The IntI1 integron integrase preferentially binds single-stranded DNA of the attC site

Abstract

IntI1 integrase is a member of the prokaryotic DNA integrase superfamily. It is responsible for mobility of antibiotic resistance cassettes found in integrons. IntI1 protein, as well as IntI1-COOH, a truncated form containing its carboxy-terminal domain, has been purified. Electrophoretic mobility shift assays were carried out to study the ability of IntI1 to bind the integrase primary target sites attI and aadA1 attC. When using double-stranded DNA as a substrate, we observed IntI1 binding to attI but not to attC. IntI1-COOH did not bind either attI or attC, indicating that the N-terminal domain of IntI1 was required for binding to double-stranded attI. On the other hand, when we used single-stranded (ss) DNA substrates, IntI1 bound strongly and specifically to ss attC DNA. Binding was strand specific, since only the bottom DNA strand was bound. Protein IntI1-COOH bound ss attC as well as did the complete integrase, indicating that the ability of the protein to bind ss aadA1 attC was contained in the region between amino acids 109 and 337 of IntI1. Binding to ss attI DNA by the integrase, but not by IntI1-COOH, was also observed and was specific for the attI bottom strand, indicating similar capabilities of IntI1 for binding attI DNA in either double-stranded or ss conformation. Footprinting analysis showed that IntI1 protected at least 40 bases of aadA1 attC against DNase I attack. The protected sequence contained two of the four previously proposed IntI1 DNA binding sites, including the crossover site. Preferential ssDNA binding can be a significant activity of IntI1 integrase, which suggests the utilization of extruded cruciforms in the reaction mechanisms leading to cassette excision and integration.

FIG. 1

(A) Simplified diagram of the type I integron in Tn21 to illustrate integron organization and location of the att sites. (B) Nucleotide sequence of the IntI1 primary sites (attI and attC) in plasmids pSU18R2 and pSU18R3. The DNA sequences show the GTTAG crossover site at the right (3′) end. This orientation defines the top and bottom nomenclature for the ssDNAs. Filled vertical arrows indicate the crossover sites, and open horizontal arrows point to the four putative pentanucleotide components (boxed and numbered) of each att site. The numbers below the bottom strand of the aadA1 attC site start at the beginning of the reverse M13 priming site of the pSU18 vector sequence and correspond to numbers in Fig. 6. The shaded area indicates the region protected by IntI1 from DNase I attack.

FIG. 2

Western blot analysis of the IntI1 protein purified fractions. (A) Coomassie blue-stained polyacrylamide-SDS gel. (B) Western blot of a similar gel immunoprobed with the anti-IntI1 antibody. Lanes 1, control strain extract; lanes 2, purified IntI1-COOH protein; lanes 3, purified IntI1 protein.

FIG. 3

Gel mobility shift assays with dsDNA fragments containing aadA1 attC (left) or the attI site (right). Lanes 1, control dsDNA; lanes 2, dsDNA plus E. coli C41 control extract; lanes 3, dsDNA plus pure IntI1-COOH; lanes 4, dsDNA plus purified IntI1. F, free dsDNA; B, DNA-protein complex.

FIG. 4

Gel retardation assays with ssDNA fragments containing either the top or the bottom strands of the aadA1 attC site (left) or the attI site (right). The substrate DNA used is indicated at the top of the figure. Protein extracts in each lane are as follows: left panel, lane 1, control DNA; lane 2, E. coli C41 control extract; lane 3, IntI1-COOH; lane 4, IntI1; lane 5 E. coli C41 control extract; lane 6, IntI1-COOH; lane 7, IntI1; lane 8, control DNA; right panel, lane 1, DNA alone; lane 2, purified IntI1; lane 3, IntI1-COOH; lane 4, E. coli C41 control extract; lane 5, IntI1; lane 6, IntI1-COOH; lane 7, E. coli C41 control extract; lane 8, control DNA. F, free DNA.

FIG. 5

Competition of IntI1 binding to the bottom strand of the aadA1 attC site and supershift experiments with anti-IntI1 antibody. Lane 1, control DNA; lane 2, IntI1 control binding reaction; lanes 3 to 6, competition by the unlabeled attC bottom strand; addition of 2-, 10-, 20-, and 40-fold excess unlabeled bottom-strand aadA1 attC, respectively; lanes 7 and 8, competition by attC top strand; addition of an identical (lane 7) or twofold (lane 8) excess amount of the top attC aadA1 strand; lanes 9 and 10, heterologous DNA competition; addition of an identical (lane 9) or twofold (lane 10) excess amount of the unrelated competitor (an oligonucleotide mix); lanes 11 to 13, supershift assays with specific antisera; lane 11, incubation of IntI1 with polyclonal anti-IntI1 antibody before binding reaction; lane 12, incubation with polyclonal anti-IntI1 antibody after binding reaction; lane 13, incubation with preimmune serum after binding reaction. ds, dsDNA; ss, ssDNA; I and II, different protein-DNA complexes.

FIG. 6

DNase I footprinting analysis of the aadA1 attC bottom strand. The ssDNA was labeled at the 5′ end, incubated with purified IntI1 protein, subjected to limited cleavage by DNase I, and electrophoresed on a 6% sequencing gel. Lane 1, DNase I cleavage pattern of the DNA alone; lane 2, binding reaction without DNase I treatment; lane 3, DNase I cleavage pattern of the DNA incubated with 0.1 μg of IntI1; lane 4, DNase I cleavage pattern of the DNA incubated with 0.5 μg of IntI1. Nucleotide numbering starts from the 5′ end of the labeled fragment, as indicated in Fig. 1. Locations of the pentanucleotides previously proposed as IntI1 binding sites are shown by vertical arrows and numbered as in Fig. 1.