Very low efficiency adapter ligation with amplicons and PCR free

[Kurt Lamour]

My first post/question. Trying and failing (4X now) to make PCR-free Illumina libraries with PCR amplicons (gel purified and sized 135-175bp), using KAPA PCR-free kit. My input is 20ng/ul based on Qbit = 1ug total. Using 15uM adapter = 0.68uM final concentration. qPCR tells me I'm getting 1nM properly adapted fragments. Should I double adapter concentration? Any clues why my efficiency is so poor?
Egel pic (with hard to read tapestation) attached

[nucacidhunter]

I wonder if you have asked KAPA techsupport.

[Kurt Lamour]

talk to them daily over the past few days. They are friendly, but no suggestions I can implement

[Kurt Lamour]

If I hadn't been successful at this a year ago (with their old kit, not hyper) - I'd think I'm just crazy. Went ahead an re-annealed my adapters today.. for good measure and am currently making yet another library with 1/3 increase in adapter concentration

[bunce]

Hi Kurt,
Not sure if this will help but here is my 2cents worth.... The entire ligation reaction will be inefficient, most of the template will not ligate 'correctly'. The proportion you do get to work (as you point out in your last post) is heavily influenced by the adapter/amplicon ratio. The solution is to try out different proportions and see how the qPCR looks (but you might need to size select to get a 'true' reading - see below)

In our lab we use the Kozarewa (http://www.ncbi.nlm.nih.gov/pubmed/21431776) method - took us a bit to dial it in but we routinely start with 1ug amplicon and end up with ~10 times the amount we need to feed the MiSeq. We use a Pippin prep post ligation to size select the library and this helped a lot to clean up the artefacts.

Best of luck with the workflow! Cheers Mike

[nucacidhunter]

1- I assume that PCR amplicon has been size selected using gel, so there is no carryover of polymerase or other impurities. It is important that DNA is washed properly and eluted with appropriate buffer at this step according to kit manual (avoiding EDTA containing buffer).

2- This is a one tube protocol and after adapter ligation you should have at least the input amount of DNA present. Fluorometric quantification immediately following ligation and after ligation clean up would give some idea on DNA loss. We can ignore contribution of adapters because they are only partially double stranded.

3- Since input DNA is already size selected before end repair and A tailing reaction, there should not be large loss.

4-More than 15-20% DNA loss would be indicator of inefficiency during ligation reaction clean up step.

5- If loss is not significant, protocol is followed and thermocycler performance is within specification, then KAPA should provide some assistance. Either end repair and A tailing step or ligation is inefficient.

[Kurt Lamour]

This is all really helpful for me to consider (nuca and Mike) - quick question for Mike - is it crucial to use the Pippen? In the past, when I've had success, I never had enough yield to afford a size selection following the ligation and bead cleanup. I do a 0.8X bead cleanup to get rid of adapters/dimers (because I don't have Pippen).

[nucacidhunter]

Trying and failing (4X now) to make PCR-free Illumina libraries with PCR amplicons (gel purified and sized 135-175bp), using KAPA PCR-free kit

You already have size selected your input, so there is no need for size selection with any means. Removing un-ligated adapter is suffice.

[pmiguel]

My first post/question. Trying and failing (4X now) to make PCR-free Illumina libraries with PCR amplicons (gel purified and sized 135-175bp), using KAPA PCR-free kit. My input is 20ng/ul based on Qbit = 1ug total. Using 15uM adapter = 0.68uM final concentration. qPCR tells me I'm getting 1nM properly adapted fragments. Should I double adapter concentration? Any clues why my efficiency is so poor?
Egel pic (with hard to read tapestation) attached

PCR amplicons? Are your PCR primers 5' phosphorylated? If not then the KAPA kit would need to include a PNK step, or the ligase would not have a 5'phosphate to ligate the 3' hydroxyl of the adapters to.

--
Phillip

[nucacidhunter]

By definition all End Repair mixes used for Illumina library prep contain PNK for 5' phosphorylation. But it may not work if you have some kind of modification in primer 5' end such as blocking.

[pmiguel]

By definition all End Repair mixes used for Illumina library prep contain PNK for 5' phosphorylation. But it may not work if you have some kind of modification in primer 5' end such as blocking.

By "definition", eh? I don't think you'll get much mileage in science relying on what should be in a given kit component "by definition".

Another possibility is that because no fragmentation was done, the OP may have skipped the "End Repair".

--
Phillip

[nucacidhunter]

By "definition", eh? I don't think you'll get much mileage in science relying on what should be in a given kit component "by definition".

Another possibility is that because no fragmentation was done, the OP may have skipped the "End Repair".

--
Phillip

Thanks for advice on my liberal use of vocabulary. Kapa hyper or other PCR free kits specification has mentioned PCR amplicons as starting material. I agree that Kurt may have skipped end repair and A tailing reaction(s) required before adapter ligation step.

[jiggydancer]

I'm going to assume he ran the protocol properly.

In which case the problem is that there isn't enough adapter

My experience with amplicons is that they actually generate much more library than a comparable input of fragmented genomic DNA.

I think 15 uM of adapter is appropriate for 50-100 ng input of amplicon, not 1000 ng.

Lower your input and try again.

[Kurt Lamour]

Just to finish this one up - after talking to KAPA and trying lots of different things - the thing that worked was to re-anneal my adapters. I did this by ramping slowly (0.1C/sec) from 95C (incubate 5 min) to 4C. fixed everything. Appreciate thoughts on site!

[SylvainL]

When I was still at the bench, and when we wanted to be sure that primers will anneal completely, we used to boil the mix (50-50) for 10 min, then transfer the tube in a room temp water-bath, cover it with aluminium and let it overnight... Acrylamid gels (both primers were radioactive) proved that we got 100% double strands...

my2c