Our protocol allows for the creation of restriction enzymes with altered, more stringent sequence specificities using in vitro compartmentalization and a unique selection strategy in directed evolution. Potentially it is quite universal, as it should be applicable to any restriction endonuclease and selection can be directed toward any more stringent version of an original cognate sequence. If newly created variants are expressed by in vitro transcription-translation, the protocol is suitable not only for narrowing down specificity, but also for altering specificity.
To decide sites for subsaturation mutagenesis, choose mutagenesis frequencies according to the hypothetical importance of the sites, keeping limits on the overall library complexity in mind. To begin the synthesis, insert columns that have been packed with new resin into the synthesizer. Synthesize oligonucleotides in all columns up to the triplet, immediately preceding the second subsaturation mutagenesis site counting from the three-prime end.
Synthesize, leaving a five-prime trityl group at the end. The protecting group will be removed at the beginning of the next synthesis cycle. Open the synthesis columns and briefly centrifuge into dry 1.5 milliliter tubes to collect the resin.
Pull the CPG synthesis support and mix by vortexing. Re-partition the mixed CPG resin into new synthesis columns. Avoid introducing humidity because it will decrease the overall yield.
Continue the synthesis, starting from the subsaturation mutagenesis site triplet. Assign columns to randomized NNS triplets or wild-type triplets according to the desired mutagenesis frequency. If additional subsaturation sites are present, proceed only to the triplet preceding the next subsaturation mutagenesis site.
If no more subsaturation sites are present downstream, complete the synthesis, leaving a five-prime trityl group at the end. Use wide bore pipette tips to prepare an oil surfactant mixture by adding 225 microliters of Span 80 and 25 microliters of Tween 80 to five milliliters of mineral oil in a 15 milliliter conical tube. Mix thoroughly by gentle inversion 15 times.
The mixture at this step should be translucent. For each library, transfer 950 microliters of the oil surfactant mixture to a two milliliter round bottom cryogenic vial. Label with a library name and transfer to ice.
Put one small cylindrical stirring bar into each vial. Prepare an in vitro transcription-translation reaction mixture according to the manufacturer's suggestions. Supplement the mixture with magnesium chloride to a final concentration of 1.5 millimolar.
Dispense 50 microliter aliquots of the reaction mixture into 1.5 milliliter tubes on ice. Add 1.7 femtomoles of the library to the reaction mixture on ice. Prepare water-in-oil emulsion consecutively for each library by first placing a small beaker filled with ice on a magnetic stirrer with the stirring speed set to 1150 RPM.
Then transfer a cryogenic vial with 950 microliters of oil surfactant mixture and a small stirring bar to an ice-cold beaker on the magnetic stirrer. Check that the stirring bar is spinning. Add five 10-microliter aliquots of the in vitro library transcription-translation mixture over a two-minute period in 30 second intervals and continue stirring for an additional minute.
Transfer the vial with the emulsion to an ice container. At this point, the liquid should be opaque, not translucent. Then, proceed with the next library.
After all the libraries are processed, start the incubation of all the libraries according to the kit manufacturer's recommendations. Transfer the vials to the temperature optimal for the engineered endonuclease for an additional two hours before placing them on ice for at least 10 minutes. Transfer the emulsions from the cryogenic vials into 1.5 milliliter tubes.
Add one microliter of 5 molar EDTA. And centrifuge them at 13, 000 times g for five minutes at room temperature. If after centrifuging you cannot see the water-oil interface clearly, freeze the mixture before aspiration of the upper oil phase.
Move the upper oil phase with a pipette and discard. Immediately perform extraction by adding 150 microliters of phenol chloroform and 100 microliters of 10 millimolar Tris-HCl to the aqueous phase. Vortex and then perform phase separation via 30-second centrifugation at 13, 000 times g.
Collect the upper aqueous phase. Precipitate the DNA by adding 15 microliters of three-molar sodium acetate, 2.5 to five micrograms of glycogen, and 375 microliters of ethanol. After incubating the samples at minus-20 degrees Celsius for one hour, centrifuge for 15 minutes at 13, 000 times g, four degrees Celsius.
Discard the supernatant and briefly wash the pellet with one milliliter of cold 70%ethanol. Dry the DNA glycogen pellet in a speed vac or air dryer for longer than five minutes. Then dissolve the pellet in 50 microliters of 10 millimolar Tris-HCl.
Next, add five microliters of streptavidin magnetic beads, prepared according to the manufacturer's instructions. Mix for one hour at room temperature, preferably in a carousel mixer or by gentle vortexing. Transfer the tubes to a magnetic stand to separate the beads.
Then collect the liquid enriched in DNA without biotin. This protocol is a tool to increase the frequency of desired variants of engineered restriction endonucleases by depleting inactive enzymes and endonucleases with unchanged wild-type sequence specificity. Successful screening can identify up to 20%of promising variants.
Variants are labeled as promising variants if they produce a cleavage pattern distinct from the wild-type enzyme. Variants that might have altered sequence preference are also labeled. Shown here is an unsuccessful screening with the majority of variance inactive and one variant with apparently unaltered cleavage pattern.
In this case, the libraries are most probably dominated by inactive variants that escaped the streptavidin capture selection step. It is critical to carefully design the selected and counter-selected sequences and to limit library diversity by mutagenesis strategy that generates substitutions in only a few selected positions. The best variants selected should be thoroughly characterized for binding and cleavage kinetics on preferred and not cleaved sequences based on the results from the initial screening.
In our test case, mutagenesis sites were selected based on a homology model. Surprisingly, a crystal structure later showed that some altered residues are most likely not in direct contact with DNA.
