Control of serine integrase recombination directionality by fusion with the directionality factor

Abstract Bacteriophage serine integrases are extensively used in biotechnology and synthetic biology for assembly and rearrangement of DNA sequences. Serine integrases promote recombination between two different DNA sites, attP and attB, to form recombinant attL and attR sites. The ‘reverse’ reaction requires another phage-encoded protein called the recombination directionality factor (RDF) in addition to integrase; RDF activates attL × attR recombination and inhibits attP × attB recombination. We show here that serine integrases can be fused to their cognate RDFs to create single proteins that catalyse efficient attL × attR recombination in vivo and in vitro, whereas attP × attB recombination efficiency is reduced. We provide evidence that activation of attL × attR recombination involves intra-subunit contacts between the integrase and RDF moieties of the fusion protein. Minor changes in the length and sequence of the integrase–RDF linker peptide did not affect fusion protein recombination activity. The efficiency and single-protein convenience of integrase–RDF fusion proteins make them potentially very advantageous for biotechnology/synthetic biology applications. Here, we demonstrate efficient gene cassette replacement in a synthetic metabolic pathway gene array as a proof of principle.


Figure S1
Vectors used for construction of recombination reaction substrates. (A) pFM141 (the vector used to construct in vivo reaction substrates). Recombination (att) sites for φC31 integrase were introduced by replacing the stuffer sequence flanked by XbaI and NotI sites (Site A), and EcoRI and SacI sites (Site B) with synthetic double-stranded oligonucleotides (see Table S1). Bxb1 integrase substrates were made similarly from pMS183Δ, a variant of FM141 in which sites A and B are both flanked by EcoRI and SacI sites. (B) pFM122 (the vector used to construct in vitro reaction substrates).
Recombination sites were cloned by replacement of the stuffer sequences between SpeI and NotI sites (Site A), and EcoRI and SacI sites (Site B) with double-stranded oligonucleotides containing att sites as shown in Table S1, flanked with the appropriate restriction sites.

Table S1
Sequences of the recombination (att) sites for φC31 integrase and Bxb1 integrase, used in this study. The central 2-bp overlap sequences of the att sites are highlighted in red font. The flanking restriction sites used in cloning double-stranded oligonucleotides into pFM141 and pFM122 (shown in Figure S1) are highlighted in blue; bases in italics are not in the synthetic oligonucleotides used for cloning the sites. Note that SpeI and XbaI-cut sites have compatible ends. The sequences as shown here give substrates with head-to-tail sites (i.e. recombination results in resolution/deletion).
To make inversion substrates, the orientation of the sequence (black letters) in one of the doublestranded oligonucleotides (attB or attL) was reversed.

Table S2
Sequences of plasmids used in this study.
(1) Full sequence of pFEM33, the plasmid used for low-level expression of φC31.Int-gp3 fusion protein in E. coli. pFEM33 has a pMB1 origin of replication and an ampicillin-resistance gene. The φC31 integrase sequence is highlighted in cyan, and the gp3 sequence is highlighted in grey. The linker between these sequences is described in the main text ( Figure 1C). The restriction sites referred to in the main text are highlighted in yellow: NdeI, CATATG; SpeI, ACTAGT; Acc65I, GGTACC; BglII, AGATCT; XhoI, CTCGAG. Stop codons (TAA) are in red font. The expression plasmid for φC31 integrase is similar, except that the gp3 sequence and the integrase-gp3 linker are sequences are absent.
(2) The coding sequence for Bxb1 integrase, and (3) the coding sequence for gp47. Sequences of the expression plasmids for Bxb1 integrase, Bxb1.Int-gp47, and gp47 can be obtained by replacement of the relevant coding sequences in pFEM33.
(4) Full sequence of pFM141, the vector used to construct in vivo recombination substrates (see Figure S1). Restriction sites are highlighted in yellow: XbaI, TCTAGA; NotI, GCGGCCGC; EcoRI, GAATTC; SacI, GAGCTC. The region between the two att sites containing the galK gene (highlighted in cyan) is deleted or inverted by recombination, along with half of each att site.