Problems with Modified Paired End Sample Prep Protocol

[vas72985]

Hello everyone,

I am having some trouble with the sample prep I am doing pre-Illumina sequencing. I have modified the standard protocol, which I will attempt to explain as clearly as possible. I begin by fragmenting genomic DNA to about 2kb in size. I then do what essentially amounts to T-linker ligation. First, I do a primer extension reaction using a primer that is specific to a fairly abundant region of DNA in the genome (eg, Alu or LINE-1). This selectively adds 3' Adenine overhangs to those fragments that contain the sequence I want to select for, allowing the T-linker ligation to occur in a selective manner. I ligate a duplex T-linker to my fragments. Then I perform a nested PCR, first using the same primer used for the primer extension (and a linker-specific primer). Then, in the second (nested) round of PCR, I add the necessary sequences required for PE cluster generation and sequencing by adding them to the 5' end of my internal (nested) primers.

This second PCR reaction is what goes onto an agarose gel to be size-selected. I cut bands at about 500bp and 750bp and then amplify using the PE Illumina Primers and Phusion Mastermix (as per the Illumina protocol) for 10 cycles to enrich for the DNA. However when I run the samples on both a bioanalyzer and a gel, the fragments that are created by the PCR are not the expected 500 and 750 bp in size, but rather are only about 250 bp (and both are the same). I am at a loss for what might be happening here, and would appreciate any and all suggestions.

Hope I haven't made the explanation of my modifications too complicated.

Thanks!

[ECO]

Thinking out loud here...

Do you have Sanger sequence of any of your fragments to ensure your strategy is working as planned?

What controls have you done? One that comes to mind is trying the PCR with only the ligated adapter primer (I bet you get the same result).

If I understand your setup, you are A-tailing every fragment in your pool....making every fragment a template for ligation. I'm not sure I understand the use of the primer extension to obtain specificity, as Taq or Klenow(exo-) will happily A-tail all blunt fragments.

Thus every fragment will contain the T-adapter priming site at their termini. You could be amplifying (or trying to) panhandle molecules which are simultaneously overwhelming your specificity and self-inhibiting their own amplification (based on the Tm of the regions and the annealing temperature you're using.

This is essentially what's going on in the GenomeWalker kits from Clontech...I suggest carefully reading their notes on adapter design to avoid this...if it's what you're experiencing.

[ECO]

Oh and I'm moving this to the Sample Prep forum.

[vas72985]

I have Sanger sequenced my products and do in fact get what I expect, one set of primer on one side and one set on the other.

I am not ligating my T-linkers to both sides (or at least I dont think I am, and the results of the Sanger sequencing I've done doesn't indicate I am). My protocol only adds my duplex T-linker to the 3' end of the strand since that is the direction of the primer extension (essentially a single round of PCR) and thus where the 3' A overhang is. What I do is a 94 degree denaturation step, a single annealing, a single extension, and then an incubation at the extension temperature to make sure the As have been added. I guess it's possible that the Taq is adding As to the other ends (not in the direction of the primer I have laid down, but again, my Sanger sequencing data to not make this seem like it's occurring with great frequency).

I'll read through the information you have provided and let you know what I make of it. Thanks for the prompt reply!