© 2000 Oxford University Press Nucleic Acids Research, 2000, Vol. 28, No.

7 e18

A method for quantification of absolute amounts of

nucleic acids by (RT)–PCR and a new mathematical
model for data analysis
Huong L. Vu*, Serge Troubetzkoy1, Huan H. Nguyen, Michael W. Russell and Jiri Mestecky
Department of Microbiology and 1Department of Mathematics, University of Alabama at Birmingham, Birmingham,
AL 35294-2170, USA

Received January 5, 1999; Revised November 8, 1999; Accepted December 16, 1999

ABSTRACT An internal standard is a nucleic acid fragment of closely

similar sequence to the amplified fragment of a natural target
Accurate quantification of nucleic acids by competitive
of interest, but differs from it in size. Internal standards are
(RT)–PCR requires a valid internal standard, a reference
commonly described by different terms such as ‘competitor’ or
for data normalization and an adequate mathematical ‘mimic’; the term ‘mimic’ will be used in this report. Recently,
model for data analysis. We report here an effective RNA mimics have been generated by PCR-assisted mutagenesis,
procedure for the generation of homologous RNA which may involve searching for and testing a ‘spacer gene’ or
internal standards and a strategy for synthesizing the direct synthesis of the whole single-stranded (ss)DNA
and using a reference target RNA in quantification of template, chemical synthesis of long overlapping primers,
absolute amounts of nucleic acids. Further, a new digestion with restriction enzymes and cloning the mutant
mathematical model describing the general kinetic DNA into a transfer vector (4–7). We propose a simpler and
features of competitive PCR was developed. The more effective strategy for generating homologous RNA
model extends the validity of quantitative competitive mimics with an internal deletion. Our strategy is based on the
(RT)–PCR beyond the exponential phase. The new combination of PCR–ligation–PCR (PLP) (8) to generate
mutant DNA, with a PCR-assisted approach for adding a phage
method eliminates the errors arising from different
RNA polymerase promoter (9).
amplification efficiencies of the co-amplified
Until recently, quantification of target nucleic acids by QC
sequences and from heteroduplex formation in the (RT)–PCR was calibrated by reference to the amount of a
system. The high accuracy (relative error <2%) is defined mimic. Since PCR is highly sensitive, a small difference
comparable to the recently developed real time in physical properties of the target and its mimic results in a
detection 5′-nuclease PCR. Also, corresponding significant difference in their total amplification efficiency.
computer software has been devised for practical Unless this difference is determined, QC (RT)–PCR is valid
data analysis. only for relative quantification (comparison of the nucleic acid
levels between samples) but not for absolute quantification
(determination of absolute amounts of nucleic acids) (10,11).
INTRODUCTION Because RNA requires special handling and storage and is
The intensive development of technologies for quantification usually present in minute quantities, relative quantification is
of nucleic acids over the last decade reflects its importance in not the best choice for samples obtained from different sources
diagnostics and in biomedical research. Despite the fact that a or at different times. We have synthesized a target RNA identical
number of methods are available for this purpose, such as to the amplified fragment of a natural target of interest. Such an
northern blotting, the RNase protection assay (RPA), quantitative identical target RNA can exactly simulate the natural target
competitive reverse transcription–PCR [QC (RT)–PCR] (1) during (RT)–PCR and, hence, can serve as a reference for
and real time detection 5′-nuclease (RTDN) PCR (2,3), the absolute quantification; here it will be referred to as a ‘reference
procedures involved are rather cumbersome. Each of these target’.
methods has its advantages as well as disadvantages. In The accuracy and reproducibility of QC (RT)–PCR have
hybridization methods like northern blotting and RPA, the been the topic of long-lasting discussions (10–14). We have
reaction kinetics are easier to determine, but the sensitivity is developed a mathematical model for data analysis which
not sufficient for most practical applications. In contrast, due to resolves this problem. Until recently, the generally accepted
the amplifying effect, PCR-based methods are more sensitive. mathematical model in traditional QC (RT)–PCR and in
However, the kinetics of PCR are more complex and hence the RTDN PCR has been the linear model (linear regression) (15)
use of PCR as a quantitative method is not straightforward. on log–log or semi-log scales. The main difference between
Generally, a quantitative (RT)–PCR method consists of four these two PCR systems is the time point at which the data are
steps: generation of internal standards, (RT)–PCR, detection of measured. In RTDN PCR each data point is collected as soon
products and data analysis. as it is detectable, i.e. within the exponential phase. In post-PCR

*To whom correspondence should be addressed. Tel: +1 205 934 2233; Fax: +1 205 934 3894; Email: [email protected]
e18 Nucleic Acids Research, 2000, Vol. 28, No. 7

Validation of RNA mimic and reference target RNA by

competitive RT–PCR
Total RNA was isolated from the human promonocytic cell
line U937 stimulated with 2 µg/ml phorbol myristate acetate
(PMA) for 4 h by phenol chloroform extraction with RNA
STAT-60 reagent (Tel-Test Inc., Friendswood, TX). The RNA
mimic was diluted with yeast RNA (Ambion) in DEPC-treated
H2O. Experiments were carried out with 50 and 100 ng of total
Figure 1. DNA template design. Fragments UMD and UD correspond to the RNA from U937 cells (or 0.1 × 107 and 0.2 × 107 copies of the
templates for the reference target and the mimic, respectively.
reference target RNA) and the RNA mimic in a 2- or 1.5-fold
dilution series. All RT and PCR were performed with reagents
from the GeneAmp RNA PCR Kit (Perkin-Elmer, Foster City, CA)
according to the producer’s protocols, with minor modifications.
analysis, all PCR products in a dilution series are analyzed at PCR products were detected by 2% agarose gel electrophoresis
the same time point, when the reactions have reached a plateau and ethidium bromide staining. The gel was scanned on a
(the saturated phase). Since the linear model assumes the data digital imaging system (AlphaImager 2000; Alpha Innotech)
to be within the exponential phase (16), only RTDN PCR can by the 1D-Multi (Line Densitometry) method. While making
yield accurate and reproducible results. However, RTDN PCR exposure adjustments, the color saturation was checked to
presents a new, more complex reaction system and requires new ensure that the image was not over-saturated or under-exposed.
costly equipment. Our strategy is based on a new mathematical
model describing the general kinetic features of competitive Mathematical modeling
PCR which extends the validity of QC (RT)–PCR beyond the The aim of a mathematical model for PCR is to describe the
exponential phase. Also, corresponding computer software has relationship between the input and output amounts of the
been devised for practical data analysis. amplified sequence in a PCR tube. The amplification of a
single target can be described by the following equation:
MATERIALS AND METHODS T1 = T0·(1 + E1) 1
where T0 is the initial number of target molecules, T1 is the
Generation of RNA mimics number after the first amplification cycle and E is the
Design and synthesis of DNA templates. Two primer pairs amplification efficiency of the target. The index refers to the
corresponding to different exons, P1 and P2 and P3 and P4, cycle number. A similar amplification equation can be written
were chosen from the sequence for human interleukin 1β for the corresponding mimic:
mRNA (GenBank accession no. M15330) by means of the M1 = M0·(1 + F1) 2
computer program Oligo 4.0 (National Biosciences, Plymouth,
MN) to amplify two PCR fragments, U (upstream) and D The amplification efficiency is defined as the fraction of the
sequence amplified per total input amount and represents the
(downstream), separated by the intervening sequence M (Fig. 1).
yield of the chemical reaction, which depends on the reaction
The upper primer P1 of fragment U was appended to an RNA
rate. A reaction rate of any order can be described as a power
polymerase promoter (SP6 or T7). Fragments U and D were
function of the concentrations of reactants. Under constant
phosphorylated by forward reaction with T4 polynucleotide
reaction conditions E or F can be considered as a power
kinase and then ligated together by T4 DNA ligase (Ambion, function of only T0, or only M0, respectively. Thus, T1 and M1
Austin, TX) to form a new fragment UD. Extended PCR was are polynomial functions of T0 and M0, respectively. After n
performed on the ligated UD fragment with upper primer P1 cycles, equations 1 and 2, subjected to n iterations, can be
and lower primer P4. This PCR product UD, therefore, was written as follows:
deleted with respect to sequence M in comparison with the
natural target. Since it had an RNA polymerase promoter Tn = P(T0) 3
attached in the correct sense, it would serve as a DNA template Mn = Q(M0) 4
for transcription or could be used directly as a DNA mimic.
where P and Q are polynomial functions of T0 and M0,
respectively.
In vitro transcription. Transcription of the PCR product (UD) Since in this case the reaction conditions are functions of
was carried out with the BIOTINscript kit (Ambion) followed time, it would be inaccurate to quantify an initial amount of
by digestion with DNase I. The transcripts were subsequently nucleic acid from the PCR product using equations 3 and 4.
purified on RNeasy columns (Qiagen, Chatsworth, CA), quan- However, the requirement of constant reaction conditions can
tified by measuring OD260 in a GeneQuant spectrophotometer be fulfilled by means of competitive PCR, as shown in the next
(Pharmacia, Cambridge, UK). paragraph.
Generation of reference target RNA
The rational model as a general mathematical model for
The DNA template for this transcript is the fragment UMD competitive PCR. In a competitive PCR tube, the target and its
(Fig. 1) which was generated from wild-type RNA by RT–PCR mimic react at the same time and compete for the same
with upper primer P1 and lower primer P4. In vitro transcription reagents. The reaction rate depends on the collision frequency
was performed as described above. of the reacting molecules. Therefore, the amplification
ii
 Nucleic Acids Research, 2000, Vol. 28, No. 7 e18

efficiency for a sequence depends on its relative abundance

(referred to the other sequence) rather than on its absolute
amount in the reaction system.
Let us define:
the input relative molar fraction of the target
1/X = T0/M0 = Y0 definition 1a
and of the mimic
X = M0/T0 = 1/Y0 definition 1b
the output relative molar fraction of the target after n cycles
Y = Tn/Mn = Yn definition 2a
and of the mimic
1/Y = Mn/Tn = 1/Yn definition 2b
In each QC (RT)–PCR experiment two dilution series are set
up, where the time parameter (the number of cycles) is fixed
and the master mixture used in all the reaction tubes came from
the same source. In each dilution series the initial target
concentration is constant while the initial concentration of mimic
is the only independent variable. Therefore, the co-amplification
of a target and its mimic can be described by equations similar
to equations 3 and 4 as follows:
Y = Yn = P(Y0) = P(1/X) = G(X) 5
1/Y = 1/Yn = Q(1/Y0) = Q(X) 6
where P and Q are polynomials of the same degree and G is a
rational function.
Both equations 5 and 6 describe the relationship between the
input and output amounts of the co-amplified sequences by a
function of only one unknown. However, since rational
functions are in this case more suitable for data analysis, both
graphically and numerically, than polynomial functions,
equation 5 was chosen as a general mathematical model for QC
(RT)–PCR. Note that the model assumes identical reaction
conditions for the dilution series (i.e. the same RT–PCR master
mixture), but does not refer to any specific reaction kinetics.
Therefore, it is general and can be used beyond the linear range
of PCR.
Figure 2a and b shows real examples, where PCR data were
fitted to this model. Since T0 is a constant, the graph Y versus
M0 is equivalent to equation 5.
Figure 2. Fitting QC (RT)–PCR data for (a) the natural and (b) the reference
As simple as it may seem, the model captures the essential targets by the program PCRFIT according to the rational model. M0 is the
nature of competitive PCR. The following illustrate how the mimic input amount. Y is the ratio between the scanned densities of target and
mathematical characteristics of the model match the physical mimic bands. A black circle is a data point, a blue circle denotes the equimolar
properties of the process. point EP, a red square is an outlier eliminated from input data. LT and LM are
the lengths of the target and mimic amplified fragments, respectively.
(i) Monotonically decreasing function. In the dilution series,
as the mimic concentration X increases, Y decreases.
(ii) Asymptotic convergence. This function has the x- and y-axes
as asymptotes: as X approaches 0, Y approaches ∞, which mimic have the same amplification efficiency, then equality of
means there is target without mimic; as X approaches ∞, the input amounts (M0 = T0) implies equality of the output
Y approaches 0, there is mimic without target in the reaction. amounts (Mn = Tn).
(iii)The fixed point and the apparent equimolar point. If P = Q If P ≠ Q, then the input equimolar point X = 1 is not a fixed
then equations 5 and 6 become: point. The target and its mimic are co-amplified by reactions of
Y = P(1/X) 7 the same order but their physical properties are different.
Therefore, P and Q are polynomials of the same degree but
1/Y = P(X) 8 with different coefficients and, hence, the target and mimic
For X = 1, equations 7 and 8 imply Y = 1/Y, i.e. Y = 1. Thus, the have different amplification efficiencies. Equality of the input
input equimolar point X = 1 is also the fixed point Y(1) = 1, amounts does not imply equality of the output amounts.
whose physical meaning is as follows: if the target and the However, since the function Y(X) decreases monotonically,
iii
e18 Nucleic Acids Research, 2000, Vol. 28, No. 7

equation 5 can be solved uniquely for X at the output equimolar imply the relationship between the outputs at skew-symmetrical
point Y = 1. At this apparent equimolar point, where X = XEP points:
and Y = 1, the input amounts of the target and the mimic are in Yi = P(1/Xi) = P(Xj) = 1/Yj 12
a ratio that just compensates for the difference in their
amplification efficiencies, yielding equal output amounts. In or on a log–log scale
practice, since the detected output signal is proportional to the logXi = –logXj 13a
length of the molecules, a correction for the difference in and
length between target and mimic is taken into account. Let C logYi = –logYj 13b
be the ratio between the lengths of the target and the mimic
amplified fragments (in our case C = 1.2). Then, graphically, Equations 13a and 13b mean that logY is a skew-symmetrical
the apparent equimolar point corresponds to the point of inter- function of logX. In the exponential phase, the graph logY
section of the line Y = Tn/Mn = C and the curve of equation 5, versus logX is a straight line with slope –1 and passes through
where M0 = MEP, XEP = MEP/T0. the origin. In the saturated phase, the graph is no longer a line.
However, the deviations of the function logY(X) from the ideal
The linear model and special cases. Although the linear model line caused by the saturation effect are skew-symmetrical with
has been the only mathematical model used for data evaluation respect to the equimolar point. As a consequence of this, the
in competitive PCR so far, its nature has not been studied regression line rotates around the equimolar point. One can
thoroughly. For the saturated phase of PCR, the model shows view the rotation as being caused by two ‘forces’ acting in
special behavior: its error is random (low reproducibility) and opposite directions on the ends of the line segment. If these
seems not to exceed a certain limit (~25%). The reason for this ‘forces’ are equal, then the line would rotate around the
may be explored by the following analysis. equimolar point. The rotation does not displace the equimolar
According to the linear model, the graph of equation 5 on a point and, hence, does not affect the result. The same result can
log–log scale be obtained even from an asymmetrical dilution series whose
total deviations on both sides of the equimolar point are
logY(X) = logG(X) 9 balanced (the balanced model). The skew-symmetrical model
should be a straight line. Let us analyze two typical but special is actually a special case of the balanced model. Therefore, if
cases when the model can give correct results. the dilution series were fortuitously balanced around the
equimolar point, the linear model would give a correct answer.
The exponential phase. During this phase of the reaction, the If P ≠ Q, then the regression line is also shifted as mentioned
amplification efficiency is supposed to be constant over the above. The center of rotation then would be the apparent
whole range of X. Then, equation 5 has the simplified form: equimolar point. In practice, for unbalanced dilution series, the
center of rotation shifts away from the apparent equimolar
Y = Yn = k·Y0 = k/X 10 point, causing an error of unpredictable nature. The fact that
where k is a constant. On a log–log scale equation 10 becomes: the non-linearity of the function logY(X) results partly in
rotation instead of mere translation of the regression line
logY(X) = logk – logX 11a explains why the error of the model is partly compromised.
or Figure 3 shows the translation and the rotation of the
logY(M0) = logk + logT0 – logM0 11b regression line in a real example.
The corresponding graph, logY versus logM0, is a line with Applications of the mathematical model for QC (RT)–PCR
slope –1. The constant logk shifts the graph along the vertical
axis. For relative quantification, this translation does not Relative quantification of nucleic acids. As implied from the
matter. The essential simplification here is the assumption of rational model, under identical reaction conditions the
independence of the amplification efficiency from the relative apparent equimolar point XEP is a constant, typically not 1. For
abundance of the co-amplified sequences. In the saturated two samples with input target amounts (T0)1 and (T0)2 it holds
phase, when some of the reagents become limiting factors, that:
co-amplification of the target and the mimic is of a competitive XEP = (MEP)1/(T0)1 = (MEP)2/(T0)2 14
nature, this assumption is no longer valid. or
R = (T0)1/(T0)2 = (MEP)1/(MEP)2 15
The balanced model. Suppose that the target and the mimic
have identical amplification efficiency, i.e. P = Q, then their Equation 15 tells us that the ratio R between the input target
co-amplification is described by equations 7 and 8. Let us also amounts in two samples can be determined from the mimic
consider a special model: a dilution series ideally ‘skew- amounts at the apparent equimolar points. This also explains
symmetrical’ around the equimolar point X = 1. This means why relative QC (RT)–PCR is theoretically always correct,
that for each PCR tube with any input relative molar fraction X regardless of the difference between the amplification efficien-
there is a corresponding PCR tube with input relative molar cies of the target and mimic. If the theoretical ratio Rth is
fraction 1/X. For example, tubes i and j, with Xi = 2 and Xj = 0.5, known in advance then the relative error of the assay can be
are skew-symmetrical points in the series. Since the relative estimated:
abundance of the target and the mimic in one tube are reciprocal, ERR% = [(R – Rth)/Rth]·100 definition 3
the target has the same amplification efficiency as that of the
mimic in the corresponding skew-symmetrical tube, rather Absolute quantification of nucleic acids. The strategy for
than in the same tube, as previously thought. Equations 7 and 8 absolute quantification arises directly from the relative
iv
 Nucleic Acids Research, 2000, Vol. 28, No. 7 e18

Figure 3. Fitting QC (RT)–PCR data (the same data as in Fig. 2b) by linear
regression according to the linear model. Line 1, log10Yideal = –log10M0 – 0.355,
r = 1.000, the ideal (hypothetical) regression line with a slope of –1, valid only
for the exponential phase; line 2, log10Yreal = –1.325 log10M0 – 0.414, r = 0.993,
the real regression line, translated and rotated due to the saturation effect; line 3,
log10Y = logC = log101.2 = 0.079. The apparent equimolar points, EP1 and EP2,
are the points of intersection of line 3 and the corresponding line (1 and 2,
respectively). The estimation error is given by the distance between EP1 and EP2.

quantification between a natural target and its identical

reference target. The principle is as follows: two dilution series
are set up in a QC (RT)–PCR experiment, one with a constant
natural target amount (T0)1 and the other with a constant
reference target amount (T0)2. Since (T0)2 is known in advance
and the ratio R between (T0)1 and (T0)2 is determined as
described above, (T0)1 can be calculated from equation 15.
Computer software for QC (RT)–PCR and practical data
analysis
To implement our mathematical model we have modified a
computer program for general least squares fit (maximum like-
lihood) (17). We show here how the computer program Figure 4. Graphical detection of outliers by the linear model on a natural
(PCRFIT 1.0, written in Microsoft Visual C++ 5.0 under logarithm scale (a), and by the rational model (b). Point B, a bad outlier, is
clearly distinguishable from point C, a slight outlier, in (b) but not in (a).
Windows 95) can help to perform the following tasks:
• recognize and eliminate biased data points (outliers);
• evaluate the consistency of the data;
• estimate the amounts of nucleic acids by determining the
apparent equimolar point. target and mimic bands), the lengths of the target and the
mimic amplified fragments.
Competitive PCR may produce outliers that cannot be recog-
(ii) The program returns the graph, the apparent equimolar
nized either on the gel or by the graph of the linear model.
point EP (M0 = MEP and Y = Tn/Mn = C), the χ2 value and
Since the regression line is plotted between the data points, it
the list of the calculated values of Y dependent on M0 for as
can be hard to distinguish ‘good’ points from ‘bad’ (Fig. 4a). In
many points as the user wishes.
contrast, when the data are plotted as Y versus M0, a hyperbola-
(iii) For typically accurate data, the graph is a smooth hyperbola-
like curve results. In most cases, one can readily discern the like curve (Fig. 2a and b) and the χ2 value is of the order of
biased data points lying off the curve (Fig. 4b). PCRFIT can 10–3 or less (see Results and Discussion below). If the χ2
help to confirm this by returning the χ2 value. After eliminating value is higher than the critical value (which is experimentally
the suspected outlier, the χ2 value should drop considerably. determined) and the graph appears atypical, then either
The following is a brief summary of the interaction between some outliers are present or the entire data set is not
the user and PCRFIT. sufficiently consistent.
(i) The user provides input data which contains the values of If some outliers occur, they can be easily located by
the mimic dilution series M0, the corresponding measured inspecting the graph and eliminated by double-clicking on
values of Y (the ratio between the scanned densities of them (they can always be recovered by being double-clicked
v
e18 Nucleic Acids Research, 2000, Vol. 28, No. 7

Table 1. Estimation of the 2-fold difference in the levels of target RNA (human interleukin 1β mRNA in U937
total RNA and the reference target) by the two mathematical models

*Results obtained after eliminating the outliers.

again). If the χ2 value does not drop under the critical value reactions in a dilution series was six. Since accurate estimation
and the graph still looks unsatisfactory after elimination of by the rational model requires at least five good data points, a
the suspected outliers, then the experimentally biased data dilution ratio of 1.5 was chosen for experiments E#4–E#7 and
points are dominant. In this case, since the entire data set is nine PCR reactions in the dilution series were performed
inconsistent, checking for experimental errors is highly (allowing for PCR failure or outliers).
recommended.
(iv) The numerical result is the value M0 = MEP corresponding Mathematical models for data analysis of QC (RT)–PCR
to the output equimolar point. Two alternative mathematical models, the linear model (linear
regression) and the rational model (by means of the computer
RESULTS AND DISCUSSION program PCRFIT), were used for analysis of the obtained data.
QC (RT)–PCR assays were performed for the natural and the Data fitting was performed both with and without outliers
reference targets as shown in Figures 5 and 6, respectively. A (Table 1). Data consistency was evaluated by the χ2 criterion.
2-fold difference in target RNA levels was estimated and the The experimental ratio, R, was determined from equation 15,
results are summarized in Table 1. In experiments E#1 and E#2, where the mimic amounts at the apparent equimolar points,
the dilution ratio for the mimic was chosen to be 2, as usually MEP, were estimated by the corresponding mathematical
recommended, and the corresponding number of PCR model. The relative error, ERR%, shown as an absolute value,
vi
 Nucleic Acids Research, 2000, Vol. 28, No. 7 e18

no outlier observed in E#7, in which the template concentration

was lowest (Table 1).
In experiments E#4A and B, the inconsistency of the data,
due to pipetting error, is revealed graphically as atypical curves
and numerically as high χ2 values, even after eliminating the
obvious outlier.
Unlike E#4, where the error arose randomly during the
Figure 5. Detection of relative levels of human interleukin 1β mRNA by competitive measurement of input parameters, in E#6 the error was related
RT–PCR. Lanes 1 and 2, mimic alone, 0.1 × 107 and 6.4 × 107 molecules,
respectively; lanes 3 and 10, total U937 RNA alone, 50 and 100 ng, respectively;
systematically to the product detection step. Two data points
lanes 4–9, the first series containing a constant 50 ng of total U937 RNA with were missing in this dataset because the two strongest target
the mimic in 2-fold dilutions from 0.1 × 107 to 3.2 × 107 molecules; lanes 11–16, bands completely overshadowed the weaker mimic bands,
the second series containing a constant 100 ng of total U937 RNA with the which, though clearly visible, could not be scanned by the
mimic in 2-fold dilutions from 0.2 × 107 to 6.4 × 107 molecules; lane 17, DNA imaging system. The divergence of strong fluorescence signals
marker.
of the target bands from the linear range caused a slight under-
estimation of the data (point C in Fig. 4b) and, hence, of the
result (R = 1.92). This type of error is of a systematic character.
It is important to emphasize that although χ2 is a reliable
criterion for data consistency, in most cases it fails to indicate
systematic errors.
Strategy for absolute quantification of nucleic acids
The strategy for absolute quantification of nucleic acids is
based on relative quantification between the natural and the
reference targets. Therefore, the validity of the method relies
Figure 6. Simulation of the natural target’s behavior in competitive RT–PCR on the identical amplification characteristics of the two targets
with the reference target RNA. Lane 1, DNA marker; lanes 2–10, the first and the accuracy of the relative quantification, as discussed
series containing a constant 2 × 106 molecules of the reference target with the previously.
mimic in 1.5-fold dilutions from 4 × 107 to 1.56 × 106 molecules; lanes 11–19, The amplification characteristics of the two targets can be
the second series containing a constant 1 × 106 molecules of the reference target
with the mimic in 1.5-fold dilutions from 2.67 × 107 to 1.04 × 106 molecules;
considered as identical for the following reasons: (i) the two
lane 20, water instead of RNA (negative control). amplified fragments are identical sequences; (ii) the RT–PCR
products of both targets appeared identical (Figs 5 and 6),
suggesting that there is no inhibiting factor which could affect
the two amplifications differently; (iii) the two experiments E#5
was then evaluated according to definition 3, where the theoretical and E#7 for the two targets show comparable data evaluated by
ratio Rth = 2. the rational model (Table 1 and Fig. 2).
The data in Table 1 show that the relative error of estimation
by the linear model ranged randomly from 4 (E#4) to 24% Standard curve. QC (RT)–PCR is based on determination of
(E#7). These results are in accord with other reports (14,18). the apparent equimolar point (EP), where Y = C and M0 = MEP.
In contrast, the rational model clearly distinguishes four If it was based on other points (Y ≠ C and M0 ≠ MEP), then the
categories of experimental datasets: (i) absence of outliers error is generally greater. Figure 7a and b shows this
(E#7); (ii) exceptional occurrence of outliers (E#5); (iii) incon- relationship for real data. The graphs corresponding to
sistent dataset (E#4); (iv) consistent dataset with a small different data sets seem to have similar patterns. Although the
systematic error (E#6). error was found to also be minimal at another point, EO, it is
The accuracy of the rational model was excellent (relative only at the apparent equimolar point EP that the accuracy is
error ~1% with low χ2 values) for good experimental data excellent and reproducible. The reason is that the apparent
equimolar point is the only fixed point of the function. At the
without outliers or after eliminating the exceptional outliers
fixed point, the value of the function does not depend on its
(E#7 and E#5, respectively). The fact that the two experiments
coefficients. In other words, the apparent equimolar point is the
were performed with different kinds of target RNA (natural
only point in the dilution series where inter-assay comparison of
and synthetic, respectively), using different batches of the the amount of nucleic acid in different tubes is valid. The use
mimic and the other reagents, implies a high reproducibility of of standard curves, which depends on unwarranted extension
the method. The origin of the exceptional outliers is the question of this validity to other points, is therefore not recommended.
that has concerned us from the beginning of this study. We
applied the theory of dynamic systems to study the mathematical PCR product detection. Our results indicate that the combination
model of competitive PCR. However, our study implies that of agarose gel electrophoresis and an imaging system is a
chaotic phenomena are not involved in this dynamic system. reliable detection method. The relative quantification of
Due to the amplifying effect of PCR, a number of factors such nucleic acids using internal standards rules out all variations in
as contamination, incorrect pipetting and insufficient mixing scanning that may occur. Although the absolute values of data
of reagents usually contribute to the outliers. According to our points may vary from one scan to another (19), the final
experience, the probability of an outlier occurring is higher in relative result is not significantly affected by this variability.
experiments starting with high DNA concentrations. There was Other more complex detection methods, such as using different
vii
e18 Nucleic Acids Research, 2000, Vol. 28, No. 7

PCR yielding mimics free of bacterial nuclease contamination;

no multiple rounds of PCR are required to produce DNA
templates; flexibility in designing the size of PCR products
resulting in easier product separation and distinction of
contaminating genomic DNA from RT–PCR products. For the
generation of RNA mimics, the crucial step is transcription.
Experience in the synthesis of more than 30 RNA probes for
non-isotopic RPA, based on the same PCR strategy for adding
a phage promoter, revealed no problems with transcription of
templates up to 600 bp, which is sufficient for the present
application.
In conclusion, our study has clearly shown that the new
mathematical model significantly improves the data analysis,
and therefore the accuracy and reproducibility, of QC (RT)–PCR.
Also, the mathematical model strongly supports our practical
strategies for generation of internal standard RNA and quanti-
fication of absolute amounts of nucleic acids by QC (RT)–PCR
using the conventional PCR system. As no special costly
equipment is required, the method is suitable for all laboratories
having basic facilities for molecular biology procedures.
Furthermore, our computer software is suitable for implementing
quantitative PCR as a new feature for automated PCR systems.

ACKNOWLEDGEMENTS
The authors wish to thank Drs Robert Kralovics and Roman
Tuma for their helpful discussions regarding molecular biology
methods and computer software development, respectively. This
work was supported by NIH Training Grants T32 HL07553 for
H.L.V. and T32 AI07150 for H.H.N. and research grants
AI 28147 and DE 09691.
Figure 7. Dependence of the estimation error, ERR%, on variable Y, the ratio
of target to mimic output signals. The absolute value of ERR% is minimal at NOTICE
the output equimolar point EP (Y = C = 1.2) and at another point EO. Only EP,
not EO, is reproducible [compare (a) and (b)]. A negative value of ERR% The computer software has been developed by personal efforts,
means underestimation and vice versa. and should be referred as PCRFIT 1.0 copyright © 1999 by
Huong Lan Vu and Vu Quoc Nguyen. PCRFIT is available for
users from: Maria Crenshaw, Department of Microbiology,
University of Alabama at Birmingham, BBRB #757, 845 19th
labeled primers, would be necessary to increase the sensitivity, Street South, Birmingham, AL 35294, USA. Tel: +1 205 934
but not the accuracy, of the method. 2225; Fax: +1 205 934 3894; Email: [email protected]. The
cost for PCRFIT is $475.00 plus shipping costs. A demonstration
Generation of RNA mimics version is available upon request, free of charge plus shipping
The fact that the accuracy of the new method is comparable costs.
with that of the RTDN PCR method, where heteroduplex
formation between the target and the mimic sequences (20) REFERENCES
does not interfere with the accuracy, suggests that hetero- 1. Wang,A.M., Doyle,M.V. and Mark,D.F. (1989) Proc. Natl Acad. Sci.
duplex formation does not present a problem in our system. If USA, 86, 9717–9721.
heteroduplexes occurred, they would have a loop of 96 nt (Fig. 1), 2. Heid,C.A., Stevens,J., Livak,K.J. and Williams,P.M. (1996) Genome Res.,
i.e. almost 25% of their total length. Such heteroduplexes would 6, 986–994.
3. Gibson,U.E., Heid,C.A. and Williams,P.M. (1996) Genome Res., 6,
not migrate at the same rate as either of the homoduplexes or 995–1001.
faster than both of them. Since no unexpected PCR product of 4. Vanden,H.J., Tyson,F.L. and Bell,D.A. (1993) Biotechniques, 14,
a larger size than the mimic was detected (Figs 5 and 6), hetero- 395–398.
duplex formation can be considered as negligible. This is an 5. Fille,M., Shanley,J.D. and Aslanzadeh,J. (1997) Biotechniques, 23,
34–36.
advantage of a deletion in the middle of the sequence, in 6. Tarnuzzer,R.W., Macauley,S.P., Farmerie,W.G., Caballero,S.,
comparison with a deletion at one end, where heteroduplexes Ghassemifar,M.R., Anderson,J.T., Robinson,C.P., Grant,M.B.,
may migrate like the longer homoduplexes and complicate the Humphreys,B.M., Franzen,L. et al. (1996) Biotechniques, 20, 670–674.
separation process. 7. Jin,C.F., Mata,M. and Fink,D.J. (1994) PCR Methods Appl., 3, 252–255.
8. Ali,S.A., Sarto,I. and Steinkasserer,A. (1997) Biotechniques, 22,
Other advantages of the new strategy for the generation of 1060–1062.
RNA mimics are: regular cloning is efficiently replaced by 9. Mullis,K.B. and Faloona,F.A. (1987) Methods Enzymol., 155, 335–350.

viii
 Nucleic Acids Research, 2000, Vol. 28, No. 7 e18

10. Raeymaekers,L. (1995) Genome Res., 5, 91–94. 17. Press,W.H., Teukolsky,S.A., Vetterling,W.T. and Flannery,B.P. (1992)
11. Peccoud,J. and Jacob,C. (1996) Biophys. J., 71, 101–108. Numerical Recipes in C. The Art of Scientific Computing.
12. Hengen,P.N. (1995) Trends Biochem. Sci., 20, 476–477. Cambridge University Press, New York, NY.
13. Delassus,S. (1997) Eur. Cytokine Netw., 8, 239–244. 18. Zimmermann,K. and Mannhalter,J.W. (1996) Biotechniques, 21,
14. Lader,E. (1998) Ambion Technotes, 5, 1. 268–272.
15. Wiesner,R.J. (1992) Nucleic Acids Res., 20, 5863–5864. 19. Sundfors,C. and Collan,Y. (1995) Cell. Mol. Biol., 41, 671–681.
16. Wiesner,R.J., Beinbrech,B. and Ruegg,J.C. (1993) Nature, 366, 416. 20. Henley,W.N., Schuebel,K.E. and Nielsen,D.A. (1996) Biochem. Biophys.
Res. Commun., 226, 113–117.

ix