Molecular and Cellular Probes (1992) 6, 251-256

Use of modified nucleotides and uracil-DNA glycosylase (UNG) for the control of contamination in the PCR-based amplification of RNA

Jenny Pang, John Modlin and Robert Yolken* Department of Pediatrics, Johns Hopkins University, 600 N . Wolfe Street, CMSC 1109, Baltimore, Maryland 21205, U .S.A . (Received 10 December 1991, Accepted 3 February 1992)

The inadvertent carryover of amplified fragments of nucleic acids (amplicons) is a potential source of contamination in the polymerase chain reaction . Recently, a method has been developed to generate amplicons with deoxyuracil triphosphate (dUTP) and to specifically hydrolyze these amplicons with uracil-DNA glycosylase (UNG) following the completion of the assay . We evaluated this system for the specific amplification of RNA from coxsackievirus A3 and B3 . We found that RNA from both viruses could be amplified with dUTP, although the use of this triphosphate in place of TTP resulted in some loss of assay sensitivity. We also found that the dUTP-containing amplicons could be efficiently hydrolyzed by UNG, resulting in a 10,000,000-fold reduction in amplicon concentration with little effect on the native nucleic acid . The dUTP-UNG method has a great deal of potential for reducing amplicon contamination during the routine performance of nucleic acid amplification reactions .

KEYWORDS: Polymerase chain reaction, coxsackievirus, contamination, viral diagnosis .

INTRODUCTION The polymerase chain reaction (PCR) is attaining

of contamination in the clinical laboratory environ-

widespread usage for the amplification and detection

ment . Specifically, if the amplicons are inadvertently

of microbial nucleic acids in experimental settings .' , '

introduced into reaction vessels containing negative

PCR utilizes oligonucleotide primers and a thermo-

samples, the subsequent performance of PCR reac-

stable DNA polymerase (generally Taq polymerase) in

tions will result in their further amplification and the

a cyclical fashion to achieve the logarithmic amplifi-

generation of false-positive signals .'-10 The occur-

cation of small numbers of target ." PCR conditions

rence of contamination is a particular problem in the

need to be modified for the detection of RNA targets

detection of RNA templates since the need for the

such as messenger RNA or RNA viral genomes . Since

performance of an RNA-DNA conversion step intro-

Taq polymerase has little activity for RNA, the RNA

duces an additional enzymatic reaction and hence an

must be converted to cDNA by reaction with reverse

additional opportunity for the occurrence of cross-

transcriptase prior to Taq-catalyzed amplification .'

contamination .

The use of enzymatic target amplification allows

Several isolation and sterilization strategies have

the PCR assays to attain high degrees of sensitivity .

been proposed in order to minimize the occurrence

However, the amplified segments of DNA (referred to

of false-positive reactions due to amplicon contami-

as 'amplicons'), which function as templates for

nation .""' The efficacy of these regimens has been

further rounds of amplification, can serve as sources

limited by the fact that DNA is extremely stable and

*Author to whom correspondence should be addressed . 0890-8508/92/030251 + 06 $03 .00/0

251

© 1992 Academic Press Limited

252

J. Pang et al.

that strict isolation regimens are difficult to enforce in clinical laboratory settings . Cross-contamination and the generation of false-positive reactions constitute major impediments to the application of PCR in clinical diagnostic settings . Recently, methods have been devised for the specific hydrolysis of DNA which has been amplified by PCR . These systems make use of the incorporation of modified nucleotides during the course of the PCR reaction and the specific chemical or enzymatic degradation of the modified bases ."," One system

and 10 pl 10% sodium docecyl sulphate were added to 200 pl samples of each viral dilution . Each specimen was then extracted with 1 volume of 1 :1 :08, phenol : chloroform : isoamyl alcohol, and reextracted with 1 :0 . 7 1 x TE buffer :10% SDS . Aqueous phases were saved and combined with 2 . 5 volumes of 100% ethanol and 0 . 25 volume of 7 . 5 M ammonium acetate . To optimize RNA precipitation, aqueous phases were incubated overnight at - 20°C .

makes use of the incorporation of deoxyuridine triphosphate (dUTP) during the course of the Taq polymerase catalyzed enzymatic amplification . After the performance of PCR and the measurement of the amplicons, nucleic acids containing dUTP are specifically hydrolyzed by reaction with uracil-DNA glycosylase (UNG) . Since this enzyme does not efficiently utilize thymidine as a substrate, most naturally occurring DNA species will not be cleaved by this enzyme ." The use of dUTP and UNG in the performance of PCR reactions have been reported to reduce amplicon contamination and allow for the detection of HIV in low concentrations . However, the sensitivity and specificity of the dUTP-UNG system for RNA detection have not been characterized . The detection of RNA is important for the diagnosis of infections with viruses which contain RNA genomes and which do not undergo endogenous conversion to cDNA as part of their life cycle. We utilized RNA from two such viruses of the family picornaviridae, coxsackievirus A3 and coxsackievirus B3, to evaluate the parameters of the dUTP-UNG system for the specific amplification of viral RNA .

Source of oligonucleotide primers

MATERIALS AND METHODS

The sequences used were those described by Rotbart et al ." The sequence of the downstream and upstream primers were 5'-ACCGACGAATACCACTGTTA and 5'-CCTCCGGCCCCTGAATGCGGCTAAT, respectively . The primers were synthesized by an automated oligonucleotide synthesizer using standard methodologies.

Conversion of viral RNA to CDNA Viral RNA, prepared as described above, was washed once with 100% ethanol and resuspended in 10 pl 1 X TE buffer containing 40 units of RNasin (Promega). The reverse transcription reaction was performed by the addition of the following to the RNA : 10 pl of 5 x M-MLV reverse transcription buffer (250 mm TrisHCI, pH 8. 3, 375 mm KCL, 15 MM MgCl 2) ; 10 pl of diethylpyrocarbonate-treated (DEPC) H 2 0 ; 2 . 5 pl of 20 gm downstream oligonucleotide primer ; 2 . 5 pl each of 10 mm ATP, CTP, GTP, and TTP ; 5 pl of 10 mm DTT; and 200 units of M-MLV reverse transcriptase (BRL, Gaithersburg, MD) . Samples were incubated for 90 min at 37 ° C and the resulting cDNA was further amplified by the PCR reactions described below .

Source and cultivation of virus Performance of the PCR Coxsackievirus A3 (CA3) and coxsackievirus B3 (CB3) stocks were originally obtained from the American Type Culture Collection, Rockville, MD and propagated in RD and Vero cell lines, respectively . The titre of the initial virus stocks by cell culture was 4 x 10 5 ml"' and 8 x 108 ml"', respectively . Uninfected RD and Vero cells were used as negative controls for reactions involving the amplification of CA3 and CB3 respectively.

Extraction of viral RNA A 2 pg aliquot of yeast tRNA (BRL, Gaithersburg, MD)

All PCR reagents were purchased from Perkin Elmer Cetus Corporation, Norwalk, CT . The PCR reactions were performed in 50 pl volumes by proportionately halving the contents of a standard 100 pl reaction containing : 10 pl of 10 x PCR Buffer II ; 2 pl each of 10 mm ATP, 10 mm CTP, 10 mm GTP, 10 mm TTP ; 4 gl of each downstream and upstream primers; 4 pl of 25 mm MgCl 2 , 0 . 5 pl of Taq polymerase, 10 pl of cDNA template, and DEPC-treated H 2 O to the final volume . For some experiments dUTP was substituted for dTTP and the concentration of MgCl 2 was increased, as described below . The following conditions were utilized for the

253

Control of contamination in RNA amplification

performance of the PCR reaction : four cycles of 1 min denaturation (94° C), 1 min annealing (55 ° C), and 1 min primer extension (72°C) followed by 25 cycles of 30 s denaturation, 30 s annealing, and 1 min primer extension . The products of the amplification were separated by electrophoresis on a 6% polyacrylamide gel and visualized following staining with silver nitrate as previously described . Uninfected RD or Hela cell cDNA, the contents of reverse transcription mixture, and DEPC-treated H2 0 were analysed as controls for contamination at the extraction, reverse transcription, and PCR steps, respectively . Reactions with uracil-DNA glycolase (UNG) were performed following the manufacturers' instructions .

Results Optimization of concentrations of MgCI 2 and dUTP Initial experiments were directed at the optimization of the PCR conditions utilizing dUTP instead of the TTP which is usually employed along with the other deoxynucleotides . We found that we could achieve optimal sensitivity using dUTP at a concentration of 400 µnn resulted along with MgCI2 at a concentration of 2 lam . The use of lower concentrations of dUTP and MgCI 2 resulted in a substantial decrease in reaction sensitivity, while the use of higher concentrations resulted in the generation of extraneous bands which compromized reaction specificity (Fig . 1) . Further experiments were thus performed utilizing dUTP at a concentration of 400 gm and MgCI 2 at a concentration of 2 mm .

TTP 200 µM

Reaction of amplified products with uracil N-glycosylase (UNG) We then measured the ability of UNG to cleave amplified DNA containing dUTP . For these experiments, RNA was isolated from CB3 by phenol-chloroform extraction, reverse transcribed as described above, and amplified by PCR using 400 µnn UTP and 2 mm MgCl 2 . The amplified products, which contained approximately 20 ng of DNA, were reextracted with phenol-chloroform and precipitated with ethanol . The DNA pellet was washed with 100% ethanol, resuspended in 1 X TE buffer, and serially diluted 10-fold with DEPC-treated H 2 0 . Two aliquots of diluted products were re-amplified in a second PCR reaction . One set was pre-treated with UNG for 10 min at room temperature at a final concentration of 10 units ml -1 (0 . 5 units/ 50Itl reaction volume) . The other set was untreated . Both sets were then incubated at 95 ° C for 13-15 min and amplified by a second PCR, using dUTP at a concentration of 400 nn-t . As depicted in Fig . 2, the DNA re-amplified in the absence of UNG displayed reaction product even if the input amplicons were diluted as much as 1 :109 . On the other hand, when the dUTP containing amplicons were re-amplified in the presence of UNG, no product was visualized when the amplicons were diluted more than 10-fold . A faint band was noted in the 1 :10 dilution of amplicons and a strong band was noted in the reaction to which undiluted amplicons had been added . The use of UNG thus resulted in a greater than 10,000,000 (10')-fold reduction in the amount of amplicon which could be detected following re-amplification.

dUTP 600 /LM

dUTP 400 ft.M

Fig . 1 . The effect of oligonucleotide and MgCl, concentration on the amplification of coxsackievirus B3 . Coxsackievirus B3 RNA was prepared and diluted as described in the text, and amplified by PCR using TTP or dUTP at the indicated concentration along with an oligonucleotide mixture containing dCTP, dATP and dGTP at concentrations of 200 gm . The reactions employing TTP was performed utilizing MgCI 2 at a concentration of 1 pm while the reactions with dUTP were performed using MgCI 2 at a concentration of 2 Rm . The products were separated by polyacrylamide gel electrophoresis and visualized following staining with silver nitrate as described in the text . The numbers above the lanes indicate the approximate number of plaque forming units of virus added to the PCR reaction . The lanes denoted M and N indicate phi X 174 markers and negative controls respectively .

J . Pang et al.

254 UND

-2

-3

-4

-5

-6

-7

-8

M

-9

N

Fig. 2 . Polyacrylamide gel electrophoresis depicting the efficiency of uracil N-glycosylase for the hydrolysis of UTP-containing PCR products . CVB3, 106 pfu, was amplified using 400 tM UTP PCR. UTP-containing product from this reaction was serially diluted 10-fold and incubated with (bottom) or without (top) 1 unit/100 pl UNG for 10 min at room

temperature prior to the performance of a second PCR reaction . The numbers at the top of the gel indicate the logo dilution of the amplicons prior to the second amplification . The lanes denoted M and N indicate phi X 174 markers and negative controls respectively .

Effect of UNG on the sensitivity of the PCR reaction

ance of PCR reactions ." We found that dUTP can be substituted for TTP in the performance of PCR reactions which utilize Taq polymerase . We also found

We then examined the effect of UNG on the sensitivity of the PCR reactions for the detection of CA3 and CB3 . RNA from these viruses was amplified using dUTP and identical reaction conditions with the exception that UNG was added at a concentration of 10 units ml - ' to one set of reactions and omitted from the other reaction set . As depicted in Fig. 3, the PCR performed without UNG resulted in the detection of as few as 2 pfu of the CA3 virus . On the other hand, when UNG was added to the reaction, 10-fold more virus (20 pfu) was required to achieve visualization of the virus-specific band . A similar reduction in sensitivity was noted in the assay for the detection of coxsackievirus B3, especially when the efficiency of PCR amplification performed with dUTP and UNG was compared to that performed with TTP in the absence of added glycosidase (Fig . 4) .

Discussion Our experimental data confirm and expand previous studies on the use of dUTP and UNG in the perform-

that UNG can cleave amplicons which contain dUTP with a high degree of efficiency, resulting in a greater than 10,000,000-fold decrease in potentially contaminating amplicons . However, we also found that there are some limitations in the use of dUTP and UNG for PCR based amplifications . Firstly, we found that the concentration of reactants may need to be modified in order to optimize the use of dUTP as a substrate for the reactions . Specifically, we found that higher concentrations of MgCl, and dUTP resulted in increased amplification, a fact which is probably related to the fact that dUTP is a somewhat less efficient substrate for the Taq polymerase reactions . More importantly

we found that, even under optimal conditions, the addition of UNG to the PCR reaction resulted in a 0 . 51 log 1 0 fold reduction in sensitivity for the detection

of virion RNA . This decrease was almost certainly due to the fact that the UNG has some degree of reactivity for unmodified nucleotides . Reaction with UNG can thus result in the cleavage of some of the target cDNA and a resulting decrease in sensitivity . Fortunately, due to the high degree of sensitivity inherent in the PCR reactions, it is possible that some

Control of contamination in RNA amplification

255

-UNG 10 2

N

10 2 .5

10 3

+UNG 10 °

10 4

M

10 3

10 2 ' 5

10 2

Fig. 3 . Polyacrylamide gel comparing the sensitivity of PCR using UTP for the detection of coxsackievirus B3 . A stock of coxsackievirus B3 was prepared and diluted as described in the text . Aliquots were amplified by PCR in the presence (+UNG, right) and the other absence (-UNG, left) of added UNG enzyme, utilizing dUTP along with the other deoxynucleotides . The products were separated by polyacrylamide gel electrophoresis and visualized following staining with silver nitrate as described in the text . The numbers above the lanes indicate the approximate number of plaque forming units of virus added to the PCR reaction . The lanes denoted M and N indicate phi X 174 markers and negative controls respectively .

M

20

DUTP/+UNG 2

0.2

0-02

M

20

TTP/-UNG 2 N 0. 2

0 .02

Fig. 4. Polyacrylamide gel comparing the sensitivity of PCR using UTP for the detection of coxsackievirus A3 . A stock of coxsackievirus A3 was prepared and diluted as described in the text . Aliquots were amplified by PCR in the presence (+UNG, left) and the other absence (- UNG, right) of added UNG enzyme . The reaction without UNG was also performed using TTP in place of dUTP . The products were separated by polyacrylamide gel electrophoresis and visualized following staining with silver nitrate as described in the text . The numbers above the lanes indicate the approximate number of plaque forming units of virus tested in the indicated dilutions . The lanes denoted M and N indicate phi X 174 markers and negative controls respectively .

256

Ji. Pang et al.

PCR systems will have sufficient efficiency for microbial detection even if the use of UNG results in decreased reaction sensitivity . However, it should be noted in this regard that the sensitivity of PCR systems for the detection of RNA templates is often somewhat less favourable than that for the detection of DNA . This decreased sensitivity is due to the fact that the conversion of RNA to cDNA by reverse transcriptase serves as a limiting step for further amplification; i .e. RNA that is not converted to DNA cannot be amplified by the cyclical reactions using Taq polymerase . For example, we have found that PCR for the detection of RNA directly isolated from rotavirus virions is > 10-fold less sensitive than the detection of homologous cDNA cloned into a bacterial plasmid . In the current study, the PCR using UNG required 1000 pfu of CB3 virus for detection, making it substantially less sensitive than tissue culture for the detection of that virus . The development of more efficient methods for the amplification of RNA would improve the overall sensitivity of PCR-based assays for the detection of viruses with RNA genomes . Finally, although the use of UNG resulted in a substantial decrease in the amount of extraneous amplicons which are re-amplified by subsequent PCR reactions, it did not result in a total elimination of amplicon regeneration . It is thus possible that high degrees of contamination will result in the generation of a false-positive signal, even in the presence of UNG. It should also be noted that a system for the specific removal of amplicons would not be expected to eliminate contamination with native DNA or RNA from true-positive samples . Care must thus still be employed in the handling of samples in the laboratory environment . Despite these limitations, it is likely that the use of the dUTP-UNG and similar systems will be useful for the performance of PCR reactions in some clinical diagnostic settings . Such settings would include those in which the disadvantages of the lower sensitivity and higher cost of the reagents will be counterbalanced by the decrease in the occurrence of falsepositive reactions due to amplicon contamination . Clinical studies are required to determine the sensitivity, specificity, and cost-effectiveness of the dUTPUNG systems for the detection of viruses in human body fluid samples . Additional efforts should be directed at the development of efficient techniques for the prevention of specimen contamination . Such techniques, as well as others directed at increasing the practicality of PCR and other nucleic acid amplification systems should hasten the application of these

techniques to the diagnosis and eventual control of human diseases .

ACKNOWLEDGEMENT This work was supported by NIH Contract 1 UO1 Al 30420.

REFERENCES 1 . Yolken, R . H . (1988). Nucleic acids or immunoglobulins : which are the molecular probes of the future? Molecular and Cellular Probes 2, 87-96 . 2 . Coutlee, F ., Viscidi, R. P ., Saint-Antoine, P ., Kessous, A . & Yolken, R . H . (1991) . The polymerase chain reaction : a tool for the understanding and diagnosis of HIV-1 infection at the molecular level . Molecular and Cellular Probes 5, 241-59 . 3 . Ou, C.-Y ., Kwok, S ., Mitchell, S . W ., Mack, D . H . et al. (1988) . DNA amplification for direct detection of HIV 1 in DNA of peripheral blood mononuclear cells . Science 239, 295-7 . 4 . Erlich, H . A., Gelfand, D . H . & Saiki, R. K . (1988) . Specific DNA amplification . Nature 331, 461-2 . 5 . Saiki, R. K ., Gelfand, D . H ., Stoffel, S ., Scharf, S . J . et al . (1988) . Primer-directed enzymatic amplification of DNA with thermostable DNA polymerase . Science 239, 48791 . 6 . Wilde, J ., Yolken, R ., Willoughby, R. & Eiden, J . (1991) .

7. 8.

9.

10.

11 . 12 .

Improved detection of rotavirus shedding by polymerase chain reaction . Lancet i, 323-6. Kwok, S . & Higuchi, R . (1989). Avoiding false positives with PCR. Nature 339, 237-8 . Shuldiner, A . R ., Nirula, A . & Roth, J . (1990) . RNA template-specific polymerase chain reaction (RS-PCR) : a novel strategy to reduce dramatically false positives . Gene 91 139-42 . Deragon, J. M ., Sinnett D ., Mitchell, G ., Potier, M . & Labuda, D . (1990). Use of gamma irradiation to eliminate DNA contamination for PCR . Nucleic Acids Research 18 6149. Porter-Jordan, K . & Garrett, C . T . (1990). Source of contamination in polymerase chain reaction assay . Lancet i, 1220 . Sarkar, G . & Sommer, S . (1990). More light on PCR contamination . Nature 347, 340-1 . Cimino, C . D ., Metchette, K . C ., Tessman, J . W ., Hearst, J . E . & Isaacs, S . T . (1991). Post-PCR sterilization : a method to control carryover contamination for the polymerase chain reaction . Nucleic Acids Research 19, 99-107 .

13 . Longo, M . C ., Berninger, M . S . & Hartley, J . L . (1990) . Use of uracil DNA glycosylase to control carry-over contamination in polymerase chain reactions . Gene 93 125-8 . 14 . Rotbart, H . A . (1991) . Nucleic acid dectection systems for enteroviruses . Clinical and Microbiological Reviews 4, 156-68 .