15 replies to this topic

[KimWG]

Hi All

I am working with some VERY difficult template in a multiplex PCR (6 primer pairs, with products from 125-495 bp in size).  I have spent A LOT of time optimizing conditions -- I think they're as good as they're going to get.  My difficult template is bat feces, and I'm trying to amplify insect DNA for dietary studies.

Anyway, like I said -- I don't think my PCR's are going to get much better.  I find that I sometimes have VERY faint bands -- whether this is due to the presence of inhibitors, low template quality, or amplification of non-target insect DNA is still up in the air.  (It probably is all three.)  I often find myself inspecting photos of my gels with extreme magnification and meticulousness, because sometimes my bands are so faint that they could be easily overlooked.

I'm making my gels with TBE, 2% agarose, and Invitrogen SYBR safe stain.  I usually use small wells and load 5ul product from a 20ul reaction.  What can I do to improve the visualization of weak product?  Would you more experience folks suggest increasing the amount of product I load?  The amount of stain included in the gel?  A different buffer?  Should I switch to EtBr?  (This would cause some hazardous waste issues to arise because our lab doesn't use EtBr anymore.)  Post-staining the gel?  (Has not been very successful for me so far.)  Downloading all files to my computer and manipulating the files in photoshop to somehow better visualize the results?  

I'll be grateful for any tips you all can suggest.

Thanks!
Kim

(P.S. If you want I can include files of the gel pics to demonstrate just how weak the bands are.)

[hobglobin]

KimWG, on Apr 28 2010, 08:10 PM, said:

I'm making my gels with TBE, 2% agarose, and Invitrogen SYBR safe stain.  I usually use small wells and load 5ul product from a 20ul reaction.  What can I do to improve the visualization of weak product?  Would you more experience folks suggest increasing the amount of product I load?  The amount of stain included in the gel?  A different buffer?  Should I switch to EtBr?  (This would cause some hazardous waste issues to arise because our lab doesn't use EtBr anymore.)  Post-staining the gel?  (Has not been very successful for me so far.)  Downloading all files to my computer and manipulating the files in photoshop to somehow better visualize the results?

That would have been my first idea, to increase the volume of PCR product, to have more DNA on the gel...at least worth a try.
Reducing the "background noise" of stain, though I've no idea if it's a problem with SYBR. This can be also done with some of the gel documentation software, by changing camera chip sensitivity, playing with brightness, contrast, shutter speed, digitise borders...
And getting rid of inhibitors with a different DNA extraction method, or additional washing steps, if you didn't try out already, might be an idea. Some kits are especially for feces DNA extraction...
Reducing the amount of template DNA in PCR to reduce inhibitors might also be possible, if the latter won't work.

Edited by hobglobin, 28 April 2010 - 11:53 AM.

[Lapsang]

Yes, I'd agree with loading more sample to the wells (as long as this doesn't mean you don't have enough to use for other purposes you had in mind).

I guess you're already using quite a high exposure time?

And as long as the bands can all be seen normally (even if faint), enhancing them on your computer should be easy if you wanted to.

Edited by Lapsang, 28 April 2010 - 12:37 PM.

[phage434]

Also, narrow wells filled with sample are better than wide wells with the same volume.

You could try silver staining the gels, or switching to PAGE gels and silver staining those gels.  Silver stains are far more sensitive than other stains.

A flat bed laser scanning gel imager might produce better results, such as the Typhoon imagers.

You could also try Agilent or Caliper microchannel gel running systems, such as the Bioanalyzer.

[KimWG]

Thanks for your suggestions all -- I will be trying them out shortly.  I've been running 20ul reactions in order to economize but I may double that in order to increase the amount loaded.  I'm not familiar with silver staining, I will definitely look into it.

[georgiadave]

I have found that thinner gels give better resolution.  Also, don't forget to wash your tray, comb, etc... after every run, it cuts down on background stain considerably (i do post stain with EtBr).

[KimWG]

Has anyone here tried the SYBR Green I and/or SYBR Gold for staining gels?  According to the Invitrogen website both are more sensitive that EtBr and the SYBR Safe stain.  (More expensive, too, go figure!)

[HomeBrew]

Have you tried increasing the number of cycles in the PCR?

What method are you using to extract the DNA from the fecal samples?  We've had good luck extracting DNA for use as PCR template from mouse and human fecal samples using the ExtractMaster fecal DNA extraction kit from Epicentre.

[gebirgsziege]

Together with my previous posters I think you could probably optimise the DNA extraction and therefore reduce background.  But it is difficult to detect the DNA of the prey in droppings. How old were your samples when they were collected? I had the experience that the insect DNA degrades rapidly which is probably causing your problems. Bat feces is usually not a too bad template (bad but there are worse things to get DNA out from).

You are using a six primer multiplex....would it be possible to "split" into two multiplex PCRs, so you can modify the conditions for the primers better and improve the yield? Or is ther a possibility for a nested approach (like unspecific primers for your gene first for 5 - 15 cycles, cleanup, and then multiplexing)?

For better visualising: loading the whole 20µL on the gel might help or longer staining times. Small combs, thin gels, and not too much stain to reduce the background.

[KimWG]

HomeBrew, on Apr 29 2010, 03:10 AM, said:

Have you tried increasing the number of cycles in the PCR?

What method are you using to extract the DNA from the fecal samples?  We've had good luck extracting DNA for use as PCR template from mouse and human fecal samples using the ExtractMaster fecal DNA extraction kit from Epicentre.

Thanks for the tip.  I have tried the Qiagen and Zymo stool kits, but with limited success.  The Qiagen kit *does* work for some samples that just won't amplify with anything else (inhibitors) but since it is designed for use with much more feces than I generally have to play with (they want about 200 mg; my bats weigh about 4 mg, so that's a pretty big dump for a pretty tiny bat) I've been reluctant to use it.  I've consistently had the best results with the Qiagen DNEasy kit, and/or the Omega Insect DNA kit.  I've also been able to get DNA to amplify with the Epicentre plant kit (insects have lots of polysaccharides, as do plants) at times, so I will definitely check it out.

[KimWG]

gebirgsziege, on Apr 29 2010, 04:36 AM, said:

Together with my previous posters I think you could probably optimise the DNA extraction and therefore reduce background.  But it is difficult to detect the DNA of the prey in droppings. How old were your samples when they were collected? I had the experience that the insect DNA degrades rapidly which is probably causing your problems. Bat feces is usually not a too bad template (bad but there are worse things to get DNA out from).

You are using a six primer multiplex....would it be possible to "split" into two multiplex PCRs, so you can modify the conditions for the primers better and improve the yield? Or is ther a possibility for a nested approach (like unspecific primers for your gene first for 5 - 15 cycles, cleanup, and then multiplexing)?

For better visualising: loading the whole 20µL on the gel might help or longer staining times. Small combs, thin gels, and not too much stain to reduce the background.

Thanks... I do feel like I've spent a lot of time (and even more money) trying to find the best extraction protocol, only to return after 2-3 months of tinkering to what I was using in the first place!  My samples are not that fresh (between 2-3 years old), and they were stored in a substandard freezer in a tropical country in a house with regular power outages, so... some of it is just the nature of the beast.  I may try a nitrogen tank in the future (another researcher did do this for her plant samples).

I guess I could split up the multiplex, although the six different primers seem to play nice together with regard to to positive control.  I think I will try another suggestion of increasing the number of cycles.  My concern with a nested approach is that I would have to amplify a fairly large chunk of the COII gene to get all of the primers in there, and I don't want to go much about 500 bp becuse the recovery does drop -- I've had very poor luck trying to amplify anything longer than 700 bp.  But it is certainly worth a shot.  Thanks!

[leelee]

Everyone else has really good suggestions for trying to optimise your results, so I can't add anything there.
But I did want to say, DON'T PHOTOSHOP your gel pic if you are going to use it for publication- its just plain dodgy!!

Good luck getting it to work

[KimWG]

leelee, on Apr 29 2010, 07:02 PM, said:

Everyone else has really good suggestions for trying to optimise your results, so I can't add anything there.
But I did want to say, DON'T PHOTOSHOP your gel pic if you are going to use it for publication- its just plain dodgy!!

Good luck getting it to work

So I can't draw the bands in that I want?  

[lab rat]

Maybe this is a silly question, but can you add the SYBR to the loading buffer and samples, and then load them on a plain gel? Maybe that could decrease your background and improve visualization of the samples.

edited for grammar.

Edited by lab rat, 29 April 2010 - 08:37 PM.

[leelee]

KimWG, on Apr 30 2010, 12:13 PM, said:

leelee, on Apr 29 2010, 07:02 PM, said:

Everyone else has really good suggestions for trying to optimise your results, so I can't add anything there.
But I did want to say, DON'T PHOTOSHOP your gel pic if you are going to use it for publication- its just plain dodgy!!

Good luck getting it to work

So I can't draw the bands in that I want?  

hahahahaha apparently not- go figure??