Nucleic Acids Research

¹Department of Pathology and Laboratory Medicine, ²Department of Urology, ³Department of Biochemistry and Biophysics and the ⁴Cancer Center, University of Rochester, 601 Elmwood Avenue, Rochester, NY 14642, USA

Received October 30, 1997; Revised and Accepted December 9, 1997

ABSTRACT

A variety of techniques are currently available for detecting point mutations in DNA. These techniques are frequently not sensitive enough to be applied as quantitative assays in evaluation of relative occurrence of alleles in cases of polymorphism or when variations in allelic gene expression are being evaluated at the level of RNA. We report here the establishment of an iterative gap ligation (IGL) assay that is both quantitative and sensitive. The design of the assay is such that ligation of an upstream to a downstream primer across a single nucleotide gap will only occur if the gap is filled with a deoxynucleotide complementary to the wild-type or mutant sequence. Under conditions in which excess upstream primer saturates the template concurrently with limiting amounts of downstream primer quantitative ligation is absolutely dependent on provision of the appropriate gap filling nucleotide. When gap ligation occurs in a single incubation, or cycle, the amount of ligated product is a linear function of the relative amount of mutant sequence, with a sensitivity and detection limit of ~3% over a range of relative concentrations of 0-100%. When the reaction occurs over multiple cycles, or iterations, gap ligation becomes a non-linear function such that small changes in the relative proportions of alleles produce a disproportionately large amount of ligation. As a consequence, the sensitivity and limits of detection of the assay improve to 0.2% after only 8 cycles. The development of this assay provides a unique means of quantifying allelic polymorphisms in both DNA and RNA (after initial amplification by PCR or RT-PCR) and should be applicable to any experimental settings in which nucleic acids from tissues or mixed populations of cells are being evaluated.

INTRODUCTION

Tumor suppressor gene and oncogene products are of fundamental importance to normal cell growth and differentiation as well as in disease (1). Molecular assays that can assess heterozygosity in these genes or in their transcripts have been important for evaluating tumor cell burden in mixed cell populations and in determining the efficacy of therapy. Several of these techniques have been used extensively in identifying and studying familial forms of cancer at the level of cancer susceptibility genes and gene mutations.

Point mutations have been analyzed by a variety of techniques, including restriction fragment length polymorphism (RFLP; 2), single-strand conformation polymorphism (SSCP) (3), denaturing gradient gel electrophoresis and chemical mismatch cleavage (CMC; 4). Although these assays are frequently able to detect single base changes in DNA, they are not generally applicable since they are so highly dependent on the flanking sequence context of the mutation. In addition, these assays are only semi-quantitative, although some modifications have been published which may allow for a fully quantitative analysis (5).

Several techniques have been developed to quantify the relative amounts of point mutations in a known sequence. Poisoned primer extension is a modified sequencing assay in which a dideoxynucleotide complementary to either the wild-type or mutant allele is added in place of the appropriate deoxynucleotide in an otherwise normal primer extension reaction (6). The amount of mutation is calculated directly from the relative amounts of extension products that stop at either the wild-type nucleotide or the point mutation. Although this assay is easy to perform, it is not universally applicable. It is highly dependent on sequence context in that the same nucleotide as the mutation cannot occur between the 3'-end of the primer and the mutation. Moreover, the resolution of the primer extension product is ideal when the mutation resides >4 nt 5' of the 3'-end of the primer.

Another quantitative assay is solid phase mini-sequencing, which uses single nucleotide extension of biotinylated primer immobilized on a streptavidin-coated solid support (7,8). The relative amount of mutant allele is calculated from standard curves. This assay has excellent sensitivity and is being developed for automation (9).

Alternatively, the ligase chain reaction (LCR) has also been developed for detection of point mutations (10-16). In LCR a pair of oligodeoxyribonucleotide primers is annealed to complementary target DNA sequence juxtaposed to the nucleotide of interest. Ligation of the oligonucleotides can only occur when the 5'-end of the upstream oligo and the 3'-end of the downstream oligo are base paired to the target DNA sequence. Amplification is exponential, but a point mutation at the junction of the two oligos prevents efficient ligation and, hence, amplification. Different pairs of primers can be used to test for specific point mutations. LCR has the advantage of being applicable to all sequences, if the sequence is known. Several approaches have improved the limits of detection of LCR, such as incorporation of PCR followed by limited amplification with LCR (17). Although LCR has generally not been used for quantitative analysis, a recent report claims success in this regard (18). The ligation reaction has also been used on pre-amplified template, such as detection of the presence or absence of mutant sequences (19-23). Another report claims to have quantitatively measured the relative amounts of mutations by a ligation reaction, although no methods were reported in this manuscript (24).

Specificity of LCR has been improved by a modification known as gap LCR (25). In gap LCR a pair of primers is annealed to the target DNA with a gap of one to several bases corresponding to the region of interest. A non-strand displacing thermostable DNA polymerase is then used to fill the `gap' between the two primers by extension of the downstream primer under conditions in which only those deoxynucleotide triphosphates predicted by the template are supplied in the reaction. The downstream extension product and the upstream primer are then ligated by a thermostable DNA ligase and the resulting products are resolved by denaturing polyacrylamide gel electrophoresis. Extension and ligation are therefore dependent on the availability of the appropriate deoxynucleotide triphosphates. A single base change in the target DNA prevents adequate extension of the downstream primer and this is detected as absence of ligation product.

In the present study we develop an assay to detect and quantify a single base change, such as has been reported in the mRNA encoding the WT-1 Wilms' tumor suppressor, which occurs through a uridine to cytidine RNA editing event (26). Our assay uses modified gap LCR methods of pre-amplified template, is very sensitive, simple to implement and can be used for the quantitative analysis of any point mutation. This is the first universally applicable quantitative assay based on the gap ligation reaction.

MATERIALS AND METHODS

Isolation of human WT-1 sequence

RNA was isolated from kidney tumor samples, kindly provided by Drs Rabinowitz and Hulbert (Department of Urology, University of Rochester Medical Center). Reverse transcriptase PCR techniques were employed to isolate a 500 bp fragment of the WT-1 gene in which the unmodified editing site (CTC) was centrally located. First-strand cDNA was generated from 1 µg oligo(dT)-primed total RNA using AMV reverse transcriptase (Promega, WI) according to the manufacturer's recommendations. PCR amplification of the WT-1 sequence encompassing the editing site was generated from the first-strand cDNA using primers WT-1 Forward and WT-1 Reverse (sequences below). PCR was performed with Taq DNA polymerase (Promega) according to the manufacturer's recommendations. Thermocycling conditions were: 1(pre-incubation) cycle at 94°C for 3 min; 5 cycles at 94°C for 1 min, 52°C for 2 min, 72°C for 2 min; 30 cycles at 94°C for 1 min, 60°C for 1 min, 72°C for 1.5 min. PCR products were gel isolated (Qiagen, CA) and subcloned into pcDNA III (Invitrogen, CA) at BamHI and EcoRI restriction sites. All DNA sequences were confirmed by dyedeoxy terminator chemistry (ABI Prism^tm) according to the manufacturer's recommendations and the products analyzed in the Core Nucleic Acid Facility (University of Rochester Medical Center).

Site-directed mutagenesis

An edited version (CCC) was generated using primers PK-1 and PK-2 by site-directed mutagenesis (QuikChange^tm Site-Directed Mutagenesis Kit) according to the manufacturer's recommendations (Stratagene, CA). Clones were sequenced by dyedeoxy terminator chemistry according to the manufacturer's recommendations (ABI Prism^tm). Plasmid DNAs were purified by a modified alkaline lysis technique (Qiagen Inc., CA).

Synthesis of edited and unedited single-stranded target DNAs

To generate single-stranded cDNA templates, asymmetric PCR amplification was used with the 5'-primer (WT-1 Forward) in a 100-fold molar excess of the 3'-primer (WT-1 Reverse), 100 and 1 µm respectively. Forty nanograms of template DNA was used. PCR was performed with Taq DNA polymerase (Promega) according to the manufacturer's recommendations. Thermocycling conditions were: 35 cycles at 95°C for 30 s, 60°C for 30 s and 72°C for 90 s. An aliquot of 5 µCi [³H]TTP (New England Nuclear, MA) was included in each asymmetric PCR reaction for subsequent quantitation of the product, giving a final specific activity PCR reaction mix of 0.2 µCi/mM TTP. Single-stranded products were purified from the double-stranded products by agarose gel electrophoresis and subsequent electrophoresis of the DNA out of agarose slices into buffer contained within dialysis bags. Confirmation of single-stranded cDNA was performed by restriction enzyme digestion using a double-stranded DNA-specific enzyme. The concentration of single-strand cDNA was calculated from the specific activity in the PCR reaction, the number of dATP residues in the complementary strand (excluding the primer) and the d.p.m. of [³H]TTP incorporated.

Iterative gap ligation reaction

WTLig-1 (22 nt) was synthesized, purified by PAGE and 5' phosphorylated (Genosys, TX). WTLig-2 (25 nt) was synthesized (Cyclone II, Millagen), gel isolated, then end-labeled using 3000 Ci/mmol [[gamma]-³²P]ATP (NEN, MA) and T4 polynucleotide kinase (US Biochemical, OH) with the manufacturer's buffers and conditions.

Gap ligation reactions were performed in buffer containing 20 mM Tris-HCl, 25 mM potassium acetate, 10 mM magnesium acetate, 10 mM DTT, 1 mM NAD, 0.1% Triton X-100, pH 7.6 (25°C). Forty femtomoles of single-stranded template were used with 0.4 fmol ³²P-labeled downstream primer (WTLig-2), 7.5 pmol upstream primer (WTLig-1),200 nMgap filling nucleotide (dATP or dGTP), 0.25 U Taq DNA polymerase (Chimerx, WI) and 10 U Taq DNA ligase (New England Biolabs, MA) in a total reaction volume of 50 µl.

Cycling conditions included incubation at 50°C for extension and ligation, followed by melting at 90°C, using an Omn-E thermal cycler (Hybaid) equipped with a heated lid. Incubation times varied according to the number of cycles: 1 cycle of 90 min; 2 cycles of 50 min each; 4 cycles of 25 min each; 8 cycles of 15 or 12 min each. Reaction products were precipitated with ethanol and resolved in a 10% denaturing polyacrylamide gel. Gels were visualized by autoradiography, but quantified by laser scanning phosphorimager densitometry (Molecular Dynamics, CA). Percentage ligation was calculated as [ligation product/(unligated primer + ligation product)].

Regression curves were fitted using Statview (Abacus Concepts, CA) and Cricket (Computer Associates, CA) software. Sensitivity (the ability to discriminate between differences in relative amounts of template based upon the experimental data and the fitted curves) and detection limits (the least amount of template able to be discerned) were calculated by two-tailed Student's t-test ([alpha]/2 = 0.025).

Figure 1. Annealing sites of gap ligation primers on the asymmetric PCR template. A select region of single-stranded PCR template corresponding to the region of the WT-1 gene flanking a point mutation (the edited site) is shown. Gap ligation primers (WTLig-1 and WTLig-2, shown as black arrows) are both phosphorylated at their 5'-ends; the asterisk on WTLig-2 indicate ³²P-labeling. Their annealing position allows for a single nucleotide gap to be present. The open letters T/C at the gap indicate the allelic variants (point mutations) corresponding to the unedited and edited versions respectively of the WT-1 mRNA.

Oligodeoxyribonucleotides

Primers used in this study were: WT-1 Forward, 5'-CTC GGA TCC ACC TGG AAT CAG ATG AAC-3'; WT-1 Reverse, 5'-CTC GAA TTC TGT ATG AGT CCT GGT GTG-3'; PK-1, 5'-GTG TGT TGC GGG TAG GGG ACG CCT CGG GTT ATG-3'; PK-2, 5'-CAT AAC CCG AGG CGT CCC CTA CCC GCA ACA CAC-3'; WTLig-1, 5'-GGA TGG GCG TTG TGT GGT TAT C-3'; WTLig-2, 5'-GTG TAT TCT GTA TTG GGC TCC GCA G-3'.

Figure 2. Gap ligation reaction visualized by autoradiography. Asymmetric PCR products (target sequences) corresponding to the edited and unedited allelic versions were synthesized and annealed to WTLig-1 and WTLig-2 as described in Materials and Methods. Parallel reactions were run containing either the appropriate (edited/dGTP, unedited/dATP) or inappropriate (edited/dATP, unedited/dGTP) target sequence/gap filling nucleotide combinations under reaction conditions described in Materials and Methods for four cycles. Positions indicated to the right of the actual autoradiograph correspond to unreacted WTLig-2 (25mer), the extension products of WTLig-2 that have not ligated (26mer with dATP, 28mer with dGTP) and extended WTLig-2 that has ligated to WTLig-1 (48mer with dATP, 49mer with dGTP). Additional minor bands probably result from non-processive 3' -> 5' exonuclease contamination and do not interfere with quantifying the amount of ligation.

RESULTS AND DISCUSSION

Specificity of iterative gap ligation

Asymmetric PCR products were synthesized as described in Materials and Methods and designated as the edited allele (CCC) or the unedited allele (CTC) based on the presence of a naturally occurring polymorphism at nt 2830, which arises from a U -> C mRNA editing event (25). The primers (WTLig-1 and WTLig-2) were designed to anneal to both allelic forms of the asymmetric PCR products flanking the edited nucleotide while leaving a single nucleotide gap corresponding to the polymorphic nucleotide (Fig. 1). WTLig-1 and WTLig-2 were both phosphorylated, but WTLig-2 was ³²P-end-labeled. Several bands are seen in the autoradiograph depicted in Figure 2. Unligated downstream primer (WTLig-2) was identified as a 25mer (upstream primer was not visualized since it was not ³²P-labeled). In the presence of dATP a 26mer band could be seen that was consistent with a single nucleotide extension of WTLig-2. Extension was limited to a single nucleotide because the template sequence read after the primer is (C/T)CCT. In the presence of dGTP a28mer extension product of WTLig-2 was also identified, which occurs when WTLig-1 was not concurrently annealed to the template. This was a rare event, since annealing was strongly favored thermodynamically (190-fold molar excess of WTLig-1 relative to template, in a reaction occurring significantly below the calculated melting temperature). Extension products >1 nt were not included in the calculation of percent ligated primer. Primer ligation across the gap in the presence of dATP on unedited template was evident as an electrophoretic mobility shift of [³²P]WTLig-2 from that of a 25mer to a 48mer, the combined length of WTLig-1 and WTLig-2 plus 1 nt. In the presence of dGTP on edited template a 49mer was the predominant ligation product and resulted from a 1 nt extension of WTLig-1, predicted by the sequence. Minor bands representing incorrect single nucleotide extensions were seen as a 49mer (+dATP) and a 50mer (+dGTP). Several other bands were reproducibly observed, which likely result from contaminating 3' -> 5' non-processive exonuclease. This was suggested by the finding that in the presence of dATP and unedited template the ligated product (48mer) had a 1 nt deletion (47mer) while the unligated primer had a 2 nt deletion (23mer). Since the 23mer could not participate in ligation it was excluded in the calculation of reaction product. In the absence of dNTP no ligation products were observed, indicating that neither template- nor nucleotide-independent ligation occur (data not shown).

When dATP was provided as the gap filling nucleotide, ligated product was only observed with the unedited allele, which was supplied in the reaction as single-stranded DNA template. Similarly, preferential ligation of primers was observed with the edited allele when dGTP was provided. Moreover, the data clearly demonstrate that efficient ligation did not occur when an inappropriate gap filling deoxynucleotide was provided. These results demonstrate that conditions have been established for gap ligation to allow specific detection of cDNAs differing by only a point mutation.

The ability of gap ligation to detect different alleles relies on: (i) correct extension of the downstream primer (WTLig-2); (ii) correct ligation of the correctly extended WTLig-2 to the upstream primer (WTLig-1), which only occurrs efficiently in the presence of the deoxynucleotide triphosphate complementary to the template strand at the gap. Two important requirements of this polymerization are enzyme fidelity and the absence of 5' -> 3' exonuclease activity, which would otherwise displace the upstream primer. Taq DNA polymerase (Chimerx, WI) has been reported previously to satisfy these requirements (21) and was used in the IGL reactions reported here. In addition, since Taq fidelity is partly dependent upon the concentration of deoxynucleotide triphosphate present, we determined by titration that a concentration of 200 nM (dATP or dGTP) optimizes assay fidelity (data not shown). Increasing reaction temperature may also increase fidelity, but since the melting temperature of primer to template is inversely proportional to a logarithmic function of primer concentration (27), 50°C was found to be the optimal temperature at which complete annealing could be assured under dilute WTLig-2 (0.4 fmol) conditions.

Time dependence of the gap ligation reaction

The gap ligation reaction was performed in the presence of both the appropriate and inappropriate gap filling nucleotide for the edited and unedited templates. Reaction tubes were pulled at various times during a single incubation and analyzed for the amount of ligation. Data from a single experiment are shown in Figure 3A. A good linear fit was made between ligation and log(time), with r² of 0.987 and 0.971 for edited and unedited templates respectively. Ligation in the presence of dATP was relatively less efficient than in the presence of dGTP, perhaps for reasons intrinsic to the polymerase and/or ligase enzymes. The slopes of these curves decreased with increasing amounts of downstream primer or with decreasing amounts of free nucleotide (data not shown). Also apparent in Figure 3A is the very low efficiency of false ligation, which occurs in the presence of inappropriate nucleotide. False ligation was <2% in these and other experiments. In the titration experiments described below false ligation is represented as the y-intercept and does not contribute to the error of the fitted lines. The curves represented in Figure 3A were used to estimate optimal incubation times under cycling conditions. The effect of cycling on ligation is shown in Figure 3B. These curves demonstrated that cycling drives the gap ligation reaction to completion, without an increase in false ligation. Therefore, reaction conditions were evaluated that might maximize the dynamic range of the assay.

Figure 3. The effect of incubation length and iteration on the gap ligation reaction. (A) Various time points of a single cycle of gap ligation reaction, using edited and unedited template in the presence of dGTP or dATP, were resolved by denaturing gel electrophoresis. These data were quantified by phosphorimaging scanning densitometry and are depicted graphically as amount of ligation as a function of log(time). Correlation coefficients (r²) are 0.987 and 0.971 for edited (+dGTP) and unedited (+dATP) respectively. Very little false ligation (edited template + dATP or unedited template + dGTP) is observed, typically ~1%. These curves were subsequently used to estimate best incubation times under cycling conditions. Each data point represents a single determination. (B) The amount of ligation was assayed, using edited and unedited template in the presence of dGTP or dATP, after 1, 2, 3, 5 and 8 cycles at 50°C × 12 min under conditions described in Materials and Methods. These data were quantified by phosphorimaging scanning densitometry and are depicted graphically. Each data point represents a single determination. The curve is drawn freehand.

Quantitative range of assay

Having demonstrated a high degree of specificity, we tested our assay's ability to quantify polymorphisms. To accomplish this, we established the primer to template molar ratios and cycling conditions described in Materials and Methods that would enable a predictable amount of [³²P]WTLig-2 to be converted to ligated product, relative to an experimentally established molar ratio of alleles in any given reaction. Gap ligation was performed on mixtures of edited (Fig. 4A) and unedited (Fig. 4B) single-stranded DNA templates (ranging from 0-100%) in the presence of the appropriate gap filling deoxynucleotide triphosphate, and the effects of cycling were observed. All quantification was done by PhosphoImager.

Figure 4. IGL reaction over a broad range of relative allele concentrations (0-100%). Mixtures of asymmetric PCR products corresponding to the indicated percentages of (A) edited allele relative to the unedited allele (determined by adding known amounts of edited and unedited) in the presence of dGTP and (B) unedited relative to the edited allele in the presence of dATP were subject to iterative gap ligation, resolved by denaturing gel electrophoresis and quantified by phosphorimage scanning densitometry, as described in Materials and Methods. These data are depicted graphically as amount of ligation as a function of relative amount of edited template (A) and of unedited template (B). The results of two experiments for each cycle are shown. Incubation times were 90, 50, 25 and 12 min for the 1, 2, 4 and 8 cycle experiments respectively. Correlation coefficients (r²) are all >= 0.990. The results of two experiments for each cycle are shown. The error in many points is smaller than the symbol used to represent the average value.

Using only a single cycle a linear relationship was observed between the amount of measured ligated product and the known relative amount of each allele (r² = 0.994 and 0.967 for the edited and unedited templates respectively). The data demonstrate that this assay had a broad range with good linearity, establishing its potential for quantitative analysis. Moreover, a two-tailed Student's t-test estimated a sensitivity of 2.4 and 3.1% and a limit of detection of 3.0 and 3.8% for the edited and unedited templates respectively. The most significant modification contributing to the extended linear range was the use of rate-limiting quantities of [³²P]WTLig-2. In pilot experiments (data not shown) gap ligation was performed under varying ratios of WTLig-1:WTLig-2:single-stranded DNA template. A molar ratio of ~10^4-5:1:10² was determined to provide optimal signal and fidelity. Under these conditions all single-stranded DNA templates underwent saturation hybridization to WTLig-1, ensuring that virtually all extension of WTLig-2 occurred in the context of appropriately positioned WTLig-1, ready for ligation.

Effects of cycling

Figure 4 also demonstrates the effects of cycling. With increasing numbers of cycles the observed ligation became a non-linear function of relative template concentration, with a disproportionate amount of ligation occurring at low levels of relative template concentration. This increased signal provided an opportunity to increase the assay's sensitivity in determining relative template concentrations of 50% or less. The assay's sensitivity was always increased by iterations, because one or the other allele must always occur in a proportion of 50% or less.

Limits of detection of iterative gap ligation reaction

We next evaluated the detection limit of IGL in analyses wherein a narrower range (0-10%) of allelic mixtures were quantified. A linear response was observed between production of ligated primers in the presence of the appropriate gap filling deoxynucleotide triphosphate (Fig. 5), because so small a range of relative template concentrations was used. A two-tailed Student's t-test demonstrated an increasing sensitivity (2.3 to 0.2%) and decreasing detection limit (1.9 to 0.2%) with increasing iterations (see Table 1).

Figure 5. IGL reaction over a narrow range of relative allele concentrations (0-10%). Mixtures of asymmetric PCR products corresponding to the indicated narrow range (0-10%) of edited allele relative to the unedited allele were subject to iterative gap ligationin the presence of 500 nM dGTP, resolved by denaturing gel electrophoresis and quantified by phosphorimage scanning densitometry, as described in Materials and Methods. These data are depicted graphically as amount of ligation observed as a function of relative amount of edited template. The results of two experiments for each cycle are shown. The error in many points is smaller than the symbol used to represent the average value.See Table 1 for results.

Conclusions

This is the first fully quantitative, universally applicable assay based on gap ligation methodology. We have demonstrated that this modified gap ligation reaction is highly specific, with a linear signal over a broad range of relative allelic concentrations when no cycling is employed. We have also demonstrated that by increasing the number of iterations to seven (i.e. 8 cycles) the sensitivity and the limits of detection improve by an order of magnitude. We did not test the assay using more than seven iterations, but sensitivity could be theoretically improved further by additional cycles. The assay is easy to perform with reagents available in most molecular biology laboratories. Importantly, as the assay is not dependent upon any flanking sequences, it is applicable for assaying relative concentrations of any point mutation. In addition, microdeletions of one to several nucleotides can be quantified by this assay, since these would produce ligation in the absence of free nucleotide.

Table 1 . Sensitivity and detection limits as a function of cycles at relative edited template concentrations of 0-10% in the presence of dGTP

No. cycles	Min/cycle	r2	Sensitivitya (%)	Detection limita (%)
1	90	0.967	2.3	1.9
2	50	0.964	1.0	0.9
4	25	0.994	0.3	0.3
8	15	0.996	0.2	0.2

^aDetermined by two-tailed Student's t-test ([alpha]/2 = 0.025).

We have developed this assay in the context of a naturally occurring polymorphism in the WT-1 tumor suppressor gene product at nt 2830, which arises from a U -> C mRNA editing event (26). The potential application of this assay in the quantification of other polymorphisms in the DNA from tissues of mixed populations of cells as a means of evaluating cell type heterogeneity is readily apparent. Moreover, it is anticipated that IGL will prove important for quantifying RNAs expressed from alleles differing by only a single or a few nucleotides or in determining the occurrence of base modification in RNA arising from RNA editing.

ACKNOWLEDGEMENTS

The authors wish to thank Drs W.C.Hulbert and R.Rabinowitz (Department of Urology, University of Rochester Medical Center) for kindly providing specimens of tumor tissue and Jenny M.L.Smith for preparation of the figures. This work was supported in part through a Public Health Services grant (DK 43739) and grants from The Council for Tobacco Research and the Rochester Area Pepper Center for Research on Aging, awarded to H.C.S. J.S. was supported through a Wilmot Postdoctoral Cancer Research Fellowship.

REFERENCES

1 Brown,M.A. and Solomon,E. (1997) Trends Genet., 13, 202-206. MEDLINE Abstract

2 Cavenee,W.K., Dryja,T.P., Phillips,R.A., Benedict,W.F., Godbout,R., Gallie,B.L., Murphree,A.L., Strong,L.C. and White,R.L. (1983) Nature, 305, 779. MEDLINE Abstract

3 Nakabayshi,Y. and Nishigaki,K. (1996) J. Biochem., 120, 320-325.

4 Haris,I.I., Green,P.M., Bentley,D.R. and Giannelli,F. (1994) PCR Methods Applicat., 3, 268-271.

5 Tanno,Y., Yoneda,M., Tanaka,K., Tanaka,H., Yamazaki,M., Nishizawa,M., Wakabayashi,K., Ohama,E. and Tsuji,S. (1995) Muscle Nerve, 18, 1390-1397. MEDLINE Abstract

6 Davies,M.S., Wallis,S.C., Driscoll,D.M., Wynne,J.K., Williams,G.W., Powell,L.M. and Scott,J. (1983) J. Biol. Chem., 264, 13395-13398.

7 Syvanen,A.-C., Sajantila,A. and Lukka,M. (1993) Am. J. Hum. Genet., 52, 46-59. MEDLINE Abstract

8 Syvanen,A.-C. (1994) Clin. Chim. Acta, 226, 225-236. MEDLINE Abstract

9 Ihalainen,J., Siitari,H., Laine,S., Syvanen,A.-C. and Paoltie,A. (1994) BioTechniques, 16, 938. MEDLINE Abstract

10 Wu,D.Y. and Wallace,R.B. (1989) Genomics, 4, 560-569. MEDLINE Abstract

11 Barany,F. (1991) Proc. Natl. Acad. Sci. USA, 88, 189-193.

12 Backman,K. (1989) European patent application EP-A-0320 308 (priority filing 1987).

13 Laffler,T.G., Carrino,J.J. and Marshall,R.L. (1993) Annls Biol. Clin., 50, 821-826.

14 Bond,S., Carrino,J., Hample,H., Hanley,H., Reinhardt,L. and Laffler,T. (1990) Papillomaviruses in Human Pathology. Recent Progress in Human Epidermoid Precancers. Raven Press, New York, NY, Vol. 78, p. 425.

15 Dille,B.J., Butzen,C.C. and Birkenmeyer,L.G. (1993) J. Clin. Microbiol., 31, 729-731. MEDLINE Abstract

16 Kalin,I., Shephard,S. and Candrian,U. (1992) Mutat. Res., 283, 119-123. MEDLINE Abstract

17 Wiedmann,M., Barany,F. and Batt,C.A. (1993) Appl. Environ. Microbiol., 59, 2743-2745. MEDLINE Abstract

18 Minamitani,S., Nishiguchi,S., Kuroki,T., Otani,S. and Monna,T. (1997) Hepatology, 25, 216-222. MEDLINE Abstract

19 Wu,D.Y. and Wallace,R.B. (1989) Gene, 76, 245-254. MEDLINE Abstract

20 Samiotaki,M., Kwiatkoski,M., Parik,J. and Landegren,U. (1994) Genomics, 20, 238-242. MEDLINE Abstract

21 Landegren,U., Kaiser,R., Sanders,J. and Hood,L. (1988) Science, 241, 1077-1080. MEDLINE Abstract

22 Frenkel,L.M., Wagner,L.E., Atwood,S.M., Cummins,T.J., and Dewhurst,S. (1995) J. Clin. Microbiol., 33, 342-347. MEDLINE Abstract

23 Redston,M.S., Papadopoulos,N., Caldas,C., Kinzler,K.W. and Kern,S.E. (1995) Gastroenterology, 108, 383-392. MEDLINE Abstract

24 Powell,S.M., Petersen,G.M., Krush,A.J., Booker,S., Jen,J., Giardiello,F.M., Hamilton,S.R., Vogelstein,B. and Kinzler,K.W. (1993) New Engl. J. Med., 329, 1982-1987. MEDLINE Abstract

25 Abravaya,K., Carrino,J., Muldoon,S. and Lee,H. (1995) Nucleic Acids Res., 23, 675-682. MEDLINE Abstract

26 Sharma,P.M., Bowman,M., Madden,S.L., Rauscher,F.J. and Sukumar,S. (1994) Genes Dev., 8, 720-731. MEDLINE Abstract

27 Innis,M.A., Gelfand,D.H. and Sninsky,J.J. (eds) (1995) PCR Strategies. Academic Press, San Diego, CA, pp. 78-80.

---

*To whom correspondence should be addressed at: Department of Biochemistry and Biophysics, University of Rochester, 601 Elmwood Avenue, Rochester, NY 14642, USA. Tel: +1 716 275 4267; Fax: +1 716 273 1027; Email: hsmith@bphvax.biophysics.rochester.edu
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals Comments and feedback: www-admin{at}oup.co.uk
Last modification: 6 Feb 1998
Copyright© Oxford University Press, 1998.

CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract Print PDF (123K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (1) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Stewart, J. Articles by Smith, H. C. Search for Related Content PubMed PubMed Citation Articles by Stewart, J. Articles by Smith, H. C. Social Bookmarking          What's this?