Nearest-neighbor amino acids of specificity-determining residues influence the activity of engineered Cre-type recombinases

Abstract

The tyrosine-type site-specific DNA recombinase Cre recombines its target site, loxP, with high activity and specificity without cross-recombining the target sites of highly related recombinases. Understanding how Cre achieves this precision is key to be able to rationally engineer site-specific recombinases (SSRs) for genome editing applications. Previous work has revealed key residues for target site selectivity in the Cre/loxP and the related Dre/rox recombinase systems. However, enzymes in which these residues were changed to the respective counterpart only showed weak activity on the foreign target site. Here, we use molecular modeling and dynamics simulation techniques to comprehensively explore the mechanisms by which these residues determine target recognition in the context of their flanking regions in the protein-DNA interface, and we establish a structure-based rationale for the design of improved recombination activities. Our theoretical models reveal that nearest-neighbors to the specificity-determining residues are important players for enhancing SSR activity on the foreign target site. Based on the established rationale, we design new Cre variants with improved rox recombination activities, which we validate experimentally. Our work provides new insights into the target recognition mechanisms of Cre-like recombinases and represents an important step towards the rational design of SSRs for applied genome engineering.

Figure 1

Nucleotides and amino acids at the protein-DNA interface of Cre/loxP and Dre/rox recombinase systems. (a) Sequences of loxP and rox target sites. The three nucleotides that differ between loxP and rox are highlighted in color. The numbering of the individual bases is provided for the upper and lower DNA strand, respectively. (b) Protein amino acids facing the altered DNA bases of loxP and rox. (c) Snapshot of the cleaving monomer of the Cre/loxP complex (PDB 1Q3U). The recognition regions at the protein-DNA interface are displayed in magenta (helix B), orange (helix D), green (helix J) and purple (beta 4). The area of interest in this study (i.e. PDI_BJ area; the protein-DNA interface formed by amino acids of helix B and J and nucleotides 10/66, 11/65, and 12/64) is zoomed in, and relevant bases and amino acids are labeled. The numbering used is based on the Cre/loxP crystal structure (for details see Fig. S1). Figure generated using Pymol (version 2.1, https://pymol.org/).

Figure 2

Analysis of protein-DNA interactions in Cre/loxP (a) Projection of hydrogen bond interactions spotted in the last frame of the Cre/loxP MD trajectory. The base-specific interactions with the minor groove are labeled with an asterisk (*). Hashes (#) denote the interactions with the major groove. Nucplot was used to generate this image. (b) Detailed view of interactions observed in the Cre/loxP crystal structure and MD simulation in the PDI_BJ area. Hydrogen bonds are shown in black dashed lines. The specificity-determining residues are highlighted with a yellow box. Water positions predicted by WaterMap (version 1.0, https://www.schrodinger.com/) are shown in orange spheres, and water-mediated contacts are depicted by orange lines. For comparison, a network of water molecules found in this area in another Cre/loxP crystal structure (PDB 3C29) is shown superimposed with the Cre/loxP MD-refined structure (cyan spheres and lines). Pymol was used to generate the image (version 2.1, https://pymol.org/).

Figure 3

Analysis of protein-DNA interactions observed in the MD simulation of Dre/rox (a) Projection of hydrogen bond interactions spotted in the last frame of the Dre/rox MD trajectory. The base-specific interactions with the minor groove are labeled with an asterisk (*). Hashes (#) denote the interactions with the DNA major groove. Nucplot was used to generate the image. (b) Detailed view of interactions observed in the PDI_BJ area of the Dre/rox complex based on MD simulations. The specificity-determining residues are highlighted with a yellow box. Hydrogen bonds are shown in black dashed lines and van der Waals interactions in dotted spheres. Pymol was used to generate the image (version 2.1, https://pymol.org/).

Figure 4

Detailed view of interactions observed in the (a) mCre_K/loxP complex and (b) mCre_K/rox complex at the PDI_BJ area based on MD simulations. The mutated amino acids with respect to wild type Cre are labeled with a yellow box. The hydrogen bonds are shown in black dashed lines and the van der Waals interactions in dotted spheres. A predicted water position using WaterMap (version 1.0, https://www.schrodinger.com/) is represented by an orange sphere and its interactions in solid orange lines. Pymol was used to generate the image (version 2.1, https://pymol.org/).

Figure 5

Detailed view of interactions observed at the PDI_BJ area in the MD simulations of the newly designed Cre variants with rox. (a) mCre1 (K43R, R259P, G263K, T258A) (b) mCre2 (K43R, R259P, G263K, T258L) (c) mCre3 (K43R, R259P, G263K, T258L, E262L) and (d) mCre4 (K43R, R259P, G263K, T258A, E262I). The mutated amino acids with respect to wild type Cre are labeled (yellow box). The hydrogen bonds are shown in black dashed lines and van der Waals interactions in dotted spheres. Pymol was used to generate the image (version 2.1, https://pymol.org/).

Figure 6

Net free energy component analyses of the studied complexes; net_pol(GB/PB): net polar contributions from MM-GBSA/MM-PBSA (net_pol(GB/PB) = ΔE_ele + ΔG_GB/PB), net_npol(GB/PB): net non-polar contributions from MM-GBSA/MM-PBSA (net_npol(GB/PB) = ΔE_vdW + ΔG_SA(GB/PB)).

Figure 7

Rationally designed Cre mutants show increased recombination activity in E.coli. (a) Schematic drawing of the plasmid assay. Important regions in the plasmids are indicated. Note the reduced size of the plasmid after recombination. The restrictions sites (BsrGI and XbaI) used for cloning indicated Cre-recombinase variants are depicted. The rox target sites are shown as red triangles. CmR, chloramphenicol resistance gene; ori, origin of replication; AraC, arabinose operon regulatory gene. (b) Agarose gels of three independently picked clones showing BsrGI and XbaI digested plasmids carrying indicated recombinases. The amount of arabinose added to the growth medium is presented below each band in μg/ml l+-arabinose. The line with two triangles indicates no recombination, whereas the line with one triangle marks the recombined band. Quantifications of the ratios of band intensities (in percent of recombination) are shown to the right for each mutant. The amount of arabinose added to the growth medium is shown on the X-axis. Error bars depict standard deviation from the three independent experiments.