Re. Southern hyb. - Stringency question

Wed Sep 28 13:09:29 EST 1994

Yng Chen writes:

> In Southern hybridization experiments how important is hybridization
> temperature?  What are the effects of rasing or lowering the hybridization
> temperature?

One generally hybridizes 15-20 C below the Tm to get favorable kinetics.  So
the exact temp. is not critical with respect to enforcing stringency.  The 
only exception is if you're trying to do something at very low stringency,
then you may need to lower the hybridization temp. below standard conditions.

> How does one determine the proper washing conditions (assuming 100%
> match in base paring). [for oligo probes]
> If you get a weak signal and you know that you have a lot of DNA
> on the membrane and that you have good specific activity on your
> probe, how do you know if this is because your washing conditions
> are too strigent or that the signal is non-specific.

You start with a calculation, and then confirm by doing a temperature
series against a cloned control target.  If you're trying to discriminate
a one base change, then you'll also need a control with the one base
mismatch.  If you can't come by the controls, then you increase stringency
until you get a single band (assuming you're after a unique target), and
then try to confirm its authenticity.  A common confirmation would be to
try to hit the same band with two different oligo probes.

> In Current Protocols In Molecular Biology they listed formula for 
> melting temperature of hybrid nucleic acids:  <snip>

Yes, these are reasonable estimates.  I prefer to use NBI's oligo
program (no affiliation) which takes the exact sequence into account.
However, the simple formulaes are OK unless you need high precision.

> Now a question specific to my experiment:

> I have two identically blotted Southern blots, and I have two 30mer 
> DNA probes both labelled with 32P to 2*10^8 dpm/ug.  Both probes also
> have about 70% GC.

A 30mer at 70% G+C is going to be highly prone to raise background
by chance matching.  When you're going after a perfect match I 
recommend making the probes 15-20 n.  When expecting mismatches, I 
recommend 36 n minimum.  
Hybridization with oligos at low concentration tends to be kinetically
limited, so also pay attention to keeping the chemical conc. in the
hyb. sol'n as high as practical.
There can be big differences in labelling efficiency.  Make sure you
actually measure sp. act. and don't just believe it because the 
protocol says so.
> I hybridized each probe on a separate membrane under the same
> conditions and washed under the same conditions.  The first probe
> gave me a good signal after a 4 hour exposure on film, and the
> 2nd probe gave me a very weak signal after 24 hours exposure.

If you're going against a single copy mammalian gene, you are fortunate
to get a weak signal in 24 hrs.   If you're going against ug amounts
of cloned DNA the signal should blow you away and most of what I'm
saying would be irrelevant.
A lot depends on what your target is:
how complex is it and are you going after unique or repetitive DNA?   Also
check each probe for strong hairpins or self complementarity.  The probe
can compete itself out if it pairs to itself.
> Both probes were pre-hybridized and hybridized in 5xSSC / 5xDenhardt / 
> 1%SDS / (100ug/ml) denatured sheared salmon sperm DNA at 54 degrees.  
> They were both washed in .2xSSC / .1%SDS at 54 degrees for 1 hour.

You hybridized about 35 C below Tm.  A lot of people would have even
gone down to room temp.  I wouldn't usually worry about this except
the probes seem awful GC rich for 30mers.   At least make sure
you're above the Tm of any self complementarity in the probe.
You should confine the salmon sperm DNA to the prehyb.  Otherwise
you are at risk that a chance match of your oligo to some salmon
sperm repeat could suck your probe out of the hyb. mix.  Most
people only use the salmon sperm competitor for high MW. probes.
You washed about 15 C below Tm.  This is a reasonable place to
start, but now you need to go up some.  One wash usually isn't
enough.  4-5 washes of short duration are much better than one
long wash.  Each wash only need be long enough that the temp.
inside the wash solution actually reaches the target temp.

Hope this helps,
Steve Hardies, Dept. of Biochemistry, Univ. of Texas HSC at San Antonio
Hardies at

More information about the Methods mailing list