INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The transposition of mobile genetic elements requires specific interactions between an element-encoded protein, the transposase, and DNA sequences located at the termini of the element. The expression of the transposase and therefore transposition itself are precisely regulated. In some instances, the transposition reaction is affected by other proteins, including those encoded by the host cell (14).

Class II (Tn3-like) transposable elements, or transposons (13), exhibit replicative transposition: a two-step process that requires both a transposase and a site-specific recombinase. These transposons are also characterized by the presence of terminal inverted repeats (TIRs) that typically range in length from 38 to 40 bp (19). Transposition is initiated by the specific binding of the transposase to the TIRs, the generation of staggered nicks at the target site, and the formation of a cointegrate intermediate containing two copies of the transposon oriented in the same direction. Resolution of the cointegrate intermediate is catalyzed at a specific site by the transposon-encoded recombinase/resolvase. The site of cointegrate resolution, referred to as the internal resolution site (IRS), includes a recombinase binding site exhibiting dyad symmetry (19).

Tn5401 is a 4,837-bp Tn3-like transposable element derived from the gram-positive Bacillus thuringiensis strain EG2158 (7) and is characterized by the presence of 53-bp TIRs (3). The tnpI and tnpA genes of Tn5401, encoding the TnpI recombinase and TnpA transposase proteins, respectively, appear to be transcribed as an operon from a single promoter, designated P_R. The structural organization of Tn5401 (Fig. 1A) resembles that of B. thuringiensis transposon Tn4430, but the sequences of the two transposons are highly divergent (3, 15).

View larger version (23K): [in this window] [in a new window]   FIG. 1.   Structural organization of Tn5401 (A) and nucleotide sequences of the tnpI-tnpA promoter region (B) and Tn5401 TIRs (C). Nucleotide positions in the promoter region are based on the published sequence of Tn5401 (3). The conserved 12-bp TnpI recognition sequence ATGTCCRCTAAY is indicated by the arrows in panels B and C. The terminal 38-bp sequence with homology to Tn3-like transposons is shown in panel C. The 35 and 10 regions of the tnpI-tnpA (PR) promoter (single underline) and the divergent orf1 (PL) promoter (double underline) are shown, along with the corresponding transcriptional start sites (*). Ribosome binding sites are in boldface.

The TnpA transposase and TIRs of Tn5401 show significant sequence similarity to those of Tn3 (3). In contrast, the TnpI recombinase of Tn5401, like that of Tn4430, is related to the phage integrase family of site-specific recombinases and is not related to the Tn3 family of resolvases (1, 3, 15). In this respect, both Tn5401 and Tn4430 are unique among Tn3-like transposable elements. The IRS region of Tn5401 has been localized to a 111-bp region 5' to the tnpI gene (4). The TnpI recombinase of Tn5401 has been shown to bind to four copies of a conserved 12-bp sequence element (ATGTCCRCTAAY) located within this IRS region (Fig. 1B, elements a to d), including two that comprise a dyad. Deletion of even one of these elements abolishes TnpI-mediated recombination in vivo. This loss of site-specific recombination is correlated with reduced TnpI binding in gel mobility shift assays (4).

Transcription of tnpI and tnpA appears to be repressed by TnpI. Disruption of tnpI, but not tnpA, increased the level of mRNA transcripts originating from P_R and the divergent promoter P_L (3). Direct repeats of the 12-bp TnpI recognition sequence are positioned adjacent to the 10 and 35 regions of the promoter, P_R, that directs the transcription of tnpI and tnpA (Fig. 1B). Thus, TnpI binding to this region may prevent transcription from P_R, perhaps by denying RNA polymerase access to the promoter.

Interestingly, DNAse I footprinting experiments (4) demonstrated that TnpI also binds to a copy of its 12-bp recognition sequence located within the 53-bp TIRs of Tn5401 (Fig. 1C). The adjacent and outermost 38-bp sequences of the TIRs are similar to the 38- to 40-bp TIRs of other Tn3-like transposons which each contain the site for transposase binding. This unique juxtaposition of TnpI and TnpA binding sites, in effect generating an unusually long TIR sequence, suggests that TnpI may itself regulate the activity of TnpA.

In this paper, we confirm that TnpI serves as a negative regulator of tnpA expression and show that the TnpI protein is required for the binding of TnpA to the TIRs of Tn5401. In addition, the effects of tnpI and tnpA mutations on Tn5401 transposition activity were assessed by using a coupled transposition-conjugation assay system. The role of TnpI as both a negative and positive regulator of TnpA function is discussed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and plasmids. Escherichia coli DH5 (GIBCO BRL, Gaithersburg, Md.) was used as a host strain for molecular cloning experiments. Table 1 describes the B. thuringiensis and E. coli strains and plasmids used in this study. Strain EG10368 is a derivative of the acrystalliferous (non-crystal-producing) B. thuringiensis strain HD73-26 (9) and contains a cryptic 4.9-MDa plasmid. Strain HD73-27R is an acrystalliferous derivative of HD73 that is resistant to the antibiotic rifampin. Strain EG2243 is a transconjugant derivative of strain HD73-26 that contains a self-transmissible 50-MDa plasmid and that is resistant to the antibiotic streptomycin. The temperature-sensitive transposon vector pEG922 contains a derivative of Tn5401 carrying a tetracycline resistance gene (tet) on an ~7-kb SacI fragment. Plasmid pEG922 tnpI contains a frameshift mutation in the tnpI gene of Tn5401, while pEG922 tnpA contains a deletion within the tnpA gene (3). Plasmid pEG853 is an E. coli-B. thuringiensis shuttle vector that contains the B. thuringiensis plasmid replication origin ori60 (2). The Tn5401-tet plasmids pEG941, pEG941 tnpA, and pEG941 tnpI were constructed by isolating the ~7-kb SacI fragments containing Tn5401-tet from pEG922, pEG922 tnpA, and pEG922 tnpI, respectively, and inserting the SacI fragments into the unique SacI site of shuttle vector pEG853. Plasmids pEG941, pEG941 tnpA, and pEG941 tnpI were introduced into the B. thuringiensis host strain EG2243 to yield the B. thuringiensis strains EG11193, EG11194, and EG11195, respectively (Table 1). These streptomycin-resistant (Str^r) strains were used as donor strains in broth mating experiments designed to measure transposition frequency. Plasmids pEG941, pEG941 tnpA, and pEG941 tnpI were also introduced into the related B. thuringiensis host strain EG10368 to yield strains EG7690, EG12153, and EG12154, respectively (Table 1). These strains were used for primer extension (RNA) and Western blot (protein) analyses of tnpA expression in B. thuringiensis. The bacteriophage lambda P_R expression vector pKJB856, maintained in the DH5 derivative JMB110, was kindly provided by Kenneth Buckley (6). The DH5 recombinant strain EG7686, containing the tnpI expression vector pEG937, has been described previously (4). The DH5 recombinant strain EG7687 contains the tnpA expression vector pEG938 (see Results).

DNA manipulations and analyses. Standard recombinant DNA procedures were performed essentially as described by Sambrook et al. (18). B. thuringiensis strains were transformed with the electroporation protocol described by Mettus and Macaluso (17). Resident B. thuringiensis plasmids were resolved on 0.52% agarose gels by a modified Eckhardt lysis procedure (10). Transcription from the P_R promoter was monitored by primer extension analysis of total RNA as previously described with the oligonucleotide primer pr10: 5'-CTTCTTGAGATAAGCTAG-3' (3). DNase I footprinting analyses of TnpI- and TnpA-DNA complexes were performed as previously described (4). PCRs were performed with Taq polymerase (Perkin-Elmer Corp., Foster City, Calif., Promega Corp., Madison, Wis.) or Taq polymerase plus Taq Extender (Stratagene, La Jolla, Calif.) under standard conditions as recommended by the manufacturers. The cycling regimen was 94°C for 30 s, 46°C for 30 s, and 72°C for 1 min, for 30 cycles.

Preparation of TnpI and TnpA soluble extracts. Both the tnpI and tnpA genes were expressed in E. coli DH5 by using the expression vector pKJB856. This vector contains the bacteriophage lambda P_R promoter, a multiple cloning site, and a temperature-sensitive allele of the cI repressor gene. Transcription from P_R is repressed at 30°C and derepressed at 37 to 41°C (6). The construction of the tnpI expression vector pEG937 and expression of the tnpI gene in E. coli EG7686 have been described previously (4). To express the tnpA gene in E. coli, a 4.8-kb NsiI restriction fragment from pEG911-2 (3) containing the entire tnpA gene was ligated to pKJB856, previously cleaved with PstI. The resulting plasmid, pEG938, contains the tnpA gene positioned in the proper orientation 3' to the P_R promoter for expression. The DH5 derivative harboring plasmid pEG938 was designated EG7687. A DH5 derivative harboring the expression vector pKJB856 (DH5/pKJB856) was also constructed to serve as a control. Strains EG7686, EG7687, and DH5/pKJB856 comprise a set of isogenic strains that differ only with respect to the presence or absence of TnpI and TnpA protein production. Induced 50-ml cultures of the TnpI- and TnpA-producing strains EG7686 and EG7687 were harvested by centrifugation at 3,200 × g (maximum) in a swinging-bucket rotor for 20 min at 4°C. The cell pellets were resuspended in 2 ml of lysis buffer (25 mM HEPES-NaOH [pH 7.0], 50 mM KCl, 10% glycerol, 1 mM EDTA, 1 mM EGTA, and 20 µg of leupeptin, 2 µg of pepstatin A, 20 µg of aprotinin, and 100 µg of phenylmethylsulfonyl fluoride per ml) containing 1 mg of lysozyme per ml. After a 15-min incubation at room temperature, the cells were placed on ice for 5 min and disrupted by sonication. The cell lysates were centrifuged at 3,200 × g (maximum) for 20 min at 4°C. The supernatants were transferred to fresh tubes and centrifuged at 23,700 × g (maximum) for 20 min at 4°C. The clarified supernatants were divided into 50- to 500-µl aliquots and stored at 70°C. Control extracts containing only the expression vector pKJB856 were also prepared from the DH5 recombinant strain. Protein concentrations in soluble extracts were determined to be approximately 10 mg/ml as measured with the bicinchoninic acid protein assay system (Pierce, Rockford, Ill.). TnpI and TnpA proteins produced by induced cultures of strains EG7686 and EG7687, respectively, were detected on Western blots with anti-TnpI and anti-TnpA antibodies and standard procedures as described below.

Preparation of TnpI and TnpA antibodies. Induction of tnpI and tnpA expression in strains EG7686 and EG7687 results in the accumulation of polar inclusion bodies composed of the TnpI and TnpA proteins. Inclusions from lysed cultures were pelleted by centrifugation at 3,200 × g (maximum) in a swinging-bucket rotor for 20 min at 4°C. The inclusions were suspended in 5 ml of 50 mM Tris-HCl (pH 7.5)-100 mM NaCl-0.005% Triton X-100 and pelleted again by centrifugation. The inclusions were resuspended in 2.5 ml of 50 mM Tris-HCl (pH 7.5)-100 mM NaCl-0.005% Triton X-100 and loaded on a 55 to 78% Renografin 76 step gradient (TnpA) or a 55 to 78% Renografin 76 linear gradient (TnpI). The gradients were centrifuged at 18,000 rpm in an SW28 rotor for 16 to 20 h at 17°C. The TnpI inclusions were recovered from the bottom of the gradient, whereas the TnpA inclusions were recovered from the 55- to 78% interface. The inclusions were washed three times with 0.5 ml of water and resuspended in 0.5 ml of 0.005% Triton X-100. Rabbit polyclonal antibodies directed against TnpI and TnpA were obtained from Rockland, Inc. (Gilbertsville, Pa.), by using the purified inclusions as a source of antigen.

Detection of TnpA protein in B. thuringiensis. The EG10368 recombinant strains EG7690, EG12153, and EG12154, containing pEG941, pEG941 tnpA, and pEG941 tnpI, respectively, comprise a set of isogenic B. thuringiensis strains that differ only with respect to the mutations in tnpI and tnpA. These strains were grown in Luria broth to mid-log phase (Klett reading, 150; red filter). The cells from 1-ml aliquots were harvested by centrifugation in a microcentrifuge and frozen at 70°C. The cell pellets were thawed and suspended in 500 µl of buffer (50 mM glucose, 20 mM Tris-HCl, 10 mM EDTA [pH 8.0]) containing 4 mg of lysozyme per ml and incubated at 37°C for 30 min. Aliquots of the suspension were treated with sodium dodecyl sulfate (SDS) sample buffer and heated at 100°C for 5 min, and the solubilized proteins were resolved on an SDS-10% polyacrylamide gel. The resolved proteins were transferred to nitrocellulose filters by electrophoresis and probed with anti-TnpA antibodies by standard procedures. The filters were blocked in 5% nonfat dry milk-150 mM NaCl-10 mM Tris-HCl (pH 7.5 to 7.8) and washed with 150 mM NaCl-10 mM Tris-HCl (pH 7.8)-0.1% (wt/vol) globulin-free bovine serum albumin (BSA) containing 0.2% (vol/vol) Tween 80. Alkaline phosphatase-conjugated anti-rabbit immunoglobulin G (Sigma A-8702) was used as the secondary antibody. Antigen-antibody complexes were visualized on the filters with a nitroblue tetrazolium/BCIP (5-bromo-4-chloro-3-indolylphosphate) substrate kit obtained from Pierce (no. 34041).

DNA binding assay. PCR was used to amplify DNA fragments containing intact TIRs or truncated TIRs (TIR) of Tn5401. Primers pr14 (5'-ACGTCTATAATGCAGCTAAAGATGGG-3') and pr17 (5'-TGTGCGAATAATGTCCGC-3') were used to amplify an ~150-bp fragment containing the entire TIR sequence: 5'-GGGGTATGTGTAGCAATGGAACAGAATCACGCAACAAGCATTAGCGGACATTA-3' (TnpI recognition sequence underlined). The TIR was amplified with the opposing primers pr14 and pr16 (5'-GCTAATGCTTGTTGCGTG-3') and is missing a portion of the TnpI recognition sequence: 5'-GGGG TATG TG TAGCAATGGAACAGAATCACGCAACAAGCAT TAGC - 3'. A biotin moiety was attached to the 5' end of pr14 to permit recovery of the biotinylated fragments through binding to streptavidin-conjugated paramagnetic particles (SPP; Promega Corp.). The fragments obtained from the PCRs (100-µl volume) were purified with the QIAquik PCR purification kit (Qiagen, Inc., Santa Clarita, Calif.) and suspended in 100 µl of sterile water. Ten microliters of the amplified TIR and TIR fragments was incubated in a solution containing 20 mM Tris-HCl, 100 mM KCl, 10 mM MgCl₂, 1 mM EDTA, 5% glycerol, 100 µg of sheared calf thymus DNA per ml, and 100 µg of BSA per ml with 10 µl of the soluble TnpI and/or TnpA extracts. After a 30-min incubation at room temperature, 10 µl of the SPP (1 mg/ml in 1× phosphate-buffered saline [PBS]) was added to the binding reaction mixtures. The suspensions were incubated for 15 min at room temperature to ensure binding of the SPP to the biotinylated DNA fragments. The SPP-DNA-protein complexes were recovered from the suspensions with a magnet and washed twice with 500 µl of binding buffer and twice with BSA-free binding buffer. The SPP-DNA complexes were finally suspended in 20 µl of SDS sample buffer and heated at 100°C for 5 min, and the associated proteins were resolved by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). The recovered TnpI and TnpA proteins were detected by Western blot analysis with antisera raised against the TnpI and TnpA inclusions. The resolved proteins were transferred to nitrocellulose filters by electrophoresis and probed with antibodies by standard procedures. The filters were blocked in 5% nonfat dry milk-80 mM NaCl-50 mM Tris-HCl (pH 8.0)-2 mM CaCl₂-0.2% NP-40 and washed with PBS (150 mM NaCl, 16 mM Na₂HPO₄, 4 mM Na₂HPO₄) containing 0.2% Tween 20. Alkaline phosphatase-conjugated anti-rabbit immunoglobulin G (Sigma; A-8702) was used as the secondary antibody. Antigen-antibody complexes were visualized on the filters with a nitroblue tetrazolium/BCIP substrate kit obtained from Pierce (no. 34041).

Transposition assay. A single chloramphenicol-resistant (Cm^r) colony of EG11193, EG11194, and EG11195 was grown in 1× brain heart infusion-0.5% glycerol (BHIG) containing 3 µg of chloramphenicol per ml until late log stage to provide donor cells for conjugation. The rifampin-resistant recipient strain HD73-27R was grown in BHIG. Equal volumes (4 µl) of donor and recipient cells were mixed in 1 ml of BHIG and grown overnight at 30°C with gentle agitation. The broth matings were subsequently plated on Luria-Bertani (LB) plates containing 10 µg of tetracycline and 10 µg of rifampin per ml to determine the transfer frequency of the Tet^r marker, on LB plates containing 3 µg of chloramphenicol and 10 µg of rifampin per ml to determine the background conjugal transfer frequency of the recombinant plasmid, on LB plates containing 10 µg of rifampin per ml to determine the number of recipient cells, and on LB plates containing 100 µg of streptomycin and 10 µg of tetracycline per ml to determine the number of donor cells. Transposition frequency was expressed as the frequency of Tet^r Rif^r transconjugants (per donor or per recipient cell) minus the frequency of Cm^r Rif^r transconjugants (per donor or recipient cell).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of tnpI and tnpA mutations on transposition. With the Tn5401 transposon vector pEG922, a transposition frequency of 10² to 10⁴ in B. thuringiensis was previously determined for Tn5401-tet (3). Transposition frequencies obtained with pEG922 tnpA (Tn5401-tet tnpA) were 10- to 1,000-fold lower than those obtained with pEG922, indicating that disruption of tnpA impaired transposition. The effect of a tnpI mutation on transposition frequency could not be measured reliably because the transposon plasmid, pEG922 tnpI, shows a 40-fold-lower copy number than either pEG922 or pEG922 tnpA and exhibited instability (3). In this study, Tn5401-tet and its tnpI and tnpA derivatives were subcloned into pEG853, a shuttle vector whose copy number and stability are not affected by either the tnpI or tnpA mutation (5) and which contains a chloramphenicol resistance gene cat. The transposition frequencies of Tn5401-tet and its tnpI and tnpA derivatives, contained on the pEG853-based plasmids pEG941 (Fig. 2), pEG941 tnpI, and pEG941 tnpA, respectively, were then compared by measuring the transposition of the Tet^r marker (contained within Tn5401) to a resident 50-MDa plasmid and its subsequent transfer to strain HD73-27R in broth mating experiments (Fig. 3). This can be accomplished because (i) Tn5401-tet transposes readily to large plasmids (3) and (ii) the conjugal transfer frequency of the 50-MDa plasmid is very high (>0.1 transconjugant/recipient cell [5, 9]), whereas the transfer frequency of the pEG853-based plasmids containing Tn5401-tet is very low (10⁵ to 10⁶ transconjugant/recipient cell [5]). Thus, for Tn5401-tet on plasmid pEG941, the frequency of Tet^r transfer to HD73-27R via transposition to the 50-MDa plasmid should be much higher than the frequency of Cm^r transfer to HD73-27R via transfer of pEG941. It should be noted that the mechanism by which pEG941 and other small recombinant plasmids are able to transfer from donor to recipient is not understood. Transposition frequency was expressed as the frequency of Tet^r Rif^r transconjugants (per donor or per recipient cell) minus the frequency of Cm^r Rif^r transconjugants (per donor or per recipient cell). The results of three independent experiments are shown in Table 2. Average transposition frequencies of 1.4 × 10³ to 4.1 × 10³ were obtained for Tn5401-tet (EG11193), while average transposition frequencies of 10⁶ were obtained for Tn5401-tet tnpI (EG11195) and Tn5401-tet tnpA (EG11194). These results are in good agreement with those obtained previously by using the transposon vector pEG922 and its tnpI and tnpA derivatives (3).

View larger version (27K): [in this window] [in a new window]   FIG. 2.   Structural map of the transposon plasmid pEG941. The modified transposon Tn5401-tet is flanked by 53-bp TIRs. A tet gene fragment was inserted into the unique ClaI site of Tn5401 (Fig. 1A) to generate Tn5401-tet. tet, tetracycline resistance gene from pBC16; cat, chloramphenicol acetyltransferase gene from pC194; ori60, B. thuringiensis plasmid replication origin.

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Coupled transposition-conjugation assay. In this example, the Strr strain EG11193 serves as the donor and the Rifr strain HD73-27R serves as the recipient. Solid box, Tn5401-tet; shaded box, chloramphenicol acetyltransferase gene (cat). See the text for a description of the assay.

Effect of tnpI on tnpA expression. Previous analyses indicated that disruption of the tnpI gene on the high-copy-number transposon vector pEG922 results in increased transcription of tnpI and tnpA from P_R (3). To confirm these findings, total RNA was isolated from mid-log-phase cultures of strains EG7690 (pEG941), EG12153 (pEG941 tnpA), and EG12154 (pEG941 tnpI) and transcription from P_R was monitored by primer extension analysis (Fig. 4A). A faint mRNA transcript corresponding to initiation at P_R could be detected in EG7690 (wild-type [wt]) and EG12153 (tnpA). A more abundant transcript corresponding to initiation at P_R was detected in EG12154 (tnpI), suggesting that P_R is normally under tight regulation and that disruption of tnpI leads to increased transcription from P_R.

View larger version (45K): [in this window] [in a new window]   FIG. 4.   Effect of tnpI and tnpA mutations on tnpA expression. (A) Primer extension analysis of tnpI-tnpA mRNAs produced by mid-log-phase cultures of EG7690 (wt), EG12153 (tnpA), and EG12154 (tnpI). The primer extension product corresponds to transcripts initiated at the promoter PR (3) (Fig. 1B). (B) Western blot analysis of TnpA protein produced by mid-log-phase cultures of EG7690 (wt), EG12153 (tnpA), and EG12154 (tnpI). MW, molecular weight standards.

TnpA and TnpI binding to the TIRs of Tn5401. TnpI has been shown to bind to a 12-bp recognition sequence contained within the TIRs of Tn5401 by DNase I footprinting analysis (4). This analysis was extended to determine the conditions for TnpA binding to the TIRs. Soluble extracts were prepared from the TnpI- and TnpA-producing strains EG7686 and EG7687 (Table 1) by the procedure described for the preparation of TnpI extracts (4). As a control, a soluble extract was prepared from a DH5 recombinant strain harboring the expression vector pKJB856. Because of the isogenic nature of the strains used, the soluble extracts should differ only with respect to the presence or absence of TnpA and TnpI protein. TnpA protein was recovered from inclusion bodies produced by induced cultures of strain EG7687 (see Materials and Methods), but TnpA protein in soluble extracts could not be reliably identified by SDS-PAGE (Fig. 5A). To confirm the presence of TnpA in the EG7687 extract, proteins separated by SDS-PAGE were blotted to a nitrocellulose membrane and probed with antibodies raised against either TnpI or TnpA inclusion bodies (Fig. 5B). Although the TnpA antibodies showed a strong cross-reaction to an unidentified E. coli protein (cross-reacting material), a protein of the expected molecular mass (~100 kDa) for TnpA was detected in the EG7687 extract and not in the control or EG7686 extracts. Similarly, TnpI protein was identified in the EG7686 extract with antibodies raised against the TnpI inclusion bodies but not in the control or EG7687 extracts. These extracts (~10 mg of total protein per ml) were subsequently used as a source of TnpI and TnpA proteins for the DNA-binding studies.

View larger version (68K): [in this window] [in a new window]   FIG. 5.   Western blot analysis of TnpI- and TnpA-producing strains. Proteins from soluble extracts (100 µg of total protein) of E. coli EG7687 (lane 1), EG7686 (lane 2), and DH5(pKJB856) (lane 3) were resolved on an SDS-polyacrylamide (10%) gel and transferred to a nitrocellulose membrane. (A) SDS-polyacrylamide gel stained with Coomassie blue R-250. (B) Western blots probed with either TnpA antibodies (top panel) or TnpI antibodies (bottom panel). M, molecular mass standards in kilodaltons; CRM, cross-reacting material.

View larger version (64K): [in this window] [in a new window]   FIG. 6.   DNase I footprinting analysis of the Tn5401 TIR. A 326-bp BamHI-ScaI fragment from pEG911-3 (3), end labeled at the BamHI site, was incubated with soluble extracts (10 µg of soluble protein per ml), treated with DNase I, and purified, and the resulting DNA fragments were resolved on a 6% sequencing gel. Lanes: 1, 10 µl of control extract; 2, 10 µl of TnpI extract; 3, 10 µl of TnpI extract plus 10 µl of control extract; 4, 10 µl of TnpI extract plus 10 µl of TnpA extract; 5, 10 µl of TnpA extract plus 10 µl of control extract; 6, 10 µl of TnpA extract. The locations of the 12-bp TnpI recognition sequence and the 38-bp Tn3-like terminal repeat sequence (Fig. 1C) are indicated by the brackets. DHSS, DNase I-hypersensitive site.

View larger version (49K): [in this window] [in a new window]   FIG. 7.   DNA-binding assay. Biotinylated DNA fragments, each containing an intact TIR or TIR, were incubated with TnpA, TnpI, and/or control extracts (10- or 20-µl volumes, as indicated above the blots, containing 10 µg of soluble protein per µl). SPP were used to isolate the resulting DNA-protein complexes. Bound proteins were analyzed on Western blots with TnpI antibodies (top panel) or TnpA antibodies (bottom panel). For each blot, the first lane contains the soluble TnpI or TnpA extract as a standard.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Tn5401 transposition appears to be mechanistically similar to that of Tn3 (19). Cointegrate molecules formed as intermediates during the transposition process by a Tn3-like transposase are resolved by the site-specific recombinase TnpI (3, 4). TnpI also regulates Tn5401 transposition by serving as a repressor of tnpI and tnpA transcription (reference 3 and this study), an apparent consequence of TnpI binding to the promoter regions of tnpI and tnpA. In this study, we present evidence that TnpI regulates Tn5401 transposition at the posttranslational level by modulating the interaction of TnpA transposase with the TIRs.

The observation that the unusually long 53-bp TIRs of Tn5401 contain adjacent recognition sites for TnpA and TnpI led to the hypothesis that TnpI regulates the enzymatic or DNA-binding activity of TnpA (4). The close proximity of the binding sites suggests that the DNA-bound proteins could physically interact. Two different methods were employed to examine DNA-protein interactions. First, a conventional DNase I footprinting analysis was used to detect interactions between TnpA, TnpI, and the TIR of Tn5401. Consistent with previous studies (4), TnpI binding to the TIR results in a footprint that extends beyond its 12-bp recognition sequence (Fig. 6). Under the conditions used in our assays, the addition of TnpA protein alone afforded little, if any, protection from DNase I cleavage. A mixture of TnpI and TnpA, however, yielded an extended footprint that included and extended beyond the entire TIR sequence. The second method employed SPP to capture TnpI and TnpA proteins bound to biotinylated DNA fragments containing an intact or truncated TIR. In this instance, TnpA capture was observed only in the presence of TnpI and an intact TIR sequence.

While a number of explanations could account for these phenomena, we propose the most likely hypothesis: that TnpI promotes the binding of TnpA to the TIR. The apparent inability of TnpA to bind TIR in the presence of TnpI suggests that TnpI must bind specifically to the TIR in order to facilitate TnpA binding. It is possible that TnpA does not bind DNA and that TnpI alone contributes the DNA-binding function required for localization of a TnpI-TnpA complex at the TIR. Thus, the extended footprint shown in Fig. 6 (lane 4) could be due to further exclusion of DNase I from the TIR sequence by a larger TnpI-TnpA dimer or multimer. Similarly, the apparent binding of TnpA to the TIR shown in Fig. 7 could be the indirect effect of TnpA binding exclusively to TnpI. We think this is unlikely because (i) a derivative of Tn5401 containing a frameshift mutation in tnpI (Tn5401-tet tnpI) can generate cointegrate molecules at a low frequency in B. thuringiensis (3), indicating that TnpA must exhibit some undetectable binding to the TIRs in the absence of TnpI and (ii) the nonspecific binding of TnpI to the TIR did not result in the recovery of TnpA (Fig. 7), as would be expected if TnpA were merely binding to TnpI.

Provided TnpI is essential for the binding of TnpA to the TIR, mutations in tnpI should effectively block Tn5401-tet cointegrate formation. A coupled transposition-conjugation assay system was developed to compare the transposition frequencies for Tn5401-tet, Tn5401-tet tnpI, and Tn5401-tet tnpA in B. thuringiensis. Mutations in either tnpA or tnpI reduced the transfer of the transposon-linked Tet^r marker to the resident 50-MDa plasmid of B. thuringiensis EG2243 by 500- to 2,500-fold. Since B. thuringiensis strains carrying Tn5401-tet tnpI exhibit elevated levels of tnpA mRNA and TnpA protein, the reduction in cointegrate formation (Tet^r transfer) cannot be due to inadequate levels of TnpA but to the absence of TnpI.

Thus, the regulation of Tn5401 transposition in B. thuringiensis is controlled at both the transcriptional and posttranslational levels by the TnpI recombinase. The following scenario can be proposed. Below a critical concentration of TnpI, transcription from P_R may proceed, leading to the production of TnpI and TnpA protein. The subsequent increase in TnpI concentration inhibits transcription from P_R but allows TnpA to bind the TIRs and proceed with the formation of cointegrate intermediates that are subsequently resolved by TnpI. One outcome of this regulatory arrangement may be that cointegrate molecules will not be formed unless there is sufficient TnpI protein available to complete the transposition process. This is reflected in the low transposition frequencies observed for Tn5401-tet tnpI (reference 3 and this study).

There are reports that the function of transposases encoded by mobile genetic elements can be modulated by other proteins (14). For instance, integration host factor (IHF) has been shown to facilitate the binding of transposase to the termini of the Tn3-like transposon (21). IHF has also been shown to play a role in the transposition of IS1 (8), IS50 and Tn5 (16), IS10 (12, 20), and Tn4652 (11). Horak and Kivisaar (11) suggest that IHF binding to the termini of Tn4652, in addition to activating the transcription of tnpA, may also modulate the binding or catalytic activity of the TnpA transposase. The involvement of a transposon-encoded recombinase to modulate the DNA-binding activity of its associated transposase, as for Tn5401, has not been reported. The most closely related transposon, Tn4430, encodes a TnpI-like recombinase but does not have TnpI binding sites closely associated with the TIRs (15), suggesting that the role of TnpI in regulating Tn5401 transposition is unique.

The role of TnpI in regulating the DNA-binding activity of TnpA provides a logical checkpoint in the regulation of Tn5401 transposition but does not exclude other roles for TnpI in modulating, for instance, the catalytic activity of TnpA. It would be interesting to explore whether other class II transposons show a similar coordination in the activities of their recombinase and transposase proteins.