Help with Lysis - (Nov/25/2013 )

Thank you Wek,

Well as per the instructions I should use around 5 ug of proteinase K from 20 mg/ml stock solution. I went as I high as using 50 ul of the stock solution. That is 10 times higher. My problem is that sample stays unchanged after incubation. Proteinase K doesn't seem to digest the feather quill samples. But papers state that it should. When I proceed with phenol:choloforom extraction using the supernatant I can't get any DNA. So I think the problem is most likely the lysis step.

I don't think Proteinase K will work on stuff like keratin very fast and efficient so you have to increase incubation time for sure. Higher temperature and up to 1 % SDS (as suggested before) also increase activity of the enzyme.

How do you measure the DNA concentration that comes out of these attempts? The final concentration might be so low that it's not visible on a gel but a PCR works anyway with it (here more sample and grinding to a powder as suggested helps to get the higher possible yields).

So I'd do a PCR with sufficient template volume to be sure that there's really nothing.

No bands after pcr either. So far I realised that adding DDT to lysis step works for digesting the sample. But still no DNA yield. Maybe it is the feathers I don't know. I am still trying whatever I can think of. I hope there was a way to debug this whole process.

You had a positive control of course and also varied the DNA amount?

And do you know this paper? Perhaps it helps.

Shelley Bayard de Volo, et al. (2008): An improved extraction method to increase DNA yield from molted feathers. The Condor 110: 762-767
 

I am bumping my thread with news. Bad ones actually. Yes, I 've contacted Dr. Shelley Bayard for help. She kindly responded and tried to help me. But no matter what I try phenol:chloroform:iaa method kept failing on me. So as a last resort I bought Qiagen DNeasy Blood and Tissue Kit. Guess what? It didn't work either. I've tried all kinds of modifications to the original protocol of the kit that I could find in literature but still no DNA. I guess, somehow environmental factors affected the feathers really really bad. Now, I have to figure out what might've had an effect so big that all feathers lost all the DNA.

I'm sorry to hear that.  Exposure to heat, moisture and light can all damage DNA, but if the feathers are reasonably fresh I would expect there to still be some DNA there.

Have you tried any of the methods used to extract ancient DNA?

I second bob1's suggestion.

You could also look at DNA extraction protocols from soil bacteria.

No, not really. I've only tried two methods that are reported to work on molted feathers; Phenol:chloroform:iaa and DNeasy kit with all kinds of modifications. I am not aware of the protocols that you speak of. I only know phenol:chloroform method for ancient DNA. Plus, I've really hit the limit on my budget. I would like to keep working but I am also nearly out of samples. I guess I have to do some thinking and re-shape my project.

Samples were collected near a lake which is one of the most important wetland sites for birds. So it is possible that the effects of light and moisture are the main cultprits. I will compare the sampling sites of previous works that were successful with my site to see if the conditions of my site could be considered extreme.

Both protocol deal with removing humic acid which is an environmental contaminant that severely inhibits the PCR reaction. And the ancient DNA protocol probably has ways to amplify damaged DNA strands.

However since you are nearly out of budget and samples. Could you get some bird feathers, (chicken, or duck from a local pond). Crush the feather in liquid nitrogen. Mix the crushed mixture in maybe 100ul of TE. Mix and then pellet down the debris. Then take a small sample of that solution and make a 1 in 20 (to 1 in 50) dilution with water. And use 1ul of that diluted solution in a PCR reaction. 4min heating at 95C before going into amplification cycles the annealing cycle, extension and denaturing.  Run maybe 35 cycles and use a 50ul PCR volume.

It is long odds. But PCR buffer does have detergent in it and together with preheating at 95 C for 4min should rupture cells. The dilution would hopefully dilute out the contaminant in your sample. And lots of cycles might be able to pull out a signal