Reverse
Transcriptase Can Block
Polymerase Chain Reaction
To the Editor:
Reverse
transcription
(RT) with
subsequent amplification of cDNA by
the polymerase
chain reaction (PCR)
is
now
frequently
used
RNA. For convenience,
are usually
added
to
analyze
PCR reagents
directly
to the RT
reaction mix. Relatively high amounts
of reverse transcriptase
(25 U per reaction tube) have been used to detect
minute amounts of RNA in solution or
cell lysates (1). We have observedthat
high concentrations of some brands of
reverse transcriptase
from Moloney
murine leukemia virus (M-MuLV) or
avian myeloblastoma
virus (AMV)
prevented
amplification
of cDNA by
PCR, whereas low concentrations apparently did not interfere with PCR.
We also describe a method to circumvent this inhibition
of PCR.
Figure 1 illustrates that PCR of rat
renal NaJH antiporter
eDNA (NHE 1)
in the presence of 2 U of reverse transcriptase (M-MuLV; New England Bio-
labs, Beverly, MA) yielded a specific
amplification product, but addition of
25 U of the same enzyme completely
blocked PCR. The effect was not encountered
when only the storage
buffer was added. A similar inhibition
was seen when M-MuLV reverse transcriptase from Boehringer
Mannheim
GmbH (Mannheim,
Germany)
was
used. No inhibition was observed,
however, when 25 U of M-MuLV reverse transcriptase from other suppliers (Stratagene,
La Jolla, CA, and
Pharinacia,
Uppsala, Sweden) was
used. Inhibition of PCR by reverse
transcriptase
apparently does not depend on the substrate used because
amplification
of reverse-transcribed
hepatitis C RNA was also inhibited by
12 U of M-MuLV reverse transcriptase (A. Wicki, personal communication).
In experiments designed to determine the minimal inhibitory concentration of M-MuLV reverse transcriptase, we found that 4 U of M-MuLV
(New England Biolabs) was sufficient
to completely block PCR of -1 ng of
Na/H antiporter
DNA.
Preliminary
observations
suggest, however, that
this minimal inhibitory
concentration
may vary from batch to batch.
We also tested the ability of AMV
reverse transcriptase
to inhibit PCR.
Amplification
was not inhibited
by
12.5 U of AMV reverse transcriptase
from two different suppliers (Biofinex,
Praroman,
Switzerland, and Boehringer Mannheim
GmbH). However,
25 U of these enzymes also completely
blocked PCR. This observation confirms a recent report from another
group (3).
To gain information about the nature of the inhibition, we tested several additives for their ability to coun-
ABCDEFGHIKLMNOP
Fig. 1. Effect of reverse transcnptase source on PCR of Na/H antiporter DNA
Rat renal Na/Haritiporter DNA (NHE 1) was amplified aspreviouslydescribed(2),yielding
an amplification
product of 464 bp. Approximately1 ng of DNA was amplified with 2 U of Tag polymerase(Boehnnger
Mannheim GmbH) in 50 L of PCR buffer[200 unol/L dNTPs, 10 mmoVL Tns . HCI (pH 8.3, room
temperature), 50 mmol/L KCI, 2 mmot/L MgCI2.1 g/L gelatin,10 mLIL Triton X-100I. PCR was carried out
at 92 ‘C (initial melt, 2 mm), followed by 30 PCR cycles:92 ‘C for 10 a, 55 ‘C for 60 8, and 72 ‘C for100
a with final extension at 72 ‘C for 10 mm. A: PCRinthepresenceof2 U of M-MuLVreversetranscriptase
(New England Biolabs); B: PCR in the presence of 25 U of M-MuLV reverse transcriptase(New England
Biolabs); C PCR in the absence ofreversetranscflptase, butinthepresence of1 L of enzyme storage
buffer (New England Biolabs); D PCR in the presence of 25 U of M-MuLV reverse
transcnptase
(Stratagene):E: PCR in the presenceof 25 U of M-MuLV reverse
transcnptase
(Pharmacia); F-N: PCR in
thepresenceof 25 U of M-MuLV reversetranscriptase
(New EnglandBiolabs)
plus(F) 5 g/Lbovineserum
albumin, (G) 200 ngofyeasttRNA, (H) 200 ng of salmon sperm DNA, (I) 200 ng of Enterococcus hirae
DNA, or (K) 400, (L) 200, (M) 100,or (N) 50 ng of pBR328 DNA; a PCR in the absence of reverse
transcnptase; lambda.DNA cut with Hindlll
368 CLINICAL CHEMISTRY, Vol.39, No. 2, 1993
teract this detrimental
effect. The
addition of bovine serum albumin or
t-RNA did not overcome the inhibition, but added DNA from different
sources consistently
prevented
the inhibition of PCR (Figure 1). This suggests that reverse transcriptase
itself
or an unidentified contaminant inhibits PCR by
interaction with DNA. The
inhibition
of PCR by reverse transcriptase was not affected by varying
the PCR buffer compositions
nor by
using different enzyme batches, although the minimal
inhibitory
concentrations varied in these maneuvers
(see above).
We observed that concentrations
as
low as 4 U ofM-MuLV or 25 U of AMV
reverse transcriptase can block PCR.
The fact that two different enzymes
exhibit
inhibition
of PCR would be
compatible with a direct inhibitory
enzyme-DNA
interaction.
However,
not all M-MuLV reverse transcriptase
preparations
block
PCR, although
they all apparently represent fusion
proteins from the same clone (4). This
makes an enzyme-DNA
interaction
unlikely. Enzyme storage buffer components cannot be implicated
in the
inhibition because these are comparable among the four M-MuLV reverse
transcriptase
suppliers and because
addition of buffer only was without
effect on PCR. We must therefore consider the possibility
that contaminants of some preparations of reverse
transcriptase
interact with DNA and
thus block PCR.
In conclusion, researchers should be
aware of a possible inhibition of subsequent PCR when M-MuLV or AMV
reverse
transcriptase
is present. We
suggest that the minimal inhibitory
concentration
of a given reverse transcriptase batch be determined
individually, to avoid false-negative
results
that might seriously affect both experimental and routine testing. If greater
than minimal inhibitory concentrations of reverse transcriptase
are
needed to detect RNA, we suggest adding DNA to overcome the inhibition.
Supported by Swiss National Foundation
grants 31-25370.88 to RK. and 31-28577.90
to M.S.
References
1. Moriyama T, Murphy HR. Martin BM,
Garcia-PerezA. Detectionof specific
mENAs in single nephron segments by use
of the polymerase
chain reaction. Am J
Physiol 1990;258 (Renal Fluid Electrolyte
Physiol 27):F1470-4.
2. Krapf R, Solioz M. Purification
and
characterization of murine retroviral reverse transcriptase
expressed
in Escherichia coli. J Clin Invest 1991;88:783-8.
3. Sellner LN, Coelen RJ, Mackenzie JS.
Na/H antiporter mRNA expression in single nephron segments of rat kidney cortex.
Nucleic Acids Has 1992;20:1487-90.
4. Roth MJ, Tanese N, (loff SP. Reverse
tranacriptase
inhibits Taq polymerase
tivity. J BiolChem 1985;260:9326-35.
ac-
Catherine Fehimaim
Reto Krapf’
Marc Solioz
Inst. of Clin. Pharmacol.
University of Berne
CH-3010 Berne
Switzerland
‘Medizinische Klinik B
Kantonsspital
CH-9007 St. Gallen
Switzerland
Multianalyte
Testing
To the Editor:
As proponents of the view that simultaneous multianalyte
testing represents the next major advance in
immunoassay
methodology
(e.g., 1,
2), we welcomed the recent papers on
this topic (3, 4). Both articles rightly
caution against the problems and difficulties likely to be encountered in
developing technologies of this kind.
However, such problems are not insuperable, and undue pessimism should
not deter researchers
and manufacturers from working toward this ultimate analytical goal. Our own belief
is that the development
of sophisticated multianalyte
assay systems is
scientifically,
logistically,
and economically
essential,
and our own
(largely unpublished)
experimental
studies in this area suggest that such
development is entirely feasible. We
here briefly indicate the basis of these
convictions.
As indicated by Kricka (3), many
clinical circumstances
arise in which
the determination
of several different
analytes in individual samples is desirable. The recent report by Wald et
al. (5) of their use of a test based on
multiple analyte immunoassays
(for
a-fetoprotein,
estriol, and chorionic
gonadotropin) for the prenatal diagnosis of Down syndrome represents a
further example. The idea of simultaneously measuring several analytes is
therefore not new. The conventional
approach-which
has many wellknown limitations-has
relied on the
use of multiple labels, a concept dating from the earliest days of radioimmunoassay.
But another
pressing
stimulus to the development of multianalyte testing is the molecular het-
erogeneity of many substances of biological importance.
This phenomenon
presentsdifficulties
of increasingseriousness, which sooner or later must
be resolved. As we have emphasized
in the past (6, 7), meaningful
measurement of a mixture of substances of
different molecular structures (in the
sense that the amount of the mixture
is represented
by a single figure) is
impossible.1 Such measurements
cannot be standardized,
and their biological significance generally cannot be
interpreted.
Erythropoietin
(EPO)
represents a substance in this category that is of considerable current
interest because of its (illegal) use to
enhance athletic performance.
EPO is
present as severalisoforms,each of
which differs in structure and biological activity. Moreover, EPO prepared
by recombinant DNA techniques may
be modified to contain different proportions of these isoforms, thus altering the biological potency of the resulting mixture.
Consequently,
an
immunological
determination
of (ex-
ogenous) EPO in blood will be virtually meaningless
unless each of the
isoforms
is assayed
individually.
Many other well-known examples of
analyte
microheterogeneity
exist,
leading to similar problems of immunoassay
validity,
standardization,
and interpretation.
For this reason,
among others, the development
of
multianalyte
immunoassay
technolo-
individual
well-walls
can be elimi-
nated.
The miniaturized
technology
that
we are developing (1,2) likewise relies
on a geometrical
array; however, it is
of such difference
in principle, physical dimensions,
and diagnostic
and
logistic implications
from anything
previously
contemplated
that it represents a qualitative
rather than a
quantitative
advance in immunoassay methodology.
If successful, it will,
we believe, almost certainly
revolutionize the immunodiagnostic
field. It
relies
on antibody
“microspota,”
which, in principle, can be so small
that 100 different immunoassays
can
be encompassed
within the cross-sectional area of a human hair, thus
permitting
the
simultaneous
mea-
surement of hundreds
of different analytes in a single drop of blood. The
possibility of achieving such miniaturization relies on concepts that contradict long-established
ideas and
practice in this field, and which many
experienced
immunoassayists
regard
as counterintuitive.
For example,
some may find difficulty
in believing
that the use of a “vanishingly” small
amount of antibody located on a microspot can provide the basis for ultrasensitive,
noncompetitive
assays
that require
very short incubation
times. Yet immunoassay
design theory (encompassing
a consideration
of
diffusion kinetics) reveals that such
gies is, in our view, vital, and the
an approach yields assays of greater
ultimate emergence of such assays
sensitivity
and requiring shorter ininevitable.
cubation times than any other design.
Kncka has indicated
some existing Nor is this merely a theoretical
con“array” methodologiesthatconstitute clusion. Figure 1 shows the low-dose
simple “multianalyteassays,”albeit region of a typical thyrotropin
remost embody relatively
minor concep- sponse curve we currently obtain, intual advances. It is evident, for examdicating that microspot assay sensiple, that different reagents
can be
tivity is comparable
with, or better
placed in each well of a microtiter
than, that reported
for any other
plate, in each case with the same test
methodology,
notwithstanding
our
sample, so that the resulting “array”
present use of instrumentation
not
ofconventionalimmunoassays consti- intended for this purpose,
“conventutesa “multianalyte” assay. Nor is a
tional” fluorophors and fluorescence
major conceptual leap required
to recmeasurement
methods,
rudimentary
ognize that, by incorporating the reantibody
microspotting
techniques,
etc. The development
of specifically
agents into various solid phases, the
‘This statement sometimes provokes
dissent. The effects or activity of a mixture
of substances
in a measuring
system can be
determined, the results being expressed in
units of activity. These are not units of
amount. The relative activities of two het-
erogeneous mixtures in different measuring systems will, in general, differ; thus, it
is impossible to state without ambiguity
that the amount of one mixture is greater
than the amount of another. Only the relative
effects
of two (or more) mixtures
within a particular measuring system can
be compared.
designed multianalyte
instruments
(incorporating, if necessary, time-resolution techniques
of fluorescence
measurement)
and the use of improved solid substrates
and microspotting techniques
can be expected to
further
enhance
sensitivity
and
shorten measurement
times.
Both Kricka (3) and Kakabakos
et
al. (4) legitimately
emphasize
the
problems
facing
the
development
of
multianalyte
assays based on arrays
of this type. We agree with them on
this point, and do not wish to belittle
CLINICAL CHEMISTRY, Vol. 39, No. 2,
1993
369