Protocol Online logo
Top : New Forum Archives (2009-): : Molecular Biology

DNA extraction from very small volumes - (May/12/2010 )

Hi everyone,
I am new to this forum and have started working in a lab a few months ago. My project requires me to extract DNA from <100 uL of samples, and there might not be much bacteria in there too. Has anyone optimized very small scale bacterial DNA extraction before? What protocols would you recommend? Links would be very much appreciated :)


Hi Phenol -- welcome to the BioFourums!

What is the nature of your small samples (e.g. stool, blood, sewerage, plant material)?

Are the bacteria therein viable? Can they be grown in vitro?

What will you use the extracted DNA for?


to add to HomeBrews questions: What bacteria are you targeting for?


Thank you for the replies.

My samples are made up in mostly filtered water, and the bacteria is isolated from an aquatic environment. I am targeting the entire microbial community present in this sample, which includes bacteria and archaea.

Yes, they are viable but I probably can't culture them as I do not know the identity of the bugs in the sample and I wouldn't want any bias.. so I am looking to extract their DNA from this very small volume, PCR them and do pyrotag sequencing in order to characterize the community..


still a little bit confused; you are filtering less than 100 ÁL of samples and then extract the dna? When we extract DNA from water we usually use 100 *mL* or more (until the filter cloggs) and then extract the DNA from the filter with a standard CTAB-PCI protocol. Usually this gives a good results (although I am working on eukaryontes).

edit: as far as I understood in your case the primers will be more decisive so to avoid bias I would use more primer pairs; the method for DNA extraction: use a standard protocol used in your lab or used in similar studies.


I won't be using a filter, but different means to isolate bacteria from the environment hence the small sample volumes. I'll probably try out a commercial kit and see if that works first.