higher than the rate of recombination and fixation of the

recombinants resulting in the heterogeneity that is present. These observations

33

in addition to the observation that some organisms have completely identical
rRNA operons but others have significant heterogeneity, raises the question of
why sequence heterogeneity exist in some organisms and not in others. The
degree of heterogeneity may simply be a result of the rates of recombination and
mutation and proximity to the other operons. There may also be ecological
explanations for the heterogeneity depending on the type of selection and the
frequency and severity of population bottlenecks resulting in the fixation of more
homogenization events or the increased divergence of certain operons resulting

in more heterogeneity.

34

Chapter 3
ALTERNATE METHODOLOGIES FOR THE REMOVAL OF RIBOSOMAL RNA

OPERONS FROM THE CHROMOSOME OF ESCHERICHIA COLI
3.1 Introduction

Microbial species with multiple copies of ribosomal RNA (rRNA) operons
have been proposed to employ an ecological strategy which differs from
organisms with one or a few copies of the rRNA operon (30). Specifically,
organisms with a high number of rRNA operons are expected to have the
transcriptional potential to quickly produce rRNA and ribosomes allowing a rapid
shift-up of growth upon the influx of nutrients resulting in shorter lag times and
higher growth rates. Organisms with one or a few copies of the rRNA operon are
thought to be more efficient at growing and surviving during periods of low
sustained nutrient levels but respond slower to an influx of nutrients. While these
observations seem to hold true in many different organisms (30), a causal link
between the copy number of rRNA operons and ecological strategy has yet to be
made.

Previous work sought to establish a direct link between rRNA operon copy
number and ecological strategy by sequentially deleting two of the seven rRNA
operons of E. coli (m8 and rmA) (49). In batch culture competition experiments
of the deletion strains versus the wild-type parental strain of E. coli B, there was
a decrease in relative fitness with each sequential deletion of a rRNA operon (i.e.
wt > ArrnB = AI'ITIA > ArrnAB) supporting the hypothesis that multiple rRNA

operons is beneficial during fluctuating nutrient conditions. In chemostat

35

competitions, however, where the dilution rate of the culture vessels was 0.11
per hour (dilution rate = growth rate) simulating a constant low nutrient
environment, ArmA and ArrnAB strains unexpectedly showed a pronounced
decrease in relative fitness (50). In parallel chemostat competitions, ArmB
strains showed no loss in fitness relative to the wild-type ancestor suggesting a
specific role for rrnA under slow growth conditions. Previous comparisons of
rRNA operon promoter strengths (P1, P2) suggested that rmH, rrnE, and rmG
may differ in their regulation from the other operons under some conditions (13).
Despite the evidence of differential regulation for some of the rRNA operons, it
was still surprising to see such a pronounced growth rate specific role for rmA.

A potential reason for the operon specific nature of the decreased fitness
lay in the sequence differences between rrnB and MA, the most obvious being
the presence of tRNAGluz in MB and tRNA”e1 and tRNAAmB in MA (6). By
deleting the rrnA operon, an imbalance may arise in the cellular tRNA pool under
slow growth conditions, where the tRNA"e1 and tRNAA'a‘B may be in higher
demand relative to tRNAGm resulting in a growth rate specific effect of deleting
rrnA. There are also sequence differences in the structural genes and in the
spacer regions that could result in transcripts that are more easily processed into
mature rRNA under certain conditions, or even ribosomes produced from the
different operons with slightly different translational functions.

An alternate explanation for the results explained above is that a mutation
or other disruption occurred during the construction of the ArmA and ArmAB

strains and that the effects are not attributable specifically to the rrnA operon. In

36

order to establish a direct link between the deletion of the rrnA operon and the
drop in relative fitness in slow dilution chemostat competitions, complementation
was necessary. Reestablishing a wild-type rmA operon in the ArmA and AmiAB
strains and observing a rescue of the wild-type or ArmB fitness under the given
conditions would establish the causal link between the rrnA operon and the
observed fitness loss in chemostat competitions.

In this chapter, I outline the construction of several strains using two
different approaches to complement the An'nAB strains and the subsequent
measurements of relative fitness. The shortcomings of the strains will be
discussed and a new approach to constructing rRNA operon deletions will be
introduced that may help establish the link between rRNA operon copy number

and fitness.

3.2 Materials and Methods
Cloning. XL-PCR. mid methocfi

Genomic DNA and plasmids were harvested from overnight cultures using
the QlAamp Tissue Kit from Qiagen (Santa Clarita, CA), and the V\fizard Plus 8V
Minipreps from Promega (Madison, WI), respectively. Restriction digestions
were carried out using restriction endonucleases from New England Biolabs
(Beverly, MA) according to the manufacturers recommended conditions. Gel
electrophoresis was carried out using 0.6-1.5% SeaKem LE Agarose (FMC;

Rockland, ME) in 0.5xTBE. When necessary, DNA fragments were extracted

37

from agarose gels using the Qiaex |l Gel Extraction Kit from Qiagen (Santa
Clarita, CA).

Amplification of genomic or plasmid DNA was carried out with thh XL-
DNA polymerase from PE Applied Biosystems (Foster City, CA) using purified
genomic or plasmid DNA. The following parameters were used in a typical 50 or
100ul amplification reaction: 94°C 3 min, 94° C 30”; 60° C 1 min; 72° C 10 min:
30 cycles, 72° C 10 min, hold at 4° C.

Cultures were typically grown in Luria-Bertani (LB) with 1.5% Bacto Agar
(Difco Laboratories; Detroit, MI) added for solid media. When necessary, the
following antibiotics were added at the specified concentrations: ampicillin (amp;
60ug/ml), chloramphenicol (cam; 30ug/ml), kanamycin (kan, Song/ml). For LB
sucrose media (LBsuc), a final concentration of 6% sucrose was added to LB
with NaCl omitted (8). All competition experiments and conditioning cultures
were carried out in Davis-Mingoli broth (DM) (33) supplemented with 2.0 mg/L
thiamine hydrochloride and glucose at a final concentration of 25 mg/l (DM+25).

Southern hybridization was used to visualize the rRNA operons using a
digoxigenin-dUTP-labeled DNA probe (Roche; Indianapolis, IN) for a conserved
region of rrs (positions 8-536 of 7738) (30).
Version 1plasmid constructions

The rrnA operon of BS4548 was amplified using the upA forward ( 5’-
GCGAATTACATATGCCCTCACGCCATCCTC‘IT'ITAT-3’) and dnA reverse (5’-
CTACTCTAGACGTTCGCATCCGGCATTTTT'I'T-3’) primers. The PCR reaction

mixture was subjected to a post amplification incubation at 72° C for 10 minutes

38

with Taq polymerase to add 3’ adenine overhangs and cloned into the Invitrogen
pCR2.1 cloning vector (Carlsbad, CA). The cloned rmA operon in pCR2.1
(p2.1A2) was moved into pUC19 and the temperature sensitive vector pMAK705
on a Xbal fragment (7.3 kb) producing pUCA21 and pMAKA14, respectively.

A series of plasmids were generated starting with pUCA21 with the
promoter region, the region spanning the nsA-lTS-rrlA, the rrsA gene, and the
niA gene deleted from the rrnA operon. These constructions are depicted and
verified by restriction digestion with Xbal in Figure 3.1 and summarized in Table

3.1.

39

10 kb—> ,..
8kb—s ‘

6kb—r
5kb——~

4kb—>
3kb—~

2kb——~
1.5 kb—~r

1kb——>-

 

 

 

 

0.5 kb—~ -
B.

P1 P2 ITS ter

rrs H rrl rrf
—IEI r01 W}—
I |
ASacll 3.5 kb
AAVI1l-Rsrll 0.9 kb APshAl 0.7 kb
AXcml 1.7 kb

Figure 3.1. Xbal digestions of pUCA21 plasmid constructions. rrnA was cloned
as an Xbal fragment into pUC19 giving pUCA21 L1 and L2. pUCA21
constructions were further digested with restrction endonucleases to yield the
resulting deletion strains. A. Xbal digestions of pUCA21 plasmids. B.
Restriction sites used in the creation of the pUCA21 deletion plasmids.

40

Table 3.1 Plasmids used for strain constructionss

 

 

Plasmid Relevant genotype or Source
phenotype
pUC19 Amp’, lacZ’ Pharmacia
pCR2.1-TOPO Kan', Amp', lacZ’ Invitrogen
pMAK705 Cam', lacZ’, rep‘s C.M. Hamilton (21)
pKO3 Cam’, sacB, rep“ GM. Church (35)
pCP15 Amp', FRT-Kan'-FRT D. Ammons (2)
cassette in pUC19
pCP20 Flp”, It cl857+, A pR rep‘s, D. Ammons (2)
Amp’, Camr
pKTS102 Cam', rep", UpAanA UpAanA fragment (Xbal,
2,609 bp) cloned in pMAK705
p2.1A2 Kan', Amp', rrnA rrnA cloned in pCR2.1-TOPO
pUCA21 Amp', rrnA miA cloned into pUC19 from
p2.1A2 on a Xbal fragment
pMAKA14 Cam', rep‘s, rrnA rrnA cloned into pMAK705

pUCA21ASaclI6

pUCA21AXcm|3

pUCA21APshAI1

pUCA21AAvrll-Rsrll

Amp', rrnAASacll

Amp', rrnAAXcml

Amp', nTiAAPshAI

Amp’, rrnAAAvrIl-Rsrll

from p2.1A2 on a Xbal
fragment

pUCA21 with the Sacll
fragment removed from rrnA
(ArrsA, ITS, rriA)

pUCA21 with the Xcml
fragment removed from rrnA
(promoters and upstream FIS
sites removed)

pUCA21 with the PshAl
fragment removed from MA
(AniA)

pUCA21 with the Avrll-Rsrll
fragment removed from rrnA
(ArTsA)

 

41

Table 3.2 E. coli Strain Constructions

 

 

Strain Relevant genotype Source
and phenotype
TOP10 lac', recA, endA1 Invitrogen
REL4548 Ara' R.E. Lenski (33)
BS4548 Ara+ B.8. Stevenson (49)
B8301 (ArmAB) Ara‘, ArrnA, ArrnB B.S. Stevenson (49)
88300 (AITI’IAB) Ara', ArrnA, ArrnB B.S. Stevenson (49)
BS301p11 Ara‘, ArrnA, ArrnB, P1 transductant of B8301 moving the
Sues, Kanr sacB-neo cassette into the MA deleted
region
BS300p11 Ara', ArmA, ArTnB, P1 transductant of B8301 moving the
sacB, Kanr sacB-neo cassette into the rrnA deleted
region
B8311 Ara‘“, rrnA‘, ArrnB, B.S. Stevenson; intermediate strain for
Sucs, Kanr B8301 construction, upAzsacB-neozdnA
transduced into rrnA
B8310 Ara’, miA', ArrnB, B.S. Stevenson; intermediate strain for
Sues, Kanr B8301 construction, upAzsacB-neozdnA
transduced into rrnA .
B8200 (ArrnB) Ara”, ArrnB B.8. Stevenson
A+KOBA* Ara‘, ArTnB, allelic exchange of upAzsacB-neozdnA of
(HSBA1) rrnA reinserted BS301p11 with rmA of pMAKA14
A-KOBA* Ara', ArmB, allelic exchange of upAzsacB-neozdnA of
(HSBAZ) rrnA reinserted BS300p11 with rmA 0f pMAKA14
A+JHB122 Ara‘“, ArrnB, allelic exchange of upAzsacB-neozdnA of
rrnA reinserted B8311 with rrnA of pMAKA14
A-JHB122 Ara', ArrnB, allelic exchange of upAzsacB-neozdnA of
rrnA reinserted B8310 with rrnA of pMAKA14
A+JHAB621 Ara“, AITI‘IB, ArrnA allelic exchange of upAzsacB-neozdnA of
B8311 with upA:dnA of pKTS102
A-J HAB311 Ara', ArmB, ArmA allelic exchange of upAzsacB-neozdnA of

88310 with upA:dnA of pKTS102

 

42

Version 1 strain construction

The construction of the rrnA complementation strains A+KOBA* and A-
KOBA" was carried out as previously reported (49) and depicted in Figure 3.2.
The upAzsacB-neozdnA region of DAS113, which is comprised of upstream
sequence (upA) with homology only to the rrnA operon, was transduced by the
phage P1vir into the strains B8300 (Ara'; ArrnAB) and B8301 (Ara*;ArmAB).
Transductants were selected by kanamycin resistance, screened for sucrose
sensitivity, and verified by PCR amplification of the rrnA region producing a
specific band indicative of rrnA containing the sacB-neo cassette. The resulting
intermediate strains (B8300p11, BS301p11) were transformed with pMAKA14.
Growing the transformants at the non-permissive temperature of 42° C in the
presence of chloramphenicol forces cointegrate formation at the MA region of
the chromosome. Growth of cointegrates at the permissive temperature of 30° C
favors the excision of the plasmid due to the detrimental effects of maintaining
multiple active origins of replication within the chromosome (2, 21). Growing
cointegrates at 30° C in the presence of 6% sucrose favors plasmid excision with
the removal of the sacB-neo cassette leaving the wild-type nnA operon in its
normal chromosomal location (8, 21, 49).

Isolates with the proper phenotype (sucrose', cams, kan‘) were verified by
PCR amplification of the rrnA operon region (Figure 3.3) and Southern
hybridization (Figure 3.4). The strains that had undergone the proper exchange

are designated Ara‘KOBA* and Ara'KOBA* (Table 3.2).

43

1. P1 Transduction

—[ upA IsacB-neo ldnA]————
\ /

 

 

I upA ldnA'r
B8301 Chromosome

 

 

2, I Select transductants (sues-kanf)

Chromosome (Intermediate strain)
—[ upA [sacB-neo ldnA}

 

 

 

dnA

 

 

‘
. .
. _ .- u
- c-
.. 4, _ .. o
- .. - ‘ - a u
' "‘ " C - . v a o v - q n n o . a c I c o u o .- '

transform with pMAKA14

1 select for cointegrate formation
(growth at 42° C)

qu _II 233 ‘ ‘ --1S-0ri---Cm'..

 

 

 

I dnAIsacB-neol upA I— : ' '

4' 1 select for plasmid excision and curing
(growth at 30° C in LB+sucrose)

 

 

_~ gcamr“ '——-I upA [sacB-neo ldnA} ..... $-90.

- - ’
-
- — __________

‘4 '* III ‘- o a a a a . p o - n c o a a - — n r- o n - o v ‘-

 

 

 

 

 

KOBA* chromosome

Figure 3.2. Diagram of steps involved in the complementation of B8301 and
B8300 creating A+KOBA* and A-KOBA*. The creation of JHB and JHAB strains
follow these steps starting at step 3 using plasmids pMAKA14 and pKTS102,
respectively.

44

 

Figure 3.3. PCR amplification of the rrnA region of the complementation and
deletion strains. The rrnA region amplifies as a 7.3 kb fragment as indicated in
the BS4548 lane. The deleted rrnA region amplifies as a 2.6 kb fragment. All
band mobilities for the deletion and complementation strains show the expected
snze.

45

A- JHAB311
A+J HAB621

(\IN
ts
7+
<<

A-KOBA*
A+KOBA*

rrnB —~

rrnA —>

 

Figure 3.4. Southern hybridization of Pvull digested genomic DNA from rrnA
deletion and complementation strains probed for a conserved region of rrs. All
strains have rrnB deleted. A+KOBA* and A-KOBA* are the complementation
strains starting from 88301 and 88300 with the rrnA operon reinserted as
outlined in Figure 3.2. A+JHB122, A-JHB122, A+JHA8621, and A-JHA8311
were constructed using the intermediate strains 88311 and 88310 with the Ara+
and Ara- marker, respectively. JHB strains have rrnA reinserted into the
chromosome. JHAB strains have rrnA deleted from the chromosome.

46

Version 2 strain construction

A second set of strain constructions with rrnA deleted and complemented
started with strains 88311 and 88310. Strains 88311 and 88310 are the
intermediate strains that were originally used to generate the deletion of rrnA
which had the unexpected low fitness in chemostat competitions (49), and have
rrnA replaced by upAzsacB-neozdnA by P1 transduction. Both strains were
transformed with either pMAKA14 or pKT8102 to replace or delete the rrnA
operon, respectively. Transformants were subjected to the same regime outlined
for the Version 1 constructions outlined in Figure 3.2. Strains with both nnA and
rrnB deleted starting from the intermediate strains 88311 and 88310 are named
A+JHAB621 and A-JHA831, respectively. Strains with rrnB deleted but with a
normal intact rrnA operon starting from intermediate strains 88311 and 88310
are named A+JHB122 and A-JH8122, respectively (Table 3.2).
Chemostat competitions

Relative fitness measurements under slow growth, continuous culture
were carried out in eight parallel chemostat vessels as previously described (49).
Relative fitness was measured as the change in the ratio of the two competitors
started at a 1:1 ratio. The two competitors are differentiable by a neutral marker
(arabinose utilization) where Ara+ colonies appear white and Ara‘ colonies

appear red on TA agar (33).

47

Batch culture competitions

Separate overnight cultures of the two competitors were inoculated into L8
from 20% glycerol freezer-stocks. Conditioning cultures, consisting of 10ml of
DM+25 in a 50 ml Erlenmeyer flask, were inoculated the following day with 3
1210000 total dilution from the overnight culture. After 24 hours shaking at 37°
C, the two conditioning cultures were mixed in a 1:1 ratio and used to inoculate
replicate 50ml flasks containing 10ml of DM+25 at a final dilution of 1:100. Batch
culture competitions were carried out in 50ml Erlenmeyer flasks containing 10ml
of DM+25 and were allowed to continue for 3-5 days with a sampling and transfer
every 24 hours. Competitors were differentiated by the ability or inability to utilize

arabinose on TA plates (33, 49).

3.3 Results
Version 1 competitions

The relative fitness of A+KOBA* (HSBA1) versus 88300 (ArmAB) in a
slow dilution chemostat showed that the complemented strain A+KOBA* has a
small fitness advantage over 88300 which has both rrnA and rrnB deleted as
evidenced by the positive slope of the KOBA* vs. KOAB line in Figure 3.5. The
chemostat competition between A+KOBA* and 88200 (ArmB), however, showed
that the presence of an intact rrnA operon in A+KOBA* does not rescue the large
fitness loss associated with the deletion of rrnA. Further analysis shows that the
relative fitness of A+KOBA* is nearly identical to that of 88301 (KOAB) versus

88200 (KOB)(Figure 3.5).

48

 

 

 

 

 

 

 

I
I
I
2.0-*ivm-————~A~~—~AWW WWWW]
I
$1.57 *'*EE~A ‘ * *‘*~‘—I
g
+ T
E .
< 1.0 _ £_ 7 + ,,_ v ,W,,A V” I
%""‘"""T"T I x"- KOAB vs KOB
\ I -+- KOBA*vs KOA%
I ,- \I T |"‘~e*“-KOBA*vs KOB .
y \X .—. .4, . .. 4 ‘
0 __ * fl - _-__ e _’ _ !
O 5 10 15 20
doubfings

Figure 3.5. Chemostat competitions of complementation strain A+KOBA*. The
A+KOBA* vs. A-KOAB competition suggests the presence of rrnA provides a
small fitness advantage. The A+KOBA* vs. A-KOB competition shows the fitness
loss seen in the KOAB strain (88301, 88300) is not complemented by the
reinsertion of rrnA. The KOAB vs. KOB competition represents data from
reference (49) and is included to show the similarity to A+KOBA* vs. KOB
competition. All chemostat competitions were run in four parallel chemostat
vessels with each time point representing the mean ratio of Ara+ to Ara- with the
error bars representing plus and minus one stadard deviation. The chemostats
were run at an average dilution rate of 0.11 hrl.

49

Batch culture competition results are summarized in Figure 3.6. The
relative fitness of the complemented strain A-KOBA* (H88A2) versus the wild
type REL4548 strain is significantly lower than the strain 88300 which has both
rmA and rrnB deleted. These results along with the chemostat competition
results suggests that the large fitness loss associated with the deletion of the
rrnA operon in slow growth chemostats is not directly attributable to the presence
or absence of the rrnA operon.

Version 2 batch cultgre competitions

The results of the batch culture competitions involving the second version
of deletion and complementation strains show no clear pattern associated with
the presence or absence of the rrnA operon (Figure 3.7). The results show a
large difference between the Ara+ and Ara- strains in control competitions with
the same number of rRNA operons deleted. Surprisingly, the Ara+ strains were
always more than 7% more fit than the Ara- strains regardless of the number of
intact rRNA operons present. Due to the large amount of variability between
strains with the same number of rRNA operons, chemostat competitions were not

conducted for the JHB and JHAB strains.

50

relative fitiness (A+/ A—)

1.06: e -—

 

 

 

I
I _r_ _

 

 

 

 

AFControl vs. A-KOBA‘°

 

 

 

 

L

AFControl vs. A-KOAB

Figure 3.6. Relative fitness of A-KOBA* and A-KOAB strains. A+Control is the
parental strain (BS4548). Error bars represent +/- one standard deviation. The
A-KOBA* competition clearly indicates that the reinsertion of rrnA does not

improve fitness.

51

1.18r~ i i- 7 v 7

 

1.16l —— ~-—~~~--#— ~ If A - f A * —-—~-* j
1.14l -—~~ ea ~——— * '

 

1.1mm; # f we fl *. imam ._ - -. __l

 

 

 

 

 

1.08 } *--- ——---— -—— — -—-— .___ ~—-- - -_..._. . _..-.____- .__fi:._.___ .

 

 

 

1.10 l a. e r - sf _. ,+ -- ‘_ _,._. -11'
l
1.06L 4 a — ~~‘ +A— ,+. —— l

 

 

 

 

1.04 l.____ —~ --__ ——— ,W--- —- --

relative fitness (A+/A—)

l
w !
m.+.isfls_aw+;—~—e—-w—~l

1.00 i -— —-———--—— ___ - . - ..__. ____._ . _____ __

0.9811 A -A -f ! .. — ‘_ ...... _“_ .#_ a. _ . l
961 , . _s , 4.
A+JHAB vs A- JHAB A+JHB vs A-JHB A+JHAB vs A-JHB A+JHB vs A-JHAB

 

 

 

 

 

 

 

 

 

 

 

Figure 3.7. Relative fitness in batch culture of JHAB and JHB strains. Error bars
represent +/- standard deviation. The large difference between the control
competitions (A+JHAB vs. A-JHAB and A+JHB vs. A-JHB) indicate that a large
amount of variability is introduced during the creation of the strains.

52

3.4 Discussion
Lack of complementation in rrnA replacement strair§

The chemostat competitions with A+KOBA* (HSBA1) clearly show that the
large fitness loss in seen for 88301 and 38300 in the slow dilution chemostat is
not a direct result of the deletion of the MA operon (Figure 3.5). The relative
fitness of A+KOBA* (ArrnAB with rrnA reinserted) is only slightly better than
88300 (ArmAB) which, if the large fitness loss seen in the strains with rrnA
deleted (38300, 88301) were a result of the deletion, should have a relative
fitness near wild-type levels. The chemostat competition between A+KOBA*
(HSBA1) and 88200 (ArmB) supports the observation that inserting the MA
operon into 88301 at the original chromosomal location does not rescue the
fitness loss.

Batch culture competitions with A-KOBA* (HSBA2) showed that
reinserting the rrnA operon in 88300 caused a decrease in relative fitness
(Figure 3.6). This is contrary to what was expected and indicates that the fitness
loss in strains bearing deletions of m8 and MA are not specifically attributable
to the loss of rRNA operons (49). The chemostat and batch culture competition
results indicate that something occurred during the construction of these strains
that had a large impact on the fitness of the strains and the fitness effects

associated with the deletion of rRNA operons is lost in the background variability.

53

Rational for constructing rrnA deletions and replacements from the s_ar_n_§
transductant

The lack of complementation seen in the KOBA* competitions shows that
the large fitness loss in the rrnA deletion strains 88301 and 88300 in chemostat
competitions is attributable to something other than the presence or absence of
the rrnA operon. We proposed that the transduction event required to move the
sacB-neo cassette inserted in rrnA from DAS113 (recD’) into the well
characterized REL4548 and BS4548 (Figure 3.2) strains was the likely cause of
the large fitness loss. Generalized transduction using the bacteriophage P1
mobilizes ~80kb from the donor to the recipient (39). For most genetic
manipulations this should not be a problem because the resulting strains are not
going to be tested for the small fitness changes that are resolved by head-to-
head competitions. However, it is apparent that the fitness effects of rRNA
operon deletion will be small and the transfer of such large fragments of DNA by
P1 transduction may have a large unanticipated effect on fitness. The results
can also be interpreted as each P1 transduction decreasing fitness regardless of
what is being deleted or complemented.

Vlfith this in mind, it was proposed that the variability introduced by the
generalized transduction could be overcome by creating both deletion and
complementation strains starting from the intermediate strains 88311 and
88310, which contain the sacB-neo cassette already transduced into the rrnA
region, that were used to create the original deletions 88300 and 38301 (Table

3.2). The JHB and JHAB strains were constructed in this fashion creating the

54

deletion and complementation strains in parallel allelic exchange experiments
starting with the same transductant. The effects of deleting MA was expected to
be small, but by starting with the same intermediate strain all of the fitness
differences was expected to be attributable specifically to the presence or
absence of rrnA.

The JHB and JHAB batch culture competition results were rather
confounding and not much information regarding the fitness effects of deleting
rrnA could be gleaned from them (Figure 3.7). The most obvious difference seen
in the strains is that the Ara+ strains derived from 88311 have a higher relative
fitness regardless of the presence or absence of the rrnA operon. Another
striking observation is the difference in relative fitness of Ara+ vs. Ara- between
the two control competitions (A+JHAB vs. A-JHAB and A+JHB vs. A-JHB). The
amount of difference between the control competitions suggests that a change of
around 7% is introduced just by taking the strains through the allelic exchange
process. These results suggests that the steps involved in the allelic exchange,
which includes several population bottlenecks, introduces enough variation in
fitness to mask any specific fitness effects of rRNA operon deletion.

Eu_tt_1re directions

Clearly, the results from the complementation analysis indicates that the
approach to rRNA operon deletion must be rethought. If at all possible, P1
transduction should be avoided and the number of steps required to go from start
to finish should be limited. One approach that seems to fit both of the above

qualifications is currently being worked out in the lab. The new approach uses a

55

site specific recombinase (F lp) to remove a kanamycin resistance cassette
flanked by F RT sites (Flp recognition sequences) from a deleted region of the
chromosome (Figure 3.8).

As I stated before, the alternative protocol , developed by Ammons et. al
(2), does not use generalized transduction to generate an intermediate strain and
limits the number of steps required for the allelic exchange (Appendix 1). The
main drawback to this approach is the limited ability to complement the rm
deletions. The high degree of homology between the seven rRNA operons of E.
coli would greatly increase the probability of non-specific integration and the
presence of the partial FRT sequence after Flp mediated excision of the antibiotic
resistance cassette may disrupt the normal operon function.
Conclusions

The fitness effects of rRNA operon deletion in E. coli are elusive. The
glimpse that has been gained through this and other studies have indicated that
the fitness effects of deleting one rRNA operon from E. coli will be small. Other
groups have succeeded in generating strains that have used antibiotic insertions
and partial deletions of the rRNA operons but maintained transcriptionally active
promoters (5, 11). However, P1 transduction has been used in the creation of
these strains which, in addition to the presence of antibiotic resistance genes and
active rrn promoters, complicates the interpretations. Thus, an accurate picture
of the effects of rRNA operon deletion has yet to be shown with respect to the
ecological and evolutionary importance of maintaining multiple rRNA operons.

The use of the strategy employed by Ammons (2) to generate rRNA operon

56

deletions should give the cleanest system to elucidate the importance of multiple
rRNA operons in E. coli. Furthermore, it may be fruitful to explore the fitness
effects of rRNA operon deletion in organisms with fewer rRNA operons than the
seven which are present in E. coli. By deleting an rRNA operon from an organism
with fewer rRNA operons, the effect on fitness would presumably be larger due to
the smaller range of compensation that would be afforded with a smaller

complement of rRNA operons.

57

Figure 3.8. Ammons-Fox protocol for generating chromosomal deletions.
Plasmid and strain information is summarized in Tables 3.1 and 3.2. The first
two steps generate the required plasmids with the FRT-kan'—FRT cassette
inserted between the up/down sequence allowing site-specific integration of the
pasmid into the E. coli chromosome. The plasmid pKO3 has a temperature
sensitive origin of replication (ts-ori) and the $308 gene from Bacillus subtilis
conferring sucrose sensitivity. E. coli B strains REL4548 and BS4548 are then
transformed with pKO3-delA (or -delB) and cointegrates are isolated by growing
transformants at 42°C in the presence of kanamycin and chloramphenicol.
Growth of cointegrants at 30°C in the presence of 6% sucrose and kanamycin
selects cells with the plasmid excised and lost from the cell due to the lethal
effects of the sacB gene in the presence of sucrose. Kanamycin resistant cells
are then transformed with pCP20 carrying the Flp recombinase resulting in the
site-specific excision of the kanr gene flanked by FRT sites (Flp recognition
sequence). After Flp removal of the kanr gene, a partial FRT sequence remains
at the site of the deletion.

58

 

 

 

FR ka_nr A up [A down
< pCP15 > pUAD
\

/

Clone FRT-kanr-FRT cassette
between Aup-Adown in plasmid
pUAD

V

Move the deletion
allele into pK03

 

 

 

 

 

 

 

Transform REL strain with
pK03-delA

REL Genomic

 

DNA ——l AXue I 168 IlTSl g3s lAdownl—
Aup lFRTl

 

 

59

 

Grow transformants at 42° C with
kanamycin and chloramphenicol to
force cointegrate formation

 

 

 

 

__{ Aup [FRTL kan" IFRTIA down} ts-ori chloramphenicolr sacB
——mdownl 23$ IITSI 168 l Aup

Grow cells at 30° C in the presence of
sucrose and kanamycin to force
plasmid excision and curing and
select for the desired excision

type (kanr)

 

 

REL chromosomal up FRT [any FRT own.—

DNA
Q 238 [ITS] 16$ rAu

Transform kanr, sucroser cells with
pCP20 at 30° C. Cure cells of the
plasmid by growing at 42° C

 

 

  

 

 

F‘RTI kanr IFRT

 

 

{ Aup I lAdown}

 

60

Chapter 4
SUMMARY AND CONCLUSIONS

The initial aim of the research contained in this thesis was to complement
strains of E. coli that had chromosomal deletions of rrnA. The low relative fitness
phenotype in slow dilution chemostat seen in the rrnA deletion strains was
expected to be reversed with the reinsertion of rrnA at its normal chromosomal
position. The specific region of the rrnA operon that was required for the
complementation was then to be determined by reinserting portions of the rrnA
operon (Figure 3.1). However, the wild-type fitness was not rescued using two
different strain constructions with a wild-type rrnA operon reinserted into the
chromosome. It is likely that the use of the generalized transducing phage P1,
which transfers ~80kb of DNA (39), was responsible for a portion of the fitness
loss associated with the original rmA deletion. Another source of variation in the
relative fitness of the strain constructions is the high number of population
bottlenecks that are encountered during the deletion and complementation
procedure. An alternate procedure for generating chromosomal deletions of the
rRNA operons of E. coli that does not use P1 transduction and limits the number
of population bottlenecks is currently being pursued in the lab. Due to the lack of
interpretable results from the complementation studies, I shifted the focus of my
thesis research to the related topic of measuring the rate of and fitness effects of
recombination between rRNA operons.

Recombination between the rRNA operons of E. coli had previously been

measured (22, 24, 25, 27) but our system is unique in allowing the easy isolation

61

of recombinants with no other chromosomal alterations. We found that
recombination events involving rrsB resulting in the removal of a 3.8 kb sacB-neo
insertion occurred at a rate of ~5x10'9 per cell division. The loss of the sacB-neo
cassette from rrnB by recombination resulted in a 1% increase in fitness
compared to the parental strain (BSV1). The increase in fitness of the
recombinant isolates may be attributable to the loss of the sacB-neo cassette
from the cell, regaining the full complement of rRNA operons, or a combination of
both possibilities. The lack of any fitness cost associated with recombination
also suggests that the sequence heterogeneity is not selectively maintained and
alternate explanations for the presence of heterogeneity in the rRNA operons of
E. coli should be explored.

Our findings regarding the fitness of the recombinant isolates were placed
in the context of a conceptual model for the generation and maintenance of
sequence heterogeneity in the rRNA operons of E. coli. In this model, mutation
gives rise to heterogeneity and recombination results in the homogenization of
the sequences. However, with such small fitness effects associated with the
homogenization of the sequences, the fixation of such events in the population
would be very rare. Thus, the relative rates of recombination and mutation and
the rate and probability of fixation of the recombinant or mutant in the population
will result in operons with either more or less heterogeneity. An additional
variable in the model was added based on observations of the degree of
sequence heterogeneity in 168 and 238 sequences of published genomes. A

positive correlation exists between the number of pair-wise differences of 168

62

and 23S sequences and the chromosomal distance between the operons (Figure
2.7). Thus, heterogeneity may also persist if an operon is located further away
from other rRNA operons with the closely located operons showing higher
degrees of sequence similarity.

The redundancy of the rRNA operons in E. coli allows the cell to adjust to
the deletion of operons allowing near normal growth rates. This redundancy may
also allow mutations within the coding sequences that disrupt the secondary or
tertiary structure of the rRNA and persist until compensatory mutations occur.
Thus, the rRNA operons provide an intriguing system to broaden our
understanding of the physiological implications of multiple rRNA operons, the
dynamics and interplay of coordinately evolving gene families, and, potentially,

the conditions required to escape the homogenizing effects of recombination.

63

Literature Cited

Abdulkarim, F., and D. Hughes. 1996. Homologous recombination
between the tuf genes of Salmonella typhimurium. J Mol Biol. 260:506-22.

Ammons, D., J. Rampersad, and G. E. Fox. 1999. 5S rRNA gene
deletions cause an unexpectedly high fitness loss in Escherichia coli.
Nucleic Acids Res. 27:637-42.

Anton, A. l., A. J. Martinez-Murcia, and F. Rodriguez-Valera. 1999.
lntraspecific diversity of the 23S rRNA gene and the spacer region
downstream in Escherichia coli. J Bacteriol. 181:2703—9.

Anton, A. l., A. J. Martinez-Murcia, and F. Rodriguez-Valera. 1998.
Sequence diversity in the 168-238 intergenic spacer region (lSR) of the
rRNA operons in representatives of the Escherichia coli ECOR collection.
J Mol Evol. 47:62-72.

Asai, T., C. Condon, J. Voulgaris, D. Zaporojets, B. Shen, M. Al-Omar,
C. Squires, and C. L. Squires. 1999. Construction and initial
characterization of Escherichia coli strains with few or no intact
chromosomal rRNA operons. J Bacteriol. 181:3803-9.

Bachellier, 8., E. Gilson, M. Hofnung, and C. W. Hill 1996. Repeated
Sequences, p. 2012-2040. In F. C. Neidhardt, J. L. lngraham, K. B. Low,
B. Magasanik, M. Schaechter, and H. E. Umbarger (eds), Escherichia coli
and Salmonella typhimurium: Cellular and Molecular Biology. American
Society for Microbiology, Washington, DC.

Blattner, F. R., G. Plunkett, 3rd, C. A. Bloch, N. T. Perna, V. Burland,
M. Riley, J. Collado-Vides, J. D. Glasner, C. K. Rode, G. F. Mayhew, J.
Gregor, N. W. Davis, H. A. Kirkpatrick, M. A. Goeden, D. J. Rose, B.
Mau, and Y. Shao. 1997. The complete genome sequence of Escherichia
coli K-12. Science. 277: 1 453-74.

Blomfield, l. C., V. Vaughn, R. F. Rest, and B. I. Eisenstein. 1991.

Allelic exchange in Escherichia coli using the Bacillus subtilis sacB gene
and a temperature-sensitive pSC101 replicon. Mol Microbiol. 5:1447-57.

64

10.

11.

12.

13.

14.

15.

16.

17.

18.

Bult, C. J., 0. White, G. J. Olsen, L. Zhou, R. D. Fleischmann, G. G.
Sutton, J. A. Blake, L. M. FitzGerald, R. A. Clayton, J. D. Gocayne, A.
R. Kerlavage, B. A. Dougherty, J. F. Tomb, M. D. Adams, C. l. Reich,
R. Overbeek, E. F. Kirkness, K. G. Weinstock, J. M. Merrick, A.
Glodek, J. L. Scott, N. S. M. Geoghagen, and J. C. Venter. 1996.
Complete genome sequence of the methanogenic archaeon,
Methanococcus jannaschii. Science. 273:1058-73.

Cilia, V., B. Lafay, and R. Christen. 1996. Sequence heterogeneities
among 168 ribosomal RNA sequences, and their effect on phylogenetic
analyses at the species level. Mol Biol Evol. 13:451-61.

Condon, C., S. French, C. Squires, and C. L. Squires. 1993. Depletion
of functional ribosomal RNA operons in Escherichia coli causes increased
expression of the remaining intact copies. Embo J. 12:4305-15.

Condon, C., D. Liveris, C. Squires, l. Schwartz, and C. L. Squires.
1995. rRNA operon multiplicity in Escherichia coli and the physiological
implications of rm inactivation. J Bacteriol. 177:4152-6.

Condon, C., J. Philips, Z. Y. Fu, C. Squires, and C. L. Squires. 1992.
Comparison of the expression of the seven ribosomal RNA operons in
Escherichia coli. Embo J. 11:4175-85.

Condon, C., C. Squires, and C. L. Squires. 1995. Control of rRNA
transcription in Escherichia coli. Microbiol Rev. 59:623-45.

Crane, G. J., S. M. Thomas, and M. E. Jones. 1996. A modified Luria-
Delbruck fluctuation assay for estimating and comparing mutation rates.
Mutat Res. 354:171-82.

Deckert, G., P. V. Warren, T. Gaasterland, W. G. Young, A. L. Lenox,
D. E. Graham, R. Overbeek, M. A. Snead, M. Keller, M. Aujay, R.
Huber, R. A. Feldman, J. M. Short, G. J. Olsen, and R. V. Swanson.
1998. The complete genome of the hyperthermophilic bacterium Aquifex
aeolicus. Nature. 392:353-8.

Dennis, P. P., S. Ziesche, and S. Mylvaganam. 1998. Transcription
analysis of two disparate rRNA operons in the halophilic archaeon
Haloarcula man'smortui. J Bacteriol. 180:4804-13.

Dvorak, J., D. Jue, and M. Lassner. 1987. Homogenization of tandemly

repeated nucleotide sequences by distance-dependent nucleotide
sequence conversion. Genetics. 116:487—98.

65

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

Fleischmann, R. D., M. D. Adams, 0. White, R. A. Clayton, E. F.
Kirkness, A. R. Kerlavage, C. J. Bult, J. F. Tomb, B. A. Daugherty, J.
M. Merrick, and et al. 1995. Whole-genome random sequencing and
assembly of Haemophilus influenzae Rd. Science. 269:498-512.

Fraser, C. M., S. J. Norris, G. M. Weinstock, 0. White, G. G. Sutton, R.
Dodson, M. Gwinn, E. K. Hickey, R. Clayton, K. A. Ketchum, E.
Sodergren, J. M. Hardham, M. P. McLeod, S. Salzberg, J. Peterson, H.
Khalak, D. Richardson, J. K. Howell, M. Chidambaram, T. Utterback,
L. McDonald, P. Artiach, C. Bowman, M. D. Cotton, J. C. Venter, and
et al. 1998. Complete genome sequence of Treponema pallidum, the
syphilis spirochete. Science. 281:375-88.

Hamilton, C. M., M. Aldea, B. K. Washburn, P. Babitzke, and S. R.
Kushner. 1989. New method for generating deletions and gene
replacements in Escherichia coli. J Bacteriol. 171:4617-22.

Harvey, S., and C. W. Hill. 1990. Exchange of spacer regions between
rRNA operons in Escherichia coli. Genetics. 125:683-90.

Harvey, S., C. W. Hill, C. Squires, and C. L. Squires. 1988. Loss of the
spacer loop sequence from the rmB operon in the Escherichia coli K-12
subline that bears the relA1 mutation. J Bacteriol. 170:1235-8.

Hill, C. W., R. H. Grafstrom, B. W. Hamish, and B. S. Hillman. 1977.
Tandem duplications resulting from recombination between ribosomal
RNA genes in Escherichia coli. J Mol Biol. 116:407-28.

Hill, C. W., and B. W. Hamish. 1981. lnversions between ribosomal RNA
genes of Escherichia coli. Proc Natl Acad Sci U S A. 78:7069-72.

Hill, C. W., and B. W. Harnish. 1982. Transposition of a chromosomal
segment bounded by redundant rRNA genes into other rRNA genes in
Escherichia coli. J Bacteriol. 149:449-57.

Hill, C. W., S. Harvey, and J. A. Gray 1990. Recombination between
rRNA genes in Escherichia coli and Salmonella typhimurium, p. 335-340.
In K. Drlica, and M. Riley (eds), The Bacterial Chromosome. American
Society of Microbiology, Washington, DC.

Hughes, D. 1999. Impact of homologous recombination on genome
organization and stability. In R. L. Charlebois (ed.), Organization of the
Prokaryotic Genome. American Society for Microbiology. Washington,
DC.

66

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Jones, M. E., S. M. Thomas, and A. Rogers. 1994. Luria—Delbruck
fluctuation experiments: design and analysis. Genetics. 136:1209-16.

Klappenbach, J. A., J. M. Dunbar, and T. M. Schmidt. 2000. rRNA
operon copy number reflects ecological strategies of bacteria. Appl
Environ Microbiol. 66:1328-33.

Koch, A. L. 1971. The adaptive responses of Escherichia coli to a feast
and famine existence. Adv Microb Physiol. 6:147-217.

Kunst, F., N. Ogasawara, l. Moszer, A. M. Albertini, G. Alloni, V.
Azevedo, M. G. Bertero, P. Bessieres, A. Bolotin, S. Borchert, R.
Borriss, L. Boursier, A. Brans, M. Braun, S. C. Brignell, S. Bron, S.
Brouillet, C. V. Bruschi, B. Caldwell, V. Capuano, N. M. Carter, S. K.
Choi, J. J. Codani, I. F. Connerton, A. Danchin, and et al. 1997. The
complete genome sequence of the gram-positive bacterium Bacillus
subtilis. Nature. 390:249-56.

Lenski, R. E., M. R. Rose, 8. C. Simpson, and S. C. Tadler. 1991. Long-
term experimental evolution in Escherichia coli: l. Adaptation and
divergence during 2,000 generations. The American Naturalist. 138:1315-
1 341 .

Liao, D. 2000. Gene conversion drives within genic sequences:
Concerted evolution of ribosomal RNA genes in Bacteria and Archaea. J
Mol Evol. 51:305-317.

Link, A. J., D. Phillips, and G. M. Church. 1997. Methods for generating
precise deletions and insertions in the genome of wild-type Escherichia

coli: application to open reading frame characterization. J Bacteriol.
179:6228—37.

Lloyd, R. G., and K. B. Low 1996. Homologous Recombination, p. 2236-
2255. In F. C. Neidhardt, J. L. lngraham, K. B. Low, B. Magasanik, M.
Schaechter, and H. E. Umbarger (eds), Escherichia coli and Salmonella
typhimurium: Cellular and Molecular Biology. American Society for
Microbiology, Washington, DC.

Lovett, S. T., T. J. Gluckman, P. J. Simon, V. A. Sutera, Jr., and P. T.
Drapkin. 1994. Recombination between repeats in Escherichia coli by a
recA-independent, proximity-sensitive mechanism. Mol Gen Genet.
245:294-300.

Luria, S. E., and M. Delbruck. 1943. Mutations of bacteria from virus
sensitivity to virus resistance. Genetics. 28:491-511.

67

39.

40.

41.

42.

43.

44.

45.

46.

47.

Milkman, R. 1999. Gene transfer in Escherichia coli, p. 291-309. In R. L.
Charlebois (ed.), Organization of the Prokaryotic Chromosome. American
Society for Microbiology. Washington, DC.

Mitchell, M. B. 1955. Aberrant recombination of pyridooxine mutants of
Neurospora. Proc. Natl Acad. Sci. USA. 41:215-220.

Mitchell, M. B. 1955. Further evidence of aberrant recombination in
Neurospora. Proc. Natl Acad. Sci. USA. 41:935-937.

Mylvaganam, S., and P. P. Dennis. 1992. Sequence heterogeneity
between the two genes encoding 16S rRNA from the halophilic
archaebacterium Haloarcula marismortui. Genetics. 130:399-410.

Nakamura, Y., T. Kaneko, M. Hirosawa, N. Miyajima, and S. Tabata.
1998. CyanoBase, a www database containing the complete nucleotide
sequence of the genome of Synechocystis sp. strain PCC6803. Nucleic
Acids Res. 26:63—7.

Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D.
Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K.
Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A.
Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S.
Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-
borne pathogen Campylobacterjejuni reveals hypervariable sequences.
Nature. 403:665-8.

Segall, A. M., and J. R. Roth. 1994. Approaches to half-tetrad analysis in
bacteria: recombination between repeated, inverse-order chromosomal
sequences. Genetics. 136:27-39.

Silvestri, M. L., W. J. Siqueira, A. A. de Souza, A. P. de Souza, M. F.
Terenzi, D. Truffi, S. M. Tsai, M. H. Tsuhako, H. Vallada, M. A. Van
Sluys, S. Verjovski-Almeida, A. L. Vettore, M. A. Zago, M. Zatz, J.
Meidanis, and J. C. Setubai. 2000. The genome sequence of the plant
pathogen X ylella fastidiosa. Nature. 406: 1 51 -7.

Singer, B. S. 1988. On the role of homologous sequences in
chromosomal rearrangements. Genes Dev. 2: 1 800-1 1.

68

48.

49.

50.

51.

52.

53.

54.

Smith, D. R., L. A. Doucette-Stamm, C. Deloughery, H. Lee, J. Dubois,
T. Aldredge, R. Bashirzadeh, D. Blakely, R. Cook, K. Gilbert, D.
Harrison, L. Hoang, P. Keagle, W. Lumm, B. Pothier, D. Qiu, R.
Spadafora, R. Vicaire, Y. Wang, J. Wierzbowski, R. Gibson, N. Jiwani,
A. Caruso, D. Bush, J. N. Reeve, and et al. 1997. Complete genome
sequence of Methanobacterium thermoautotrophicum deltaH: functional
analysis and comparative genomics. J Bacteriol. 179:7135—55.

Stevenson, B. S. 2000. Doctoral dissertation. Life history implications of
ribosomal RNA gene copy number in Escherichia coli. Michigan State
University.

Stevenson, B. S., and T. M. Schmidt. 1998. Growth rate-dependent
accumulation of RNA from plasmid-borne rRNA operons in Escherichia
coli. J Bacteriol. 180:1970-2.

Tettelin, H., N. J. Saunders, J. Heidelberg, A. C. Jeffries, K. E. Nelson,
J. A. Eisen, K. A. Ketchum, D. W. Hood, J. F. Peden, R. J. Dodson, W.
C. Nelson, M. L. Gwinn, R. DeBoy, J. D. Peterson, E. K. Hickey, D. H.
Haft, S. L. Salzberg, O. White, R. D. Fleischmann, B. A. Dougherty, T.
Mason, A. Ciecko, D. S. Parksey, E. Blair, H. Cittone, E. B. Clark, M. D.
Cotton, T. R. Utterback, H. Khouri, H. Qin, J. Vamathevan, J. Gill, V.
Scarlato, V. Masignani, M. Pizza, G. Grandi, L. Sun, H. O. Smith, C. M.
Fraser, E. R. Moxon, R. Rappuoli, and J. C. Venter. 2000. Complete
genome sequence of Neisseria meningitidis serogroup B strain MC58.
Science. 287: 1 809-1 5.

Tomb, J. F., 0. White, A. R. Kerlavage, R. A. Clayton, G. G. Sutton, R.
D. Fleischmann, K. A. Ketchum, H. P. Klenk, S. Gill, B. A. Daugherty,
K. Nelson, J. Quackenbush, L. Zhou, E. F. Kirkness, S. Peterson, B.
Loftus, D. Richardson, R. Dodson, H. G. Khalak, A. Glodek, K.
McKenney, L. M. Fitzegerald, N. Lee, M. D. Adams, J. C. Venter, and
et al. 1997. The complete genome sequence of the gastric pathogen
Helicobacter pylori. Nature. 388:539-47.

Wang, Y., Z. Zhang, and N. Ramanan. 1997. The actinomycete
Thermobispora bispora contains two distinct types of transcriptionally
active 168 rRNA genes. J Bacteriol. 179:3270-6.

Yap, W. H., Z. Zhang, and Y. Wang. 1999. Distinct types of rRNA
operons exist in the genome of the Actinomycete Thermomonospora
chromogena and evidence for horizontal transfer of an entire rRNA
operon. J Bacteriol. 181:5201-5209.

69

' llillllllllllllllil '