A novel mapping method based on touchdown PCR was developed for identifying a transposon insertion site in genomic DNA using a hybrid consensus-degenerate primer in combination with a specific primer that anneals to the transposon. The method was tested using Xanthomonas citri transposon mutants. PCR products contained adjacent DNA regions that belonged to both X. citri genomic DNA and the transposon. Products were directly sequenced from PCRs using only the specific primer. Different PCR conditions were tested, and the optimized reaction parameters that increased product yields and specificity are described. Best results were obtained with the HIB17 hybrid primer, which is a 25-mer oligonucleotide having degenerate bases at 6 different positions within the last 12 bases at the 3′ end. An X. citri mutants library was produced by random transposition using the EZ::TN transposon, and we identified the insertion sites within the genome of 90 mutants. Insertions were found within both the chromosomal and the plasmid DNA in these X. citri mutants. Restriction mapping and Southern blot analysis confirmed the insertion sites for eight randomly chosen mutants. This method is a very useful tool for large-scale characterization of mutants in functional genomics studies.

Introduction

Genomic technology popularized whole genome sequencing. To date, more than a hundred microbial and eukaryotic genomes have been sequenced (1,2). This expansion increases the number of genes with unknown function or genes with just putative functions that have been deposited at GenBank® (http://www.ncbi.nlm.nih.gov). Usually, a second phase in genome projects involves the functional characterization of annotated genes. This task is achieved using several methodologies, and one of the most popular is gene disruption by marker insertion (3). This methodology has the advantage of producing a large number of mutants in a few rounds of insertion (4,5). However, finding the exact site of insertion is time-consuming because it requires mapping and isolation of flanking DNA followed by sequencing. The most frequently used approach to map these insertion mutants involves whole genome subcloning into cosmids/bacterial artificial chromosomes (BACs) devoid of marker selection genes, which are screened by the presence of the marker that had been previously inserted into the genome for gene disruption. The selected cosmid/BACs are sequenced in parts to determine the insertion position (6). This methodology precludes large-scale screening of mutants. To overcome these difficulties, most laboratories first perform phenotype screening and then map the insertion region (7). However, the latter approach has the bias of only mapping disrupted genes that cause detectable phenotype-positive mutants.

In addition, two other methods are used for mapping random insertions within genomic DNA, namely thermal asymmetric interlaced PCR (tail PCR; Reference (8) and inverse PCR (9,10). They have been demonstrated as good tools but involve multiple steps, which become very expensive for large-scale analysis of mutants. Moreover, inverse PCR is difficult to apply to genomes containing over 109 bp (11).

Here we report a new methodology for mapping a randomly inserted marker into the genome of a target cell, which allows rapid large-scale screening and identification of an insertion site. It is based on touchdown PCR (12,13) using a pair of primers where one is a hybrid consensus-degenerate oligonucleotide and the other is a sequence-specific primer that only anneals to one of the strands of the inserted marker gene, with its 3′ end pointing toward the unknown genomic sequence. The resulting PCR products are sequenced using solely the specific primer. The initial portion of the obtained sequence serves as confirmation that the PCR product has started at the desired inserted marker, and then the remaining sequence corresponds to genomic sequence and reveals where in the genome of the target cell the marker has been inserted. This methodology also has other applications such as genome walking/gap closure in a genomic sequencing project. In this case, the specific primer represents the end of known sequence at the border of a gap. To test the methodology, we mapped several mutants generated by random insertion of a transposon containing a kanamycin resistance marker gene into the Xanthomonas axonopodis pv. citri (X. citri) genome.

Materials and Methods

Bacterial Strains, Vectors, and Culture Conditions

The X. citri strain 306 (14) used in this study is a virulent citrus canker type strain, which was routinely cultured in liquid or on solid Klebs-Loeffler Bacillus (KLB) medium at 28°C. For the selection of recombinants, the KLB media were supplemented with 50 µg/mL kanamycin. Escherichia coli strain DH10B (Invitrogen, Carlsbad, CA, USA) was used as a host for the plasmids and was cultured in liquid or on solid Luria-Bertani (LB) medium at 37°C. E. coli was transformed by electroporation (15), and culture media were supplemented with 50 µg/mL kanamycin.

In Vivo Transposition

X. citri randomly inserted mutants have been produced with a transposition methodology (16) using the EZ::TN™ Transposome™ Kit (EPICENTRE, Madison, WI, USA). Transposons were introduced into X. citri cells by electroporation (17) using a Gene Pulser® (Bio-Rad Laboratories, Hercules, CA, USA). X. citri mutants were selected on solid KLB media supplemented with 50 µg/mL kanamycin after 48 h of incubation at 28°C. The mutants obtained were cultured in 96-well microplates with liquid 2× trypsin-yeast (TY) extract media supplemented with 8% glycerol and stored at -80°C.

DNA Isolation

Genomic DNA from mutants was isolated with DNAzol® (Invitrogen). Mid-scale plasmid DNA isolation from E. coli was performed using 500 mL of culture and the Plasmid Maxi Kit (Qiagen, Valencia, CA, USA). Genomic and plasmid DNA samples were analyzed on a 1% agarose gel [1× TAE (Tris-acetate-EDTA)] and stained with ethidium bromide.