Quantitative PCR

Anton Scott Goustin asg at cmb.biosci.wayne.edu
Wed Nov 15 18:25:54 EST 1995

aversa at bms.com (Cathy Aversa) wrote:
>I am looking for references/information/opinions etc. on various methods for Q-PCR.  Northerns have been useful, but I need to increase 
sensitivity when probing for certain genes and I'd also like to cut down
and/or stop using radioactivity in the process.  Any comments or 
>Cathy Aversa 

There are generally two approaches for Q-PCR.
The first approach follows the ideas first developed by Salvatore Luria
and Max Delbrück in 1937.  Here's what they developed.  If you dilute
down bacteriophage so far that you have, on the average, one healthy
virus particle per mL, and you plate 50 aliquots of this dilution on
bacteria, you will get about 15 positive cultures.  The fraction follows
the Poisson distribution.  If the fraction is high, you have 1 or more
virus particles per culture if it is low, you have less than 1 virus
particle per culture.  In the last few years, people have applied this
logic to PCR.  You can dilute down a source DNA to the point where you
have on the average ~1 amplicon (i.e., 1 copy of target DNA) per reaxn
tube; if you then do PCR (usually nested, two rounds of ~30 cycles each), then the fraction of tubes which contain a PCR product can be
converted (using the Poisson tables) into a number which tells you, for
example, that you had 1.5 amplicons per tube at the start of PCR.  In 
this way, you can quantitively estimate the concentration of amplicons
(targets) in a sample.  THE PROBLEM:  To get an accurate quantitation,
you need to do >10 tubes per sample.  This method involves no isotope.
The math is the logic of limiting dilution or endpoint analysis in
microbiology and immunology.

The second basic approach involves the use of an internal standard in
the PCR tube, which competes (for primers) with the target of unknown
concentration.  This basic approach relies on the mathematics first
used by Rosalyn Yalow in the 1950s in the radioimmunoassay; in this
assay, unlabeled (cold) proteins of known amount are added to a series
of tubes containing unknown antigen, excess antibody, and a constant
tracer amount of 125I-labeled (hot) antigen.  In the Q-PCR approach,
an internal standard is provided by adding a target amplicon in known
quantity to each PCR tube (for example, 1000 copies per tube) and doing
PCR in the presence of an unknown amount of sample amplicon.  If the
sample amplicon is a 0.5 kb of the human actin cDNA, your reference
amplicon might be a 0.7 kb of DNA, perhaps artificial, which resembles
the 0.5 kb human actin cDNA only at the ends where primers bind (the
stuffer inside presumably need only have the sample nucleotide compo-
sition as the authentic 0.5 kb amplicon.  This approach has been widely-
championed by Paul Siebert and James Larrick, in the little booklet they
wrote on Q-PCR, distributed by Clontech:
We have developed a variation on this theme, which basically involves
constructing a standard curve and using one (or duplicate or triplicate,
depending on the accuracy needed) sample dilution per sample.  It is
published in NAR:

Zachar V, Thomas RA, Goustin AS.  (1993).  Absolute quantification of 
     target DNA: a simple competitive PCR for efficient analysis of 
     multiple samples. Nucleic Acids Reseach  21(8), 2017-2018.

This article is referenced in the standard brochures on PCR distributed
by Perkin-Elmer.


More information about the Methods mailing list