PCR Tiempo Real
PCR Tiempo Real
PCR Tiempo Real
org on July 29, 2023 - Published by Cold Spring Harbor Laboratory Press
GENOME METHODS
1BioAnalytical Technology Department, Genentech, Inc., South San Francisco, California 94080;
2Applied BioSystems Division of Perkin Elmer Corp., Foster City, California 94404
We have developed a novel "real time" quantitative PCR method. The method measures PCR product
accumulation through a dual-labeled fluorogenic probe {i.e., TaqMan Probe}. This method provides very
accurate and reproducible quantitation of gene copies. Unlike other quantitative PCR methods, real-time PCR
does not require post-PCR sample handling, preventing potential PCR product carry-over contamination and
resulting in much faster and higher throughput assays. The real-time PCR method has a very large dynamic
range of starting target molecule determination (at least five orders of magnitude}. Real-time quantitative
PCR is extremely accurate and less labor-intensive than current quantitative PCR methods.
Quantitative nucleic acid sequence analysis has that it be used properly for quantitation (Raey-
had an important role in many fields of biologi- maekers 1995). Many early reports of quantita-
cal research. Measurement of gene expression tive PCR and RT-PCR described quantitation of
(RNA) has been used extensively in monitoring the PCR product but did not measure the initial
biological responses to various stimuli (Tan et al. target sequence quantity. It is essential to design
1994; Huang et al. 1995a,b; Prud'homme et al. proper controls for the quantitation of the initial
1995). Quantitative gene analysis (DNA) has target sequences (Ferre 1992; Clementi et al.
been used to determine the genome quantity of a 1993).
particular gene, as in the case of the h u m a n HER2 Researchers have developed several methods
gene, which is amplified in -30% of breast tu- of quantitative PCR and RT-PCR. One approach
mors (Slamon et al. 1987). Gene and genome measures PCR product quantity in the log phase
quantitation (DNA and RNA) also have been used of the reaction before the plateau (Kellogg et al.
for analysis of h u m a n immunodeficiency virus 1990; Pang et al. 1990). This method requires
(HIV) burden demonstrating changes in the lev- that each sample has equal input amounts of
els of virus throughout the different phases of the nucleic acid and that each sample under analysis
disease (Connor et al. 1993; Piatak et al. 1993b; amplifies with identical efficiency up to the point
Furtado et al. 1995). of quantitative analysis. A gene sequence (con-
Many methods have been described for the tained in all samples at relatively constant quan-
quantitative analysis of nucleic acid sequences tities, such as [~-actin) can be used for sample
(both for RNA and DNA; Southern 1975; Sharp et amplification efficiency normalization. Using
al. 1980; T h o m a s 1980). Recently, PCR has conventional methods of PCR detection and
proven to be a powerful tool for quantitative quantitation (gel electrophoresis or plate capture
nucleic acid analysis. PCR and reverse transcrip- hybridization), it is extremely laborious to assure
tase (RT)-PCR have permitted the analysis of that all samples are analyzed during the log phase
minimal starting quantities of nucleic acid (as of the reaction (for both the target gene and the
little as one cell equivalent). This has made pos- normalization gene). Another method, quantita-
sible many experiments that could not have been tive competitive (QC)-PCR, has been developed
performed with traditional methods. Although and is used widely for PCR quantitation. QC-PCR
PCR has provided a powerful tool, it is imperative relies on the inclusion of an internal control
competitor in each reaction (Becker-Andre 1991;
Piatak et al. 1993a,b). The efficiency of each re-
3Corresponding author. action is normalized to the internal competitor.
E-MAIL [email protected]; FAX (415) 225-1411. A known amount of internal competitor can be
986 ~ GENOME RESEARCH 6:986-994 9 by Cold Spring Harbor Laboratory Press ISSN 1054-9803/96 $5.00
Downloaded from genome.cshlp.org on July 29, 2023 - Published by Cold Spring Harbor Laboratory Press
HElD ET AL.
porter/emission intensity of quencher measured value remains at base line. W h e n sufficient hy-
prior to PCR amplification in that same reaction bridization probe has been cleaved by the Taq
tube. For the purpose of quantitation, the last polymerase nuclease activity, the intensity of re-
three data points (ARns) collected during the ex- porter fluorescent emission increases. Most PCR
tension step for each PCR cycle were analyzed. amplifications reach a plateau phase of reporter
The nucleolytic degradation of the hybridization fluorescent emission if the reaction is carried out
probe occurs during the extension phase of PCR, to high cycle numbers. The amplification plot is
and, therefore, reporter fluorescent emission in- examined early in the reaction, at a point that
creases during this time. The three data points represents the log phase of product accumula-
were averaged for each PCR cycle and the mean tion. This is done by assigning an arbitrary
value for each was plotted in an "amplification threshold that is based on the variability of the
plot" shown in Figure 1A. The ARn mean value is base-line data. In Figure 1A, the threshold was set
plotted on the y-axis, and time, represented by at 10 standard deviations above the mean of
cycle number, is plotted on the x-axis. During the base-line emission calculated from cycles 1 to 15.
early cycles of the PCR amplification, the ARn Once the threshold is chosen, the point at which
Figure 1 PCR product detection in real time. (A) The Model 7700 software will construct amplification plots
from the extension phase fluorescent emission data collected during the PCR amplification. The standard de-
viation is determined from the data points collected from the base line of the amplification plot. Ct values are
calculated by determining the point at which the fluorescence exceeds a threshold limit (usually 10 times the
standard deviation of the base line). (B) Overlay of amplification plots of serially (1:2) diluted human genomic
DNA samples amplified with 13-actin primers. (C) Input DNA concentration of the samples plotted versus Ct. All
points represent the mean of triplicate PCR amplifications, and error bars are shown (but not always visible).
the amplification plot crosses the threshold is de- ments over a very large range of relative starting
fined as CT. CT is reported as the cycle number at target quantities.
this point. As will be demonstrated, the CT value
is predictive of the quantity of input target. Sample Preparation Validation
Several parameters influence the efficiency of
CT Values Provide a Quantitative Measurement of PCR amplification: magnesium and salt concen-
Input Target Sequences trations, reaction conditions (i.e., time and tem-
perature), PCR target size and composition,
Figure 1B shows amplification plots of 15 differ-
primer sequences, and sample purity. All of the
ent PCR amplifications overlaid. The amplifica-
above factors are c o m m o n to a single PCR assay,
tions were performed on a 1:2 serial dilution of
except sample to sample purity. In an effort to
h u m a n genomic DNA. The amplified target was
validate the m e t h o d of sample preparation for
h u m a n ~-actin. The amplification plots shift to
the factor VIII assay, PCR amplification reproduc-
the right (to higher threshold cycles) as the input
ibility and efficiency of 10 replicate sample
target quantity is reduced. This is expected be-
preparations were examined. After genomic DNA
cause reactions with fewer starting copies of the
was prepared from the 10 replicate samples, the
target molecule require greater amplification to
DNA was quantitated by ultraviolet spectroscopy.
degrade enough probe to attain the threshold
Amplifications were performed analyzing [3-actin
fluorescence. An arbitrary threshold of 10 stan-
gene content in 100 and 25 ng of total genomic
dard deviations above the base line was used to
DNA. Each PCR amplification was performed in
determine the CT values. Figure 1C represents the
triplicate. Comparison of CT values for each trip-
CT values plotted versus the sample dilution
licate sample show minimal variation based on
value. Each dilution was amplified in triplicate
standard deviation and coefficient of variance
PCR amplifications and plotted as mean values
(Table 1). Therefore, each of the triplicate PCR
with error bars representing one standard devia-
amplifications was highly reproducible, demon-
tion. The CT values decrease linearly with increas-
strating that real time PCR using this instrumen-
ing target quantity. Thus, CT values can be used
tation introduces m i n i m a l variation into the
as a quantitative measurement of the input target
quantitative PCR analysis. Comparison of the
number. It should be noted that the amplifica-
mean CT values of the 10 replicate sample prepa-
tion plot for the 15.6-ng sample shown in Figure
rations also showed minimal variability, indicat-
1B does not reflect the same fluorescent rate of
ing that each sample preparation yielded similar
increase exhibited by most of the other samples.
results for ~-actin gene quantity. The highest CT
The 15.6-ng sample also achieves endpoint pla-
difference between any of the samples was 0.85
teau at a lower fluorescent value than would be
and 0.71 for the 100 and 25 ng samples, respec-
expected based on the input DNA. This phenom-
tively. Additionally, the amplification of each
enon has been observed occasionally with other
sample exhibited an equivalent rate of fluores-
samples (data not shown) and may be attribut-
cent emission intensity change per a m o u n t of
able to late cycle inhibition; this hypothesis is
DNA target analyzed as indicated by similar
still under investigation. It is important to note
slopes derived from the sample dilutions (Fig. 2).
that the flattened slope and early plateau do not
Any sample containing an excess of a PCR inhibi-
impact significantly the calculated CT value as
tor would exhibit a greater measured ~-actin CT
demonstrated by the fit on the line shown in
value for a given quantity of DNA. In addition,
Figure 1C. All triplicate amplifications resulted in
the inhibitor would be diluted along with the
very similar CT values--the standard deviation
sample in the dilution analysis (Fig. 2), altering
did not exceed 0.5 for any dilution. This experi-
the expected CT value change. Each sample am-
m e n t contains a >100,000-fold range of input tar-
plification yielded a similar result in the analysis,
get molecules. Using CT values for quantitation
demonstrating that this method of sample prepa-
permits a much larger assay range than directly
ration is highly reproducible with regard to
using total fluorescent emission intensity for
sample purity.
quantitation. The linear range of fluorescent in-
tensity measurement of the ABI Prism 7700 Se-
Quantitative Analysis of a Plasmid After
quence Detector only spans three logs, resulting
Transient Transfection
in only a 1000-fold dynamic range of input mol-
ecules. Thus, CT values provide accurate measure- 293 cells were transiently transfected with a vec-
HElD ET AL.
tor containing a partial cDNA for h u m a n factor between any two sample means was 0.95 CT. Ten
VIII, pF8TM. A series of transfections was set nanograms of total DNA of each sample were also
up using a decreasing a m o u n t of the plasmid (40, examined for [3-actin. The results again showed
4, 0.5, and 0.1 tug). Twenty-four hours post- that very similar amounts of genomic DNA were
transfection, total DNA was purified from each present; the m a x i m u m mean [3-actin CT value
flask of cells. [3-Actin gene quantity was chosen as difference was 1.0. As Figure 3 shows, the rate of
a value for normalization of genomic DNA con- [3-actin CT change between the 100- and 10-ng
centration from each sample. In this experiment, samples was similar (slope values range between
[3-actin gene content should remain constant - 3 . 5 6 and -3.45). This verifies again that the
relative to total genomic DNA. Figure 3 shows the m e t h o d of sample preparation yields samples of
result of the [3-actin DNA measurement (100 ng identical PCR integrity (i.e., no sample contained
total DNA determined by ultraviolet spectros- an excessive a m o u n t of a PCR inhibitor). How-
copy) of each sample. Each sample was analyzed ever, these results indicate that each sample con-
in triplicate and the mean [3-actin CT values of tained slight differences in the actual a m o u n t of
the triplicates were plotted (error bars represent genomic DNA analyzed. Determination of actual
one standard deviation). The highest difference genomic DNA concentration was accomplished
21.5
Sample
PCR amplifications. As shown, pF8TM purified
21 from the 293 cells decreases (mean CT values in-
2
crease) with decreasing a m o u n t s of plasmid
20.5
5 transfected. The mean CT values obtained for
20- 7 pF8TM in Figure 4A were plotted on a standard
.c_ 8
9 curve c o m p r i s e d of serially diluted pF8TM,
19.5- 10
oQ. shown in Figure 4B. The quantity of pF8TM, b,
19- found in each of the four transfections was de-
18.5 termined by extrapolation to the x-axis of the
standard curve in Figure 4B. These uncorrected
18
1.3 1 '4 115 1'8 1'7 I'.8 1'9 .... 2.1 values, b, for pF8TM were normalized to deter-
log (ng inputgenomicDNA) mine the actual a m o u n t of pF8TM found per 100
Figure 2 Sample preparation purity. The replicate ng of genomic DNA by using the equation:
samples shown in Table I were also amplified in
tripicate using 25 ng of each DNA sample. The fig- b x 100 ng actual pF8TM copies per
ure shows the input DNA concentration (I00 and a -- 100 ng of genomic DNA
25 ng) vs. CT. In the figure, the 100 and 25 ng
points for each sample are connected by a line.
where a = actual genomic DNA in a sample and
b = pF8TM copies from the standard curve. The
normalized quantity of pF8TM per 100 ng of ge-
by plotting the mean [~-actin C T value obtained nomic DNA for each of the four transfections is
for each 100-ng sample on a [3-actin standard shown in Figure 4D. These results show that the
curve (shown in Fig. 4C). The actual genomic quantity of factor VIII plasmid associated with
DNA concentration of each sample, a, was ob- the 293 cells, 24 hr after transfection, decreases
tained by extrapolation to the x-axis. with decreasing plasmid concentration used in
Figure 4A shows the measured (i.e., non- the transfection. The quantity of pF8TM associ-
normalized) quantities of factor VIII plasmid ated with 293 cells, after transfection with 40 pug
DNA (pF8TM) from each of the four transient cell of plasmid, was 35 pg per 100 ng genomic DNA.
transfections. Each reaction contained 100 ng of This results in -520 plasmid copies per cell.
total sample DNA (as determined by UV spectros-
copy). Each sample was analyzed in triplicate
DISCUSSION
HElD ET AL.
28 30,
o. ~-~40..155+ -3.71x
26-
25-1
24
22-
I-- 20J
~ 2o
18-
15J
16-
14 lO-~
4'o 4'.o o'.s o'.1
Plasmid used in transfection (lig) log (relative factor VIII copies)
< D
32 .o i n7
30
Eo
28
26
o
24-
o
22
20-
18
._~
-2 -1 40 ug 4 ug 0.5 ug 0.1 ug
log (ng input genomic DNA) n-
plasmid used in transfection 01g)
of sample. Therefore, the potential for PCR con- for each sample minimizing potential error. The
tamination in the laboratory is reduced because system allows for a very large assay dynamic
amplified products can be analyzed and disposed range (approaching 1,000,000-fold starting tar-
of without opening the reaction tubes. Second, get). Using a standard curve for the target of in-
this method supports the use of a normalization terest, relative copy number values can be deter-
gene (i.e., [3-actin) for quantitative PCR or house- mined for any u n k n o w n sample. Fluorescent
keeping genes for quantitative RT-PCR controls. threshold values, CT, correlate linearly with rela-
Analysis is performed in real time during the log tive DNA copy numbers. Real time quantitative
phase of product accumulation. Analysis during RT-PCR methodology (Gibson et al., this issue)
log phase permits many different genes (over a has also been developed. Finally, real time quan-
wide input target range) to be analyzed simulta- titative PCR methodology can be used to develop
neously, without concern of reaching reaction high-throughput screening assays for a variety of
plateau at different cycles. This will make multi- applications [quantitative gene expression (RT-
gene analysis assays much easier to develop, be- PCR), gene copy assays (Her2, HIV, etc.), geno-
cause individual internal competitors will not be typing (knockout mouse analysis), and immuno-
needed for each gene u n d e r analysis. Third, PCR].
sample t h r o u g h p u t will increase dramatically Real-time PCR may also be performed using
with the new method because there is no post- intercalating dyes (Higuchi et al. 1992) such as
PCR processing time. Additionally, working in a e t h i d i u m b r o m i d e . The f l u o r o g e n i c probe
96-well format is highly compatible with auto- m e t h o d offers a major advantage over inter-
mation technology. calating dyes--greater specificity (i.e., primer
The real-time PCR method is highly repro- dimers and nonspecific PCR products are not de-
ducible. Replicate amplifications can be analyzed tected).
HElD ET AL.
Vehar, and C. Gorman. 1986. Construction and fully automated genome-wide polymorphism screening
characterization of an active factor VIII variant lacking [Letter]. Nature Genet. 9: 341-342.
the central one-third of the molecule. Biochemistry
25: 8343-8347. Mulder, J., N. McKinney, C. Christopherson, J. Sninsky,
L. Greenfield, and S. Kwok. 1994. Rapid and simple PCR
Fasco, M.J., C.P. Treanor, S. Spivack, H.L. Figge, and L.S. assay for quantitation of human immunodeficiency virus
Kaminsky. 1995. Quantitative RNA-polymerase chain type 1 RNA in plasma: Application to acute retroviral
reaction-DNA analysis by capillary electrophoresis and infection. J. Clin. Microbiol. 32: 292-300.
laser-induced fluorescence. Anal. Biochem. 224: 140-147.
Pang, S., Y. Koyanagi, S. Miles, C. Wiley, H.V. Vinters,
Ferre, F. 1992. Quantitative or semi-quantitative PCR: and I.S. Chen. 1990. High levels of unintegrated HIV-1
Reality versus myth. PCR Methods Applic. 2: 1-9. DNA in brain tissue of AIDS dementia patients. Nature
3 4 3 : 85-89.
Furtado, M.R., L.A. Kingsley, and S.M. Wolinsky. 1995.
Changes in the viral mRNA expression pattern correlate Piatak, M.J., K.C. Luk, B. Williams, and J.D. Lifson.
with a rapid rate of CD4 + T-cell number decline in 1993a. Quantitative competitive polymerase chain
human immunodeficiency virus type 1-infected reaction for accurate quantitation of HIV DNA and RNA
individuals. J. Virol. 69: 2092-2100. species. BioTechniques 14" 70-81.
Gibson, U.E.M., C.A. Held, and P.M. Williams. 1996. A Piatak, M.J., M.S. Saag, L.C. Yang, S.J. Clark, J.C. Kappes,
novel method for real time quantitative competitive K.C. Luk, B.H. Hahn, G.M. Shaw, and J.D. Lifson. 1993b.
RT-PCR. Genome Res. (this issue). High levels of HIV-1 in plasma during all stages of
infection determined by competitive PCR [see
Comments]. Science 259: 1749-1754.
Gorman, C.M., D.R. Gies, and G. McCray. 1990.
Transient production of proteins using an adenovirus
Prud'homme, G.J., D.H. Kono, and A.N. Theofilopoulos.
transformed cell line. DNA Prot. Engin. Tech. 2: 3-10.
1995. Quantitative polymerase chain reaction analysis
reveals marked overexpression of interleukin-1 beta,
Higuchi, R., G. Dollinger, P.S. Walsh, and R. Griffith. interleukin-1 and interferon-gamma mRNA in the lymph
1992. Simultaneous amplification and detection of nodes of lupus-prone mice. Mol. Imrnunol. 32: 495-503.
specific DNA sequences. Biotechnology 10: 413-417.
Raeymaekers, L. 1995. A commentary on the practical
Holland, P.M., R.D. Abramson, R. Watson, and D.H. applications of competitive PCR. Genome Res. 5: 91-94.
Gelfand. 1991. Detection of specific polymerase chain
reaction product by utilizing the 5 ' - - 3 ' exonuclease Sharp, P.A., A.J. Berk, and S.M. Berget. 1980.
activity of Thermus aquaticus DNA polymerase. Proc. Transcription maps of adenovirus. Methods Enzymol.
Natl. Acad. Sci. 88: 7276-7280. 65: 750-768.
Huang, S.K., H.Q. Xiao, T.J. Kleine, G. Paciotti, D.G. Slamon, D.J., G.M. Clark, S.G. Wong, W.J. Levin, A.
Marsh, L.M. Lichtenstein, and M.C. Liu. 1995a. IL-13 Ullrich, and W.L. McGuire. 1987. Human breast cancer:
expression at the sites of allergen challenge in patients Correlation of relapse and survival with amplification of
with asthma. J. Immun. 155: 2688-2694. the HER-2/neu oncogene. Science 235: 177-182.
Huang, S.K., M. Yi, E. Palmer, and D.G. Marsh. 1995b. A Southern, E.M. 1975. Detection of specific sequences
dominant T cell receptor beta-chain in response to a among DNA fragments separated by gel electrophoresis.
short ragweed allergen, Amb a 5. J. Immun. J. Mol. Biol. 98:503-517.
154: 6157-6162.
Tan, X., X. Sun, C.F. Gonzalez, and W. Hsueh. 1994. PAF
and TNF increase the precursor of NF-kappa B p50
Kellogg, D.E., J.J. Sninsky, and S. Kowk. 1990.
mRNA in mouse intestine: Quantitative analysis by
Quantitation of HIV-1 proviral DNA relative to cellular
competitive PCR. Biochim. Biophys. Acta 1 2 1 5 : 157-162.
DNA by the polymerase chain reaction. Anal. Biochem.
189: 202-208.
Thomas, P.S. 1980. Hybridization of denatured RNA and
small DNA fragments transferred to nitrocellulose. Proc.
Lee, L.G., C.R. Connell, and W. Bloch. 1993. Allelic Natl. Acad. Sci. 77: 5201-5205.
discrimination by nick-translation PCR with fluorogenic
probes. Nucleic Acids Res. 21: 3761-3766. Williams, S., C. Schwer, A. Krishnarao, C. Heid, B.
Karger, and P.M. Williams. 1996. Quantitative
Livak, K.J., S.J. Flood, J. Marmaro, W. Giusti, and K. competitive PCR: Analysis of amplified products of the
Deetz. 1995a. Oligonucleotides with fluorescent dyes at HIV-1 gag gene by capillary electrophoresis with laser
opposite ends provide a quenched probe system useful induced fluorescence detection. Anal. Biochem. (in press).
for detecting PCR product and nucleic acid
hybridization. PCR Methods Applic. 4: 357-362.
Received June 3, 1996; accepted in revised form July 29,
Livak, K.J., J. Marmaro, and J.A. Todd. 1995b. Towards 1996.
References This article cites 28 articles, 11 of which can be accessed free at:
http://genome.cshlp.org/content/6/10/986.full.html#ref-list-1
License
Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the
Service top right corner of the article or click here.