Light Cycler

Dr. Duncan Clark Duncan at
Thu Oct 19 07:43:37 EST 2000

In article <39EEB969.9D0B35A6 at>, the eminent Paul
Cullen at Monash University wrote
>Why does glass in the capillaries stuff up PCR using standard buffers?
>We had a light cycler on loan and I found that I could not even
>reproduce many reactions that I carry out routinely?

Well I don't want to give away all the secrets but...

You must have BSA, (heat treated not acetylated - another story on it's
own) in the reaction mix. Glass non-specifically binds proteins and
given the huge surface area to volume ratio in the capillary plus
roughly only 1.6ng of protein from the Taq, this effectively means that
the glass wall will mop up most of the enzyme. Addition of a huge excess
BSA reduces minimizes the loss of Taq pol.

Glass also has this effect of messing with the Mg. A capillary takes out
roughly 1.5mM of Mg so you must add extra back to something like that.
Optimum will be between 2 and 5 mM, usually 3 and 4. But like any PCR
optimum Mg varies depending upon other components in the PCR reaction.

Now all the above is reasonably logical and makes some sense. 

So why does the addition of a few mM of KCl, AmmSO4 and every salt you
can imagine actually reduce the PCR yield in a glass capillary compared
to no salt? (*) This is opposite to what one finds in normal
polypropylene micro-tube PCR where 50mM KCl is normal. I have no
explanation. It doesn't really matter with standard PCR because  10
extra cycles will only take 4 mins or so. However it does matter when
you get to trying single tube RT-PCR in a capillary when you are using
two enzymes, Taq or a.n.other enzyme plus a separate RT with very
different buffer requirements. You land up with a very poor compromise
whereas with plastic you can get a very good compromise buffer.   

(*) I do now know of the odd probe technologies that love high salt in
capillaries and it is the only way they will work. The problem then is
that some probes work in that buffer and if they don't, then they work
in a totally different low salt buffer. Unfortunately there is no way
one can tell which buffer to use from the initial sequence. 
>Another interesting question is why does it take more cycles?

I guess that the PCR reaction, due to a compromise buffer, is less
efficient per cycle by maybe only 0.5%. It's a bit like oligo synthesis.
After 30 rounds the final product yield is visibly down from a reaction
with even just 0.1% more efficiency per cycle etc. Don't forget that a
capillary PCR has substantially more Taq per ul than a standard plastic
tube PCR conditions.. 

>Just curious! 

I think real-time is a great way of learning/seeing what is actually
happening in PCR itself. You can really see the kinetics and get a feel
for a reaction. All good fun. 


The problem with being on the cutting edge is that you occasionally get 
sliced from time to time....

Duncan Clark
GeneSys Ltd.
Tel: +44(0)1252376288
FAX: +44(0)8701640382

More information about the Methods mailing list