Xba star activity?

Dennis J. Templeton djt2 at po.CWRU.Edu
Wed Jan 20 10:16:49 EST 1993


In a previous article, komm at a1.mscf.upenn.edu (Barry Komm) says:

>We are currently preparing sub-clones of mouse genomic DNA fragments that
>have been placed in Lambda FIX II.  When we digest with Xba one of our
>clones reveals 6 insert fragments along with the two phage arms.  Note: We
>can take two oligos that we know only match up with our gene of interest
>and amplify the appropriate sized fragment from our genomic clone.  Our PCR
>product reveals the correct bands when restriction digested and our PCR
>product matches with a positive control product produced from mouse genomic
>DNA.  The problem is that when we transfer our DNA and probe with a DNA
>fragment that we know lies within our clone the hybridizing band is not
>visible on the ethidium-stained gel.  Any ideas?  Am I missing something
>that is obvious?
>
>******************************************************
>Thanks from,
>      Barry Komm
>      University of Pennsylvania
>      komm at a1.mscf.upenn.edu
>

I think you are missing the same obvious thing that we missed a couple of
years ago, and figured out after a tip from a friend.

Xba I is sensitive to dam methylation (or is it damned methylation??)  How
does this affect you?  The methylation state of your DNA depends on its
source:

Plasmid DNA grown in "normal" coli hosts will be almost completely
	methylated, probaly more than 99%
 
Lambda phage (because it is synthesized and packaged rapidly) is only 
	partially methylated.  My experience is that a sensitive XbaI site
	will probably be about 25% resistant, but this probably will vary 
	with growth conditions.

PCR DNA should be fully unmethylated unless you use methylated dNTP's

Therefore, an Xba band that you see in your PCR fragment may be visible in
your Lambda DNA (though probably as an overlooked partial). Plasmid
DNA will be almost completely uncut, at *some* Xba sites.  If your
subclone is *almost* fully methylated, you might see hybridization due to
the 1%, yet this faint band will not be obvious on an ethidium gel.  Are
you ignoring strong hybridization near the top of the gel??

I would switch enzymes.  After our experience with this before, we found it
hart to work with an enzyme that wouldn't cut our subcloned DNA, and didn't
want to switch to an dam-minus coli strain.

The NEB catalog has a reasonable description of methylation of restriction
sites, for further information.

good luck, 

dennis



More information about the Methods mailing list