Analytical Biochemistry 299, 24–30 (2001)doi:10.1006/abio.2001.5355, available online at http://www.idealibrary.com on

Detection of Single Base Alterations in Genomic DNAby Solid Phase Polymerase Chain Reactionon Oligonucleotide Microarrays

Martin Huber, Doris Losert, Reinhard Hiller, Christian Harwanegg,Manfred W. Mueller, and Wolfgang M. Schmidt1

VBC-Genomics Bioscience Research GmbH, 1030 Vienna, Austria

Received July 10, 2001; published online November 3, 2001

DNA microarray technology holds significant prom-ise for human DNA diagnostics. A number of technicalapproaches directed at the parallel identification ofmutations or single nucleotide polymorphisms makeuse of polymerase-based specificity, like minisequenc-ing or allele-specific primer elongation. These tech-niques, however, require separate laborious sampleamplification, preparation, and purification steps,making large-scale analyses time and cost consuming.Here, we address this challenge by applying an exper-imental setup using simultaneous solid and liquidphase PCR on polyethyleneimine-coated glass slides, anovel microarray support allowing on-chip amplifica-tion reactions with exquisite specificity. A gene-spe-cific oligonucleotide tiling array contains covalentlyattached allele-specific primers which interrogate sin-gle nucleotide positions within a genomic region ofinterest. During a thermal cycling reaction amplifica-tion products remain covalently bound to the solidsupport and can be visualized and analyzed by theincorporation of fluorescent dyes. Using the describedprocedure we unequivocally defined the presence ofpoint mutations in the human tumor suppressor genep53 directly from a natural DNA source. This semi-multiplex solid phase amplification format allowedthe rapid and correct identification of 20 nucleotidepositions from minute amounts of human genomicDNA. Our results suggest that this approach mightconstitute a vital component of future integrated DNAchip devices used in gene analysis. © 2001 Elsevier Science

Key Words: DNA microarray; solid phase PCR; SNPdetection; point mutation detection; sequencing on-a-chip.

1 To whom correspondence should be addressed at VBC-GenomicsBioscience Research GmbH, Rennweg 95b/5, A-1030 Vienna, Austria.Fax: 143 1 7966572-21. E-mail: wolfgang.schmidt@vbc-genomics.com.

24

Single nucleotide changes in human genes maycause genetic disorders and could provide importanthelp for explaining, e.g., disease susceptibility. Also,knowledge about point mutations linked to malignantdisease could be of prognostic value. Therefore, theaccurate and sensitive analysis of single base alter-ations plays a central role in the field of DNA diagnos-tics. “DNA chip” technology promises to fulfill a num-ber of prerequisites for massive parallel mutationanalysis and high-throughput scoring of single nucleo-tide polymorphisms (SNPs)2 (1, 2).

Although amplification techniques other than PCR,like rolling circle amplification (3) or serial invasivesignal amplification (4), have been described, PCR stillremains the most commonly employed experimentaltool for sample amplification and subsequent sequenceanalysis. Microarray hybridization of PCR amplifiedfragments to allele-specific oligonucleotide probes iswidely used in large-scale SNP genotyping (5). Becausethe specificity of hybridization strongly depends onnucleotide sequence and is difficult to control in a high-density array format, numerous redundant probes foreach locus of interest are required for accurate analysis(6, 7). Therefore, a number of other microarray-basedmethods for mismatch detection exploit the high spec-ificity intrinsic to DNA polymerases, such asminisequencing (8–10), solid phase primer elongation(11–13), or solid phase PCR (14–16). These methodsbased on enzymatic solid phase DNA synthesis, how-ever, require fairly high amounts of template DNAwhich must be prepared separately prior to the actualon-chip genotyping reaction. Protocols often call forsample amplification and purification steps which ren-der large-scale approaches time and cost consuming.

2 Abbreviations used: SNP, single nucleotide polymorphism; EGS,ethylene glycol-bis(succinic acid N-hydroxysuccinimide) ester; PEI,polyethyleneimine; SPEC, silane-polyethyleneimine crosslinker.

0003-2697/01 $35.00Copyright © 2001 by Academic Press

All rights of reproduction in any form reserved.

ts

(

25SOLID PHASE PCR ON OLIGONUCLEOTIDE MICROARRAYS

In this work we have addressed this challenge byapplying a novel solid phase PCR approach for thedetection of single base exchanges in a DNA microar-ray format. The experimental setup combines on-chip PCR of human genomic DNA with simultaneousnested solid phase amplification mediated by allele-specific oligonucleotide primers tethered to a glassslide. Since the analyzed sample is genomic DNA,our approach completely eliminates the need for la-borious sample setup steps. As shown for the humantumor suppressor gene p53, this approach allows uso obtain exact sequence information directly frommall amounts of DNA in fewer than 3 h.

MATERIALS AND METHODS

DNA Samples

Genomic DNA samples from colon cancer cell linesCCL-220 and CCL-228 with described missense muta-tions of the p53 gene (17) were a generous gift of Dr.Robert Mader. CCL-220 carries a mutation in codon248 (CGG to TGG), whereas CCL-228 contains a mu-tated codon 273 (CGT to CAT). In both cell lines themutations are homozygous due to loss of the wild-typeallele.

Oligonucleotides

All oligonucleotides were synthesized in house withan Expedite 8909 nucleic acid synthesizer (PerseptiveBiosystems, Foster City, CA) using standard phospho-roamidite chemistry. Oligonucleotides for microarrayattachment were synthesized with a 59 terminalCH2)6-NH2 modification (Cruachem Ltd, Glasgow,

UK) and purified by perfusion chromatography on aBioCAD Sprint system (PerSeptive Biosystems).

For analyzing the p53 mutations in codon 248 and273 (17) we designed a set of 40 primers each for exons7 and 8, starting at nucleotide positions 46–69 of exon7 and 10–35 of exon 8, respectively (Fig. 1B). Subse-quent primers were relocated one nucleotide down-stream of the previous oligonucleotide and each primerwas synthesized with a 39 G, A, T, or C terminal base.Thus, the tiling array contained 80 oligonucleotides for20 basepairs covering codons 248–250 in exon 7 andcodons 273–275 in exon 8. In addition to the aminomodification the oligonucleotides contained a 59-dT5

spacer (18); for each primer the sequence of the wild-type reference sequence is indicated by an underlinedletter: Ex7-1 59-GTTCCTGCATGGGCGGCATGAA(C/G/A/T), Ex7-2 59-TTCCTGCATGGGCGGCATGAAC(C/G/A/T), Ex7-3 59-CCTGCATGGGCGGCATGAACC(C/G/A/T), Ex7-4 59-CTGCATGGGCGGCATGAACCG(C/G/A/T), Ex7-5 59-TGCATGGGCGGCATGAACCGG(C/G/A/T),Ex7-6 59-GCATGGGCGGCATGAACCGGA(C/G/A/T),Ex7-7 59-GCATGGGCGGCATGAACCGGAG(C/G/A/T),Ex7-8 59-GCATGGGCGGCATGAACCGGAGG(C/G/A/T),

Ex7-9 59-CATGGGCGGCATGAACCGGAGGC(C/G/A/T),Ex7-10 59-GGCGGCATGAACCGGAGGCC(C/G/A/T),Ex8-1 59-ACTGGGACGGAACAGCTTTGAGGT(C/G/A/T), Ex8-2 59-CTGGGACGGAACAGCTTTGAGGTG(C/G/A/T), Ex8-3 59-GGGACGGAACAGCTTTGAGGTGC(C/G/A/T), Ex8-4 59-GGACGGAACAGCTTTGAGGTGCG(C/G/A/T), Ex8-5 59-GACGGAACAGCTTTGAGGTGCGT(C/G/A/T), Ex8-6 59-GACGGAACAGCTTTGAGGTGCGTG(C/G/A/T), Ex8-7 59-G-ACGGAACAGCTTTGAGGTGC-GTGT(C/G/A/T), Ex8-8 59-ACGGAACAGCTTTGAGGT-GCGTGTT(C/G/A/T), Ex8-9 59-CGGAACAGCTTTGAG-GTGCGTGTTT(C/G/A/T), Ex8-10 59-GGAACAGCTTT-GAGGTGCGTGTTTG(C/G/A/T).

Four liquid phase PCR primers were designed to am-plify p53 fragments of 272 bp (spanning exon 7), 284 bp(exon 8), and 673 bp (exons 7 and 8) (Fig. 1B). Theprimers were located at nucleotide positions 536–561 ofintron 6, 106–131 and 249–274 of intron 7, and 29–54 ofintron 8: I6-for 59-GGCCTCATCTTGGGCCTGTGT-TATC, I7-rev 59-GCCGGAAATGTGATGAGAGGTG-GAT, I7-for 59-TGGTTGGGAGTAGATGGAGCCTGGT,I8-rev 59-AGGCATAACTGCACCCTTGGTCTCC.

Preparation of Polyethyleneimine-ethylene glycol-bis(succinic acid N-hydroxysuccinimide) ester(EGS) Derivatized Glass Slides

Briefly, glass slides were treated with a trimethoxy-silane derivatized polyethyleneimine (PEI) and subse-quently activated using an amine reactive crosslinker.The slides are called SPEC, for silane-polyethylenei-mine crosslinker.

Standard glass slides (Melvin Brand, Sigma-Aldrich)were immersed in HCl:methanol (1:1) for 24 h at roomtemperature, washed extensively with deionized water,and dried under an air stream. The slides were thenincubated in a solution containing 3% (v/v) trimethoxysi-lylpropyl modified polyethyleneimine (Gelest, Tullytown,PA) in 95% ethanol for 1 h with vigorous agitation atambient temperature. Afterward the slides were washedin 95% ethanol and dried under an air stream. The slideswere cured at 80°C for 1 h. Activation of the slides wasachieved by treating with the homobifunctionalcrosslinker EGS (Pierce, Rockford, IL).

Coupling of Primers to the Microarrays for Use inOn-Chip PCR

For coupling to SPEC slides oligonucleotides weredissolved in 150 mM sodium phosphate buffer (pH8.5 at 24°C, supplemented with 0.1% (w/v) SDS) at aconcentration of 20 mM and spotted using an Af-fymetrix 417 Arrayer (Affymetrix, Santa Clara, CA)equipped with 125-mm pins. Spots had a diameter of190 –200 mm and the spot-to-spot distance was 300mm. The arrayed slides were incubated in a NaCl-

26 HUBER ET AL.

saturated humid chamber at 25°C for at least 12 h.To block unreacted succinimide esters, the slideswere immersed in 150 mM ethanolamine, 100 mM

FIG. 1. Schematic representation of the experimental strategy. DNAproducts are shown as straight lines. Free primers in the liquid phasearrows. (A) Mismatch detection by solid phase amplification. Genomicphase and the allele-specific amplification in the solid phase in a single rfree DNA primers for the amplification of a fragment spanning the nprimers tethered to the glass slide. The DNA polymerase extends the pfluorescently labeled extension products, indicated by green arrows.polymerase (indicated by a black dot). (B) Detection of p53 point mutaamplification of fragments suitable for the detection of mutations in codbase variation to all four possible bases interrogate nucleotide position

Tris (pH 9.0) at 55°C for 20 min. The slides were thenwashed thoroughly with deionized water and driedagain.

shown in blue; genomic DNA is represented as curved lines and PCRindicated by gray arrows. The solid phase primers are shown as blackA is used as template for the simultaneous amplification in the liquidtion on a glass microarray sealed with a coverslip. The reaction containsotide region of interest. Individual spots of the array contain nested

ectly matched primers (PM) during the cycling reaction and generatesmatched primers (MM) do not allow efficient extension by the DNAs by solid phase amplification. The liquid phase primers used for the48 and codon 273 are indicated. Solid phase primers with a 39 terminalithin the two hotspot regions in exons 7 and 8.

isareDNeacucleerf

Mistionon 2s w

T(m

t

27SOLID PHASE PCR ON OLIGONUCLEOTIDE MICROARRAYS

Solid Phase PCR

A 10-ml PCR reaction mix containing 23 HotStaraq PCR buffer, 50 mM each dNTP, 20 mM Cy3-dCTP

Amersham Pharmacia Biotech Europe, Freiburg, Ger-any), 2.5 mg/ml bovine serum albumin, 25% (v/v) Self-

Seal reagent (MJ Research, Waltham, MA), 3 unitsHotStar Taq DNA polymerase (Qiagen, Hilden, Ger-many) and the p53 primers at 0.5 mM was prepared.Human genomic DNA (30 ng) in sterile water wasadded prior to thermal cycling. After the reaction mixwas pipetted onto the oligonucleotide array, a glasscoverslip (22 3 22 mm) was mounted to seal the reac-ion. Glass slides were put into a PTC 200 In Situ slide

thermocycler (MJ Research) and cycling was carriedout according to the following scheme: 80°C for 10 min,95°C for 5 min, 20 cycles at 95°C for 30 s and 70°C for1 min, followed by 40 cycles at 95°C for 15 s and 70°C

FIG. 2. Detection of point mutations in the human p53 gene by solion-chip PCR using an oligo microarray containing 80 features of 2007 and 8. The 39 terminal nucleotide of the arrayed primers and thobtained in experiments using different DNA sources with either wilddefined the correct mutations (shown in red) in codon 248 (CGG to

for 45 s. After being cycled the slides were placed in0.13 SSC, 0.1% SDS for 10 min with gentle agitation.After the coverslips were removed the slides werewashed again for 10 min in SSC and then washed withdeionized water and dried under an air stream. Wefound that denaturing washes after the amplificationprocedure were not necessary, suggesting that no sig-nificant amount of sequence hybridization to unex-tended mismatch primers occurred.

Fluorescence Scanning and Data Analysis

Incorporation of Cy3-dCTP during the DNA amplifi-cation reaction was measured using an Affymetrix 418Array Scanner (Affymetrix). Fluorescence intensities(medians after local background subtraction) were cal-culated using Genepix 3.0 software (Axon Instruments,Foster City, CA).

hase amplification of genomic DNA. Genomic DNA was subjected todiameter with primers specific for 20 bases of the p53 gene exons

ild-type reference sequences are indicated. The fluorescent imagespe or mutated p53 genes are shown. The experiments unequivocally) and codon 273 (CGT to CAT).

d pmm

e w-ty

TGG

sst .

28 HUBER ET AL.

RESULTS AND DISCUSSION

In our study, high DNA oligonucleotide binding ca-pacity through a heat stable covalent bond and chem-ical properties of the solid support favoring DNA syn-thesis were considered essential for solid phase PCR apriori. We designed a novel glass surface consisting ofa PEI layer activated with amine-reactive crosslinkermoieties suitable for attachment of oligonucleotideprimers. The polymer PEI is known to enhance nucleicacid derivatization of nylon beads (19) and DNA hy-bridization (20). We used a propyltrimethoxylsilyl de-rivatized PEI compound to fabricate glass slides coatedwith a PEI polymer layer. The DNA oligonucleotide

TAB

Data Analysis of Solid Phase Ampli

Exon7

39Nucleotide

SampleA

SampleB CCL-228 CCL-220

1 G 0.4 0.1 0.0 0.8A 0.2 0.3 0.2 0.8T 0.2 0.0 0.3 0.4C 5.0 2.7 3.3 20.7

2 G 0.1 0.2 0.1 0.0A 0.3 0.1 0.3 0.4T 0.0 0.0 0.0 12.2C 3.9 4.4 5.5 3.3

3 G 4.8 10.7 8.3 0.6A 0.3 0.3 0.5 0.1T 0.0 0.0 0.0 1.1C 0.2 0.2 0.6 0.8

4 G 3.5 5.8 6.8 0.3A 0.2 0.2 0.0 0.0T 0.4 0.2 0.4 0.0C 0.6 0.9 1.0 0.6

5 G 0.4 0.6 0.6 0.0A 5.7 8.6 9.6 1.1T 0.0 0.2 1.0 0.0C 0.4 0.3 0.1 0.1

6 G 8.9 21.7 13.4 3.5A 0.7 1.5 0.0 0.1T 0.1 0.2 0.3 0.0C 2.8 2.9 1.9 0.6

7 G 10.6 33.2 18.8 7.2A 0.8 0.0 2.2 0.7T 0.1 0.2 0.0 0.0C 0.3 0.4 0.0 0.5

8 G 0.5 1.1 0.3 0.6A 0.0 0.0 0.0 0.0T 2.4 4.4 1.9 1.3C 12.6 31.6 19.4 11.7

9 G 0.9 2.3 0.6 1.2A 0.4 0.9 0.4 0.6T 0.7 1.1 0.0 1.2C 20.1 32.5 25.4 31.4

10 G 1.1 0.9 0.3 1.5A 0.5 0.7 1.8 0.4T 0.0 0.0 0.0 0.1C 44.5 4.6 6.5 31.1

Note. DNA templates isolated from either healthy donors (sampubjected to solid phase amplification as described in the text. 39 nuhows relative fluorescence intensities obtained from three experimhe correct sequence within the template DNA are indicated in bold

PCR primers were covalently attached to the surfacevia amide bonds, mediated by reaction with the homo-bifunctional crosslinker EGS. The procedure for thefabrication of PEI–EGS-coated glass slides (which wecall SPEC) is described under Materials and Methods.Pilot studies showed that the slides allowed consistentPCR amplification with high specificity. In detail, wedid not observe unspecific reactions resulting in tem-plate-independent primer artifacts (not shown).

In order to apply solid phase DNA amplification onSPEC slides in SNP and mutation detection applications,we designed an amplification strategy using a combina-tion of gene-specific unbound (liquid phase) primers with

1

tion Using Different DNA Sources

Exon8

39Nucleotide

SampleA

SampleB CCL-228 CCL-220

1 G 10.7 7.7 6.1 14.1A 0.9 0.6 0.3 1.0T 2.1 0.8 0.0 1.3C 0.9 0.5 0.0 0.2

2 G 0.8 0.4 0.0 1.1A 0.8 0.3 0.1 0.5T 3.2 1.1 0.5 1.2C 33.0 21.6 19.4 23.7

3 G 25.3 10.5 1.6 27.6A 2.4 0.8 6.1 2.2T 2.3 1.2 0.3 2.4C 4.5 1.6 0.7 3.9

4 G 3.7 1.2 0.0 3.3A 2.8 1.0 0.0 1.6T 28.7 15.1 1.6 23.4C 4.8 1.1 0.0 4.5

5 G 23.3 7.8 0.3 16.7A 2.5 0.7 0.1 1.3T 3.4 0.6 0.1 1.9C 3.8 0.6 0.0 1.9

6 G 4.1 1.2 0.0 2.2A 2.9 1.3 0.0 0.9T 33.9 19.4 5.7 21.2C 5.6 0.8 0.0 3.9

7 G 3.2 1.4 0.0 1.5A 5.2 1.5 0.1 3.0T 34.8 17.8 6.9 30.2C 5.3 1.3 0.0 4.6

8 G 2.1 1.2 0.0 1.5A 2.4 0.8 0.0 0.9T 33.2 17.0 6.4 26.7C 3.6 0.8 0.0 3.0

9 G 21.9 8.8 5.4 25.7A 2.9 1.5 0.0 2.4T 2.7 1.3 0.0 2.5C 2.5 0.9 0.0 2.2

10 G 3.1 1.4 0.0 2.5A 3.0 1.4 0.0 1.7T 29.8 18.6 5.1 32.9C 5.5 1.1 0.0 5.2

A, B) or human colon cancer cell lines (CCL-220, CCL-228) wereotides corresponding to the wild-type reference are bold. The tables. The highest values obtained for each nucleotide position defining

LE

fica

lescle

ent

ahhpffiispptiwit(mlre

ro

29SOLID PHASE PCR ON OLIGONUCLEOTIDE MICROARRAYS

allele-specific solid phase primers in a single reaction.The strategy is shown in Fig. 1A. The unbound PCRprimers amplify the genomic DNA template. Simulta-neously, the allele-specific solid phase nested primersinterrogate the positions of interest within the templateDNA which is generated in the liquid phase. The DNApolymerase elongates only primers with the matching 39terminal base. During cycling the primer extension prod-ucts serve as template for the second strand elongationprimed by the liquid phase primers, thus generating newtemplates for the allele-specific solid phase amplification.Because the reaction products remain covalently boundto the solid phase, incorporation of a fluorescent dyeallows subsequent detection of enzymatic activity andinterpretation of the correct sequence in the templateDNA.

This experimental setup suitable for the detection ofSNPs was exemplarily applied for the detection of pointmutations in the human tumor suppressor gene p53. Thep53 gene belongs to the most frequently mutated genes intumor cells and could therefore constitute an importantprognostic factor in human cancer (21). The vast majorityof mutations reported occur in a conserved region of thegene (exons 5–9). We built an oligonucleotide primer til-ing array covering 10 nucleotide positions within de-scribed mutation hotspot regions each of exons 7 and 8(Fig. 1B). The array contained four primers for each nu-cleotide investigated, each primer with one of the fourpossible bases at its 39 terminal end. This “complete”setup for detection of all possible single nucleotide vari-

FIG. 3. Sensitivity of solid phase amplification for the detection ofcleotide tiling subarray covering p53 codon 273. Wild-type (wt; CGTatios indicated to the left of the images and subjected to solid phasebtained for the four spots corresponding to the affected nucleotide

ants within the investigated region was useful becausemutations in virtually each nucleotide of the exon 7 and8 codons have been reported (17).

Genomic p53 fragments were amplified on-chip us-ing a set of four unbound primers in the liquid phaseallowing the direct analysis of genomic DNA. As shownin Fig. 2 this sequencing “on-a-chip” approach success-fully determined the exact nucleotide sequence withinthe two inspected regions from DNA samples contain-ing wild-type or mutant p53 alleles. Thus, these assayslso easily discriminated between mutations in theomozygous and the heterozygous status. Because aomozygous mutation renders at least two succeedingrimers within the tiling array mismatched primersor the polymerase, such a DNA lesion is easily identi-ed. As an internal control, this provided an additional

mportant advantage for the detection of homozygousingle base alterations (Fig. 2). Naturally, this com-lete sequencing on-a-chip strategy inherently de-ends on template quality and sequence properties andherefore can bring along significant amounts of spotntensity variation. However, the accurate sequenceas unambiguously derived from relative fluorescence

ntensities calculated from the spots corresponding tohe four possible variants of each nucleotide positionsummarized in Table 1). We observed perfect match to

ismatch ratios higher than 10 in most cases and of ateast 3 at only a few positions. Therefore, the procedureeliably generated the correct sequence in multiplexperiments using different DNA sources.

terozygous mutations. Detailed fluorescence images of the oligonu-nd mutant (mt; CAT) synthetic DNA templates were mixed at theplification as described in the text. Relative fluorescence intensitiesition (G to A exchange) are shown to the right.

he) aampos

imcbnsrsS(oa

1

1

30 HUBER ET AL.

Next, we investigated the amount of DNA carrying amutation within a complex genomic DNA source (liketumor tissue) required for accurate detection. Using syn-thetic DNA templates mixed at different ratios we foundthat a mutation can be unambiguously detected, if theaffected allele is present at a proportion of 20% (Fig. 3).Although the level of background signals varies withdifferent solid phase primers and likely influences themethod’s inherent sensitivity, the codon 273 CGT to CATmutation was detectable even if the mutant template waspresent at 10–15% of the total template concentration.

Although not shown here, the described procedure isapplicable equally well to SNP-scoring protocols. Inprinciple, the liquid phase on-chip amplification per-mits similar multiplexing capability like conventionalPCR amplification in a tube, thus allowing the simul-taneous analysis of sequence variation in several genes(manuscript in preparation).

In summary, we have shown that DNA oligonucleo-tides covalently bound to SPEC slides can efficientlyserve as primers for solid phase amplification providing apowerful tool for microarray-based mismatch and SNPdetection. As shown for the human p53 gene, our exper-mental approach is capable of providing sequence infor-

ation directly from a genomic DNA sample within aouple of hours. When compared to other microarray-ased methods, our approach completely eliminates theeed for time and cost consuming sample preparationteps, thus representing a useful advance for applicationsequiring rapid genotyping directly from a natural DNAource. Because the chemical steps involved in preparingPEC slides are generally applicable to silicon surfaces

22, 23), this approach also imparts a valuable componentf future integrated DNA chip devices for use in genenalysis and DNA diagnostics.

ACKNOWLEDGMENTS

We thank Prof. Dr. R. Mader and his lab for supplying us with thecell line DNA samples used in this work.

REFERENCES

1. Landegren, U., Nilsson, M., and Kwok, P.-Y. (1998) Reading bitsof genetic information: Methods for single-nucleotide polymor-phism analysis. Genome Res. 8, 769–776.

2. Tillib, S. V., and Mirzabekov, A. D. (2001) Advances in theanalysis of DNA sequence variations using oligonucleotide mi-crochip technology. Curr. Opin. Biotechnol. 12, 53–58.

3. Lizardi, P. M., Huang, X., Zhu, Z., Bray-Ward, P., Thomas, D. C.,and Ward, D. C. (1998) Mutation detection and single-moleculecounting using isothermal rolling-circle amplification. Nat.Genet. 19, 225–232.

4. Hall, J. G., Eis, P. S., Law, S. M., Reynaldo, L. P., Prudent, J. R.,Marshall, D. J., Allawi, H. T., Mast, A. L., Dahlberg, J. E.,Kwiatkowski, R. W., de Arruda, M., Neri, B. P., and Lyamichev,V. I. (2000) Sensitive detection of DNA polymorphisms by theserial invasive signal amplification reaction. Proc. Natl. Acad.Sci. USA 97, 8272–8277.

5. Southern, E. M. (1996) DNA chips: Analysing sequence by hy-bridization to oligonucleotides on a large scale. Trends Genet. 12,110–115.

6. Hacia, J. G. (1999) Resequencing and mutational analysis usingoligonucleotide microarrays. Nat. Genet. 21, 42–47.

7. Sapolsky, R. J., Hsie, L., Berno, A., Ghandour, G., Mittmann, M.,and Fan, J. B. (1999) High-throughput polymorphism screeningand genotyping with high-density oligonucleotide arrays. Genet.Anal. 14, 187–192.

8. Pastinen, T., Kurg, A., Metspalu, A., Peltonen, L., and Syvanen,A. C. (1997) Minisequencing: A specific tool for DNA analysis anddiagnostics on oligonucleotide arrays. Genome Res. 7, 606–614.

9. Head, S. R., Rogers, Y. H., Parikh, K., Lan, G., Anderson, S.,Goelet, P., and Boyce-Jacino, M. T. (1997) Nested genetic bitanalysis (N-GBA) for mutation detection in the p53 tumor sup-pressor gene. Nucleic Acids Res. 25, 5065–5071.

10. Dubiley, S., Kirillov, E., and Mirzabekov, A. (1999) Polymor-phism analysis and gene detection by minisequencing on anarray of gel-immobilized primers. Nucleic Acids Res. 27, e19.

11. Shumaker, J. M., Metspalu, A., and Caskey, C. T. (1996) Muta-tion detection by solid phase primer extension. Hum. Mutat. 7,346–354.

2. Pastinen, T., Raitio, M., Lindroos, K., Tainola, P., Peltonen, L.,and Syvanen, A. C. (2000) A system for specific, high-throughputgenotyping by allele-specific primer extension on microarrays.Genome Res. 10, 1031–1042.

3. Erdogan, F., Kirchner, R., Mann, W., Ropers, H. H., and Nuber,U. A. (2001) Detection of mitochondrial single nucleotide poly-morphisms using a primer elongation reaction on oligonucleotidemicroarrays. Nucleic Acids Res. 29, e36.

14. Adessi, C., Matton, G., Ayala, G., Turcatti, G., Mermod, J. J.,Mayer, P., and Kawashima, E. (2000) Solid phase DNA amplifi-cation: Characterisation of primer attachment and amplificationmechanisms. Nucleic Acids Res. 28, e87.

15. Strizhkov, B. N., Drobyshev, A. L., Mikhailovich, V. M., andMirzabekov, A. D. (2000) PCR amplification on a microarray ofgel-immobilized oligonucleotides: Detection of bacterial toxin-and drug-resistant genes and their mutations. Biotechniques 29,844–850.

16. Tillib, S. V., Strizhkov, B. N., and Mirzabekov, A. D. (2001)Integration of multiple PCR amplifications and DNA mutationanalyses by using oligonucleotide microchip. Anal. Biochem.292, 155–160.

17. Hernandez-Boussard, T., Rodriguez-Tome, P., Montesano, R.,and Hainaut, P. (1999) IARC p53 mutation database: A rela-tional database to compile and analyze p53 mutations in humantumors and cell lines. Hum. Mutat. 14, 1–8.

18. Southern, E., Mir, K., and Shchepinov, M. (1999) Molecularinteractions on microarrays. Nat. Genet. 21, 5–9.

19. van Ness, J., Kalbfleisch, S., Petrie, C. R., Reed, M. W., Tabone,J. C., and Vermeulen, N. M. (1991) A versatile solid supportsystem for oligodeoxynucleotide probe-based hybridization as-says. Nucleic Acids Res. 19, 3345–3350.

20. van Ness, J., Tabone, J. C., and Moynihan, K. (2000) Polyethyl-eneimine-based biomolecule arrays. U.S. Patent 6150103.

21. Sidransky, D., and Hollstein, M. (1996) Clinical implications ofthe p53 gene. Ann. Rev. Med. 47, 285–301.

22. Shoffner, M. A., Cheng, J., Hvichia, G. E., Kricka, L. J., and Wild-ing, P. (1996) Chip PCR. I. Surface passivation of microfabricatedsilicon-glass chips for PCR. Nucleic Acids Res. 24, 375–379.

23. Cheng, J., Shoffner, M. A., Hvichia, G. E., Kricka, L. J., andWilding, P. (1996) Chip PCR. II. Investigation of different PCRamplification systems in microfabricated silicon-glass chips. Nu-cleic Acids Res. 24, 380–385.