1284
Technical Briefs
Fig. 1. Results of triplex genotyping given as patterns of peaks called pyrograms (top), and genotype assignment (bottom). (Top), the x axis shows the dispensation order of dNTPs during pyrosequencing; the y axis shows peak heights representing the intensity of the light signal produced, which is proportional to the number of dNTPs incorporated into the DNA templates. (Bottom), genotypes are assigned by computer-automated comparison of peaks with predicted histograms. The individual shown in this figure was heterozygous for MTHFR C677T (u) and homozygous wild type for factor V G1691A ( ) and factor II G20210A (f).
We are grateful to Biosense for supplying the instrument and reagents and, in particular, to Dr. Elvira Meroni for excellent technical support. This work was partially supported by MURST-FIRB Grant RBAU01LSR4_001 (to F.F.).
14.
References
16.
1. Angelini A, Di Febbo C, Baccante G, Di Nisio M, Di Ilio C, Cuccurullo F, et al. Identification of three genetic risk factors for venous thrombosis using a multiplex allele-specific PCR assay: comparison of conventional and new alternative methods for the preparation of DNA from clinical samples. J Thromb Thrombolysis 2003;16:189 –93. 2. Blasczyk R, Ritter M, Thiede C, Wehling J, Hintz G, Neubauer A, et al. Simple and rapid detection of factor V Leiden by allele-specific PCR amplification. Thromb Haemost 1996;75:757–9. 3. Bortolin S, Black M, Modi H, Boszko I, Kobler D, Fieldhouse D, et al. Analytical validation of the Tag-It high-throughput microsphere-based universal array genotyping platform: application to the multiplex detection of a panel of thrombophilia-associated single-nucleotide polymorphisms. Clin Chem 2004;50:2028 –36. 4. Schrijver I, Lay M, Zehnder J. Diagnostic single nucleotide polymorphism analysis of factor V Leiden and prothrombin 20210G⬎A. A comparison of the Nanogen Eelectronic Microarray with restriction enzyme digestion and the Roche LightCycler. Am J Clin Pathol 2003;119:490 – 6. 5. Liebman H, Sutherland D, Bacon R, McGehee W. Evaluation of a tissue factor dependent factor V assay to detect factor V Leiden: demonstration of high sensitivity and specificity for a generally applicable assay for activated protein C resistance. Br J Haematol 1996;95:550 –3. 6. Patnaik M, Dlott JS, Fontaine RN, Subbiah MT, Hessner MJ, Joyner KA, et al. Detection of genomic polymorphisms associated with venous thrombosis using the Invader biplex assay. Mol Diagn 2004;6:137– 44. 7. Corral J, Iniesta J, Gonzalez-Conejero R, Vicente V. Detection of factor V Leiden from a drop of blood by PCR-SSCP. Thromb Haemost 1996;76: 735–7. 8. Zotz R, Maruhn-Debowski B, Scharf R. Mutation in the gene coding for coagulation factor V and resistance to activated protein C: detection of the genetic mutation by oligonucleotide ligation assay using a semi-automated system. Thromb Haemost 1996;76:53–5. 9. Bowen D, Standen G, Granville S, Bowley S, Wood N, Bidwell J. Genetic diagnosis of factor V Leiden using heteroduplex technology. Thromb Haemost 1997;77:119 –22. 10. Lay MJ, Wittwer CT. Real-time fluorescence genotyping of factor V Leiden during rapid-cycle PCR. Clin Chem 1997;43:2262–7. 11. Palmieri O, Toth S, Ferraris A, Andriulli A, Latiano A, Annese V, et al. CARD15 Genotyping in inflammatory bowel disease patients by multiplex pyrosequencing. Clin Chem 2003;49:1675–9. 12. Lindqvist M, Haglund S, Almer S, Peterson C, Taipalensu J, Hertervig E, et al. Identification of two novel sequence variants affecting thiopurine methyltransferase enzyme activity. Pharmacogenetics 2004;14:261–5. 13. Fakhrai-Rad H, Pourmand N, Ronaghi M. Pyrosequencing: an accurate
15.
17.
18. 19.
detection platform for single nucleotide polymorphisms. Hum Mutat 2002; 19:479 – 85. Zhang Z, Liu W, Jia X, Gao Y, Hemminki K, Lindholm B. Use of pyrosequencing to detect clinically relevant polymorphisms of genes in basal cell carcinoma. Clin Chim Acta 2004;342:137– 43. Ronaghi M, Elahi E. Pyrosequencing for microbial typing. J Chromatogr B Analyt Technol Biomed Life Sci 2002;782:67–72. Segui R, Estelles A, Mira Y, Espana F, Villa P, Falco C, et al. PAI-1 promoter 4G/5G genotype as an additional risk factor for venous thrombosis in subjects with genetic thrombophilic defects. Br J Haematol 2000;111:122– 8. Weisberg I, Tran P, Christensen B, Sibani S, Rozen R. A second genetic polymorphism in methylenetetrahydrofolate reductase (MTHFR) associated with decreased enzyme activity. Mol Genet Metab 1998;64:169 –72. Endler G, Mannhalter C. Polymorphisms in coagulation factor genes and their impact on arterial and venous thrombosis. Clin Chim Acta 2003;330:31–55. Ronaghi M, Uhlen M, Nyren P. A sequencing method based on real-time pyrophosphate. Science 1998;281:363–5. DOI: 10.1373/clinchem.2005.048124
Arrayed Primer Extension Resequencing of Mutations in the TP53 Tumor Suppressor Gene: Comparison with Denaturing HPLC and Direct Sequencing, Florence Le Calvez,1 Aune Ahman,2 Neeme Tonisson,2,3 Jeremy Lambert,1 Ste´phane Temam,4 Paul Brennan,1 David G. Zaridze,5 Andres Metspalu,2,3 and Pierre Hainaut1* (1 International Agency for Research on Cancer, Lyon, France; 2 Asper Biotech Ltd., Tartu, Estonia; 3 Institute of Molecular and Cell Biology, University of Tartu/Estonian Biocentre, Tartu, Estonia; 4 Department of Head and Neck Surgery, Institut Gustave-Roussy, Villejuif, France; 5 Institute of Carcinogenesis, Cancer Research Center, Moscow, Russia; * address correspondence to this author at: International Agency for Research on Cancer, 150, Cours Albert Thomas, F-69372 Lyon Cedex 08, France; fax 33472738322, e-mail
[email protected]) Mutations of TP53 (17p13.1; OMIM 191170; PubMed accession number X54156) are common in cancers and are
1285
Clinical Chemistry 51, No. 7, 2005
A
Homoduplexes
Heteroduplexes
B
Codon 236, TA C>TG C
⇓ 7
Absorbance (mV)
6
5
4
3
2
1 0
C
1
2
3
4
D
Codon 236, TAC>TGC A
C
G
T
Average signals
Wild-type reference
Sense Antisense
Sense Antisense
A
C
G
T
Fig. 1. Example of detection of TP53 mutation at codon 236 (TAC⬎TGC) in an archived pathology specimen. DNA extracted from a formalin-fixed, paraffin-embedded lung squamous cell carcinoma was analyzed by DHPLC and direct sequencing (A and B) or by APEX (C and D). (A), DHPLC chromatograms of exon 5– 6, showing the superposition of profiles for wild-type (blue) and tumor (red) DNA. (B), direct sequencing of exon 5– 6. A portion of the sense sequence surrounding codon 236, containing an A-to-G mutation at the second base, is shown. (C), detection of a mutation at the second base of codon 236 by APEX. The signals for all 4 fluorescent dinucleotides are shown for the sense and antisense strands. Each dinucleotide is visualized in duplicate. (D), analysis of the signals shown in A by the GENORAMA software. Left, average of the signals for each dinucleotide on the sense and antisense strands, expressed as a percentage of the highest signal detected at each position. Right, average of wild-type signals at each position. Comparison of the panels indicates the presence of a mutant, G:C, base pair against a background of the wild-type, A:T, base pair.
typically missense within exons 4 –9, impairing the capacity of p53 to transactivate genes involved in cell cycle arrest, apoptosis, and DNA repair (1 ). Functionally, mutations may differ according to their nature and position, as well as to the presence of a common polymorphism at codon 72 (arginine or a proline) in the mutant allele (2 ). Knowing TP53 mutation status has potential applications for cancer prognosis (3, 4 ) and early diagnosis (5 ), identification of mutagen “fingerprints” (1, 6 ), and prediction of therapeutic outcomes (7, 8 ). To achieve this purpose, sensitive, fast, and cost-effective methods are needed to Table 1. DHPLC and APEX detection limits for TP53 mutations in 6 cell lines. Detection limit, % mutant DNA Cell line
Exon
Codon
Mutation
DHPLC
APEX
Hs578T T47D TE11 TE6 TE1 MDA-MB 231
5 6 7 7 8 9
157 194 237 248 272 280
GTC⬎TTC CTT⬎TTT ATG⬎ATT CGG⬎CAG GTG⬎ATG AGA⬎AAA
12.5 3.125 3.125 6.25 3.125 6.25
6.25 3.125 6.25 3.125 6.25 6.25
assess the whole coding sequence plus exon/intron boundaries. Current approaches are based on mutation prescreening with single strand conformational polymorphism analysis, temporal temperature gradient electrophoresis, or denaturing HPLC (DHPLC) combined with direct sequencing of relevant PCR fragments [reviewed in Ref. (9 )]. These methods are labor-intensive, difficult to standardize, and in some cases, of limited sensitivity. In recent years, 2 microarray methods for resequencing TP53 have been described: the Affymetrix p53 GeneChip array, described elsewhere (10, 11 ), and the Arrayed Primer Extension (APEX), based on incorporation of 4 dye terminators into oligonucleotide primers that each identify a base in the target sequence (12 ). In 2002, we described an APEX array for resequencing TP53 exons 2–9, which contain 95% of known mutations in TP53 (13 ). Here we compare the sensitivity and detection limits of APEX with a standard method, DHPLC/direct sequencing, and discuss the potential of APEX for application to cancer diagnostic or prognostic purposes. Specimens in the comparison set included 6 cell lines with mutations in different TP53 exons (see Table 1 of the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/
1286
Technical Briefs
vol51/issue7/) and 60 tumor samples with TP53 mutations identified by DHPLC (exons 4 –9) and direct sequencing (29 from paraffin-embedded lung cancers and 31 frozen specimens of oral cancers). The persons performing the APEX screening (exons 2–9) were blinded to mutation status. Mutations in these specimens were representative of the diversity of TP53 mutations in human cancers. Four specimens positive by DHPLC but with no mutation found by sequencing were also tested. DNA extractions, PCR, DHPLC, and sequencing were performed according to described protocols (see Tables 2 and 3 of the online Data Supplement) (14 ). APEX probes were 25mer oligonucleotides with 12carbon amino linkers at their 5⬘ end covering the sense and antisense wild-type TP53 sequence (Genset) and were spotted on 24 ⫻ 60 mm SAL-type aminosilane/ phenylene diisothiocyanate-coated microarray slides (15 ). PCR products were concentrated and purified by use of the Jet quick PCR purification Spin Kit (Genomed). Hybridization and primer extension reactions were performed as published previously (13 ). Data were analyzed with GenoramaTM 4.2 genotyping software, using clustered signal patterns from 20 wild-type TP53 DNA sequences as statistical reference. For each position, the software calculates the distance (difference) between the sample signal and the signal pattern of the wild-type reference cluster database [algorithm described by Tonisson et al. (13 )]. The distance value is used as a measure for calling the given base. Zero distance indicates a perfect match between the given base and the wild-type reference. The default cutoff distance value of the software with cluster analysis is 30 but can be adapted by the user. A representative analysis of a TP53 exon 7 mutation at codon 236 (TAC⬎TGC) is shown in Fig. 1. To evaluate APEX detection limits, defined amounts of PCR products of DNA from 6 cell lines homo- or hemizygous for known mutations were mixed with wild-type PCR products and analyzed in parallel by DHPLC and APEX. An example of DHPLC and APEX detection limits is given in Fig. 1 of the online Data Supplement. With DHPLC, detection limits showed variations depending on mutation type and mutation sequence context (Table 1). With APEX, all mutations were detected with detection limits of 3.125%– 6.25%, consistent values reported previously for codons 273 (CGT⬎CAT) and 248 (CGG⬎TGG) (13 ). In comparison, direct sequencing does not detect most mutations in samples containing less than 25%–30% mutated DNA (16, 17 ). Of 66 TP53 mutations identified by DHPLC/ sequencing in 60 tumors, APEX identified 62 but failed to detect 4 mutations (sensitivity, 94%; see Table 4 of the online Data Supplement). All mutations occurred between exons 4 and 9. Among substitutions, 49 of 52 (94.2%) were correctly identified by APEX. Two of the mutations that escaped APEX detection (codon 91, TGG⬎TGA and codon 174 AGG⬎TGG) corresponded to positions generating background signals in one DNA strand. The third one (codon 173 GTG⬎ATG) was identified on the antisense strand only because of inefficient
hybridization at this position. APEX specificity and sensitivity may be further improved by experimentally establishing cutoff values for mutants at each position of the APEX array or by modifying probe design to minimize the formation of secondary structures interfering with DNA hybridization. The 14 other mutations detected by DHPLC/sequencing were deletions or insertions. APEX correctly identified 1 insertion (⫹ 1 bp; codons 62– 63) and detected the base change at one extremity in 12 other cases (85.7%), including 4 insertions ranging from 1 to 8 bp, 7 deletions ranging from 1 to 4 bp, and 1 rearrangement at codon 184 that was not clearly identified as an insertion or deletion by sequencing. However, in these 12 cases, APEX did not allow complete identification of the change in DNA sequence. Overall, APEX had a sensitivity of 93% in detecting insertion/deletions, but only 7% in fully identifying their positions and sizes. These limitations are intrinsic to APEX design, which is based on hybridization of wildtype sequences to complementary, immobilized oligonucleotides (18 ). In theory, the presence of a deletion should lead to incorporation of an unexpected nucleotide at the 5⬘ border of the deletion and corresponding to the first nucleotide flanking the 3⬘ border of the deletion. Presence of an insertion should generate an unexpected signal corresponding to the first inserted base at the 5⬘ border of the insertion. However, these events will be detectable only if the signals they generate differ from expected wild-type signals at these positions. Assuming equal distribution of insertions and substitutions at all base positions, there is a 1 in 4 chance for either strand that the border may be undetectable and a 1 in 16 chance that the sequence alteration escapes detection. However, in practice, the odds for nondetection are higher because insertions and deletions often occur within singlebase repeats, making it likely that the modified signal may be identical to the expected one. The nonrandom distribution of insertions/deletions in TP53 may explain the relatively poor performance of APEX in mapping both borders of insertions/deletions. Despite these limitations, APEX compares well with the Affymetrix GeneChip p53 in detecting deletions and insertions (11, 17, 19, 20 ). Indeed, the latter does not allow detection of sequence alterations unless representative mutant oligonucleotides have been spotted on the array (10, 11 ). Of the 4 specimens positive by DHPLC but negative by sequencing, APEX detected base changes in 3, including 1 specimen with 2 mutations (a 1-bp substitution in intron 5 and a silent mutation at codon 147; see Table 4 of the online Data Supplement). Finally, in 5 specimens, APEX generated several signals (2 or 3 per specimen), only one of which was confirmed by sequencing (see Table 4 of the online Data Supplement). The unconfirmed signals by DHPLC/sequencing were detected only on 1 DNA strand by APEX, suggestive of an insertion or deletion as discussed above. Thus, none of these discordant changes were single-base substitutions. We interpreted the discordant signals as “false-positive” APEX signals, which gave
Clinical Chemistry 51, No. 7, 2005
an overall positive predictive value for APEX of 92.5% (62 of 67). We found that both fresh-frozen and formalin-fixed, paraffin-embedded tissues are suitable for mutation detection by APEX. The positive predictive values for APEX were 92% (33 of 36) and 93.5% (29 of 31) for frozen and paraffin-embedded tissues, respectively. Thus, the use of paraffin archives did not lead to a higher number of ambiguous or artifactual signals. The second base of codon 72 harbors a common polymorphism in TP53 that affects a BstUI digestion site. The concordance between APEX and either restriction digestion or DHPLC/sequencing data was perfect for all 64 specimens analyzed (see Fig. 2 of the online Data Supplement). This polymorphism may play a role in cancer susceptibility (21, 22 ) and response to therapy (2 ). Recently, Bonafe et al. (23 ) reported that preferential loss of the codon 72P allele in breast tumors of heterozygous patients was associated with a significant decrease in disease-free and overall survival. Additional studies have shown an improved response to chemo-radiotherapy and a longer overall survival in patients whose squamous cell carcinomas tumor retained a wild-type 72R allele rather than a wild-type 72P allele (24 ). Thus, genotyping of TP53 at codon 72, combined with mutation detection, may be relevant for the therapeutic management of cancer. APEX also genotyped 2 other, less frequent polymorphisms at codons 213 (CGA/CGG) and 36 (CCG/CCA) for which concordance between DHPLC and APEX was also perfect. In conclusion, APEX offers a flexible, sensitive, and low-cost resequencing alternative for large-scale studies involving retrospective analysis of pathology collections as well as for application to studies for which fresh-frozen materials are available.
5. 6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
Florence Le Calvez was supported by Special Training Awards from the International Agency for Research on Cancer; Neeme Tonisson was supported as Junior Fellow by Wellcome Trust International Senior Research Grant for Central Europe no. 070191/Z/03/Z and Andres Metspalu by targeted funding from EMER (no. 0518). This research project was supported by European Community (EC) FP6 funding. This publication reflects the authors’ views and not necessarily those of the EC. The Community is not liable for any use that may be made of the information contained herein.
19.
20.
21.
22.
23.
References 1. Hainaut P, Hollstein M. p53 and human cancer: the first ten thousand mutations. Adv Cancer Res 2000;77:81–137. 2. Bergamaschi D, Gasco M, Hiller L, Sullivan A, Syed N, Trigiante G, et al. p53 polymorphism influences response in cancer chemotherapy via modulation of p73-dependent apoptosis. Cancer Cell 2003;3:387– 402. 3. Schneider PM, Stoeltzing O, Roth JA, Hoelscher AH, Wegerer S, Mizumoto S, et al. P53 mutational status improves estimation of prognosis in patients with curatively resected adenocarcinoma in Barrett’s esophagus. Clin Cancer Res 2000;6:3153– 8. 4. Samowitz WS, Curtin K, Ma KN, Edwards S, Schaffer D, Leppert MF, et al.
24.
1287
Prognostic significance of p53 mutations in colon cancer at the population level. Int J Cancer 2002;99:597– 602. Sidransky D. Emerging molecular markers of cancer. Nat Rev Cancer 2002;2:210 –9. Greenblatt MS, Bennett WP, Hollstein M, Harris CC. Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res 1994;54:4855–78. Borresen AL, Andersen TI, Eyfjord JE, Cornelis RS, Thorlacius S, Borg A, et al. TP53 mutations and breast cancer prognosis: particularly poor survival rates for cases with mutations in the zinc-binding domains. Genes Chromosomes Cancer 1995;14:71–5. Geisler S, Lonning PE, Aas T, Johnsen H, Fluge O, Haugen DF, et al. Influence of TP53 gene alterations and c-erbB-2 expression on the response to treatment with doxorubicin in locally advanced breast cancer. Cancer Res 2001;61:2505–12. Kristensen VN, Kelefiotis D, Kristensen T, Borresen-Dale AL. High-throughput methods for detection of genetic variation. Biotechniques 2001;30: 318 –22, 324, 326. Hacia JG. Resequencing and mutational analysis using oligonucleotide microarrays. Nat Genet 1999;21:42–7. Wikman FP, Lu ML, Thykjaer T, Olesen SH, Andersen LD, Cordon-Cardo C, et al. Evaluation of the performance of a p53 sequencing microarray chip using 140 previously sequenced bladder tumor samples. Clin Chem 2000;46: 1555– 61. Kurg A, Tonisson N, Georgiou I, Shumaker J, Tollett J, Metspalu A. Arrayed primer extension: solid-phase four-color DNA resequencing and mutation detection technology. Genet Test 2000;4:1–7. Tonisson N, Zernant J, Kurg A, Pavel H, Slavin G, Roomere H, et al. Evaluating the arrayed primer extension resequencing assay of TP53 tumor suppressor gene. Proc Natl Acad Sci U S A 2002;99:5503– 8. Dai M, Clifford GM, Le Calvez F, Castellsague X, Snijders PJ, Pawlita M, et al. Human papillomavirus type 16 and TP53 mutation in oral cancer: matched analysis of the IARC multicenter study. Cancer Res 2004;64:468 – 71. Guo Z, Guilfoyle RA, Thiel AJ, Wang R, Smith LM. Direct fluorescence analysis of genetic polymorphisms by hybridization with oligonucleotide arrays on glass supports. Nucleic Acids Res 1994;22:5456 – 65. Rosenblum BB, Lee LG, Spurgeon SL, Khan SH, Menchen SM, Heiner CR, et al. New dye-labeled terminators for improved DNA sequencing patterns. Nucleic Acids Res 1997;25:4500 – 4. Ahrendt SA, Halachmi S, Chow JT, Wu L, Halachmi N, Yang SC, et al. Rapid p53 sequence analysis in primary lung cancer using an oligonucleotide probe array. Proc Natl Acad Sci U S A 1999;96:7382–7. Tonisson N, Kurg A, Kaasik K, Lohmussaar E, Metspalu A. Unravelling genetic data by arrayed primer extension. Clin Chem Lab Med 2000;38: 165–70. Bosch FX, Ritter D, Enders C, Flechtenmacher C, Abel U, Dietz A, et al. Head and neck tumor sites differ in prevalence and spectrum of p53 alterations but these have limited prognostic value. Int J Cancer 2004;111:530 – 8. Wen WH, Bernstein L, Lescallett J, Beazer-Barclay Y, Sullivan-Halley J, White M, et al. Comparison of TP53 mutations identified by oligonucleotide microarray and conventional DNA sequence analysis. Cancer Res 2000;60: 2716 –22. McGregor JM, Harwood CA, Brooks L, Fisher SA, Kelly DA, O’nions J, et al. Relationship between p53 codon 72 polymorphism and susceptibility to sunburn and skin cancer. J Invest Dermatol 2002;119:84 –90. Buller RE, Sood A, Fullenkamp C, Sorosky J, Powills K, Anderson B. The influence of the p53 codon 72 polymorphism on ovarian carcinogenesis and prognosis. Cancer Gene Ther 1997;4:239 – 45. Bonafe M, Ceccarelli C, Farabegoli F, Santini D, Taffurelli M, Barbi C, et al. Retention of the p53 codon 72 arginine allele is associated with a reduction of disease-free and overall survival in arginine/proline heterozygous breast cancer patients. Clin Cancer Res 2003;9:4860 – 4. Sullivan A, Syed N, Gasco M, Bergamaschi D, Trigiante G, Attard M, et al. Polymorphism in wild-type p53 modulates response to chemotherapy in vitro and in vivo. Oncogene 2004;23:3328 –37.
DOI: 10.1373/clinchem.2005.048348
