Two band/peak corresponding to 400bp for V4 region

[Naarkhoo]

We are trying to sequence (illumina - miseq) bacterial population from mouse feces. We are amplifiying the V4 region, and the problem is that instead of having only one band/peak corresponding to about 400bp for V4 region, we have this band plus a smir on top of it, of higher molecular weight. I was wondering if this is an artifact of something else which could be amplified, like the mouse genomic DNA for example.

I was wondering about the V4 range of size variation that could be found in mouse feces(is it just a few base pair, or hundreds of it).

We are using Patrick Schloss protocol, using promers 515F and 806R targeting v4 region of 16S

515F AATGATACGGCGACCACCGAGATCTACAC XXXXXXXX TATGGTAATT GT GTGCCAGCMGCCGCGGTAA (red is targeting V4 region, blue is sequence to anneale to flow cell, green is pad, X index)

806R: CAAGCAGAAGACGGCATACGAGAT XXXXXXXX AGTCAGTCAG CC GGACTACHVGGGTWTCTAAT

Could it be because of mitochondrial genes ? I wonder how variable in size is 16S?

[kmcarr]

How did you analyze the PCR products, Bioanalyzer, gel? Whatever the method could you please post an image? It's very difficult to diagnose the issue based just on a description.

All that said I'll offer a possibility. We sometime see apparent higher molecular weight "humps" in Bioanalyzer or Caliper LabChip traces for 16S V4 amplicons. Our conclusion is that these are most likely caused by over amplification and resultant heteroduplex or other tertiary products which don't resolve. Backing off the PCR cycles typically clears them up, but even using the original material for sequencing normally works fine.

[Naarkhoo]

We did bioanalyser, here is the image, and qbit for quantification. So you think that may be due to overamplification? We did 35 cycles, as said on the protocole, so maybe it was too much!



Do you think that would be allright though? I could redo it, but as we have 350 samples that would be really time and cost consuming (of course, our main interest is having a very good data). The thing is this artifact takes for some samples more than ¾ of the concentration.

My second concern is that I pool every samples according to their global DNA concentration, but the proportion of the artifact is not the same for all samples, so the final concentration of the good product of 390 bp will not be equal for each samples in the pool.

Do you think it could be a major problem?