Adaptor ligation (View forum version)


Posted 07 June 2013 - 04:23 AM

Hi there,

I am trying since some time to ligate adaptors to blunt ended sequences, but so far no success.

I am using DNAse1 to randomly cut my sequence of interest and T4 DNA polymerase to fill the ends. Then I add two annealed adaptors, one is 5'AAACCTCTCCGGAGCCTGA 3' and /5'Phos/TCAGGCTCCGGAGAGGT 3' for the 5' end and 5'GAGGTGTCGCCGGTCA 3' and /5Phos/TGACCGGCGACACCTCTCCGGAAAA for the 3' end and ligate O/N at 16C. Next I wanted to do a PCR using the unphosphorylated adaptors as primer, but no product.

I have also tried to use Taq polymerase with random primers to add an A overhang to the digested sequence and ligation with single T containing adaptors as well as blunting and ligation of didesoxyTTP together with an A overhang, but nothing. The ligase works well with all other stuff.

Any idea what to improve? I hope it's a stupid mistake, like adaptor sequence or too high concentration of DNA...

Thanks a lot,



Posted 08 June 2013 - 01:59 AM

Blunt cloning in general is weak and slow, takes a lot of optimisation. couldn't you make randomly overhanging adaptors and ligate those onto the overhanging ends you are generating in the first place?


Posted 10 June 2013 - 06:53 AM

Thanks for your answer. It is a really good idea to use adaptors with random overhangs. How many random nt shall I use, is 3 enough for ligation? Then I may preserve the ORF... The only problem I can see is that I don't know if the overhangs I generate are 3 or more bp long. I use MnCl2 to achieve a more proximal/blunt digest.

Do you think that conditons of adaptor annealing are important? I usually mix 1ul of 100uM for/rev adaptor in water, 95C for 3mins and cool to RT with 0.5C/sec. Most people use some buffer (1x ligase buffer) and a slower cooling rate...




Posted 10 June 2013 - 12:31 PM

The buffer might be important, the dissociation rates and annealing rates of oligonucleotides are affected by the salt content. It is also a good idea to store your DNA in a buffered solution to help prevent things like acid hydrolysis.


Posted 13 June 2013 - 04:56 AM

Hi bob1, it's me again. Ligation of random overhangs did not work either. I used AAACCTCTCCGGAGCCTGANNN for 5' end and GAGGTGTCGCCGGTCANNN for 3' end together with above phosphorylated nucleotides. This time annealed slowly in 1x ligation buffer and ligated to a DNAse cut sequence directly. After O/N ligation, PCR using primer AAACCTCTCCGGAGCCTGA and GAGGTGTCGCCGGTCA (without N) gave no product at all. Actually, can I expect a PCR product if ligated? Shall I use only one random overhang? And how about purification of the phosphorylated primers, I used desalting not HPLC which some people recommend.
Sorry, I am a bit desperate :) Thanks for your help.


Posted 13 June 2013 - 06:21 AM

This problem is almost identical to the process of preparing sheared DNA fragments for Illumina sequencing. I'd suggest that you find one of the preparation kits for end repair and adapter ligation. Virtually all suppliers have these (NEB, Epicentre, LifeTech). DnaseI may be a bad way to shear DNA. Usually sonication, nebulizers, or mechanical hydroshearing are used. What is your end goal?


Posted 13 June 2013 - 07:40 AM

Hi phage434, thanks for your suggestion. I had a look into the Illumina kits, which are extremely expensive. NEB is an alternative, to me the protocol sounds like blunting, terminal deoxy transferase treatment and TA cloning. I have tried something similar and hope that you can answer one question... To avoid ligation of multiple A to the blunt DNA one uses dideoxy ATP, right? But the one hydroxyl group that is missing is required for ligation, so how can ligation happen? Or am I mistaken...? The NEB kit uses dAMP, does this prevent concatemerization as well?
My goal is to integrate a defined stretch of amino acids into a target peptide at a random position (insertional mutagenesis). The target DNA is ~750bp, circular. The nucleotides are 27nt long, and I'm struggling with it since almost a year :( Fortunately I have other projects as well. I would be very happy if anyone can suggest something.


Posted 13 June 2013 - 12:19 PM

I believe Klenow exo- in the presence of dATP will add a single A to a blunt fragment. If you have a dsDNA 27 bp fragment (made by annealing two ssDNA fragments) which has a single base T overhang at each end (and a 5' phosphate on each oligo) then I think you should be able to ligate this into your cut fragments. A major issue is getting the concentration of the adapter right -- you need MUCH less than you think, since you need about equimolar amounts of your insert and your target.

I suspect that a combination of blunt end polishing, followed by A tailing, followed by ligation would work. Do you have a way of selecting for success? How will you know you have the correct result?


Posted 14 June 2013 - 04:48 AM

Yes, I eventually realized that I used way too much adaptor and calculated the amount as equimolar to my starting DNA (I'd expect to lose quite a bit during prep). The last time i had a titration of 1-10-100-1000 fold dilution, but none gave anything.
My selection would be a PCR with AAACCTCTCCGGAGCCTGA and GAGGTGTCGCCGGTCA, which are part of the original adaptor duplex. Do you think I can get a PCR product? Maybe this is the problem that my detection method doesn't work...


Posted 14 June 2013 - 05:00 AM

phage434, you suggest to ligate the complete 27nt into linearized DNA. My plan was to add one part of it to the 5' end and the other to the 3' end, and make a PCR to select for ligated events. I am not sure how to select for ligated events otherwise... Contamination by not-DNAse cut circular DNA is major issue I guess. Hm, PCR with part of the adaptor and Dpn1 treatment to get rid of the old DNA?