Protocol Online logo
Top : Forum Archives: : Molecular Biology

genomic DNA prep (protocol check) - (Sep/04/2006 )

I like to hear about comments on my protocol (E. coli genomic DNA prep for PCR).

I am trying to prepare genomic DNA from E. coli for PCR amplification. It starts with purely mechanical lysis (e.g., bead beating with no detergent or buffers) followed by centrifugation at 12000 RPM for 10 min. The supernatant (may include sheared genomic DNA, plasmid DNA, RNA, proteins) is mixed with a binding buffer in QIAquick PCR purification kit and flown-though a spin column. The wash and elution steps are followed as described in the manual.

The flow-through fraction at elution step may contain sheared genomic DNA, plasmid DNA, and might also some RNA). Again, my goal in this prep is to amplify a template of genomic DNA.

Will this protocol work for PCR? I am very welcome for any comments or experience.

Thank you in advance.


possibly it will work, but if your chromosomal DNA has very much shearing, the result will often be smeary bands


I agree with aimikins. The DNA will be sheared. And without a buffer (or detergent at the very least) to handle the nucleases, somewhat degraded. Plus, without a RNAse step, contaiminated with RNA. The silicon with the PCR purification column will happily bind to any polynucleotide, including RNA. A RNAse step has to be included before adding the mix to the PCR purifcation column... which will only bind well to DNA stands 10kb and below.

How about subsituiting the mechanical lysis with a guanidine thiocyanate based lysis solution? It is gentler and since people use it to extract eukaryotic genomic DNA, I think it would do fine for old E.coli. Add a RNAse step after that followed by phenol chlorofrom extraction. And once that is done, percipiate the mix with 3M Sodium acetate and 100% Etanol.... good grief I feel old and boring.

As for a PCR template... I have the feeling it might work... .but I also feel the DNA wouldn't be usefull to PCR anything big.

This are my thoughts anyhow.