We use most often the second of the following three methods to get rid
of the primers:
A. QIAquick spun columns (QIAGEN), or similar "quick" columns with
silica gel membranes. Fastest, cheapest (because time is saved), but in
our hands often not good enough for our needs (a good amount of the PCR
product is often lost and primers still can be present in the eluate).
Still, you may wish to try this first, before going into more
manipulations (below). You can easily process hundreds of samples and
some people even reuse columns after elution if small-scale
contamination is acceptable.
B. Electrophoresis. The slowest and the best (from the methods suitable
for large-scale runs). We like the most the electrophoretic separation
of the PCR reaction products from the primers in regular agarose gel
(the gel has also other analytical advantages), followed by cutting a
well in front of the PCR product, removing a piece that often contains
the unincorporated primers, filling the well with melted LMPA
(low-melting point agarose), cooling down to let the LMPA solidify, and
running again for 3-10 minutes until the PCR band goes into the LMPA.
After cutting out and melting the piece of LMPA with the PCR band at 65
degrees Celsius for 15 minutes, there are two possibilities: a) you can
cool down to 41 degrees, digest with agarase for an hour and extract
with phenol-chloroform or b) purify the DNA on a "quick" column. The
former recovers most of the PCR product, the latter is lossy but quicke
(in that case the method combines electrophoresis and spun columns).
The agarose gel elctrophoresis is suitable for PCR porducts larger than
100 bp and is a method that removes the primers efficiently (unless you
overload the well).
Using electrophoresis, it is not difficult to process several dozens
of samples in a single experiment. Total time, if you have a stock of
pre-cast agarose gels, could be from two hours to a working day for a
large batch of samples. It is not expensive in terms of reagents,
because the LMPA (the most expensive reagent) is used only for the tiny
wells in front of the bands (this also allows you to run quick agarose
gels at 200V and higher, which would be impossible for a gel made
entirely of LMPA).
B. Chromatography. Regular primers (20-mer) have m.w. of about 5K, PCR
products are much bigger (e.g. 50K-200K and bigger). Size fractionation
on a Sephadex column with correctly chosen resin should therefore be an
easy and fast method to separate PCR products from primers.
Unfortunately, while PCR products are excellently separated from
nucleotides (as we all know from the Sephadex G-50 spun columns
described some 17 years ago in the manual of Maniatis), primers are 20
times or so larger than nucleotides, and still run with the PCR products
at the front on a regular Sephadex G-25 column (although some of them
are retained), however you do get a better enrichment if you use G-50
and especially G-100 (both PCR losses and primer contamination still
occur, especially if you are using the spun columns; G-100 should be
spun at very low speed). Our experience was never very good with size
fractionation on a Sephadex column. But it is faster than agarose gels,
large number of spun columns could be prepared and run for less than an
hour through the end-of-spin step, and the price is lower than the
"quick" columns. However, after the spin you may need to precipitate,
reduce the volume in a SpeedVac, etc., and this could take longer.
Other methods for preparative separation, such as PAGE and HPLC, are
quite cumbersome, time- and equipment-consuming, and I cannot recommend
their use to process bazillion of PCR products.
Regards,
victor, Alpha DNA