ã 2002 Oxford University Press
Nucleic Acids Research, 2002, Vol. 30, No. 24 e140
T7 RNA polymerase as a self-replicating label for
antigen quanti®cation
Bakhos A. Tannous, Eleftheria Laios and Theodore K. Christopoulos1,2,*
Department of Chemistry and Biochemistry, University of Windsor, Ontario N9B 3P4, Canada, 1Department of
Chemistry, University of Patras, Patras GR-26500, Greece and 2Institute of Chemical Engineering and High
Temperature Processes, PO Box 1414, Patras GR-26500, Greece
Received August 21, 2002; Revised and Accepted October 30, 2002
ABSTRACT
Enzymes are used widely as labels in binding
assays for protein analytes, because they provide
signal ampli®cation. Efforts at improving the assay
sensitivity have been focused mainly on the synthesis of novel substrates, e.g. ¯uorogenic and chemiluminogenic ones. We report the investigation of
T7 RNA polymerase (T7RP) as a label with unique
characteristics for antigen quanti®cation. In an
in vitro, coupled (one-step) transcription/translation
reaction, T7RP catalyzes the expression of an
enzyme-coding DNA template to produce free
enzyme (luciferase) in solution. We demonstrate
that the generated luciferase is linearly related to
the input T7RP in a range covering over four orders
of magnitude. It is also shown that T7RP exhibits a
signi®cant level of self-replication (100-fold) in vitro
by acting on a DNA template comprising the T7RP
cDNA downstream of a T7 promoter. By combining
the self-replication reaction with the expression of
luciferase DNA, as low as 1400 T7RP molecules are
detectable. Furthermore, the T7RP is biotinylated,
complexed with streptavidin and used for antigen
quanti®cation in a microtiter well-based assay with
high sensitivity and reproducibility.
INTRODUCTION
Whole-genome sequencing projects have led to the identi®cation of thousands of new genes. The challenge ahead is to
unravel gene function and regulation on a genome-wide scale.
Most studies of gene function are based on the comparison of
expression pro®les between control and perturbed states,
which allows for the identi®cation of genes whose expression
is induced or suppressed. DNA microarrays provide valuable
information on gene expression at the mRNA level (1,2). Gene
function, however, is manifested through the activity of the
encoded protein. mRNA abundances do not always correlate
with protein concentrations due to signi®cant post-translational regulation (3). Consequently, the direct quantitative
analysis of proteins provides more accurate information about
biological systems. Moreover, the comparison of protein
expression pro®les in patients and normal samples (differential pro®ling) reveals potential biomarkers for diagnosis,
prognosis and monitoring of disease progression, as well as
new therapeutic targets. The challenge, however, lies in the
fact that proteins present at low concentrations are usually the
ones that mediate the cellular response to various stimuli and
are involved in the early stages of pathological processes. A
recent study has shown that half of the yeast proteome was
undetectable using two-dimensional electrophoresis followed
by mass spectrometry (4). Thus, high sensitivity, along with
speci®city, are essential requirements for any new technique
in the ®eld of proteomics, because they permit quanti®cation
of minute amounts of antigen and/or the use of smaller
numbers of cells. Furthermore, these qualities must be
combined with the ability for automation and high-throughput
protein analysis, in order to exploit the information provided
by large-scale sequencing projects.
Target ampli®cation techniques analogous to PCR that offer
exquisite sensitivity to nucleic acid analysis are not available
for protein analytes. The most sensitive protein assays are
based on the interaction of the analyte with a speci®c binder
(antibody, receptor or peptide) that is linked to a signalgenerating molecule (label). The assay sensitivity is determined, mainly, by the detectability of the label and the af®nity
of the binder. DNA fragments have been used as labels that
provide signal ampli®cation through replication [PCR (5) or
rolling circle DNA replication (6)] or expression (7).
However, the most widely used labels are enzymes (alkaline
phosphatase, horseradish peroxidase, etc.) because they provide signal ampli®cation through the turnover of many
substrate molecules to detectable product. For almost 30
years, research efforts have been focused on the synthesis of
novel substrates to allow more sensitive detection of enzyme
labels. Thus, chromogenic substrates were gradually replaced
by ¯uorogenic (8) and, more recently, chemiluminogenic ones
(9). In contrast, this work introduces an enzyme label, T7 RNA
polymerase (T7RP), which (i) has the unique ability to selfreplicate in vitro and (ii) catalyzes the in vitro synthesis of a
second enzyme (®re¯y luciferase). The resulting signal
ampli®cation is due to the generation of many enzyme
molecules in solution. The assay allows for antigen quanti®cation with high sensitivity, wide dynamic range and very
good reproducibility. Because it is performed in microtiter
*To whom correspondence should be addressed at Department of Chemistry, University of Patras, Patras GR-26500, Greece. Tel: +30 2 610 997130;
Fax: +30 2 610 997118; Email: [email protected]
e140 Nucleic Acids Research, 2002, Vol. 30, No. 24
wells, it is amenable to automation and high-throughput
analysis.
MATERIALS AND METHODS
In vitro coupled transcription/translation
The reaction mixture contained rabbit reticulocyte lysate
(TNT) from Promega Corp., Madison, WI, supplemented with
amino acids, but lacked T7RP. The appropriate DNA
templates were added to the mixture.
Determination of ®re¯y luciferase
A 2 ml aliquot of the transcription/translation reaction mixture
was added to 50 ml of luciferase substrate buffer (20 mmol/l
Tricine, pH 7.8, 1.1 mmol/l magnesium carbonate pentahydrate, 2.7 mmol/l MgSO4, 0.1 mmol/l EDTA, 33 mmol/l
dithiothreitol, 270 mmol/l coenzyme A, 530 mmol/l ATP and
470 mmol/l luciferin) (10). The luminescence was monitored
for 1 min using a liquid scintillation counter (model LS-6500;
Beckman Instruments, Fullerton, CA) in the single-photon
monitoring mode.
Biotinylation of T7 RNA polymerase
An aliquot of 1 mg (1.8 mmol) of sulfo-N-hydroxysuccinimide
ester of biotin (NHS-LC-biotin; Pierce, Rockford, IL) was
dissolved in 3 ml of dimethyl sulfoxide and then diluted to
15 mmol/l in T7RP buffer (20 mmol/l magnesium phosphate,
pH 7.7, 0.1 mol/l NaCl, 1 mmol/l dithiothreitol and 1 mmol/l
EDTA). A sample of 2 ml of the NHS-LC-biotin solution was
mixed with 1 ml (6 pmol) of T7RP (Stratagene, La Jolla, CA)
and incubated for 1 h at 4°C. The volume was increased to
50 ml with T7RP buffer containing 0.2 g/l bovine serum
albumin (BSA). The biotinylated T7RP (BT7RP) was puri®ed
from free biotin by size exclusion chromatography using NAP
columns (Amersham Pharmacia Biotech, Piscataway, NJ).
The enzyme was eluted with 1 ml of sodium phosphate buffer
pH 6.8. An aliquot of 100 ml of 103 concentrated T7RP buffer
containing 1.4 g/l BSA was added to the puri®ed BT7RP
solution and the mixture was concentrated by ultra®ltration
using microcon-30 ®lters (mol. wt cut-off = 30 000; Amicon,
Beverly, MA)
Preparation of streptavidin-biotinylated T7 RNA
polymerase complex (SA±BT7RP)
Puri®ed BT7RP (3 pmol) was mixed with 4.8 pmol
streptavidin (Sigma, St Louis, MO), diluted in T7RP buffer
(®nal volume 150 ml). The complexation reaction was allowed
to proceed for 10 min at room temperature and the SA±BT7RP
complex was used without puri®cation.
Biotinylation of monoclonal anti-prostate-speci®c antigen
(PSA) antibody
The monoclonal anti-PSA antibody solution (catalog no.
8311; Diagnostic Systems Laboratories, Webster, TX) was
dialyzed overnight against 3.5 l of 0.1 mol/l sodium
bicarbonate at 4°C. A sample of 0.2 mg of the antibody was
diluted with 0.5 mol/l carbonate buffer, pH 9.1, to a ®nal
concentration of 0.5 g/l. For biotinylation, 1 mg NHS-LCbiotin was dissolved in 50 ml of dimethyl sulfoxide and a
12.5 ml (0.25 mg) aliquot was added to the antibody solution.
PAGE 2 OF 7
The mixture was incubated for 2 h at room temperature. The
biotinylated antibody was stored at 4°C and used without
puri®cation.
T7 RNA polymerase as a label for antigen quanti®cation
U-bottom, polystyrene microtiter wells (Nunc Maxisorp; Life
Technologies, Burlington, Ontario, Canada) were coated
overnight at room temperature with 25 ml of 5 mg/l capture
anti-PSA antibody (catalog no. 8301; Diagnostic Systems
Laboratories) diluted in 50 mmol/l Tris, pH 7.8, and 0.5 g/l
NaN3. Before use, the wells were washed six times with wash
solution (50 mmol/l Tris, pH 7.4, 150 mmol/l NaCl and 1 ml/l
Tween-20). A 10 ml aliquot of PSA standard (Scripps
Laboratories, CA) diluted in 50 mmol/l Tris, pH 7.8, and
60 g/l BSA, along with 15 ml of 0.5 mg/l biotinylated anti-PSA
antibody, diluted in assay buffer (50 mmol/l Tris, pH 7.8, 60 g/l
BSA, 0.5 mmol/l KCl, 0.5 g/l NaN3 and 0.5 g/l Triton X-100),
were added to each well. The immunoreaction was allowed to
proceed for 1 h with continuous shaking. At the end of the
incubation, any unbound biotinylated anti-PSA antibody was
removed by washing the wells six times as above. Afterwards,
25 ml of 2.4 nmol/l SA±BT7RP complex (diluted in T7RP
buffer containing 1% fat-free dry skim milk) was added to
each well and incubated for 10 min. The wells were then
washed six times followed by twice with 50 mmol/l potassium
acetate. Subsequently, 25 ml of transcription/translation mixture containing 52.5 fmol luciferase cDNA (Luc-DNA) (4.3 kb
plasmid containing the T7RP promoter upstream from the
®re¯y luciferase gene) was added to each well. The coupled
in vitro transcription/translation reaction was allowed to
proceed for 90 min at 30°C and the activity of synthesized
®re¯y luciferase was measured as described above.
Antigen quanti®cation using a self-replicating T7 RNA
polymerase label
The formation of the immunocomplex on microtiter wells and
the binding of the SA±T7RP complex were carried out as
described above. Subsequently, 23.5 ml of transcription/
translation mixture containing 37.5 fmol T7RP-DNA [plasmid
pT7G1 (11), a kind gift from J. A. Wolff, Departments of
Pediatrics and Medical Genetics, Waisman Center, Madison,
WI] and 150 ng of salmon testes DNA (Sigma) was added to
each well and incubated for 60 min (self-replication phase).
Afterwards, 1.5 ml of Luc-DNA (26 fmol) was added to the
wells and incubated for another 60 min (detection phase). The
activity of synthesized ®re¯y luciferase was measured as
above.
RESULTS AND DISCUSSION
T7RP was chosen for this study because it is one of the
simplest DNA-dependent enzymes, capable of transcribing a
complete gene without the need of additional proteins.
Moreover, it is a single polypeptide chain (mol. wt 98 000),
speci®c for its promoter. T7RP has been cloned and
overexpressed in Escherichia coli (12).
The principle of antigen quanti®cation using T7RP as a
label is illustrated in Figure 1. Two approaches for measuring
T7RP, based on in vitro transcription/translation (with and
PAGE 3 OF 7
Figure 1. Assay con®guration for quanti®cation of antigens using T7RP as
a label. The antigen (Ag) is bound simultaneously to an immobilized
capture antibody and a biotinylated detection antibody. BT7RP complexed
with streptavidin (SA) is then added to the immunocomplex. The bound
T7RP is determined by in vitro coupled transcription/translation. Two approaches were explored. (a) T7RP acts on ®re¯y Luc-DNA, located downstream of the T7 promoter, to produce several molecules of active luciferase
which is measured by its characteristic bioluminogenic reaction. (b) T7RP
acts on T7RP cDNA (T7RP-DNA), positioned downstream of the T7 promoter, to generate several T7RP molecules (self-replication phase) which,
in turn, act on Luc-DNA to produce luciferase (detection phase). B, biotin.
The T7 promoter is represented by a hatched square.
without self-replication), are also shown diagrammatically in
Figure 1.
Quanti®cation of T7 RNA polymerase by coupled in vitro
transcription and translation
The goal of these experiments was to establish a relationship
between the input T7RP and the synthesized protein in an
in vitro transcription/translation system. The expression of
®re¯y luciferase was chosen because this enzyme can be
detected with high sensitivity by using its characteristic
bioluminogenic reaction (13).
Various amounts of T7RP were added to a coupled (onestep) transcription/translation reaction (rabbit reticulocyte
mixture, ®nal volume 12.5 ml) that contained the ®re¯y LucDNA under the control of the T7 promoter. The reaction was
allowed to proceed for 90 min at 30°C and then the activity of
synthesized luciferase was measured by adding 2 ml of the
expression mixture to 50 ml of luciferin substrate solution. It
was observed (Fig. 2) that the luminescence was linearly
related to the number of T7RP molecules in a range extending
over four orders of magnitude (5.2 3 104±8 3 108 molecules
of T7RP). The signal-to-background (S/B) ratio for the 5.2 3
104 molecules was 2.8.
The coupled transcription/translation process consists of a
series of complex reactions that require the concerted action of
numerous factors, such as RNA polymerase, ribosomal
subunits, translation initiation, elongation and termination
factors, aminoacyl-tRNA synthetases, etc. Nevertheless, our
data demonstrate that the ®nal outcome is a simple linear
relationship between input T7RP and the in vitro synthesized
protein over a wide range of T7RP concentrations. This forms
the basis for the development of a T7RP-based signal
ampli®cation system exploiting T7RP as a label.
Nucleic Acids Research, 2002, Vol. 30, No. 24 e140
Figure 2. Establishing a quantitative relationship between the input T7RP
and the synthesized ®re¯y luciferase in a coupled in vitro transcription/
translation system with and without self-replication of T7RP. Squares, the
transcription/translation mixture contained only Luc-DNA as template (no
self-replication). Triangles, two expression reactions (12.5 ml each) were
carried out. T7RP-DNA, placed downstream of the T7 promoter, served as
the template for the ®rst reaction (self-replication). Then, 2 ml were transferred to the second expression reaction in which the Luc-DNA served as
template (detection). Circles, a single expression reaction was carried out
with delayed addition of Luc-DNA. Transcription/translation was allowed to
proceed for 60 min with T7RP-DNA as template (self-replication) and then
Luc-DNA was added and the reaction proceeded for another 60 min prior to
luciferase measurement. cpm, counts per min.
Enhancing the detectability of T7 RNA polymerase
through self-replication
A self-replication system for T7RP was designed as follows.
Two consecutive 90 min in vitro expression reactions were
carried out (12.5 ml each). In the ®rst reaction the T7RP
catalyzed the transcription of its cognate gene positioned
downstream of the T7 promoter (T7RP-DNA) (11,14) and the
generated RNA was translated simultaneously into active
T7RP molecules. The newly synthesized T7RP also acted on
the T7RP-DNA template to produce more of the enzyme (selfreplication). The T7RP was then measured by transferring 2 ml
into another expression reaction, containing 35 fmol
Luc-DNA, and monitoring the synthesized luciferase. The
extent of self-replication is a function of the T7RP-DNA level,
as indicated by the increase in the luminescence as the T7RPDNA concentration increases (Fig. 3A). At the optimum level
of T7RP-DNA, the self-replication process caused a 110-fold
increase in the signal compared to a reaction that contained no
T7RP-DNA.
Similar experiments with decreasing amounts of T7RP
(aimed at estimating the detectability of the polymerase)
revealed a low level of `illegitimate' transcription of T7RPDNA in the absence of T7RP. This was attributed to a
eukaryotic RNA polymerase activity that is present in the
rabbit reticulocyte extract and initiates a low level of
transcription of T7RP-DNA, generating a few T7RP molecules which, in turn, are ampli®ed by entering the
e140 Nucleic Acids Research, 2002, Vol. 30, No. 24
Figure 3. (A) Effect of the amount of T7RP-DNA on the extent of selfreplication of T7RP. Two consecutive transcription/translation reactions
(12.5 ml each) were performed. In the ®rst reaction T7RP acted on various
concentrations of T7RP-DNA (self-replication). Then 2 ml of the reaction
mixture was transferred into a second expression reaction containing a
constant amount of Luc-DNA (35 fmol). The synthesized ®re¯y luciferase
was measured as described in Materials and Methods. The luminescence is
plotted against the amount of T7RP-DNA in the transcription/translation
reaction. The arrow indicates the signal obtained without self-replication
(absence of T7RP-DNA). (B) Effect of the concentration of salmon DNA
on the extent of `illegitimate' expression of T7RP-DNA in the absence of
T7RP (solid line) and on the extent of T7RP-catalyzed expression (dashed
line). Experiments were performed as above using two consecutive expression reactions, the ®rst reaction containing 25 fmol T7RP-DNA and various
amounts of salmon testes DNA and the second reaction containing 35 fmol
Luc-DNA. The percent luminescence is plotted as a function of the amount
of salmon DNA in the transcription/translation reaction mixture. The value
of 100% is de®ned as the luminescence obtained with no salmon DNA
present. (C) Study of the effect of the T7RP-DNA:Luc-DNA molar ratio on
the yield of an expression reaction that combines self-replication of T7RP
and luciferase synthesis. A delayed addition protocol was performed
(60 min±60 min). Increasing amounts of T7RP-DNA were used with 8.75
(squares), 17.5 (circles) and 35 fmol (triangles) Luc-DNA. (D) Effect of the
reaction times before and after the addition of Luc-DNA (delayed addition
protocol) on the yield of an expression reaction mixture that contains T7RP
and T7RP-DNA. The reaction starts with the addition of 25 fmol
T7RP-DNA. The ®rst and second numbers of each pair on the x-axis
correspond to the incubation time before and after the addition of 35 fmol
Luc-DNA, respectively. The ®rst column (90 min) represents the signal
obtained in the absence of self-replication (no T7RP-DNA).
self-replication cycle. This activity was not detectable in the
absence of T7RP-DNA. Because illegitimate transcription
compromises the detectability of T7RP, we carried out
experiments to minimize it by adding various amounts of
salmon DNA to the rabbit reticulocyte extract. Addition of
100 ng salmon DNA suppressed illegitimate transcription by
98%, whereas it caused only a 15% decrease in the T7RPcatalyzed transcription of T7RP-DNA (Fig. 3B).
In order to estimate the detectability of T7RP in a selfreplication system, various amounts of the enzyme were added
to the ®rst expression reaction containing 25 fmol T7RP-DNA
followed by a separate expression reaction containing 35 fmol
Luc-DNA. The linearity extends from 1.4 3 103±107
PAGE 4 OF 7
molecules (Fig. 2). The S/B ratio at 1400 T7RP molecules
was 2.5.
We investigated the possibility of combining self-replication with the detection of T7RP in a single reaction mixture
containing both T7RP-DNA and Luc-DNA templates.
However, it was observed that self-replication was suppressed
dramatically due to competition between the two templates for
binding to a limited number of T7RP molecules. Therefore, a
delayed addition protocol was designed in which the T7RP
was ®rst allowed to act on its cognate gene (self-replication)
followed by the addition of Luc-DNA, in the same reaction
mixture.
The T7RP-DNA/Luc-DNA ratio, as well as the incubation
times required before and after the addition of Luc-DNA, were
optimized to ensure ef®cient self-replication and detection
with the delayed addition protocol. The T7RP-DNA/LucDNA molar ratio was studied in the range 0.05±3 at three
levels of Luc-DNA (Fig. 3C). The luminescence increases as
the ratio becomes greater, due to increased self-replication,
and reaches a plateau when the molar ratio of the two
templates becomes 1±1.5. For the same molar ratio, the signal
increases by increasing the concentration of Luc-DNA.
Various combinations of reaction times before and after
Luc-DNA addition were studied (Fig. 3D). The maximum
signal was achieved with 60 min±60 min, respectively, giving
a 50-fold enhancement over the assay that contains no T7RPDNA. Other combinations, such as 30±60, 30±90 and 60±30,
compromised either the yield of the self-replication reaction or
the detection reaction, thus giving a lower signal. In particular,
the 0±90 combination (both templates added simultaneously at
the beginning of expression) gave the lowest yield of selfreplication (Fig. 3D).
The detectability of T7RP using the optimized single
expression reaction protocol (delayed addition protocol) was
8.5 3 103 molecules, with a S/B ratio of 4.3. The
luminescence was a linear function of the amount of T7RP,
up to 107 molecules (Fig. 2).
Biotinylation of T7 RNA polymerase, complexation with
streptavidin and application to antigen quanti®cation
The effect of biotinylation on the activity of T7RP was studied
by reacting with increasing concentrations of the sulfo-Nhydroxysuccinimide ester of biotin (NHS-LC-biotin) at pH 7.7
and 9.0. All reactions were incubated for 1 h at 4°C and the
activity of T7RP was measured by in vitro coupled transcription/translation. Inactivation of T7RP becomes signi®cant at
biotin:T7RP molar ratios greater than 5 (Fig. 4A), due to
modi®cation of free amino groups that are necessary for full
activity. The inactivation was more extensive at pH 9.0 than
pH 7.7 because the biotinylation reaction is more ef®cient
when the NH2 groups are deprotonated.
A microtiter well-based `two-site' immunoassay was
developed for PSA, as a model (Fig. 1). The antigen was
bound both by an immobilized capture antibody and a
biotinylated detection antibody. The two anti-PSA antibodies
are directed to different epitopes. BT7RP was complexed to
streptavidin and added to the immunocomplex. The solid
phase-bound T7RP was measured by in vitro expression.
The complexation of streptavidin with BT7RP was studied
at SA:BT7RP molar ratios ranging from 0.5 to 12 (Fig. 4B)
and the complexes were applied directly to the assay of 20 fmol
PAGE 5 OF 7
Figure 4. (A) Study of the effect of biotinylation on the activity of T7RP.
Biotinylation was performed at various NHS-LC-biotin:T7RP molar ratios
at pH 7.7 and 9.0, as described in Materials and Methods. T7RP was then
determined by in vitro expression using 35 fmol Luc-DNA as template. The
percent luminescence is plotted versus the molar ratio NHS-LC-biotin:T7RP
at pH 7.7 (solid line) and pH 9.0 (dashed line). The value 100% is de®ned
as the signal obtained from non-biotinylated T7RP. (B) Optimization of the
streptavidin SA:BT7RP molar ratio for the preparation of the SA±BT7RP
complex. The complex was prepared as described in Materials and Methods
and 60 fmol were used (without prior puri®cation) for antigen quanti®cation
(20 fmol PSA). The solid and dashed lines correspond to the signal and
S/B ratio, respectively. The background is de®ned as the luminescence obtained in the absence of antigen. (C) Time dependence of the transcription/
translation reaction with T7RP immobilized on the solid phase. The
immunoassay was performed as described in Materials and Methods.
Following reaction with the SA±BT7RP complex, a 25 ml transcription/
translation mixture was added containing 52.5 fmol Luc-DNA as template.
Expression was allowed to proceed for various time intervals (up to
180 min). The solid and dashed lines correspond to the signal and S/B ratio,
respectively. The background is de®ned as the luminescence obtained in the
absence of antigen. (D) Effect of the concentration of SA±BT7RP complex
on the luminescence (solid line) and the S/B ratio (dashed line) obtained
from the assay of 10 fmol antigen. Luc-DNA was used as template for
T7RP. The immunoassay was performed as described in Materials and
Methods.
antigen. The luminescence reached a maximum at a
SA:BT7RP molar ratio of 1.5. The signal dropped sharply at
lower or higher ratios. When BT7RP is in excess, all four
biotin-binding sites of streptavidin are occupied and the
complexes cannot bind to the biotinylated antibody on the
solid phase. On the other hand, when streptavidin is in excess,
free streptavidin competes with SA±BT7RP complex for
binding to the well (15).
The time-course of the in vitro transcription/translation
process with immobilized T7RP was studied up to 180 min
(Fig. 4C). The luminescence increased with time but the S/B
ratio reached a plateau at 120 min. The background was
de®ned as the luminescence obtained when no antigen was
present in the well. The concentration of SA±BT7RP added to
the well also affected both the signal and the background of
the assay and therefore its detectability (Fig. 4D). The signal
increased with the SA±BT7RP concentration and a plateau
was reached at 7 nmol/l. The S/B ratio, however, was highest
Nucleic Acids Research, 2002, Vol. 30, No. 24 e140
Figure 5. Assessing the sensitivity and analytical range for antigen
quanti®cation using T7RP as a label, (squares) without self-replication
(absence of T7RP-DNA) and (circles) with self-replication of T7RP. For
self-replication, T7RP-DNA was included in the expression reaction mixture
and the delayed addition protocol was employed. The immunoassays were
carried out as described in Materials and Methods. The luminescence (corrected for the background) is plotted against the concentration of PSA present in the well. The background is de®ned as the signal obtained in the
absence of antigen. Also, shown (triangles) are data for a classical ELISA
that uses a streptavidin±alkaline phosphatase conjugate (SA±AP) for detection and p-nitrophenylphosphate as a chromogenic substrate. Following
immunocomplex formation (as described in Materials and Methods),
SA±AP was added (1000 U), instead of SA±BT7RP, and incubated for
15 min. After washing out the excess of conjugate, the substrate was added
for 30 min in the dark, followed by absorbance measurement at 405 nm.
at 3 nmol/l and then dropped because of the increasing nonspeci®c binding of the complex to the solid phase.
The sensitivity and dynamic range of the optimized
immunoassay were assessed by analyzing serial dilutions of
the antigen, in the appropriate buffer. The S/B ratio was
plotted as a function of the amount of PSA in the assay mixture
(Fig. 5). In the absence of T7RP-DNA (no self-replication),
195 amol antigen were detected with a S/B ratio of 2.3. When
the self-replication system was used (single expression
reaction with a delayed addition of Luc-DNA), as low as
12 amol PSA could be detected with a S/B ratio of 1.9. The
dynamic range of the assay extended up to 50 000 amol
(Fig. 5). For comparison of detectabilities, we also performed
a classical ELISA assay in which the immunocomplex was
detected by using a streptavidin±alkaline phosphatase conjugate (New England BioLabs) and the enzymic activity was
determined by adding p-nitrophenylphosphate as a substrate
(Sigma) and measuring the absorbance at 405 nm on a
microplate photometer (model EL-307C; BioTek Instruments,
Winooski, VT). As shown in Figure 5, a S/B ratio of 2.1 was
obtained for 7300 amol antigen. Consequently, the proposed
assay offers ~600-fold higher detectability and a much wider
analytical range.
To assess the reproducibility of the proposed immunoassay
(including all the steps, i.e. coating of the wells, immunocomplex formation, binding of SA±BT7RP, in vitro coupled
transcription/translation and luciferase measurement), we
analyzed samples containing 0.1, 1 and 10 fmol PSA. The
e140 Nucleic Acids Research, 2002, Vol. 30, No. 24
percent coef®cients of variation were 7.8, 7.5 and 8.1,
respectively (n = 7).
It should be noted that the assay is carried out in three steps,
namely immunocomplex formation, reaction with the
SA±T7RP complex and ®nally self-replication of T7RP and
luciferase synthesis. The wells are washed at the end of each
step to remove all unbound components of the samples as well
as the excess of reagents. Consequently, the transcription/
translation reaction mixture does not come into contact with
the sample components (removed during ®rst washing).
It has been shown previously that T7RP transcripts that are
neither capped nor polyadenylated can be ef®ciently translated
in eukaryotic systems (16,17). Optimizing the structure of
T7RP-DNA and Luc-DNA templates may further enhance the
sensitivity of the system. For instance, suitable enhancer and
transcription termination sequences may be incorporated to
increase the yield of both self-replication of T7RP and
luciferase synthesis. Insertion of both T7RP-DNA and
Luc-DNA templates, under the control of the T7 promoter,
into a single vector may also be investigated for higher
yields. Besides ®re¯y luciferase DNA, cDNAs for other
highly detectable proteins may be employed, e.g. green
¯uorescent protein (18), alkaline phosphatase, aequorin (19),
etc. A eukaryotic (rabbit reticulocyte) coupled transcription/
translation system was used throughout this work. However,
prokaryotic (E.coli) systems may also be tested by using DNA
templates containing a T7 promoter and a Shine±Dalgarno
sequence for ribosome binding.
We have used streptavidin as a linker between the
biotinylated detection antibody and BT7RP. An alternative
means of biotinylation would be to express a recombinant
T7RP/strep-tag (II) fusion protein. Strep-tag (II) is an eight
amino acid polypeptide, generated by combinatorial engineering, which exhibits binding properties for streptavidin or
for a mutant form of streptavidin called strep-tactin (20).
The assay is amenable to automation because it is
performed in microtiter wells. Although we used the 96-well
format, the technology is transferable to microtiter plates with
larger numbers of wells as well as to microfabricated wells for
high throughput parallel analysis of multiple proteins.
While this manuscript was in progress, another technique
for protein analysis was reported (21) that exploited the T7RP
reaction for signal ampli®cation (IDAT, immunodetection
ampli®ed by T7RP). However, the two concepts are fundamentally different. In IDAT, a double-stranded DNA fragment
(the substrate for T7RP) was used as the antibody label.
Following immunocomplex formation, the label was transcribed for 4 h by excess T7RP to generate multiple RNA
copies. Since RNA by itself is not a suitable reporter molecule,
a 32P-labeled ribonucleotide was incorporated during transcription in order to confer high detectability. The labeled
transcripts were then separated by electrophoresis on a
denaturing (urea) polyacrylamide gel and measured by
autoradiography. Conversely, the present work uses the
enzyme T7RP as a label that by acting on both T7 cDNA
and Luc cDNA templates produces the corresponding mRNAs
that upon translation generate T7RP (self-replication of T7RP)
and ®re¯y luciferase (indicator enzyme).
Although at present the proposed assay does not achieve the
reported detectability of IDAT, it offers a number of
signi®cant advantages. (i) The assay does not use radioactive
PAGE 6 OF 7
isotopes. During the last two decades there have been intense
research efforts in the ®elds of immunoassays and DNA/RNA
hybridization assays towards replacing radioactive labels with
non-radioactive alternatives, in order to avoid the hazards
associated with the use and disposal of radioactivity. As a
consequence, most clinical laboratories now use exclusively
non-radioactive assays and the use of non-radioactive detection systems in research laboratories is expanding rapidly.
(ii) In contrast to IDAT, which requires tedious denaturing gel
electrophoresis and autoradiography, the present assay is
performed entirely in microtiter wells, thereby allowing for
automation and high-throughput analysis. (iii) A quantitative
relationship is established between the luminescence signal
and the amount of antigen with a dynamic range covering
almost four orders of magnitude. (iv) Compared to IDAT, the
proposed assay is much shorter. Indeed, after immunocomplex
formation, IDAT requires a 4 h transcription step followed by
time-consuming electrophoresis and autoradiography. In contrast, the present technique requires ~2 h for quanti®cation of
the immunocomplexes. (v) Because of the self-replication
reaction, ampli®cation in the proposed system is exponential,
whereas IDAT involves linear ampli®cation of the label.
In recent years, research efforts have focused increasingly
on the identi®cation of binders, with the requisite af®nity and
speci®city, for a large number of protein analytes. Various
approaches include recombinant antibodies selected by phage
(22) or ribosome display (23), RNA or DNA aptamers (24)
and small organic compounds selected through combinatorial
library methods (25,26). The extent and/or the position of
biotinylation of proteins, nucleic acids and small molecules
can be controlled to achieve minimal interference with the
interaction between the binder and the protein analyte.
Consequently, SA±BT7RP is a universal detection reagent
that can bind with high af®nity (Kd = 10±14 M) to any
biotinylated binder±analyte complex.
ACKNOWLEDGEMENTS
We would like to thank J. A. Wollf for providing the plasmid
pT7G1. This work was supported by grants from the National
Science and Engineering Research Council of Canada
(NSERC).
REFERENCES
1. Wodicka,L., Dong,H., Mittmann,M., Ho,M.H. and Lockhart,D.J. (1997)
Genome-wide expression monitoring in Saccharomyces cerevisiae. Nat.
Biotechnol., 15, 1359±11367.
2. Schena,M., Shalon,D., Heller,R., Chai,A., Brown,P.O. and Davis,R.W.
(1996) Parallel human genome analysis: microarray-based expression
monitoring of 1000 genes. Proc. Natl Acad. Sci. USA, 93, 10614±10619.
3. Gygi,S.P., Rochon,Y., Franza,B.R. and Aebersold,R. (1999) Correlation
between protein and mRNA abundance in yeast. Mol. Cell. Biol., 19,
1720±1730.
4. Gygi,S.P., Corthals,G.L., Zhang,Y., Rochon,Y. and Aebersold,R. (2000)
Evaluation of two-dimensional gel-electrophoresis-based proteome
analysis technology. Proc. Natl Acad. Sci. USA, 97, 9390±9395.
5. Sano,T., Smith,C.L. and Cantor,C.R. (1992) Immuno-PCR: very
sensitive antigen detection by means of speci®c antibody-DNA
conjugates. Science, 258, 120±122.
6. Schweitzer,B., Wiltshire,S., Lambert,J., O'Malley,S., Kukanskis,K.,
Zhu,Z., Kingsmore,S.F., Lizardi,P.M. and Ward,D.C. (2000)
Immunoassays with rolling circle DNA ampli®cation: a versatile
PAGE 7 OF 7
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
platform for ultrasensitive antigen detection. Proc. Natl Acad. Sci. USA,
97, 10113±10119.
Chiu,N.H.L. and Christopoulos,T.K. (1999) Two-site expression
immunoassay using a ®re¯y luciferase-coding DNA label. Clin. Chem.,
45, 1954±1959.
Christopoulos,T.K. and Diamandis,E.P. (1996) Fluorescence
immunoassays. In Diamandis,E.P. and Christopoulos,T.K. (eds),
Immunoassay. Academic Press, San Diego, CA, pp. 309±335.
Kricka,L.J. (1996) Chemiluminescence immunoassay. In Diamandis,E.P.
and Christopoulos,T.K. (eds), Immunoassay. Academic Press, San Diego,
CA, pp. 337±353.
Titus,D.E. (1991) Promega Protocols and Applications Guide, 2nd Edn.
Promega Corp., Madison, WI.
Deng,H. and Wolff,J.A. (1994) Self-amplifying expression from the T7
promoter in 3T3 mouse ®broblasts. Gene, 143, 245±249.
Davanloo,P., Rosenberg,A.H., Dunn,J.J. and Studier,F.W. (1984)
Cloning and expression of the gene for bacteriophage T7 RNA
polymerase. Proc. Natl Acad. Sci. USA, 81, 2035±2039.
Wilson,T. and Hastings,J.W. (1998) Bioluminescence. Annu. Rev. Cell.
Dev. Biol., 14, 197±230.
Dubendorff,J.W. and Studier,F.W. (1991) Creation of a T7 autogene.
Cloning and expression of the gene for bacteriophage T7 RNA
polymerase under control of its cognate promoter. J. Mol. Biol., 219,
61±68.
Diamandis,E.P. and Christopoulos,T.K. (1991) The biotin-(strept)avidin
system: principles and applications in biotechnology. Clin. Chem., 37,
625±636.
Gurevich,V.V., Pokrovskaya,I.D., Obukhova,T.A. and Zozulya,S.A.
(1991) Preparative in vitro mRNA synthesis using SP6 and T7 RNA
polymerases. Anal. Biochem., 195, 207±213.
Nucleic Acids Research, 2002, Vol. 30, No. 24 e140
17. Craig,D., Howell,M.T., Gibbs,C.L., Hunt,T. and Jackson,R.J. (1992)
Plasmid cDNA-directed protein synthesis in a coupled eukaryotic in vitro
transcription-translation system. Nucleic Acids Res., 20, 4987±4995.
18. Tsien,Y. (1998) The green ¯uorescent protein. Annu. Rev. Biochem., 67,
509±544.
19. White,S.R. and Christopoulos,T.K. (1999) Signal ampli®cation system
for DNA hybridization assays based on in vitro expression of a DNA
label encoding apoaequorin. Nucleic Acids Res., 27, e25.
20. Skerra,A. and Schmidt,T.G.M. (1999) Applications of a peptide ligand
for streptavidin: the strep-tag. Biomol. Eng., 16, 79±86.
21. Zhang,H.T., Kacharmina,J.E., Miyashiro,K., Greene,M.I. and
Eberwine,J. (2001) Protein quanti®cation from complex protein mixtures
using a proteomics methodology with single-cell resolution. Proc. Natl
Acad. Sci. USA, 98, 5497±5502.
22. de Wildt,R.M.T., Mundy,C.R., Gorick,B.D. and Tomlinson,I.M. (2000)
Antibody arrays for high-throughput screening of antibody-antigen
interactions. Nat. Biotechnol., 18, 989±994.
23. Hanes,J., Schaf®tzel,C., Knappik,A. and Pluckthun,A. (2000) Picomolar
af®nity antibodies from a fully synthetic naive library selected and
evolved by ribosome display. Nat. Biotechnol., 18, 1287±1292.
24. Hesselberth,J., Robertson,M.P., Jhaveri,S. and Ellington,A.D. (2000)
In vitro selection of nucleic acids for diagnostic applications. Rev. Mol.
Biotechnol., 74, 15±25.
25. Han,Y. and Kodadek,T. (2000) Peptides selected to bind the Ga180
repressor are potent transcriptional activation domains in yeast. J. Biol.
Chem., 275, 14979±14984.
26. Shuker,S.B., Hajduk,P.J., Meadows,R.P. and Fesik,S.W. (1996)
Discovering high-af®nity ligands for proteins: SAR by NMR. Science,
274, 1531±1534.