Mechanical constraints on Hin subunit rotation imposed by the Fis/enhancer system and DNA supercoiling during site-specific recombination

Abstract

Hin, a member of the serine family of site-specific recombinases, regulates gene expression by inverting a DNA segment. DNA inversion requires assembly of an invertasome complex in which a recombinational enhancer DNA segment bound by the Fis protein associates with the Hin synaptic complex at the base of a supercoiled DNA branch. Each of the four Hin subunits becomes covalently joined to the cleaved DNA ends, and DNA exchange occurs by translocation of a Hin subunit pair within the tetramer. We show here that, although the Hin tetramer forms a bidirectional molecular swivel, the Fis/enhancer system determines both the direction and number of subunit rotations. The chirality of supercoiling directs rotational direction, and the short DNA loop stabilized by Fis-Hin contacts limit rotational processivity, thereby ensuring that the DNA strands religate in the recombinant configuration. We identify multiple rotational conformers that are formed under different supercoiling and solution conditions.

Figure 1. Crosslinking of Invertasomes Assembled during Hin-catalyzed Site-specific DNA Recombination

(A) Pathway of the site-specific DNA inversion reaction by Hin. Hin dimers bind the hixL and hixR recombination sites, and Fis dimers bind two sites on the recombinational enhancer segment. The Hin-hix and Fis-enhancer segments assemble into an invertasome complex at the base of a supercoiled DNA branch with the help of the DNA bending protein HU. The Hin tetramer catalyzes a two base staggered cleavage resulting in each of the four subunits engaged in a serine-phosphodiester linkage with the 5′ DNA ends. Consistent with the data in this paper, the top pair (yellow and purple) of DNA-linked Hin subunits undergo a single clockwise rotation to generate heterodimers. Ligation of the DNA ends generates the inverted DNA product. (B) Structural model of the synaptic Hin tetramer linked to the 5′ ends of four DNA duplexes based on the crystal structure of γδ resolvase (2GM4). The view highlights the “flat interface” whereby the top pair of synapsed subunits (yellow and purple) can rotate relative to the bottom pair (green and blue). Residues discussed in this paper that support efficient crosslinking are denoted with red spheres. (C) Subunit rotation mechanism. Panels i and ii: Hin tetramer model in the orientation as in (B) and after a 90° rotation about the x-axis, respectively. The long E helices from each subunit are highlighted. Orange spheres represent Cys134, red spheres represent Cys94, and gray spheres represent Cys101. A 90° clockwise (iii) or counterclockwise (iv) rotation of the yellow and purple subunits generates helix E-aligned conformers and positions pairs of cysteines for crosslinking. Additional 90° rotations generate the recombinant DNA-aligned conformation (v). (D) Strategy used for cysteine-dependent crosslinking and ³²P-end-labeling of Hin-DNA covalent complexes formed on supercoiled DNA. The plasmid substrates (see Figure S1 for details) contain two hixL sites that are flanked by EcoR1 sites; for hixL1 the distance of EcoR1 sites from the Hin cleavage sites is 50 bp and for hixL2 the distances are either 50 bp or 14 bp. The enhancer is spaced from 99 bp to 155 bp from the center of hixL1 in most substrates. (i) Hin and Fis are incubated with the supercoiled plasmid substrate under ethylene glycol Mg²⁺-free conditions to generate DNA-cleaved invertasomes. (ii) At specified times after Hin addition, reactions are subjected to cysteine-specific crosslinking for 20-60 sec followed by (iii) digestion with EcoR1 and end-filling with α-³²P-dATP and (iv) SDS-PAGE. Crosslinks between different pairs of subunits are depicted. (E) Crosslinking of Hin-DNA complexes formed on supercoiled plasmid DNA by Hin mutants containing cysteines at different locations. Hin cysteine mutants (labeled at the bottom of each panel) were incubated with Fis and pRJ2372 for 20 min and subjected to 1 min crosslinking reactions with diamide (0 Å), BMOE (8 Å), BMH (16 Å), or a no crosslinker control (none); no Hin was added to the reaction labeled “-Hin.” Autoradiographs of the SDS gels are shown. The fastest migrating species is the monomeric Hin-(³²P)54 nt covalent complex, and the slower migrating species is the crosslinked dimeric Hin-(³²P)54 nt product. An additional product present with Hin-S94C, D96C, S99C (trace), and M101C (trace) are crosslinked dimers in which only one Hin subunit is covalently associated with DNA. (+) indicates the location of a minor Hin-DNA cleavage band that appears in different amounts independent of crosslinking, and (*) indicates the location of a DNA-only band that is derived from the invertible segment.

Figure 2. Crosslinking between Cysteines Located at the C-terminal End of Helix E

(A) Schematic representation of the Hin tetramer undergoing clockwise or counterclockwise subunit rotations to generate the observed crosslinks between cysteines at residue 134 (see Movie S1A and B). The cartoons are modeled after Figure 1C (ii-v) with the catalytic and DNA binding domains represented by large and small ovals separated by the E helices. The different length DNA segments generated after EcoR1 digestion are also depicted. Red stars denote the locations of Q134C on each subunit, and the boxed cartoons depict the crosslinked products resolved by SDS-PAGE. (B) Crosslinking of Fis-activated Hin-Q134C reactions on supercoiled pRJ2330. Crosslinking was for 20 sec with BMOE (8 Å spacer) at the designated times after Hin addition. The portion of the SDS-polyacrylamide gel showing the crosslinked products is shown (see Figure S2 for the autoradiograph of the full gel). The predominant crosslinked hetero-diprotomer species containing the Hin-54 nt (³²P-labeled) and Hin-18 nt (unlabeled) migrates fastest, the crosslinked homo-diprotomer containing labeled 54 nt segments migrates near the top, and the crosslinked homo-diprotomer containing unlabeled 18 nt segment is not visible in the autoradiograph. Control lanes on the right show: (1) the location of the crosslinked 54-54 nt homo-diprotomer from reactions performed on pRJ2372, (2) the products obtained without crosslinker added, and (3) a reaction without Hin added. The asterisk marks a vector DNA band from the invertible segment. (C) Crosslinking of reactions with the Fis-independent mutant Hin-H107Y/Q134C on open-circular pRJ2330. (D) Plot of the relative proportion of the 54-18 nt hetero-diprotomer crosslinked products with respect to Hin incubation time obtained from quantifying the results of panels B (filled circles, negatively supercoiled DNA) and C (open circles, open circular DNA).

Figure 3. Crosslinking between Cysteines Located at Residue 94

(A) Models of the Hin tetramer undergoing clockwise or counterclockwise subunit rotations to generate the observed crosslinks between cysteines at residue 94 (see Movie S2A and B). The outlay of the figure is similar to Figure 2A. Although the helix E-aligned rotational conformers are depicted here, clockwise rotations of 90°-155° and counterclockwise rotations of 25°-90° are predicted to support crosslinking (Figure S6B). (B) Crosslinking of Fis-activated Hin-S94C reactions on supercoiled pRJ2330 performed as in Figure 2B. The predominant labeled crosslinked product is the 54-54 nt homo-diprotomer species. (C) Crosslinking of reactions with the Fis-independent mutant Hin-H107Y/S94C on open-circular pRJ2330. (D) Plot of the relative proportion of the hetero-diprotomer crosslinked products with respect to Hin incubation time obtained from quantifying the results of panels B (filled circles, negatively supercoiled DNA) and C (open circles, open circular DNA). Data (gray circles) from Hin-H107Y/S94C crosslinking reactions on 112 bp and 40 bp hix fragments are also included (see Figure S4A for the gel).

Figure 4. Crosslinking between Cysteines Located at Residue 101

(A) Models of the Hin tetramer undergoing clockwise subunit rotation to generate the observed crosslinks between cysteines at residue 101 (see Movie S3). Cys101 residues from subunits bound to synapsed hix sites (e.g., yellow and blue subunits) are within crosslinking distance to generate hetero-diprotomer products at rotations from 0° -80° (Figure S6C). Cys101 from subunits originally bound as dimers would not become sufficiently close to form homo-diprotomer crosslinked products (e.g., between yellow and green subunits) until almost a complete 180° rotation. (B) Crosslinking of Fis-activated Hin-M101C reactions on negatively supercoiled pRJ2330 performed as in Figure 2A. (C) Crosslinking of reactions with the Fis-independent mutant Hin-H107Y/M101C on open-circular pRJ2330. (+) marks background bands. (D) Plot of the relative proportion of the hetero-diprotomer crosslinked products with respect to Hin incubation time obtained from quantifying the results of panels B (filled circles, negatively supercoiled DNA) and C (open circles, open circular DNA). E. Crosslinking with Hin-H107Y/M101C on an equal mixture of 3′-³²P-labeled 112 bp and unlabeled 40 bp hix fragments. Crosslinking was for 20 sec with BMOE at the indicated times after Hin addition. 112/40 bp synaptic complexes were then isolated by native PAGE, extracted, and subjected to SDS-PAGE. The region of the gel corresponding to the crosslinked products is shown and the locations of the 54-54 nt homo-diprotomer and 54-18 nt hetero-diprotomer products are indicated. The 54-54 homo-diprotomer control lane on the right was from a reaction that only contained labeled 112 bp hix fragments. (F) Plot of the relative proportion of the hetero-diprotomer crosslinked products with respect to Hin incubation time obtained from quantifying the results of panels E.

Figure 5. Subunit Rotation on Positively Supercoiled DNA

(A) Agarose gel of pRJ2385 preparations: lane 1, negatively (-) supercoiled DNA extracted from cells; lane 2, relaxed DNA generated with topoisomerase I; lane 3, positively (+) supercoiled DNA generated with reverse gyrase. (B and C) DNA cleavage and inversion reactions by Hin-wt, respectively. Fis-activated reactions on (+) supercoiled pRJ2385 were for 10, 30, and 60 min; -Fis and all Hin reactions on (-) supercoiled DNA were for 10 min. Cleavage products were electrophoresed directly after proteinase K treatment, and inversion reactions were first cleaved with restriction enzymes to detect the inverted and parental products (Johnson and Bruist, 1989). (D and E) Fis-activated Hin-Q134C crosslinking reactions on (+) and (-) supercoiled pRJ2385, respectively. (F) Plots of the relative proportion of the hetero-diprotomer crosslinked products as a function of Hin incubation time obtained from quantifying the results of panels D (filled circles, (+) supercoiled DNA) and E (open circles, (-) supercoiled DNA). (G and H) Fis-activated Hin-S94C crosslinking reactions on (+) and (-) supercoiled pRJ2385, respectively. (I) Plots of the relative proportion of the hetero-diprotomer crosslinked products as a function of Hin incubation time obtained from quantifying the results of panels G (filled circles, (+) supercoiled DNA) and H (open circles, (-) supercoiled DNA).

Figure 6. Fis-Hin Contacts Limit Processivity of Hin Subunit Rotations

(A and B) Hin crosslinking of Hin-Q134C and S94C, respectively, activated by wild-type and D20N mutant Fis. Hin and Fis were incubated with pRJ2330 for 20 min to accumulate cleaved invertasomes and then subjected to crosslinking with BMOE. Control lanes are: 1, no crosslinker added; 2, no Fis added; 3, Fis-WT + Hin-Q134C or S94C reaction with pRJ2372 to mark the 54-54 nt homo-diprotomer crosslinked product. The bar graphs give the means and standard deviations of the hetero-diprotomer crosslinked products generated from Fis-WT and D20N reactions determined from 5 experiments. (C) Fis-activated Hin-Q134C crosslinking reactions performed as in Figure 2B but with pRJ2123, which contains the enhancer 5 bp closer to hixL1 than pRJ2330. (D and E) Plots of reactions with Hin-Q134C and S94C, respectively. Filled circles are data from pRJ2123 (enhancer out-of-phase, from panel C and Figure S4B, respectively) and open circles are data from pRJ2330 (enhancer in-phase, from Figures 2B and 3B, respectively). (F) Fis structure (PDB code 1F36) highlighting important residues for Hin activation or positions where cysteines were introduced for crosslinking experiments. (G) Fis-Hin crosslinking. Fis-Q21C + Hin-WT were incubated with pRJ2372 for 20 min to accumulate cleaved invertasomes followed by crosslinking for 30 sec with the cysteine-lysine heterobifunctional agents SIA (1.5 Å spacer), AMAS (4.4 Å spacer), GMBS (6.8 Å spacer), or no crosslinker. After quenching, the DNA was cleaved with EcoR1 and end-labeled with ³²P-dATP. The locations of the Hin-DNA(³²P) complex and Hin-DNA(³²P) crosslinked with one Fis subunit are shown; the minor slowest migrating labeled species may represent 2 Fis subunits crosslinked to a Hin-DNA(³²P) complex. Similar crosslinking reactions using AMAS were performed on pRJ2372 (enhancer in-phase) and pRJ2118 (enhancer out-of-phase) in the presence of Fis-Q21C, WT (control), and R71C (control). (H) Bar graph giving the mean and standard deviations of the percent of Fis-Hin crosslinks relative to the total number of Hin-DNA(³²P) complexes obtained with pRJ2330 (enhancer in-phase) and pRJ2123 (enhancer out-of-phase) determined from 8 reactions where the crosslinking times varied from 30 sec to 5 min. (I) Plot comparing Fis-activated Hin-Q134C crosslinking reactions on pRJ2340, containing the enhancer 972 and 730 bp from the hix sites (long spacing, filled circles; Figure S4C), with reactions on pRJ2330 (short spacing, open circles, from Figure 2B).

Figure 7. Effects of core nucleotide homology and reaction conditions on rotation conformations

(A) Schematic representation of synapsis and DNA exchange between hixL(WT) and hixL(A/T) recombination sites. After a single-round DNA exchange one of the two core nucleotides cannot base-pair, which prevents ligation. (B and C) The single-round subunit exchange property of Fis-activated Hin reactions is not disturbed by the formation of mismatched DNA recombinant products. Hin crosslinking was performed on Fis-activated Hin-Q134C (B) and S94C (C) reactions using pRJ2330 [hixL(WT) × hixL(WT)] or pRJ2383 [hixL(WT) × hixL(AT)]. No significant differences in the ratios of crosslinked products at any of the time points are observed. (D) Ethylene glycol influences the distribution of rotational conformers. Fis-activated reactions on Hin-M101C were preformed on pRJ2383 for 20 min in reaction buffer containing: lanes 1 and 2, 25% ethylene glycol plus 2 mM EDTA; lanes 3 and 4, 25% ethylene glycol plus 10 mM MgCl₂; lanes 5 and 6, part of the reaction in lanes 1 and 2 was diluted in Mg²⁺-free EDTA buffer to a final concentration of 5% ethylene glycol; lanes 7 and 8, same as lanes 5 and 6 except the dilution buffer contained 10 mM MgCl₂; lanes 9 and 10, independent reaction were performed in the same manner as in lanes 7 and 8; lanes 11 and 12, ethylene glycol (25% final concentration) was added back to part of the reaction shown in lanes 10 and 11. The reactions were quenched without (-) or after (+) BMOE crosslinking for 1 min. Note that ligation, which would normally occur upon dilution of the ethylene glycol and addition of Mg²⁺ and would reverse the Hin-DNA linkage to generate DNA inversion products, is inhibited because the substrate contains non-identical core nucleotides. The ratios of crosslinked hetero-diprotomer/homo-diprotomer products averaged from 2 independent experiments are given below the lanes.