Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Plasmid replication in
thermophilic bacteria
Tom Hargreaves
200374823

Submitted in part fulfilment for the requirements of the degree of
Natural Sciences
March 2010
Supervisor: Dr. C. D. Thomas
Faculty of Biological Sciences Undergraduate School
University of Leeds, Leeds LS2 9JT

1

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Abstract
pSTK1 is a 1883 bp cryptic plasmid found in Bacillus stearothermophilus. A putative
replication-initiation (Rep) protein has been identified in an open reading frame, and
expressed in E. coli. Previous topoisomerase assays have cast doubt over the effectiveness
of this protein at binding pSTK1, leading to this study. SELEX was used for selection of highaffinity DNA sequences for both the pSTK1 Rep protein and the well-characterized RepD
from pC221. A YF mutant that result in inability to nick without harming DNA binding was
used. Electrophoretic mobility shift assays were performed to confirm binding of the selected
sequences. These assays were inconclusive, and the DNA sequences produced were hard
to interpret.

Abbreviations used
aa: amino acid
Absx: absorbance at x nm.
bp: base pair
ds: double-stranded
dso: double-strand origin
EtBr: ethidium bromide
ICR: inverted complementary repeat
MW: molecular weight
ODx: absorbance at x nm.
ORF: open reading frame
ori: origin of replication
SELEX: Selective Enrichment of Ligands by Exponential Enrichment
sso: single-strand origin

Acknowledgements
Thanks are due to everyone who helped to make this project possible. It wasn't their fault.

2

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Introduction and previous work
pT181 is a family of high-copy-number plasmids found in gram-positive bacteria (notably
Staphylococcus aureus) whose rolling-circle replication mechanism has been extensively
studied (Kahn, 2005). Each pT181-family plasmid encodes a Rep protein responsible for
initiation of replication by nicking of the plasmid and recruitment of PcrA helicase.
pSTK1 (figure 2) is a 1883 bp cryptic plasmid found in Bacillus stearothermophilus. pSTK1
has an open reading frame that codes for a 269-aa protein fragment that has sequence
similarities with the active site of pT181-family Rep proteins (table 1). When expressed in E.
coli, this protein was insoluble, but extending the ORF at the N-terminal end resulted in a
343-aa soluble protein hereafter called BstRep.
The interest in this particular plasmid comes from its origin in a thermophile. No Rep protein
structures are currently known. Any pSTK1 Rep protein must be sufficiently robust to function
at the high temperatures favoured by the host organism (60-65 °C), which might make it
amenable to crystallization for subsequent X-ray diffraction experiments.
In 1993, Narumi et. al constructed a shuttle vector pSTE33 from pSTK1 and the common E.
coli vector pUC19 (Narumi, 1993), with the intention that it should be capable of replication in
both B. stearothermophilus and E. coli.
However, after acquiring a copy of pSTE33 from RIKEN it was found not to be viable in B.
stearothermophilus, and furthermore had slight deviations from the published pSTK1
sequence, including a missing GT adjacent to the nick site. Nonetheless BstRep was found
to nick both the published sequence and the truncated pSTE33 sequence (Thomas, 2009).
Protein
RepD,E,I,N
RepC
RepJ
BstRep

Plasmid
Host
pC221 et al. S.aureus
pT181
S.aureus
pUB112
S.aureus
pSTK1
B.stearothermophilus
Identity

Sequence
TKYFGVRDSDRFIRIYNKKQE
TKYFGVRDSNRFIRIYNKKQE
TKYFGSRDSNRFIRIYNKKKE
TLYFGAPSSDIQVRFYEKNVQ
T.YFG...S....R.Y.K...

Table 1: Sequences of the active sites of some pT181-family Rep proteins, with BstRep for
comparison.

3

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Rep proteins perform nicking of the circular dsDNA strand and recruit (host-encoded) PcrA
helicase. The origin of replication (ori) contains the nick site which is flanked by a region of
inverted complementary repeats known as ICR II. ICR II is highly conserved among pT181family plasmids. Specificity is determined by a nearby ICR region called ICR III to which the
Rep protein binds prior to nicking at ICR II. Each plasmid type contains a different ICR III
(and a gene for a Rep protein that binds to it), but all of them have another ICR region
immediately adjacent to ICR II.
That ICR III is the binding site for pT181-family Rep proteins has been confirmed by
mutagenesis (Thomas et. al, 1995). Unlike these plasmids, pSTK1 does not have an ICR
region immediately adjacent to the nick site (figure 1). It is not immediately obvious what the
expected binding site of BstRep would be; while pSTK1 does not have a good adjacent ICR
region, there does exist a widely spaced ICR region somewhat further away that is a
potential candidate for Rep binding. It is also unclear whether binding to pSTK1 occurs with
any reasonable efficiency, given that pSTE33 has not been successfully reintroduced into B.
stearothermophilus.
Thus, this study aims to determine the binding site of BstRep. SELEX (Stoltenburg et al.,
2007) was used to find dsDNA aptamers that BstRep will bind to.
A Y191F1 BstRep mutant was used. In RepD, such a mutation prevents nicking at the active
site Y191 but not (non-covalent) DNA binding. It is hoped that a similar prevention of nicking
in BstRep would allow useful aptamer selection.

(a)

ICR I
nick site (ICR II)
binding site (ICR III)
----- -> <- -----========>
| <======== ------>
<-----AATTACTTACAAAATAAGGATTTAGACAATTTTTCTAAAACCGGCTACTCT'AATAGCCGGTTAAGTGGTAATTTTTTTACCACCCCTCAACCAGAATT

(b)
nick site
possible binding site?
---------->
|
<--------------->
<-----AAATTGGAAAAAATTCAGAGTCTACCCCCGTTGTGT'AACACGGGGGTAGAAAGTACAAGTCAAAGTGGTCTGAAGCCTTGTGTAGACTGGCTTCAAGT

Figure 1: Nick site and surrounding ICR regions for (a) pC221 oriD; (b) putative pSTK1 ori.
1 Residue 191 refers not to the location in BstRep (or RepD!) but to the analogous location in other
Rep proteins. See Appendix A.
4

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Purpose

5'

3'

SELEX

AATTCTAATACGACTCACTATAGGGAGAAGGGCACGGCACGTAGGCAACTA

TTATGAGTGCACGGGCGGCA

Sequencing

GATGTGCTCCATGGCGATTAAGTTGG

TCATTAATGCAGCTGGCACG

Table 2: PCR primers used.

orf3
(rep)

pSTK1
1883 bp

sso

(rep?)
dso

Figure 2: Map of pSTK1. Open reading frames are represented by arrows. The putative
double-strand origin (dso) is the location of the BstRep nick site.

5

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 3: Overview of experimental procedures
6

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Materials and methods
Experiments were carried out as shown in figure 3. Buffers used were as follows:
•

“K0”: 50 mM Tris (pH 7.5).

•

“K200”: 50 mM Tris, 200 mM KCl (pH 7.5).

•

“K600”: 50 mM Tris, 600 mM KCl (pH 7.5), filtered under vacuum with 1μm filter.

•

“K1M”: 50 mM Tris, 1 M KCl (pH 7.5).

•

TAE (50x): 242.2 g/l Tris base, 50 ml/l glacial acetic acid, 18.6 g/l disodium EDTA, pH
8.3. Filtered under vacuum with 0.2 μm filter.

•

Te buffer: 10 mM Tris, 1 mM EDTA, pH 8.0.

•

NEBuffer 4: 50 mM potassium acetate, 20 mM Tris-acetate, 10 mM magnesium
acetate, 1mM dithiothreitol (New England Biolabs).

LB was used as a growth medium for agar plates: 10 g/l tryptone, 5 g/l yeast extract, 10 g/l
NaCl, pH 7.5.
2YT was used as a growth medium for cell culture solutions: 16 g/l tryptone, 10 g/l yeast
extract, 5 g/l NaCl, pH 7.5
1. Binding of protein to magnetic beads.
2 ml of 40 mg/ml 1 μm-diameter Talon beads (Invitrogen) were used. The beads were
first washed with buffer, a magnet placed against the tube and the supernatant
removed. The protein was washed twice with 1 ml of K600 buffer to remove
preservatives, then added to the beads. The following steps were repeated three
times:
1. The tube was rolled gently for 20 minutes.
2. The magnet was placed against the tube, the unbound supernatant was collected,
centrifuged for 2 minutes and the Abs260 and Abs280 checked.
3. 300 μl of K600 buffer was added. (200 μl in the final iteration.)
2. Extraction of protein on heparin sepharose column.
The column volume was 0.3 ml. The salt concentration of the protein solution was
adjusted to 200 mM, and the protein applied to the column. The unbound fraction was
collected and its Abs280 measured. The column was washed twice with 2 ml K200, the
wash fractions collected and their absorbance measured. The protein was eluted with

7

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

2 ml K600.
3. DNA SELEX.
15 rounds of SELEX were carried out. A pool of synthetic oligonucleotides with
random 50 bp region was used (the same pool was used for all experiments).

The

sequences of the flanking regions were the same as those of the PCR primers (see
Table 2).
SELEX services were kindly provided by Dr. David Bunka.
4. Preparation of competent E. coli cells.
A single colony of Escherischia coli strain DH5α was picked, and incubated in 5 ml
2YT broth overnight at 37 °C. A ~1 ml aliquot corresponding to an OD600 of 0.05 was
removed, placed in a flask containing 50 ml 2YT broth, and incubated at 37 °C.
Growth was monitored at 20-minute intervals until the OD600 reached 0.5. The
doubling time was calculated. The cell culture solution was cooled to 4 °C, centrifuged
at 3000rpm for 10 minutes, and the supernatant discarded. The pellet was
resuspended in 25ml sterile 100 mM CaCl2, kept at 4 °C for 30 minutes, then
centrifuged at 3000rpm for 10 minutes. The supernatant was discarded. The pellet
was resuspended in 5ml sterile 100 mM CaCl2 and stored overnight at 4 °C.
5. Ligation of DNA into plasmid vector.
pGEM T-Easy (Promega) was used as a vector. Four ligations were carried out as per
the manufacturer's recommendations: a positive control with provided test insert, a
negative control with no insert, and two reaction mixtures with varying amounts of
sample DNA.
6. Transformation of E. coli.
The four 10 μl ligation mixes and a fifth control mix containing 2ng supercoiled
plasmid DNA from pCER19 were placed in microcentrifuge tubes. 10 μl Te buffer and
200 μl of competent cells were added, and the tubes incubated on ice for 30 minutes.
The cells were heat shocked for 2 minutes at 42 °C. 500 μl of 2YT broth was added,
the tubes inverted once then incubated at 37 °C for 90 minutes. 550μl was carefully
removed from the top of the liquid, and the remaining liquid was agitated. The cells
were transferred to agar plates containing 50 μg/ml ampicillin, and incubated
overnight at 37 °C.
7. Plasmid miniprep.
1.5 ml of bacterial culture was collected in a microcentrifuge tube. Plasmid DNA was

8

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

prepared using a Qiagen QIAprep Spin Miniprep kit as per the manufacturer's
instructions. Two modifications to the protocol were made. Firstly, after the wash
buffer (the ethanol-containing “buffer PE”) was removed by centrifugation, the spin
columns were placed under a fan for 3 minutes to evaporate any residual wash buffer.
Secondly, DNA was eluted in a final volume of 100 μl instead of 50 μl.
8. Agarose gel electrophoresis.
A solution of 1.2% agarose in TAE buffer was prepared (approx. 100ml per gel) and
heated until fully dissolved. Ethidium bromide was added for a total concentration of
1μg/ml (from stock 10mg/ml). The agarose was allowed to cool to 55 °C, then poured
and allowed to set. After samples were loaded, 100V DC (current limited to 100 mA)
was applied across the gel for 80-90 minutes. The gel was removed and
photographed under UV light.
9. DNA sequencing.
30 μl of DNA at a concentration of approximately 50ng/μl was used. Primers used
were from oriDF and CER0372R; sequences are in Table 2. DNA sequencing
services were provided by the Faculty of Biological Sciences, University of Leeds,
using an ABI 3130xl capillary sequencer.
10. Restriction digest with PvuII.
5μl of each DNA sample was incubated with 0.155 μl PvuII-HF (New England
Biolabs) in NEBuffer 4 (total volume 10 μl) for 60 minutes at 37 °C, followed by
inactivation of the enzyme by incubation at 80 °C for 20 minutes.
11. Electrophoretic mobility shift assay (EMSA).
10 μl of each of the PvuII restriction digest products were incubated with 10 μl protein
or K0 (control) for 60 minutes at 60 °C (BstRep) or 37 °C (RepD). The samples were
run on an agarose gel without EtBr to avoid interference with bound protein during
electrophoresis: the gel was subsequently immersed in a 1 μg/ml EtBr solution for 30
minutes before being photographed.
12. Quantitation of DNA.
A 5 μl aliquot of DNA sample was run on an agarose gel along with reference
samples containing 25 ng, 50 ng, 100 ng and 200 ng of supercoiled plasmid DNA.
The relative intensities of the bands were compared quantitatively using ImageQuant
software.

9

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Results
BstRep is 343 amino acids long, but was prepared with a 20-aa N-terminal His6 tag for a total
molecular weight of 42.1 kDa (Appendix A). The stock BstRep was at a concentration of 0.64
mg/ml, stored in 1 mM Tris, 1 mM EDTA, 10% ethanediol, 500 mM KCl at pH 7.5.
The protein was bound to beads. The absorbance of the unbound fractions indicated that
very little protein had bound (data not shown). This was surmised to be due to the
competitive chelation of the Co2+ ions in the beads by the EDTA present in the buffer. Thus
the unbound fractions were pooled and the EDTA was removed by elution on a heparin
sepharose column.
The protein was again bound to the beads. The beads were submitted for SELEX.
The SELEX procedure as performed by Dr. David Bunka (Bunka et. al., 2006):
1. Immobilize the protein of interest on beads.
2. Add solution containing random oligonucleotide pool.
3. Repeat multiple times:
1. Add protein.
2. Allow DNA to bind to protein immobilized on beads.
3. Remove unbound solution.
4. Perform PCR on remaining oligonucleotides.
The PCR products after 15 rounds of SELEX were run (also by Dr. Bunka) on a
polyacrylamide gel, and the ~120bp products extracted. The result is a selection of
“aptamers”: oligonucleotides that exhibit unusually-high affinity for the protein under
consideration.
The PCR product after 15 rounds of SELEX contained 19 ng of DNA in a 5 μl aliquot, for a
concentration of 3.75 ng/μl. This was ligated into the plasmid vector pGEM T-Easy, which
was used to transform E. coli cells. Two ligations were carried out containing 1 μl and 3 μl of
DNA for an approximate ratio of DNA to vector of 1:1 and 3:1 respectively. 24 colonies were
picked and grown, the first 12 (A-L) were from the 1:1 ligation and the other 12 (M-X) were
from the 3:1 ligation. Plasmid DNA was extracted.

10

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 4: Restriction digest of pGEM vector with inserts from BstRep SELEX.

Of the 24 plasmid DNA samples, restriction digest analysis with PvuII (figure 4) showed that
16 contained an insert (furthermore, an insert that lacked a PvuII site). These 16 were
sequenced.
The result was 16 pairs of DNA sequences, one from each primer. Having two versions of
each sequence, one in each direction, allowed for confirmation of the fidelity of the
sequencing. Only one of the 16 sequences (“A”) had any significant dropouts or mismatches
between versions, and it was nonetheless possible to deduce the correct sequence of A with
reasonable certainty.

11

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 2: multiple sequence alignment of the random region of
the 16 BstRep SELEX sequences (A-X) and their complements (rev suffix). The forward and reverse sequences are nearly
disjoint, indicating that the primer direction is significant. The
markedly skewed proportion of AC bases is also apparent.

Figure 5: multiple sequence alignment of the random region of the 16 BstRep SELEX sequences (AX) and their reverse complements (-rev suffix). Sequence length is in the righthand column. The
forward and reverse sequences are nearly disjoint, indicating that the primer direction is significant.
The markedly skewed proportion of AC bases is also apparent.

Sequence analysis.
The sequences were oriented so that the SELEX primer sequences were all facing in the
same direction, and multiple alignment was performed using ClustalX (figure 5).
No obvious consensus sequence is apparent from the sequence alignment; nonetheless the
sequences are decidedly non-random. The possibility that BstRep's region of specificity is
non-contiguous was tested by supposing that it encompasses exactly N bases in a
contiguous M-base region. An exhaustive search of all possibilities was undertaken by
assigning a score to each possibility P according to how well it fits with each sequence S.
Two different scoring algorithms were used in an attempt to reduce artifacts.

12

Plasmid replication in thermophilic bacteria

(a)
CACCACNNNCCNNNNN
CCNNCACNNNCCNANN
CACNNNCAACCNNNNN
ACACNNNCAANCNNNN
CACNACNNACCNNNNN
CACCNNNNACCNNCNN
ACNCNACCNNCNNNAN
ACNCNACCNNCNNANN
CNCCNCCANNNNCCNN
CNNNCACCACNNNCNN
CACCNNCANANNNCNN
CNCCANCANNNNCCNN
ACACNNNCAACNNNNN
CNCNACCNNCNNAANN
CNCCNNCANANNCCNN

Tom Hargreaves, February 2010

(b)
CACCACNNNCCNNNNN
ACNCNNCCNNCNNANA
ACNCNACCNNCNNNNA
CCNNCACNNNCNNAAN
ACNCCNCCNNCNNNNA
CNCCACNNNNCNCCNN
CCNCCACNNNCCNNNN
CCNNCACNNNCCNANN
ACNCNACCNNCNNANN
ACNCNACNNNCNNANA
CACNCNNCCNNCNNAN
ACNCCACCNNNNNNNA
ACNCCACCNNCNNNNN
ACNCNANCNNCNNANA
ACNNNNCACNNACNNA

104
104
104
103
103
103
103
103
103
103
103
103
103
103
103

1888
1880
1872
1856
1848
1824
1824
1808
1808
1792
1792
1792
1776
1776
1768

Table 3: Top-matching consensus sequences for the forward-direction BstRep SELEX sequences, for
N=8, M=16 (i.e. 8 significant bases are chosen out of 16). (a) additive scoring; (b) exponential scoring
(see text for details of scoring algorithms). Perfect scores would be 128 and 2048 respectively.

A possibility is scored as follows. For each S: P is matched with both S and its reverse
complement at every position. Bases that match score 1, up to a maximum of N for a perfect
match. Let P(S) be the highest score when P is matched with S. Then, for additive scoring, P
scores ΣP(S), the sum of all highest scores. For exponential scoring, P scores Σ2P(S). Additive
scoring ranks reasonable overall matches

highly, whereas exponential scoring prefers

matches that are perfect for some sequences but that fare poorly on others.
The same consensus sequence is ranked first by both algorithms; however, a large number
of unrelated sequences rank similarly so this may be misleading. All of the top ten sequences
contain only A and C.
Since AC-richness is the only clear trend, pSTK1 was searched for AC-rich (or GT-rich)
regions. While there are such regions, there are none in the vicinity of the nick site.

13

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 6: restriction map of pGEM T-Easy vector showing PvuII sites.

Restriction digests and gel shift assay.
The restriction endonuclease PvuII cuts at CAG'CTG sites; the resulting fragments have
blunt ends. pGEM T-Easy is 3015 bp long prior to insertion and has two PvuII sites
positioned as shown in figure 6. Since the inserts used were about 120 bp, this resulted in
fragments of ~2500 bp and ~570 bp, with the insert located within the 570 bp fragment.
An electrophoretic mobility shift assay (EMSA) was used to test binding of BstRep protein to
the restriction fragments (figure 7). Such a large insert-containing fragment was desirable
because the behaviour of the Rep proteins at binding small DNA fragments was unknown.
330nM BstRep was used, with an ionic strength of ~45mM.
No shifts were observable, however the lanes containing protein showed bands still in the
well, indicating total immobility.

14

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 7: BstRep mobility shift assay. Each sample occupies two adjacent lanes, without and with
protein added respectively.

15

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

RepD.
The entire experiment was repeated using RepD instead of BstRep. The RepD used was
wild-type with a 20aa N-terminal tag, with a molecular weight of 39.6 kDa.
EDTA extraction and SELEX were performed as before. Two ligations were carried out, but
both used 3 µl of DNA due to the concentration of the PCR product being substantially lower.
24 colonies were picked; the first 12 (A-L) were taken from the PCR product after 10 rounds
of SELEX, and the other 12 (M-X) from the PCR product after the full 15 rounds as before 2.
17 colonies out of 24 showed an insert after digestion with PvuII (figure 8), and were
sequenced. Only 10 out of 17 DNA sequences were viable.
Figure 9 shows the 10 sequences and their reverse complements aligned with ClustalX as
before. Table 4 shows the results of a consensus sequence search.
Gel mobility shift assays (figure 10) did not show any observable shift, however as before,
samples with protein contained species that were effectively immobile in the gel, perhaps
pointing to aggregation. Furthermore, control experiments (figure 10(d)) showed that no shift
was observable even with a sequence that RepD is known to bind to (pCER19 plasmid
containing the oriD sequence), and similarly with BstRep. Even without any DNA present,
BstRep was immobile (and visible!) in the gel, thus casting doubt over all previous mobility
shift assays.

2 It seemed like a good idea at the time.
16

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 8: PvuII restriction digest of vector containing RepD SELEX inserts.

Figure 9: multiple sequence alignment of the 10 RepD SELEX sequences (A-X) and their reverse
complements (-rev suffix). Sequence length is in the righthand column. A repeated CCGG motif is
apparent.

17

Plasmid replication in thermophilic bacteria

(a)
CCNNGNANNNCANCNG
CNNGNNANANANNCGG
CNNGGNANANNNNCGG
CNGNANNNNACCNGGN
CNNGNNANNCANNCGG
GCAANNNACCNGNNNN
CNNNGNANNNCACCNG
CNNGNNANANNNCCGG
CNNNNNANANANCCGG
GNAANNNANCCGGNNN
GNAANNCANCNGGNNN
CNNNGNANNANACCNG
CNNGNNNNNCANCCGG
CNNNNNANNCANCCGG
GNAANNNACCNGGNNN
GNAANANACNCGNNNN

Tom Hargreaves, February 2010

(b)
CNNGNNANANNNCCGG
CNNGNNANNCNCCNGG
CNNGNAANANNNNCGG
CNNGNNANANANNCGG
GCAANNNACCNGNNNN
CNNGNNANANNCNCGG
CCNGNNANANNNNCGG
CNNGNNANNCANCNGG
CNNGNANNNCACCNGN
CNNGNNANACNNNCGG
CCNGNNNNNCANCNGG
GNAANANANNCGGNNN
CNNGNANNNNACCNGG
CCNGNANNNNANCNGG
CNNGNANNNCANCNGG
GNNANANNCCCGGNNN

67
67
67
67
67
67
67
67
67
67
67
67
67
67
67
67

1536
1456
1456
1440
1440
1424
1424
1424
1424
1408
1408
1408
1408
1408
1408
1408

Table 4: top-matching consensus sequences for RepD SELEX sequences, for N=8, M=16 (i.e. 8
significant bases are chosen out of 16). (a) additive scoring; (b) exponential scoring (see text for
details of scoring algorithms). Perfect scores would be 80 and 2560 respectively. Note that in (a) there
were 41 joint-top-scoring sequences: the selection is illustrative.

18

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Figure 11: mobility shift assay for RepD. (a) lanes 1 and 14, size markers; lanes 2-13, samples A-F,
without10: mobility shiftrespectively. (b) lanes 5 and 16, size size markers; lanes 2-13, samples A-F,
Figure and with RepD assay for RepD. (a) lanes 1 and 14, markers; lanes 1 and 2, sample G, with
and without RepD respectively; lanes 3 lanes 5sample H, without and with RepD respectively; lanes 6without and with RepD respectively. (b) and 4, and 16, size markers; lanes 1 and 2, sample G, with
15, samplesRepDM, O and P, without 3 and 4, sample H, without and with RepD respectively; lanes 6and without J, L, respectively; lanes and with RepD respectively. (c) lanes 1 and 12, size markers;
lanes 2-9, samples Q, S, T and U, without with with RepD respectively; lanes 10 and 11, positive
15, samples J, L, M, O and P, without and and RepD respectively. (c) lanes 1 and 12, size markers;
control sequence from S, T and U, without and with RepD respectively; lanes 10 and 11, lanes 1 and
lanes 2-9, samples Q, pCER19 containing oriD, without and with RepD respectively. (d) positive
14, sizesequence from pCER19 containing oriD, without andand with RepD respectively; lanes 1 and
control markers; lanes 2-9, samples B, D, F and P, without with RepD respectively. (d) lanes 10-11,
positive control sequence from pCER19 D, F and P, oriD, without with RepD respectively; lanes 10-11,
14, size markers; lanes 2-9, samples B, containing without and and with RepD respectively; lanes
12-13, negative sequence from pCER19 containing oriD,oriD, with and with RepD respectively; lane
positive control control sequence from pCER19 without without and with RepD respectively; lanes
15, negative control sequence from pCER19 without oriD, with BstRep;with RepD respectively; (no
12-13, negative control sequence from pCER19 without oriD, with and lane 16, BstRep alone lane
DNA).
15, negative control sequence from pCER19 without oriD, with BstRep; lane 16, BstRep alone (no
DNA).

19

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

Discussion
Plasmids are a major vector for the spread of antibiotic resistance within pathogenic bacteria.
Emerging bacterial strains such as methicillin-resistant Staphylococcus aureus (MRSA) are a
growing clinical problem. Structural studies of the methods of replication and conjugation of
resistance plasmids are of potential interest as ways of inhibiting the proliferation of such
strains are sought.
BstRep is an interesting target for structural determination due to its thermophilic nature. If
the structure of BstRep can be elucidated it may shed light on the action of Rep proteins in
general.
Without confirmed protein binding, the DNA sequences presented here are meaningless.
However, assuming that binding is nonetheless occurring, it is possible to attempt to make
some infererences. The BstRep experiments show no clear consensus sequence and a very
puzzling orientation correlation with respect to the primers (perhaps BstRep is binding to the
primers?). The RepD experiments show a slightly clearer consensus sequence but not one
that matches well with the known binding site of RepD.
Obvious further work would include making the mobility shift assay work. Further further work
could include mutagenesis of pSTK1 to determine the nature of the BstRep binding site.

Conclusion
SELEX produced some DNA sequences, but their meaning is deeply unclear. Mobility shift
assays were inconclusive, thus it is still unknown whether these sequences are actually
bound by their respective proteins.

20

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

References
BUNKA, D.H.J., STOCKLEY, P.G., 2006. Aptamers come of age - at last. Nature Reviews
Microbiology, 4, 588-596.
KHAN, S. A. 2005. Plasmid rolling-circle replication: highlights of two decades of research.
Plasmid, 53, 126-136.
NAKAYAMA, N., NARUMI, I., NAKAMOTO, S. & KIHARA, H. 1993. Complete nucleotide
sequence of pSTK1, a cryptic plasmid from Bacillus stearothermophilus TK015.
Biotechnology Letters, 15, 1013-1016.
NARUMI, I., NAKAYAMA, N., NAKAMOTO, S., KIMURA, T., YANAGISAWA, T. & KIHARA, H.
1993. Construction of a new shuttle vector pSTE33 and its stabilities in Bacillus
stearothermophilus, Bacillus subtilis, and Escherichia coli. Biotechnology Letters, 15,
815-820.
STOLTENBURG, R., REINEMANN, C. & STREHLITZ, B. 2007. SELEX--A (r)evolutionary
method to generate high-affinity nucleic acid ligands. Biomolecular Engineering, 24,
381-403.
THOMAS, C. D., NIKIFOROV, T. T., CONNOLLY, B. A. & SHAW, W. V. 1995. Determination
of Sequence Specificity between a Plasmid Replication Initiator Protein and the Origin
of Replication. Journal of Molecular Biology, 254, 381-391.
THOMAS, C. D., LYNCH, G. P. 2009. Unpublished observations.

Appendix A
Sequence of BstRep (363aa; N-terminal tag underlined, “Y191F” mutation highlighted in bold):
1

MGSSHHHHHH SSGLVPRGSH MSGLKPCVDW LQVTFKTGQD SVKKCVEKLE KVFEILGLNE

61

AEFLPLKNGK YGYKQGVAFQ GNPVLAVYYD GADDMGIHVE MTGQGCRLFE LHTSINWYEL

121

FYRLVYEYEV NITRLDVAVD DFKGYFKINT LVKKLKDDEV TSRFKKARHI ENIVIEGGET

181

IGHTLYFGAP SSDIQVRFFE KNVQMGMDID VWNRTEIQLR DDRAHVVAQI IADDVLPLGE

241

IVAGLLRNYI QFRTRKATDK NKKRWPLARF WLNFLGDVQP LRIAKQMPKT SIEKKYRWID

301

SQVSKSFFMI YYCLNEEEKQ RFIDDVLAEG ASKLTKADLQ VINQFKSKNI TYDEMIKIIR

361

QSK

Predicted MW is 42085.40 Da.

Appendix B
Source code to the brute-force consensus searching (“consensus.pl”), window overlap determination
(“window.pl”),

and

other

custom

software

used

in

this

study

are

available

http://sphere.chronosempire.org.uk/~HEx/bioc3160/, as well as being reproduced (in part) below.

21

at

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

#!/usr/bin/perl -wl
use strict;
use Bio::SeqIO;
use Getopt::Long;
use Math::Combinatorics;
my $format = 'fasta';
GetOptions('f|format:s' => \$format);
my @bases=qw[A C G T N];
my @tests;
use constant MAXRESULTS => 100;
use Inline C=> Config =>
CCFLAGS => '-O3';
use Inline C => <<'EOF';
int run_test (char *test, char *seq) {
int len=strlen(test);
int maxscore=0;
int seqlen=strlen(seq);
while (*seq) {
int score=0;
int n;
for (n=0; n<len; n++) {
char *b = test[n];
if (b != 'N' && (n < seqlen)) {
if (b==seq[n])
score++;
}
}
if (score > maxscore)
maxscore = score;
seq++;
seqlen--;
}
}
EOF

return maxscore;

my $len=shift; # total bases specified (4^n combinations)
my $size=shift; # space in which the bases occur
sub num2seq {
my $a;
my $num=$_[0];
my $numns=0;
for (0..$len-1) {
my $n=$num % 4;
$a.=$bases[$n];
$num/=4;
}
$a;
}
my %seq;
my %revseq;
foreach my $f (@ARGV) {
my $in = new Bio::SeqIO(-file => $f, -format => $format);
while (my $seq = $in->next_seq) {
my $name=$seq->primary_id;
my $s=$seq->primary_seq->seq;
$seq{$name}=$s;
my $rev = $seq->revcom ();
my $r=$seq->primary_seq->seq;
$revseq{$name}=$r;
}
}
my (%results,%results2);
my (@results, @results2);

22

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

my $num=4**$len;
my $tests=0;
my $m=Math::Combinatorics->new(count => $len-1, data => [1..$size-1]);
while (my @combo=$m->next_combination) {
$tests++;
for (0..$num-1) {
my @test=('N') x $size;
@test[0,@combo] = split//, num2seq($_);
my $test=join"", @test;
my ($r1,$r2) = test($test);
if (!@results || $r1 > $results{$results[0]}) {
$results{$test}=$r1;
push @results, $test;
@results = sort { $results{$a}<=>$results{$b} } @results;
if (@results > MAXRESULTS) {
delete $results{shift @results};
}
}

}

if (!@results2 || $r2 > $results{$results[1]}) {
$results2{$test}=$r2;
push @results2, $test;
@results2 = sort { $results2{$a}<=>$results2{$b} } @results2;
if (@results2 > MAXRESULTS) {
delete $results2{shift @results2};
}
}

}
print "Len: $len size: $size";
print "Total tests: ". $tests*$num;
print
print
print
print

"Additive scoring (top ".MAXRESULTS."):";
map {"$_\t$results{$_}\n"} reverse @results;
"Exponential scoring (top ".MAXRESULTS."):";
map {"$_\t$results2{$_}\n"} reverse @results2;

sub test {
my $test=shift;
my $score1;
my $score2;
for my $name (keys %seq) {
my ($seq,$rev)=($seq{$name}, $revseq{$name});
my $sscore=run_test ($test, $seq);
my $rscore=run_test ($test, $seq);
$sscore=$rscore if $rscore>$sscore;
$score1+=$sscore;
$sscore=1<<$sscore;
$score2+=$sscore;

}

}
return ($score1,$score2);

Listing 1: consensus.pl. Note that this code is naïve and egregiously inefficient (running time for the
results shown in the text was approximately 24 hours).

23

Plasmid replication in thermophilic bacteria

Tom Hargreaves, February 2010

List of COSHH forms signed

[this page intentionally left blank]

24