Detection limits of 32P ?

afc at afc at
Fri Apr 16 10:09:55 EST 1993

In article <bheymann-150493171011 at>, bheymann at (Bernard Heymann) writes:
> In article <11857.jdboer at>, jdboer at SOL.UVIC.CA ("Johan de Boer")
> wrote:
>> We have some dispute about how much 32P end-labelled DNA fragment can be
>> detected by autoradiography. I am sure there is someone who has a pretty
>> good idea. Any takers?
>> Johan de Boer
> Johan
> The detection limits of autoradiography depends on how long you expose
> the film to the gel. I know of instances where people have developed their
> autoradiograms for months, although that is hardly practical. The second
> variable is of course the specific activity of the 32P used.

The half life of 32P is two weeks.  That means that after
two weeks the exposure is half of what it would be if you left it for
eternity.  Leaving a 32P autorad more than a couple of weeks is not
productive, since the signal won't increase significantly.

The area that the signal occurs in is important. 1 CPM in a 1 micron spot
would be easy to detect in an in situ.  The same number of counts
on a 1 cm bacterial colony would probably be undetectable.

Another point is that you can't detect anything less than background.  So
if your filter is not washed well, the background might swamp out the

Finally, if sensitivity is critical, spend ~$100k and get an imaging
system that is several times more sensitive than autoradiography :^).

Andrew Cockburn

> The easiest way to find out what the limit is for your protocol with
> 32P of a specific activity is to compare your autoradiogram with
> scintillation counting of the same sample. You can then calculate how
> much labeled DNA would give you a detectable band within a reasonable time.
> I hope this info is of use :)
> Bernard Heymann
> bheymann at

More information about the Methods mailing list