Oncogene
SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Archive
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Subscribe
Advertising
naturereprints
Contact NPG
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
 
14 March 2002, Volume 21, Number 12, Pages 1909-1921
Table of contents    Previous  Article  Next   [PDF]
Technical Report
An endonuclease/ligase based mutation scanning method especially suited for analysis of neoplastic tissue
Jianmin Huang1, Brian Kirk1, Reyna Favis1, Thierry Soussi2, Philip Paty3, Weiguo Cao1,a and Francis Barany1

1Department of Microbiology, Box 62, Hearst Microbiology Research Center, Strang Cancer Prevention Center, Joan and Sanford I Weill Medical College of Cornell University, Room B-406, 1300 York Avenue, New York, NY 10021, USA

2Laboratoire de Génotoxicologie des tumeurs, Institut Curie, 26 rue d'Ulm, 75248 Paris, France

3Colorectal Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, Room C-983, 1275 York Avenue, New York, NY 10021, USA

Correspondence to: F Barany, Department of Microbiology, Box 62, Hearst Microbiology Research Center, Strang Cancer Prevention Center, Joan and Sanford I Weill Medical College of Cornell University, Room B-406, 1300 York Avenue, New York, NY 10021, USA. E-mail: barany@med.cornell.edu

aCurrent address: Department of Genetics and Biochemistry, South Carolina Experiment Station, Clemson University, 122 Long Hall, Clemson, SC 29634-0324, USA.

Abstract

Knowledge of inherited and sporadic mutations in known and candidate cancer genes may influence clinical decisions. We have developed a mutation scanning method that combines thermostable EndonucleaseV (Endo V) and DNA ligase. Variant and wild-type PCR amplicons are generated using fluorescently labeled primers, and heteroduplexed. Thermotoga maritima (Tma) EndoV recognizes and primarily cleaves heteroduplex DNA one base 3' to the mismatch, as well as nicking matched DNA at low levels. Thermus species (Tsp.) AK16D DNA ligase reseals the background nicks to create a highly sensitive and specific assay. The fragment mobility on a DNA sequencing gel reveals the approximate position of the mutation. This method identified 31/35 and 8/8 unique point mutations and insertions/deletions, respectively, in the p53, VHL, K-ras, APC, BRCA1, and BRCA2 genes. The method has the sensitivity to detect K-ras mutations diluted 1 : 20 with wild-type DNA, a p53 mutation in a 1.7 kb amplicon, and unknown p53 mutations in pooled DNA samples. EndoV/Ligase mutation scanning combined with PCR/LDR/Universal array proved superior to automated DNA sequencing for detecting p53 mutations in colon tumors. This technique is well suited for scanning low-frequency mutations in pooled samples and for analysing tumor DNA containing a minority of the unknown mutation.

Oncogene (2002) 21, 1909-1921 DOI: 10.1038/sj/onc/1205109

Keywords

mutation detection; endonuclease V; thermostable ligase; mismatch recognition

Introduction

Cancers arise from the accumulation of inherited and sporadic mutations in cell cycle, DNA repair, and growth signaling genes. Knowledge of the inherited or sporadic change in tumor suppressor genes or oncogenes could influence prevention and treatment decisions. For example, certain specific single nucleotide polymorphisms (SNPs) in cancer genes such as BRCA1, BRCA2 or APC are present at low frequency (1-5%) in certain ethnic groups. These carriers are at significantly higher risk of developing breast, ovarian, prostate or colon cancers (Abeliovich et al., 1997; Beller et al., 1997; Berman et al., 1996; Laken et al., 1997; Oddoux et al., 1996; Roa et al., 1996; Struewing et al., 1995, 1997). Similarly, those with inherited mutations in the von Hippel-Lindau gene (VHL) would benefit from early screening and preventive surgery at the appropriate time (Chen et al., 1996; Stolle et al., 1998; Libutti et al., 2000; Mohr et al., 2000; Sgambati et al., 2000; Walther et al., 1999). Sporadic mutations in the p53 gene influence both clinical outcome and response to therapy (Broll et al., 1999; Bunz et al., 1999; Dameron et al., 1994; Heide et al., 1997; Prives and Hall, 1999; Tortola et al., 1999; Zou et al., 2000). The precise nature of the p53 mutation presents both challenges and opportunities for alternate treatment strategies in specific cancers (Wang et al., 1998b; Aurelio et al., 1998, 2000; Foster et al., 1999; Webley et al., 2000). These diverse genes highlight the clinical need to accurately identify often unknown inherited aberrations or infrequently represented mutations in mixed populations of DNA molecules.

A variety of methods have been developed to successfully scan for unknown mutations. Unfortunately, current technologies used for in vitro scanning are typically lacking in either throughput capacity or sensitivity to rapidly detect a range of mutations. This can result in false-negatives and lower throughput, since many of these techniques cannot detect tumor mutations in the presence of contaminating stromal cells, or be amenable to the use of pooled samples. Cycle-sequencing can detect any mutation and its position (Innis et al., 1988; Tabor and Richardson, 1995), but this method has low sensitivity which prevents an accurate analysis of pooled samples. Automated fluorescent dideoxysequencing often failed to detect germline mutations in DNA repair genes (Yan et al., 2000a,b). Variation detection arrays (VDA) use standard hybridization microarrays to scan large sequence blocks in given genes, however, false positive rates of 11-21% have been observed, and there are particular difficulties in detecting frameshift mutations (Cargill et al., 1999; Halushka et al., 1999; Wang et al., 1998a; Hacia et al., 1996; Hacia, 1999). A direct comparison between commercially available p53 hybridization chips and automated fluorescent dideoxysequencing showed they miss 23% and 25% of mutations in tumors, respectively (Ahrendt et al., 1999). These techniques have reached their theoretical limits due to the dilution of mutant DNA in tumor samples by stromal cell infiltration.

Unknown mutations may be detected by resolving homoduplex and heteroduplex DNA based on their differing electrophoretic migration using single-stranded conformational polymorphism (SSCP) (Hayashi, 1991; Korn et al., 1993; Makino et al., 1992; Suzuki et al., 1990), denaturing-gradient gel electrophoresis (DGGE) (Fahy et al., 1997; Fodde and Losekoot, 1994; Guldberg and Guttler, 1994; Ridanpaa and Husgafvel-Pursiainen, 1993; Ridanpaa et al., 1995), constant denaturing capillary electrophoresis (CDCE) (Chen and Thilly, 1994; Khrapko et al., 1994), dideoxy fingerprinting (ddF) (Sarkar et al., 1992), and restriction endonuclease fingerprinting (REF) (Liu and Sommer, 1995), and more recently, denaturing high-performance liquid chromatography (DHPLC) (Underhill et al., 1996). Each of these approaches has their advantages, nevertheless, none of the approaches has the combined ability to rapidly detect low level mutations, distinguish missense from silent polymorphisms, and locate the position of the mutation. Polymorphisms may also be identified by cleavage of mismatches in DNA-RNA hybrids via RNase A mismatch cleavage (Myers et al., 1985; Perucho et al., 1989; Winter et al., 1985), as well as in DNA-DNA homoduplexes via chemical cleavage (Cotton et al., 1988; Hansen et al., 1995; Haris et al., 1994), or enzymatic cleavage by T4 Endonuclease VII, CEL1, or MutY (Giunta et al., 1996; Oleykowski et al., 1998; Xu et al., 1996; Youil et al., 1995). These enzymatic cleavage approaches identify the approximate position of most polymorphisms. However, these enzymes often also nick matched DNA leading to high backgrounds and limiting their usefulness in identifying DNA lesions in solid tumors. In an ideal mutation scanning method, no or little cleavage should be observed in a completely matched DNA, while a wide variety of mismatches should result in cleavage (Taylor, 1999).

To overcome background signal from enzymatic cleavage, we developed a mutation scanning method which combines the mismatch recognition and nicking ability of thermostable Endonuclease V (EndoV) with the high fidelity of a thermostable DNA ligase to proofread and reseal background nicks at duplex DNA (see Figure 1). Endonuclease V recognizes a wide variety of DNA lesions, predominantly deoxyinosine, xanosine, uracil, and apurinic sites, and under select buffer conditions, base mismatches and insertions/deletions (Huang et al., 2001; Yao and Kow, 1994; Liu et al., 2000). The enzyme nicks DNA one base 3' to the mismatch in heteroduplexed DNA (Figure 1). As with all mismatch recognition enzymes, EndoV will also nick matched regions of the DNA, but in contrast to other enzymes, EndoV leaves a ligatable end with a 5' phosphate (Yao et al., 1994; Yao and Kow, 1994). We have cloned a thermostable ligase from Thermus species AK16D that exhibits 2.5 to 5-fold better discrimination at the 3' penultimate position compared to other ligases (Tong et al., 1999). By using these two enzymes sequentially, we can religate spurious nicks, while maintaining the desired nicks at the polymorphic sites, resulting in a highly sensitive mutation scanning assay. In this report, we describe an optimized scheme for detection of cancer-associated mutations using Tma. EndoV in conjunction with Tsp. AK16D Ligase. PCR fragments up to 1.7 kb were tested and correct mutations can be detected in the entire fragment size range. The approximate position of the mutation is determined from the fragment length, and confirmed by sequencing. The assay can detect 98% of the typical mutations found in the human genome, with only two GC-rich sequences refractory to detection. It is also highly sensitive and can identify a mutant sequence at a 1 : 20 dilution in wild-type DNA, making it amenable to pooling strategies.

Results

Strategy, optimization and characterization of EndoV/ligase mutation scanning assay

Initial characterization of Tma EndoV activity was performed on fluorescently labeled synthetic substrates containing mismatched bases. Optimal cleavage at the mismatch with minimal non-specific cleavage was achieved in a 10 mM HEPES (pH 7.4) buffer containing, 1 mM DTT, 2% glycerol, and 5 mM MgCl2 (Huang et al., 2001). Analysis of cleavage patterns of all 12 natural mismatched base-pairs suggests that purine bases are preferentially cleaved, showing a general hierarchy of A=G>T>C (Table 1). This hierarchy is based on the amount of cleavage product observed for each strand of the mismatched heteroduplexed substrates.

To test the utility of the EndoV/ligase scheme for finding mutations in the genome, cell line DNA containing either wild-type, G12V, or G12D mutations in the K-ras gene were PCR amplified (see Table 2 for list of PCR primers used). Mutant amplicons were mixed with wild-type fragments, denatured, reannealed, and treated with EndoV. The single nucleotide change G12V generates G/A and T/C mismatches, while G12D generates G/T and A/C mismatches. Initial results in the above buffer yielded weak bands, however, cleavage of the heteroduplex DNA was enhanced by the addition of 5% DMSO and 1-1.5 M N,N,N-trimethylglycine (betaine, data not shown). The EndoV cleaved both strands for G12V/G12wt and G12D/G12wt heteroduplexes, resulting in products of approximately 159 nt and 116 nt (Figure 2). Without the addition of DNA ligase, a faint, larger, non-specific cleavage fragment was observed (Figure 2). However, this fragment was significantly reduced when the cleavage products were incubated with Tsp. AK16D ligase (compare first four lanes with last 12 lanes). Thus, non-specific cleavage nicks can be preferentially sealed by the DNA ligase, which acts as a proofreading enzyme in this assay.

Some low molecular weight non-specific fragments still persist following the ligation step (bottom, Figure 2). They could be reduced by filtration of the PCR sample prior to heteroduplex formation using Microcon 30 spin columns (data not shown). Filtration removes most unused primers, which can hybridize to the ends of single-stranded PCR products during the heteroduplex formation step. EndoV cleaves at the single-strand region of these PCR fragments to generate low molecular weight non-specific bands. This artifact can be alleviated by either removing unused primers, or designing the primers 30 bp outside the anticipated region of mutation scanning, such that lower molecular weight fragments do not interfere with interpretation of the results.

Most of the current gel-based mutation scanning methods are limited to short PCR fragments. To assess the application of the EndoV/ligase assay on long PCR fragments, 1.7 kb fragments containing either wild-type or p53 R248W mutations were amplified. Fragments containing this mutation generated a unique 413 nt cleavage product that was not observed in the pure wild-type sample (Figure 3). In addition, treatment of the cleavage products with ligase resulted in reduction of most non-specific cleavage bands. Remaining bands (i.e. in lanes 5, 7 and 8) represent putative SNPs, but this needs to be confirmed by sequencing. As is often the case with Cright arrowT transitions, cleavage of only one strand (containing the Gright arrowA change) by EndoV was observed.

To determine the sensitivity of the EndoV/ligase assay, PCR fragments containing K-ras G12V, G12D, and G12A mutations were diluted with a wild-type PCR fragment in ratios ranging from 1 : 1 to 1 : 100. The relative fluorescence intensity of EndoV cleavage products as a function of mutant-to-wild-type fragment ratios were determined (Figure 4). Using the criterion that cleavage signals twofold above background are significant, these results indicate that the cleavage signal can be detected in samples diluted 20-fold with wild-type PCR fragments. This result suggests that this assay is compatible with sample pooling.

Detection of various point mutations in p53 and VHL genes

To study the versatility of Tma EndoV in scanning various point mutations, we applied this technique to both tumor and germline samples containing different known point mutations in p53 and VHL (Figure 5 and Table 3). These genes and exons were chosen because mutant sequences are difficult to detect (Sgambati et al., 2000; Ahrendt et al., 1999). Addition of ligase significantly reduced non-specific cleavage background (compare p53 mutation detection in Figure 5, lanes 13-18 with 19-24). EndoV/Ligase mutation scanning detected mutations in 9/11 of the p53 samples and 12/13 of the VHL samples. The refractory sequences were transition mutations in GC-rich DNA (gRcg and rcRc, where R=purine, and the underlined base is the position of the mutation). The 500 bp VHL exon 1 fragment did not amplify efficiently with the standard PCR methodology, possibly due to the very high GC content in the middle of this exon. Therefore, we diluted the PCR products 100-fold and used 1 mul of this diluted reaction as template for a second round of amplification under the same PCR conditions, but for only 15 cycles. This resulted in efficient amplification and detection of the point mutations. Overall, with the exception of a few refractory sequences, the scanning assay detected point mutations in GC-rich DNA.

Detection of the Tright arrowA mutation in codon 1307 of the APC gene

To validate the utility of this assay on mononucleotide repeats, we applied this method to germline samples scanning the APC I1307K mutation, which creates an A8 mononucleotide repeat (Laken et al., 1997). The DNA mutations in these samples were previously verified by PCR/LDR (Zirvi et al., 1999). Nine samples from patients carrying the mutation and four from normal individuals were assayed. Mutations in samples from all of the carriers were detected, with no false positives detected in the normal samples (data not shown). It was found that cleavage activity is higher in the presence of 1.0 M betaine instead of 1.5 M betaine (data not shown). This is presumably due to the higher AT content of exon 15 in APC than previous genes we assayed (see Discussion). Therefore, the optimum amount of betaine added to a reaction mixture may vary depending on the AT content of the PCR fragment being amplified.

Detection of small deletion and insertion mutations in BRCA1, BRCA2 and VHL genes

In order to study the utility of this assay to detect small deletions or insertions, segments of BRCA1 and BRCA2, VHL genes containing 1-3 nt insertions/deletions were amplified. Cleavage of single-base (five cases including p53 samples in Table 4 below), two-base (two cases) or three-base (one case) insertions or deletions generally gave robust signal, with the exception of the single-base insertion and deletion found in BRCA1 and BRCA2 (Figure 5 and Table 3). In the case of BRCA2 6174delT, addition of ligase eliminates a background band that is stronger than the correct cleavage product at the mismatch (compare lanes 9 and 10, with the incorrect band at 140 bp with lanes 11 and 12 showing only the correct band at 151 bp). For the BRCA1 185delAG mutation, multiple strong bands on both strands are observed presumably reflecting alternate structures of the heteroduplexed substrates (lanes 2 and 4). Thus, insertions and deletions that commonly create frameshift mutations are distinguished by the EndoV/ligase scanning assay.

Detection of p53 mutations in colon tumor samples

An analysis of p53 mutations in DNA isolated from 50 Stage I-IV colon adenocarcinomas using a combination of both PCR/LDR/Universal array and EndoV/Ligase mutation scanning proved more accurate than automated sequencing alone (see Table 4, Favis et al., unpublished work; Favis et al., 2000; Gerry et al., 1999). Of the 23 samples, 26 p53 mutations were identified by PCR/LDR combined with EndoV/ligase, while eight samples were missed by automated sequencing (65% true positive, 35% false negative). Confirmation of the exact sequence in the samples called positive with EndoV/Ligase mutation scanning, required gel purification of some PCR fragments, and re-sequencing of both strands with manual reading. Significantly, EndoV/Ligase mutation scanning scored all four frameshift mutations, which accounts for almost 20% of the samples with p53 mutations. Such frameshift mutations are beyond the detection capacity of commercially available p53 hybridization chips (Ahrendt et al., 1999).

The above samples were pooled in subsets to test the ability of the assay in a mixed sample format. These tumor samples were frozen directly in liquid nitrogen and tissue sections were not microdissected prior to extraction of nucleic acids. Microscopic examination of sections revealed stromal cell infiltration ranging from about 10-50%. Mutation of one p53 allele is invariably accompanied by loss of heterozygosity (LOH) of the other allele, occurring either through partial to complete chromosome loss or through mitotic nondisjunction (i.e. both mutant chromosomes migrate to one daughter cell) (Thiagalingam et al., 2001). Thus, DNA from the above samples would have a range of mutant p53 : wild-type allele of from about 90 : 10% (for non-disjunction with 10% stroma) to 33 : 67% (for chromosome loss with 50% stroma). When the above samples were re-analysed for p53 mutations in pools of three samples together, the EndoV/ligase mutation scanning assay could still distinguish the presence of all the mutants. Several of the pooled sample bands were even stronger than from individual samples (suggesting the original sample was mostly mutant DNA), while a few bands were weaker but could still be detected (suggesting the original sample had substantial stromal contamination). The ability to detect mutations in pools of five or even 10 samples demonstrates the sensitivity of the assay. This high sensitivity is potentially useful in routine analysis of individual clinical samples in which relatively few neoplastic cells are present.

Discussion

Mutation scanning requires robust techniques with high specificity and sensitivity. Current enzymatic methods often suffer from low specificity or low sensitivity (Taylor, 1999). Endonuclease V exhibits broad mismatch cleavage activities and generates a ligatable nick at the 3' penultimate position (Yao and Kow, 1994). DNA ligases vary substantially as to their mismatch ligation fidelity, but a thermostable ligase from Thermus species AK16D demonstrates superior fidelity in sealing a matched nick at the 3' penultimate position (Tong et al., 1999). By combining the cleavage/ligation activities of these two enzymes, we developed a mutation scanning method where ligase proofreading reduced background and increased sensitivity.

The utility and versatility of this assay has been demonstrated on over 80 samples containing known and unknown SNPs, or mutations in p53, VHL, K-ras, APC, BRCA1, and BRCA2 (see Tables 3 and 4). Most point mutations (31/35 unique mutations) and all frameshift mutations (8/8 unique insertions or deletions) could be detected with this method (Figures 2 and 5, Tables 3 and 4). The four refractory mutations, K-ras G13D(Gright arrowA), p53 R175H(Gright arrowA), R273C(Cright arrowT) and VHL G164D(Gright arrowA) are all transition mutations of the form gRcg and rcRc (Table 3). These refractory polymorphisms appear on average only 2% of the time in human DNA, as judged by their frequency in 6000 random SNPs culled from the public SNP database. Further, combining PCR/LDR detection of the known refractory sequences with EndoV/Ligase mutation scanning can provide essentially complete coverage of all point and frameshift mutations in a given gene (Table 4).

For detectable mutations resulting from a transition, the purine strand at the mismatch is always cleaved, while the pyrimidine strand is generally not cleaved. Of the 22 transition mutations assayed, in only six cases were both strands cleaved (Table 3). Among the seven transversions assayed, both the top and bottom strands containing both purine and pyrimidine bases were cleaved for all but one case studied. (The top strand of F190L produced no observable product exactly 56 bp in length, but there were some non-specific bands around this position.) We did not distinguish between purine and pyrimidine strand cleavage for transversions. These results follow the general hierarchy for Tma EndoV cleavage when using synthetic oligonucleotides as substrates (Table 1, Huang et al., 2001). The intensity of cleavage of one strand follows A=G>T>C, in which G, A, T and C indicates the base in the mismatch (Tables 1 and 3). Tma EndoV cleavage of the strand containing C in the mismatch was usually very poor and sometimes not observed at all.

The cleavage activity of Escherichia coli (E.coli) EndoV is affected by the sequence content surrounding the mismatch. The presence of two G/C pairs, one 5' and the other 3' immediately adjacent to the mismatch abolishes the ability of the enzyme to cleave the mismatch (Yao and Kow, 1994). It appears that Tma EndoV cleavage activity is not significantly constrained by this neighbor effect. We assayed 14 mutations flanked with two G/C pairs, one 5' and the other 3' to the mismatch. Of these, Tma EndoV cleaved 10 mutations, often giving robust signal. Besides the structural differences between the two EndoV proteins, assaying at higher temperature (65°C) and the addition of betaine or DMSO to the reaction buffer may have facilitated Tma EndoV cleavage of DNA containing mismatches flanked by two G/C pairs. Betaine has been used to equalize the melting temperatures of DNA fragments having different GC content (Rees et al., 1993), as well as an additive to facilitate the PCR amplification of DNA regions with high GC content (Henke et al., 1997). Knowledge of the AT content of the fragment should guide selection of a betaine concentration that provides optimal cleavage while minimizing background.

A two or three base deletion/insertion resulted in cleavage products with very strong signal for both strands (BRCA1 185 delAG and VHL S200 insAGA), and a strong single band in one case (p53 Q167 delGT). An A deletion in VHL also generated a strong signal for both strands. An A or T deletion in p53 generated a moderate signal for one strand, while a C insertion in BRCA1 and a T deletion in BRCA2 only resulted in a weak signal for one strand, and no signal for the other. These results suggest that hybrids with multiple base insertions/deletions are good substrates for EndoV, whereas one base insertions or deletions are less efficiently cleaved and surrounding sequence may have an effect. Nevertheless, EndoV/ligase scanning identified all the frameshift mutations tested.

With this method we can detect a mutation in PCR fragments as long as 1.7 kb. In addition to the known mutation, multiple cleavage bands were observed even after the ligation step. Since these additional bands are also present in one but not the other wild-type DNA samples (WT1, Figure 3), this result is consistent with additional polymorphic sites being present in the tumor sample. These combined results indicate that this assay can detect a mutation in fragments up to 1.7 kb in length, and suggests that multiple mutations can be identified in a single amplicon. In addition, dilution experiments with K-ras and pooling experiments with p53 from tumor samples demonstrate that this assay is amendable to pooling. Therefore, the throughput of EndoV/ligase mutation scanning has the potential of being increased by both scanning larger regions of DNA in a single reaction as well as scanning multiple samples within the same reaction.

Tests for germline BRCA1, BRCA2, APC, and VHL mutations provide highly accurate disease diagnosis or prediction of future disease (Abeliovich et al., 1997; Beller et al., 1997; Berman et al., 1996; Laken et al., 1997; Oddoux et al., 1996; Roa et al., 1996; Struewing et al., 1995, 1997; Chen et al., 1996; Stolle et al., 1998). For example, these DNA diagnostic tests enable physicians to reassure patients with an initial incorrect clinical diagnosis of von Hippel-Lindau disease, to provide molecular, and thus more accurate diagnosis of von Hippel-Lindau disease, and to guide clinical management of members of families affected with VHL. Tests for somatic VHL mutations may be useful in classifying sporadic renal cancers. Although we encountered one refractory mutation using EndoV/Ligase scanning, this could be detected in a parallel PCR/LDR test.

The EndoV/Ligase mutation scanning method may become clinically useful for high throughput detection of somatic p53 mutation in tumors. To date, various technologies used for pre-screening or screening are not sensitive enough, nor do they provide information about the precise p53 mutation (Soussi and Béroud, 2001). Consequently, the majority of studies of mutational status of p53 were performed on exons 5 through 8 in the common belief that only a few per cent of mutations were missed. In a recent study, we have demonstrated that more than 15% of mutations are found outside exons 5-8 (Soussi and Béroud, 2001). The most frequent missing mutations are located in exons 4, 10, and 9, respectively. Furthermore, the pattern of these mutations is different from those found in exon 5-8, suggesting that they could have an important impact on prognosis (Soussi and Béroud, 2001). Screening exon 4 to 10 with the EndoV/Ligase mutation scanning assay should lead to an improvement in the sensitivity of p53 mutation detection. This may be further enhanced by performing the EndoV/Ligase mutation scanning assay on p53 cDNA.

SNPs may serve as markers of disease, or may associate directly with the disease gene (Risch and Merikangas, 1996). Identification of SNPs and mutations in large stretches of DNA and in a large number of samples requires novel methods to replace classical ones, which were not designed for high-throughput screening. In a recent survey to discover new SNPs in 106 genes with 114 independent alleles, VDA was used alone or in combination with DHPLC. The approach was shown to be systematic and comprehensive with a discordancy rate of 21% compared to direct DNA sequencing (Cargill et al., 1999; Halushka et al., 1999). A larger scale survey of all exons and flanking intronic regions of 313 human genes identified 3899 SNPs in 82 samples using automated DNA sequencing (Stephens et al., 2001). Although these approaches are ideal when surveying a broad spectrum of the population for common variants or cosmopolitan SNPs (Stephens et al., 2001), identifying low frequency disease causing mutations with high penetrance will require screening dozens of candidate genes in hundreds of affected individuals and matched controls. Endo V/ligase mutation scanning supports fragment pooling 5-10-fold, as well as the ability to distinguish mutations in large fragments that would normally require four or more sequencing runs. This translates into a potential 10-40-fold increase in throughput. Confirming the precise nature of newly identified mutations can be achieved using automated sequencing, while the subsequent large-scale clinical studies to validate disease association can be facilitated by PCR/LDR/Universal Array (Gerry et al., 1999; Favis et al., 2000).

Materials and methods

Materials

All routine chemical reagents were purchased from Sigma Chemicals (St. Louis, MO, USA) or Fisher Scientific (Fair Lawn, NJ, USA). DNA sequencing kits, GeneScan-500 (TAMRA) Size Standard, and PCR kits were purchased from Applied Biosystems Division of Perkin-Elmer Corporation (Foster City, CA, USA). Deoxyribonucleoside triphosphate (dNTPs), bovine serum albumin (BSA), ATP, 7-deaza-dGTP were purchased from Boehringer-Mannheim (Indianapolis, IN, USA). TaqPlus Precision PCR kit was purchased from Stratagene (La Jolla, CA, USA). Proteinase K was purchased from QIAGEN (Valencia, CA, USA). Protein assay kit was from Bio-Rad (Hercules, CA, USA). Deoxyoligonucleotides were ordered from Integrated DNA Technologies Inc. (Coralville, IA, USA). Microcon 30 filters were purchased from Millipore (Bedford, MA, USA). DNA polymerase I (Klenow fragment) was purchased from New England Biolab (Beverly, MA, USA). Sep-Pak Cartridge C-18 was purchased from Waters (Milford, MA, USA). Centri-SePTM spin columns were purchased from Princeton Separation (Adelphia, NJ, USA). Thermotoga maritima Endonuclease V and Thermus species AK16D DNA ligase were purified as described (Huang et al., 2001; Tong et al., 1999).

For detecting K-ras mutations, genomic DNA was extracted from cell lines as described (Khanna et al., 1999). Cell lines HT29 and SW1417 contain the wild-type K-ras gene, while SW620 and SW480 contain the G12V (Gright arrowT) mutation. The ratio of wild-type to mutant sequence in the genomic DNA extracted from cell lines was determined using PCR/LDR: LS180,1 : 1.8 (wt: G12D, Gright arrowA); SW1116, 1 : 0.7 (wt: G12A, Gright arrowC) and HCT15, 1 : 1.1 (wt: G13D, Gright arrowA). Genomic DNA containing germline mutations, or tumor DNA containing sporadic mutations was extracted as described (Khanna et al., 1999). A blinded set of 50 adenocarcinomas comprised of a mix of stages I, II, III and IV were evaluated for p53 mutations using both PCR/LDR Universal Array (Favis et al., unpublished results) and EndoV/Ligase mutation scanning (this work). Tumors were frozen in liquid nitrogen directly after surgical resection. The tissue was not microdissected, i.e. five core samples were punched from various positions in the tumor and subjected to DNA extraction as described above.

PCR amplification

DNA sequences of PCR primers used in this study are listed in Table 2. 5' end ((3',6'-dippivaloylfluoresceinyl)-6-carboxamidohexyl)-1-O-(2-cyanoethyl)-(N,N-diisopropyl)-phosphoramidite (6-FAM) or 4,7,2',7'-tetrachloro-(3',6'-dippivaloylfluoresceinyl)-6-carboxamidohexyl-1-O-(2-cyanoethyl)-(N,N-diisopropyl)-phosphoramidite (TET) labeled PCR primers were purified by separation on a 10% denaturing polyacrylamide gel (7M urea) as described (Applied Biosystems Inc. FC, CA, USA in 1992). PCR reactions (50 mul) contained 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 200 muM of each dNTP, 0.2 muM of each primer pair, 2.5 mM MgCl2, 2.5 unit AmpliTaq DNA polymerase or AmpliTaq Gold DNA polymerase, and 100 ng genomic DNA. In the case of VHL exon 1, which has a high GC content, 2% dimethyl sulfoxide (DMSO) was included in the PCR reaction mixture. Addition of DMSO alone has been shown to improve amplification of some GC-rich sequences by disrupting base-paring (Pomp and Medrano, 1991). PCR amplification conditions for each fragment were as follows (gene, exon, polymerase, denaturation, cycles, cycling conditions): (K-ras, exon 1, AmpliTaq DNA polymerase, 94°C for 2 min, 30 cycles, 94°C for 15 s, 60°C for 2 min), (BRCA 1 exon 2 and exon 20, BRCA 2 exon 11 and p53 exon 7, AmpliTaq Gold DNA polymerase, 95°C for 10 min, 35 cycles, 94°C 30 s, 60°C for 30 s, 72°C for 1 min), (APC exon 15, AmpliTaq Gold DNA polymerase, 95°C for 10 min, 30 cycles, 94°C for 30 s, 63°C for 75 s), (VHL exon 1 and 3, AmpliTaq DNA polymerase, 95°C for 2 min, 35 cycles, 94°C for 20 s, 66°C for 30 s, 72°C for 1 min), (VHL exon 2, AmpliTaq DNA polymerase, 95°C for 2 min, 35 cycles, 94°C for 20 s, 60°C for 30 s, 72°C for 1 min), (p53 exons 5, 6 and 8, AmpliTaq DNA polymerase, 95°C for 2 min, 35 cycles, 94°C for 20 s, 65°C for 2 min).

For amplification of the 1.7 kb segment in p53, the PCR mixtures (50 mul) contained 1´ TaqPlus Precision buffer, 200 muM dNTP, 0.2 muM of primers, 100 ng DNA template and 2.5 unit TaqPlus precision DNA polymerase (Stratagene, La Jolla, CA, USA). The PCR procedure included a predenaturation step at 95°C for 2 min, 35 cycles of two-step amplification with each cycle consisting of denaturation at 94°C for 20 s and annealing-extension at 68°C for 3 min. All PCR reactions above were completed with a final extension step at 72°C for 7 min.

Preparation of heteroduplexed DNA substrates

To remove Taq DNA polymerase, 4 mul of 20 mg/ml proteinase K (QIAGEN) was added to the PCR mixtures (50 mul) and incubated at 70°C for 10 min. Proteinase K was inactivated by incubating at 80°C for 10 min. Amplicons containing wild-type sequence were added in approximately equal ratios when missing from the sample (i.e. pure mutant cell line DNA). The mixed PCR fragments, were heated at 94°C for 1 min to denature the DNA, and then cooled at 65°C for 15 min and at room temperature for 15 min to allow efficient formation of heteroduplex DNA.

EndoV/ligase mutation scanning assay

The reaction mixtures (20 mul) contained 10 mM N-(2-Hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) (HEPES) (pH 7.4), 1 mM dithiothreitol (DTT), 2% glycerol, 5 mM MgCl2, 5% DMSO, 1-1.5 M N,N,N-trimethylglycine (betaine), 100 ng heteroduplexed PCR products, and 500 nM purified Tma Endonuclease V. The reaction mixtures were incubated at 65°C for 1 h. Fifteen mul of reaction mixtures from EndoV cleavage were supplemented with 2 mul of 10´supplemental buffer (200 mM Tris-HCl, pH 8.5, 12.5 mM MgCl2, 500 mM KCl, 100 mM DTT and 200 mug/ml BSA), 1 mul of 20 mM NAD+, and 2 mul of 60 nM Tsp. AK16D DNA ligase. The resulting mixtures were incubated at 65°C for 20 min and terminated by adding equal volumes of GeneScan stop solution (50 mM ethylenediaminetertraacetic acid (EDTA), 1% blue dextran and 80% formamide). After denaturing at 94°C for 1 min, 3 mul of the mixtures were loaded onto a 6% acrylamide/Bisacrylamide (19 : 1) 0.2 mm thick denaturing gel containing 6M urea and electrophoresed for 1 h at 1000 V in TBE buffer at 45°C in an ABI-377 sequencer. Results were analysed using the Genescan program (Perkin Elmer).

Detection of mutations in long PCR fragment

For long range mutation scanning, the PCR fragments were filtered twice with Microcon 30 filters (10 mM Tris pH 7.5), prior to the Endo V cleavage reactions. This removed excess dNTP and primers while concentrating the sample twofold. After Endo V reactions, the fragments were washed in the same fashion to remove betaine and DMSO. The ligation reactions were carried out at 65°C for 20 min in the presence of 6 nM Tsp.AK16D ligase in the buffer containing 20 mM Tris-HCl (pH 8.5), 5 mM MgCl2, 50 mM KCl, 10 mM DTT, 1 mM NAD+ and 20 mug/ml BSA.

Acknowledgements

We thank Berton Zbar, Alan Friedman, Donald Bergstrom, Arnold Levine, David Thaler, Richard Cunningham, Daniel Notterman, Ken Offit and Alvaro Monteiro for helpful discussion and Norman Gerry, Carrie Shawber, Jing Lu, Jie Tong and Andrew Grace for their suggestions and technical assistance. We thank Yoke Kow for providing E.coli Endo V and helpful suggestions. We thank Laura Schmidt of the National Cancer Institute-Frederick Cancer Center for providing germline samples containing VHL mutations. We thank Harry Ostrer and Carol Oddoux from the Human Genetics Program at New York University Medical College for providing blinded genomic DNA samples from Ashkenazi Jewish individuals previously typed for the APCI1307K allele. We thank Steven Narod of the University of Toronto for providing samples containing the Ashkenazi Jewish BRCA1 and BRCA2 founder mutations. This work was supported by grants from the National Cancer Institute (P01-CA65930 and RO1-CA81467).

References

Abeliovich D, Kaduri L, Lerer I, Weinberg N, Amir G, Sagi M, Zlotogora J, Heching N, Peretz T. (1997). Am. J. Hum. Genet., 60: 505-514. MEDLINE

Ahrendt S, Halachmi S, Chow J, Wu L, Halachmi N, Yang S, Wehage S, Jen J, Sidransky D. (1999). Proc. Natl. Acad. Sci. USA, 96: 7382-7387. Article MEDLINE

Aurelio ON, Cajot JF, Hua ML, Khwaja Z, Stanbridge EJ. (1998). Cancer Res., 58: 2190-2195. MEDLINE

Aurelio ON, Kong XT, Gupta S, Stanbridge EJ. (2000). Mol. Cell. Biol., 20: 770-778. MEDLINE

Beller U, Halle D, Catane R, Kaufman B, Hornreich G, Levy-Lahad E. (1997). Gynecol. Oncol., 67: 123-126. Article MEDLINE

Berman DB, Costalas J, Schultz DC, Grana G, Daly M, Godwin AK. (1996). Cancer Res., 56: 3409-3414. MEDLINE

Broll R, Stark A, Windhovel U, Best R, Strik MW, Schimmelpenning H, Schwandner O, Kujath P, Bruch HP, Duchrow M. (1999). Eur. J. Cancer, 35: 1083-1088. MEDLINE

Bunz F, Hwang PM, Torrance C, Waldman T, Zhang Y, Dillehay L, Williams J, Lengauer C, Kinzler KW, Vogelstein B. (1999). J. Clin. Invest., 104: 263-269. MEDLINE

Cargill M, Altshuler D, Ireland J, Sklar P, Ardlie K, Patil N, Lane CR, Lim EP, Kalayanaraman N, Nemesh J, Ziaugra L, Friedland L, Rolfe A, Warrington J, Lipshutz R, Daley GQ, Lander ES. (1999). Nat. Genet., 22: 231-238. Article MEDLINE

Chen F, Slife L, Kishida T, Mulvihill J, Tisherman SE, Zbar B. (1996). J. Med. Genet., 33: 716-771. MEDLINE

Chen J, Thilly WG. (1994). Environ. Health Perspect., 3: 227-229.

Cotton RGH, Rodrigues NR, Campbell RD. (1988). Proc. Natl. Acad. Sci. USA, 85: 4397-4401. MEDLINE

Dameron KM, Volpert OV, Tainsky MA, Bouck N. (1994). Science, 265: 1582-1584. MEDLINE

Fahy E, Nazarbaghi R, Zomorrodi M, Herrnstadt C, Parker WD, Davis RE, Ghosh SS. (1997). Nucleic Acids Res., 25: 3102-3109. MEDLINE

Favis R, Day JP, Gerry NP, Phelan C, Narod S, Barany F. (2000). Nat. Biotechnol., 18: 561-564. Article MEDLINE

Fodde R, Losekoot M. (1994). Hum. Mutat., 3: 83-94. MEDLINE

Foster BA, Coffey HA, Morin MJ, Rastinejad F. (1999). Science, 286: 2507-2510. Article MEDLINE

Gerry NP, Witowski NE, Day J, Hammer RP, Barany G, Barany F. (1999). J. Mol. Biol., 292: 251-262. Article MEDLINE

Giunta C, Youil R, Venter D, Chow CW, Somers G, Lafferty A, Kemper B, Cotton RG. (1996). Diagn. Mol. Pathol., 5: 265-270. Article MEDLINE

Guldberg P, Guttler F. (1994). Nucleic Acids Res., 22: 880-881. MEDLINE

Hacia J. (1999). Nat. Genet., ((Supplement)) 21: 42-47. Article MEDLINE

Hacia JG, Brody LC, Chee MS, Fodor SP, Collins FS. (1996). Nat. Genet., 14: 441-447. MEDLINE

Halushka MK, Fan JB, Bentley K, Hsie L, Shen N, Weder A, Cooper R, Lipshutz R, Chakravarti A. (1999). Nat. Genet., 22: 239-247. Article MEDLINE

Hansen LL, Justesen J, Kruse TA. (1995). PCR Primer: a Laboratory Manual Diefenbach CW and Dveksler GS Cold Spring Harbor Laboratory Press: New York, pp 275-286.

Haris II, Green PM, Bentley DR, Giannelli F. (1994). Pcr Methods Appl., 3: 268-271. MEDLINE

Hayashi K. (1991). PCR Methods Appl., 1: 34-38. MEDLINE

Heide I, Thiede C, Sonntag T, de Kant E, Neubauer A, Jonas S, Peter FJ, Neuhaus P, Herrmann R, Huhn D, Rochlitz CF. (1997). Eur. J. Cancer, 33: 1314-1322. MEDLINE

Henke W, Herdel K, Jung K, Schnorr D, Loening SA. (1997). Nucl. Acids Res., 25: 3957-3958.

Huang J, Lu J, Barany F, Cao W. (2001). Biochemistry, 40: 8738-8748. Article MEDLINE

Innis MA, Myambo KB, Gelfand DH, Brow MAD. (1988). Proc. Natl. Acad. Sci. USA, 85: 9436-9440. MEDLINE

Khanna M, Park P, Zirvi M, Cao W, Picon A, Day J, Paty P, Barany F. (1999). Oncogene, 18: 27-38. MEDLINE

Khrapko K, Hanekamp JS, Thilly WG, Belenkii A, Foret F, Karger BL. (1994). Nucl. Acids Res., 22: 364-369.

Korn SH, Moerkerk PT, de Goeij AF. (1993). J. Clin. Pathol., 46: 621-623. MEDLINE

Laken SJ, Petersen GM, Gruber SB, Oddoux C, Ostrer H, Giardiello FM, Hamilton SR, Hampel H, Markowitz A, Klimstra D, Jhanwar S, Winawer S, Offit K, Luce MC, Kinzler KW, Vogelstein B. (1997). Nat. Genet., 17: 79-83. MEDLINE

Libutti SK, Choyke PL, Alexander HR, Glenn G, Bartlett DL, Zbar B, Lubensky I, McKee SA, Maher ER, Linehan WM, Walther MM. (2000). Surgery, 128: 1022-1027, discussion 1027-1028. Article MEDLINE

Liu J, He B, Qing H, Kow YW. (2000). Mutat. Res., 461: 169-177. MEDLINE

Liu Q, Sommer SS. (1995). Biotechniques, 18: 470-477. MEDLINE

Makino R, Yazyu H, Kishimoto Y, Sekiya T, Hayashi K. (1992). PCR Methods Appl., 2: 10-13. MEDLINE

Mohr VH, Vortmeyer AO, Zhuang Z, Libutti SK, Walther MM, Choyke PL, Zbar B, Linehan WM, Lubensky IA. (2000). Am. J. Pathol., 157: 1615-1621. MEDLINE

Myers RM, Larin Z, Maniatis T. (1985). Science, 230: 1242-1246. MEDLINE

Oddoux C, Struewing JP, Clayton CM, Neuhausen S, Brody LC, Kaback M, Haas B, Norton L, Borgen P, Jhanwar S, Goldgar D, Ostrer H, Offit K. (1996). Nat Genet, 14: 188-190. MEDLINE

Oleykowski CA, Bronson Mullins CR, Godwin AK, Yeung AT. (1998). Nucleic Acids Res., 26: 4597-4602. Article MEDLINE

Perucho M, Forrester K, Almoguera C, Kahn S, Lama C, Shibata D, Arnheim N, Grizzle WE. (1989). Cancer Cells, 7: 137-141.

Pomp D, Medrano JF. (1991). Biotechniques, 10: 58-59. MEDLINE

Prives C, Hall PA. (1999). J. Pathol., 187: 112-126. Article MEDLINE

Rees WA, Yager TD, Korte J, von Hippel PH. (1993). Biochemistry, 32: 137-144. MEDLINE

Ridanpaa M, Burvall K, Zhang LH, Husgafvel-Pursiainen K, Onfelt A. (1995). Mutat. Res., 334: 357-364. MEDLINE

Ridanpaa M, Husgafvel-Pursiainen K. (1993). Hum. Mol. Genet., 2: 639-644. MEDLINE

Risch N, Merikangas K. (1996). Science, 273: 1516-1517. MEDLINE

Roa BB, Boyd AA, Volcik K, Richards CS. (1996). Nat. Genet., 14: 185-187. MEDLINE

Sarkar G, Yoon HS, Sommer SS. (1992). Genomics, 13: 441-443. MEDLINE

Sgambati MT, Stolle C, Choyke PL, Walther MM, Zbar B, Linehan WM, Glenn GM. (2000). Am. J. Hum. Genet., 66: 84-91. MEDLINE

Soussi T, Béroud C. (2001). Cancer, 1: 233-240. MEDLINE

Stephens JC, Schneider JA, Tanguay DA, Choi J, Acharya T, Stanley SE, Jiang R, Messer CJ, Chew A, Han JH, Duan J, Carr JL, Lee MS, Koshy B, Kumar AM, Zhang G, Newell WR, Windemuth A, Xu C, Kalbfleisch TS, Shaner SL, Arnold K, Schulz V, Drysdale CM, Nandabalan K, Judson RS, Ruano G, Vovis GF. (2001). Science, 293: 489-493. Article MEDLINE

Stolle C, Glenn G, Zbar B, Humphrey JS, Choyke P, Walther M, Pack S, Hurley K, Andrey C, Klausner R, Linehan WM. (1998). Hum. Mutat., 12: 417-423. Article MEDLINE

Struewing JP, Abeliovich D, Peretz T, Avishai N, Kaback MM, Collins FS, Brody LC. (1995). Nat. Genet., 11: 198-200. MEDLINE

Struewing JP, Hartge P, Wacholder S, Baker SM, Berlin M, McAdams M, Timmerman MM, Brody LC, Tucker MA. (1997). N. Engl. J. Med., 336: 1401-1408. MEDLINE

Suzuki Y, Orita M, Shiraishi M, Hayashi K, Sekiya T. (1990). Oncogene, 5: 1037-1043. MEDLINE

Tabor S, Richardson CC. (1995). Proc. Natl. Acad. Sci. USA, 92: 6339-6343. MEDLINE

Taylor GR. (1999). Electrophoresis, 20: 1125-1130. Article MEDLINE

Thiagalingam S, Laken S, Willson JK, Markowitz SD, Kinzler KW, Vogelstein B, Lengauer C. (2001). Proc. Natl. Acad. Sci. USA, 98: 2698-2702. Article MEDLINE

Tong J, Cao W, Barany F. (1999). Nucl. Acids Res., 27: 788-794.

Tortola S, Marcuello E, Gonzalez I, Reyes G, Arribas R, Aiza G, Sancho FJ, Peinado MA, Capella G. (1999). J. Clin. Oncol., 17: 1375-1381. MEDLINE

Underhill PA, Jin L, Zemans R, Oefner PJ, Cavalli-Sforza LL. (1996). Proc. Natl. Acad. Sci. USA, 93: 196-200. Article MEDLINE

Walther MM, Reiter R, Keiser HR, Choyke PL, Venzon D, Hurley K, Gnarra JR, Reynolds JC, Glenn GM, Zbar B, Linehan WM. (1999). J. Urol., 162: 659-664. MEDLINE

Wang DG, Fan JB, Siao CJ, Berno A, Young P, Sapolsky R, Ghandour G, Perkins N, Winchester E, Spencer J, Kruglyak L, Stein L, Hsie L, Topaloglou T, Hubbell E, Robinson E, Mittmann M, Morris MS, Shen N, Kilburn D, Rioux J, Nusbaum C, Rozen S, Hudson TJ, Lander ES. (1998a). Science, 280: 1077-1082. Article MEDLINE

Wang Y, Blandino G, Oren M, Givol D. (1998b). Oncogene, 17: 1923-1930. MEDLINE

Webley KM, Shorthouse AJ, Royds JA. (2000). J. Pathol., 191: 361-367. Article MEDLINE

Winter E, Yamamoto F, Almoguera C, Perucho M. (1985). Proc. Natl. Acad. Sci. USA, 82: 7575-7579. MEDLINE

Xu JF, Yang QP, Chen JY, van Baalen MR, Hsu IC. (1996). Carcinogenesis, 17: 321-326. MEDLINE

Yan H, Kinzler KW, Vogelstein B. (2000a). Science, 289: 1890-1892. Article MEDLINE

Yan H, Papadopoulos N, Marra G, Perrera C, Jiricny J, Boland CR, Lynch HT, Chadwick RB, de la Chapelle A, Berg K, Eshleman JR, Yuan W, Markowitz S, Laken SJ, Lengauer C, Kinzler KW, Vogelstein B. (2000b). Nature, 403: 723-724. Article MEDLINE

Yao M, Hatahet Z, Melamede RJ, Kow YW. (1994). Ann. NY Acad. Sci., 726: 315-316. MEDLINE

Yao M, Kow YW. (1994). J. Biol. Chem., 269: 31390-31396. MEDLINE

Youil R, Kemper BW, Cotton RG. (1995). Proc. Natl. Acad. Sci. USA, 92: 87-91. MEDLINE

Zirvi M, Nakayama T, Newman G, McCaffrey T, Paty P, Barany F. (1999). Nucl. Acids Res., 27: 40-.

Zou Z, Gao C, Nagaich AK, Connell T, Saito S, Moul JW, Seth P, Appella E, Srivastava S. (2000). J. Biol. Chem., 275: 6051-6054. Article MEDLINE

Figures

Figure 1 Endonuclease V/DNA ligase mismatch scanning assay for scoring unknown mutations. Heteroduplexes are formed from PCR amplicons containing both normal and variant sequence. Mixed amplicons of variant sequence can be obtained by amplifying heterozygous germline samples, or from tumor samples where stromal cell contamination provides sufficient amounts of wild-type DNA, or by mixing the PCR products from unknown and wild-type samples in a 1 : 1 ratio. Tma EndoV nicks DNA one base 3' to the mismatch (big triangle), and it can also generate non-specific nicks in homoduplex DNA with minor activity (small triangle). Addition of Thermus sp. AK16D DNA ligase (solid circle) seals these non-specific nicks, providing a proofreading mechanism to improve signal-to-noise. Both top and bottom strand PCR primers are 5' end-labeled with different fluorescent dyes (6-FAM and TET, respectively) allowing for cleavage products to be distinguished on a denaturing polyacrylamide gel. The approximate position of the mutation can be determined from the resultant fragment lengths

Figure 2 Detection of mutations in K-ras gene codon 12 using Tma EndoV and Tsp. AK 16D DNA ligase. PCR fragments of K-ras exon 1 from genomic DNA of pure wild-type (G12) and pure mutant (G12V) were used as controls. The mixtures of PCR fragments of wild-type and pure mutant (G12V)(1 : 1 ratio) or wild-type and pure mutant (G12D)(1 : 1 ratio) were used as DNA substrate samples in the Tma EndoV cleavage assay. DNA duplexes (100 ng) were cleaved in a buffer containing 500 nM Tma EndoV and other reagents as described in Materials and methods. Increasing amounts of Tsp. AK16D DNA ligase were added to seal the nicks generated by non-specific cleavage, as described in Materials and methods. The ligation mixtures were electrophoresed in a 6% denaturing polyacrylamide gel. Top and bottom strand PCR primers were 5' end-labeled with TET and 6-FAM (showing green and blue, respectively, using filter C setting on an ABI 377 DNA sequencer). The green band was generated from cleavage of the top strand, and the blue band was generated from the cleavage of the bottom strand

Figure 3 Using Tma EndoV and Tsp. AK16D DNA ligase to detect the p53 R248 Cright arrowT mutation in a 1775 bp PCR fragment. Wild-type PCR fragments from two different individuals, WT1 and WT2, were used as normal controls. Mutant PCR fragments containing R248 Cright arrowT mutations from two different tumors were used as samples in the mutation scanning assay. The reaction mixtures were electrophoresed in a 6% denaturing polyacrylamide gel. Lanes 1 to 4 without ligase, and lanes 5 to 8, with 6 nM Tsp. AK16D DNA ligase. M1: GeneScan 2500 molecular standard, M2: GeneScan 500 molecular standard

Figure 4 Sensitivity of the Endo V/DNA ligase reaction. Wild-type and mutant K-ras exon 1 was PCR amplified using genomic DNA from the following cell lines: SW480 (pure G12V mutant), LS180 (1.8 pure G12D mutant : 1 wild-type), SW1116 (0.7 pure G12A mutant : 1 wild-type) and HT29 (pure wild-type). Mutant amplicons were diluted with wild-type amplicons to produce mutant : wild-type ratios ranging from 1 : 1 to 1 : 100 (note that for G12A, the actual value of the first ratio in the graph is 0.7 : 1). The heteroduplex PCR fragments were cleaved by Tma EndoV and ligated with Tsp. AK16D. The signals that migrated similarly to cleavage products from the wild-type homoduplex were used as wild-type (wt) controls. Two background peaks located near either the top or bottom strand cleavage products were used as background signals to compare signal and noise for the appropriate cleavage product. The average background signal (represented by dashed lines) was calculated as the mean value of background signals from seven samples with different mutant : wild-type ratios and one wild-type sample. Bar graphs indicate the relative fluorescence intensity of top strand (striped bars) and bottom strand (solid bars) cleavage products for each mutant : wild-type ratio. The relative fluorescence intensity is defined as the area under a signal's peak as determined by GeneScan analysis software. Bkgd-top: average background signals for the top strand cleavage products; bkgd-bot: average background signals for the bottom strand cleavage products. The mutation, nucleotide change, and the mismatched base pairs are indicated for each panel. The position of each mutation is underlined in the mismatch base pairing

Figure 5 Detection of small insertions/deletions in BRCA1, BRCA2 and point mutations in p53. Lanes 1-12 show Endo V/Ligase reactions that detect the three founder mutations in BRCA1 and BRCA2. Lanes 13-24 show assays for mutations in exon 7 of p53. The results of incubation with and without DNA ligase are shown. Mutation tested and presence or absence of ligase are indicated above gel lanes. To estimate site of cleavage, forward and reverse PCR primers were 5' end labeled with TET and 6-FAM, respectively. Multiple cleavage bands with 185 delAG mutation reflects different conformations of heteroduplexed DNA. Arrows indicate positions of bands of interest

Tables

Table 1 Summary of Tma endonclease V cleavage of heteroduplexed synthetic single-base mismatched substratesa

Table 2 PCR primers to amplify exons of the K-ras, APC, p53, BRCA1, BRCA2, and VHL genes

Table 3 Summary of Tma EndoV/Ligase mutation scanning of cancer genes

Table 4 Detecting p53 mutations in tumor samples: Comparisons of using a combined analysis of PCR/LDR and Endo V/ligase mutation scanning to Dideoxysequencing

Received 13 August 2001; revised 15 October 2001; accepted 29 October 2001
14 March 2002, Volume 21, Number 12, Pages 1909-1921
Table of contents    Previous  Article  Next    [PDF]