To access the private area of this site, please log in.

Username: 

Password:  

 

Chapter 141

Transposons Currently in Use in Genetic Analysis of Salmonella Species

ELLIOT ALTMAN, JOHN R. ROTH, ANDREW HESSEL, and KENNETH E. SANDERSON

 

INTRODUCTION

This chapter discusses the uses of transposons in genetic analysis, with emphasis on work done in Salmonella typhimurium (official designation, Salmonella enterica serovar Typhimurium) and on the methods available for use in Salmonella spp. Many of these methods, however, can be applied to other genera as well.

 

TYPES OF TRANSPOSONS

The values of transposable elements for genetic analysis have been amply described elsewhere (2, 10, 15, 16). As these elements have become more widely used, the array of variants developed for specific purposes has increased. Below, we list and describe some of the attributes of the transposons that we have found most useful. We then describe current methods for making insertion mutations with these elements.

 

Derivatives of Tn10

Although the Tn10 element has been widely used in its original, unmodified form, it is being increasingly superseded by a variety of elements that have the advantages of smaller size and the stability that is achieved by removal of transposase activity. The original Tn10 element (9.3 kb) includes flanking copies of IS10 that are capable of independent transposition. Although transposition of the entire transposon is not a problem for genetic analysis, the insertion sequence (IS) elements transpose extremely frequently and cause a significant elevation of mutation frequencies (21, 24). A second disadvantage of the original Tn10 is seen when deletions are sought using the Bochner selection (3). The large terminal repeats are prone to "imprecise excision," which leads to loss of tetracycline resistance but leaves small segments of the element at the insertion sites. These "imprecise excisants" must be screened through when one is searching for deletions that extend outside of the element into adjacent chromosomal sequences. The smaller element Tn10d-Tet (see below) is not subject to this imprecise excision, and thus, a much higher fraction of clones surviving the Bochner selection are true deletions which extend into the adjacent chromosome. Below, we describe a series of Tn10 derivatives that are in current use.

Tn10d-Tet.

Tn10d-Tet is 2.9 kb in size. It differs from the original element by two deletion mutations, each of which removes the bulk of one of the IS10 elements and extends into the central, single-copy region (9). The modified element does not make transposase but has 70 bp of inverse-order IS10 sequence repeated at each end, thus providing sites at which transposase can act (9, 25). Although transposase must be provided to achieve transposition, insertion mutations are completely stable in the absence of an external source of transposase.

Recently, it was observed that the divergent promoters of the tetracycline resistance determinants direct regulated transcripts from both ends of the inserted element into the adjacent chromosomal region (23). This fact has not yet been routinely exploited but shows promise of being widely applicable.

Tn10d-Kan, Tn10d-Cam, Tn10d-Gen.

The Tn10d-Kan, Tn10d-Cam, and Tn10d-Gen elements are similar to Tn10d-Tet and contain the terminal 70-bp ends of the original Tn10 element. Each has a different drug resistance determinant placed between these flanking repeats. While Tn10d-Kan (1.8 kb, kanamycin resistance) and Tn10d-Cam (1.5 kb, chloramphenicol resistance) have been used for some time (8, 25), the last element, Tn10d-Gen (2.0 kb, gentamicin resistance), was constructed only recently by Higgins et al. (P. Higgins, X. Yang, Q. Fu, and J. Roth, submitted for publication) and adds a new resistance to the list of available drug markers.

Tn10d-Cam/Gen.

Tn10d-Cam/Gen, which also was constructed by Higgins (Higgins et al., submitted for publication), is identical to the Tn10d-Cam described above except that a gentamicin resistance determinant has been inserted in the middle of the coding sequence of the chloramphenicol resistance gene. This element, which is 3.4 kb in size, confers only gentamicin resistance. Its value is that it can be used as a donor to convert the resistance of any existing Tn10d-Cam insertion mutation from Cam^r to Gen^r.

Generating Mutations with Tn10 Derivatives.

We found that the new low-specificity transposase mutant of Kleckner and coworkers, which is carried on pNK2881, provides an excellent way of isolating new insertion mutations with less tendency to populate the "hot spots" favored by the original Tn10 (15). We make mutations by a transduction cross (using P22) in which the donor carries the defective element on an F' pro lac plasmid derived from Escherichia coli. The recipient carries plasmid pNK2881, which expresses the low-specificity transposase. Selection is for the relevant drug resistance. Since the donor element is inserted in material that has no homolog in the recipient, there is no inheritance by recombination. Each transductant requires an act of transposition of the defective element from the transduced fragment into the recipient chromosome. We generally pool these resistant transductants, grow P22 on the pool, and use this pool to transduce individual insertion mutations into a clean genetic background before screening for mutations of interest. Donor strains useful for these procedures are TT10423 proAB47/F' pro ⁺ lac ⁺ zzf-1831::Tn10d-Tet, TT17394 proAB47/F' pro ⁺ lac ⁺ zzf-3734::Tn10d-Kan, TT10605 proAB47/F' pro ⁺ lac ⁺ zzf-1837::Tn10d-Cam, NH2048 proA36 strA1/F' pro ⁺ lac ⁺ Tn10d-Gen, and NH2030 metA22 metE551 trpD2 ilv-452 galE leu pro strA120 hsdLT6 hsdSA29 hsdB/pPH620 (this strain contains Tn10d-Cam/Gen on an Amp^r plasmid instead of on the F' pro⁺ lac ⁺ plasmid). A useful recipient strain is TT17437 (LT2)/pNK2881 (Amp^r).

In order to avoid lysogenization, it is important to use a high-transducing, low-integrating mutant of P22 (such as P22HT105/int-1) when lysates are being formed.

Use of Tn10 Elements for PCR Priming Sites.

The elements described above are extremely useful in conjunction with PCR amplification. This technique can be used to clone the region between inserted elements and to physically map the elements in a known sequence. Listed below are oligonucleotide primers we have found to work well with Tn10 derivatives (P. Chen, M. Ailion, T. Bobik, G. Stormo, and J. Roth, submitted for publication).

  Universal (outward directed). The universal primer sequence is just within the inverse (70-bp) repeats found at both ends of all of the Tn10 derivatives described above. It is directed outward, so PCR reactions primed by this oligonucleotide proceed out of the element into adjacent chromosomal sequences. The sequence is 5' GACAAGATGTGTATCCACCTTAAC 3'.

  Unique (left end). The unique (left-end) primer is unique to the left end of the Tn10d-Tet element. It primes replication that extends out of the end of the element where the tetA gene (tetracycline resistance) is located. Replication proceeds in the same direction as transcription of tetA. The sequence is 5' ACCTTTGGTCACCAACGCTTTTCC 3'.

  Unique (right end). The unique (right-end) primer is unique to the right end of the Tn10d-Tet element. It primes replication that extends out of the end of the element where the tetR gene (repressor of tetA) is located. Replication proceeds in the same direction as transcription of tetR. The sequence is 5' TCCATTGCTGTTGACAAAGGGAAT 3'.

 

Derivatives of Phage Mu.

Most of the derivatives of Mu elements were constructed by Malcolm Casadaban and coworkers. The elements were designed to be delivered by Mu virions using wild-type Mu helper phages. Most include a lac operon without a promoter and are designed to form fusions of the included lac operon to a promoter located near the target site in the bacterial chromosome. As described below, some elements lack both a promoter and a translation start site, so they express lac only if they fuse the lacZ gene to an expressed chromosomal open reading frame.

MudI (37.2 kb, Operon Fusions) and MudII (35.6 kb, Gene Fusions).

MudI and MudII are Casadaban’s original constructions; they form operon fusions and gene fusions, respectively (4, 5). The elements are transposition proficient and are induced by high temperature to form complete phages.

MudA (37.2 kb, Operon Fusions) and MudB (35.6 kb, Gene Fusions).

MudA and MudB are derivatives of the original Amp^r constructions which have been made transposition defective by the addition of two amber mutations (12). Both are stable in strains lacking an amber suppressor but are made transposition proficient by the presence of a suppressor. Each element is 36 to 37 kb and thus can be transduced only rarely by a P22 virion (44 kb). At high transducing phage multiplicities, these elements can be transduced by two cooperating P22 phages (13).

MudJ (11.3 kb, Operon Fusions) and MudK (9.7 kb, Gene Fusions).

MudJ and MudK are the original Kan^r mini-Mud constructions of Casadaban and coworkers (6). They contain deletions of the transposase functions and thus are incapable of transposing without an outside source of transposase.

MudF (14.0 kb, Lac⁺ without Fusions).

MudF confers Kan^r and contains a complete lac operon, including the lacI gene (22). It gives a Lac⁺ phenotype no matter where it inserts. This element provides a useful way of making any Salmonella strain Lac⁺ with a wild-type, inducible lac operon.

MudCam (2.9 kb, a Simple Drug Resistance Element).

MudCam includes no lac sequences and contains only a simple drug resistance determinant placed between the ends of phage Mu (7). It can be used as an independent transposable element for making mutations and has special value as a vector for placing a simple drug resistance insertion at the site of a known lac fusion insertion. Because this element shares sufficient homology with the ends of MudA, MudB, MudF, MudJ, and MudK, it can be used as a donor in transduction crosses in which the selected Cam^r marker replaces the lac and Kan^r (or Amp^r) sequences of the other elements. It should be possible, in principle, to use this element as an intermediate in converting a gene fusion insertion to an operon fusion insertion (to our knowledge, this has not yet been done).

MudSac (5.4 kb, a Counterselectable Kan^r Mini-Mud).

MudSac was constructed by M. Lawes and S. Maloy (submitted for publication) for use with the Mud-P22 chromosome mapping method of Benson and Goldman (1). The element is transposition defective and includes the Bacillus subtilis sacB (secretory levansucrase) gene, which makes cells sensitive to sucrose. Transpositions of this element are generated via standard techniques (see below) by selecting for insertion mutations that inherit Kan^r. Because MudSac insertions can be counterselected on sucrose, their map positions can be easily determined using Mud-P22 mapping techniques. The MudSac element should also prove to be useful for generating deletions near the insertion site since one can select for spontaneous sucrose-resistant derivatives of the insertion mutation.

Isolating Insertions of Transposition-Defective Mud Elements.

We have found that "transitory cis complementation" is a convenient way of providing transposase to the defective elements described above (14). This procedure involves the use of strains that contain the defective element inserted in the bacterial chromosome adjacent to a tranposition-proficient MudI prophage. The insertions are placed so that the end of the MudI prophage including the transposition functions is closest to the defective element. When P22 phage is grown on such a strain, the transducing virions that happen to package the defective element frequently include the transposition functions of the adjacent proficient prophage. Because of the sizes of the Mu phages, it is not possible for a single P22 particle (44 kb) to include both the defective prophage and the complete helper genome (37 kb).

The lysate is used to transduce a recipient strain selecting for inheritance of the drug resistance encoded by the defective element. The transduced fragment expresses transposition functions (from the proficient element) which allow the defective element to tranpose to the recipient chromosome. This method is particularly effective because the Mu transposase functions tend to act preferentially in cis. The appropriate strains for this procedure are TT10288 hisA9944::MudI hisD9953::MudJ, TT10381 hisA9944::MudI hsiD1284::MudK, TT16528 hisA9944::MudI hisD9953::MudCam, TT18310 hisA9944::MudI hisC10081:: MudF, and MST3281 hisA9944::MudI his::MudSacI.

PCR Primers for Mu Ends.

The following primers were designed for stimulating PCR reactions from the ends of all Mu-derived elements described above (Chen et al., submitted): for the left end, 5' ATCCCGAATAATCCAATGTCC 3'; for the right end, 5' GAAACGCTTTCGCGTTTTTCGTGC 3'.

 

TnPhoA

TnPhoA, a derivative of transposon Tn5, was constructed by C. Manoil and J. Beckwith to detect insertions into genes which code for exported proteins (18). The element is transposition proficient and includes at one end a gene for alkaline phosphatase which lacks a promoter, translation start, and signal sequence. Alkaline phosphatase is normally active only if it can be exported to the periplasm. Thus, in order for the protein encoded by the element to be expressed, the element must insert, in the proper orientation and reading frame, into a gene that has a signal sequence. Such insertion mutants are identified on medium including the chromogenic phosphatase substrate 5-bromo-4-chloro-3-indolyl phosphate (X-P). Use of this element in Salmonella spp. requires elimination of the normal acid phosphatase, which can also hydrolyze X-P. Finding insertions of TnPhoA into the chromosome requires screening a large number of insertions, since only rare insertions meet all of the criteria for activity listed above. We have done this screening by using the P22 "locked-in" phages of Youdarian et al. (27). This locked-in prophage is placed near a TnPhoA insertion in the his operon. When induced, this prophage packages the adjacent material preferentially, and the lysate includes a high titer of particles that include the TnPhoA element; the donor strain carries a plasmid which overproduces the P22 tail protein, which is normally produced at a low level during induced phage growth. This lysate is used to transduce a recipient lacking acid phosphatase (a phoN mutant). Selection for His⁺ is maintained to avoid the high level of recombinants that inherit the donor his::TnPhoA element by recombination. Most transductant colonies are white on X-P medium, but rare blue insertions in which the element has inserted into a gene with a signal sequence can be identified (28). The necessary donor strain for this procedure is TT15088 hisD10088::TnPhoA hisHA9556::MudP/pPB13. The recipient strain is any strain with a phoN mutation; the insertion phoN51::Tn10d-Tet (e.g., TT13216) can be transduced into the desired strains.

 

MudP and MudQ (Transposable Locked-In P22 Prophages)

MudP and MudQ permit one to place a "P22 prophage" at virtually any site in the chromosome. Since the prophage lacks P22 attachment sites, induction stimulates the phage to replicate and package DNA processively out of the prophage and into the adjacent bacterial chromosome in a unidirectional manner. This results in a lysate that is very highly enriched for transducing particles that include DNA from a very small region of the chromosome (27). These lysates provide a rich source of DNA which is useful for both cloning and high-frequency transduction of markers in that region. These phages have been used to a create a new method for mapping mutations in the Salmonella chromosome (1).

MudP and MudQ elements are essentially P22 prophages lacking the attachment site and flanked by the ends of phage Mu; a chloramphenicol resistance determinant is included between the Mu ends. The two phages differ in the orientation of the P22 sequences between the Mu ends. The element transposes as a unit when Mu transposition functions are provided; thus, the entire element can be inserted at any site into which Mu transposition is possible. Alternatively, the MudP and MudQ elements can be transduced into a recipient strain carrying a known Mud element (either the original MudI element or any of the standard mini-Mu elements). The MudP or MudQ element then recombines with the recipient element by using the flanking Mu sequences as regions of homology. This results in the replacement of the recipient element with MudP or MudQ.

The chromosomal mapping method based on these phages employs a set of strains that contain mapped MudP or MudQ prophages. Lysates of these strains are used in spot transduction tests to repair the mutation that is to be mapped. If the unknown mutation is not counterselectable, a Tn10 insertion in or near the mutant gene can be used in conjunction with the Bochner method of counterselecting tetracycline-resistant strains to permit selective repair by the donor locked-in lysates (3). Since the donor lysates are enriched for only several minutes of chromosome, high-frequency (about 200 kb according to some estimates [17]) repair of the recipient mutation by one or more of the donor mapping strains allows one to determine the chromosomal position of the unknown mutation to within a few map minutes.

Donor strains useful in transductional conversion of a recipient Mud element to a MudP or a MudQ prophage are the following: TT12915 (=PY13518) leuA414(Am) r^–m⁺ (Fels2^–)/ F'114ts lac ⁺ zzf-20::Tn10 zzf-3551::MudP and TT12916 (=PY13757) leuA414(Am) r^–m⁺ (Fels2–)/F'114ts lac ⁺ zzf-20::Tn10 zzf-3553::MudQ.

A set of 30 strains in current use for mapping (known affectionately as the "First String Mud-P22 Collection" [N. Benson, personal communication]) is listed in Table 1.

 

TRANSPOSON-INDUCED MUTATIONS INS. TYPHIMURIUM

Transposon-induced mutations have been isolated in S. typhimurium by many investigators as part of a wide range of studies. Most have been isolated by the techniques described above. We present a list of strains of S. typhimurium which carry insertions of transposons of several different types (Table 2). The strains of S. typhimurium which carry these transposon insertions along with linkage data as to the exact chromosomal locations of the insertions were generously provided to the Salmonella Genetic Stock Center (SGSC) by many different investigators. Table 2 is an update of a previously published table (19). Many of these strains were also listed by Berg and Berg (2).

These strains are listed in the order of the map location of the transposon insertion, with thr used as the starting point. These locations have been determined by several techniques. Most have been located next to specific genes by P22 transduction; the distance (in kilobases) from the transposon insertion to the known gene or the distance between two different transposons can be calculated from transduction data with the formula of Wu (26) as modified to take account of the size of the transposon (19). Some Tn10 insertions have been located on the chromosome by using the rapid mapping method developed by Benson and Goldman with MudP22 (1). Tn10 has sites for the rarely cutting endonucleases XbaI and BlnI; locations of Tn10 insertions on the chromosomal genomic cleavage maps for these enzymes were determined by using pulsed-field gel electrophoresis (17, 20).

These strains are available to any researcher. Table 2 gives only a limited amount of information on each strain; the complete genotype for any strain can be obtained from the SGSC or from the laboratory in which the strain originated. Requests for information and for strains can be made to the SGSC by telephone (403-220-6792), fax (403-289-9311), or e-mail (kesander@ acs.ucalgary.ca). In addition, electronic versions of the strain list can be obtained as follows. A delimited text file and FileMaker Pro (Version 2) database containing complete descriptions of the strains listed in Table 2 are available directly via the University of Calgary Gopher server or by contacting the SGSC. Further details about the structure of the database are described in chapter 136 of this volume. The files can be obtained through Internet as follows: access the University of Calgary Gopher, and then select the following in order: "Faculty and Department information," "Department of Biological Sciences," "Salmonella Genetic Stock Centre." On-line files are updated at regular intervals.

ACKNOWLEDGMENTS

We are greatly indebted to the many investigators who provided the strains listed in Table 2 along with published or unpublished genetic information. We thank Tim Galitsky and Nick Benson for their critical comments regarding the manuscript.

Work in Calgary was supported by an Infrastructure Grant and an Operating grant from the Natural Sciences and Engineering Research Council of Canada and by grant R01AI34829 from the National Institute of Allergy and Infectious Diseases. Work in Utah was supported by grant GM27068 from the National Institutes of Health.