Tn1549 Transposition and the Tyrosine Recombinase Chemistry
(A) Schematics of Tn1549 transposon movement. Tn1549 encodes genes responsible for transposition (including Int), conjugation and mobilization (light gray boxes) and confers vancomycin resistance (VanB^R, pink). Excision (i) in the donor cell (purple) creates the CI, with IR_L and IR_R (block arrows) joined by 5–7 nt heteroduplex; the donor DNA site is resealed. A single CI strand is transferred (ii) to a recipient bacterium (blue) by conjugation and replication re-creates the double-stranded CI, now with homoduplex at the crossover region. Integration (iii) in the recipient genome generates a new vancomycin resistant cell.
(B) Tyrosine recombinases (gray ovals) recombine DNA substrates stepwise: (i) two proteins cleave one strand of each dsDNA using a tyrosine nucleophile, creating a covalent 3′-phosphotyrosyl bond (black dots) and a free 5′-hydroxyl (5′OH) group; (ii) the 5′ ends swap places and resolve the protein-DNA link on the partner strands, generating a four-way Holliday junction (HJ) intermediate; (iii) isomerization activates the second protein pair to cleavage (iv) and exchange (v) the other DNA strands, generating the final recombined products. Energetically inexpensive strand exchange requires homology in the substrates (red).

Transposon End Recognition by the Tn1549 Int
(A) Crystal structure of the R225K Int^82N-CI5 DNA complex. Two Int molecules (A, B, two shades of blue) bind at IR_L and IR_R (gray), with the crossover region (red) at the center. Bases are numbered from the IR boundaries, numbers increase with the distance, negative numbers upstream, bottom strand marked with prime. CB and CAT are connected by a flexible linker. αA-D, αI, αJ, and the β-hairpin insertion (pink) interact with the DNA; αM is swapped between the two subunits.
(B) Close-up of the base-specific DNA contacts of N150 with three terminal bases (sticks in atomic coloring).
(C) Close-up of the β-hairpin. Q249, R252, and T254 (sticks) contact the DNA phosphate backbone.
See also and and and .

Reconstitution of the Int^82N-CI5 Complex and Its Structural Features, Related to
(A) Domain composition of Int. The N-terminal Arm-Binding domain (AB) (gray box) is followed by the Core Binding domain (CB) (αA-D, light blue boxes) and the Catalytic domain (CAT, consisting of αE-M and β1-4, dark blue boxes and arrows, respectively). The linker between CD and CAT is shown in black and the β-hairpin insertion is indicated as a gray box with dark purple arrows (β2-β3) inside. The catalytic residues are marked with black ovals and labeled.
(B) In vitro reconstitution of Int activity. Left: DNA cleavage assays with Int^82N (aa 82-397) and Int^FL (aa 1-397) showing that both constructs cleave ‘suicide’ CI5 DNA. The substrates contain a nick in the DNA backbone downstream of the expected cleavage position (see schematics). Upon cleavage, 2 nucleotides diffuse away, trapping the covalent protein-DNA intermediate (gray oval attached to DNA) that can be resolved from unmodified Int on SDS-PAGE. Reactions with (+) or without (-) DNA are analyzed on a 12% SDS-PAGE gel. Relevant gel parts are cropped and shown next to each other. Asterisks indicate the band corresponding to the covalent protein-DNA product. Right: Results of DNA strand exchange assays with Int^82N and Int^FL show that both constructs create strand exchange products in vitro. Strand exchange reactions using 5′-³²P radiolabeled (star) suicide CI5 DNA substrates (gray in the schematic) run on a denaturing 12% TBE-Urea PAGE gel. Following DNA cleavage, nucleophilic attack by the 5′OH of an unlabeled partner substrate (black) leads to a recombined product (black/gray bar). Ligation of the labeled substrate strand (20 nt) with unlabeled DNA results in a larger product (43 nt, marked with asterisk). The DNA band below the substrate corresponds to the cleavage product (18 nt). WT: wild-type protein, R225K: catalytic mutant, (-) no protein in the reaction.
(C) Analytical size exclusion chromatography (SEC) of Int^82N with CI5 DNA at different protein:DNA ratios (on a Superdex 200 PC 3.2/30 column, GE Healthcare) reveals homogeneous complex at 2:1 ratio. Elution profiles for free CI DNA (peak at 1.50 mL), Int^82N alone (1.51 mL), complex at ratio 1 Int^82N: 1 DNA (1.41 mL and 1.55 mL), and complex at ratio 2 Int^82N: 1 DNA (1.41 mL) are compared.
(D) Close-up of the active site in the Int^82N-CI5 complex shows good quality experimental electron density. Simulated annealed composite omit map (2Fo-Fc, contoured at 1σ level) is shown for the catalytic residues (in sticks and atomic coloring) and DNA.
(E) Structural superposition of Tn1549 Int, Cre and λInt. One subunit is shown in cartoon representation (with helices shown as cylinders) for Int^82N (blue), λInt^75N (dark cyan, PDB: 1z19, aa 75-356) and Cre recombinase (light cyan, PDB: 1q3u). The overall fold is conserved among the three proteins, only the β-hairpin insertion is unique to Tn1549 Int.
(F) Comparison of the dimeric Int^82N-CI5 complex (blue) with the tetrameric pre-cleavage λInt^75N-COC’ complex (PDB: 1z19, cyan and gray). In the Int^82N-CI5 complex, the C-terminal alpha helix is reciprocally exchanged between the two protein subunits, whereas the λInt^75N-COC’ structure shows a cyclic exchange in homo-tetrameric assemblies.

Structural Characteristics of Transposon End Binding, Related to
(A) Schematic view of the protein-DNA contacts in the Int^82N-CI5 crystal structure.
The bases at the IRs are labeled in uppercase and the crossover region in lower case letters. IR_L is bound by Molecule A and IR_R by Molecule B (as indicated at the bottom). Blue arrows mark hydrogen bonding with the phosphate backbone and orange polygons indicate hydrophobic contacts. N150, the only residue performing base-specific contacts, is highlighted in blue with the interacting residues marked with blue squares. The residues involved in base flipping are shown in pink (R153) and orange (Y160).
(B) Surface representation for the Int^82N subunit bound to IR_L DNA colored by electrostatic surface potential (blue, positive; red, negative; −5 to +5 kT/e isosurfaces) calculated using APBS in Pymol (Schrödinger). N- and C- termini are marked. Int^82N interacts extensively with DNA covering 8170 Å² buried surface area at IR_L and 7810 Å² at IR_R (calculated by PISA).
(C) Graphical representation of roll (gray) and twist (blue) parameters for each local base pair step in CI5 DNA. Pink arrows in the plot indicate the last step in the A-tracts, with a marked decrease in the roll angle. The final A-s (A8’ and A8 in IR_L and IR_R, respectively) have C3′ endo sugar pucker, a non-canonical backbone conformation in normal dsDNA that is frequently associated with DNA bending (). The crossover region (covered with a gray bar) was omitted from the calculation due to the flipped bases. The DNA sequence is shown under the plot, with pink boxes indicating the steps with a decreased roll angle. See also .
(D) Phylogenetic tree of Tn916-like family transposases shows three different subgroups: Tn1549, Tn916 (both with two consecutive tyrosines at the catalytic pocket, YY), and a third clade with the first tyrosine substituted by a tryptophan (called WY). XY marks a distinct related clade with only one conserved tyrosine. The red circles outside the tree indicate the number of distinct genomes in which the specific sequence is found. The largest circle marks sequences present in more than 100 genomes. Among these, Tn916 and Tn1549 Int-s are the most abundant, found in 3412 and 520 genomes, respectively.
(E) Comparative sequence analysis reveals a high level of conservation for the β-hairpin (left) and for the residues involved in base flipping (R153 and Y160) among Tn916-like CTns, but not in the XY clade (right). N150, which recognizes the IR ends, is also highly conserved in the Tn916-like family.
(F) Sequence alignment of Tn916-like Y-transposases and more distantly related tyrosine recombinases. The β-hairpin is present in Tn916-like Y-transposases, but absent in other tyrosine recombinases. The IntΔβ construct was created by replacing the β-hairpin insertion (H248-P263, pink frame) with a flexible two amino acid linker (GG).
(G) Comparison of DNA binding by Int^82N and IntΔβ. Electrophoretic Mobility Shift Assay (EMSA) was performed with constant concentration of radiolabeled CI5 DNA (1 μM) and increasing concentrations of Int proteins (as indicated above the gel). Complexes are run on a native gel (TBE 4%–12% polyacrylamide gel). Schematics on the side mark the putative composition of each band. At 1.5 μM of protein, the amount of free DNA is higher for IntΔβ than for Int^82N, indicating some decrease in DNA affinity upon deletion of the β-hairpin.
(H) Results of in vitro cleavage (left) and strand exchange (right) assays with suicide CI5 DNA substrates demonstrating that the β-hairpin insertion is important for Int activity. See schematics in B for the assay design. Cleavage assays are analyzed on SDS-PAGE, separating free Int and the covalent Int-DNA intermediate (asterisk). Strand exchange assays with radiolabeled suicide CI5 DNA substrates are analyzed on a denaturing TBE-Urea PAGE gel. Star denotes 5′-³²P. Ligation of the labeled substrate strand (gray, 20 nt) with unlabeled DNA (black) results in a larger product (gray/black, 43 nt). The DNA band below the substrate corresponds to the cleavage product (18 nt). Int^FL and Int^82N readily catalyze CI DNA cleavage and strand exchange, but IntΔβ is compromised. Catalytic mutant R225K Int^82N is shown as negative control.

Int Triggers DNA Base Flipping, Unwinding, and Melting
(A) Close-up of the crossover (red) in the Int^82N-CI5 DNA complex. R153 invades the DNA after each IR, stacks with A5′/A5 and flips out t4′/c4 that are stabilized by Y160 (sticks).
(B) 2AP fluorescence spectroscopy monitoring base flipping upon Int binding. Top: fluorescence emission spectra of Int^82N-CI5 complexes (2:1 ratio) with 2AP in the IR (IR), in the crossover region (Co), and at the flipped base (F). Bottom: fluorescent signal of CI5-F with different amounts of Int^82N (gray) and R153A mutant (blue) at λem = 371 nm. Data are represented as mean ± SEM ().
(C) Base flipping is required for efficient strand exchange. Results of strand exchange assays with Int^82N and base-flipping mutants (see also ). Cleavage of the radiolabeled CI5 DNA (gray) and strand exchange with an unlabeled partner substrate (black) leads to a recombined product (gray/black) that is detected on denaturing PAGE.
(D) Base flipping facilitates Tn1549 excision in vivo. Excision of mini-Tn1549 under Int^FL and Xis expression in E. coli is followed by PCR, monitoring formation of the CI and its loss from the donor plasmid (DP, RDP) (). Agarose gel showing PCR products from samples taken at 0 and 3 hr of Xis/Int^FL expression. R153A and R153A-Y160A mutants show a strong reduction of excision. Superfluous lanes were removed. Controls: R225K inactive mutant or no protein expressed (−).
(E) Superposition of the R225K Int^82N-CI6a (yellow) and Int^82N-CI5 (blue) complex crystal structures (root-mean-square deviation [RMSD] for Cα atoms 0.54 Å) shows similar DNA distortions and a slight movement of Molecule B. The crossover DNA (red) has a high disorder in the electron density map of CI6a (composite omit map at 1σ) and some bases could not be located (inset; missing bases in brackets).
See also , , and .

Int-Mediated Melting of the Crossover Region in Diverse CI DNA Substrates, Related to
(A) Close-up of the crossover region (red) showing how DNA distortions destabilize base-pairing interactions. Several bases (a0-a3′-t1 and g3/g0’) form interdigitating base stacking instead of regular Watson-Crick base pairs. Simulated annealed composite omit map (2Fo-Fc, contoured at 1σ level) shows reduced quality electron density map at this region.
(B) DNA binding assay with 2-aminopurine (2AP) modified CI5 DNA variants (CI5-F: 2AP at the flipped base, CI5-IR: 1nt inside IR and CI5-Co: 1nt inside the crossover region). Int^82N binds to all 2AP-modified DNA variants.
(C) DNA binding assays with Int^82N, R153A and R153A-Y160A mutants. Complexes with a constant concentration of radiolabeled CI5 DNA and different protein concentrations (0.75, 1.5 and 3 μM) are run on a native gel (TBE 4%–12% polyacrylamide gel). All mutants form complexes with DNA.
(D) Int^82N, R153A and R153A-Y160A mutants show similar DNA cleavage activity on suicide CI5 DNA. Gray circle attached to DNA marks the band (asterisk) corresponding to the covalent cleavage product on a 12% SDS-PAGE gel. See schematics in B.
(E) Tn1549 CI sequences used in this study. IR_L and IR_R of Tn1549 are highlighted in gray. Bases are numbered as described in A. The crossover regions differ between the three sequences and are colored in red with lowercase letters. All CI sequences were derived from in vivo studies - CI5 from and the CI6 sequences from .
(F) Int^82N binding to different CI DNA sequences. Complexes show similar band shift, forming stable complexes with 2:1 protein:DNA ratio on a native gel (TBE 4%–12% PAGE gel).
(G) Int^82N can cleave CI DNA independently of the crossover region sequence and length. DNA cleavage reactions with suicide CI DNA are analyzed on a 12% SDS-PAGE gel. Relevant gel segments are shown side by side. The covalent intermediate product is marked on the side.
(H) Confirmation of base flipping in different CI DNA sequences. Fluorescence emission spectra of Int^82N-CI5 DNA and Int^82N-CI6b DNA complexes with 2AP modification at the flipped-out base are compared. y axis shows the corrected fluorescence intensity using an excitation wavelength (λex) of 320 nm. Complexes were prepared at 2 Int^82N:1 DNA molar ratio. The fluorescence spectrum of the Int^82N-CI5 complex with 2AP inside the IR is shown as control (see B).

Structural Insights into the Recognition of Different Crossover Sequences and Lengths in Diverse Transposon CIs, Related to
(A) Superposition of the Int^82N-CI6b complex (cyan) and the Int^82N-CI6a complex (yellow), with Molecule A as a reference, shows high similarity (RMSD 0.34 Å for Cα). Both crossover regions show high disorder and the central bases could not be built. The DNA appears to form interdigitating base stacking as in the Int^82N-CI5 complex, although its path could not be fully resolved.
(B) Comparative schematic representation of the Int^82N-DNA contacts in the CI5, CI6a and CI6b structures. The DNA recognition pattern is conserved independent of the sequence and length of the crossover region. The CI6 IR_L sequence is shown. IR_R differs in the base highlighted in white. Blue arrows indicate hydrogen bonding with the phosphate backbone (contact cut-off 3.5Å). Base-specific DNA contacts of N150 are highlighted in blue with the interacting bases marked. Interactions that are present in some but not all complexes are marked with superscript: The interaction with R252 is shifted one nucleotide downstream in the CI6a structure; T147 forms hydrophobic contacts in CI5; N101 makes a water-mediated contact in CI6a; K188 contacts the T-3 phosphate in IR_R in CI6a and in both IRs in CI6b; R95 contacts (#) are absent at IR_R.
(C) Close-up of the β-hairpin insertion shows two alternative conformations for the final small loop (E262-T265, sticks) in the electron density maps in the CI6 structures (shown for CI6a). Simulated annealed composite omit map (2Fo-Fc, contoured at 1σ level) is shown for the loop.
(D) Simulated annealed composite omit electron density maps reveal higher disorder in the CI6 structures. The map is shown at 1σ level, together with a cartoon representation for the three CI DNA molecules: CI5 (gray), CI6a (yellow) and CI6b (cyan). Bases at the crossover region are shown in atomic representation in red. Missing bases are indicated with gray letters in red brackets.

Int Cleaves inside the Conserved IRs
(A) Close-up of the R225K Int^82N-CI5 active site. All catalytic residues (sticks) assemble around the phosphate group of T-1. Y379 (dark blue) points toward the phosphate (4.9–5.4 Å distance, dashes); Y380 (gray) is anchored in a conserved hydrophobic pocket on the subunit interface, hydrogen bonding with M359 at the base of this pocket (black dashes) and with L377 and K362 in the backbone.
(B) Sequence LOGO of Tn916-like CTn transposases (top; see D for phylogenetic tree) shows that both tyrosines are conserved, but Y379 can be replaced with a tryptophan (W). Canonical tyrosine recombinases only have one tyrosine as seen in the structural alignment (bottom).
(C) PTO mapping at IR_R (left) or IR_L (right) in ligation or strand exchange assays, respectively. PTO at T-1/T-1′ inside the IRs blocks DNA cleavage and strand exchange by Int^82N. Control: no PTO; (−): no protein.
(D) Strand ligation assays show that a terminal thymine at the incoming 5′OH is needed for strand exchange. 5′-radiolabeled IR_L (yellow, left) or IR_R (orange, right) half-site substrate is cleaved and ligated to an unlabeled partner substrate. Reactions with Int^82N and partner substrates with different base at the 5′OH (T/C/A/G) analyzed on denaturing PAGE. T^P: 5′-phosphorylated DNA; (−) no protein.
See also .

Characterization of the Int Active Site and Its DNA Cleavage and Strand Exchange Specificity, Related to
(A) Close-up of the active site in the wild-type Int^82N-CI5 complex structure shows the same architecture as in the R225K mutant. Simulated annealed 2Fo-Fc composite omit map contoured at 1σ level. As before, Y379 points toward the DNA at the phosphate of T-1/T-1’ (at 5.2-5.8 Å distance). R225 forms a hydrogen bond with the scissile phosphate (T-1/T-1’). The structures superpose with RMSD of 0.5020 Å for Cα atoms in Molecule A.
(B) In vitro activity assays with tyrosine mutants Y379F and Y380F, and the double mutant 2YF with both tyrosines mutated to phenylalanine. DNA cleavage reactions (left) with suicide CI5 DNA are analyzed on 12% SDS-PAGE gel. Asterisk marks the covalent cleavage product. Both Y379F and Y380F mutants cleave DNA, whereas the double mutant is inactive. Strand exchange assays (right) with suicide CI5 DNA are run on denaturing 12% TBE-Urea gel. Irrelevant gel segments were eliminated. The recombination product is indicated with an asterisk. Here, the two single mutants show different behavior. (-): control lane without protein in the reaction.
(C) Phosphorothioate (PTO) modification at position -1 inside the transposon blocks DNA cleavage and strand exchange in CI6a DNA. PTO modification is introduced at T-1/T-1’, at the neighboring position a0/g0’, or both (-1’,0’ and -1,0) (as for CI5 DNA in C). Reactions are run on denaturing 12% TBE-Urea gel. Substrates and products are marked with schematics as in H and C. (-): reaction without protein; control: reaction without PTO modification.
(D) Ligation assays with truncated partner DNA confirm the need for cleavage 5′ of T-1/T-1’. Schematics show the experimental design with IR_R labeled. Two unlabeled IR_L partner substrates were tested: one with T-1’ and the other without (marked with red box). Reactions are run on denaturing 12% TBE-Urea PAGE gel with radiolabeled IR_R (on the left) and IR_L (on the right). Products are only observed with T-1/T-1’ present. (-) reactions without protein. T^P indicates reactions with the 5′ phosphorylated partner substrates.
(E) Thymine at the 5’ is also necessary for strand exchange with CI6a DNA. Ligation reactions are visualized on denaturing 12% TBE-Urea PAGE gel. Reactions with radiolabeled IR_L (left) and IR_R (right) are shown.
(F) A single stranded DNA (ssDNA) competes with Int-mediated recombination. Int^82N at 40 μM concentration was incubated with 0.5 μM CI5 DNA and increasing concentrations (0.2-1.7 μM, as marked above the gel) of unlabeled ssDNA that mimic the crossover region sequence (5nt length for cross5 and 10nt for cross10, both ending with the essential thymine at the 5′). The CI5 DNA substrate has a radiolabeled 20nt strand (short gray bar), that upon ligation with its partner substrate creates a 43nt recombination product (black/gray bar). Addition of cross5 or cross10 ssDNA blocks correct strand exchange and results in alternative products of 23 nt or 28 nt, respectively (white/gray bar). Controls: 5′P, 5′-phosphorylated DNA; -Int^82N, no protein in the reaction.

Int Resolves the HJ Intermediate
Schematics of the HJ and its resolution. Results of HJ resolution assays with IR_L and/or T_R arms radiolabeled (^∗) on denaturing PAGE gel. Int^82N -mediated cleavage and strand exchange (at the arrowheads) leads to the final integration products (IR_L-T_L, 44 nt and T_R-IR_R, 47 nt).
See also .

Characterization of Int’s HJ Resolution Activity, Related to
(A) Int^82N binds to the synthetic HJ, forming stable complexes at 4:1 ratio. HJ substrates were radioactively labeled at the IR_L or T_R arm or both (see schematics in ). Complexes were prepared at a constant concentration of radiolabeled HJ (1 μM) and different protein concentrations (2 and 4 μM), and run on a native PAGE gel (6% TBE gel). 44bp dsDNA is shown as a control and migrates faster than the HJ.
(B) Int^82N resolves HJ intermediates toward recombined products. Native gel with the purified radiolabeled DNA products of the HJ resolution assay shown in . Reactions with (+) and without (-) Int^82N confirm that Int^82N has resolved the HJ to dsDNA.
(C) Int^82N can resolve HJ intermediates back to substrates. HJ was radioactively labeled on the IR_R arm, the T_L arm or both as indicated (stars). Solid arrowheads mark the cleavage positions. The sizes of the individual arms, as well as the lengths of the substrates (44 nt and 47 nt) and recombinant products (42 nt and 49 nt) are indicated in the schematics. Reactions with (+) or without (-) Int^82N are visualized on a 12% TBE-Urea gel. Upon Int^82N activity, the sizes of the labeled strands are changed to the expected product sizes. Size markers (49, 47, 44 and 42 nt) are shown on the left.
(D) Native gel with the purified DNA products from C confirms HJ resolution to dsDNA substrates.

The Role of Int’s C-Terminal Helix
(A) The Int^82N dimer interface. Molecule B is shown as surface (gray); residues interacting with the C-terminal helix of Molecule A (blue) are colored: orange, hydrophobic; red, hydrophilic interaction (see ). C-terminal truncations are marked.
(B) 390C forms higher oligomers with CI5 DNA on EMSA. Schemes indicate putative complex stoichiometries.
(C) 390C is hyperactive in strand exchange. Superfluous lanes were removed.
(D) Synthetic peptide antagonizes Int activity. Strand exchange reactions with different peptide amounts (at constant 33 μM Int^82N and 1 μM ³²P-labeled CI5 DNA). Cleavage and recombination products decrease gradually.
See also .

Int’s C-Terminal Helix Mediates Protein Oligomerization and Catalytic Activity, Providing a Target for Inhibitor Design, Related to
(A) Conservation of the C-terminal helix αM and its binding site in Tn916-like CTn transposases. Top: Surface residues interacting with αM in the Int^82N dimer are colored by conservation (exact % identity values are listed in ). Bottom: Sequence LOGO for αM.
(B) Size exclusion chromatography (SEC, top) and small angle X-ray scattering (SAXS, bottom) reveal dimers for Int^82N and monomers for C-terminally truncated variants. In SEC, the elution profiles for Int^82N (black line), 390C (gray line), 384C (green dots) and 381C (green dashes) are compared on Superdex 200 10/300 column. C-terminally truncated variants show an elution volume corresponding to a molecular weight of ∼36 kDa, whereas Int^82N elutes as a dimer at ∼72 kDa. For SAXS, pairwise distance distribution function p(r) is plotted. Maximal particle dimensions (D_max) are 95Å for 390C and 130Å for Int^82N. The sharp drop in D_max determination indicates conformational flexibility (especially for the monomeric 390C construct; see for details), likely reflecting free movement of the two Int domains in absence of DNA.
(C) C-terminally truncated Int variants form higher oligomers upon DNA binding. Complexes were prepared at a constant concentration of radiolabeled CI5 DNA with different concentrations of 381C and 390C and run on a native gel (TBE 4%–20% PAGE gel). The presumed protein-DNA composition is shown schematically for each band.
(D) Truncated variants are hyperactive in DNA cleavage. Reactions with suicide CI5 DNA are analyzed on a 12% SDS-PAGE gel. Asterisk marks the covalent cleavage product.
(E) 390C and 384C variants are hyperactive in strand exchange under low protein concentration. Strand exchange reactions with Int^82N, 390C, 384C and 381C constructs are run on denaturing 12% TBE-Urea PAGE gel. The recombined product is indicated with a black/gray bar. The 381C construct shows a great reduction in strand exchange activity.
(F) Strand exchange assays in the presence of a scrambled peptide do not show a dose dependent inhibition as seen for the native peptide (D). Reactions with different peptide concentrations (as marked above the gel) at a constant concentration of protein (33 μM), radiolabeled suicide CI5 DNA (1 μM) and unlabeled DNA (75 μM) are visualized on denaturing 12% TBE-Urea gel. Products are marked with schematics. ‘-Int^82N’, reaction without protein.
(G) The C-terminal tail mimicking peptide affects Int^82N-CI5 complex formation. Complexes were prepared with constant peptide and DNA concentration with different protein amounts (as marked). Complexes are run on a native gel (TBE 4%–12% PAGE). Addition of the peptide reduces protein-DNA complex formation.
(H) The peptide compromises assembly of higher order Int-CI complexes. Complexes were prepared at constant concentration of the peptide and CI5 DNA with increasing concentrations of the C-terminally truncated 390C Int variant (as marked), and run on a native gel (TBE 4%–12% PAGE).

Model for Tn1549 Integration
Int (gray ovals) binds to CI DNA (black circle) as a dimer and actively opens the crossover region (red) (i). Int is in an inactive conformation. Synapsis between CI and target DNA (blue) captures an AT-rich chromosomal site (ii). Following structural rearrangement and activation, Int-mediated cleavage and ligation of one strand pair generates an HJ intermediate (iii). Int cleaves 1nt inside the transposon end (top inset) and gets covalently attached to the IR DNA (black dot). The target site is cleaved at a conserved T. The strands are exchanged; base pairing at the 5′Ts helps join the partner strands (bottom inset). HJ resolution integrates Tn1549 in the recipient genome (iv). Again, Int cleaves upstream of a conserved T in the CI and the target and base pairing promotes strand ligation.
See also .