Synapsis and DNA cleavage in phiC31 integrase-mediated site-specific recombination

Abstract

The Streptomyces phage phiC31 encodes an integrase belonging to the serine recombinase family of site-specific recombinases. The well studied serine recombinases, the resolvase/invertases, bring two recombination sites together in a synapse, and then catalyse a concerted four-strand staggered break in the DNA substrates whilst forming transient covalent attachments with the recessed 5' ends. Rotation of one pair of half sites by 180 degrees relative to the other pair occurs, to form the recombinant configuration followed by ligation of the DNA backbone. Here we address the nature of the recombination intermediates formed by phiC31 integrase when acting on its substrates attP and attB. We have identified intermediates containing integrase covalently attached to cleaved DNA substrates, attB or attP, by analysis of complexes in gels and after treatment of these complexes with proteinases. Using a catalytically inactive integrase mutant, S12A, the synaptic complexes containing integrase, attP and attB were identified. Furthermore, we have shown that attB mutants containing insertions or deletions are blocked in recombination at the stage of strand cleavage. Thus, there is a strict spacing requirement within attB, possibly for correct positioning of the catalytic serine relative to the scissile phosphate in the active site. Finally, using integrase S12A we confirmed the inability of attL and attR or other combinations of sites to form a stable synapse, indicating that the directionality of integrative recombination is determined at synapsis.

Figure 1

Detection of recombination intermediates by φC31 integrase and susceptibility to proteinase K. Labelled attB fragment (lanes 5–11) was incubated without (lane 5) or with integrase (lanes 6–11) with no further additions (for 30 min and 2 h; lanes 6 and 7, respectively) or with 192 bp cold attP (for 30 min and 2 h; lanes 8 and 9, respectively) or 192 bp cold attP^TA (for 30 min and 2 h; lanes 10 and 11, respectively) under suboptimal conditions for recombination. The shifted attB fragment in the presence of integrase alone is labelled Int:attB II (see also Fig. 4). The positions of the recombination products either free (attL in lane 1; attR in lane 3) or complexed with integrase (Int:attL in lane 2 and Int:attR in lane 4) are also shown. When attB, integrase and the cold partner fragment were present, supershifts (labelled CC and SC) and recombination products (free and shifted) were observed in addition to free attB and shifted attB.

Figure 2

Susceptibility of complexes to subtilisin and release of cleaved substrates. (A) Representation of the two radiolabelled substrates attB and attP. The bases filled in by the Klenow reaction are shown in grey and those in bold are radiolabelled. The black arrows indicate the proposed positions of cleavage by integrase and concomitant formation of a covalent bond to the recessed 5′ ends. The grey arrows on the attB sequence are the positions of cleavage by StyI. (B) Disruption of complexes by subtilisin (lanes 4 and 6) and sizing of the cleaved substrates using StyI cut attB as a marker (lane 5). Labelled attB (lanes 1–5) and attP (lanes 6–9) and cold partner fragments (192 bp attP, lanes 3 and 4; 193 bp attB, lanes 6 and 7) were used as substrates for integrase (lanes 2–4 and 6–8).

Figure 3

Identification of the covalently bound cleaved intermediate of attB. A reaction containing labelled attB, integrase and 192 bp cold partner attP was loaded on a non-denaturing gel (as shown for lane 9, Fig. 1), the lane excised and incubated with proteinase K. The gel slice was then inserted horizontally along the top of a second non-denaturing gel and subjected to electrophoresis. On the same gel a reaction containing the same components, i.e. labelled attB, integrase, 192 bp cold partner attP was incubated, treated with proteinase and then run in a marker lane (M) next to the gel slice. No correction was made for the additional gel matrix from the gel slice so there is a shift in mobility of free attB, cleaved attB and attL and attR (shown with arrows) compared to the marker lane.

Figure 4

Interaction with attP and attB by the integrase mutant S12A. (A) Affinity of IntS12A for attB and attP compared to the wild-type integrase. The major complexes are labelled ‘Int:attB II’ and ‘Int:attP II’, the minor complexes ‘Int:attB I’ and Int:attP I’ [after Thorpe et al. (11)]. (B) Supershift complexes formed with Int S12A. Labelled attB was incubated with wild-type (lanes 2–9) and mutant integrases (lanes 10–15) in the presence of no cold attP (lanes 2 and 3), 75 bp cold attP (lanes 4, 7, 10 and 13), 94 bp cold attP (lanes 5, 8, 11 and 14), or 192 bp cold attP (lanes 6, 9, 12 and 15) and then untreated (lanes 2, 4–6 and 10–12) or treated (lanes 3, 7–9 and 13–15) with proteinase K. The mobilities of the different complexes and recombination products changes with the size of the cold attP fragment added. The SC and CC complexes are shown by asterisks and filled circles, respectively. The shifted attL and/or attR fragments are shown by open circles and the recombinant products attL and attR are shown by filled and empty triangles, respectively.

Figure 5

Insertions or deletions in attB abolish recombination but have little effect on affinity of integrase. (A) Sequence of attB and the mutant attB sites used in this study. The plasmids that encode the sites are also shown. (B) Recombination reactions by wild-type and mutant sites. Reactions containing two supercoiled plasmids, one encoding attP (pRT700 or pRT702) and the other encoding attB (wild-type or mutant sites) were incubated for 1 h at 30°C with varying integrase concentrations, heat treated and then digested with HindIII. Linearized parental plasmids and fragments containing recombination products were separated by agarose gel electrophoresis. (C) Affinity of wild-type integrase for the mutant sites attB^MSX7 and attB²¹¹⁶.

Figure 6

Insertion mutants attB^MSX7 and attB²¹¹⁶ can synapse with attP and integrase but are inhibited in strand cleavage. (A) Enrichment of the SC complex in reactions containing 194 bp attB^MSX7 and attB²¹¹⁶ as cold partner fragments. Labelled attP was incubated in the absence (lane 1) or presence of wild-type integrase (lanes 2–9) or IntS12A (lanes 10 and 11). Partner fragments (∼200 bp fragments) were added as shown and either not treated further or treated with protease. (B) Labelled substrate attB²¹¹⁶ also accumulated SC. (C) 2D gel analysis of the intermediates obtained with labelled attB²¹¹⁶. The experiment was performed as described for Figure 3.

Figure 7

Use of IntS12A to detect SC with all combinations of att sites. AttB, attP, attL and attR were labelled and incubated without (lanes 1, 6, 11 and 16) or with (lanes 2–5, 7–10, 12–15 and 17–20) integrase S12A and all four ∼200 bp partner fragments as shown. The positions of the free probe (attP, attB, attR or attL), the shifted att sites (Int:att) and the SC are shown with arrows. Only combinations of attP and attB generated SC.