WO1991015601A1

WO1991015601A1 - Modified rna template-specific polymerase chain reaction

Info

Publication number: WO1991015601A1
Application number: PCT/US1991/002211
Authority: WO
Inventors: Alan R. Shuldiner; Jesse Roth
Original assignee: The United States Of America, As Represented By The Secretary, U.S. Department Of Commerce
Priority date: 1990-04-05
Filing date: 1991-04-04
Publication date: 1991-10-17
Also published as: AU7653691A

Abstract

The present invention relates to methods of detecting an RNA sequence by tagging the sequence with a unique random nucleotide sequence during reverse transcription. The unique nucleotide sequence is then utilized to selectively amplify the resulting DNA sequence. The present invention reduces the number of false positives obtained as a result of contaminating DNA.

Description

MODIFIED RNA TEMPLATE-SPECIFIC POLYMERASE CHAIN REACTION The present application is a continuation-in- part application of Serial No. 07/504,591 filed April 5, 1990, which is hereby incorporated in its entirety.

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to a method for detecting an RNA sequence. More specifically, the present invention relates to a method of amplifying an RNA sequence using a modification of the polymerase chain reaction.

2. Background Information

The polymerase chain reaction (PCR) method, developed by Perkin-Elmer-Cetus Corporation, is a popular and extraordinarily powerful technique for the amplification of DNA sequences. It has wide-ranging applications including molecular biology, medical diagnostics, genetics, forensics, and archeology [Saiki et al., Science 230, 1350 (1985); Saiki et al., Science 239, 487 (1985); Kogan et al., N. Engl. J. Med. 316, 656 (1987); Higuchi et al., Nature 332, 543 (1988); and Paabo et al., J. Biol. Chem. 264, 9709 (1989)]. When coupled with reverse transcription (RT-PCR) , this technique can detect as few as 1 to 100 copies of a specific RNA from single cells or small numbers of cells. [Kawasaki et al., Proc. Natl. Acad. Sci. U.S.A. 85, 5698 (1988); Rappolee et al., Science 241, 708 (1988); Rappolee et al., Science 241, 1823 (1988); Sarkar et al., Science 244, 331 (1988); and Froh an et al., Proc. Natl. Acad. Sci. U.S.A. 85, 8998 (1988)]. Unfortunately, the exquisite sensitivity of this technique presents one of its severe shortcomings, false positives resulting from contamination with minute quantities of DNA [Kwok et al.. Nature 339, 237 (1989); Sakar et al.. Nature 343, 27 (1989)].

Potential sources of contaminating DNA may include:

1) endogenous sources such as small quantities of genomic DNA which may copurify with RNA, and

2) exogenous sources such as cDNA, plasmid DNA, or DNA fragments amplified in previous PCRs (i.e. carryover) .

While the frequency of false positives can be reduced somewhat by instituting and maintaining special techniques (e.g. premixing and aliquoting reagents; use of disposable gloves and positive displacement pipettes; and adding the experimental sample last) , contamination still remains a major problem, especially when low abundance RNA transcripts are being sought [Kwok et al., Nature 339, 237 (1989); Lo, Y.-M., et al.. Lancet 2, 699 (1988)]. Conventional RT-PCR amplifies equally well DNA derived from an RNA template or from a DNA template. Therefore, small quantities of contaminating DNA from virtually any source may easily result in false positives. Assuming approximately 4 pg of genomic DNA per mammalian haploid cell, and a sensitivity of 1 to 100 copies, conventional RT-PCR would result in false positives from only picogram quantities of contaminating genomic DNA.

It is possible to avoid false positives caused by amplification of genomic DNA which may copurify with RNA if the target sequence to be amplified by RT-PCR spans an intron. However, this experimental design is not always possible since 1) some genes do not contain introns in convenient regions, and 2) the genomic structure of many target genes are not yet known.

In the laboratory of the present inventors, RT-PCR was recently used to detect small quantities of Xenopus insulin mRNAs in unfertilized eggs and early embryos. Despite the fact that numerous precautions were taken to exclude contamination of Xenopus insulin cDNAs which had been previously cloned in our laboratory [Shuldiner et al., J. Biol. Chem. 264,9428 (1989)], frequent false positives precluded meaningful ■ interpretation of the experiments.

The present invention, RNA template-specific PCR (RS-PCR) and modified RS-PCR, provides methods of detecting minute quantities of RNA without the problems of false positives associated with RT-PCR. In the present methods the reduction in the frequency of false positives is achieved without sacrificing sensitivity obtained with conventional RT-PCR. SUMMARY OF THE INVENTION

Accordingly, it is an object of the present invention to provide a method of detecting an RNA sequence which reduces the number of false positives resulting from DNA contamination in the sample (i.e., previously cloned cDNAs, genomic DNA or carryover of DNA amplified in previous PCRs) . The present method increases the accuracy of the procedure without sacrificing sensitivity.

It is another object of the present invention to provide a method of detecting an RNA sequence which obviates the necessity to choose a target RNA sequence which spans an intron. Various other objects and advantages of the present invention will be apparent from the following description of the invention and the drawings.

In one embodiment, the present invention relates to a method of detecting an RNA sequence. The method comprises: i) reverse transcribing the RNA sequence from an oligonucleotide primer hybridized thereto which oligonucleotide primer (d₁₇-t₃₀) comprises: a) on the 3¹ end thereof (segment d₁₇) , a nucleotide sequence complementary to a region of the RNA sequence to be detected; and b) on the 5* end thereof (segment t₃₀) , a unique random nucleotide sequence or tag whereby a single stranded DNA sequence is produced which has at its 5' end the unique sequence; ii) hybridizing, at a temperature high enough to preclude annealing of the d₁₇ segment of the dι₇-t₃₀ primer to possible contaminating DNA, but low enough to allow annealing, an upstream oligonucleotide primer (U₃₀) , to a region of said DNA sequence to which it is complementary, a predetermined distance upstream from soi iii) extending the primer (u₃₀) so that a DNA molecule is produced having at its 3* end a nucleotide sequence complementary to said unique nucleotide sequence (segment t₃₀) ; iv) denaturing the double-stranded DNA molecule produced in step (iii) ; v) hybridizing, at a temperature high enough to preclude annealing of the d₁₇ segment of the d₁₇-t₃₀ primer to possible contaminating DNA, but low enough to allow annealing the upstream PCR oligonucleotide primer (u₃₀) to the region of said DNA sequence to which it is complementary and, hybridizing, at a temperature high enough to preclude annealing of the d₁₇ segment of the d₁₇-t₃₀ primer to possible contaminating DNA, but low enough to allow annealing of a PCR oligonucleotide primer (t₃₀) comprising all or a portion of said unique nucleotide sequence, to the 3¹ end of said DNA sequence to which it is complementary; vi) extending the primers (u₃₀) and (t₃₀) thereby producing two DNA molecules; and vii) detecting the presence or absence of the amplified DNA sequence; wherein the d₁₇ segment of the oligonucleotide primer d₁₇-t₃₀ does not hybridize to contaminating DNA at the annealing temperature of the PCR, and oligonucleotide primer u₃₀ and oligonucleotide primer t₃₀ do hybridize to their appropriate DNA templates at the annealing temperature of the PCR.

BRIEF DESCRIPTION OF THE DRAWINGS FIGURE 1 shows diagrammatically the RNA template-specific PCR method (RS-PCR) .

FIGURE 2 compares the sensitivity of conventional reverse transcriptase - PCR (RT-PCR) and novel RS-PCR when beginning with an RNA template.

FIGURE 3 compares conventional RT-PCR and novel RS-PCR when DNA rather than RNA is used as a starting template to mimic DNA contamination. FIGURE 4 shows the effect of changing the nucleotide sequence of the unique segment of oligonucleotide primer d₂o^_ ₂i• FIGURE 5 shows schematically the modified RNA template-specific PCR.

FIGURE 6 compares conventional RT-PCR and modified RS-PCR. RT primer d_7t30: PCR primers t₃₀ and u₃₀ (lanes 1-4) ; RT primer d₃₀t: PCR primers d₃₀ and u₃₀ (lanes 6-9) . Lane 0 is a Haelll digest of PhiX174 DNA, while lane 5 is RS-PCR in the absence of any template. FIGURE 7 shows PCR carryover contamination is ignored with modified RS-PCR. Lanes 1 and 3; RT primer d₁₇-t₃₀, PCR primers u₃₀ and t₃₀. Lanes 2 and 4; RT primer d₁₆t'₃₀, PCR primers u₃₀ and t'₃₀.

FIGURE 8 shows the region of Xenopus insulin RNA that was used as the target RNA to test the modified RS-PCR procedure. Reverse transcription primer d₁₇-t₃o contained a 17 base sequence at its 3' end (segment d₁₇) that was complementary (antisense) to a region of Xenopus insulin RNA in the 3' untranslated region (nucleotides 404-420), and 30 bases at its 5¹ end (segment t₃₀) that were unique in sequence. Upstream primer u₃₀ is identical (sense) to Xenopus insulin RNA in the coding region (nucleotides 59-88) . DETAILED DESCRIPTION OF THE INVENTION The present invention, RNA template-specific PCR (RS-PCR) and modified RS-PCR, relates to a targeted amplification method which distinguishes RNA in the sample from contaminating DNA and amplifies only sequences derived from RNA. Minute quantities of cDNA, plasmid DNA or carryover DNA amplified in previous PCRs can be important sources of contamination when using conventional RT-PCR. The present invention reduces the number of false positives obtained as a result of contaminating DNA. Furthermore, the present invention obviates the necessity of choosing a target RNA sequence which spans an intron in order to distinguish the reverse transcribed DNA from contaminating genomic DNA. In addition, the modified RS-PCR eliminates the need for removal of the primer after reverse transcription, such as by ultrafiltration.

The RS-PCR method of the present invention is shown schematically in Figure 1.

In the first step, a first oligonucleotide primer designated d₂₀-t₂₁ in Figure 1 (advantageously, of about 41 nucleotides) is hybridized to the RNA sequence to be detected. Primer d₂o-t₂ι comprises on the 3• end, a nucleotide sequence (advantageously, about 20 nucleotides) complementary to the 3¹ end of the RNA sequence whose presence is to be detected (segment d₂o) and on the 5' end, a unique random nucleotide sequence or tag (advantageously, about 21 nucleotides) (segment t₂ ) . While the 3¹ end of the primer hybridizes to the RNA sequence, the 5' end of the primer remains unhybridized as no complementary sequence exists within the sample.

Once primer d₂₀-t₂ has been hybridized to the 3¹ end of the RNA sequence, reverse transcriptase is used to extend the primer. The resulting single -(-) stranded DNA segment is thus tagged at its 5' end with the unique sequence t₂₁ of original primer d₂₀-t₂₁. This unique 5' sequence (t₂₁) distinguishes between DNA generated from the RNA-template and possible contaminating DNA. It is preferable for the unique sequence to be composed of approximately equal amounts of each nucleotide (i.e. about 25% of each nucleotide) . Furthermore, it is preferable to choose a unique sequence which is unlikely to have significant secondary structure, and does not contain significant complementarity at its 3• end with the 3' end of the upstream primer (for example, primer u₂ι) . The sequence can also, be selected so as to contain a convenient restriction enzyme recognition site if desired. One skilled in the art can easily generate by computer appropriate sequences, 5•-GACAAGCTTCAGGTAATCGAT-3' and 5*-CCGAATTCTGTAGTCCGTCA-3* being two examples.

Prior to amplification, excess primer d₂₀-t₂₁ is removed by ultrafiltration through a Centricon 100 device (Amicon, Danvers, MA) or similar device.

In the second and third steps of the present method, the DNA segment resulting from the previous step is amplified using the PCR technique (see U.S. Patents 4,683,202 and 4,683,195). Two oligonucleotide primers designated u₂₁ and t₂₁ in Figure 1 are utilized to amplify the DNA. Upstream oligonucleotide primer u₂₁ (advantageously, about 21 nucleotides) comprises a nucleotide sequence complementary the single stranded DNA segment produced in Step 1, a predetermined distance upstream from primer d₂₀t₂₁. Oligonucleotide primer t₂ι (advantageously, about 21 nucleotides) comprises the unique nucleotide sequence with which the segment of DNA was tagged during reverse transcription. The two primers are added to the sample and the PCR is carried out.

In the second step of the present method (PCR cycle 1), primer u₂₁, which is complementary to a region of the single stranded DNA segment produced in Step 1, a predetermined distance upstream from primer d₂₀t₂₁, hybridizes thereto and is extended therefrom creating a complementary strand of DNA which includes at its 3• end a sequence complementary to the unique sequence. Primer t₂₁ is not utilized in the first PCR cycle since no complementary sequence is present in the sample. However, primer t₂ι is used in the second PCR cycle and all cycles thereafter.

In the third step of the present method, the double stranded DNA segment resulting from the first PCR cycle, is denatured prior to the second PCR cycle. For the second cycle and all subsequent PCR •cycles, primer u₂ι which is complementary to the 3¹ end of the single - (-) stranded DNA hybridizes thereto and is extended therefrom. At the same time, the primer t₂₁ hybridizes to its complementary sequence at the 3 ' end of the single - (+) strand of DNA and is extended therefrom. All DNA synthesis occurs in the 5¹ to 3' direction.

The modified RS-PCR method of the present invention is shown schematically in Figure 5.

The modified RS-PCR method eliminates the need to remove the first oligonucleotide primer, designated d₇-t₃₀ in Figure 5, by selecting oligonucleotide primers d₁₇-t₃₀, t₃₀ and d₃₀, so that differential hydridization occurs under the PCR conditions. The primers are selected so that the d₇-t₃₀ primer and the d₃₀ and the t₃₀ primers anneal under different temperatures.

In the first step, as with the RS-PCR method, a first oligonucleotide primer designated d₁₇-t₃₀ in Figure 5 (advantageously, of about 47 nucleotides) is hybridized to the RNA sequence to be detected. Primer d₁₇-t₃₀ comprises on the 3' end, a nucleotide sequence (advantageously, about 17 nucleotides) complementary to the 3' end of the RNA sequence whose presence is to be detected (segment dι₇) , and on the 5* end, a unique random nucleotide sequence or tag (advantageously, about 30 nucleotides) (segment t₃₀) .

The primer should be selected so that the length of the d segment is such that it will not anneal efficiently to any DNA contaminants at the elevated annealing temperatures used in Steps 2 and 3. One skilled in the art can easily generate by computer suitable d₁₇-t₃o primers including, for example, 5¹- gaacatcgatgacaagcttaggtatcgatatgatggaattgccttga-3' and 5*-cttatacggatatcctggcaattcggacttgcatgatggaattgcc-3' . Once primer dι₇-t₃₀ has been hybridized to the RNA sequence, reverse transcriptase is used to extend the primer thereby creating a single -(-) stranded DNA segment which is tagged at its 5* end with the unique sequence, the t₃₀ segment, of original primer. This unique 5¹ sequence, as with the RS-PCR method, distinguishes between DNA generated from the RNA- template and possible contaminating DNA.

In the second step of the present method (PCR cycle 1) , oligonucleotide primer designated u₃₀ in Figure 5 (advantageously, about 30 nucleotides) hybridizes to the single stranded DNA generated in Step 1 a predetermined distance upstream from primer d₁₇-t₃₀, and is extended therefrom creating a complementary strand of DNA which includes at its 3* end a sequence complementary to the unique sequence. The primer u₃₀ comprises a nucleotide sequence complementary to the single stranded DNA segment produced in Step 1, a predetermined distance upstream from primer d₁₇t₃₀. The annealing stage of the PCR cycle is carried out at a temperature high enough to preclude annealing of the d₁₇ segment of the reverse transcription primer d₇-t₃₀ a to contaminating DNA, but low enough to allow annealing of PCR primer u₃₀, for example temperatures of 42" C or greater.

In the third step of the present method, the double stranded DNA segment resulting from the first PCR cycle, is denatured prior to the second PCR cycle. For the second cycle and all subsequent cycles, primer u₃o which is complementary to the single - (-) stranded DNA hybridizes thereto and is extended therefrom. At the same time, a second oligonucleotide primer, designated t₃₀ in Figure 5 (advantageously, about 30 nucleotides) is added to the sample. The primer t₃₀ comprises the unique nucleotide sequence with which the segment of DNA was tagged during reverse transcription. When the primer is added to the sample it hybridizes to its complementary sequence at the 3¹ end of the single - (+) strand of DNA and is extended therefrom. The annealing stage of all PCR cycles is conducted at a temperature high enough to preclude annealing of the d₁₇ segment of the reverse transcription primer d₁₇-t₃₀ a to contaminating DNA, but low enough to allow annealing of PCR primers u₃o and t₃₀.

With RS-PCR, sequences derived from RNA that are tagged with the unique sequence (t₃₀) during reverse transcription (step 1) are amplified preferentially during PCR (steps 2 and 3) . The original RS-PCR method requires ultrafiltration after reverse transcription to remove excess RT primer [Shuldiner et al.. Gene 91, 139 (1990) ] . The modified RS-PCR cirumvents this step by increasing the length of the RT and PCR primers, and increasing the PCR annealing temperature. The primers are selected so that the RT primer, d₁₇-t₃o, hybridizes to the RNA template under the reverse transcription conditions but does not hybridize to possible DNA contaminants under the PCR conditions.

The longer length of the u₃₀ and t₃₀ primers allows annealing to occur at an increased temperature, that is temperatures up to about 72" C. Annealing of the 17 base d₇ segment of the RT primer d₁₇-t₃₀ occurs efficiently during reverse transcription at 37°C, but not at the higher PCR annealing temperature. Thus, when Steps 2 and 3 are carried out at a temperature of 42" C or greater (preferably 65° or greater), remaining d₁₇-t₃₀ primer does not anneal to possible DNA contaminants while the u₃₀ and t₃₀ primers will anneal and be extended.

In both RS-PCR methods described above, each cycle of PCR involves primer hybridization, extension to yield double stranded DNA and denaturation. After the first PCR cycle, both the (+) and (-) strands of the DNA serve as templates from which a new strand of DNA is created. This leads to logarithmic expansion of the tagged segment of DNA.

Contaminating DNA lacks the unique nucleotide sequence. Thus, during the PCR the 3* end of the single - (-) strand of DNA serves as a template for primer u₂₁ or u₃₀ (but the 3' end of the -(+) strand can not act as template for unique primer t₂ι or t₃o since there is no complementary sequence) . This allows only linear amplification which, as one skilled in the art knows, does not produce enough DNA to result in a false positive when detecting the presence of the logarithmically amplified PCR product.

Potential contamination arising from carryover of PCR products from previous experiments in which a different unique sequence was used is virtually eliminated when the present invention is used. With the methods of the present invention, no false positives were observed in over 20 independent experiments. The criteria for selecting the unique sequence of the primer used for reverse transcription and subsequent PCR is that the sequence selected is not present in the sample i.e. is unique. Therefore, the sequence used can be changed periodically. Changing the unique sequence prevents amplification of carryover PCR products. Thus, the methods of the present invention are particularly useful in a clinical laboratory setting where many samples and automation make careful laboratory hygiene more difficult. The present invention is as sensitive as the well known PCR and RT-PCR procedures. Therefore, the small quantity of RNA needed is not affected. However, the present invention has the advantage of being more accurate. In the following non-limiting examples, a segment of Xenopus insulin RNA is amplified by the present methods. The methods are applicable to the amplification of other RNAs.

Examples RS-PCR

Example 1: RNA Template-Specific Polymerase Chain Reaction (RS-PCR) Xenopus insulin mRNA was amplified using novel RS-PCR, which involves first reverse transcribing Xenopus pancreatic RNA using an oligonucleotide 41-mer as a primer (oligonucleotide d₂o-t₂ι) whose nucleotide sequence contained 20-bases at the 3*-end which were complementary to a region of Xenopus insulin mRNA (segment d₂₀) , and 21-bases at the 5*-end which consisted of a unique random sequence selected by computer or similar method (segment t₂₁) (FIGURE 1) followed by PCR amplification of the DNA segment. As a first step, total RNA from Xenopus pancreatic tissue was prepared by the guanidinium isothiocyanate method [Chirgwin et al.. Biochemistry 18, 5294 (1979)]. RNA was reverse transcribed at 42"C for one hour in a 25 μl reaction mixture containing Tris-HCl (50 mM, pH adjusted to 8.7 at room temperature) , NaCl (100 mM) , MgCl₂ (6 mM) , dithiothreitol (10 mM) , dNTP's (1 mM each), RNasin (1 μl; Promega Biotec; Madison, WI) , oligonucleotide d₂₀- t₂₁ (5•-GACAAGCTTCAGGTATCGATTTGCATGATGGAATTGCCTTG-3^» ; 0.5 μM) , and AMV-reverse transcriptase (10 units; Promega Biotec) .

This reverse transcription step resulted in single-(-) stranded DNA which had been "tagged" at its 5¹ end with a unique 21-nucleotide sequence or tag (segment t₂₁) .

After reverse transcription, the oligonucleotide primer d₂₀-t₂₁ was efficiently removed (>99.9%) using a Centricon 100 ultrafiltration device (Amicon; Danvers, MA) according to manufacturer's recommendations. Then PCR was performed using as primers oligonucleotide t₂₁, a 21-mer containing the same unique nucleotide sequence as in segment t₂₁ of oligonucleotide d₂₀-t₂ι and oligonucleotide u₂₁, a 21-mer complementary to the first strand, 244 bp upstream from oligonucleotide ₂ι« PCR amplification was performed in a 50 μl reaction volume containing Tris-HCl (10 mM, pH adjusted to 8.3 at room temperature), KC1 (50 mM) , MgCl₂ (1.5 mM), gelatin (0.01%), dNTP's (200 μM each), oligonucleotide t₂₁ (5•-GACAAGCTTCAGGTAATCGAT-3• ; 0.5, μM) , oligonucleotide u₂₁ (5^«-GAGGCTTCTTCTACTACCCTA-3• ; 0.5 μm) and Taq polymerase (1 units; Perkin Elmer-Cetus Corp., Emeryville, CA) . The reaction mixture was covered with paraffin oil (approximately 50 μl) , heated to 94"C for 5 minutes, followed by PCR (45-60 cycles). Each cycle consisted of annealing (55'C, 1.5 min) , extension (72"C, 1.5 min) and denaturation (94"C, 1 min) except for the last cycle, in which the extension time was increased to 15 minutes to insure completeness of extension. Twenty microliters of the reaction mixtures were loaded onto a composite gel consisting of 1% agarose and 2% Nusieve GTG (FMC Bioproducts; Rockland, ME) in Tris-borate-EDTA buffer, electrophoresed, stained with ethidium bromide, and visualized by UV transillumination.

Since logarithmic amplification is dependent upon nucleotide sequences corresponding to d₂₀, t₂ and u₂₁, only sequences derived from Xenopus insulin RNA which had been reverse transcribed with oligonucleotide d₂₀-t₂₁ were amplified logarithmically, and contaminating DNA, which lacks the oligonucleotide t₂₁ sequence was not amplified logarithmically. Example 2: Comparison of novel RS-PCR and conventional RT-PCR using an RNA template. To test whether novel RS-PCR was as sensitive as conventional RT-PCR, Xenopus pancreatic RNA which had been reverse transcribed with oligonucleotide d₂₀- t₂₁ and ultrafiltered, was subjected to either conventional RT-PCR (oligonucleotides d₂₀ and u₂₁) , or novel RS-PCR (oligonucleotides t₂₁ and u₂₁) . PCR with either of these two oligonucleotide pairs resulted in similar sensitivity (FIGURE 2) .

Xenopus pancreatic RNA (1 ng) was reverse transcribed and ultrafiltered according to the methods of Example 1. For the conventional RT-PCR, 60 cycles of the

PCR were performed on serial ten-fold dilutions of reverse transcribed and ultrafiltered pancreatic RNA with oligonucleotide primers d₂₀ and u₂ (FIGURE 2, lanes 1-5) using the conditions described in Example 1. For the novel RS-PCR comparison, identical serial dilutions of the reverse transcribed and ultrafiltered pancreatic RNA was amplified by PCR using oligonucleotide primers t₂₁ and u₂ι (FIGURE 2, lanes 6-

9). The predicted 244-bp and 265-bp amplified bands observed on the ethidium bromide-stained gel hybridized strongly to a radiolabeled full-length Xenopus insulin cDNA probe [Southern J. Mol. Biol. 98, 503 (1975)]. PCR with either the oligonucleotide pair d₂₀ and u₂ or the pair t₂₁ and u₂ι resulted in similar sensitivity. Conventional RT-PCR with or without removal of excess oligonucleotide d₂₀-t₂₁ by Centricon 100 ultrafiltration resulted in similar sensitivity, as did reverse transcription with oligonucleotide d₂₀ as the primer rather than oligonucleotide d₂₀-t₂₁. These results suggest that neither Centricon 100 ultrafiltration or reverse transcription using an oligonucleotide with a random 21-nucleotide overhang at its 5• end result in a significant decrease in sensitivity. Example 3: Comparison of novel RS-PCR and conventional RT-PCR using a DNA template By contrast to the Example 2 where the sensitivity of the reaction was not affected by the use of the unique nucleotide sequence, novel RS-PCR was approximately 10 to 1000-fold less affected by the presence of DNA contaminants (i.e., Xenopus insulin cDNA) than conventional RT-PCR even after 60 cycles (FIGURE 3) .

Full-length Xenopus insulin cDNA (300 pg) was "reverse transcribed" with oligonucleotide d₂₀-t₂₁, excess oligonucleotide d₂₀-t₂ι removed by ultrafiltration, and PCR (60 cycles) was accomplished as described in the above Examples. Results of the conventional RT-PCR performed on serial ten-fold dilutions of the "reverse transcribed" and ultrafiltered Xenopus insulin cDNA using oligonucleotides d₂₀ and u₂₁ is shown in FIGURE 3, lanes 1-5. Novel RS-PCR of identical serial ten-fold dilutions of "reverse transcribed" and ultrafiltered Xenopus insulin cDNA using oligonucleotides t₂₁ and u₂₁ is shown in lanes 6-10 of the same figure. In theory, with the RS-PCR method, only RNA that had been primed with oligonucleotide d₂o-t₂₁ during RT should have been amplified during PCR. However, it was found that when relatively large quantities of DNA template (>10 pg or approximately 1 x 10⁷ molecules) were used, detectable amplification was observed (lane 6 in Fig. 3) . It has been determined from separate experiments that this phenomenon was caused by two mechanisms; i) at relatively high DNA concentrations, RT acted as a DNA polymerase and incorporated oligonucleotide d₂₀-t₂ι into the so-called first strand, and ii) the minute quantities of oligonucleotide d₂₀-t₂₁ that remained behind after ultrafiltration incorporated into DNA during early PCR cycles which could then be amplified efficiently in RS-PCR.

Example 4: Effect of changing the sequence of the unique segment t_2. of oligonucleotide d₂₀- t₂₁ on conventional RT-PCR and novel RS- PCR. In order to evaluate the ability of the RS-PCR method to eliminate problems of carryover contamination of amplified DNA from previous RS-PCR experiments which had been tagged with a different unique sequence t₂₁ the following experiment was conducted. Xenopus pancreatic RNA (1 ng) was reversed transcribed with either oligonucleotide 41-mer d₂₀-t₂₁ (FIGURE 4, lanes 1, 2, 3 and 7), or oligonucleotide 41- mer d₂₀-t'₂₁ (5^»- CCGAATTCTGTAGTCCGTCATTGCAGATGGAATTGCCTTG-3•) (FIGURE 4, lanes 4-6) . After ultrafiltration, PCR (45 cycles) was accomplished as described in the previous Examples using oligonucleotide pairs t₂₁ and u₂₁ (FIGURE 4, lanes 1 and 4), t'₂₁ (5'-CCGAATTCTGTAGTCCGTCA-3') and u₂₁ (FIGURE 4, lanes 2 and 5), d₂₀ and u₂₁ (FIGURE 4, lanes 3 and 6), or t₂₁ and u'₂₁ (5'-TGACCTTTCCAGCACTTATC-3') (FIGURE 4, lane 7) . As expected, the RNA that had been reversed transcribed with oligonucleotide d₂₀-t₂₁ was amplified only when oligonucleotide t₂₁ was used during PCR, but not when an unrelated unique 21-mer (oligonucleotide t'₂₁) was used. Conversely, reverse transcription of Xenopus pancreatic RNA with oligonucleotide d₂₀-t'₂₁, could only be amplified by the corresponding unique 21- mer, oligonucleotide t'₂₁, and not by the unrelated random 21-mer, oligonucleotide t₂₁. As expected, when conventional RT-PCR was used (i.e. oligonucleotide d₂₀) , amplification occurred regardless of whether reverse transcription primers d₂₀-t₂₁ or d₂₀-t'₂₁ were used.

MODIFIED RS-PCR Example 5: Modified RS-PCR

Oligonucleotides were synthesized on a Coder 300 automated DNA synthesizer (E.I. Du Pont Company;

Wilmington, DE) , and purified with NENsorb Prep columns (New England Nuclear; Boston, MA) according to the manufacturer's directions (see Table 1 below). Xenopus insulin (sense) RNA was prepared by ligating an 890 bp Xenopus insulin cDNA [Shuldiner et al., J. Biol. Chem. 264, 9428 (1989)] into pSP71 (Promega Biotec; Madison, WI) .

After linearization of the recombinant plasmid with Bglll, T7 RNA polymerase (Promega Biotec) was used for in vitro transcription to generate Xenopus insulin (sense) RNA. The RNA was purified by oligo-dT cellulose chromatography (Bethesda Research Laboratories) . Only full-length RNA was retained by the column since the 3' end contained a long poly-A tail. The RNA was quantitated by UV absorbance at 260 nm. RNA was diluted to the appropriate concentration in water containing yeast tRNA (100 μg/ml) (Bethesda Research Laboratories) . DNA templates used to demonstrate RNA specificity were either a double-stranded 890 bp Xenopus insulin cDNA insert [Shuldiner et al., J. Biol. Chem. 264, 9428 (1989)], or a 377 bp Xenopus insulin RS-PCR product that had been subjected to ultrafiltration with a Millipore-MC-100 device (Millipore; Bedford, MA) to remove excess primers. DNA templates were quantitated by comparison to a known quantity of a Haelll digest of PhiX174 (Bethesda Research Laboratories) after agarose gel electrophoresis.

Table I. Primer sequences used to compare improved RS-PCR to conventional RT-

PCR.

5 Primer Sequence

^d17 ^t30 5'-GAACATCGATGACAAGCTTAGGTATCGATATGATGGAATTGCCTTGA-3' ^fc30 5'-GAACATCGATGACAAGCTTAGGTATCGATA-3' ^d16^t'30 5'-CTTATACGGATATCCTGGCAATTCGGACTTGCATGATGGAATTGCC-3'

10 t

30 5^»-CTTATACGGATATCCTGGCAATTCGGACTT-3' ^d30 5'-GCATGATGGAATTGCCTTGAAGGTGCCTTG-3 '

^U30 5'-ATGCAGTGTCTGCCCCTGGTTCTTGTCCTC-3 ^»

Reverse transcription of serial ten-fold dilutions of Xenopus insulin RNA (10⁷ to 10⁴ copies) was accomplished at 37° C in a final volume of 20 μl containing KC1 (50 mM) , Tris-HCl (10 mM; pH 8.3 at 25" C) , MgCl₂ (1.5 mM) , gelatin (0.01 mg/ml), dNTPs (200 μM each) , RNasin (40 U; Promega Biotec) , AMV- reverse transcriptase (7 U; Promega Biotec) , and primer.d₁₇-t₃₀ (0.5 μM) .

Primer d₁₇-t₃₀ (Table I) was a 47-mer whose sequence contained 17 bases at its 3•-end that were complementary to a region of Xenopus insulin mRNA, designated segment d₁₇, and 30 bases at its 5'-end that were unique in sequence, designated segment t₃₀. Thus, reverse transcription yields single-stranded DNA that contains a unique 30 base "tag" (segment t₃₀) at its 5^« end (FIGURE 5).

The second strand was synthesized during the first cycle of PCR in which 5 μl of the RT reaction mixture from step 1 was used directly in a final volume of 50 μl containing KC1 (50 mM) , Tris-HCl (10 mM; pH 8.3 at 25° C) , MgCl₂ (1.5 mM) , gelatin (0.01 mg/ml), dNTPs (200 μM each), upstream primer u₃₀ (0.5 μM) , downstream primer t₃₀ (0.5 μM) and Taq polymerase (1.5 U; Perkin Elmer-Cetus; Emeryville, CA) .

Upstream (sense) primer u₃₀ was a 30-mer corresponding to Xenopus insulin cDNA that was 347 bp upstream from the sequence corresponding to segment dι₇, while downstream primer t₃₀ was a 30-mer whose sequence was identical to segment t₃₀ of RT primer d₁₇-t₃₀ (see Table I) . With these primers, sequences derived from RNA that had been tagged with unique sequence (t₃₀) during reverse transcription were amplified logarithmically preferentially, while contaminating DNAs, lacking the unique tag, were not amplified logarithmically (FIGURE 5) [Shuldiner et al.. Gene 91, 139 (1990)].

After covering the PCR reaction mixture with parafin oil (approximately 50 μl) , 35 cycles of PCR were performed, each cycle consisting of denaturation (94° C, 1 mih) and annealing/extension (70° C, 2 min) . In the first cycle, the denaturation time was increased to 5 min, and in the last cycle, the annealing/extension time was increased to 10 min to ensure completeness of the extension.

Twenty microliters of the PCR reaction mixture was electrophoresed on a composite gel consisting of 1% agarose (Bethesda Research Laboratories) and 2% Nuseive GTG (FMC Bioproducts; Rockland, ME) . DNA was visualized by ethidium bromide staining and UV transillumination. Example 6: Comparison of modified RS-PCR and conventional RT-PCR To compare the sensitivity of modified RS- PCR to conventional RT-PCR, serial ten-fold dilutions of Xenopus insulin RNA (10⁷ to 10⁴ molecules) were amplified using either modified RS-PCR (RT primer d₁₇-t₃₀; PCR primers u₃₀ and t₃₀) (FIGURE 6, panel a, lanes 1-5) , or conventional RT-PCR (RT primer d₃₀; PCR primers u₃₀ and d₃₀) (FIGURE 6, panel a, lanes 6-

9). Modified RS-PCR was equally sensitive to conventional RT-PCR when beginning with an RNA template. By contrast, when Xenopus insulin double- stranded DNA (10⁷ copies) was used as starting template to mimic DNA contamination, conventional RT- PCR resulted, as expected, in a strong signal (FIGURE 6, panel b, lane 11) , while the modified RS-PCR method virtually ignored theDNA template (FIGURE 6, panel b, lane 10) . When larger amounts of DNA were used (i.e., > 10⁸ copies), a faint signal was detected with RS-PCR [Shuldiner et al.. Gene 91, 139 (1990)]. To mimic RS-PCR carryover contamination, RS-

PCR was performed with two 377 bp Xenopus insulin RS- PCR products (approximately 10⁸ copies) that were identical to each other except each contained a different unique tag (sequence t₃₀ or t'₃₀ (Table I and FIGURE 7) .

Double-stranded Xenopus insulin DNA containing either tag sequences t₃₀ (FIGURE 7, lanes 1 and 2), or t'₃₀ (FIGURE 7, lanes 3 and 4) were subjected to improved RS-PCR as described expect 30 cycles were performed. Amplification of each DNA template occurred efficiently when the primers matched the unique tag present in the PCR product (FIGURE 7 lanes 1 and 4) . However, when RS-PCR primers were used that did not match the unique tag present in the PCR product, no amplification occurred (FIGURE 7, lanes 2 and 3). Thus, carryover contamination of RS-PCR products in which one unique tag was used is virtually eliminated when RS-PCR is performed with a different unique tag. * * * * *

The entire contents of all publications cited hereinabove are hereby incorporated by reference.

While the foregoing invention has been described in some detail for purposes of clarity and understanding, it will be clear to one skilled in the art from a reading of this disclosure that various changes in form and detail can be made without departing from the true scope of the invention.

Claims

WHAT IS CLAIMED IS:

1. A method of detecting an RNA sequence comprising the steps of: i) reverse transcribing the RNA sequence from an oligonucleotide primer hybridized thereto which oligonucleotide primer (d₂₀-t₂₁) comprises: a) on the 3• end thereof, a nucleotide sequence complementary to the 3• end of the RNA sequence (segment d₂₀) ; and b) on the 5' end thereof, a unique random nucleotide sequence or tag whereby a single stranded.DNA sequence is produced which has at its 5* end said unique sequence (segment t₂₁) ; ii) removing excess oligonucleotide primer (d₂₀-t₂₁) ; iii) hybridizing an upstream oligonucleotide primer (u₂₁) complementary to the 3¹ end of said DNA sequence thereto; iv) extending the primer (u₂₁) so that a DNA molecule is produced having at its 3' end a nucleotide sequence complementary to said unique nucleotide sequence; v) denaturing the double-stranded DNA molecule produced in step (iv) ; vi) hybridizing the oligonucleotide primer (u₂₁) to the 3' end of said DNA sequence to which it is complementary and hybridizing an oligonucleotide primer (t₂₁) comprising all or a portion of said unique nucleotide sequence, to the 3' end of said DNA sequence to which it is complementary; vii) extending the primers (u₂₁) and (t₂₁) thereby producing two DNA molecules; and viii) detecting the presence or absence of the amplified DNA sequence.

2. The method according to claim 1 further comprising before step (viii) repeating steps v-vii multiple times.

3. The method according to claim 1 wherein said oligonucleotide primer (d₂₀-t₂₁) is about 41 nucleotides in length.

4. The method according to claim 1 wherein said oligonucleotide primers (u₂₁) and (t₂₁) are each about 21 nucleotides in length.

5. The method according to claim 1 wherein said unique random nucleotide sequence is comprised of approximately 25% of each nucleotide.

6. The method according to claim 1 wherein said RNA sequence is a segment of mRNA.

7. The method according to claim 6 wherein said segment of mRNA is a segment of insulin mRNA.

8. The method according to claim 7 wherein said insulin mRNA is Xenopus insulin mRNA.

9. The method according to claim 1 wherein said unique nucleotide sequence is 5'- GACAAGCTTCAGGTAATCGAT-3' .

10. The method according to claim 1 wherein said unique nucleotide sequence is 5'- CCGAATTCTGTAGTCCGTCA-3^» .

11. The method according to claim 1 wherein said segment of RNA is contaminated with DNA.

12. A method of detecting an RNA sequence comprising the steps of: i) reverse transcribing the RNA sequence from an oligonucleotide primer hybridized thereto which oligonucleotide primer (d₁₇-t₃₀) comprises: a) on the 3' end thereof (segment d₁₇) , a nucleotide sequence complementary to a region of the RNA sequence; and b) on the 5' end thereof (segment t₃₀) , a unique nucleotide sequence or tag whereby a single stranded DNA sequence is produced which has at its 5' end said unique sequence; ii) hybridizing an upstream oligonucleotide primer (u₃₀) , complementary to said DNA sequence, to a region of said sequence, at a temperature selected so that said d₁₇ segment does not anneal to contaminating DNA but so that said primer u₃₀ does anneal; iii) extending the primer (u₃₀) so that a DNA molecule is produced having at its 3' end a nucleotide sequence complementary to said unique nucleotide sequence (segment t₃₀) ; iv) denaturing the double-stranded DNA molecule produced in step (iii) ; v) hybridizing the oligonucleotide primer (u₃₀) to said DNA sequence to which it is complementary and hybridizing an oligonucleotide primer (t₃₀) comprising all or a portion of said unique nucleotide sequence, to a region of said DNA sequence to which it is complementary, wherein said hybridization is carried out at a temperature selected so that said d₁₇ segment does not anneal to contaminating DNA but so that said primers u₃₀ and t₃₀ do anneal; vi) extending the primers (u₃₀) and (t₃₀) thereby producing two DNA molecules; and vii) detecting the presence or absence of the amplified DNA sequence;

13. The method according to claim 12 wherein said hybridization is carried out at a temperature 42* C or greater.

14. The method according to claim 12 further comprising before step (vii) repeating steps iv-vi multiple times.

15. The method according to claim 12 wherein said hybridization in steps ii and v occurs at a temperature between 65" and 72° C.

16. The method according to claim 12 wherein said oligonucleotide primer (d₁₇-t₃₀) is about 47 nucleotides in length.

17. The method according to claim 12 wherein said oligonucleotide primers (u₃₀) and (t₃₀) are each about 30 nucleotides in length.

18. The method according to claim 12 wherein said unique random nucleotide sequence is comprised of approximately 25% of each nucleotide.

19. The method according to claim 12 wherein said RNA sequence is a segment of mRNA.

20. The method according to claim 18 wherein said segment of mRNA is a segment of insulin mRNA.

21. The method according to claim 19 wherein said insulin mRNA is Xenopus insulin mRNA.