Radioactive probes by PCR

Anita Gould anita at
Sun Dec 25 15:08:51 EST 1994

In article <asheri-251294092545 at> asheri at (Ella) writes:

   Does anyone have a good protocol for labelling PCR fragments with 32-P and
   using them as probes for Southern hybridization?

I happen to have some recent posts on this subject stashed away --
here they are.  Haven't tried it myself yet.  NB: These can also be
found by searching the newsgroup archives -- see the FAQ for details.
Good luck!


From: jspaffor at (J. David Spafford)
Newsgroups: bionet.molbio.methds-reagnts
Subject: Re: PCR probes (32 P) Q&A
Date: Fri, 15 Apr 94 23:39:05 GMT
Organization: University of Alberta
Message-ID: <jspaffor.1116840785D at NEWS.SRV.UALBERTA.CA>
References: <1994Apr13.125331.5635 at>

In Article <1994Apr13.125331.5635 at>, futers at 
>Hi netters,
>Has anyone performed a PCR reaction in the presence of a 32P dNTP and
>then used the product to probe a Southern blot?
>If there is any interest I will post a summary.
>                       Many thanks,
>                         Simon    (futers at

I have used the random priming method from BRL to screen libraries and
Southern Blots, but I prefer hot PCR.

I stopped using random priming because it doesn't work well with the small
sized probe I am using.  50% of random primer labelling products are <= 50%
the size of the original template.  With small pieces (<200 bp), the
template needs to be ligated together to create a larger-sized priming area.

With hot PCR probes and degenerate primers, I have been able to compare
Southern Blots using PCR probes from cloned templates vs PCR probe from
genomic DNA template.

Per blot, I have used the following hot PCR probe labelling protocol:

0.2 ug - 1 ug DNA template
PCR buffer mix
1.5 mM - 3.0 mM MgCl2
Taq polymerase

0.2 ul 100 mM dGTP
0.2 ul 100 mM dATP
0.2 ul 100 mM dTTP
0.2 ul 10 mM dCTP
in 95 ul reaction

5 ul 50 uCi 32P labelled dCTP (370 MBq/ml//111TBq/mM//3000Ci/mM from Dupont NEN)

run 10-15 cycles
separate unincorporated by (for example):  isopropanol precipitation
                                           Pharmacia NICK spin columns

I haven't spent much time optimizing the protocol, but it appears to work.
I would appreciate some suggestions for improving the protocol if someone
has some.

Thanks in advance,
J. David Spafford
Department of Zoology, University of Alberta
jspaffor at

From: drmax at (Marianna Max)
Newsgroups: bionet.molbio.methds-reagnts
Subject: Re: PCR probes (32 P) Q&A
Date: 18 Apr 1994 00:16:31 GMT
Organization: Northwestern University, Evanston IL USA
Message-ID: <2osjgv$24f at>
References: <1994Apr13.125331.5635 at> <jspaffor.1116840785D at NEWS.SRV.UALBERTA.CA>

Someone mentioned that one advantage to PCR probes was that one didn't have
to purify the insert away from vector. For most uses this works but if you
are screening a library or a southern that contains any of your vector
sequence be aware that your entire plasmid will be linearly amplified to
some extent and in some cases enough to cause background problems.



From: drm21 at (David Micklem (WCI))
Newsgroups: bionet.molbio.methds-reagnts
Subject: Re: PCR products in Southerns
Date: 27 Jul 1994 22:13:07 GMT
Organization: U. of Cs
Subject: Re: PCR products in Southerns
Date: 27 Jul 1994 22:13:07 GMT
Organization: U. of Cambridge, England
Message-ID: <316m5j$1r4 at>
References: <CtKu7p.49s at> <316ch7$gm9 at>

In article <316ch7$gm9 at> sywang at (Shiao Y. Wang) writes:
>Judy Broom (judy at wrote:
>: I would like to use PCR products as probes in Southerns (using 
>: the Boehringer Random Primed DNA Labeling Kit).  What purification 
>: of the PCR product is necessary?  Someone recommended the Bio Rad Pr
>: ep-A-Gel system to me - would a crude ethanol precipitation do in
>: stead?  (I tried probing with NO purification, and it sort of worked... )
>Why don't you just make your probe by PCR directly? Gibco/BRL sells a kit
>for doing so but the components are those used for normal PCR. That way you
>can skip the purification step altogether. I'm trying this but am having
>difficulty getting good incorporation. Will post a success story soon I hope.

Yes! I do this all the time now, and find it works better than random priming.
The protocol I have (which I got from Michael Akam's lab) is very simple,
 and certainly doesn't require any form of kit!:
1) Do a standard PCR of the bit you want to label. I usually only do 20 cycles.
2) If the band you get is reasonably clean, use 0.5ul in a second PCR reaction
  WITHOUT gel purifying first.  Cold nucleotides remaining from the initial PCR
  dilute the hot label, giving a more stable probe. (See 3) if you want a
screaming hot probe which blows itself to bits v quickly).
   THis labelling PCR is done in 40ul with 4ul(or whatever) of a stock of
   2mM each dA,dG,dTTP replacing your normal 2mM dNTP mix, and 4ul of 32P dCTP
(2-3000Ci/mMole, about 40uCi).  After 10-15 cycles, incorporation should be
   about 50-60%.

3) If your band isn't clean enough, or you need an incandescent probe, then
  gel purify by your favorite method first.  THe probe wont all be full length
  but will be 5-10* hotter than by 2).  
  Use about 1% of the DNA from the first reaction in a 20ul PCR mix.  Use 1ul
  of the dAdGdTTP mix (ie 0.5*normal conc) instead of dNTPs, and 4ul 32PdCTP
  as above.  Five cycles should be sufficient, but I don't really ever use this
 method as I find that 2) above works great for everything I've tried. (eg.(
 screening genomic libraries for homologous sequences).

That my 50pence worth - hope it works for you....


Wellcome/CRC Institute,       Time flies like an arrow...
Tennis Court Road,              
Cambridge CB2 1QR             Fruit flies like a banana.
Tel: [+44] (0)223 334129      Email:drm21 at
Fax: [+44] (0)223 334089

Newsgroups: bionet.molbio.methds-reagnts
From: kvernick at
Subject: re: PCR for radiolabelling probes 
Message-ID: <1994Sep30.172413.24494 at>
Organization: National Institutes of Health
References: <940929140630.20205ab4 at> 
Date: Fri, 30 Sep 1994 20:19:49 GMT

Richard P. Phipps wrote:

> I once tried labeling a PCR product (350 bp) by reamplifying (PCR) the
> fragment in the presence of CTP, GTP, TTP and 32P-ATP.  However, this 
> procedure did not appear to yield a probe which was hotter than if this 
> fragment was randomly-labeled following a standard protocol.
> Could anyone enlighten me on the advantages
> /disadvantages of these two procedures?

PCR labelling works fine, you can achieve very high specific activity,
but there are are some important considerations which affect
specificity and spp activity. A protocol was reported (Schowal &
Sommer (1989) Anal Biochem 177,90-94) which works well but if you are
using a PCR product as a template for labelling, you must gel-purify
it first. The article describes why. A recent optimization of PCR
labelling was reported (BRL Focus rst. The article describes why. A recent optimization of PCR
labelling was reported (BRL Focus 16, 45-48) which I haven't tried
yet, but it seems carefully done. They will also sell you their
optimized system (ahem...) for a slight fee...

Ken Vernick
kvernick at

                                   -Anita Gould
                                    anita at

More information about the Methods mailing list