Hi,
I was recently told that Gibson assembly does not really need the use of Taq ligase. Once transformed, the bacteria can ligate the nicks on the vector with their endogenous enzymes.
Has anybody tested this and compared with the standard Gibson assembly?
Cheers
Martiniano
I don't know the answer. But I bet it's "possible", in the sense of getting a discoverable assembly, under carefully selected circumstances.
Generally Gibson Assembly is a numbers game, where you try to shift the efficiency of assembly of your desired construct versus undesired constructs to the point where you can discover the desired construct by screening a reasonable number of colonies. Omitting the ligase will certainly knock down the efficiency of your desired construct, but it won't affect the efficiency of some unwanted constructs, which don't require ligase, like spontaneous recombination products or template plasmid carryover.
So, I guess you won't necessarily make the method stop working under all circumstances, but you'll reduce its usefulness and versatility to a rather dissatisfying level.
What John Schloendorn said. Also keep in mind that Gibson assembly uses T5 exonuclease for generating overhangs, and nicked DNA is one of the substrates it can start digesting at, so there's a chance you could lower the yield of fully synthesized product pretty substantially by not having an enzyme that can seal those nicks.
Thank for the answers, but I was hopping to get an empirical reply.
I theory, the T5 activity will cease pretty fast due to thermal inactivation while the DNApol (Typically a phusion proof-reading enzyme) activity will continue and fill up any gaps leaving just a nick that can be repaired by the endogenous bacterial machinery. In that sense I would expect the reaction to work exactly like the site directed mutagenesis protocol where you regenerate a whole vector leaving just two nicks. I don´t think the efficiency should be much lower, but I´m hoping somebody did the trial and got some results to share.