16S library issues

[anatgonca]

Hello to all of you.
I am a newbie in the genomics field, so I address here my problem hoping for some feedback and solving ideas.
I am developing a project for 16S V4 region sequencing to understand the diversity of bacterial community in fish intestines. I use feces and mucus as sample, extract genomic DNA with column purification (kit specific for stool) and after quantify prepare the library with NextFlex kit from Bioscientific. I optimized the PCR cycle number and the template concentration to amplify, and I got acceptable amplicon concentrations (measured by Qubit). The problem starts when I check on a gel or by bioanalyzer I will always have (or almost always) 2 bands/peaks) (1 at 300 bp and other at 400bp). The target amplicon should have about 300 bp so I did a gel purification of the target band but then the concentrations are very low and MiSeq doesn't detect signal and fails to clusterize. I asked BioScientific tech help and they reply it should be from any kind of eukaryotic cells around, but I still think it should not derive in two bands but one.
The options I thought were to prepare a first PCR with universal primers (the same of the kit without illumina adapters) and used as template for the kit library prep, or gel band purification, and I was advice to go with the gel band purification. I did and its always too low yield for Miseq to recognize even with PhiX control added.
So I was wondering if anybody has a clue of what might be happening and any idea how to overpass this. Increasing DNA template concentration is no good since it reaches some kind of limit for the PCR of the kit and does not bring anything new in therms of concentration.
I thank you in advance and sorry for the long text.
Ana

[pmiguel]

I suggest just running the sample with both bands. Even if you have to throw away 1/2 of your data, you will be better off than doing runs with that low an amount of DNA.

--
Phillip

[cliffbeall]

We sometimes see amplification of human DNA with 16S primers, though not at a very high level. In the cases we have seen it's to a random location that happens to be a (possibly partial) match to the primers. You may have just gotten unlucky with your host organism if it has a good match or something that is repeated in the host genome.

I think you might:
1) Try optimizing the PCR conditions.
2) Use phosphorothioate modified primers - for some non-specific amplicons it appeared that the proofreading polymerase had chewed off a base or two at the 3' end of the primer. You might be able to avoid that by modifying the last phosphate bond.
3) Sequence everything and bioinformatically remove the non-specific.

[kmcarr]

...
The problem starts when I check on a gel or by bioanalyzer I will always have (or almost always) 2 bands/peaks) (1 at 300 bp and other at 400bp). The target amplicon should have about 300 bp so I did a gel purification of the target band but then the concentrations are very low and MiSeq doesn't detect signal and fails to clusterize.

I think you isolated the wrong product. The target V4 region (515->806) is ~290bp but you have to consider the added length for the required clustering, priming and indexing bases. Bioo is always very opaque about details of their kits, but assuming they are constructing an final amplicon similar to those described by the Rob Knight or Patrick Schloss labs the expected size is ~400bp.

I'm not sure what the 300bp product you are seeing is but as Phillip suggested, just throw all your product on the sequencer and see what happens.

[Bioo Scientific]

Hi Ana,

You can see the desired product size of a NEXTflex 16S V4 Amplicon-Seq library on page 11 of the protocol. Sorry to hear about the poor results. I recommend you contact us at [email protected] so we can send you additional reagents for the failed reactions.

[anatgonca]

Hi Ana,

You can see the desired product size of a NEXTflex 16S V4 Amplicon-Seq library on page 11 of the protocol. Sorry to hear about the poor results. I recommend you contact us at [email protected] so we can send you additional reagents for the failed reactions.

Thank you all for your replies. I checked before how it should look the amplicon from NEXTflex v4 kit and for that reason I contacted your tech help. The answer I got was that it was weird. That's it. They recommended purify the band but as I explained that protocol would result in very poor concentration. But thanks for your reply I will contact bioscientific to see if I could get some more help or some reagents in replacement of the samples used.
Thank you
Ana

[SNPsaurus]

I know some fish 16S regions give mtDNA artifact byproducts and yield two bands. Selecting slightly different primers helped.

[anatgonca]

I know some fish 16S regions give mtDNA artifact byproducts and yield two bands. Selecting slightly different primers helped.

Thanks! The problem about selecting different primers is that the libraries are prepared with a kit...which includes the primers with the illumina adapters etc so changing these would change kit performance. Still, Do you have any primer recommendation? There are several in the literature so I was wondering if you would have a personal choice that would not have this double band problem. Thanks in advance.
Ana

[gsshankar]

Hi Anatgonga,

I am a postdoc working on soil metagenomics. I saw your post on seqanswers. I am facing the exact same problem that you encountered. I am using the 16s Metagenomics workflow/protocol laid out by Illumina. We are using the Greg Caporaso (515f and 806r) primers with the Illumina adapters attached to them for the amplicon PCR. I am getting a faint non-specifc band (close to 300 p) every time I run the pcr.

Could you please let me know what did you end up doing. Did you run everything in sequencer or did you do gel purification of the desired band?

Thank you very much for your time and help,

Shankar

[anatgonca]

Hi gsshankar

Unfortunately we didn't find a solution to our problem. After cutting the band and purify the concentration was too little even after pcr amplification and we could not cluster properly in the miseq. We ended up doing illumina way and go with their primers and adapters for region V3 and V4 that gave a 600 bp amplicon with excellent quality. Shame we had to spend a lot of money and time in something it just didn't work. But I am still going to give it a second trial in a month or so and will keep you updated. Good luck with your project and if you find a solution also please share.

[microgirl123]

I run a lot of amplicon libraries for different labs, and most of them have problems getting rid of secondary bands with 16s amplicons. Generally, people choose to sequence everything and then discard reads that don't align.

[adam.geber]

Out of curiosity, for folks using the 515f/806r Caporaso primers are you having trouble getting your libraries to cluster despite seeding at the expected concentration? I'm currently doing a gel-based purification to get rid of 300bp mtDNA amplicons after 16S PCR and this yields libraries of decent concentration (>4ng/uL) according to Qubit readings but very low molarity (<500pM) by library quant qPCR. I'm not sure what to make of this discrepancy.