PCR help
February 24, 2012 9:17 PM   Subscribe

Molecular Biology / PCR Gel Electrophoresis Help? Help in elucidation of source of spurious banding?

Hi all,

I am looking for a certain bacterium in soil samples using a degenerative primer set and getting a lot of spurious banding. Attached are images of my gels, the two bright bands are positive controls at ~500 bp. The samples repeat every 15 wells, using different dilutions. I am running a "touchdown" PCR which lowers the annealing temp 1 degree ever 5 cycles starting at 62. These results are perfectly characteristic of the type of results I am getting these past few months. I would like to sequence the fainer bands, but they are so haphazard I am unsure if this is a good course of action. Any help would be much appreciated. Thanks everyone!
posted by Benzle to Technology (8 answers total) 2 users marked this as a favorite
First things of any thing, what are you looking for? A species specific amplification at 500 bp?
If you're looking for an enrichment at 500 bp you've got it in multiple samples. What is it you want to sequence or discern? Relative abundance of that species? It's impossible to tell from these given what a band implies as far as positive vs. negative results. Do you need to do QT PCR, rather than touchdown?
If you're sure that only species specific sequence will produce that 500 bp band, you can simply gel purify and sequence that band... otherwise. without the givens, I can't recommend any particular further action.
Finally, and without malice or judgement: shouldn't you have a mentor or local guidance for your interpretation of these gels? Strangers on the internet are all well and good for some things, but one would hope that you're in a context with more primary experience in this assay.
posted by Cold Lurkey at 9:44 PM on February 24, 2012

"Soil microbial communities are believed to be the most complex on earth. ...with an estimated tens of thousands of genotypes per gram of soil."
Williamson SJ, and Yooseph S. 2012. From bacterial to microbial ecosystems (metagenomics). Methods Mol Biol 804: 35–55.

Degenerative primers are by their very nature unspecific, and so I think that trying to isolate a specific bacterium using them is unlikely to be successful — especially when the sample probably has a very large numbers different species of bacteria in it.

Also, I agree with Cold Lurkey that you'd be better off asking the people you work with than AskMeFi or other internet fora.
posted by davemessina at 2:08 AM on February 25, 2012

What are your controls? You've mentioned positive controls at 500bp, but nothing about negative controls.
I'm hoping you have template alone, and primer alone, and maybe -dNTPs?

The smallest bands could either be primer dimer or RNA.

As mentioned above, is there a way to use less degenerate primers?
posted by sciencegeek at 5:14 AM on February 25, 2012

Some good advice here already. If you can avoid using degenerate primers that would be helpful. Two things you could try are to systematically vary the number of PCR cycles and the amount of product you are loading on a gel.

It may be that you will be stuck with byproduct no matter what, in which case you just want to maximize your 500bp product and get the best separation before doing a gel extraction. On the second gel, Lane 13 looks the most promising from the standpoint of matching your positive control (You overloaded your positive control samples so its difficult to compare). You could vary the conditions around those in Lane 13 to get fewer byproduct bands and to maximize your specific product. You'll also want to run the gel for longer as the separation isn't so good here.

When soliciting help it's good to provide as much info as possible up front. Are you using a 2% gel? What voltage and for how long? How many cycles of PCR? How degenerate are your primers? How is your sample purified?

Good luck!
posted by euphorb at 6:57 AM on February 25, 2012

I'm a little surprised that the bands are pretty distinct instead of a giant smear. That suggests each one band is one of several pretty specific genes that happened to get amplified, major species as it were.

You can cut out each band and gel extract each PCR product. There are a variety of kits, enzymes, and old school extraction methods. Sequence them all.

If you're looking for one specific gene, decrease the degeneracy.
posted by porpoise at 5:17 PM on February 25, 2012

Oh, had a thought; if you're screening for alternate splices or homologues or something, make a cDNA library. It easy - isolate total mRNA from your soil sample, and just add some linkers on your degenerate primers. Do an RT PCR with Superscript, digest the product to open up the linkers, and generate a library. Transform, plate, pick a bunch of colonies and have them all sequenced with an insert-upstream sequencing primer present on your library-generating plasmid backbone. Cheap process.

If you needed to isolate invdividual colonies/clones expressing variants of the gene that you're interested in, memail me, we can work out an arrangement that involves me getting money for me doing that for you.
posted by porpoise at 9:18 PM on February 25, 2012

You could try starting the touchdown at a higher temp - I got good results by starting around 68, and dropping a degree every two cycles. But I think the problem is going to be the primers - if your positive control with single band is the bacterium you're interested in alone, then the bands in the other wells must be because you're amplifying something similar from the other bacteria in the sample, and the best way to solve that is to use more specific primers.
posted by penguinliz at 4:42 PM on February 26, 2012

Not being entirely clear on what you're trying to do, here are some comments and suggestions, from extended personal experience with various PCR problems in a range of experiments:

-More specific primers, with less degeneracy. penguinliz makes a good point about the controls. Along those same lines, the lowest MW bands in your gels make me wonder whether you're getting primer-dimers and other weirdness. Have you checked for hairpins, self- and hetero-dimerizing energies? Nothing's a guarantee, but that'll help you avoid primers that are extra-likely to be frustrating. (IDT's OligoAnalyzer is pretty good.) And while I assume you are, make sure you're already following all the usual primer advice you can (18-21nt (unless there's some reason not to, I use the latter by default for more specificity), avoid runs and repeats, 1bp GC clamp on the end but not more than GCs in the last 5bp, etc.)

-Alternately, if you have little flexibility on this set of primers: nested PCR (if this is compatible with your experiment.) Look for the _best possible_ primers for a region surrounding your gene. PCR out that larger chunk. Then run this PCR, with less background template to cause you problems.

-Higher touchdown start temp. I've gone as high as 68 or 67°C; I drop .5°C every cycle for about 10°C, then do normal another normal 15 cycles at that annealing temp. Longer extension time and higher cycle number are both associated with more unwanted products. Another thing to try, if you have access to a machine that can do gradient _and_ incrementing/decrementing temperature at the same time (some, including the common MJ Research/Bio-Rad machines, don't seem to be able to) is gradient touchdown PCR (in other words, doing tdPCR over a bunch of different starting Ta/final Ta ranges, at once, in case one is really clearly better.) I'll often do gradient screens the first time I do a given PCR reaction (with small 10µL reactions, from a mastermix split into 4 or 5 replicate reactions) and it really can be a timesaver.

-You're working with bacteria, and it sounds like you're likely starting from the genomic DNA, not reverse-transcribing from mRNA or something. Any chance they have pain-in-the-ass high-GC% genomes? If so, DMSO, glycerol, betaine, formamide, etc. may help out. Similarly, there are various polymerase variants (I've worked with KOD and Phusion variants for this, but there are others) that are better for difficult templates. Both of these things can make a big difference. Other PCR components - [Mg2+], [primer] are worth looking at too. While I rarely futz with the former, higher primer concentration can definitely up non-specific products and primer dimers.

-All this other advice to the side, if you're just looking to detect something or not, just run larger lots of your reactions on a 2% gel, gel purify, and sequence the better-looking bands. It's not that time- or money-intensive a process these days.

And yeah, while we all end up struggling through frustrating experiments without much help sometimes, you should definitely find another lab that does this stuff and at least ask them questions (or possibly collaborate with them) if this isn't the sort of thing your lab has much experience with.
posted by ubersturm at 6:16 PM on February 28, 2012

« Older I need pretty hands.   |   What's it like being a UN Volunteer? Newer »
This thread is closed to new comments.