Genetic Mapping by Duplication Segregation in Salmonella enterica

Corresponding author: Josep Casadesús, Departamento de Genética, Facultad de Biología, Universidad de Sevilla, Apartado 1095, Sevilla 41080, Spain., genbac{at}cica.es (E-mail)

Communicating editor: P. L. FOSTER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

MudP and MudQ elements were used to induce duplications in Salmonella enterica by formation of a triple crossover between two transduced fragments and the host chromosome. The large size (36 kb) of MudP and MudQ is a favorable trait for duplication formation, probably because homology length is a limiting factor for the central crossover. Additional requirements are a multiplicity of infection of 2 or higher in the infecting phage suspensions (which reflects the need of two transduced fragments) and an exponentially growing recipient (which reflects the need of a chromosome replication fork). We describe a set of 11 strains of S. enterica, each carrying a chromosomal duplication with known endpoints. The collection covers all the Salmonella chromosome except the terminus. For mapping, a dominant marker (e.g., a transposon insertion in or near the locus to be mapped) is transduced into the 11-strain set. Several transductants from each cross are grown nonselectively, and haploid segregants are scored for the presence of the marker. If all the segregants contain the transduced marker, it maps outside the duplication interval. If the marker is found only in a fraction of the segregants, it maps within the duplicated region.

A classical review of genetic duplications in bacteria (ANDERSON and ROTH 1977 ) envisaged that a potential use of duplications might be genetic mapping. The logic of such a mapping system is simple: a marker introduced into a preexisting duplication will segregate together with the duplication, while a marker outside the duplication will not (Fig 1). In practice, the development of a procedure for mapping by duplication segregation requires the following tools: (i) a method to construct duplications with known endpoints; (ii) a procedure to introduce the marker to be mapped into one copy of the duplication; (iii) a screen to detect segregation of the duplication; and (iv) a screen to score cosegregation of the marker under study. All these tools are available for Salmonella, combining P22 HT transduction (SCHMIEGER 1972 ), the use of transposable elements as genetic tools (KLECKNER et al. 1977 ; MALOY et al. 1996 ), and a number of procedures developed by John Roth and co-workers for the construction and analysis of genetic rearrangements (reviewed by ROTH et al. 1996 ).

View larger version (22K): In this window In a new window Download PPT slide   Figure 1. Logic of genetic mapping by duplication segregation.

This study describes the construction of duplications covering all regions of the Salmonella enterica chromosome except the terminus, where duplications were not obtained. Observations made by other authors suggest that duplications involving the terminus may be inviable (D. R. HILLYARD and J. R. ROTH, personal communication). All duplications have been constructed with the procedure of HUGHES and ROTH 1985 , except that the Mud-P22 prophages MudP and MudQ (YOUDERIAN et al. 1988 ) have been used instead of MudA elements. MudP and MudQ are hybrids between Salmonella phage P22 and coliphage Mu (YOUDERIAN et al. 1988 ). Each consists of about two-thirds of the phage P22 genome placed between the ends of Mu; the construction includes also a chloramphenicol resistance marker. The MudP and MudQ constructs contain the P22 immunity region but lack the P22 sieA (superinfection exclusion) gene; this permits DNA injection by phage P22 (POTEETE 1988 ). MudP and MudQ elements also lack the P22 mnt gene, which encodes a repressor for the maintenance of lysogeny (BOTSTEIN et al. 1975 ). Absence of mnt permits the expression of lytic genes in superinfecting P22 phage (BOTSTEIN et al. 1975 ; POTEETE 1988 ). Thus MudP and MudQ lysogens can be lysed by P22, and the resulting lysates are suitable for generalized transduction, since the 36-kb MudP and MudQ elements can be accommodated in a P22 head (CASJENS and HAYDEN 1988 ).

We have constructed a collection of MudP/Q-held duplications that permits mapping of a dominant mutation to any of 11 chromosomal intervals. Mapping is carried out by P22 HT-mediated transduction, followed by segregation analysis. This technically simple method may be complementary to mapping with lysates of "locked-in" Mud-P22 prophages and can sort out intricacies derived from bidirectional packaging of markers flanking a Mud-P22 hybrid (BENSON and GOLDMAN 1992 ; FLORES and CASADESUS 1995 ). The duplication collection can be also used for dominance/recessivity tests and other operations of genetic analysis. We also discuss experimental procedures and conditions that facilitate the construction of MudP/Q-induced duplications; use of such procedures may be helpful to add novel duplication-bearing strains to the current 11-strain set.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Bacterial strains:
Strains of S. enterica, serovar Typhimurium (usually abbreviated as S. typhimurium) listed in Table 1 are derived from the standard, wild-type strain LT2. Strain MST1202 was a gift from Stanley R. Maloy, University of Illinois, Urbana, Illinois. Strain TT16714 of S. typhimurium was obtained from Rafael Camacho, Universidad Nacional Autónoma de México, Mexico City, Mexico. The virulent strain S. typhimurium SL1344 (HOISETH and STOCKER 1981 ) was provided by Francisco García-del Portillo, Universidad Autónoma de Madrid, Cantoblanco, Spain. Strain 3246 of S. dublin was provided by Timothy S. Wallis, Institute for Animal Health, Compton, England. Strain SAO44 of S. abortusovis was obtained from Salvatore Rubino, Istituto di Microbiologia, Università degli Studi di Sassari, Sassari, Italy. Strain G9/4223 of S. gallinarum was a gift from John E. Olsen, Royal Veterinary and Agricultural University, Fredericksberg, Denmark. Mud1-8[AmpLac] (HUGHES and ROTH 1984 ) and MudI1734[KanLac] (CASTILHO et al. 1984 ) are transposition-deficient Mu derivatives. These elements have been renamed MudA and MudJ, respectively (HUGHES and ROTH 1985 ). MudP and MudQ are P22-Mu hybrids that carry a chloramphenicol-resistance marker (YOUDERIAN et al. 1988 ). The nomenclature for duplications follows the rules of HUGHES and ROTH 1985 .

Culture media:
The E medium of VOGEL and BONNER 1956 was used as minimal medium. The carbon source was 0.2% glucose. Auxotrophic requirements and antibiotics were used at the concentrations described by MALOY et al. 1996 . The rich medium was Difco nutrient broth (NB; 8 g/liter) with added NaCl (5 g/liter). Solid media contained Difco agar at 1.5% final concentration. Green plates were prepared according to CHAN et al. 1972 , except that methyl blue (Sigma, St. Louis) substituted for aniline blue.

Bacteriophages and transduction:
The transducing phage was P22 HT 105/1 int201 (SCHMIEGER 1972 ; G. ROBERTS, unpublished data), henceforth referred to as "P22 HT." Phage sensitivity was tested by cross-streaking with the clear-plaque mutant P22 H5. Phage NBP182 erf (am1173) is a derivative of P22 HT 105/1 provided by Nicholas R. Benson, Sidney Kimmel Cancer Center, San Diego, California. Erf^- (essential recombination function) mutants of P22 are unable to grow on RecA^- strains (BOTSTEIN and MATZ 1970 ). Transducing lysates were prepared according to DAVIS et al. 1980 . The titers of phage suspensions were calculated from counts of plaque-forming units on strain LT2. The multiplicity of infection (MOI) is the number of input phage per colony-forming unit. For transduction, 0.2 ml of the recipient culture was mixed with 0.2 ml of phage suspension, diluted as needed. The mixtures were preincubated for 35 min at 37° with gentle shaking before spreading on nutrient agar (NA) supplemented with the appropriate antibiotic. To obtain phage-free isolates, transductants were purified by streaking on green plates.

Construction of "enlarged" MudJ elements:
The MudJ elements carried by strains TT12313 and TT12316 were "enlarged" with a Tn10, introduced into their respective lacZ genes by homologous recombination. For this purpose, a P22 high-transducing (HT) lysate grown on strain MST1202 was used to transduce TT12313 and TT12316, selecting tetracycline resistance. Individual Tc^r transductants were made phage free and patched with toothpicks to appropriate media. Isolates with the appropriate phenotype (Km^r Tc^r Ap^s Lac^-) were propagated as strains SV4018 (derived from TT12313) and SV4019 (derived from TT12316). In both strains, the kanamycin resistance marker of MudJ and the tetracycline resistance marker of Tn10 proved to be 100% linked by cotransduction analysis.

Converting MudJ elements to MudP or MudQ:
The MudJ element of strain TT12318 and the MudA element of strain TT16714 were both converted to MudQ elements to generate strains SV4130 and SV4131, respectively. SV4131 is actually the product of a second transductional cross using LT2 as the recipient; for this reason, it lacks the argR9002::Tn10dTet mutation. Replacement of MudA or MudJ with MudQ was performed according to BENSON and GOLDMAN 1992 , using TT12916 as the donor strain. For tailing, the lysate obtained upon induction of TT12916 was mixed with an excess of P22 tails (YOUDERIAN et al. 1988 ).

Transductional test for the classification of duplication-carrying isolates:
Individual transductants putatively carrying induced duplications were made phage free and lysed with P22 HT. The lysates were used to transduce LT2, selecting Cm^r transductants on NA-chloramphenicol plates. One hundred or more Cm^r transductants were then replica printed to minimal plates. Reconstruction of the duplication was detected by the formation of Cm^r prototrophic transductants.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Properties of P22 HT lysates grown on MudP/Q lysogens:
Because MudP/Q elements lack the sieA and mnt genes of P22, MudP/Q lysogens can be lysed by infecting P22 HT phage. The resulting lysates have titers similar to those obtained on nonlysogenic hosts. If used for transduction, these lysates yield Cm^r transductants at frequencies around 10^-5 per pfu. If the transduced MudP/Q insertion caused auxotrophy, all the transductants were auxotrophs. These observations suggest that MudP/Q elements can be transduced as ordinary transposon insertions. In addition, the following experiments indicated that lytic growth of P22 on a MudP/Q lysogen does not cause recombinational rearrangements at high frequency:

1. Twenty independent Cm^r transductants, generated upon transduction of the insertion putA1019::MudP to strain NB2, proved to be inducible by mitomycin C and able to package the nearby serC locus. Thus the MudP element appeared intact after transduction.

2. Formation of P22 HT derivatives able to perform specialized transduction of the Cm^r marker did not occur. In these experiments, six independent lysates of TT15240 were used to transduce a RecA^- recipient (SV1235), selecting chloramphenicol resistance. Cm^r transductants were not obtained. Recombination is probably constrained by the position of the MudP chloramphenicol resistance determinant, which may lack homology for crossover at one side.

3. Evidence that recombination between MudP/Q elements and superinfecting phage occurred at low frequency was provided by the formation of Erf⁺ recombinants during lytic growth of a P22 Erf^- mutant on a MudP lysogen. Lysates of NBP182, an Erf^- derivative of P22, were obtained on a pair of largely isogenic strains, TT15240 and TT6934. The resulting phage suspensions were then plaqued on RecA⁺ and RecA^- hosts (LT2 and SV1235). Only Erf⁺ phage can form plaques on SV1235, because P22 Erf^- mutants are unable to grow on RecA^- strains (BOTSTEIN and MATZ 1970 ). A higher frequency of Erf⁺ phage was detected in the lysate of TT15240 (3 x 10^-8 vs. 4 x 10^-9 recombinants per pfu), providing evidence that the infecting Erf^- mutant had recombined, albeit at low frequency, with the MudP (Erf⁺) prophage.

Single- vs. double-lysate transduction as a screen for the formation of directed duplications:
The method of HUGHES and ROTH 1985 for the construction of directed duplications employs two MudA elements that provide homology for a crossover between two fragments independently transduced into a single cell. If host sequences adjacent to each MudA element form two additional crossovers with a replication fork in the recipient chromosome, the triple recombination event results in the duplication of the chromosomal region between the original MudA insertions and the reconstruction of a single MudA element at the joint point of the duplicated segments. The resulting recombinants can be selected by their ability to form prototrophic, ampicillin-resistant colonies (HUGHES and ROTH 1985 ). We applied this procedure to duplication formation using MudP and MudQ prophages (YOUDERIAN et al. 1988 ). The advantages expected from the use of these elements will be discussed below.

As a screen to detect the occurrence of duplications induced by MudP/Q elements, we compared the frequencies of duplications obtained in ordinary, one-lysate transductions to the frequency obtained upon two-lysate transduction. The screen relies on the consideration that one-lysate transduction cannot induce duplication formation. However, prototrophic Cm^r transductants can be formed by one-lysate transduction if the incoming fragment carrying a MudP/Q element recombines with one copy of a duplicated gene already present in the infected culture (ANDERSON and ROTH 1979 , ANDERSON and ROTH 1981 ). Thus an excess frequency of prototrophic transductants obtained upon two-lysate transduction (compared to one-lysate transduction) provides evidence of MudP/Q-directed duplications. In all cases, the frequency of duplications was measured as the ratio of the number of prototrophic Cm^r transductants over the total number of Cm^r transductants.

Because the incidence of spontaneous duplications varies among different chromosomal regions (ANDERSON and ROTH 1981 ), comparisons between one-lysate and two-lysate transductions were performed in the following chromosomal intervals: thr–proA [centisomes (CS) 0-7], proA–purE (CS 7-12), purE–purB (CS 12-27), hisH–cysA (CS 44-53), and cysG–ilvA (CS 75-85). An experiment for each chromosomal interval is shown in Table 2. A significant increase in the frequency of Cm^r prototrophs by two-lysate transduction was observed in all intervals assayed except cysG–ilvA. In the latter interval, the absence of differences may be explained by the higher incidence of spontaneous duplications affecting the ilvA region. Chromosomal regions near the origin of replication undergo high frequencies of spontaneous duplication (ANDERSON and ROTH 1979 , ANDERSON and ROTH 1981 ). The ilvA gene lies at CS 85, 3, relatively close to the origin of replication of the Salmonella chromosome, located at CS 84, 9 (SANDERSON et al. 1995 ).

Duplications across the terminus (between purE and trp) were not obtained (data not shown). This failure is in accordance with previous observations (D. R. HILLYARD and J. R. ROTH, personal communication).

In favorable cases, the two-lysate/one-lysate screen provides evidence that directed duplications have been formed. However, the screen does not identify the individual transductants that carry them: isolates carrying spontaneous duplications and those bearing duplications directed by MudP/Q are all Cm^r prototrophs. As described below, each type can be distinguished by its segregation pattern and by its behavior as a transductional donor.

Segregation patterns of spontaneous and directed duplications:
Cm^r prototrophic transductants generated by recombination of a transduced fragment bearing the Mud element with a preformed, spontaneous duplication will segregate Cm^s protrotrophs and Cm^r auxotrophs (ANDERSON and ROTH 1981 ). In contrast, merodiploids carrying Mud-held, directed duplications will segregate only chloramphenicol-sensitive protrotrophs (HUGHES and ROTH 1985 ). To analyze the segregation pattern of individual transductants, five or more Cm^r prototrophic transductants from each cross were inoculated in NB, grown to full density, diluted, and spread on nutrient agar. One hundred individual colonies derived from a single transductant were replica printed to minimal plates, minimal plates with chloramphenicol, NA plates, and NA plates with chloramphenicol. The data shown in Table 3 can be summarized as follows:

1. Isolates carrying duplications of either type (spontaneous or directed) were obtained in all crosses. In the examples shown in Table 3, all (5/5) Cm^r transductants from donors TT15224 and TT15231 carried directed duplications. The incidence of directed duplications was lower in other crosses: only one transductant (III) from donors TT15251 and TT15254 and two from donors TT15264 and TT15269 (IV and V) carried a directed duplication.

   
   

   View this table: In this window In a new window   Table 3. Identification of isolates carrying duplications (spontaneous or directed) by their segregation patterns

2. All isolates carrying duplications underwent segregation when grown in nonselective medium (NB), irrespective of the type of duplication carried. However, the frequency of segregants varied largely among different isolates. In the examples shown in Table 3, the slowest segregation rates (reflected in the higher number of Cm^r prototrophs) were found in crosses involving TT15269. One explanation might be that duplication of this region confers a selective growth advantage (ANDERSON and ROTH 1981 ).

A cross-check for isolates putatively carrying induced duplications was their ability to yield Cm^r prototrophic transductants if used as donors. In these crosses, transduction of the duplication joint point reconstructs the duplication by a double crossover with a replicating chromosome (HILL et al. 1969 ; ANDERSON et al. 1976 ). In every cross, all transductants putatively carrying directed duplications (as judged from their segregation pattern) also passed the transductional test. Thus the congruence and reliability of both tests is high.

Effect of the multiplicity of infection on duplication formation:
Formation of a triple crossover between two MudA elements and a replicating chromosome requires two independently transduced fragments (HUGHES and ROTH 1985 ). To investigate whether our MudP/Q-based procedure for duplication formation involved a similar mechanism, we studied the effect of MOI upon duplication formation. The underlying rationale was that, if duplication formation does require two transducing fragments, it should show a second-order dependence upon the MOI of the lysates (MIESEL and ROTH 1996 ). On these grounds, we performed transductional crosses in which a constant number of recipient cells was transduced with two lysates at various MOIs. To minimize the bias caused by spontaneous duplications, these experiments were carried out at the chromosomal region of CS 7-12, where the frequency of spontaneous duplications is low. The donor strains were TT15229 and TT15232; the recipient was LT2. A prediction is that the number of transductants (T) obtained will be a function of the phage multiplicity (P) raised to the power of the number of transducing particles required (n),

where k is a proportionality constant. Plotting the logarithm of T vs. the logarithm of P will generate a straight line, whose slope will be the number of particles required to generate a transductant (HUGHES et al. 1987 ; MIESEL and ROTH 1996 ):

Data plotted in Fig 2 indicate that the slope is 2. This result confirms that two transducing particles, each carrying a MudP/Q element, are required to generate a duplication. Because MOI values were calculated from plaque-forming units and P22 HT suspensions contain 50% of transducing particles (MALOY et al. 1996 ), the actual MOI is likely higher than 2. Thus the possibility that additional infecting phage might be involved cannot be ruled out, as suggested by MIESEL and ROTH 1996 for ordinary, one-fragment transduction.

View larger version (9K): In this window In a new window Download PPT slide   Figure 2. Effect of the multiplicity of infection on the formation of directed duplications. MOI is the multiplicity of infection, calculated from counts of plaque-forming units. T is the number of prototrophic Cmr transductants. Results are averages from four independent experiments; in each experiment, four assays (plates) per MOI were performed. Donors were TT15229 (proA692::MudQ) and TT15232 (purE2154::MudQ); the recipient was LT2.

Effect of marker size on the formation of directed duplications:
Because of the headful packaging mechanism of P22, all fragments transduced by P22 HT have a similar length, 44 kb (CASJENS and HAYDEN 1988 ). Thus every fragment carrying a Mud element will contain host sequences at one or both sides; the occurrence and length of these sequences will be constrained, among other factors, by the size of the Mud element itself. If large Mud elements are used (e.g., MudA, 38 kb, or MudP/MudQ, 36 kb) transducing fragments containing the entire element can be expected to be rare. In contrast, small elements such as MudJ (11 kb) will have more chances of being incorporated completely into the transducing particle. Transduced fragments carrying small elements will also contain flanking host DNA sequences at one of both sides, while fragments carrying large elements will likely contain flanking DNA at only one side. With these constraints in mind, one can envisage two possibilities:

1. That a limiting factor for the formation of a directed duplication is the length of the sequences required to form the flanking crossovers, as shown for ordinary, one-fragment transduction (MIESEL and ROTH 1994 ). If such were the case, directed duplications would be more efficiently formed by small elements.

2. That a limiting factor for the formation of a directed duplication is the length of the DNA sequence involved in the central crossover (e.g., the Mud element). If such were the case, directed duplications would be more efficiently formed by large elements. Furthermore, if a large size favors incomplete packaging of Mud elements, selection of Cm^r prototrophic transductants will enforce triple crossover (HUGHES et al. 1987 ).

To examine these possibilities, we compared the frequencies of duplication formation between purE to purB (centisomes 12 to 25) using markers of different sizes: MudJ (11 kb), a MudJ derivative enlarged with a Tn10 insertion (11 + 10 kb = 21 kb) and MudP (36 kb). Formation of directed duplications was assessed by comparing double-lysate transductions to single-lysate transductions. The latter measures the background frequency of spontaneous duplications (see above and ANDERSON and ROTH 1981 ). The results, shown in Fig 3, were unambiguous: transduction with two lysates increased the frequency of Pur⁺ Cm^r isolates in a size-dependent fashion, suggesting that triple crossover is facilitated by the use of large elements. Support for the conclusion that the element size is a limiting factor for triple crossover formation was also obtained in the hisC–cysJ interval: use of two MudA insertions (carried by strains TT9686 and TT9636) yielded duplications at a frequency 17-fold higher than two MudJ elements (carried by SV2090 and SV3038; data not shown). Size dependence admits two nonexclusive interpretations: (i) that a large size increases the homology length required for the central crossover and (ii) that a large size favors incomplete packaging, thus making the central crossover essential for the formation of Cm^r prototrophs.

View larger version (12K): In this window In a new window Download PPT slide   Figure 3. Effect of marker size on the formation of directed duplications. The frequency of duplications is calculated as the quotient between the number of Cmr protrotrophic transductants and the number of Cmr transductants. Because of the high numerical dispersion typical of these experiments, the median is shown instead of the average, and error bars are not included. Each median has been calculated from more than six independent experiments. P22 HT donor lysates were from the following strains: A, TT12313 (purE2154::MudJ); B, TT12316 (purB1879::MudJ); C, SV4018(purE2154::MudJ [lacZ::Tn10]); D, SV4019 (purB1879::MudJ [lacZ::Tn10]); E, TT15235 (purE2154::MudQ); and F, TT15242 (purB1879::MudP). The recipient was LT2.

Effect of the growth rate of the recipient culture on duplication formation:
Besides the central crossover, directed duplication of a chromosomal region requires two flanking crossovers, each with a different daughter DNA duplex after passage of a replication fork (HUGHES and ROTH 1985 ). If this model applies to duplication formation induced by MudP/Q, the number of transductants should be a function of the number of replication forks present in the cell. In turn, the number of replication forks is a function of growth rate (SCHAECHTER et al. 1958 ). On these grounds, we compared the frequencies of duplication formation in exponential and stationary cultures. We chose a region in which the frequency of spontaneous duplications is low (CS 7-12) and another region in which spontaneous duplications are common (CS 72-83). To measure the background level of spontaneous duplications, transductions with single lysates were performed. The exponential cultures used had optical densities (OD₅₆₀) of 0.5–0.8 and were concentrated 10-fold before being mixed with the phage suspension. The stationary cultures had optical densities of 2. The conclusion from these experiments, summarized in Fig 4, is that both spontaneous and directed duplications occur more frequently in exponential cultures. Thus the formation of both duplication types may require a replicating chromosome.

View larger version (20K): In this window In a new window Download PPT slide   Figure 4. Effect of growth rate on the formation of directed duplications. The median of more than six independent experiments is shown (see the legend of Fig 3). Donors were the following: A, TT15229 (proA692::MudQ); B, TT15232 (purE2154::MudQ); C, TT15264 (cysG1573::MudP); and D, TT15269(ilvA2648::MudP). The recipient was LT2.

Effect of marker orientation on the formation of directed duplications:
A critical factor for duplication formation by MudA elements is their orientation; in fact, duplication formation by two MudA elements can be used to ascertain the direction of transcription of a gene or operon (HUGHES and ROTH 1985 ). An appropriate orientation is also required for duplication formation by MudP/Q elements. As shown in Table 4, double-lysate transductions with two MudP or two MudQ elements in opposite orientation failed to increase the frequency of Cm^r prototrophic transductants over those obtained with single lysates. However, the symmetrical nature of the MudP and MudQ constructs (YOUDERIAN et al. 1988 ; BENSON and GOLDMAN 1992 ) permits a more versatile use of these elements in duplication formation. A given MudP element can recombine either with a MudP element in the same orientation or with a MudQ element in the opposite orientation (Table 4). Recombination between a MudP and a MudQ element can be expected to generate hybrid elements of either type, whose ends will be both MuL or both MuR. These rearragements are irrelevant for the formation of the duplication and for the stability and segregation capacity of the resulting merodiploid.

Construction of a collection of MudP/Q-held duplications covering the Salmonella chromosome:
Using the procedures described above, 11 strains carrying duplications with known endpoints were constructed (Fig 5). Seven of these strains carry duplications held by a MudP element; 4 carry duplications held by MudQ. The collection covers all the Salmonella chromosome except the purB–trp region (CS 27-38). The longest duplication (purE–purB) is 15 centisomes long; the shortest are cysA–purG (3 centisomes) and proA–purE (4 centisomes). The remaining duplications are all 7–10 centisomes long. All the duplication-carrying strains are stable if maintained in chloramphenicol-containing media. Long-term preservation can be carried out in frozen stocks, and massive segregation during recovery is not observed. An alternative preservation procedure is the storage of a P22 lysate grown on the merodiploid strain; high-titer lysates sterilized with chlorophorm are able to transduce the duplication joint point (and thus to reconstruct the duplication) after prolonged storage at 4°. For reconstruction, the original duplication-bearing strain is used as a donor in a transductional cross mediated by P22 HT, selecting Cm^r transductants in minimal medium (HILL et al. 1969 ; ANDERSON et al. 1976 ).

View larger version (8K): In this window In a new window Download PPT slide   Figure 5. A diagram of the 11-duplication set that covers most of the Salmonella chromosome (not drawn to scale). The endpoints of each duplication are indicated by the number of the corresponding centisomes (CS).

Duplications held by MudP/Q elements also could be reconstructed in genetic backgrounds other than LT2, such as the standard pathogenic strain S. typhimurium SL1344, and strains belonging to other Salmonella serovars of medical or veterinary importance, such as S. dublin 3246, S. abortusovis SAO44, and S. gallinarum G9/4223. Cm^r prototrophic transductants were obtained for all recipients tested, and the segregation patterns of individual transductants confirmed that all carried a directed duplication (data not shown). DNA fragments transduced by phage P22 undergo massive degradation at their ends before integration (EBEL-TSIPIS et al. 1972 ; MIESEL and ROTH 1994 ). Thus transduction of a duplication joint point to a genetically different host can be expected to introduce a small amount of heterologous DNA in the recipient chromosome aside from the MudP/Q element. A potential application of these constructions is the genetic analysis of chromosomal regions that are highly polymorphic among Salmonella serovars (CONNER et al. 1998 ).

Mapping by duplication segregation:
Mapping is carried out in three steps. First, the marker to be mapped is transferred by P22 HT transduction to all members of the 11-strain collection. One or more isolates from each cross are then allowed to segregate. Finally, segregants are scored for the presence of the marker to be mapped. The interpretation of the results is straighforward: if all the Cm^s segregants contain the transduced marker, it maps outside the duplication. If the marker is found only in a fraction of the segregants, it maps within the duplicated region (Fig 1).

In the example shown in Table 5, the insertion ara-651::Tn10 was transduced into the 11 duplication-carrying strains, selecting Tc^r transductants on NA with tetracycline and chloramphenicol. Three individual Tc^r Cm^r transductants from each cross were grown to full density in NB to permit segregation. The cultures were then diluted and spread on NA plates to obtain single colonies. One hundred colonies from each culture were patched to the following media: (i) NA with chloramphenicol and tetracycline, to detect nonsegregating Cm^r Tc^r colonies; (ii) NA with tetracycline, to detect the presence of segregants (which are Cm^s); and (iii) NA, to detect Cm^s Tet^s segregants (in which the duplication and the Tn10 element have cosegregated). Cm^s Tc^s segregants were found only in isolates derived from SV1604, indicating that the mutation ara-651::Tn10 maps in the thr–proA interval between centisomes 0 and 7. This mapping location is correct, since the araBAD operon maps at CS 2.5 (SANDERSON et al. 1995 ). Strains SV1603 and SV4142, whose duplications lie at each side of the duplication of SV1604, have been included in Table 5 as representatives of the 11 strains that do not yield Cm^s Tc^s segregants.

In duplications affecting certain chromosomal regions, segregation is slow even in the absence of selection (see Table 3). In such cases, the number of Cm^r segregants found upon nonselective growth can be insufficient to score cosegregation of the transduced marker. To solve this problem, longer segregation periods can be allowed (e.g., during 40–60 generations, which can be achieved by two or more serial cultures in NB). Slow segregation is the only potential complication of the method, which is otherwise unambiguous.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Genetic mapping in Salmonella can be performed by several procedures: cotransduction of nearby markers by high-transducing derivatives of phage P22 (SCHMIEGER 1972 ), conjugal transfer of DNA via Hfr formation (DEMEREC et al. 1955 ), and packaging of discrete chromosomal stretches upon induction of locked-in Mud-P22 prophages (BENSON and GOLDMAN 1992 ). Cotransduction with P22 HT is probably the easiest and most reliable mapping method ever described (MALOY et al. 1996 ). However, it can be applied only to closely linked markers. For large-scale mapping, mobilization of the Salmonella chromosome by the Escherichia coli F sex factor has been used (SANDERSON and MACLACHLAN 1987 ). F can be transferred to Salmonella, and Hfr strains can be obtained upon F integration into the chromosome (SANDERSON and MACLACHLAN 1987 ). However, the efficiency of Hfr formation in Salmonella is reduced; one cause may be the absence of insertion elements IS2 and IS3 from the chromosome of S. typhimurium LT2 (LAM and ROTH 1983 ; MAHILLON and CHANDLER 1998 ). The development of transposon technology introduced a number of refinements for Hfr formation in Salmonella, such as the use of transposable elements as portable regions of homology to direct F integration (KLECKNER et al. 1977 ; CHUMLEY et al. 1979 ). However, Hfr mapping in Salmonella can still encounter problems derived from the instability of the Hfr donors.

A mapping procedure that can efficiently substitute for Hfr transfer in S. typhimurium was developed by BENSON and GOLDMAN 1992 . The method employs MudP and MudQ, two hybrids between Salmonella phage P22 and coliphage Mu (YOUDERIAN et al. 1988 ). MudP and MudQ are inducible by DNA-damaging treatments. Upon induction, the entire prophage genome is replicated in situ but cannot excise. As a consequence, replication forks initiated at the P22 replication origin invade neighboring host DNA, causing a selective amplification of the chromosomal regions that flank the prophage (YOUDERIAN et al. 1988 ; BENSON and GOLDMAN 1992 ). Replicated regions are then packaged into P22 capsids by the "headful" mechanism characteristic of phage P22 (CASJENS and HAYDEN 1988 ). The first headful will package a portion of the prophage and some adjacent host DNA; further headfuls will consist of chromosomal DNA only (BENSON and GOLDMAN 1992 ). The packaging direction depends on the orientation of the prophage itself. The MudP and MudQ constructs are identical except that each carries the central, P22-derived region in opposite orientation (YOUDERIAN et al. 1988 ). As a consequence, MudP will package in one direction and MudQ in the other when inserted at the same locus (BENSON and GOLDMAN 1992 ). Use of a set of MudP and MudQ lysogens scattered along the Salmonella chromosome permits mapping with a resolution of 5 min, provided that a selection procedure is available (BENSON and GOLDMAN 1992 ). Bidirectional packaging of markers flanking a MudP/Q prophage is one of the few drawbacks of this technique (BENSON and GOLDMAN 1992 ; FLORES and CASADESUS 1995 ).

Mapping by duplication segregation, the procedure described in this study, was envisaged long ago (ANDERSON and ROTH 1977 ) but has remained unexploited. The procedure is based on the rationale that a marker introduced into a preexisting duplication will segregate together with the duplication, while a marker outside the duplication will not (Fig 1). To assay this mapping procedure in Salmonella, we constructed a collection of duplications held by MudP/Q elements. Duplication construction was achieved by adapting the elegant method of HUGHES and ROTH 1985 , originally devised for MudA elements, to MudP and MudQ. The choice of these elements over other transposons (e.g., MudA) was based on the following criteria: (i) MudP and MudQ have a large size, which is a desirable trait to construct directed duplications (Fig 3); (ii) large collections of MudP and MudQ inserts exist (BENSON and GOLDMAN 1992 ; N. R. BENSON, unpublished data); (iii) because MudP and MudQ carry a chloramphenicol resistance marker, duplications held by these elements can be used to map mutations tagged with tetracycline, ampicillin, and kanamycin resistance markers; and (iv) for the selection of rare recombinants, chloramphenicol resistance is a highly reliable marker, because it does not involve the production of extracellular enzymes.

The extensive homology between Mud-P22 elements and P22 DNA permits recombination with superinfecting phage during lytic growth and/or transduction. Thus, isolates carrying transduced MudP/Q elements should not be used for induction of locked-in prophages. An additional reason to discourage induction of such lysogens is the presence of the mitomycin-inducible Fels prophage in the LT2 background. Last, it seems also advisable that the P22 HT lysates used for such transductions are not further propagated, because they might contain P22 HT recombinant derivatives. Aside from these caveats, transduction of MudP/Q elements by P22 HT appears to be a safe procedure, at least for operations like those described below, in which MudP/Q elements are used simply to provide homologies for recombination.

Duplication formation induced by MudP/Q elements followed relatively simple rules. One was the need to co-infect the recipient culture with two P22 HT lysates, each at a MOI of >2. This condition seems to reflect the need of two transduced fragments (Fig 2), as previously proposed for MudA elements (HUGHES and ROTH 1985 ; HUGHES et al. 1987 ). Use of an exponentially growing culture increased the overall yield of duplications, both spontaneous and induced, suggesting that formation of both types requires a replication fork (Fig 4). The populations of Cm^r prototrophic transductants were always heterogeneous and contained both spontaneous and induced duplications (Table 3). However, each type could be easily distinguished by its segregation pattern (ANDERSON and ROTH 1977 , ANDERSON and ROTH 1979 ; HUGHES and ROTH 1985 ; ROTH et al. 1996 ). Directed duplications could be formed by two MudP or MudQ elements in the same orientation or by a MudP and a MudQ element in opposite orientation (Table 4). This orientation dependence for successful crossover confirmed that duplication formation was actually directed by the MudP/Q elements involved in the cross.

Each of the 11 strains shown in Fig 5 is a merodiploid carrying a duplication held by a MudP/Q element. If maintained under selection, these strains are stable; however, segregation can be easily achieved by growth under conditions that do not select the duplication. Segregation occurs fast in all strains but SV4195. For mapping, a dominant marker (e.g., a transposon insertion) is transferred by P22 HT transduction into the 11-merodiploid set. The resulting isolates contain the duplication and the marker to be mapped. These isolates are then grown nonselectively (e.g., in NB) to permit segregation, and haploid segregants are scored for the presence of the insertion mutation. If all the segregants contain the insertion, it maps outside the duplication interval. If the insertion is found only in a fraction of the segregants, it maps within the duplicated region (Fig 1).

Aside from mapping, chromosomal duplications with known endpoints can be used for other operations of genetic analysis. For instance, the introduction of a mutation into one copy of a chromosomal duplication permits complementation analysis: the phenotype of a dominant mutation will still be observed in the merodiploid, while that of recessive mutation will require the formation of haploid segregants. An advantage of this complementation procedure (e.g., over complementation with plasmids) is that it does not alter gene dosage. Duplications can also be useful to ascertain whether a mutation is lethal. For this purpose, a plasmid-borne allele can be recombined into one copy of a duplication; the segregation pattern of the resulting heterozygote will indicate whether haploid segregants carrying the mutation are viable. Another potential application of duplications is the analysis of gene fusions, especially for the study of genes that regulate their own transcription. In such cases, use of a duplication permits monitoring of gene expression in the presence of the normal dosage of the wild-type allele.

	ACKNOWLEDGMENTS

We are grateful to John Roth, Kelly Hughes, Stan Maloy, and Andrés Garzón for helpful discussions, to Amando Flores for early experiments of duplication formation, and to David Cano for critical reading of the manuscript. Strains were kindly provided by John Roth, Stan Maloy, John Olsen, Salvatore Rubino, Rafael Camacho, Francisco García-del Portillo, Tim Wallis, Nick Benson, and by the Salmonella Genetic Stock Centre, University of Calgary, Canada. The assistance of Ana Moreno, José Córdoba, and Luis Romanco is also acknowledged. This study was supported by grant PM97-0148-CO2-02 from the Dirección General de Enseñanza Superior of the Government of Spain and by grant PL96/1743 from the FAIR Program of the European Union.

Manuscript received July 10, 2000; Accepted for publication October 16, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ANDERSON, R. P. and J. R. ROTH, 1977  Tandem genetic duplications in phage and bacteria. Annu. Rev. Microbiol. 31:473-505[Medline].

ANDERSON, R. P. and J. R. ROTH, 1979  Gene duplication in bacteria: alteration of gene dosage by sister-chromosome exchanges. Cold Spring Harbor Symp. Quant. Biol. 43:1083-1087.

ANDERSON, R. P. and J. R. ROTH, 1981  Spontaneous tandem genetic duplications in Salmonella typhimurium arise by unequal recombination between rRNA (rrn) cistrons. Proc. Natl. Acad. Sci. USA 78:7113-7117.

ANDERSON, R. P., C. G. MILLER, and J. R. ROTH, 1976  Tandem duplications of the histidine operon following generalized transduction in Salmonella typhimurium.. J. Mol. Biol. 105:201-218[Medline].

BENSON, N. R. and B. S. GOLDMAN, 1992  Rapid mapping in Salmonella typhimurium with Mud-P22 prophages. J. Bacteriol. 174:1673-1681[Abstract/Free Full Text].

BOTSTEIN, D. and M. J. MATZ, 1970  A recombination function essential to the growth of bacteriophage P22. J. Mol. Biol. 54:417-440[Medline].

BOTSTEIN, D., K. K. LEW, V. JARVIK, and C. A. SWANSON, JR., 1975  Role of antirepressor in the bipartite control of expression and immunity by bacteriophage P22. J. Mol. Biol. 91:439-462[Medline].

CASJENS, S. and M. HAYDEN, 1988  Analysis in vivo of the bacteriophage P22 headful nuclease. J. Mol. Biol. 199:467-474[Medline].

CASTILHO, B. A., P. OLFSON, and M. J. CASADABAN, 1984  Plasmid insertion mutagenesis and lac gene fusion with Mini-Mu bacteriophage transposons. J. Bacteriol. 158:488-495[Abstract/Free Full Text].

CHAN, R. K., D. BOTSTEIN, T. WATANABE, and Y. OGATA, 1972  Specialized transduction of tetracycline by phage P22 in Salmonella typhimurium. II. Properties of a high frequency transducing lysate. Virology 50:883-898[Medline].

CHUMLEY, F. G., R. MENZEL, and J. R. ROTH, 1979  Hfr formation directed by Tn10.. Genetics 91:639-655[Abstract/Free Full Text].

CONNER, C. P., D. M. HEITHOFF, S. M. JULIO, R. L. SINSHEIMER, and M. J. MAHAN, 1998  Differential patterns of acquired virulence genes distinguish Salmonella strains. Proc. Natl. Acad. Sci. USA 95:4641-4645[Abstract/Free Full Text].

DAVIS, R. W., D. BOTSTEIN and J. R. ROTH, 1980 Advanced Bacterial Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

DEMEREC, M., I. BLOMSTRAND, and Z. E. DEMEREC, 1955  Evidence of complex loci in Salmonella.. Proc. Natl. Acad. Sci. USA 41:359-364[Free Full Text].

EBEL-TSIPIS, J., M. S. FOX, and D. BOTSTEIN, 1972  Generalized transduction by bacteriophage P22 in Salmonella typhimurium. II. Mechanism of integration of transducing DNA. J. Mol. Biol. 71:449-469[Medline].

FLORES, A. and J. CASADESÚS, 1995  Suppression of the pleiotropic effects of HisH and HisF overproduction identifies four novel loci on the Salmonella typhimurium chromosome: osmH, sfiW, sfiX, and sfiY.. J. Bacteriol. 177:4841-4850[Abstract/Free Full Text].

GARZÓN, A., D. A. CANO, and J. CASADESÚS, 1995  Role of Erf recombinase in P22-mediated plasmid transduction. Genetics 140:427-434[Abstract].

HILL, C. W., D. SCHIFFER, and P. BERG, 1969  Transduction of merodiploidy: induced duplication of recipient genes. J. Bacteriol. 99:274-278[Abstract/Free Full Text].

HOISETH, S. K. and B. A. STOCKER, 1981  Aromatic-dependent Salmonella typhimurium are non-virulent and effective as live vaccines. Nature 291:238-239[Medline].

HUGHES, K. T. and J. R. ROTH, 1984  Conditionally transposition-defective derivative of MudI(AmpLac). J. Bacteriol. 159:130-137[Abstract/Free Full Text].

HUGHES, K. T. and J. R. ROTH, 1985  Directed formation of deletions and duplications using Mud(Ap, lac). Genetics 109:263-282[Abstract/Free Full Text].

HUGHES, T. K., B. M. OLIVERA, and J. R. ROTH, 1987  Rec dependence of Mu transposition from P22-transduced fragments. J. Bacteriol. 169:403-409[Abstract/Free Full Text].

KLECKNER, N., J. ROTH, and D. BOTSTEIN, 1977  Genetic engineering in vivo using translocatable drug-resistance elements: new methods in bacterial genetics. J. Mol. Biol. 116:125-159[Medline].

LAM, S. T. and J. R. ROTH, 1983  IS200: a Salmonella-specific insertion sequence. Cell 34:951-960[Medline].

MAHILLON, J. and M. CHANDLER, 1998  Insertion sequences. Microbiol. Mol. Biol. Rev. 62:725-774[Abstract/Free Full Text].

MALOY, S. R., W. J. STEWART and R. K. TAYLOR, 1996 Genetic Analysis of Pathogenic Bacteria. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

MIESEL, L. and J. R. ROTH, 1994  Salmonella recD mutations increase recombination in a short sequence transduction assay. J. Bacteriol. 176:4092-4103[Abstract/Free Full Text].

MIESEL, L. and J. R. ROTH, 1996  Evidence that SbcB and RecF pathway functions contribute to RecBCD-dependent transductional recombination. J. Bacteriol. 178:3146-3155[Abstract/Free Full Text].

POTEETE, A. R., 1988 Bacteriophage P22, pp. 647–682 in The Bacteriophages, Vol. 2, edited by R. CALENDAR. Plenum Press, New York.

ROTH, J. R., N. BENSON, T. GALITSKI, K. HAACK, J. G. LAWRENCE et al., 1996 Rearrangements of the bacterial chromosome: formation and applications, pp. 2256–2276 in Escherichia coli and Salmonella: Cellular and Molecular Biology, Ed. 2, edited by F. C. NEIDHARDT, R. CURTISS, III, J. L. INGRAHAM, E. C. C. LIN, K. B. LOW, B. MAGASANIK, W. S. REZNIKOFF, M. RILEY, M. SCHAECHTER and H. E. UMBARGER. American Society for Microbiology, Washington, D.C.

SANDERSON, K. E., and R. P. MACLACHLAN, 1987 F-mediated conjugation, F⁺ strains and Hfr strains of Salmonella typhimurium and Salmonella abony, pp. 1138–1144 in Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, edited by F. C. NEIDHARDT, J. L. INGRAHAM, K. B. LOW, B. MAGASANIK, M. SCHAECHTER and H. E. UMBARGER. American Society for Microbiology, Washington, D.C.

SANDERSON, K. E., A. HESSEL, and K. E. RUDD, 1995  Genetic map of Salmonella typhimurium, edition VIII. Microbiol. Rev. 52:485-532.

SCHAECHTER, M., M. W. BENTZON, and O. MAALØE, 1958  Synthesis of deoxyribonucleic acid during the division cycle of bacteria. Nature 183:1207-1208.

SCHMIEGER, H., 1972  Phage P22 mutants with increased or decreased transducing abilities. Mol. Gen. Genet. 119:75-88[Medline].

VOGEL, H. and D. BONNER, 1956  Acetylornithase of Escherichia coli: partial purification and some properties. J. Biol. Chem. 218:97-106[Free Full Text].

YOUDERIAN, P., P. SUGIONO, K. L. BREWER, N. P. HIGGINS, and T. ELLIOTT, 1988  Packaging specific segments of the Salmonella chromosome with locked-in Mud-P22 prophages. Genetics 118:581-592[Abstract/Free Full Text].

This article has been cited by other articles:

				C. B. Garcia-Calderon, M. Garcia-Quintanilla, J. Casadesus, and F. Ramos-Morales Virulence attenuation in Salmonella enterica rcsC mutants with constitutive activation of the Rcs system Microbiology, February 1, 2005; 151(2): 579 - 588. [Abstract] [Full Text] [PDF]

				A. I. Prieto, F. Ramos-Morales, and J. Casadesus Bile-Induced DNA Damage in Salmonella enterica Genetics, December 1, 2004; 168(4): 1787 - 1794. [Abstract] [Full Text] [PDF]

				D. A. Cano, G. Dominguez-Bernal, A. Tierrez, F. G.-d. Portillo, and J. Casadesus Regulation of Capsule Synthesis and Cell Motility in Salmonella enterica by the Essential Gene igaA Genetics, December 1, 2002; 162(4): 1513 - 1523. [Abstract] [Full Text] [PDF]