Specialized transduction: an efficient method for generating marked and unmarked targeted gene disruptions in Mycobacterium tuberculosis, M. bovis BCG and M. smegmatis

Stoyan Bardarov1, Svetoslav Bardarov, Jr1, Martin S. Pavelka, Jr4, Vasan Sambandamurthy1,2, Michelle Larsen1,2, JoAnn Tufariello1,3, John Chan1,3, Graham Hatfull5 and William R. Jacobs, Jr1,2

Dept of Microbiology and Immunology1 and Howard Hughes Medical Institute2, Albert Einstein College of Medicine, Bronx, NY 10461, USA
Division of Infectious Diseases, Department of Medicine, Montefiore Medical Center, Bronx, NY, USA3
Department of Microbiology and Immunology, University of Rochester Medical Center, 601 Elmwood Ave., Rochester, NY 14642, USA4
Department of Biological Sciences, Bacteriophage Institute, University of Pittsburgh, 365A Crawford Hall, Pittsburgh, PA 15260, USA5

Author for correspondence: William R. Jacobs, Jr. Tel: +1 718 430 2888. Fax: +1 718 430 8844. e-mail: jacobsw{at}hhmi.org


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The authors have developed a simple and highly efficient system for generating allelic exchanges in both fast- and slow-growing mycobacteria. In this procedure a gene of interest, disrupted by a selectable marker, is cloned into a conditionally replicating (temperature-sensitive) shuttle phasmid to generate a specialized transducing mycobacteriophage. The temperature-sensitive mutations in the mycobacteriophage genome permit replication at the permissive temperature of 30 °C but prevent replication at the non-permissive temperature of 37 °C. Transduction at a non-permissive temperature results in highly efficient delivery of the recombination substrate to virtually all cells in the recipient population. The deletion mutations in the targeted genes are marked with antibiotic-resistance genes that are flanked by {gamma}{delta}-res (resolvase recognition target) sites. The transductants which have undergone a homologous recombination event can be conveniently selected on antibiotic-containing media. To demonstrate the utility of this genetic system seven different targeted gene disruptions were generated in three substrains of Mycobacterium bovis BCG, three strains of Mycobacterium tuberculosis, and Mycobacterium smegmatis. Mutants in the lysA, nadBC, panC, panCD, leuCD, Rv3291c and Rv0867c genes or operons were isolated as antibiotic-resistant (and in some cases auxotrophic) transductants. Using a plasmid encoding the {gamma}{delta}-resolvase (tnpR), the resistance genes could be removed, generating unmarked deletion mutations. It is concluded from the high frequency of allelic exchange events observed in this study that specialized transduction is a very efficient technique for genetic manipulation of mycobacteria and is a method of choice for constructing isogenic strains of M. tuberculosis, BCG or M. smegmatis which differ by defined mutations.

Keywords: mycobacteriophage, homologous recombination, allelic exchange

Abbreviations: AES, allelic exchange substrate; MCS, multiple cloning site


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Tuberculosis continues to be a major public health problem. With the rapid increase of drug-resistant strains of Mycobacterium tuberculosis, the need for an improved vaccine to combat the disease has become an international priority (see http://www.who.int/inf-fs/en/fact104.html). The determination of the genome sequence of two strains of M. tuberculosis (Cole et al., 1998 ; Fleischmann, 2002 ), available in two public domains (http://genolist.pasteur.fr/TubercuList and http://www.tigr.org), promises to accelerate the increase in our knowledge concerning the fundamental biology of these organisms. However, an efficient methodology to generate defined mutations is urgently needed for the rapid development of the functional genomics of the clinically important pathogenic mycobacteria in order to understand the genetic basis for their tropism, virulence and persistence in the host.

Mutant isolation and gene transfer strategies have been successfully used for the fast-growing mycobacteria such as Mycobacterium smegmatis (Boshoff & Mizrahi, 2000 ; Braunstein et al., 2001 ; Frischkorn et al., 1998 ; Knipfer et al., 1997 ; Pavelka & Jacobs, 1999 ). However, in the slow-growing mycobacterial species the construction of genetically defined isogenic strains containing single or multiple mutations has been notoriously difficult. This is primarily due to the high frequency of illegitimate recombination in these organisms (Aldovini et al., 1993 ; Kalpana et al., 1991 ) as well as their intrinsic tendency to grow in aggregates (clumps), which makes the isolation of individual clones problematic. The difficulties encountered in early attempts at allelic exchange led to the general conclusion that homologous recombination in the slow-growing mycobacteria is inefficient (McFadden, 1996 ).

Several groups have reported successful gene disruptions in M. tuberculosis and Mycobacterium bovis BCG using short (Azad et al., 1996 ; Kalpana et al., 1991 ; Reyrat et al., 1995 ) or long linear DNA fragments (Balasubramanian et al., 1996 ) as allelic exchange substrates (AESs) for homologous recombination. A ‘suicide’ vector approach, using recombinant plasmids unable to replicate in mycobacteria, was extensively used to achieve allelic exchange in both fast- and slow-growing mycobacteria (Berthet et al., 1998 ; Fitzmaurice & Kolattukudy, 1998 ; Knipfer et al., 1997 ; Parish et al., 1999 ; Pavelka & Jacobs, 1996 ; Pelicic et al., 1996a , b , 1997 ; Sander et al., 1995 ). A two-step selection method using selectable and counterselectable markers, positioned on either replicating or non-replicating plasmids, has been also successfully used in M. smegmatis (Knipfer et al., 1997 ; Pelicic et al., 1996a ), M. bovis BCG and M. tuberculosis (Hinds et al., 1999 ; Parish et al., 1999 ; Parish & Stoker, 2000 ; Pavelka & Jacobs, 1999 ). Unfortunately, the suicide vector approach (using a non-temperature-sensitive plasmid) is dependent upon the delivery of the AES by electroporation. Because homologous recombination frequencies are very close to the efficiency at which plasmids can be electroporated into slow-growing mycobacteria, the suicide vector approach is limited to those cases where high transformation efficiencies can be obtained. It has been proposed that this electroporation limitation, not inefficient homologous recombination, is the reason for earlier difficulties encountered in allelic exchange experiments in slow-growing mycobacteria (Pavelka & Jacobs, 1999 ). The use of conditionally replicating temperature-sensitive plasmid replicons as delivery vectors has greatly improved reproducibility of allelic exchange in the slow-growing mycobacteria (Pelicic et al., 1997 ), although growth of the cultures at low temperature is required, which may not be expedient.

General transduction, which is based on the natural genetic exchange of DNA information, is an alternative strategy to efficiently introduce homologous DNA into the recipient cells by bacteriophages (Lenox, 1955 ; Masters, 1996 ; Zinder & Lederberg, 1952 ). Transductional transfer of AESs has greatly facilitated the generation of specific mutations and the functional analysis of the genomes of Escherichia coli and Salmonella. However, while the transfer of DNA by genetic transduction has been reported for M. smegmatis (Sundar Raj & Ramakrishnan, 1970 ), it has not been reported yet for the slow-growing mycobacteria such as M. bovis BCG and M. tuberculosis. The ability to transfer DNA to slow-growing mycobacteria by a highly efficient, phage-based method would overcome the electroporation limitations described above for the plasmid transformation methods of allelic exchange.

This report describes a novel genetic method for mycobacteria for the generation of targeted deletion mutations by allelic exchange using in vitro-generated specialized transducing mycobacteriophages. The utility and reproducibility of this method have been demonstrated by the construction of seven isogenic auxotrophic mutant strains of M. smegmatis, three substrains of M. bovis BCG and three strains of M. tuberculosis. The effectiveness of this method has also been shown by several other researchers, who successfully engineered numerous targeted gene disruptions in M. tuberculosis (Glickman et al., 2000 ; Raman et al., 2001 ; Sirakova et al., 2001 ; Steyn et al., 2002 ), using the reagents described or their derivatives, prior to submission of this article. We also demonstrate the efficient elimination of the resistance gene (unmarking the mutation) by using a plasmid expressing the {gamma}{delta}-TnpR site-specific resolvase, which acts on the directly repeated res sites flanking the resistance gene. Based on the observed high frequency of allelic exchange and the subsequent efficient removal of the marker gene we conclude that this is a powerful genetic method for engineering targeted marked and unmarked mutations in various mycobacterial species, particularly in the slow-growing pathogenic mycobacteria and those in which it is difficult to achieve efficient plasmid transformation.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, media and growth conditions.
The E. coli strains were grown in Luria–Bertani (LB) broth or on LB agar (Difco) for the amplification of recombinant clones, plasmid isolation and transformation, and in TYM broth (1% tryptone, 0·5% yeast extract, 0·5% NaCl, 0·2% maltose) for transduction with {lambda}-packaged cosmids. For the preparation of electrocompetent cells M. smegmatis mc2155 was grown in LB containing 0·5% (w/v) Tween 80 (LBT) (Pavelka & Jacobs, 1999 ). When required, the following antibiotics were used at the specified concentrations: carbenicillin (50 µg ml-1), kanamycin (25 µg ml-1) and hygromycin B (50 µg ml-1 for E. coli and 150 µg ml-1 for M. smegmatis). Mycobacterial strains (except for M. smegmatis) were grown in basal Middlebrook 7H9 (Difco) supplemented with 1x ADS [0·5% bovine serum albumin, fraction V (Boehringer Manheim), 0·2% glucose and 0·085% NaCl], 0·2% glycerol and 0·1% Tween 80 (7H9ADSTW broth) (Jacobs et al., 1991 ). Complete medium consisted of 7H9ADSTW supplemented with individual L-amino acids at a final concentration of 40 µg ml-1. Pantothenic acid and nicotinamide (Sigma) were used at 20 µg ml-1. For transduction experiments, cultures of the M. bovis BCG or M. tuberculosis strains were prepared by inoculating 10 ml 7H9ADSTW broth with 1 ml frozen stock into 30 ml plastic culture bottles and incubated at 37 °C in an incubator-shaker. For propagation of mycobacteriophages in M. smegmatis mc2155, 7H9 basal medium supplemented with 0·4% glycerol and 2 mM CaCl2 (7H9-Gly-CaCl2) was used as a bottom agar, and 0·6% agar in water supplemented with 2 mM CaCl2 as a top agar. High-titre mycobacteriophage lysates were prepared from phage propagated in M. smegmatis grown on 7H9-Gly-CaCl2 solid medium at 30 °C. Phage lysates were stored in MP buffer (50 mM Tris/HCl pH 7·6, 150 mM NaCl, 10 mM MgCl2 and 2 mM CaCl2). Adsorption of the phages was performed for 30 min at 37 °C for M. smegmatis and for 3 h at 37 °C for M. tuberculosis. Outgrowth of the infected cells was performed by transferring the phage/cell mixture to a roller bottle containing 50 ml complete 7H9ADSTW pre-warmed to 37 °C.

DNA manipulations.
DNA manipulations were done essentially as described by Sambrook et al. (1989) . High-molecular-mass chromosomal DNA from the mycobacterial strains was purified by the CTAB method as described previously (Balasubramanian et al., 1996 ). Phage DNA from mycobacteriophages was purified as previously described (Jacobs et al., 1991 ) with slight modifications. High-titre phage lysates (30 ml) were layered onto a 5 ml cushion of 50% glycerol in MP buffer in a 35 ml nitrocellulose centrifuge tube (Beckman) and centrifuged at 25000 r.p.m. for 2 h in an SW27 rotor at 15 °C. Phage pellets were resuspended in 500 µl MP buffer to which 25 µl (0·05 vol.) of STEP lysis solution (0·4 M Na2EDTA, 1% SDS, 50 mM Tris/HCl pH 8·0, 200 µg Proteinase K ml-1) was added and the suspension was incubated for 30 min at 56 °C. Following extraction with phenol/chloroform (1:1) and chloroform/isoamyl alcohol (24:1), phage DNA was precipitated by the addition of 2 vols 100% ethanol, washed in 70% ethanol, air-dried and resuspended at a concentration of approximately 200 µg ml-1 in TE buffer (10 mM Tris/HCl pH 8·0, 1 mM Na2EDTA).

Construction of the recombinant cosmids containing allelic exchange substrates (AESs).
Cosmid vector pYUB572 is a derivative of pYUB328 (Balasubramanian et al., 1996 ), in which 2·2 kb fragment containing one of the {lambda}-cos sites was removed by cleavage with MunI and Csp45I. Cosmid pYUB854 is a derivative of pYUB572, in which the bla gene was removed by digestion with BspHI and replaced with a res–hyg–res gene cassette flanked by multiple cloning sites (MCSs). Plasmid pYUB870 is a derivative of pMV261 (Stover et al., 1991 ) in which the {gamma}{delta}-resolvase gene (tnpR) from transposon Tn1000 was cloned under the control of the mycobacterial hsp60 promoter. It also contains the sacB gene, which provides a negative selection for the loss of the helper plasmid when KanR cultures are plated on sucrose-containing media (Fig. 1). The phasmid phAE87 is a derivative of the conditionally replicating mycobacteriophage PH101(ts) (Bardarov et al., 1997 ). The phasmid phAE159, which permits a larger cloning capacity, is a derivative of phAE87 (J. I. Kriakov & W. R. Jacobs, Jr, unpublished results). The plasmid pYUB619 (Pavelka & Jacobs, 1999 ) was used as a source of the M. smegmatis {Delta}lysA4::res–hyg–res deletion allele. A 4130 bp BamHI–NotI fragment was cloned into BspHI-digested pYUB572 by blunt-end ligation to generate two cosmids, pYUB804 and pYUB805, which differ in the orientation of the {Delta}lysA4::res–hyg–res with respect to the BamHI site in the cosmid. Plasmid pYUB665 (Pavelka & Jacobs, 1999 ) was used as a source of the M. tuberculosis {Delta}lysA5::res–aph–res gene. A 4198 bp BclI–AscI fragment from pYUB665, containing the {Delta}lysA5::res–hyg–res gene flanked by ~1 kb of DNA sequence on each side, was cloned by blunt-end ligation into BspHI-digested cosmid pYUB572, to generate pYUB586. To generate the panC deletion mutation in M. smegmatis, the sequence database (http://www.tigr.org) was utilized to generate two primer pairs, which were employed to amplify the upstream LEFT and the downstream RIGHT arms flanking the panC gene. Flanking arms were cloned directionally into cosmid pYUB854 to generate the recombinant cosmid pYUB2500. PCR amplification of the entire operons using single primer pairs was the method used to generate the leuCD and nadBC AESs. PCR products were cloned into pBluescript KSII cloning vector. Deletions in the operons were generated with the appropriate restriction endonuclease and then marked with res–aph–res ({Delta}leuCD) or res–hyg–res ({Delta}nadBC). The AESs thus generated were then cloned by replacing the bla gene of the cosmid pYUB572. PCR amplification of the upstream and the downstream flanking DNA sequences using the M. tuberculosis genome sequence database (http://genolist.pasteur.fr/TubercuList) was used to generate the homologous AESs for the generation of deletion mutations in the panCD, Rv0867c and Rv3291c genes. After sequencing, the resulting DNA fragments were cloned directionally into cosmid vector pYUB854 to generate deletion mutants marked with the res–hyg–res gene cassette.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 1. Maps of the cosmids pYUB572 and pYUB854, and the helper plasmid pYUB870. (a) Cosmid vectors pYUB572 and pYUB854. Common features are the presence of origin of replication for E. coli (oriE), a single {lambda}-cos site, and PacI sites for efficient introduction of the cosmid into the mycobacteriophage shuttle phasmids. The res–hyg–res gene cassette in the cosmid pYUB854 is flanked by the {gamma}{delta}-res sites for subsequent removal of the selectable marker gene and by convenient MCSs for directional cloning of the flanking homologous recombination substrates. (b) The helper plasmid pYUB870 for transient expression of the {gamma}{delta}-tnpR gene. The plasmid has the aph gene conferring resistance to kanamycin, Phsp60sacRB and Phsp60tnpR.

 
Construction of specialized transducing mycobacteriophages.
Concatamers of phAE87 were prepared by self-ligation of purified phage DNA, which was then digested with PacI. To generate each specialized transducing phage, the PacI-digested recombinant cosmid carrying the appropriate AES was used to replace the pYUB328 cosmid in phAE87 or phAEB159 in an in vitro {lambda}-packaging reaction (GIGAPackII, Stratagene). After transducing E. coli HB101 and plating the transductants on selective media containing kanamycin or hygromycin, phasmid DNA was prepared from the pooled antibiotic-resistant transductants and electroporated into M. smegmatis mc2155. All transducing phages were plaque-purified and tested to confirm the temperature-sensitive phenotype. The correct structure of the AES in the specialized transducing phage was confirmed by PCR with locus-specific primers and by Southern blot analysis.

Transduction protocol.
M. smegmatis mc2155 was grown in LBT to an OD600 of 1·0 (~6x108 c.f.u. ml-1). M. bovis BCG strains and M. tuberculosis were grown in 7H9ADSTW to an OD600 of ~0·8–1·0. Ten millilitres of the culture was centrifuged and resuspended in 10 ml washing medium (7H9ADS without Tween 80) and incubated as a standing culture at 37 °C for 24 h. This incubation was included to remove traces of the Tween 80 detergent, which can inhibit phage infection. After this incubation period the cells were again centrifuged and resuspended in 1·0 ml 7H9ADS broth without Tween 80, pre-warmed at 37 °C, and mixed with specialized transducing phage at an m.o.i. of 10. The cell/phage mixture was incubated at the non-permissive temperature (37 °C) for 30 min (M. smegmatis) or 3 h (BCG and M. tuberculosis), after which the mixture was inoculated into 50 ml LBT (M. smegmatis) or complete 7H9ADSTW (BCG and M. tuberculosis) pre-warmed at 37 °C. Outgrowth of the cultures was performed for 30 min (M. smegmatis) or 24 h (BCG and M. tuberculosis) at 37 °C. Cells were then pelleted by centrifugation, resuspended in 1 ml PBS-TW (0·1% Tween 80 in phosphate-buffered saline) and plated on 7H9ADSTW complete medium containing kanamycin (25 µg ml-1) or hygromycin (150 µg ml-1 for M. smegmatis and 75 µg ml-1 for M. tuberculosis). Auxotrophic analysis was performed by plating the transductants on complete medium as well as on minimal medium. Transduction frequencies were calculated by dividing the number of HygR or KanR colonies obtained minus the number of spontaneous drug-resistant colonies from control cells receiving no phage by the total number of viable cells. The frequency of allelic exchange was calculated as the percentage of auxotrophs in the population of antibiotic-resistant transductants.

PCR analysis and Southern blotting.
PCR amplification was performed with AmpliTaq polymerase (Perkin-Elmer) under standard conditions. Primer concentrations and cycling conditions were adjusted depending on the size of the amplified product. All PCR reactions were performed in a Perkin-Elmer 9600 thermal cycler. Southern blotting was done by the alkali-denaturing procedure. DNA was transferred to HyBond-N+ membrane (Amersham) by the capillary method. Hybridization and detection were done with a chemiluminescent detection system (ECL, Amersham) as recommended by the manufacturer, In some cases probes were labelled with [{alpha}-32P]dCTP using Ready-To-Go DNA Labelling Beads (Amersham Pharmacia Biotech).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Specialized transduction methodology
In this study the primary goal was to develop a highly efficient system for the delivery of homologous DNA substrates in order to facilitate the generation of targeted gene disruptions in a wide array of mycobacterial species. Previous studies have shown that conditionally replicating shuttle phasmids are an efficient system for the generation of large libraries of independent transposon mutants in mycobacteria, as they provide the means to efficiently deliver transposon constructs to virtually every cell in the bacterial population (Bardarov et al., 1997 ; Cox et al., 1999 ). We reasoned that this recombinant DNA transfer system could be modified into a system for the generation of targeted deletion mutations in mycobacteria by allelic exchange, using in vitro-generated specialized transducing mycobacteriophages.

The specialized transducing mycobacteriophages consist of two basic components: a cosmid vector containing the AES and a conditionally replicating shuttle phasmid vector, which is a derivative of the broad-host-range TM4 phage (PH101ts). Specialized transduction involves five basic steps as outlined in Fig. 2: (1) construction of an AES in a PacI-containing E. coli cosmid; (2) cloning of the recombinant cosmid into the conditionally replicating shuttle phasmid; (3) transfection of M. smegmatis at the permissive temperature of 30 °C, to generate mycobacteriophage-packaged shuttle phasmids; (4) phage infection (transduction) of the mycobacteria at the non-permissive temperature of 37 °C; and (5) unmarking the deletion mutation by transient expression of tnpR.



View larger version (34K):
[in this window]
[in a new window]
 
Fig. 2. Schematic of one-step gene replacement in mycobacteria using in vitro-constructed specialized transducing phages.

 
(1) Construction of the AES in a PacI-containing E. coli cosmid. To make a targeted gene disruption of your favourite gene (yfg), it is necessary to generate an AES that consists of a selectable marker gene and the regions of DNA flanking yfg. Such a construct is simply made by using either the cosmid pYUB572 or the cosmid pYUB854, which have several features in common: (a) an origin of replication for E. coli (oriE); (b) a single {lambda}-cos site allowing in vitro packaging into phage {lambda} heads; (c) a mycobacterial selectable marker gene such as aph or hyg flanked by {gamma}{delta}-res sites for subsequent removal of the selectable marker gene; (d) MCSs flanking the selectable marker gene for directional cloning of the flanking homologous recombination substrates; and (e) a unique PacI site for the efficient introduction of the cosmid into the mycobacteriophage shuttle phasmid.

The AES can be generated in the cosmid pYUB572 by traditional cloning methods, where a DNA fragment containing yfg, with 700–1000 bp of upstream and downstream flanking DNA sequence, is amplified by PCR and cloned into any plasmid vector of choice. The desired deletion mutation in yfg is then generated by digestion with the chosen restriction enzyme and marked by insertion of the reporter gene. The marked homologous DNA substrate is then cloned into the cosmid vector pYUB572, by replacing the bla gene and plating on appropriate selective media. The cosmid cloning vector pYUB854, depicted in Fig. 1 and in Fig. 2 (Step 1), provides an alternative strategy for the engineering of AES. The LEFT and RIGHT flanking DNA arms (700–1000 bp long) are generated by PCR using two sets of primers. The region between the LEFT 3' (reverse) and the RIGHT 5' (forward) primers is the portion of the gene which is deleted. To help directional cloning of the DNA arms, the primer pairs may be engineered to contain restriction endonuclease sites corresponding to those found in the MCSs flanking the hygromycin-resistance gene cassette. Once this cloning is complete, the LEFT and RIGHT flanking DNA arms will be in their original genome orientation separated by the res–hyg–res cassette. The res–hyg–res cassette contains the DNA-binding sites (res sites) for the site-specific resolvase {gamma}{delta}-TnpR (Hatfull, 1988 ; Reed, 1981 ). When the resolvase gene (tnpR) is provided by a plasmid and transiently expressed in the mutant host strain, the site-specific recombination between the res sites results in unmarking of the deletion by precise excision of the hyg gene cassette.

(2) Introduction of the recombinant cosmid into the conditionally replicating TM4 shuttle phasmid. The second step involves cloning of the recombinant cosmid, containing the AES, into the conditionally replicating shuttle phasmid vector phAE87 or phAE159. Shuttle phasmid vectors are useful in that they allow all of the DNA manipulations to be performed in E. coli. However, there is an upper limit to the size of the allelic exchange construct that can be cloned into the mycobacteriophage genome. This limit is set by the size of DNA that can be packaged into {lambda} phage heads (about 50 kb) for the in vitro packaging and E. coli transduction step, and also by the size of the DNA that can be packaged into TM4 phage heads when transfected into M. smegmatis. The shuttle phasmid phAE87 contains a deletion of about 300 bp and is able to accommodate a maximum of 6·0 kb of exogenous DNA (cosmid vector size included). Any larger-sized phage molecules are very unstable in E. coli in their cosmid form or as mycobacteriophages in M. smegmatis. The shuttle phasmid phAE159 contains a deletion of 5·6 kb and thus is able to accommodate larger DNA inserts (J. I. Kriakov & W. R. Jacobs, Jr, unpublished results).

(3) Transfection of M. smegmatis at the permissive temperature of 30 °C to generate mycobacteriophage-packaged shuttle phasmids. This step involves the conversion of the recombinant shuttle cosmids purified from E. coli HygR transductants into mycobacteriophage-packaged DNA molecules. This is achieved by transfection of M. smegmatis cells with purified cosmids and plating for phage plaques at the permissive temperature (30 °C). Transfection frequencies of 103–104 p.f.u. per µg DNA are routinely obtained. High-titre transducing lysates (1010–1011 p.f.u. ml-1) can be readily obtained by propagation of the mycobacteriophage in M. smegmatis.

(4) Phage infection (transduction) of mycobacteria at the non-permissive temperature of 37 °C. In this step the AES, generated as part of the specialized transducing phage, is transferred with high efficiency into the recipient mycobacteria by transduction at the non-permissive temperature. Since upon infection of the recipient cells at non-permissive temperature (37 °C) phage replication is restricted, a large number of abortive transductants are accumulated. Allelic exchange occurs as a result of a double crossover between the homologous DNA arms flanking the disrupted gene. When plated on selective medium, antibiotic-resistant transductants are obtained with a mean frequency of 10-5–10-7 per total number of cells transduced. Of these antibiotic-resistant transductants, ~95% show the desired mutant phenotype when hyg is used as the reporter gene.

(5) Unmarking the deletion mutation by transient expression of {gamma}{delta}-tnpR. For unmarking of the deletion mutations generated by specialized transduction the helper plasmid pYUB870, expressing the tnpR gene under the control of hsp60 promoter, was constructed. The sacB gene included in the plasmid provides negative selection for the spontaneous loss of the helper plasmid when plated on media containing sucrose.

Specialized transduction in M. bovis BCG and M. tuberculosis
To test the specialized transduction system for the construction of isogenic mutant strains containing defined mutations in slow-growing mycobacteria a number of different auxotrophic mutations in several substrains of M. bovis BCG and M. tuberculosis were constructed (Table 1). For the construction of the lysine auxotrophs in M. bovis BCG three different substrains – Pasteur, Copenhagen and Moreau – were infected with the specialized transducing phage phAE128 containing the M. tuberculosis {Delta}lysA5 allele (Pavelka & Jacobs, 1999 ) marked with the res–aph–res gene cassette flanked by approximately 1 kb homologous DNA. Since the time required for optimal recombination in the slow-growing mycobacteria was unknown, cell/phage mixtures were incubated at 37 °C for different times before plating on complete medium containing kanamycin (Table 2). KanR colonies were obtained at frequencies in the range of 5x10-5 to 1x10-6 of the input cells for all three BCG strains at all three incubation times. Cell titres of the control cultures, not infected with the phage, determined at the start of infection and 24 h later, revealed no significant change in cell numbers (data not shown). Surprisingly, no difference in the number of KanR colonies was observed between the cells infected with the transducing phage and the control cultures not infected with the transducing phage. Therefore to test for allelic exchange events, all of the KanR transductants were screened for lysine auxotrophy. The highest percentage of auxotrophs was observed in BCG Pasteur and BCG Copenhagen after 24 h outgrowth time (34% and 29%, respectively), where the percentage of the auxotrophs nearly tripled between the 6 h outgrowth time and the 24 h outgrowth time. Although an unexpectedly low frequency of allelic exchange was observed for the BCG Moreau strain, this increase in the number of auxotrophs after 24 h outgrowth time (from 2% to 6%) was also observed. Analysis of the lysine auxotrophs by PCR, using locus-specific primers (10 auxotrophic clones for each strain) confirmed that the mutation was due to an allelic exchange of the wild-type allele with the {Delta}lysA5::res–aph–res allele (data not shown). The results from PCR analysis were confirmed by Southern blotting (Fig. 3). For all three BCG strains, the lysine auxotrophs had a single hybridization band, which was of the predicted size of the {Delta}lysA5::res–aph–res allele. In contrast, the parent strains or the KanR prototrophic strains had a single band that corresponded to the wild-type lysA+ allele. A high proportion of spontaneous KanR colonies was also observed in all three substrains of BCG when this gene cassette was used to genetically mark the deletion mutation in the leuCD operon (data not shown). In these experiments when KanR transductants were screened for an auxotrophic phenotype true allelic exchange events were observed at a frequency varying between 10% and 35% of the population of KanR transductants, depending on the number of transductants screened.


View this table:
[in this window]
[in a new window]
 
Table 1. Mutant alleles in M. bovis BCG, M. tuberculosis and M. smegmatis generated by specialized transduction

 

View this table:
[in this window]
[in a new window]
 
Table 2. Auxotroph analysis of BCG KanR colonies following phAE128 transduction at different outgrowth time intervals

 


View larger version (40K):
[in this window]
[in a new window]
 
Fig. 3. A composite of Southern blot analysis of M. bovis BCG and M. tuberculosis auxotroph mutants. (a) Lanes 1–4, genomic DNA from BCG Pasteur (lane 1) and {Delta}lysA::res–aph–res auxotroph mutants of: BCG Pasteur (lane 2), BCG Copenhagen (lane 3) and BCG Moreau (lane 4) digested with ApaLI and probed with the lysA PCR product from BCG Pasteur wild-type genomic DNA. The wild-type fragment is the expected 1·1 kb, while the genomic DNA from the auxotrophic strains has the expected 2·2 kb fragment. Lanes 5–8, genomic DNA from BCG Pasteur wild-type (lane 5) and {Delta}leuCD::res–aph–res auxotroph mutants of BCG Pasteur (lane 6), BCG Copenhagen (lane 7) and BCG Moreau (lane 8) digested with SacII and probed with the 3·3 kb PstI–HindIII fragment from pYUB804. The wild-type fragment is the expected 1·9 kb while the genomic DNA from the auxotrophic strains has the expected 3·4 kb fragment. Lanes 9–12, genomic DNA from BCG Pasteur wild-type (lane 9) and {Delta}nadBC::res–hyg–res auxotroph mutants of BCG Pasteur (lane 10), BCG Copenhagen (lane 11) and BCG Moreau (lane 12) digested with BsaAI and probed with the 1·7 kb KpnI–SacI fragment from pYUB862. (b) Lanes 1 and 2, genomic DNA from M. tuberculosis H37Rv wild-type (lane 1) and the {Delta}nadBC::res–hyg–res auxotroph mutants (lane 2) digested with BsaAI and probed with the 1·7 kb KpnI–SacI fragment from pYUB862. The wild-type fragment is the expected 1·6 kb while the genomic DNA from the auxotrophic strains has the expected 2·9 kb fragment. Lanes 3 and 4, genomic DNA from M. tuberculosis H37Rv wild-type (lane 3) and the {Delta}panCD::res–hyg–res auxotroph mutant (lane 4). Genomic DNAs were digested with BssHII, and probed with a 716 bp downstream region flanking the panCD operon. The wild-type fragment is the expected 1·48 kb, while the genomic DNA from the auxotrophic strain has the expected 2·65 kb fragment.

 
Due to the very high frequency of spontaneous kanamycin resistance observed in the slow-growing mycobacteria (10-6 of the input cells, which is nearly identical to the frequency of homologous recombination), in all further experiments the hygromycin-resistance gene cassette rather than the aph gene was used as the selectable marker. This is because this gene cassette was found to provide a much more stringent selection. Using this reporter gene, we constructed specialized transducing phages to engineer {Delta}panCD::res–hyg–res deletion mutations in M. bovis BCG Pasteur (Table 1). A transduction frequency of 10-6 of the input cells for the HygR marker was observed, whereas virtually no HygR colonies were observed when the control cells, not infected with the transducing phage, were plated on hygromycin selective media. Using this selectable marker gene we successfully engineered targeted deletions in five different chromosomal loci of M. tuberculosis substrains H37Rv, Erdman and CDC1551 (Table 1). In the course of constructing these mutant strains in numerous transduction experiments we consistently obtained high numbers of HygR transductants, ranging from experiment to experiment between 10-5 and 10-6 of the input cells (50 to more than 300 per experiment). As expected, in all cases the HygR transductants tested (typically 3–20 clones in each experiment) had the predicted auxotrophic phenotype. The correct structure of the deleted alleles was confirmed by PCR using locus-specific primers (data not shown) or Southern blotting (Fig. 3).

Specialized transduction in M. smegmatis
In M. smegmatis mc2155 specialized transduction was used to generate {Delta}lysA and {Delta}panC auxotrophic mutants marked with the hygromycin reporter gene. M. smegmatis cells were adsorbed with the transducing phages and phage/cell mixtures were plated on complete medium containing hygromycin. When HygR transductants were screened for auxotrophy, a high percentage (95% on average) of the transductants were lysine auxotrophs. Southern analysis confirmed that the lysine-auxotrophic phenotype was due to an allelic exchange of the wild-type lysA+ allele with the mutated {Delta}lysA4::res–hyg–res allele (data not shown). Comparable transduction frequencies of the hygromycin resistance marker were obtained when specialized transduction was employed to generate a {Delta}panC deletion mutation in M. smegmatis. In this experiment when several representative HygR transductants were tested on minimal medium they all showed an auxotrophic phenotype, requiring supplementation of the media with pantothenate for optimal growth. Analysis by PCR using locus-specific primers, and Southern analysis (data not shown), confirmed the predicted configuration of the {Delta}panC deletion allele.

Construction of unmarked deletion mutations in M. tuberculosis
One of the benefits of engineering unmarked mutations in M. tuberculosis is that multiple mutations can be generated in a single strain by using a three-step process of (1) mutagenesis (2) unmarking the deletion mutant and (3) introducing a second mutation using the original selectable marker. The specialized res–hyg–res gene cassette, described above, contains the specific DNA binding sites (res) for a site-specific {gamma}{delta} resolvase, the product of the tnpR gene of E. coli transposon Tn1000(AmpR) (Berg et al., 1992 ; Hatfull, 1988 ; Reed, 1981 ). Transient expression of the resolvase gene promotes site-specific recombination between the res sites. The final result is a precise excision of the hygromycin-resistance gene, leaving the deletion site unmarked. We constructed a helper plasmid expressing the {gamma}{delta} resolvase under transcriptional control of the hsp60 promoter. We also included the sacB gene in this plasmid to provide a negative selection for spontaneous plasmid loss. To generate unmarked deletion mutations we used the helper plasmid pYUB870 to unmark the M. tuberculosis {Delta}Rv0867c::res–hyg–res mutant strain. After transformation by electroporation with pYUB870 and plating onto medium containing kanamycin, a total of 20–30 KanR colonies were obtained and screened by a pick-and-patch method (streaking on 7H10 agar alone and on 7H10 agar with 50 µg hygromycin ml-1) for hygromycin sensitivity. A hygromycin-sensitive clone was grown in liquid medium in the absence of antibiotic selection, and genomic DNA was prepared and tested by Southern blot analysis for loss of the hygromycin-resistance gene cassette. The analysis of one such clone is shown in Fig. 4. Southern blotting with different gene probes as well as with a probe specific for the hyg gene all showed that the res–hyg–res cassette had been excised, giving the correct fragment sizes. This clone was then plated onto sucrose-containing medium and single KanS HygS colonies obtained.



View larger version (54K):
[in this window]
[in a new window]
 
Fig. 4. Deletion of Rv0867c from the chromosome of M. tuberculosis by specialized transduction, and excision of the res–hyg–res gene cassette by tnpR to generate an unmarked deletion strain. (a) Schematic diagram of the Rv0867c region of the chromosome of M. tuberculosis (Erdman). Also shown is the position of the three hybridization probes: the Rv0867c ORF; an internal portion of the hyg gene; and the upstream flanking region of the Rv0867c gene. (b) Southern blot hybridizations of SphI-digested chromosomal DNA from M. tuberculosis wild-type (lanes 1), the {Delta}Rv0867c:: res–hyg–res strain (lanes 2) and the {Delta}Rv0867c-resolved strain (lanes 3). The upstream flank probe yields hybridizing bands of the predicated sizes in all three strains, while the Rv0867c gene probe hybridizes only to the wild-type strain. The hyg gene probe hybridizes only to the {Delta}Rv0867c::res–hyg–res strain. The SphI restriction sites are indicated, as are the sizes of fragments hybridizing to the three probes. Deletion of Rv0867c with its replacement by the res–hyg–res cassette results in a smaller hybridizing band, since the cassette introduces an additional SphI site. The loss of both the res–hyg–res cassette and the additional SphI site mediated by the resolvase results in an only slightly larger hybridizing band.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Several groups have demonstrated the use of suicide plasmid vectors to deliver AESs for insertional mutagenesis in both fast- and slow-growing mycobacteria (Berthet et al., 1998 ; Frischkorn et al., 1998 ; Knipfer et al., 1997 ; Parish & Stoker, 2000 ; Pavelka & Jacobs, 1996 ; Pelicic et al., 1997 ; Pelicic et al., 1996a , b ). One of the limitations in using transformation as a means to deliver the AES for generation of targeted gene disruptions is the high frequency of illegitimate recombination in these bacterial species and the comparatively low transformation efficiency, especially in the slow-growing pathogenic mycobacteria.

Based on our previous experience in using conditionally replicating shuttle phasmid vectors (Hatfull, 1996 ; Hatfull et al., 1994 ; Jacobs et al., 1991 ) for efficient delivery of transposon constructs and generation of transposon libraries in both M. bovis BCG and M. tuberculosis (Bardarov et al., 1997 ; Cox et al., 1999 ) we sought the opportunity to extend their usage by modifying them into in vitro-generated specialized transducing phages. Specialized transduction as a method for delivery of recombinant DNA into mycobacteria has several useful features. Firstly, the specialized transduction phage system described in this study can be readily generalized to any set of genes by virtue of the facile cloning properties of the shuttle phasmids and the properties of the conditionally replicating mycobacteriophage vector. Once a specific mutated allele is engineered into the phage, that mutated allele can be introduced into any fast- or slow-growing mycobacterial species. The conditionally replicating shuttle phasmids used in this system as AES delivery vectors can be propagated in variety of ways: they can be electroporated into E. coli, where they behave as multicopy plasmids; they can be packaged in vitro into {lambda} phage heads and transduced in E. coli; the recombinant phasmids can be transfected into M. smegmatis, where at the permissive temperature (30 °C) they behave as lytic phages and high-titre transducing phage lysates could be produced; and they can be transduced with high efficiency into the mycobacterial hosts at non-permissive temperature (37 °C), where they behave as non-replicating circular replicons giving rise to progeny of abortive transductants. Secondly, the specialized transducing mycobacteriophages used in this study provide a natural counterselection mechanism for the integration of the whole phage, or parts of it, into the host chromosome by single crossover or illegitimate recombination. If such an integration event occurs, expression of the phage genes, as a part of the chromosome, will be deleterious to the host cell. Thirdly, the broad host range of the mycobacteriophage TM4 allows specialized transduction to be used for efficient delivery of AESs for the generation of targeted gene disruptions into many genetically and clinically important mycobacterial species. Once a specific mutation is generated in a defined genetic background the mutant strain becomes a useful reagent for engineering of series of isogenic mutant strains that differ by that single mutated gene. Moreover, it should be rather straightforward to extend this genetic strategy to many other bacterial species for which specific phages exist.

We have demonstrated the utility and reproducibility of specialized transduction as a highly efficient method for the delivery of homologous DNA substrates for allelic exchanges in both fast- and slow-growing mycobacteria. Using the M. tuberculosis sequence database (Cole et al., 1998 ), seven deletion mutant alleles were engineered. When the mutant alleles were introduced by transduction into the host mycobacteria the observed recombination frequencies were in the 10-6 range; thus, these mutants could not be constructed by electroporation with a suicide plasmid, since the best electroporation efficiencies we were able to obtain with M. tuberculosis are in the 10-5 range. These results amply demonstrate the potential of using in vitro-generated specialized transducing mycobacteriophages for efficient gene replacement in mycobacteria. Ultimately, such a system should prove to be very useful for the construction and development of M. tuberculosis-based vaccine strains with numerous defined non-revertible mutations.

We also constructed a plasmid, pYUB870, for the delivery and transient expression of the {gamma}{delta} site-specific resolvase, which acts on the directly repeated res sites flanking the antibiotic-resistance marker gene. Removal of the antibiotic marker was readily achieved by incubation of a few KanR clones obtained after transformation with the expression plasmid and plating on sucrose-containing media. Most of the KanS HygS clones had lost both the hygromycin-resistance gene and the plasmid.

In summary we have developed a novel genetic method for a single-step, highly efficient delivery of AESs for the generation of targeted gene disruptions in both fast- and slow-growing mycobacteria. Seven different mutations have been engineered in three substrains of M. bovis BCG and three strains of M. tuberculosis. The present study amply demonstrates the power of specialized transduction as a natural mycobacterial genetic transfer system of AESs and generation of targeted gene disruptions that should be applicable to a wide variety of mycobacterial species.


   ACKNOWLEDGEMENTS
 
This work was supported by grants from the National Institutes of Health AI26170, AI46690, GMG2410, AI49375 and AI28927.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Aldovini, A., Husson, R. N. & Young, R. A. (1993). The uraA locus and homologous recombination in Mycobacterium bovis BCG. J Bacteriol 175, 7282-7289.[Abstract]

Azad, A. K., Sirakova, T. D., Rogers, L. M. & Kolattukudy, P. E. (1996). Targeted replacement of the mycocerosic acid synthase gene in Mycobacterium bovis BCG produces a mutant that lacks mycosides. Proc Natl Acad Sci USA 93, 4787-4792.[Abstract/Free Full Text]

Balasubramanian, V., Pavelka, M. S.Jr, Bardarov, S. S., Martin, J., Weisbrod, T. R., McAdam, R. A., Bloom, B. R. & Jacobs, W. R.Jr (1996). Allelic exchange in Mycobacterium tuberculosis with long linear recombination substrates. J Bacteriol 178, 273-279.[Abstract]

Bardarov, S., Kriakov, J., Carriere, C., Yu, S., Vaamonde, C., McAdam, R. A., Bloom, B. R., Hatfull, G. F. & Jacobs, W. R.Jr (1997). Conditionally replicating mycobacteriophages: a system for transposon delivery to Mycobacterium tuberculosis. Proc Natl Acad Sci USA 94, 10961-10966.[Abstract/Free Full Text]

Berg, C. M., Vartak, N. B., Wang, G., Xu, X., Liu, L., MacNeil, D. J., Gewain, K. M., Wiater, L. A. & Berg, D. E. (1992). The m {gamma}{delta}-1 element, a small {gamma}{delta} (Tn1000) derivative useful for plasmid mutagenesis, allele replacement and DNA sequencing. Gene 113, 9-16.[Medline]

Berthet, F. X., Lagranderie, M., Gounon, P. & 9 other authors (1998). Attenuation of virulence by disruption of the Mycobacterium tuberculosis erp gene. Science 282, 759–762.[Abstract/Free Full Text]

Boshoff, H. I. M. & Mizrahi, V. (2000). Expression of Mycobacterium smegmatis pyrazinamidase in Mycobacterium tuberculosis confers hypersensitivity to pyrazinamide and related amides. J Bacteriol 182, 5479-5485.[Abstract/Free Full Text]

Braunstein, M., Brown, A. M., Kurtz, S. & Jacobs, W. R.Jr (2001). Two nonredundant SecA homologues function in mycobacteria. J Bacteriol 183, 6979-6990.[Abstract/Free Full Text]

Cole, S. T., Brosch, R., Parkhill, J. & 39 other authors (1998). Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544. [An erratum appears in Nature 396, 190.].

Cox, J. S., Chen, B., McNeil, M. & Jacobs, W. R.Jr (1999). Complex lipid determines tissue-specific replication of Mycobacterium tuberculosis in mice. Nature 402, 79-83.[Medline]

Fitzmaurice, A. M. & Kolattukudy, P. E. (1998). An acyl-CoA synthase (acoas) gene adjacent to the mycocerosic acid synthase (mas) locus is necessary for mycocerosyl lipid synthesis in Mycobacterium tuberculosis var. bovis BCG. J Biol Chem 273, 8033-8039.[Abstract/Free Full Text]

Fleischmann, R. D., Alland, D., Eisen, J. A. & 23 other authors (2002). Whole genome comparison of Mycobacterium tuberculosis clinical and laboratory strains. J Bacteriol (in press).

Frischkorn, K., Sander, P., Scholz, M., Teschner, K., Prammananan, T. & Bottger, E. C. (1998). Investigation of mycobacterial recA function: protein introns in the RecA of pathogenic mycobacteria do not affect competency for homologous recombination. Mol Microbiol 29, 1203-1214.[Medline]

Glickman, M. S., Cox, J. S. & Jacobs, W. R.Jr (2000). A novel mycolic acid cyclopropane synthetase is required for cording, persistence, and virulence of Mycobacterium tuberculosis. Mol Cell 5, 717-727.[Medline]

Hatfull, G. F. (1996). The molecular genetics of Mycobacterium tuberculosis. Curr Top Microbiol Immunol 215, 29-47.[Medline]

Hatfull, G. F., Salvo, J. J., Falvey, E. E., Rimphanitchayakit, V. & Grindley, N. D. (1988). Site-Specific Recombination by the {gamma}{delta} Resolvase. Cambridge: Cambridge University Press.

Hatfull, G. F., Barsom, L., Chang, L., Donnelly-Wu, M., Lee, M. H., Levin, M., Nesbit, C. & Sarkis, G. J. (1994). Bacteriophages as tools for vaccine development. Dev Biol Stand 82, 43-47.[Medline]

Hinds, J., Mahenthiralingam, E., Kempsell, K. E., Duncan, K., Stokes, R. W., Parish, T. & Stoker, N. G. (1999). Enhanced gene replacement in mycobacteria. Microbiology 145, 519-527.[Abstract]

Jacobs, W. R.Jr, Kalpana, G. V., Cirillo, J. D., Pascopella, L., Snapper, S. B., Udani, R. A., Jones, W., Barletta, R. G. & Bloom, B. R. (1991). Genetic systems for mycobacteria. Methods Enzymol 204, 537-555.[Medline]

Kalpana, G. V., Bloom, B. R. & Jacobs, W. R.Jr (1991). Insertional mutagenesis and illegitimate recombination in mycobacteria. Proc Natl Acad Sci USA 88, 5433-5437.[Abstract]

Knipfer, N., Seth, A. & Shrader, T. E. (1997). Unmarked gene integration into the chromosome of Mycobacterium smegmatis via precise replacement of the pyrF gene. Plasmid 37, 129-140.[Medline]

Lenox, E. S. (1955). Transduction of linked genetic characters of the host by bacteriophage P1. Virology 1, 190-206.

Masters, M. (1996). Generalized transduction. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 2421–2441. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

McFadden, J. (1996). Recombination in mycobacteria. Mol Microbiol 21, 205-211.[Medline]

Parish, T. & Stoker, N. G. (2000). Use of a flexible cassette method to generate a double unmarked Mycobacterium tuberculosis tlyA plcABC mutant by gene replacement. Microbiology 146, 1969-1975.[Abstract/Free Full Text]

Parish, T., Gordhan, B. G., McAdam, R. A., Duncan, K., Mizrahi, V. & Stoker, N. G. (1999). Production of mutants in amino acid biosynthesis genes of Mycobacterium tuberculosis by homologous recombination. Microbiology 145, 3497-3503.[Abstract/Free Full Text]

Pavelka, M. S.Jr & Jacobs, W. R.Jr (1996). Biosynthesis of diaminopimelate, the precursor of lysine and a component of peptidoglycan, is an essential function of Mycobacterium smegmatis. J Bacteriol 178, 6496-6507.[Abstract]

Pavelka, M. S.Jr & Jacobs, W. R.Jr (1999). Comparison of the construction of unmarked deletion mutations in Mycobacterium smegmatis, Mycobacterium bovis bacillus Calmette-Guérin, and Mycobacterium tuberculosis H37Rv by allelic exchange. J Bacteriol 181, 4780-4789.[Abstract/Free Full Text]

Pelicic, V., Reyrat, J. M. & Gicquel, B. (1996a). Generation of unmarked directed mutations in mycobacteria, using sucrose counter-selectable suicide vectors. Mol Microbiol 20, 919-925.[Medline]

Pelicic, V., Reyrat, J. M. & Gicquel, B. (1996b). Positive selection of allelic exchange mutants in Mycobacterium bovis BCG. FEMS Microbiol Lett 144, 161-166.[Medline]

Pelicic, V., Jackson, M., Reyrat, J. M., Jacobs, W. R.Jr, Gicquel, B. & Guilhot, C. (1997). Efficient allelic exchange and transposon mutagenesis in Mycobacterium tuberculosis. Proc Natl Acad Sci USA 94, 10955-10960.[Abstract/Free Full Text]

Raman, S., Song, T., Puyang, X., Bardarov, S., Jacobs, W. R.Jr & Husson, R. N. (2001). The alternative sigma factor SigH regulates major components of oxidative and heat stress responses in Mycobacterium tuberculosis. J Bacteriol 183, 6119-6125.[Abstract/Free Full Text]

Reed, R. R. (1981). Resolution of cointegrates between transposons {gamma}{delta} and Tn3 defines the recombination site. Proc Natl Acad Sci USA 78, 3428-3432.[Abstract]

Reyrat, J. M., Berthet, F. X. & Gicquel, B. (1995). The urease locus of Mycobacterium tuberculosis and its utilization for the demonstration of allelic exchange in Mycobacterium bovis bacillus Calmette-Guérin. Proc Natl Acad Sci USA 92, 8768-8772.[Abstract]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sander, P., Meier, A. & Bottger, E. C. (1995). rpsL+: a dominant selectable marker for gene replacement in mycobacteria. Mol Microbiol 16, 991-1000.[Medline]

Sirakova, T. D., Thirumala, A. K., Dubey, V. S., Sprecher, H. & Kolattukudy, P. E. (2001). The Mycobacterium tuberculosis pks2 gene encodes the synthase for the hepta- and octamethyl-branched fatty acids required for sulfolipid synthesis. J Biol Chem 276, 16833-16839.[Abstract/Free Full Text]

Steyn, A. J. C., Collins, D. M., Hondalus, M. K., Jacobs, W. R.Jr, Kawakami, R. P. & Bloom, B. R. (2002). Mycobacterium tuberculosis WhiB3 interacts with RpoV to affect host survival but is dispensable for in vivo growth. Proc Natl Acad Sci USA 99, 3147-3152.[Abstract/Free Full Text]

Stover, C. K., de la Cruz, V. F., Fuerst, T. R. & 11 other authors (1991). New use of BCG for recombinant vaccines. Nature 351, 456–460.[Medline]

Sundar Raj, C. V. & Ramakrishnan, T. (1970). Transduction in Mycobacterium smegmatis. Nature 228, 280-281.

Zinder, N. & Lederberg, J. (1952). Genetic exchange in Salmonella. J Bacteriol 64, 679-699.

Received 27 June 2002; revised 11 July 2002; accepted 15 July 2002.