Dear Ferdinand:

Having only 5 months and 75 amino acids – this is very challenging, I hope you have a good screening or selection system.

To your question: I would guess you can randomize more than 3 positions at the same time. You have to make sure that a suitable primer is long enough or still binds correctly to your wildtype sequence. You can also try to use more than one primer in one PCR or performing two or more PCRs and then assemble the different PCR products to obtain your full gene or vector.

Regarding the choice of the best degenerate codon: In my old lab we had the option to identify so called allowed and not allowed amino acid for an interesting position by a (structural) alignment of similar enzymes, and based on the allowed amino acids appropriate primers/codons were selected, which reduces the size of the library (“…only those amino acids were considered for library creation that appeared frequently in structurally equivalent positions. Thus, the number of mutants to be screened could be substantially reduced while the number of functionally intact variants was increased.”, Jochens & Bornscheuer, ChemBioChem 2010, Vol 11, p1861-1866).

You might also check other publications of the Bornscheuer group and of Prof. Reetz (CASTING, ISM) were you should get a clue.

Good luck!

The thing to keep in mind is how thoroughly you want to assess each position in terms of saturation. If you wanted to do complete saturation mutagenesis of 5 positions there are 20e5 possible variants, and since you will get duplications you need well over 20,000,000 clones just for those 5 positions.

Well with NDT you have only 12 AA so you land around 7,000,000 clones (95% coverage), but you are right it defeats to purpose to keep the libarys small and i only save only a few transformations anyway. My screening system allows a very high throughput im just restricted by the transformation efficiency into my screening strain. I will use NNK and screen only half of each library (around 40,000 colonies) and see were I land with it, if i see no improvement i can always screen more.

Hello Ferdinand

I suggest asking yourself the question:

Are all of these 75 aa's of equal importance?

Try to make a ranking based on available data (experimental, in silico, literature etc.) and see if you can limit it to a range of 10-30 residues for the beginning.

In addition, recent publications by Yuval Nov indicate that the traditional screening effort (90-95% coverage) is very conservative and possible could be lowered maintaining the chance to find an improved variant within the targeted (and hopefully by PCR produced) sequence space. Don't forget to make controls if your PCR successfully created the achieved degeneration, otherwise you look for something that isn't there.

See Yuvals work here:

Nov Yuval (2013), "Fitness Loss and Library Size Determination in Saturation Mutagenesis." PLoS ONE, 8(7): e68069.

Nov Yuval (2012), "When Second Best is Good Enough: Another Probabilistic Look at Saturation Mutagenesis." Applied and Environmental Microbiology, 78(1), 258-262

Use his great program to play a little bit with your numbers.

http://stat.haifa.ac.il/~yuval/toplib/

Additionally, don't use NNK. If you want to mutate to all other 19 codons use either Tangs-Trick or the 22-codon-trick. Both will remove redundancy and therefore lower your screening effort. You could also create your own, self-made degeneracy by mixing primers to include those residues you may know already of playing an important role. (e.g NDC+Trp, 12:1 mixing)

Which trick you use depends on your codon locations within the gene and how much money your project includes for primer purchase. There are other more complex possibilities available like MAX, Omnichange, TrimerDimer etc.

Find the original papers of the two Tricks here:

http://pubs.acs.org/doi/abs/10.1021/sb300037w http://www.biotechniques.com/BiotechniquesJournal/2012/March/Construction-of-small-intelligent-focused-mutagenesis-libraries-using-well-designed-combinatorial-degenerate-primers/biotechniques-327414.html

One more thing. If you system is expressed in E.coli use NDC and NNS instead of NNT and NNK to avoide rare codons and achieve a better annealing during the PCR . The latter will help in achieving a higher yield of the designed degeneracy.

And finally, answering your original question:

Yes you can!

It is actually a very good idea. It also provids an outlook of greater success, as if only 1-3 residues are simultaneously mutated.

Have a look on this work by L. Parra and M. Reetz

http://onlinelibrary.wiley.com/doi/10.1002/cbic.201300486/abstract;jsessionid=25F91ADF5F66BF458F10E015262F52F0.f01t02

and also this article by A. Sandström and J. Bäckvall

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3252943/

Good Luck and Happy Screening!