Construction of arsB and tetH Mutants of the Sulfur-Oxidizing

Bacterium, Acidithiobacillus caldus by Marker Exchange.

Leonardo J. van Zyl, Jolanda M. van Munster, and Douglas E. Rawlings∗.

5

Department of Microbiology, University of Stellenbosch, Private Bag X1, Matieland, 7602,

South Africa

Author for correspondence: Douglas E. Rawlings. Tel: +27-21- 808 3071, Fax: +27-21-808

3680 e-mail: der@sun.ac.za 10

Section: Genetics and molecular biology

Abstract 15

Acidithiobacillus caldus is a moderately thermophilic, acidophilic bacterium that has been

reported to be the dominant sulfur-oxidizer in stirred tank processes used to treat gold-

bearing arsenopyrite ores. It is also widely distributed in heap-reactors used for the

extraction of metals from ores. Not only are these bacteria commercially important but 20

they have an interesting physiology the study of which has been restricted by the non-

availability of defined mutants. A recently reported conjugation system based on the broad

host-range IncW plasmids pSa and R388 was used to transfer mobilizable, narrow host-

range, suicide plasmid vectors containing inactivated and partially deleted chromosomal

genes from Escherichia coli to At. caldus. Through the dual use of a selectable kanamycin 25

resistance gene and a hybridization probe made from a deleted portion of the target

chromosomal gene, single and double recombinant mutants of At. caldus were isolated.

The functionality of the gene inactivation system was shown by the construction of At.

caldus arsB and tetH mutants and the effect of these mutations on cell growth in the

presence of arsenic and by means of tetrathionate oxidation was demonstrated. 30

ACCEPTED

Copyright © 2008, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.Appl. Environ. Microbiol. doi:10.1128/AEM.01235-08 AEM Accepts, published online ahead of print on 25 July 2008

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

Introduction

35

The creation of null mutants (non-functional gene product) of any organism is pivotal

to our understanding of individual gene function as it provides a baseline of activity to

which comparisons can be made. We have developed a method for generating such mutants

in the acidophilic, moderately thermophilic (45ºC) bacterium Acidithiobacillus caldus (8,

14) through allelic exchange (homologous recombination). Together with the iron 40

oxidizing bacterium Leptospirillum ferriphilum, At. caldus has been reported to dominate

the microbial populations in commercial arsenopyrite concentrate bio-oxidation tanks that

operate at 40ºC and pH 1.6 as well as in pilot-scale stirred-tank bioleaching operations at

45ºC (5, 20, 24). At. caldus has also been found in heap reactors (12, 25) making the study

of this organism of economic importance. These bacteria are also interesting to study from 45

a fundamental biology point of view due to their unusual metabolism and physiological

characteristics (8).

The development of genetic systems for members of the genus Acidithiobacillus and

other acidophilic bacteria found biomining environments has been particularly challenging.

Nevertheless, plasmid transfer, and expression of heterologous genes have been reported 50

for three species of acidithiobacilli, Acidithiobacillus ferrooxidans (16, 18, 22, 23),

Acidithiobacillus thiooxidans (13, 30) and At. caldus (15, 19, 32). In spite of this, there has

been only one report of the construction of a null mutant among these bacteria, that being a

recA mutant of At. ferrooxidans (18). Attempts to create other mutants have so far been

unsuccessful. Although the ability to express or over-express a chosen gene(s) in a 55

particular organism is useful for strain improvement, the lack of a more complete suite of

genetic tools such as the construction of knockout mutants has hampered the study of the

Acidithiobacilli.

The genome sequence of At. caldus is not yet available, however, two sets of genes

that are candidates for the construction of mutants have been reported. These are the genes 60

conferring arsenic resistance (15) and the genes for tetrathionate utilization (26).

The chromosomally encoded arsenic resistance operon from At. caldus has been

described previously (15) and consists of three genes: arsR, arsB and arsC (Fig. 1). The

product of the arsB gene is the arsenite exporting membrane-located pump while the arsR

and arsC genes encode for an arsenite sensitive regulator and an arsenate reductase, 65

respectively. Inactivation of the arsB should therefore result in an arsenic sensitive

phenotype unless another resistance mechanism is present.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

Although the ability of At. caldus to utilize reduced inorganic sulfur compounds

(RISC) as energy source is well known, the exact mechanism by which it decomposes these

compounds and derives energy from them remains elusive. Tetrathionate (S4O62-) can be 70

used as its sole energy source and the enzyme thought to be responsible for its breakdown

is the tetrathionate hydrolase (tetH) (6). A gene cluster from At. caldusT (KU) containing a

homolog of this gene was recently cloned and sequenced. Identification of the gene as a

tetrathionate hydrolase was made on the basis of sequence similarity to other known

tetrathionate hydrolases and the tetrathionate hydrolase activity was linked to this gene by 75

N-terminal sequencing of the protein responsible (26).

Here we report the inactivation of the chromosomally encoded arsenic resistance gene

arsB in At. caldusT through marker exchange mutagenesis, to serve as a model system for

the knockout of other genes of interest from this microorganism. We have further

demonstrated the usefulness of this technique by constructing a second At. caldusT 80

knockout mutant, that of the tetH gene.

MATERIALS AND METHODS

Media, bacterial strains and plasmids. Bacterial strains and plasmids used in this 85

study are listed in Table 1. Escherichia coli strains were grown in Luria–Bertani (LB) broth

medium (27), with ampicillin (200 µg/ml), kanamycin (50 µg/ml), naladixic acid (35

µg/ml) or trimethoprim (50 µg/ml) added as required. To culture At. caldus strains in liquid

medium, a 10X 9K basal salts solution (29) was made and the pH adjusted to 2.5 using

concentrated H2SO4 and autoclaved. When Na2S2O3·5H2O was used as the sulfur source, 90

the 10X stock was diluted to a 1X solution and the pH adjusted to 4.7 using 10N NaOH to

which other components would be added to make the growth media. A 20% (wt/vol) stock

solution of Na2S2O3·5H2O was made and filter sterilized of which 500 µl was added to

every 100 ml of 1X basal salts to give a final concentration of 0.1% (wt/vol). To this 50 µl

of 1000 x trace elements (ZnSO4⋅7H2O 1 g, CuSO4⋅5H2O 0.1 g, MnSO4⋅4H2O 0.1 g, 95

CoSO4⋅7H2O 0.1 g, Cr2(SO4)3⋅15H2O 0.05 g, Na2B4O7⋅10H2O 0.05 g, Na2MoO4⋅2H2O 0.05

g, NaVO3 (optional) 0.01 g dissolved in 100 ml dH2O, 53 µl H2SO4 added and autoclaved)

was added per 100 ml media. The final pH of the media was ± 4.5. The cultures were

incubated in shake flasks at 37ºC with active aeration (100 rpm). When K2S4O6 served as

sulfur source (K2S4O6 is less acid liable than Na2S2O3) the 10X basal salts stock was diluted 100

to 1X and the pH adjusted to 2.5 instead of 4.7 using concentrated H2SO4.

Solid medium for culturing At. caldus was made by diluting the 10X basal salts stock

to 2X solution and the pH adjusted to 4.7 using 10N NaOH and autoclaved. This made up

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

half of the final volume of the final media. The other half consisted of a 1.75% (wt/vol)

Phytagel (Sigma-Aldrich) solution made up in dH2O and autoclaved. The solutions were 105

allowed to cool to 50ºC. To the 2X basal salts solution was added, Na2S2O3·5H2O to a final

concentration of 0.2% wt/vol and trace element solution, 100 µl for every 100 ml of the 2X

basal salts solution. When used as selective media kanamycin was added to a final

concentration of 100 µg/ml. The basal salts and phytagel solutions would then be mixed

and the plates poured. 110

Conjugation. At. caldusT was cultured on the higher pH (4.5) media and at 37ºC to

pre-adapt it to the conditions of the conjugation experiments. Cells from late stationary

phase cultures (5 days) served as recipient. E. coli HB101 to be used in the conjugation

experiments were initially cultured in LB broth but then inoculated into 50 ml of the pH 4.5

media with thiosulphate and trace elements added to the same final concentration as for the 115

At. caldus cultures, but also supplemented with 0.05% (wt/vol) yeast extract and cultured

O/N at 37ºC. This was done to pre-adapt the donor to the mating media. Cells from donor

(50 ml) and recipient (500 ml) cultures were collected by centrifugation and washed twice

in 2 ml of the 1X high pH basal salts solution. Donor and recipient cells were then

separately re-suspended in 250 µl of the salts solution each and mixed in a 1:1 ratio. Of 120

this 500 µl mixture, 100 µl was spread evenly onto a Supor® 0.2 µm filter (PALL®

Gelman Laboratory) which was placed on the mating medium. Thus five filters were used

per mating to accommodate the whole mating mixture. The mating medium was the same

as the solid media for culturing At. caldus described above but supplemented with 0.05%

(wt/vol) yeast extract and 0.5X 10-4M diaminopimelic acid (18). After 5 days of incubation 125

at 37ºC, the cells were harvested by scraping growth from the filters with a loop and

washed twice in 2 ml of 1X high pH basal salt solution. Following mating and to provide

the opportunity for the generation of a mutant, the cells were collected, washed, and

inoculated into several 500 ml high pH liquid cultures with selection for kanamycin

resistance (100 µg/ml). This was done for several reasons. First, to avoid inoculation with 130

too many cells thereby making it difficult to judge whether growth had taken place.

Second, to negate the protective effect that cells appear to afford one another when

inoculating or plating too many cells for a given antibiotic concentration. Third, the cells

were cultured with selection to raise the number of mutants (single or double crossover) to

a detectable level as the combined mating and crossover frequencies could be well below 135

our detection limit (approximately 10-7 transconjugants / recipient). Cultures were incubated

at 37ºC for 6-7 days in shake flasks with vigorous aeration.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

In an attempt to determine the rate at which the suicide vector (pOTF101) was

transferred from E. coli to At. caldus, the cells were collected after mating, washed, and a

serial dilution plated onto selective and non-selective plates. The plates were incubated for 140

5 to 6 days at 37ºC. The frequencies of plasmid transfer were expressed as the “apparent

transfer frequency”, i.e. the number of transconjugant colonies that grew on selective

medium per recipient divided by the number of colonies that grew on non-selective

thiosulfate medium. Donor bacteria were counter selected by the absence of a carbon

source in the selective medium. 145

PCR. Polymerase chain reactions were performed using BIOTAQ™ DNA

polymerase from BIOLINE according to the manufacturer’s recommendations. In general

50 ng of DNA was used in a 50 µl reaction volume containing 2 mM MgCl2, 0,25 µM of

each primer, 200 µM of each dNTP, and 1 U TaqI polymerase. Reactions were carried out

in a Hybaid Sprint thermocycler, with an initial denaturation at 94ºC for 60 s, followed by 150

25 cycles of denaturation (30 s at 94ºC), an annealing step of 30 s, and a variable

elongation step at 72ºC. Annealing temperatures and elongation times were altered as

required. The 557-bp origin of transfer from the IncW plasmid R388 was amplified using

primers R388oriTF (5’-TATAGAATTCAGCTCGCCTTGCAAGTCG-3’) and R388oriTR

(5’-TCGCGAATTCAAGGTCGTTTGCCTGCAT-3’) respectively. This product was 155

cloned using the pGEM® cloning kit from Promega. Primers TetH Fwd (5’-

TAGAACCAAGGACAGC-3’) and TetH Rev (5’-AACATCGGCACAGAGA-3’) were

used to amplify the 2.3-kb fragment from the At. caldus chromosome that contains part of

the tetrathionate hydrolase operon used in the making of the suicide vector (Fig. 1A). The

fragment was then cloned using pGEM. Primers KmFor (5’-TTGCACGCAGGTTCTCC-160

3’) and KmRev (5’-TCGGGAGCGGCGATACC-3’) were used to amplify a 714-bp

fragment internal to the kanamycin resistance gene from Tn5. Fig. 2 shows how the

oligonucleotide pair ORF1-RTRev (5’-GATCGCGCAGCCAGAGTT-3’) and ORF5-

RTRev (5’-GTTTGGCAGGGATTGCGG-3’) allowed the insertion junction between

ORF1 and the interrupted arsB to be checked. Similarly Fig. 2 indicates how the primer 165

pair arsB-Crossover (5’-GTTTGATCGCTATGCCC-3’) and ORF7-Crossover (5’-

GTCTGCACGGACTGCAT-3’) can be used to check the insertion junction between ORF7

and the interrupted arsB.

Construction of the At. caldus arsB and tetH mutants. The origin of transfer was

excised from pGEM using the EcoRI restriction endonuclease and cloned into the EcoRI 170

site on pUC19 to give pOT. A 2.2-kb PstI-PvuII fragment from pAtcars4 (Table 1)

containing part of the chromosomally encoded arsenic resistance operon was cloned into

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

the PstI and blunted SacI sites of pOT to give pOTars (Fig. 1A). This construct was

digested with KpnI which removed two ±200-bp fragments from the centre of the arsB

gene, which were later used as a probe for detecting single and double crossover mutants, 175

and the ends filled in using the T4 DNA polymerase (Fermentas, Vilnius Lithuania). The

±1.45-kb XbaI-XhoI fragment from pSKM2 carrying the kanamycin resistance cassette

from Tn5 was also blunt-end cloned into the blunted KpnI sites of arsB to give parsB::Km

(Fig. 1B). The 2.3-kb DraI-EcoRI fragment containing part of the tetrathionate hydrolase

operon was excised from pGEM. The EcoRI site was blunted and the fragment cloned into 180

the SmaI site of pOT to give pOTtet (Table 1). This construct was digested with BstEII,

which removed 659-bp from the middle of the tetH, and the ends filled in using T4

polymerase. The ±1.45-kb NotI-SalI fragment from pSKM2 carrying the kanamycin

resistance cassette from Tn5 was also blunt-end cloned into the blunted BstEII sites of tetH

to give ptetH::Km (Fig. 1B). 185

Plasmids parsB::Km and ptetH::Km were each transformed into E. coli HB101 cells

which contained pR388 by using kanamycin and trimethoprim resistance selection. The

ability of both these suicide vectors to be mobilized was checked by transferring them by

conjugation on Luria agar from E. coli HB101-R388 to E. coli CSH56 with selection for

nalidixic acid and kanamycin resistance. The mobilizable suicide vectors were then 190

transferred from E. coli HB101-R388 to At. caldusT under the conditions described above.

DNA manipulations and sequencing. Plasmid preparation, restriction

endonuclease digestion, gel electrophoresis, ligation and Southern / colony-blot

hybridization were performed using standard methods or the manufacturers’

recommendations (27). Ultrapure plasmid DNA was obtained using the Wizard® Plus SV 195

miniprep DNA purification system from Promega. Total DNA from At. caldus was

prepared as previously described (15). Large scale plasmid preparations were made using

the Nucleobond® AX kit from Machery-Nagel. The sequence of the origin of transfer

amplification product, cloned into pGEM, was determined using an ABI PRISM™ 377

automated DNA sequencer. The sequence was analysed using the PC based DNAMAN 200

(version 4.1) package from Lynnon BioSoft.

Arsenic resistance assay of the wild-type At. caldusT strain versus the arsB

double crossover mutant strain. To test for growth of At. caldus in the presence of

arsenite, cells were cultured in the pH 4.5 thiosulphate medium containing 0.0, 0.25, 0.50,

0.75 and 1.0 mM arsenite. Early stationary phase cultures were diluted 2000-fold into fresh 205

medium containing As(III), incubated for 12 days and the cell density determined by

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

OD600 measurement. Growth in the presence of arsenate was not tested, as the phosphate

in the growth medium would contribute to apparent arsenate resistance (28).

Growth curve of the wild-type At. caldusT strain versus the tetH double mutant

crossover strain on tetrathionate. To determine the ability of the mutant to grow in 210

tetrathionate-containing media, cells from both the mutant and wild type were initially

cultured in the pH 4.5 thiosulphate-containing medium until early stationary phase. The

cells were then diluted 2000-fold in fresh tetrathionate media (pH 2.5) and the growth

monitored by measuring the change in optical density at 600 nm over 5 days. The growth of

the wild type was compared to that of the double crossover mutant. 215

Screening for single or double crossover mutants. Screening of transconjugants

to determine whether they were single or double crossovers was done by colony-blot

Southern hybridization. After plating a serial dilution of the 500 ml culture (inoculated with

cells directly after mating) on selective plates, Kmr colonies were picked and streaked on

fresh selective plates which were incubated at 37ºC for 5 days. The colony-blot was 220

performed using standard methods. The membrane was first probed against a 714-bp

internal fragment of the kanamycin resistance cassette under stringent conditions. This

indicated whether enough cell mass had transferred to the membrane and whether proper

lysis of cells had taken place so as to give a strong signal. The blot was stripped of the

kanamycin probe and reprobed this time using either the two ±200-bp arsB KpnI or the 225

659-bp tetH BstEII internal fragments under stringent conditions. The colonies that gave a

strong signal with both probes were considered to be single crossover mutants while those

that gave a strong signal for the kanamycin probe but gave no signal with the arsB probe

were possible double crossover mutants (Fig. 2 and Fig. 4).

230

RESULTS

Construction of the At. caldusT arsB and tetH mutants. To test the procedure for

the construction of a null mutant, we chose the arsB gene from the chromosomally encoded

arsenic resistance operon found in At. caldus (Fig. 1A). The arsenic resistance operon was 235

selected as it is one of the few areas of the At. caldus genome sequence that was publicly

available (15). Since the At. caldus ars operon has been well characterized with respect to

which other ORF’s are in its vicinity and how these genes are transcribed, we had

confidence that the inactivation of the arsB gene would not result in a lethal mutation (15).

Furthermore, inactivation of the gene should be easily confirmed by a simple growth assay 240

in the presence of arsenite.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

To make the At. caldus arsB mutant, a mobilizable suicide vector was constructed,

parsB::Km (see Materials and Methods). The suicide vector was based on pUC19 as this

plasmid has a narrow host-range ColE1 type replicon. The suicide vector carried a copy of

the arsB gene with an internal fragment replaced by the kanamycin resistance cassette from 245

E. coli transposon Tn5 (Fig. 1B). The conjugative plasmid pR388 was used to mobilize the

suicide vector from E. coli HB101 to At. caldusT as described above. After several

unsuccessful attempts to detect mutants directly after mating by plating on selective media,

it was decided to inoculate cells directly after mating into fresh liquid media with selection

to try and amplify the number of mutants to a detectable level. After 6-7 days of incubation 250

at 37ºC, growth was observed and the cells were harvested, washed and serial dilutions

plated on kanamycin selective plates.

Of the colonies that grew on selective media, 140 were picked and streaked onto fresh

solid selective media. Colony-blot Southern hybridization was used to screen the

transconjugants to determine which of these were possibly single or double crossover 255

mutants. All 140 colonies probed using a 714-bp PCR fragment internal to the kanamycin

resistance gene gave a signal suggesting that all were transconjugants. When the blot was

then stripped of the kanamycin probe and reprobed with the two ±200-bp KpnI fragments

internal to the arsB gene, six colonies did not give a positive hybridization signal using this

second probe. As the probe fragments should not be present on the chromosome of double 260

crossover mutants, those colonies that gave a signal with the kanamycin gene but not the

internal arsB fragment were considered to be potential double crossover mutants (Fig. 2).

The ratio of single to potential double crossover mutants was therefore approximately 1 to

23.

A second suicide vector targeting the tetH gene was constructed (ptetH::Km) in a 265

manner similar to the vector used in disruption of the arsB gene (Fig. 1B). This vector

carried a copy of the tetH interrupted by the kanamycin resistance cassette from E. coli

transposon Tn5 and could be mobilized by the IncW plasmid R388. Having determined

that an “enrichment” step was required to amplify mutants to a detectable level, cells were

collected after conjugation and inoculated into 500 ml liquid culture with selection and 270

incubated at 37ºC for 6-7 days after which growth was observed. Cells were harvested,

washed and a serial dilution plated on selective media.

All of the 249 colonies picked gave a signal following colony-blot Southern

hybridization using the 714-bp PCR fragment internal to the kanamycin resistance gene as

probe, suggesting that all were transconjugants. The blot was then stripped of the 275

kanamycin probe and reprobed with the 659-bp BstEII fragment internal to the tetH gene.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

Only one colony did not give a positive hybridization signal using this second probe

indicating that this was a possible double crossover mutant. The ratio of single to potential

double crossover mutants was therefore 1 in 249.

To determine the frequency at which the suicide vector was mobilized from E. coli to 280

At. caldus, the vector pOTF101 was constructed. The vector consisted of the pTC-F14 IncQ

replicon and kanamycin resistance cassette, cloned into pOT instead of the 2.2-kb PstI-

PvuII fragment from the At. caldus chromosomally encoded arsenic operon (Table 1). As

plasmid pTC-F14 originated from At. caldus (7) this would allow the plasmid to replicate in

At. caldus and permit us to determine the frequency at which the IncW oriT is transferred 285

from E. coli to At. caldus. As with the mutant construction matings, plasmid R388 was

used as the conjugative plasmid to transfer pOTF101 from E. coli HB101 to At. caldus.

Mating frequencies varied considerably within the range 10-5 and 10-7 transconjugants per

recipient.

Genetic characterization of arsB and tetH mutants. To demonstrate that the 290

potential double and single crossover mutants identified by colony-blot Southern

hybridization had a gene layout that would be predicted after homologous recombination

(Fig. 2), genomic DNA was prepared from randomly selected clones. These were

compared by Southern blot hybridization and PCR analysis to DNA from the wild-type

strain. 295

A Southern blot was prepared using genomic DNA from wild-type At. caldus as well

as putative double and single crossover arsB mutants digested with restriction

endonucleases BstEII, SalI and PvuI (Fig. 3). When an internal fragment of the arsB gene

was used as the probe (Fig. 3A), the chromosomal DNA from the putative double crossover

mutant did not give a signal indicating the absence of these fragments in double crossover 300

mutants and confirming the colony-blot Southern hybridization result. When wild type

DNA was the target, this probe gave 2.6-kb BstEII, ±7-kb SalI and 3.5-kb PvuI signals,

whereas 5.4-kb BstEII, 9.2-kb SalI and 3.5-kb PvuI signals were observed for the putative

single crossover mutant (Fig. 2 and 3A). When using an internal fragment of the

kanamycin gene as a probe, the wild type DNA did not give a signal as expected. Genomic 305

DNA from the putative double crossover mutant gave signals of 3.5-kb, 3.8-kb and 4.4-kb

for BstEII, SalI and PvuI digested DNA respectively (Fig. 2 and 3B). In contrast, signals of

3.5-kb, 9.2-kb and 4.4-kb were obtained for BstEII, SalI and PvuI digests respectively when

using genomic DNA from the putative single crossover mutants. From this data it was

concluded that the single crossover took place according to option B in Fig. 2. 310

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

Using PCR primers internal to ORFs 1 and ORF5 that flank the arsBRC operon, the

At. caldus wild-type and clone of the wild-type genes gave a similar sized (±2.7-kb) PCR

fragment as predicted (Figs. 2 and 4). The putative single and double crossover mutants

gave similar sized but larger PCR fragments (±3.7-kb) due to the replacement of a small

piece of the arsB gene with a larger fragment containing the kanamycin resistance gene. 315

This also indicated that the single crossover mutant was option “Single B” rather than

option “Single A” (Figs. 2 and 4). Although there are two places where the ORF5 and

arsB-Crossover primers can anneal, the elongation times for the PCR reactions were chosen

so as to give fragments in the <5.0-kb size range only. When using PCR primers internal to

arsB and ORF7, similar sized (±2.4-kb) fragments were obtained for the wild type and 320

cloned wild type DNA as expected (Fig. 4). An identical size fragment was also obtained

for the single crossover (Fig. 4), as predicted from single crossover option “Single B” (Fig.

2). A larger ±3.2-kb fragment (Fig. 4) was obtained for the double crossover mutant as

predicted in Fig. 2.

A Southern blot was also prepared using genomic DNA from wild-type At. caldus as 325

well as putative tetH double crossover mutant digested with restriction endonucleases KspI

and a DraI+SalI double digest (Fig. 5). When a fragment internal to the tetH was used as

the probe, the chromosomal DNA from the putative double crossover mutant did not give a

signal (Fig. 5A). When wild type DNA was the target, this probe gave a 3.4-kb KspI and

4.2-kb DraI+SalI signals (Fig. 5A). When using an internal fragment of the kanamycin 330

gene as a probe, the wild type DNA did not give a signal as expected. Genomic DNA from

the putative double crossover mutant gave signals of 4.2-kb KspI and 5-kb DraI+SalI

digested DNA respectively (Fig. 5B).

Taken together these results suggest that the selected arsB potential single crossover

clone was a single crossover mutant and has a gene arrangement similar to “Single B” as 335

shown in Fig. 2. It also shows that the selected potential double crossover clone is indeed a

double crossover mutant and that in this mutant the arsB gene has been interrupted by the

kanamycin resistance cassette. Similarly, the potential tetH double crossover mutant was

confirmed and that the tetH has been disrupted by the kanamycin resistance cassette in this

clone. 340

Arsenite resistance in At. caldusT wild type and its arsB disrupted mutant. To

compare the ability of the wild type and arsB double crossover mutant to cope with As(III)

toxicity we compared their growth after 12 days in liquid media at a range of As(III)

concentrations (Fig. 6A). The arsB double crossover mutant displayed a greatly reduced

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

capacity in dealing with As(III) compared to the wild type and grew poorly, or not at all, in 345

> 0 mM As(III) (Fig. 6A).

Growth of the At. caldusT wild type and its tetH disrupted mutant on

tetrathionate. To determine whether or not the proposed tetH is responsible for the

decomposition of tetrathionate to other sulfur intermediates allowing the bacterium to

utilize it as sole energy source, the wild type and double crossover mutants were initially 350

cultured on thiosulphate media and then switched to media containing only tetrathionate as

sulfur source. Figure 6B shows that whereas the wild type cells grew relatively well in

tetrathionate medium, the mutant did not grow to a level that was detectable using optical

density measurements. To determine if any mutant growth had taken place, the CFU of

both wild type and mutant cultures was determined on day 5. The CFU for the mutant 355

(tetH::Km) was 1.3 X 106 CFU/ml ± 0.8 X 106 while the CFU for the wild type was 1.6 X

108 CFU/ml ± 0.4 X 108. As the CFU of the cultures after inoculation was ± 8 X 104

CFU/ml, growth of the mutant had taken place although approximately 100-fold reduced

compared with wild type cells.

360

DISCUSSION

This is the first report of the construction of a null mutant of the bacterium At. caldus

using marker exchange mutagenesis. Inactivation of the arsenite pump arsB was used as a

model to demonstrate how this system could be used to study other genes of interest from 365

the organism. Although plasmid transfer to At. caldus has been reported (15, 19, 32), the

ability to generate mutants has not been demonstrated previously, possibly due to the

difficulty in detecting these low frequency events. Interruption of the At. caldusT arsB gene

with the kanamycin resistance cassette was confirmed using molecular analysis and

resulted in a marked increase in sensitivity to arsenite displayed by the arsB double mutant 370

compared to the wild type strain. This also demonstrated that there are no other detectable

mechanisms of arsenite resistance in At. caldusT and shows the benefit of being able to

create null mutants.

The transfer frequency of the pOTF101 plasmid varied widely between 10-5 and 10-7

transconjugants per recipient. As the protocol for all conjugation experiments was 375

standardized this result shows that, under these mating conditions, even slight variations

can have a significant effect on the frequency of plasmid transfer and therefore the probable

success rate of constructing and selecting mutants by homologous recombination. For E.

coli the frequency at which homologous recombination occurs has been found to be

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

approximately 10-3 to 10-4 less than the transformation frequency (10). If one speculates 380

that this same proportion holds for At. caldus, at a transformation efficiency of 10-5

transconjugants per recipient, the best case scenario would result in a single crossover

recombinant occurring at a rate of 10-8 per recipient. The occurrence of double crossover

recombinants would be even lower with only 1 in 23 of the crossovers being double

crossovers in the case of the arsB mutant. This frequency could however be much lower in 385

the worst case scenario (10-7 transconjugants per recipient and 10-4 recombination

efficiency) which then makes it exceedingly difficult to isolate a double crossover mutant

and probably shows why the “enrichment” step after conjugation was required to isolate

mutants. The ratios of 1 in 23 for the arsB or 1 in 249 for the tetH reported in this study are

not accurate estimates. The numbers of double crossovers were too small to allow for 390

statistical validity and could be further skewed as this ratio was not measured directly after

mating, but only once the cells had been cultured long enough to bring the recombinants

into detection range. If, either the single or double crossover mutant has a slightly different

growth rate compared to the other, the ratio would be affected.

Reasons why there are relatively more double crossover mutants for the arsB than the 395

tetH could be partly due the smaller total area available for recombination on the tetH

vector (1807bp for arsB and 1622bp for tetH) as well as the relative sizes of the areas on

either side of the kanamycin gene for the two different vectors (802bp-1005bp for the arsB

and 929bp-693bp for the tetH) (1, 10, 11). Besides this it has also been reported that

depending on the locus being targeted, the frequency of recombination can vary by several 400

orders of magnitude (21).

Efforts have been made to study the metabolism of this organism with respect to how

it oxidizes RISC’s and how it generates energy, in the form of ATP, from this process (2, 4,

6, 9, 26). The ability to knockout genes involved in these processes has however been

lacking, making it difficult to confirm their involvement or to identify other pathways 405

involved. We have therefore, constructed a null mutant for the chromosomally encoded

tetrathionate hydrolase gene, tetH, which is thought to be responsible for the hydrolysis of

tetrathionate to thiosulfate, sulfur, and sulfate which At. caldus uses as energy sources (2,

26). The observation using cell counts that the tetH mutants were able to grow on

tetrathionate media, although very poorly, could be due to one of two reasons. There may 410

be an additional, less efficient mechanism of tetrathionate hydrolysis or alternately the

tetrathionate in the medium may have been unstable such that there was a natural

conversion to other RISC’s allowing growth to take place. These possibilities have yet to be

investigated and work is underway to characterize this mutant physiologically.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

The addition of this technique to the suite of tools already available for At. caldus can 415

be considered a significant milestone towards our ability to study the organism.

ACKNOWLEDGEMENTS

This work was funded by grants from the National Research Foundation (Pretoria), BHP-420

Billiton Johannesburg Technology Centre and the BioMinE project 500329 of the EU

framework 6. We thank Mark Dopson and Olena Rzhepishevska (University of Umeå) for

providing us with the sequence data used for the construction of the tetH mutant.

REFERENCES 425

1. Biswas, I., A. Gruss, D. S. Ehrlich, and E. Maguin. 1993. High-efficiency gene

inactivation and replacement system for gram-positive bacteria. J. Bacteriol. 175: 3628-

3635

430

2. Bugaytsova, Z., and E. B. Lindström. 2004. Localization, purification and properties of

a tetrathionate hydrolase from Acidithiobacillus caldus. Eur. J. Biochem. 271: 272-280.

3. Datta, N., and R. W. Hedges. 1972. Trimethoprim resistance conferred by W plasmid

in Enterobacteriaceae. J. Gen. Microbiol. 72: 349-355. 435

4. Dopson, M., and E. B. Lindström. 1999. Potential role of Thiobacillus caldus in

arsenopyrite bioleaching. 1999. Appl. Environ. Microbiol. 65: 36-40.

5. Dopson, M., and E. B. Lindström. 2004. Analysis of community composition during 440

moderately thermophilic leaching of pyrite, arsenical pyrite and chalcopyrite. Microb.

Ecol. 48: 19-28.

6. Dopson, M., E. B. Lindström, and K. B. Hallberg. 2002. ATP generation during

reduced inorganic sulfur compound oxidation by Acidithiobacillus caldus due to electron 445

transport phosphorylation. Extremophiles. 6: 123-129.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

7. Gardner, M. N., S. M. Deane, and D. E. Rawlings. 2001. Isolation of a new broad-

host-range IncQ-like plasmid, pTC-F14, from the acidophilic bacterium Acidithiobacillus

caldus and analysis of the plasmid replicon. J. Bacteriol. 183: 3303-3309. 450

8. Hallberg, K. B., and E. B. Lindström. 1994. Characterization of Thiobacillus caldus

sp. nov., a moderately thermophilic acidophile. Microbiology. 140: 3451-3456.

9. Hallberg, K. B., M. Dopson, and E. B. Lindström. 1996. Reduced sulfur compound 455

oxidation by Thiobacillus caldus. J. Bacteriol. 178: 6-11.

10. 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: 4617-4622. 460

11. Hasty, P., J. Rivera-Pérez, and A. Bradley. 1991. The length of homology required

for gene targeting in embryonic stem cells. Mol. Cell. Biol. 11: 5586-5591

12. Hawkes, R.B., P.D. Franzmann, and J.J. Plumb. 2006. Moderate thermophiles 465

including “Ferroplasma cyprexacervatum” sp. nov., dominate an industrial scale chalcocite

heap bioleaching operation. Hydrometallurgy. 83: 229-236.

13. Jin, S. M., W. M. Yan, and Z. N. Wang. 1992. Transfer of IncP plasmids to extremely

acidophilic Thiobacillus thiooxidans. Appl. Environ. Microbiol. 58: 429-430. 470

14. Kelly, D. P., and A. P. Wood. 2000. Reclassification of some species of Thiobacillus

to the newly designated genera Acidithiobacillus gen. nov., Halothiobacillus gen. nov. and

Thermithiobacillus gen. nov. Int. J. Syst. Evol. Microbiol. 50: 511-516.

475

15. Kotze, A. A., M. I. Tuffin, S. M. Deane, and D. E. Rawlings. 2006. Cloning and

characterization of the chromosomal arsenic resistance genes from Acidithiobacillus caldus

and enhanced arsenic resistance on conjugal transfer of ars genes located on transposon

TnAtcArs. Microbiology. 152: 3551-3560.

480

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

16. Kusano, T., K. Sugawara, C. Inoue, T. Takeshima, M. Numata, and T. Shiratori.

1992. Electrotransformation of Thiobacillus ferrooxidans with plasmids containing a mer

determinant. J. Bacteriol. 174: 6617-6623.

17. Lacks, S., and B. Greenberg. 1977. Complementary specificity of restriction 485

endonucleases of Diplococcus pneumoniae with respect to DNA methylation. J. Mol. Biol.

114: 153–168.

18. Liu, Z., N. Guiliani, C. Appia-Ayme, F. Borne, J. Ratouchniak, and V. Bonnefoy.

2000. Construction and characterization of a recA mutant of Thiobacillus ferrooxidans by 490

marker exchange mutagenesis. J. Bacteriol. 182: 2269-2276.

19. Liu, X., J. Lin, Z. Zhang, J. Bian, Q. Zhao, Y. Liu, J. Lin, and W. Yan. 2007.

Construction of conjugative gene transfer system between E. coli and moderately

thermophilic, extremely acidophilic Acidithiobacillus caldus MTH-04. J. Microbiol. 495

Biotechnol. 17: 162-167.

20. Okibe, N., M. Gericke, K. B. Hallberg, and D. B. Johnson. 2003. Enumeration and

characterization of acidophilic microorganisms isolated from a pilot plant stirred-tank

bioleaching operation. Appl. Environ. Microbiol. 69: 1936-1943. 500

21. Parish, T., and N. G. Stoker. 2000. Use of a flexible cassette method to generate a

double unmarked Mycobacterium tuberculosis tlyA plcABC mutant by gene replacement.

Microbiology 146:1969–1975.

505

22. Peng, J.-B., W.-M. Yan, and X.-Z. Bao. 1994. Plasmid and transposon transfer to

Thiobacillus ferrooxidans. J. Bacteriol. 176: 2892-2897.

23. Peng, J.-B., W.-M. Yan, and X.-Z. Bao. 1994b. Expression of heterogenous arsenic

resistance genes in the obligately autotrophic biomining bacterium Thiobacillus 510

ferrooxidans. Appl. Environ. Microbiol. 60: 2653-2656.

24. Rawlings, D. E., N. J. Coram, M. N. Gardner, and S. M. Deane. 1999. Thiobacillus

caldus and Leptospirillum ferrooxidans are widely distributed in continuous flow

biooxidation tanks used to treat a variety of metal containing ores and concentrates, p. 777-515

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

786. In R. Amils and A. Ballester (ed.), Biohydrometallurgy and the environment toward

the mining of the 21st Century, Part A. Elsevier, Amsterdam, The Netherlands.

25. Rawlings, D.E., and D.B. Johnson. 2007. The microbiology of biomining:

development and optimization of mineral-oxidizing microbial consortia. Microbiology 153: 520

315-324.

26. Rzhepishevska, O. I., J Valdés, L. Marcinkeviciene, C. A. Gallardo, R. Meskys, V.

Bonnefoy, D. S. Holmes, and M. Dopson. 2007. Regulation of a novel Acidithiobacillus

caldus gene cluster involved in metabolism of reduced inorganic sulfur compounds. Appl. 525

Environ. Microbiol. 73: 7367-7362.

27. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory

manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

530

28. Silver, S., K. Budd, K. M. Leahy, W. V. Shaw, D. Hammond, R. P. Novick, J. R.

Willsky, M. H. Malamy, and H. Rosenburg. 1981. Inducible plasmid-determined

resistance to arsenate, arsenite, and antimony(III) in Escherichia coli and Staphylococcus

aureus. J. Bacteriol. 172: 424–430.

535

29. Silverman, M. P., and D. G. Lundgren. 1958. Studies on the chemoautotrophic iron

bacterium Ferrobacterium ferrooxidans. I. An improved medium and a harvesting

procedure for securing high cell yields. J. Bacteriol. 77: 642-647.

30. Tian, K. L., J. Q. Lin, X. M. Liu, Y. Liu, C. K. Zhang, and W. M. Yan. 2003. 540

Conversion of an obligate autotrophic bacteria to heterotrophic growth: expression of a

heterogeneous phosphofructokinase gene in the chemolithotroph Acidithiobacillus

thiooxidans. Biotechnol. Lett. 25: 749-754.

31. Yannisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning 545

vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene

33: 103-119.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

32. Zhao, Q., X. M. Liu, Y. Zhan, J. Q. Lin, W. M. Yan, J. Bian, and Y. Liu. 2005.

Construction of an engineered Acidithiobacillus caldus with high-efficiency arsenic 550

resistance. Wei Sheng Wu Xue Bao. 45: 675-679.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

TABLE 1. Bacteria and plasmids used in this study

α DSMZ, Deutsche Sammlung von Mikroorganismen und Zellkulturen

Strain or plasmid Genotype or Description Source or reference

Strains

At. caldus

KU (DSMZ α

8584) Kingsbury Coal Spoil, UK DSMZ

E. coli

DH5α F’/endA1 hsdR17(rK-mK

+) supE44 thi-1 recA1 gyrA(Nal

r)

relA1 ∆(lacZYA-argF)U169 (φ80dlac∆(lacZ)M15)

Promega Corp., Madison, Wis.

HB101 F- ∆(mcrC-mrr) hsdS20(rB- mB-) recA13 ara-14 proA2 lacY1

λ- galK2 rpsL20(Sm

r) Xyl-5 mtl-1 leuB6 thi-1 supE44

(17)

CSH56 F- ara ∆(lacpro) supD nalA thi Cold Spring Harbor, NY

Plasmids

pAtcars4 Asr Ap

r; 10-kb Sau3A fragment of At. caldus #6 cloned into

the BglII site of pEcoR252

(15)

pSKM2 Apr Km

r; ±1.45-kb HindIII-SmaI kanamycin resistance

cassette from Tn5 cloned into the HindIII-SmaI sites on

pBluescript(SK)

Stellenbosch University lab

collection (Douglas Rawlings )

pGEM-T Apr; T-tailed PCR product cloning vector Promega Corp.

pUC19 Apr; lacZ’,ColE1 replicon, cloning vector (31)

pOT Apr; the 557bp origin of transfer region from pR388 cloned

into the EcoRI site of pUC19

This study

pOTars Apr; a 2234bp PstI-PvuII fragment from pAtcars4 containing

part of the chromosomally encoded arsenic resistance operon

cloned into the PstI and blunted SacI site of pOT

This study

pOTtet Apr; a 2281bp DraI-EcoRI (blunted) fragment containing

part of the chromosomally encoded tetrathionate hydrolase

operon cloned into the SmaI site of pOT

This study

parsB::Km Apr Km

r; the kanamycin resistance cassette from Tn5 cloned

into blunted KpnI sites internal to the arsB gene on pOTars

This study

ptetH::Km Apr Km

r; the kanamycin resistance cassette from Tn5 cloned

into blunted BstEII sites internal to the tetH gene on pOTtet

This study

pTC-F101 Kmr; 6434bp HindIII-SphI pTC-F14 (IncQ) replicon

fragment joined to the ±1.45-kb HindIII-SmaI kanamycin

resistance cassette from Tn5

(7)

pOTF101 Apr Km

r; pOT joined to pTC-F101 using the unique HindIII

site on each plasmid

This study

pR388 TraW IncW Tpr Su

r (3)

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

Figure legends

FIG 1. A) Arrangement of the arsenic resistance genes and the tetrathionate hydrolase

operon on the At. caldus genome. Solid boxes below ORFs indicate the regions used in

construction of the suicide vectors (parsB::Km and ptetH::Km). The fragments removed

between the KpnI sites (in bold) internal to the arsB and between the BstEII sites (in

bold) in the tetH were used as probes to distinguish between single and double crossover

mutants. The positions of the primers used to amplify the 2.3-kb fragment (TetH For and

TetH Rev) used in construction of ptetH::Km are indicated. B) The parsB::Km and

ptetH::Km suicide vectors used to generate doublecross over mutants. The 427bp KpnI

fragment internal to the arsB and the 659bp BstEII fragment internal to the tetH gene

were replaced with the kan gene.

FIG 2. Construction of single and double crossover recombination mutants. On

introduction of the non-replicating vector parsB::Km into At. caldus KU, a single

recombination event results in one of two possible single crossover strains, Single A or

Single B. A second recombination event can result either in restoration of the wild type

gene or a mutant double crossover strain. The wild type should not be observed due to

selection with kanamycin. Annealing sites for primers used to characterize the various

recombinants are indicated. Solid boxes indicate where the arsB probe hybridizes while

hatched boxes indicate the hybridization position of the kanamycin probe.

FIG 3. Southern blot analysis of BstEII-, SalI- and PvuI-digested DNA isolated from At.

caldus KU-WT, a double crossover arsB mutant and a single crossover arsB mutant. The

blots were probed with either a fragment internal to the arsB gene (A) or a fragment

internal to the kanamycin resistance gene (B). The numbers between the panels indicate

the sizes (in kilobases) of some fragments from the λ PstI molecular size marker.

FIG. 4. PCR analyses of selected single and double crossover mutants of At. caldus KU

and pAtcars4 using primer pairs ORF1-ORF5 and arsB-ORF7 (Fig. 2). See text for

details.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

FIG 5. Southern blot analysis of KspI- and DraI+SalI-digested DNA isolated from At.

caldus KU-WT and a double crossover mutant. The blots were probed with either a

fragment internal to the tetH gene (A) or a fragment internal to the kanamycin resistance

gene (B). The numbers between the panels indicate the sizes (in kilobases) of some

fragments from the λ PstI molecular size marker.

FIG. 6. A) Growth of At. caldus KU-WT (�), and KU-arsB::Km (�) after 12 days in the

presence of various concentrations of arsenite. Cell densities were determined (OD600)

and represented as a percentage of growth in the absence of arsenite. B) Growth of At.

caldus KU-WT (�), and KU-tetH::Km (�) on tetrathionate containing media over 5

days. Each data point represents duplicate results of at least two experiments. Error bars

indicate SD.

ACCEPTED

on April 10, 2019 by guest

http://aem.asm

.org/D

ownloaded from

pasrB::Km

bla

IncW-oriT

lacZoriE

ORF6’

ORF5arsR’

arsB’ arsB’

kan

ptetH::Km

doxD’

tetH’tetH’

kan

lacZ

bla

IncW-oriT

oriE

1

Pvu

II(3

53

6)

6500

Bst

EII

(53

9)

Pst

I(9

49

)

Pst

I(1

302

)P

vuI

(14

58

)

Kp

nI

(210

4)

Kp

nI

(230

2)

Kp

nI

(253

1)

Bst

EII

(31

39

)

Kpn

I(4

081

)

Pvu

I(4

984

)S

alI

(50

07

)

arsBarsRarsC ORF5 ORF6 ORF7 ORF8ORF1

Pvu

II(4

09

0)

1 8827

tetHisac isacdoxD nodTrsrS rsrR

Ksp

I(3

240

)

Bst

EII

(62

69

)

Sm

aI

(8417

)

Ksp

I(2

72)

Dra

I(1

734

)

Ksp

I(7

764

)

Sm

aI

(7711

)

Bst

EII

(2663

)

Bst

EII

(2801

)

Ksp

I(6

210

)

Bst

EII

(2670

)

Bst

EII

(3322

)

SalI

(59

28

)

Ksp

I(3

676

)

Eco

RI

(401

5)

Sm

aI

(0)

TetH For TetH Rev

A)

B)

Fig 1

ACCEPTED on A

pril 10, 2019 by guesthttp://aem

.asm.org/

Dow

nloaded from

1 6500arsBarsRarsC ORF5 ORF6 ORF7 ORF8ORF1

1 12793

SalI

(74

74

)

Bst

EII

(53

9)

Pvu

II(9

82

9)

Pvu

I(6

828

)

Kpn

I(1

037

4)

Pvu

II(1

03

83

)

Bst

EII

(31

39

)

Pvu

I(1

458

)

Bst

EII

(94

32

)

Kpn

I(2

104

)

SalI

(11

300

)

Kpn

I(2

302

)

Pvu

II(8

05

9)

Kpn

I(2

531

)

Pvu

II(7

69

4)

Pvu

I(1

1277

)

Pst

I(9

49

)

Pst

I(1

302

)

Pvu

I(4

113

)

Pvu

I(5

009

)

Pvu

II(4

08

4)

Pvu

II(6

44

8)

arsBarsRarsC ORF5 ORF6 ORF7 ORF8ORF1 arsB’kanarsB’arsR’ORF6’ORF5 Vector

ORF1-RTRev arsB-Crossover ORF5-RTRev ORF7-Crossover

1 12793

Bst

EII

(53

9)

Kpn

I(1

037

4)

Pvu

II(1

03

83

)

SalI

(11

300

)

Pvu

I(1

1277

)

Pst

I(9

49

)

Pst

I(1

302

)

Pvu

II(9

82

9)

Bst

EII

(94

32

)

Pvu

I(7

751

)

Kpn

I(8

397

)

Kpn

I(8

595

)

Kpn

I(8

824

)

SalI

(21

04

)

Pvu

I(1

458

)

Bst

EII

(40

62

)

Pvu

II(2

68

9)

Pvu

II(2

32

4)

Pvu

I(5

045

)

Pvu

I(5

932

)

Pvu

II(5

00

7)

Pvu

II(7

37

1)

arsB’arsRarsC ORF5 ORF6 ORF7 ORF8ORF1 kan arsBarsR’ORF6’ORF5 VectorarsB’

ORF1-RTRev arsB-Crossover ORF5-RTRev ORF7-CrossoverORF5-RTRevarsB-Crossover

arsB-CrossoverORF5-RTRev

Single recombination event

Second recombination event

OR

1 7423

Bst

EII

(53

9)

Pst

I(9

49

)

Pst

I(1

302

)P

vuI

(14

58

)

SalI

(21

04

)

Bst

EII

(40

62

)

Pvu

II(4

45

9)

Kpn

I(5

004

)

Pvu

II(5

01

3)

Pvu

I(5

907

)S

alI

(59

30

)

Pvu

II(2

32

4)

Pvu

II(2

68

9)

arsB’arsRarsC ORF5 ORF6 ORF7 ORF8ORF1 arsB’kan

ORF1-RTRev ORF5-RTRev ORF7-CrossoverarsB-Crossover

Double

Single B

Single A

pasrB::Km

bla

IncW-oriT

lacZoriE

ORF6’

ORF5arsR’

arsB’ arsB’

kan

1

Pvu

II(3

53

6)

6500

Bst

EII

(53

9)

Pst

I(9

49

)

Pst

I(1

302

)P

vuI

(14

58

)

Kpn

I(2

104

)K

pn

I(2

302

)K

pn

I(2

531

)

Bst

EII

(31

39

)

Kpn

I(4

081

)

Pvu

I(4

984

)S

alI

(50

07

)

arsBarsRarsC ORF5 ORF6 ORF7 ORF8ORF1

Pvu

II(4

09

0)

ORF1-RTRev ORF5-RTRev ORF7-CrossoverarsB-Crossover

WT

OR

Fig 2

ACCEPTED on A

pril 10, 2019 by guesthttp://aem

.asm.org/

Dow

nloaded from

Pvu

I

Bst

EII

SalI

Pvu

I

SalI

Bst

EII

Pvu

I

SalI

Bst

EII

λP

stI

KU-WT Double Single

BA

Pvu

I

Bst

EII

SalI

Pvu

I

SalI

Bst

EII

Pvu

I

SalI

Bst

EII

λP

stI

KU-WT Double Single

14.0

5.04.5

2.8

11.5

1.7

Fig 3

ACCEPTED on A

pril 10, 2019 by guesthttp://aem

.asm.org/

Dow

nloaded from

Fig 4

pA

tcar

s4

λP

stI

pA

tcar

s4

Do

ub

le

Sin

gle

KU

-WT

Do

ub

le

Sin

gle

ORF1 and ORF5 arsB and ORF7

2.84kb

5.08kb4.5kb

14kb

1.99kb

1.70kb

KU

-WT

ACCEPTED on A

pril 10, 2019 by guesthttp://aem

.asm.org/

Dow

nloaded from

Fig 5

B

Dra

I+

SalI

λP

stI

KU-WT Double

Dra

I+

SalI

Ksp

I

Ksp

I

Dra

I+

SalI

λP

stI

KU-WT Double

Dra

I+

SalI

Ksp

I

Ksp

I

A

2.8

14.011.5

1.7

5.04.5

ACCEPTED on A

pril 10, 2019 by guesthttp://aem

.asm.org/

Dow

nloaded from

Fig 6

KU-WT

KU-tetH::Km

0

0.01

0.02

0.03

0.04

0.05

0.06

0 24 48 72 96 120O

D600nm

Time in hours

B

KU-WT

KU-arsB::Km

0

20

40

60

80

100

120

140

160

OD

600nm

as

a %

of

0m

M A

sIII

0 0.25 0.5 0.75 1

AsIII concentration in mM

A

ACCEPTED on A

pril 10, 2019 by guesthttp://aem

.asm.org/

Dow

nloaded from