Library on gel but not on Bioanalyzer

[pauji]

I'm having a strange problem with my library prep for Illumina. Unfortunately I have no been able to find any information on this, not here nor from Agilent, so i have to assume that i'm the only one with this problem...
maybe i can get some feedback from you guys

My post-PCR RAD libraries (i.e. dsDNA 300-500bp) are clearly visible on my agarose gels
but hardly or mostly not at all on the DNA1000 Bioanalyzer chips

The Bioanalyzer runs look fine, i.e. the ladder is excellent and I cannot detect any problems with the lower and higher marker peaks, i.e. the matrix and peak detection seems to work fine. Between those peaks though the electrogram is completely flat, no DNA at all by now I have at least 20 different PCR products that I can't visualize on the BioAnalyzer. So my PCR apparently works but without the Bioanalyzer I can't sequence

Has anybody seen/heard of this before? any idea what I can do?? I guess the problem is most likely with the Bioanalyzer but...
any comments appreciated!!
cheers
P

[vkkolla]

Hi P, try it on DNA High sensitivity chip.
Cheerz..

[pmiguel]

There are lots of possibilities. Here is a list in order of what seems most likely to me:

1. You are diluting your DNA before loading it on the BioAnalyzer DNA chip. If so, reduce the dilution.
2. Your agarose gel is more sensitive than the BioAnalyzer chip. This could happen if your BA reagents are expired, or near their expiration date, or if you gel electrophoresis parameters are tuned well and you have a good image documentation set up. Hard to say though because you gave absolutely none of your gel electrophoresis parameters. If you know the mass of your MW ladder for the gel, you could estimate the concentration of your PCR products. Then see if what you loaded is within the detection limits of a DNA chip.
3. Your agarose gel loading dye or PCR mix is contaminated with something you are interpreting as your PCR fragment. Try running a "mock" lane, where you set up a PCR reaction, with no template DNA. Then, with no thermal cycling, prep that the same way you do real samples and load it on the gel. Use the same tips and pippeters! (They may be the source of this contaminating band.)
4. Your PCR products are single-stranded. This could happen if one of your primers ran out before the other. All subsequent amplification cycles would linearly increase the concentration of only one of the strands. Single-stranded molecules migrate more slowly on BioAnalyzer chips than with gel electrophoresis. You could test this by running some of your samples on an RNA chip.
5. Okay, this is the most crazy one, and almost certainly not what you are seeing. Years ago I loaded some uncut plasmid preps on a DNA chip. I could see these clearly on an agarose gel. But saw nothing on the DNA chip. Maybe the supercoiling caused them not to bind any of the dye? Anyway, if your bands are actually somehow became circular and supercoiled... This is testable, but so far-fetched I am not going to comment further.

[pauji]

Thanks for your input guys, i really appreciate!

as of your comments:
a) will do! couldn't do it so far because we don't have them in my lab and agilent was unwilling to give me a sample...


b) lots of good ideas, thanks!
1. no dilution.

2. yep, my agarose gel is more sensitive than the BA chip! i just don't get why
the DNA1000 is supposed to detect 0.5ng/ul, Etbr-agarose gels usually require ~10ug... according to my guesstimates based on ladder intensity i have plenty of DNA to detect with the BA but i don't! i considered old reagents but the beautiful marker peaks and BA ladder seemed to indicate that everything is working fine... still this is the most parsimonious explanation for me.
positive gels were obtained with very different gel parameters, i.e. 1-2%, 100-200V, 30-60min, different loading dyes...

3. yep, i considered that too but that'd mean tons of contamination at just the right size + my negative controls are fine (i.e. empty) + i should detect the contamination on the BA as well + no clear contamination pattern etc. etc.

4. i read about ssDNA BA problems but i though PCR products are by definition dsDNA... single-stranded PCR products seemed like an oxymoron too me! definitely a cool idea, thanks! i obviously can't exclude that but it seems unlikely given that a) the band is just the right size on the gel (i.e. not faster) and i have nothing on the BA (i.e. no big unexplained band)

5. also a very cool idea that i hadn't considered!

as of now, i looks like i'm having a fairly rare problem that might cannot be easily explained.
i think i'll ask a friend in another lab to run a chip for me on another BA with other reagents... i.e. changing everything and hope that everything changes :-) i'll report back!

cheers!
P

[Susanne]

In case you are dealing with a smear, your sensitivity of 0.5 ng/µl for the DNA1000 doesn't hold true, as these are defined to be in a (sharp) peak.
For a smear, the sensitivity will be lower (maybe ~10fold).

[miss_shell]

I agree with Vkkolla. We run all of our post-PCR samples on a high sensitivity chip. Should solve your problem.

[pauji]

hey guys

looks like the price goes to... vkkolla & miss_shell for suggesting the high sensitivity chip... it showed beautiful libraries, which i accepted instantly

after reading up a bit more here, i found out that most people do qc with BA chip but don't trust its quantification at all. subsequent quantifications of my libraries with the qubit system were 2-10x higher than BA quantifications and coincided much better with gel observations. I thus went with that and am now waiting for the results from sequencing.

it doesn't all make sense to me but it doesn't have to as long as i works ;-) otherwise, i'll be back soon...

wish me good luck :-)
cheers
P