three-point ligation - (Jun/14/2007 )

I searched the forums and did not find anything on this. I have a big insert, which I could not PCR all 4kb of in one shot, different parts required different buffers etc., so I split the PCR work into 2. Primers were designed to amplify past a central BsgI cutsite, which I chose as the midpoint of the thing. I cut the vector with Sph1 and cut the insertds with Sph1 qnd BSG1. I plan on mixxing both fragments witht he vector nd adding ligase to let the ligase peice together and assemble the fragments. Has anyone done anything like this? Any suggestions?

Thank you.

Yup, done 3-way, 4-way and 5-way ligation.

I usually keep the mole ratio of vector:InsertA:InsertB at 1:1:1. I have varied the vector:inserts ratio before, but have not seen any difference. So I keep to 1:1:1 ratio.

For a 3way ligation you need to be able to screen a large number of colonies. Colony PCR and a multichannel pipette will come in handy. While the efficiency of 3 way ligation is rather good, (1:7, to 1:20), the process is rather stochastic, you might have no colonies for the first 20 colonies, then 3 in the next 5.

However there is a slight problem with the current ligation strategy. (well actually it is an outcome) because both insert A and insert B share both the BsgI and SphI cut site, there are now a large number of ways that the insert could ligate to each other.
So be aware that efficiency will be lower (need to screen lots. say 84 colonies), and you have to bare in mind the possibility of 3 inserts in your vector. The later restriction digest should screen for that.

As for the 4kB PCR. Cheamps does raise a point.

It is the wierdest thing. The gene is actually 7069 bp. My PI designed primers to amplify all of it at once. I PCR'd using them and using phusion polymerase or Finnzymes EXT Dynazyme, any number of combinations of additives, glycerol, DMSO, betaine, HF buffer, GC buffer, different gradients etc, no band was visible above 3 kb. I hypothesized that there was either repetitive DNA, or inverted repeat DNA somewhere in the sequence. And tried splitting the gene into quarters, redsigned primers and restarted.

I was able to get the first half using Phusion and HF buffer. The next quarter using Phusion with GC buffer and the last quarter. using just plain old Taq.

My control using Phusion to amplify 10Kb worked just fine, but for some reason this one could not be done in one PCR, unless you know some tricks I do not.

there is emulsion PCR. which does handle internal repeats better then regular PCR. But since you already have the fragments made, I say go with that.

50% + 25% +25% + vector (pBS or something with colour testing)

4 way ligation.... If you have enough DNA, a multichannel pipette and the ends of the fragments have restirction sites that are non compatible (ie they can only assemble one way), you could try assembling your entire gene in a vector like pBS (something colour testable). Once build, could then move the completed gene into the desired vector.

Basically, you make a water in oil emulsion of the PCR solution.

The PCR solution is thus broken up into many small water droplets floating in oil. The concentration of template DNA is set such that on average each droplet has only a single template molecule. The PCR reaction is run. Emulsion are very stable and don't break down on standing or even heating.

As each droplet has but a single template and the products of the PCR reaction do not pool together, the rate at which shorter PCR product amplify is reduced. This allows for more longer PCR to be produced.

The difficult is to make a stable emulsion. You need a homogeniser.