Library PCR enrichment questions (with image!)

[ZAAB]

Hi All:
I'm preparing libraries (five of them)...all are PCR products to begin with...so size is very uniform, and I avoided the fragmentation and end repair. I picked up at the A-tailing, moved forward to ligating adapters, gel purified 'em, and moved on to PCR.

Attached is a gel showing sets of five:

First set: Starting material (insert).
Second set: After ligation of adapters, gel purified (thought I did a good job).
Third set: after 10 rounds of PCR.

Ladder is 766, 500, 350, 300, 250, 200(bright), 150, 100, 75, 50, 25.

My Questions: What are the large bands? I can see my correct bands within the haze, floating at ~300. Should I gel purify again? Anyone seen this before?

Thanks.
ZAAB

[Heisman]

How long was the adapter ligation step? Looks like you have concatemers of inserts and/or adapters. If you run the gel a lot longer and extract the bands at the correct desired sizes, you'll probably be fine.

[ZAAB]

reply

Thanks for the reply.
Adapter ligation step, per the manual, is 10 minutes...(seems WAY too short for me).
Recommendations? I was thinking o/n (like I do for any plasmid constructs).

I've run the whole PCR rxn out on a gel to perform another gel extraction...

Funny thing, though...With SYBR gold as the stain, opposed to the EtBr used for diagnosis (shown in attached gel on original post), the remaining PCR rxn runs as one fat band, just like the adapter ligation purification.

So I am now running an EtBr preparatory gel...and will cut out the correct bands.

Thanks
Z

[Heisman]

If the PCR works then the adapter ligation had to work to some extent.

Is this an Illumina library prep? I've routinely used 15-30 minute adapter ligations with no issues.

How much starting material did you use? And what manual did you follow? Were there any modifications?

SYBR gold stains single-stranded DNA a lot better than EtBr, I believe, but otherwise I can't think of a reason why it would run differently.

[ECO]

Looks like insert concatemers to me...

The mobility differences could perhaps be because SYBR was post-stained (not affecting mobility) and EtBR was in the gel?

[ZAAB]

Nah...SYBR or EtBr were both in the gel prior to running it.

15-30minute ligations...OK. thanks. And yes, I agree that some degree of ligation occurred since there's an increase in the amount of correct size fragment.

Its the TruSEQ Illumina prep kit. Following it verbatim other than the Minelute (vs Ampure) because I determined that the insert size was below the threshold of Ampure...nice waste of cash buying the ampure.

Maybe the DNA size is big enough now, but don't think I'll waste my time extracting DNA from the Ampure washes.

Thanks Guys.
Z

[sehrrot]

Dear ZAAB

I've also seen this issue for long time. Please read my recent post, which I saw same thing on bioanalyzer. http://seqanswers.com/forums/showthread.php?t=11886

I also assume, as others, that it might be a problem of long ligation incubation. When I firstly made Truseq library, I did 30 minute incubation as I thought it's too short as well as old PE library protocol recommends. But I saw very definite dimer on bioA profile and gel. (I also used to assume that it might be a secondary problem in the bioA but bioA reagent includes DMSO to prevent 2nd structure and gel showed big fragment, which means that it's not simply random hybridized fragment.)


As I wrote in my recent post in this forum, I did PCR optimisation and got conclusion, which I should use around 3 ng PCR input to avoid this big fragment issue. I guess if we just use whole genomic sequencing, it's not gonna be a problem as bridging amplficiation will sort out big fragments by size. But if we use exome sequencing, which needs hybridisation with probes, it might be a problem that probes can bind to big fragment more than aimed fragment (actually, I've seen this one and probes was bound to big fragment so overall capture efficiency reduced).

I still don't know what exactly this big fragment is. I cut this big fragment and did qPCR. the result was really lower molarity than I put. Also, on the sequencer, it didn't make enough cluster generation. So, I guess truseq library prep provides very strong ligation step, as a result, some fragment can be ligated randomly to other library if ligation time is longer.