RATIONALSELECTIONOF PCR-BASEDPLATFORMSFOR … · 2016. 6. 7. · ResolutionMelting(HRM). 1 RATIONALSELECTIONOF PCR-BASEDPLATFORMSFOR PHARMACOGENOMICTESTING A.DEMONACO 1,A.D’ORTA

Abbreviations: Single nucleotide polymorphisms (SNPs);Single-strand Conformational Polymorphism (SSCP); AlleleSpecificAmplification (ASA);Amplification Refractory Mu-tation System (ARMS); Restriction fragment length poly-morphism analysis (RFLP); Peptide Nucleic Acid (PNA);

Matrix-Assisted Laser Desorption/Ionization time of flight(MALDI TOF); Fluorescent Resonance Energy Transfer(FRET); Locked nucleic acid (LNA); Oligonucleotide liga-tion assay (OLA); Rolling Circle Amplification (RCA); HighResolution Melting (HRM).

1

RATIONAL SELECTION OFPCR-BASED PLATFORMS FORPHARMACOGENOMIC TESTING

A. DE MONACO1, A. D’ORTA2, C. FIERRO3, M. DI PAOLO4, L. CILENTI5,R. DI FRANCIA1

1Hematology-Oncology and Stem Cell Transplantation Unit, National Cancer Institute, Naples, Italy.2DD Clinic, Caserta, Italy.3UOS Hematology and Cellular Immunology, Azienda dei Colli Monaldi Hospital, Naples, Italy.4CETAC Research Center, Caserta, Italy.5Burnett School of Biomedical Sciences, College of Medicine, University of Central Florida,Orlando, Florida, USA.

Corresponding Author: A. De Monaco, MD; e-mail: [email protected]

AABBSSTTRRAACCTT:: Background: Genotyping is crucial to the identification of genetic markers underlyingdevelopment of neoplastic diseases and individual variations in responses to specific drugs. Cost-and time-effective technologies able to accurately identify genetic polymorphisms will dramaticallyaffect routine diagnostics processes and future therapeutic developments. However, such methodsneed to fulfill the principles of analytical validation to determine their suitability to assess nucleotidepolymorphisms in target genes. Approach: This article reviews the recent developments of technologies for genotyping of sin-

gle nucleotide polymorphisms (SNPs). For the appropriate choice of any method, several criteriamust be considered: i) known or unknown genetic variations in a given cancer gene; ii) needs oftesting within pharmacogenomics studies; iii) diffusion and availability of large platforms and re-quired equipments; iv) suitability of tests for routine diagnostics; v) capacity of methods to offer aspecific and sensitive detection of mutant alleles within great excess of wild-type alleles in a givensample; vi) suitability for high-throughput implementation.Content: This review is intended to provide the reader with a better understandingof the various

technologies for pharmacogenomics testing in the routine clinical laboratory. A brief overview is givenon the available technologies for detection of known mutations together with a precise descriptionof the homogeneous technologies and platforms currently employed in genotyping analysis.

Based on the criteria proposed here, potential users may evaluate advantage and limitations ofthe different analytical platforms and possibly identify the most appropriate one according to spe-cific operative settings and diagnostic needs.

KEY WORDS: Genotyping methods, Analytical validations, Molecular diagnostics.

WCRJ 2014; 1 (4): e391

INTRODUCTION

Recent research demonstrates that certain geneticpolymorphisms are linked to significant variationsamong individuals in the response, in terms of ac-tivity and toxicity, to a given drug1. To confirm thecandidate genetic markers emerging from suchstudies, there is a commensurate need for pharma-cogenomic laboratories to design and validate tar-geted genotyping assays capable of rapidly identifythe individual Single Nucleotide Polymorphism(SNP) of interest within confirmatory clinical stud-ies and in the routine clinical practice. In recentyears, a number of increasingly complex tech-nologies have been applied to the qualitative andsemi-quantitative detection of polymorphisms andmutations in DNA (for simplification, we shallmainly refer topoint mutations, though in general,small deletions or insertions can be as efficientlydetected by the methods described here)2.

Traditional techniques for SNP genotyping de-tection by Single-Strand Conformational Poly-morphism (SSCP) and Heteroduplex analysis havenow been largely replaced by high-throughputmethods including “in silico” discovery platforms.These latter methods generate much more data andare easier to automate.

A recent breakthrough in high-throughput strate-gies is represented by DNA chip technology, whichallows the combined detection and identification ofmutations3. However, for many applications appro-priate chips will be only available in the forthcom-ing years. Thus, conventional screeningmethods forpoint mutations and small deletions will most prob-ably keep their place in the diagnostic laboratory fora reasonable amount of time. Costs, however, areprojected to be high and assay performance and re-sults interpretation will remain strictly dependent onthe availability of highly qualified and well-trainedpersonnel. We will highlight some of the most pop-ular homogeneous technologies that are currentlyused in specialized laboratory, making the transitionfrom the research setting to the clinical laboratoryand discuss key aspects in their validation for geno-typing in pharmacogenomics4.

Homogeneous methods are essentially “single-tube” assays in which all of the processes requiredfor target amplification and detection occur in asingle “closed-tube” reaction (except for Pyrose-quencing), without a solid phase. Combining thethermal cycling system with the signal detectionsystem allows the on-line monitoring of the PCRamplification process5. In this review post PCRagarose gel-based detection methods, will be con-sidered as homogeneous.

The advantages of homogeneous methods in-clude reduced riskof cross-contamination, time-ef-

fectiveness and practicability. These methods areusually amenable to automation and high-through-put processingwith 96-well plates, the current in-dustry standard, but can be further implemented by384-well plate capabilities. Maximum automationcan be achieved by fully integrated systems with ro-botic processing of 96- or 384-well plates through-out the stages of DNA extraction, PCR set-up,amplification, detection, and data interpretation6.

Many of these strategies are now commerciallyavailable and this field is characterized by intensecompetition mixed to many examples of produc-tive cooperation and cross licensing7. No singlegenotyping platform stands out as ideal and it islikely that many of the different technologies de-scribed in this article will be employed in com-bined studies aimed to find disease genes andnovel drug targets.

NEEDS TO DETECT GENETIC VARIATIONS IN CANCER CHEMOTHERAPY

Pharmacogenomic approaches have been applied tomany existing chemotherapeutic agents in an effortto identify relevant inherited variations that may bet-ter predict patient response to treatment and toxic-ity8. Genetic variations which can alter the aminoacid sequence of the encoded protein, include nu-cleotide repeats, insertions, deletions, translocationsand SNPs. Genetic polymorphisms in drug metab-olizing enzymes like Cytochrome P450 family, drugtransporters like Multidrug Resistance-1, and othermolecular targets have been actively explored withregard to functional changes in phenotype (alteredexpression levels and/or activity of the encoded pro-teins) and their contribution to variable drug re-sponse9,10. Clinically relevant examples of geneticdefects highlighting the relevance of cancer phar-macogenomics in optimizing cancer chemotherapyby improving its efficacy and safety are given inTable 1. A new generation of anticancer drugs hasbeen recently designed with high specificity towardtumour cells, providing a broader therapeutic win-dow with less toxicity as compared to conventionalchemotherapy; these drugs represent a new andpromising approach to targeted cancer therapy11.

New agents are designed to interfere with a spe-cific molecular target, usually a protein with a crit-ical role in tumour growth or progression (i.e. atyrosine kinases). There are multiple types of othertargeted therapies already clinically available, in-cluding monoclonal antibodies, antisense in-hibitors, proteasome inhibitors, enzyme-activitymodifiers and immuno-modulatory drugs. Obvi-ously, any of these new agents may exert a selec-tive pressure on tumour cells that elaborate

2

RATIONAL SELECTION OF PCR-BASED PLATFORMS FOR PHARMACOGENOMIC TESTING

TABLE 1. MOST COMMON GENETIC ABNORMALITIES IN CANCER GENESAND THEIR EFFECT IN CHEMOTHERAPY OUTCOMES.

GENE Polymorphism Molecular Drug Effect on (nucleotide effect therapy translation)

Cytochrome Various nucleotide Decreased enzyme Cyclophosfamide Inter-individual variabilityP450 family translation activity Etoposide in Pharmacokinetics

PaclitaxelTPMT2, 3A, 3C Various Rapid degradation 6-MP Hematopoietic toxicity

Polymorphism ThioguanineUGT1A 28 TA repeats in Low expression Irinotecan Neutropenia toxicity

5’ promoterMDR1 (C3435T) Low expression Various Drug resistance TYMS 3 tandem repeats High expression

5-FU, Methotrexate Drug resistance DHFR (T91C) Increase enzyme Methotrexate Drug resistance

activityMTHFR (C677T) Decreased enzyme Methotrexate Toxicity

activityc-KIT (T1982C) Constitutive signal Imatinib Desensitizes activity

(T81421A) activation in GISTc-KIT D816V Imatinib Good response in

Semaxinib t(8;21)-positive AMLEGFR L858R Gefitinib Good response in NSCLC

ErlotinibABL T(9;22) Constitutive signal Imatinib Good response in CML

BCR/ABL activation Dasatinibfusion gene Nilotinib

ABL T315I Imatinib Drug resistanceM351T

RARα T(15;17) Block of maturation All Trans Retinoic Good response in AML-M3 PML/RARα of Myeloid cells acid (ATRA) subtypes fusion gene

Abbreviations: TPMT = thiopurine methyltransferase; UGT1A1 = UDP-glucuronosyltransferase 1A1; MDR1 = multidrug re-sistance 1; TYMS = thymidylate synthase; DHFR = Dihydrofolate reductase; MTHFR = 5,10-methylene tetra hydrofolate re-ductase; EGFR = Epidermal Grow Factor Receptor; 5-FU = 5-fluorouracil; 6-MP = 6-mercaptopurine; GIST = Gastro-intestinalStromal Tumor; AML = Acute Myeloid Leukemia; NSCLC = Non-Small Cell Lung Cancer; CML = Chronic Myeloid Leukemia

3


a subset of patients affected by Non Small CellLung Cancer (NSCLC). The mechanism of actionof both these agents is the selective inhibition of ki-nase activity of Epidermal Growth Factor Receptor(EGFR). Recently, it has been reported that detec-tion of specific point mutations of the EGFR gene intumour cells may allow to discriminate gefinitib-re-sponding patients (EGFR mutated), from non-re-sponders (EGFR wild type), in NSCLC17. Theavailability of this kind of biomarkers could thenrepresent a useful tool for investigating drug resist-ance to specific types of targeted therapies.

GENOTYPING METHODOLOGIES

New in vitro diagnostic assays and the multiplexassay technologies have been developed to re-spond to rapid advances in the understanding ofgenomic variation affecting drug responses. Thesemolecular assays guide the therapeutic treatmentof many diseases because they give information

strategies to survive and proliferate in their pres-ence. The same basic principle applies to proteinkinase inhibitors; the best understanding of thisproblem at a molecular level derives from studieson imatinib resistance in Chronic MyelogeneousLeukemia (CML) patients carrying the BCR/ABLfusion gene. These imatinib-resistant leukemic cellclones, develop following a single nucleotide mu-tation in ABL kinase domain (with consequentamino acid substitution), but can be efficiently sup-pressed by second-generation tyrosine kinase in-hibitors (i.e. dasatinib, nilotinib bosutinib)12,13.These latter agents maintain a full antineoplasticactivity on almost all imatinib-resistant mutantleukemic cell clones14.

Imatinib is also able to inhibit the activity ofother tyrosine kinase such as those encoded by c-KIT and FLT3 genes in patients with Acute Myel-ogeneous Leukemia and in GastrointestinalStromal Tumours (GISTs)15,16.

Similarly to imatinib, two other biological drugs(gefitinib and erlotinib) showed a clinical activity in


about: molecular subtypes of disease (that requiredifferential treatment), which drug has the great-est possibility of managing the disease, and whichpatients are at the highest risk of adverse reactionsto a given drug therapy18.

Mutation-detecting technologies can be dividedinto two major categories depending on the capac-ity to screen for new mutations or to identify al-ready known mutations.

Until a short time ago the only platform cuttingedge considered was Matrix-Assisted Laser Des-orption/Ionization time of flight (MALDI TOF),able to fulfill both discrimination between alterna-tive alleles and detection of both alleles in a singlestep assay. Alternatively, there were several allelediscrimination methods that combine PCR-basedmethods with hybridization probes. The most usedhomogeneous platforms for the detection of knownSNPs can be operatively classified in two majorcategories of PCR-based methods (Table 2): i)agarose gel-based detection; ii) fluorescence-baseddetection19. For the unknown mutations, platformsmust be able to perform both screening and detec-tion. The techniques listed in Table 2 show manyoverlaps and attempts to compare each other mayresult difficult and unproductive. In almost all as-says, DNA amplification is required.

DNA microarrays and next-generation se-quencing (NGS) are the two most important tech-nologies for high-throughput genomic analysis.DNA microarray technology has been developedand consolidated as a routine tool in research lab-oratories and is now transitioning to the clinic.NGS technologies have emerged that enable thesequencing of large amounts of DNA in paralleland they are suitable to different applications, suchas whole or targeted genome sequencing, and RNAsequencing (RNA-seq).

Because of the cost, the last application is stillslow to replace DNA microarray transcriptomeprofiling analysis, especially in our country.

A real progress may soon be reached with the ad-vent of PCR-microarray platforms combining multi-genic analysis with real-time detection. Theirsample-to-result characteristic and simple use will en-able them to bridge the technical gap between re-search and clinics. The miniaturization, integrationand automation of these tools increase accuracy andreproducibility, making them more suitable for rou-tine use. With these advances, genome-based testshave the potential to become a standard tool for main-stream diagnostics, in order to monitor disease onsetand progression, facilitate individualized patient ther-apy and, ultimately, improve patient outcomes20.

Non-PCR-based technologies such as the lig-ase chain reaction, Rolling Circle Amplification(RCA) and Invader® assays (Third Wave Tech-nologies, Madison, WI, USA) are able to genotypedirectly from genomic DNA (i.e. without PCR am-plification) and are amenable to be applied as ho-mogeneous detection methods. DNA chip-basedmicroarray, Golden Gate® Assay and mass spec-trometry genotyping technologies are the latest de-velopment in the genotyping arena. These newertechnologies are currently less widely used in theclinical laboratory setting than PCR-based meth-ods. Due to their wide diffusion, special attentionwill be paid to performance and quality assessmentin all of the homogeneous methodologies.

HOMOGENOUS METHODS FOR DETECTION OF KNOWN MUTATIONS

The choice of a specific genotyping detectionassay for identification of mutations is strongly de-pendent from the type of mutation and its alleleheterogeneity. In general, homogeneous systemsincrease throughput, reduce the chance of crosscontamination and are amenable to automation, butrequire more fluorescently-labelled probes, in-creasing costs and reduced multiplexing capabili-ties. When a large panel of SNPs assays needs tobe developed and budget for instruments are lim-

4

TABLE 2. COMMONLY USED METHODSFOR GENOTYPING AT MOLECULAR LEVEL

Methods for detection and screening for unknown mutations

Screening– Heteroduplex DNA assay (melting curve) – Denaturing-HPLC– Denaturing gradient gel electrophoresis (DGGE)– Single strand conformation polymorphism (SSCP)Detection and screening– Conventional sequencing– High Throughput sequencing

Methods for detection of known mutationsHomogenous: gel based detection– Allele Specific Amplification (ASA)– Restriction Fragment Length Polymorphism (RFLP)– Peptide nucleic acid-mediated Clamping PCRHomogenous: Fluorescent-based detection– FRET probe Allelic Discrimination (Hyb Probe®TaqMan®, Beacons® Scorpions®)

– Locked Nucleic Acid (LNA) probe– Oligo ligation assay (SNPlex®)– Invader® Assay– Pyrosequencing*– High resolution melting (HRM)Heterogenous– Gene Chip technology– Maldi-TOF Mass Spectroscopy– Golden Gate®Assay

*Required pre-PCR step

ited, methods based on conventional PCR fol-lowed by a gel-based detection assay should bepreferred over fluorescent hybridization-basedmethods. Another advantage of electrophoretic de-tection systems is the possibility to direct check forthe appropriate size of amplicons. We define thesemethods as low-throughput, due to their time-con-suming and labour-intensive characteristics. How-ever, fluorescent-based detection systems havebeen developed for application in routine labora-tories, due to their high specificity, high sensitivityand medium/high-throughput.

Gel loading-based detection. Low- throughput

Allele Specific Amplification (ASA)

The method is based on a PCR performed in twoparallel reactions. In the first reaction, the 5’ primeris complementary to the wild-type sequence; in thesecond reaction, the 5’ primer is complementary tothe mutant or polymorphic sequence. Assumingthat elongation occurs only when primer and targetsequence match completely, only one allele of ei-ther mutant or wild-type DNA is amplified. Twodifferent approaches have been described. The firstapproach is based on the lack of primer elongationdue to a mismatch at the far 3’-end of the primer.This method has been named ARMS “Amplifica-tion Refractory Mutation System” and developedby DxS Diagnostics (Manchester, UK)21. In thesecond approach, the mismatch is located withinthe primer, preventing primer annealing when mis-pairing occurs; methods based on this principle aredefined AS-PCR “Allele Specific-PCR”22.

Assuming a homozygous situation, lack of am-plification will occur in one of the reactions whenPCR is performed with different pairs of 5’primers, one complementary and the other notcomplementary to the represented allele. By mul-tiplex, PCR developed by ARMS, different allelescan be distinguished in a single PCR, by using twoannealing temperatures and four primers23. How-ever, elongation of mismatched bases can beavoided when appropriate primers and reactionconditions are applied. Specificity of primer ex-tension may be improved by appropriate adjust-ment of experimental conditions and a web-basedAS primer design application called WASP[http://bioinfo.biotec.or.th/WASP]. This softwareoffers a tool for designing AS primers for SNPs.By integrating the database for known SNPs (usinggene ID), it also facilitates the awkward process ofgetting flanking sequences and other related infor-mation from public SNP databases24.

From these simple PCR-based methods, mostof other above described technologies havestemmed: Real Time SYBR Green is currentlyused to enhance throughput; pyorosequencing ofASA PCR amplicon, could be used to enhance ac-curacy25,26. The specificity and sensitivity of themethod is strongly influenced by the ratio of mu-tant to wild-type DNA. Limit of detection andidentification of a homozygous or heterozygousstate is the main application of the ASA. Severalauthors have reported the detection of few tumorcells carrying mutations in the presence of a largenumber of normal cells27.

Performance and quality assessment is crucial,since the possibility of false positive or negativeresults is the major limitation of ASA. False-posi-tive results may be due to contamination or artifi-cial point mutation introduced by mis-annealing ofthe primers specific for the mutated allele. Opera-tional guidelines to avoid contaminations and ap-propriate of assay conditions should be strictlyfollowed. In addition, positive target alleles mustbe included as controls.

PCR-restriction fragment length polymorphism analysis (RFLP)

This is a commonly used method including a gelelectrophoresis-based technique, such as PCR, cou-pled with RFLP analysis. Specific DNA sequencescan be amplified. The PCR products are then di-gested with appropriate restriction enzymes and vi-sualized by staining the gel after electrophoresis. Ifthe genetic polymorphism produces a gain or lossof the restriction site, a different restriction diges-tion pattern can be recognized28. A major limitationof the PCR-RFLP method is the absolute require-ment that the polymorphisms alter a restriction en-zyme cutting site29. For same point mutation orSNPs that reside in sequences one nucleotide awayfrom endonuclease restriction sites, allele-specificprimers introducing a point mutagenesis may beused to generate artificial mutation sites for RFLP30.Detection limit of simple RFLP analysis is of onemutant cell out of 50 to 100 non-mutant cells31.

For RFLP analysis, a specificity of 100% can beachieved when appropriate restriction enzymes areused. As quality controls, different allelic variantsor wild type and mutant DNA must be included ineach analysis. Recognition sequences may be de-stroyed by errors of the Taq polymerase. In general,errors due to mis-incorporations will become de-tectable only when high numbers of PCR cyclesand/or sensitive detection methods are used. Themethod has to be adjusted to conditions such that nofalse-positive results are obtained when variable


5


amounts and different proportions of wt and mutantDNA are analyzed. Improvement of specificity anddouble-checking for questionable results can beachieved by sequencing of the PCR product.

PNA-mediated Clamping PCR

The peptide nucleic acid (PNA)-based PCR pro-cedure has been developed for the selective en-richment of mutant alleles-specific ampliconswithin a large excess of wild type alleles32. PNA isa synthetic DNA analog in which the normal phos-phodiester backbone is replaced by a non-polar 2-aminoethylglycine chain, while its attachednucleobases complement DNA or RNA in the A-Tand G-C geometry33. Two important features makePNA a superior PCR clamp oligonucleotide forspecific alleles: i) PNA cannot serve as a primerfor polymerization, nor can it be a substrate for ex-onuclease activities of Taq polymerase ii) the melt-ing temperature (Tm) of a perfectly matchedPNA-DNA duplex is higher than that of DNA-DNA of the same length, but a single mismatchdestabilizes the PNA-DNA hybrids, causing a Tmshift of 10-18°C. Therefore, PNA can specificallyblock primer annealing and/or chain elongation ona perfectly matched template without interferingwith templates carrying mismatched bases34. Inthis way, a target mutant DNA can be specificallydetected in a large excess of wild type DNA. De-tection limit of analysis is of one mutant cell overone hundred-thousand wild type cells35. To im-prove sensitivity and throughput, it should be pos-sible to carry out a large-scale screening in anautomated manner by using matrix-assisted laserdesorption/ionization time of flight (MALDI TOF)mass spectrometry36. In addition, the large Tm dif-ference between perfectly matched and mis-matched hybrids makes PNA a good sensor ofpoint mutations. For example, a PNA sensor probehas been used to detect GNAS gene mutationsafter conventional PCR37.

The use of melting curve analysis in combina-tion with fluorescent probes provides a powerfultool for the detection of single base alterations. Thehybridization probe system is most widely used forthis purpose. It consist in a pair of oligonucleotides– the anchor and the sensor – each labeled with adifferent fluorescent dye, such that fluorescenceenergy transfer occurs between the two when theyanneal adjacent sites of a complementary PCRstrand. The melting curve profile of the sensorprobe (designed to anneal to the variable region),allows homogeneous genotyping in a closed tube38.Recently, fluorescent PNA probe was combinedwith PNA-mediated PCR clamping for detection

of variant BCR/ABL allele in leukemia, and of K-Ras mutation in pancreatic cancer39,40.

The key feature of this procedure is that a PNAoligomer bound with fluorophore serves both asPCR clamp and sensor probe, which allows thediscrimination of sequence alterations in mutantcodons from the wild-type sequence38.

Fluorescent-based detection. Medium/high-throughput

Fluorescent Resonance Energy Transfer (FRET) based allelic discrimination

Many of the probe-based systems rely on the prin-ciple of FRET for signal generation. FRET in-volves the non-radioactive transfer of energy froma donor molecule to an acceptor molecule. Briefly,if two fluorophores with an overlapping spectra ofemission are physically close, the wavelength ofthe light emitted from the first fluorophore afterexcitation, is adsorbed from the second fluo-rophore, causing its excitation. Using FRET hy-bridization probes, a donor and an acceptorfluorophore present in two different probes, co-lo-calize, after hybridization, to an adjacent regionon the target molecule. The donor fluorophore isexcited by instrument light source transferring en-ergy from the donor to the acceptor, and causingan increase of measurable fluorescence of the ac-ceptor fluorophore. Based on this chemistry sev-eral platform for genotyping have been developed:Hyb-Probe® (Roche Diagnostic, Indianapolis, IN,USA) TaqMan® (Applied Biosystems, Foster City,CA, USA), Beacons® (Public Health Research In-stitute, Inc. NJ, USA) and Scorpions® (DxS Man-chester, UK).a) Hyb-probe, developed for the use with dedicate

instrumentation “LightCycler” (Roche Diagnos-tic) utilizes a blue-light-emitting diode with themeasurement of fluorescence by three photo de-tection diodes with different wavelength filters.The rapid heating and cooling, allows amplifica-tions to be completed in less than 20 minutes.Using this platform with allele specific primers,Agarwal et al41 described a meaningful improve-ment of the MTHFR and TYMS genotyping.Comparing this technique to conventional RFLP,the authors obtained a 100% concordance in testresults and concluded that Hyb-probe assay is re-liable, economical, and can be performed by lesstrained technologists.

b) An alternative method for polymorphism de-tection is the TaqMan-based allelic discrimina-tion assay, which combines the use of astandard pair of PCR primers, designed to am-

6

plify the region containing the polymorphismof interest, with two different detection probes(one with the wild type sequence and the othercontaining the mutated nucleotide). Each de-tection probes is labeled on 5’-end by a “re-porter” emitting fluorescent dye (different fromthe other) and a quencher on its 3’-end. DuringPCR, the probes specifically annealed to theircomplementary sequence, are cleaved by TaqDNA polymerase (5’ exonuclease), causing theseparation of the reporter dye from thequencher. The relative fluorescence emittedfrom both probes (wt and mutated) is detectedby the instrument and plotted in a 2D clusterplot allowing quantitation of the amount ofeach specific allele present in the analyzed sam-ples. Homozygosity shows increased fluores-cence in one channel, while heterozygosityexhibits intermediate fluorescence in bothchannels.Generally, in an allelic discrimination assay thedefault minimum quality value required for anacceptable genotype call is set to 95 (qualityrange). However, once the accuracy of a geno-typing assay employing allelic discriminationis established, the stringency for these qualityvalues can be increased or decreased, as appro-priate.While this method can be considered ofmedium/high throughput, the ability to analyzemultiplex samples in single tube is limited bythe restricted number of fluorescent dyes withnon-overlapping spectra42. An interesting vari-ant of this technique has been obtained by theFluorescence Polarization detection, which isable to eliminate the need of a quencher dyewith the reduction of the probes price43. How-ever, probe design is largely empirical and op-timization times are significantly increased.Therefore, the optimal use of this platform isprobably achieved when a relatively smallnumber of SNPs must be assayed on a largenumber of samples.

c) Molecular beacons are oligonucleotide probeswith two complementary DNA sequencesflanking the target DNA sequence and with adonor-acceptor dye pair at opposite ends ofeach probe. When the probe is not hybridized tothe target, it adopts a hairpin-loop conforma-tion so that the reporter and quencher dyes areclose together, so that no donor fluorescence isgenerated. When the probe hybridizes to thetarget sequence, the two dyes are separated andthe fluorescence is dramatically increased44.Since the mismatched probe-target hybrids dis-sociate at a consistent lower temperature thanmatched ones, the different Tm increases the

specificity of molecular beacons. In a typicalSNP genotyping, two molecular beacons withsequence matching to the wild-type and variantalleles respectively, are used in the same PCRreaction. The use of two probes labeled withdifferent fluorophores emitting fluorescence atdistinct optical wavelengths, allows simultane-ous discrimination of the possible allelic com-binations.

d) Scorpion is a single bi-functional molecule con-taining a PCR primer covalently linked to aprobe. The molecules are oligonucleotide witha “Stem-Loop” tail containing a fluorophore,which interacts with a quencher to reduce flu-orescence. The Stem-Loop tail is separatedfrom the PCR primer sequence by a chemicalmodification of oligonucleotide, called “PCRstopper” that prevents the copying of the stem-loop sequence during polymerization started byScorpion primer. During the annealing phase ofPCR, the probe sequence in the Scorpion tailcurls back to hybridize to the complementarytarget sequence in the PCR product (as the tailof the scorpion and the amplicon are now partof the same strand of DNA, the interaction isintermolecular). This hybridization event opensthe hairpin loop causing the increase of fluo-rescent signal because the fluorophore is notquenched anymore. The PCR stopper, locatedbetween the primer and the stem sequence, pre-vents read-though of the hairpin loop. Since thehybridization event is generated by a singlemolecule with two functions (primers andprobe), the scorpion system is more effectivethan the other homogeneous probe systems.The reaction is instantaneous and occurs priorto any other competing or side reactions (e.g.amplicon re-annealing or inappropriate targetfolding), resulting in stronger signals, more re-liable probe design, shorter reaction time andbetter discrimination. This contrasts to the bi-molecular collisions required by other tech-nologies such as Taqman or MolecularBeacons45.The possibility to use Scorpion primers foreachpossible mutations in a single multiplex reac-tion, reduce scoring mistakes in the presence ofa negative results46. Moreover, Scorpion chem-istry, suitable for several thermal cycling plat-forms, is cheaper, because it only requires aconventional PCR machine combined to a flu-orescent plate reader. Therefore, Scorpion tech-nology can be easily adapted to high throughputanalysis for large-scale screening programmesby using 96-well plate formats and kits stan-dardized are likely to become available in thenext future.


7


LNA

Locked nucleic acid (LNA) is a nucleic acid ana-logue displaying a very high affinity towards com-plementary DNA and RNA. Structural studiesdemonstrated that the LNA is a DNA mimic, fit-ting seamlessly into an A-type duplex geometry.Several reports indicated LNA as a most promisingmolecule for the development of oligonucleotide-based unambiguous scoring of SNP47. Many SNPassays using LNA technology have been designedand implemented. Because the difference of melt-ing temperature between a perfect match and a sin-gle-nucleotide mismatch is larger for LNA-DNAheteroduplex than DNA-DNA homoduplex, thediscrimination of a SNP is easier using this chem-istry. LNA technique allows the sensitive detectionof rare mutations in a tissue sample containing anexcess of wild-type DNA. During PCR, LNA se-quence selectively blocks amplification of wild-type DNA, while allowing the amplification of themutant codon48. Currently, several LNA genotyp-ing assays have been reported for the screening ofi) factor V Leiden mutation, ii) apolipoprotein B(apoB) R3500Q mutation and iii) two mutations inapolipoprotein E49-51. In these assays, 8mer LNA-capture probes (complementary to either the wildtype or the mutated genomic sequence) are cova-lently attached to individual wells of a microtiterplate and, after hybridization with PCR amplicons,scored colorimetrically with an ELISA-like tech-nique49. The assays have been carefully validatedand results were highly consistent with DNA se-quencing. Immobilized LNA probes may also besuccessfully used in a multiplex SNP genotypingassay performed on a microarray platform52.

PCR-Invader® Assay

The PCR-Invader® (developed by Third WaveTechnologies) is a homogeneous assay. It is a ro-bust SNP genotyping method that does not requireallele-specific dye-labelled probe for each everySNP marker. The use of two generic dye-labelledprobes is sufficient for all SNP markers. This de-tection method is based on the FRET signal gener-ated by cleavage of a doubly labelled fluorescentprobe53. Briefly,PCR product is incubated with twoallele-specific oligonucleotides (called Invaderoligonucleotide) and with the primary probe. TheInvader oligonucleotide anneals to the downstreamportion of the polymorphic site and while the 3’ re-gion of the primary probe is complementary to theupstream region of the polymorphic site. When thepolymorphism is complementary to the primaryprobe (the opposing base), the probe overlaps the 3’

end of the Invader oligonucleotide and forms astructural sequence containing a specific site for re-striction by the Cleavase® enzyme which releasethe 5’ arm of the primary probe54. The cleaved 5'arm of the Invader oligonucleotide, leads to thecleavage of the doubly labelled signal probe by theCleavase enzyme. Since the signal probe is labelledat the 5’ end with a fluorophore and internally witha quencher, the cleavage removes the 5’ fluorophoreand enhances fluorescence. In the mutant allele pri-mary probe and PCR product do not match to thenucleotide being genotyped, as a result no overlap-ping flap structure is formed and no cleavage of pri-mary probe occurs. Several properties ofPCR-Invader technology make it suitable for high-throughput genotyping, as demonstrated by an as-sociation study of 463 SNPson 33 candidate genes,performed to identify a genetic marker able to pre-dict clinical response to IFN-α therapy55,56. Thisstudy successfully proved that SNPs in the 5’-flanking region of signal transducer and activator3 (STAT3) had themost significant association withresponsiveness. One of the major disadvantages ofthe current technology is the need to assay the twoalleles of each SNP in separate reaction wells.Thisreaction format makes this assay time-consumingand labour-intensive. Furthermore, genotype mis-calling can occurwhen one of the two reactions ofthe sample does not work, leading to a heterozy-gous individual being mistyped as homozygous.

Oligonucleotide Ligation Assay (SNPlex)

The SNPlex system (developed by AppliedBiosystems) uses oligonucleotide ligation assay(OLA) and capillary electrophoresis (CE) to ana-lyze bi-allelic SNP genotypes.

The assay workflow for the SNPlex Genotyp-ing System involves the following seven steps, de-signed for easy automation, which can becompleted within two days http://www.pubmed-central.nih.gov/articlerender.fcgi?artid=2291745&rendertype=figure&id=f1: 1) OLA reaction: allele-specific oligonucleotide and locus-specific oligonu-cleotide probes hybridize to the genomic targetsequence and in presence of exactly matched se-quence of SNP site; 2) purification of OLA reac-tion by exonucleolytic digestion of excess probesand linkers. This step is necessary to ensure the ef-ficiency of the subsequent PCR reaction; 3) the re-action involving the simultaneous PCR allowingthe amplification of purified ligation products usinga single pair of PCR primers, one of which is bi-otinylated; 4) capturing of biotin-labelled PCRproducts in streptavidin coated microtiter plates.After a washing step, the non-biotinylated strands

8

are removed, leaving single-stranded ampliconsbound to the microtiter plate; 5) fluorescently la-belled universal “Zip-Chute” probes hybridize tothe bound single-stranded amplicons. Since eachZipChute probe is complementary to a sequencecontained in the ASO (called ZipCode sequence), adouble number of ZipChute probe for each SNP inscreening is required; 6) the specifically boundZipChute probes are eluted into CE buffer; and 7)detection is obtained by CE. This approach is wellsuited for SNP genotyping efforts in which high-throughput and cost effectiveness are essential. TheSNPlex genotyping system offers a high degree offlexibility and scalability, allowing the selection ofcustom-defined sets of SNPs for medium- to high-throughput genotyping projects.

Based on the same principles of the OLA, Faruqiet al57 used a PCR free Rolling Circle Amplification(RCA) in combination with FRET detection, togenotype 10 SNPs in 192 samples. The methodgave quantitative results when a real-time PCR in-strument was used, and the specificity of the assaywas quite high: matching allele specific probes wereligated 100,000-fold faster than mismatched probes.Genotypes called by RCA were identical to thosedetermined by RFLP or minisequencing with a fre-quency of 93%. The RCA assay is commerciallyavailable as part of the SNiPer™ system (GEHealthcare-Amersham Pharmacia Biotech, UK).

Pyrosequencing

Pyrosequencing detects de novo incorporation of nu-cleotides. The incorporation process releases a py-rophosphate, which is converted to ATP in thepresence of adenosine 5'-phosphosulfate that in turnstimulates luciferase58. A charge couple device(CCD) camera detects the light produced by the lu-ciferase-catalyzed reaction. The height of each peakcorrelates to the light signal and is proportional tothe number of nucleotides incorporated. During re-action, after single dNTP incorporation, both ATPand unincorporated dNTPs are degraded by a pyrase,the light is switched off and the cycle re-starts. Thenovel DNA strand, complementary to the target tem-plate, is built up and the nucleotide sequence is de-termined from the signal peak in the pyrogram.Current instrumentation, produced by PyroMark®(Biotage, Uppsala, S.), can detect 500 SNPs/h post-PCR in a 96-well plates format. Pyrosequencing ishighly specific and automated genotype calling isalso allowed for the quantitative nature of the assay.Multiplexed reactions (up to 4-plex) have been suc-cessfully designed. It has also been described a 3-primer system using a common biotinylated primer,with a further reduction of the cost of each assay59.

One more advantage of Pyrosequencing is the abil-ity to detect additional sequence flanking the poly-morphic site (typically 6-30 bases). Therefore, thismethod can be applied for discovery of unknownSNPs and turns useful for analyzing those sequencecontaining complex secondary structures that renderdifficult the application of conventional sequencingapproaches. The major disadvantage of this methodlays in the requirement a post-PCR cleanup step forremoving unincorporated nucleotides, primers, andsalts. Moreover, the presence of >10 nucleotides ho-mopolymer tracts, could complicate the analysis dueto the non-linear light response after incorporationof 5-6 identical dNTPs. A further advantage of Py-rosequencing over other genotyping methods is thatcan it be used to characterize the entire haplotype,not just individual SNPs.

Sivertsson et al60 used pyrosequencing tech-nique for the simultaneous genotyping and screen-ing for new polymorphisms of 82 samples. Twomutation hot spots containing 5 and 7 knownSNPs, respectively, were analyzed in parallel byPyrosequencing and the more labor-intensiveSSCP/cycle-sequencing technique. There was a100% concordance between the two methods, butthe Pyrosequencing assay identified also fournovel sequence variation.

High resolution melting (HRM)

HRM represent an extension of previous heterodu-plex DNA dissociation or melting analysis. It was re-cently introduced as a technique to genotype knownSNPs within small amplicons. This technique is aclosed-tube method, applicableby heteroduplex-de-tection DNA dyes, which can be used at saturatingconcentrations without inhibiting PCR steps61. Thesethird generation fluorescent dsDNA dyes, such asSYTO 9 (Invitrogen Corp., Carlsbad, CA, USA), LCGreen (Idaho Technologies, Salt Lake City, UT,USA) and Eva Green (Biotium Inc, Hayward, CA,USA), devoid of inhibitory ability towards amplifi-cation reactions, can be used at higher concentrationsresulting in a greater saturation of the dsDNA sam-ple. Greater dye saturation produces a higher fidelityof measured fluorescent signals, apparently becausethere is less dynamic dye redistribution to non-de-natured regions during melting phase and becausedyes do not favour higher melting temperature prod-ucts61. The other basic requirement of this techniqueis the presence of a HRM instrument that collectsfluorescent signals with higher optical and thermalprecision. Wild type and mutant samples are distin-guished by melting temperature (Tm) shift. An al-tered curve shape, distinguish heterozygous samplesfrom homozygous ones, better than Tm. The advan-


9


tages of this approach are that labelling of eachprimer (with dye) is not needed and PCR amplifica-tion and melting analysis can be performed in thesame tube/capillary, minimizing specimen handlingand reducing the possibility of error and sample con-tamination. HRM is easy, rapid and not expensiveand has a relevant accuracy for mutational analysisin clinical practice, mainly for genotyping of geneticdisorders and for the identification of somatic muta-tions in human cancers.

The most relevant screening tests with HRM incancer studies were the detection of: i) large se-quence aberrations of FLT3 gene in AML; ii) of c-Kit mutation in GIST; iii) mutation of EGFR andHER2 gene, in lung and in head-neck cancer and c-Kit and BRAF gene activating mutations inmelanoma62-65. HRM was also tested in pancreaticcancer for the mutational analysis of p53, K-ras,BRAF, and EGFR genes, performed on bile ductbrushing specimens66. Finally, HRM was also pro-posed as a rapid and sensitive technique for the as-sessment of DNA methylation67.

COMPARISON OF GENOTYPING METHODOLOGIES

Homogeneous methods for the detection of knownpoint mutations and small deletions or insertionsare summarized in Table 3.

None of the genotyping methods appears idealfor all situations, and then the technique used mustbe driven by project requirements. Each assay hasadvantages and disadvantages and reaction condi-tions must be standardized for each technique.

Fluorescent-based detection systems havebeen developed for most of the assays described,resulting in an easy application in routine labora-tories, due to their high specificity, low detectionlimit, moderate assay stability (decreasing fluo-rophore activity) and medium/high-throughput.Furthermore, these systems are specific, sensitiveand highly reproducible, but usually require morefluorescently labelled probes, with the increaseof cost and limitation in multiplexing capabilities.When a large number of assays must be devel-

10

TABLE 3. COMPARISON OF METHODS FOR DETECTION OF KNOWN POINT MUTATION IS BASED ON ANALYTICAL VALIDATION.

Genotyping Specificity Sensitivity Assay* Equipment#methods (Mut/Wt) stability required

ASA -Medium 1/100 High Gel electrophoresis-High§ (sequencing) 1/10000 system

RFLP -Medium 1/1000 Low (restriction Gel electrophoresis-High (sequencing) 1/10000 enzyme) system

PNA-mediate -High 1/100000 High Gel electrophoresisPCR calmping -Very High (Maldi-Tof) 1/1000000 systemFRET Allelic High 1/100 Middle Dedicate instrumentationDiscrimination (probe fluoro-labeled) and software

a) LC probeb) TaqMan

5’nuclease (End point detection)

c) Beacons probe

d) Scorpions probe Hyb probe

LNA probe Very High 1/10000 Middle (probe Common Fluorescent-fluoro-lebelled) detecting instrumentation or

plate readerInvader Assay Very High 1/100 Middle (cleavase

enzyme)OLA (SNPlex) Very High 1/100 Middle (probe fluoro- Dedicate instrumentation

ebelled) and software

Pyrosequencing Very High 1/10000 Middle (luciferase Pyrosequencerrelated enzyme) dedicate software

HRM Medium 1/100 High Common Fluorescent-detecting

§High/very high specificity, if combined to other detection platform (i.e. sequencer or MALDI-TOF) *Referred to reagent stability: low = restriction enzyme; Middle = dye-labeled oligonucleotide; High = basic oligonucleotide #PCR thermal cycler and other common diffuse instrument, are not included in the estimate equipment

oped and optimization time is limited, methodsbased on simple gel electrophoresis detection arepreferable. The flexibility of the system may alsobe an important factor to be evaluated, if othertypes of assays or applications are desired. Forexample, the throughput of Pyrosequencing islimited by the time required for the luminescencecascade after the addition of each nucleotide. Onthe other hand, this method is more flexible thanother assays because it can be used in a wider va-riety of polymorphisms including, short inser-tions/deletions, triallelic SNPs and because it canbe used for SNP discovery and CpG methylationanalysis.

Nearly all methods require a separate PCR am-plification step for the highest specificity and sen-sitivity. This requirement limits throughput andincreases the cost per reaction. Multiplex PCRhelps to increase throughput and decrease cost, butit is still not possible to develop robust and highlymultiplexed reactions quickly. At present, PNA-mediated PCR Clamping analysis appears the tech-nique best suited for the amplification of lowabundance mutated alleles in great excess of non-mutated ones (very high sensitivity). In additionthese technique is flexible and could be coupledwith Real time-PCR and other post PCR methodslike sequencing, MALDI-TOF68,69. AS-PCR analy-sis, if compared to RFLP, is technically simplerand sufficiently specific (higher specific if com-bined to sequencing of PCR products). Both ofthem do not require any specialized equipment orreagents, but suffer from a concomitant significantdecrease in throughput and are laborious. No at-tempt has been made to assess the cost per reac-tion for each of the described methods, due todifference in manufacturer instrumentation modeland reagents. Moreover, the costs may consistentlyvary between different laboratories due to the abil-ity to produce “in house” many of the reagentsneeded70.

CONCLUSIONS AND FUTURE PERSPECTIVE

All the homogeneous technologies described here,present advantages and disadvantages as summa-rized in Table 4. The comparisons among the dif-ferent genotyping approaches are only made tohighlight the differences in performance amongthese platforms and to draw attention to a core setof selected criteria, before developing SNP assayon a given platform. Decision criteria for the ra-tional selection of on homogeneous platform forSNP detection, mainly depend on specific aims ofthe different diagnostic laboratories. If the lab is

built up to genotype a large number of SNPs in asmall number of sample and if it has specializedpersonnel with a low budget, the most suitableplatforms can be those for which the requiredreagents can be produced “in house” (i.e. AS-PCR,RFLP and PNA). Otherwise, if genotyping testingneeds additional information on mutant allele bur-den quantification, allele-specific detection plat-form are to be preferred (i.e. HybProbe, TaqMan,Beacons, and Scorpions). Finally, if the genotypingpanel is narrow and genotyping samples numer-ous, high-throughput platforms appear more suit-able (Invader, SNPlex). Pyrosequencing is the onlyplatform available for the detection of short se-quence context (i.e. nucleotide deletion or inser-tion). PNA and LNA chemistries are ideal to beused for detection of a rare mutant allele in a largevolume of wt allele. Each technology designed todetect genetic abnormalities continues to evolvequickly. All present and future technological im-plementations for the detecting mutations, whichyield to altered drug response, will always beamenable to analytical validation. The homoge-neous PCR-based methods described in this reviewhave been validated and are well known in theworld of molecular diagnostic. The expected per-formance of an assay can be estimated and eachtest can be monitored by validated QCs proceduresin the global context of external Quality control as-surance programs.

The cornerstone of pharmacogenomics andpharmacogenetics on the future of health care, isthe ability to identify genetic variations (SNP, shortdeletions, translocations and insertions) that alteran individual response to a given drug, and trans-lating SNP test from research to clinical practice.In this scenario, the present process is a multifac-eted task that needs the successful cooperation ofthe diagnostic, pharmaceutical, medical and publichealth fields.

At the present SNP testing is a small and spe-cialized sector, in the context of global diagnosticsindustry, comprising less than 5% of molecular di-agnostics segment71. Over the next few years, theemergence of molecular resistance in the new ther-apies as results of genomic alteration (i.e. kinaseinhibitors), will drives diagnostics company to de-velop new test able to produce results indicativefor tailoring patient’s treatment. Hopefully, the fu-ture implementation of the methods for genotyp-ing, will result in personalized treatment andeventually, in shifting the balance from disease re-lapse towards disease eradication72. Therefore,pharmaceutical and biotechnology companiesshould join each other, in order to develop a com-mercial test suitable for routine diagnostics inpharmacogenomics.


11


AKNOWLEDGEMENTSThe authors are grateful to the “Italian Associationof Pharmacogenomics and Molecular Diagnos-tics” for the use of its bibliography resources.

CONFLICT OF INTERESTSThe Authors declare that they have no conflict ofinterests.

REFERENCES

1. Ma Q, Lu AY. Pharmacogenetics, pharmacogenomics, andindividualized medicine. Pharmacol Rev 2011; 63: 437-459.

2. Isler JA, Vesterqvist OE, Burczynski ME. Analytical valida-tion of genotyping assays in the biomarker laboratory.Pharmacogenomics 2007; 8: 353-368

3. Clark MJ, Chen R, Lam HY, Karczewski KJ, Chen R, EuskirchenG, Butte AJ, Snyder M. Performance comparison of exome DNAsequencing technologies. Nat Biotechnol 2011; 29: 908-901.

4. De Monaco A, Faioli D, Di Paolo M, Catapano O, D’OrtaA, Del Buono M, Del Buono R, Di Francia R. Pharmacoge-nomics markers for prediction response and toxicity incancer therapy WCRJ 2014; 1: e276.

5. Di Francia R, Frigeri F, Berretta M, Cecchin E, Orlando C,Pinto A, Pinzani P. Decision criteria for rational selection ofhomogeneous genotyping platforms for pharmacoge-nomics testing in clinical diagnostics. Clin Chem Lab Med2010; 48: 447-459.

6. Foy CA, Parkes HC. Emerging homogeneous DNA-based tech-nologies in the clinical laboratory. Clin Chem 2001; 47: 990-1000.

7. Kitzmiller JP, Groen DK, Phelps MA, Sadee W. Pharma-cogenomic testing: relevance in medical practice: whydrugs work in some patients but not in others. Cleve ClinJ Med 2011; 78: 243-257.

8. Harrak M, Khabbal Y, Amarti A, El Hassouni M, OuldimK. Pharmacogenetics and prediction of side effects ofdrugs. Ann Biol Clin (Paris) 2014; 72: 405-412.

12

TABLE 4. ADVANTAGE AND DISADVANTAGES OF DESCRIBED PLATFORMS.

Platforms Advantages Disadvantages

AS-PCR Larger diffuse methods Low specificity and sensitivityCombining capabilities Low throughputAutomation feasibleLow cost “in house” set up

RFLP Low cost Data interpretationSimple instrumentation Low specificity and sensitivityAllelic discrimination Low throughput

Intense time-labourPNA-mediate PCR clamping- Combining capabilities No allelic discrimination

Automation feasible Low throughputLow cost “in house” set upHigh sensitivity

FRET Allelic Discrimination Common instrumentation Low sensitivitya) LC probe Medim/high throughput Moderate multiplex capabilitiesb) TaqMan 5’nuclease Allelic discrimination Possible data mis-interpretation

(End point detection) Quantification feasiblec) Beacons probed) Scorpions probeLNA probe Combining capabilities No allelic discrimination

Automation feasible Labor intense for optimization protocolHigh sensitivityCommon instrumentation

OLA (SNPlex) High throughput Dedicate software and reagentsAllelic discrimination No combining capabilitiesHigh sensitivity

Invader assay High throughput Dedicate instrument, software and reagentsAllelic discrimination (2 tube assay) No combining capabilitiesHigh sensitivity

Pyrosequencing High throughput Dedicate instrument, software and reagentsAllelic discrimination No combining capabilitiesHigh sensitivity No multiplex Short sequence context for each sample, good for deletion and insertion

HRM Common instrumentation Low sensitivityMedim/high throughput No multiplex capabilities Allelic discrimination Possible data mis-interpretationLow cost

Major selected criteria include: analytical validation, larger diffuse methods and instruments, allelic discrimination, platformcombine capabilities, intense time-labor, methods allows the specific and sensitive detection of mutant alleles in great excess ofwt alleles, throughput

9. Bozina N, Bradamante V, Lovri� M. Genetic polymorphismof metabolic enzymes P450 (CYP) as a susceptibility fac-tor for drug response, toxicity, and cancer risk. Arh HigRada Toksikol 2009; 60: 217-242.

10. Brambila-Tapia AJ. MDR1 (ABCB1) polymorphisms: func-tional effects and clinical implications. Rev Invest Clin2013; 65: 445-454.

11. Weng L, Zhang L, Peng Y, Huang RS. Pharmacogeneticsand pharmacogenomics: a bridge to individualized can-cer therapy. Pharmacogenomics 2013; 14: 315-324.

12. Gray NS, Fabbro D. Discovery of allosteric bcr-abl in-hibitors from phenotypic screen to clinical candidate.Methods Enzymol 2014; 548: 173-188.

13. Radi M, Schenone S, Botta M. Allosteric inhibitors of Bcr-Abl: towards novel myristate-pocket binders. Curr PharmBiotechnol 2013; 14: 477-487.

14. Ai J, Tiu RV. Practical management of patients withchronic myeloid leukemia who develop tyrosine kinase in-hibitor-resistant BCR-ABL1 mutations. Ther Adv Hematol2014; 5: 107-120.

15. Tomasson MH, Xiang Z, Walgren R, Zhao Y, Kasai Y, MinerT, Ries RE, Lubman O, Fremont DH, McLellan MD, PaytonJE, Westervelt P, DiPersio JF, Link DC, Walter MJ, GraubertTA, Watson M, Baty J, Heath S, Shannon WD, NagarajanR, Bloomfield CD, Mardis ER, Wilson RK, Ley TJ. Somaticmutations and germline sequence variants in the ex-pressed tyrosine kinase genes of patients with de novoacute myeloid leukemia. Blood 2008; 111: 4797-4808.

16. Le Cesne A, Blay JY, Reichardt P, Joensuu H. Optimizing ty-rosine kinase inhibitor therapy in gastrointestinal stromaltumors: exploring the benefits of continuous kinase sup-pression. Oncologist 2013; 18: 1192-1199.

17. Yu S, Zhang B, Xiang C, Shu Y, Wu H, Huang X, Yu Q, YinY, Guo R. Prospective Assessment of Pemetrexed or Peme-trexed Plus Platinum in Combination With Gefitinib or Er-lotinib in Patients With Acquired Resistance to Gefitinibor Erlotinib: A Phase II Exploratory and Preliminary Study.Clin Lung Cancer 2014 pii: S1525-7304 00195-8

18. Koch WH. Technology platforms for pharmacogenomicdiagnostic assays. Nat Rev Drug Discov 2004; 3: 749-761.

19. Di Francia R, Valente D, Catapano O, Rupolo M, Tirelli U,Berretta M. Knowledge and skills needs for health pro-fessions about pharmacogenomics testing field. Eur RevMed Pharmacol Sci 2012; 16: 781-788.

20. Guarnaccia M, Gentile G, Alessi E, Schneider C, Petralia S,Cavallaro S. Is this the real time for genomics? Genomics2014; 103: 177-182.

21. Newton CR, Graham A, Heptinstall LE, Powell SJ, Sum-mers C, Kalsheker N, Smith JC, Markham AF. Analysis ofany point mutation in DNA. The amplification refractorymutation system (ARMS). Nucleic Acids Res 1989; 17:2503-2516.

22. Okayama H, Curiel DT, Brantly ML, Holmes MD, CrystalRG. Rapid, nonradioactive detection of mutations in thehuman genome by allele-specific amplification. J Lab ClinMed 1989; 114: 105-113.

23. Ferrie RM, Schwarz MJ, Robertson NH, Vaudin S, SuperM, Malone G, Little S. Development, multiplexing, andapplication of ARMS tests for common mutations in theCFTR gene. Am J Hum Genet 1992; 51: 251-262.

24. Wangkumhang P, Chaichoompu K, Ngamphiw C, Ru-angrit U, Chanprasert J, Assawamakin A, Tongsima S.WASP: a Web-based Allele-Specific PCR assay designingtool for detecting SNPs and mutations. BMC Genomics2007; 8: 275.

25. Grupe A, Germer S, Usuka J, Aud D, Belknap JK, Klein RF,Ahluwalia MK, Higuchi R, Peltz G. In silico mapping ofcomplex disease-related traits in mice. Science 2001; 292:1915-1918.

26. Pettersson M, Bylund M, Alderborn A. Molecular haplo-type determination using allele-specific PCR and pyrose-quencing technology. Genomics 2003; 82: 390-396.

27. Sun X, Hung K, Wu L, Sidransky D, Guo B. Detection oftumor mutations in the presence of excess amounts ofnormal DNA. Nat Biotechnol 2002; 20: 186-189.

28. Shi MM. Enabling large-scale pharmacogenetic studies byhigh-throughput mutation detection and genotypingtechnologies. Clin Chem 2001; 47: 164-172.

29. Shi MM, Bleavins MR, de la iglesia FA. Technologies fordetecting genetic polymorphisms in pharmacogenomics.Mol Diagn 1999; 4: 343-351.

30. Haliassos A, Chomel JC, Grandjouan S, Kruh J, Kaplan JC,Kitzis A. Detection of minority point mutations by modi-fied PCR technique: a new approach for a sensitive diag-nosis of tumor-progression markers. Nucleic Acids Res1989; 17: 8093-8099.

31. Chen J, Viola MV. A method to detect ras point mutations insmall subpopulations of cells. Anal Biochem 1991; 195: 51-56.

32. Demers DB, Curry ET, Egholm M, Sozer AC. EnhancedPCR amplification of VNTR locus D1S80 using peptide nu-cleic acid (PNA). Nucleic Acids Res 1995; 23: 3050-3055.

33. Egholm M, Buchardt O, Christensen L, Behrens C, FreierSM, Driver DA, Berg RH, Kim SK, Norden B, Nielsen PE.PNA hybridizes to complementary oligonucleotides obey-ing the Watson-Crick hydrogen-bonding rules. Nature1993; 365: 566-568.

34. Taback B, Bilchik AJ, Saha S, Nakayama T, Wiese DA,Turner RR, Kuo CT, Hoon DS. Peptide nucleic acid clampPCR: a novel K-ras mutation detection assay for colorec-tal cancer micrometastases in lymph nodes. Int J Cancer2004; 111: 409-414.

35. Hancock DK, Schwarz FP, Song F, Wong LJ, Levin BC. De-sign and use of a peptide nucleic acid for detection of theheteroplasmic low-frequency mitochondrial en-cephalomyopathy, lactic acidosis, and stroke-like episodes(MELAS) mutation in human mitochondrial DNA. ClinChem 2002; 48: 2155-2163.

36. Schatz P, Distler J, Berlin K, Schuster M. Novel method forhigh throughput DNA methylation marker evaluationusing PNA-probe library hybridization and MALDI-TOF de-tection. Nucleic Acids Res. 2006; 34: e59.

37. De Monaco A, Berretta M, Pugliese S, Valente D, Ciaf-farafa S, Di Francia R. Evaluation of genotyping Costs. EurRev Med Pharm Sci 2014; 18: 2084-2087.

38. Luo JD, Chan EC, Shih CL, Chen TL, Liang Y, Hwang TL,Chiou CC. Detection of rare mutant K-ras DNA in a single-tube reaction using peptide nucleic acid as both PCRclamp and sensor probe. Nucleic Acids Res 2006; 34: e12.

39. Kreuzer KA, Le Coutre P, Landt O, Na IK, Schwarz M,Schultheis K, Hochhaus A, Dörken B. Preexistence andevolution of imatinib mesylate-resistant clones in chronicmyelogenous leukemia detected by a PNA-based PCRclamping technique. Ann Hematol 2003; 82: 284-289.

40. Chen CY, Shiesh SC, Wu SJ. Rapid detection of K-ras mu-tations in bile by peptide nucleic acid-mediated PCRclamping and melting curve analysis: comparison with re-striction fragment length polymorphism analysis. ClinChem 2004; 50: 481-489.

41. Agarwal RP, Peters SM, Shemirani M, von Ahsen N. Improvedreal-time multiplex polymerase chain reaction detection ofmethylenetetrahydrofolate reductase (MTHFR) 677C>T and1298A>C polymorphisms using nearest neighbor model-based probe design. J Mol Diagn 2007; 9: 345-350.

42. De la Vega FM, Lazaruk KD, Rhodes MD, Wenz MH. As-sessment of two flexible and compatible SNP genotypingplatforms: TaqMan SNP Genotyping Assays and theSNPlex Genotyping System. Mutat Res 2005; 573: 111-135.


13

43. Latif S, Bauer-Sardina I, Ranade K, Livak KJ, Kwok PY. Flu-orescence polarization in homogeneous nucleic acidanalysis II: 5'-nuclease assay. Genome Res. 2001; 11: 436-440.

44. Tyagi S, Bratu DP, Kramer FR. Multicolor molecular bea-cons for allele discrimination. Nat Biotechnol 1998; 16:49-53.

45. Solinas A, Brown LJ, McKeen C, Mellor JM, Nicol J, Thel-well N, Brown T. Duplex Scorpion primers in SNP analysisand FRET applications. Nucleic Acids Res 2001; 29: E96.

46. Thelwell N, Millington S, Solinas A, Booth J, Brown T.Mode of action and application of Scorpion primers to mu-tation detection. Nucleic Acids Res 2000; 28: 3752-3761.

47. Mouritzen P, Nielsen AT, Pfundheller HM, Choleva Y,Kongsbak L, Møller S. Single nucleotide polymorphismgenotyping using locked nucleic acid (LNA). Expert RevMol Diagn 2003; 3: 27-38.

48. Dominguez PL, Kolodney MS. Wild-type blocking poly-merase chain reaction for detection of single nucleotideminority mutations from clinical specimens. Oncogene2005; 24: 6830-6834

49. Orum H, Jakobsen MH, Koch T, Vuust J, Borre MB. De-tection of the factor V Leiden mutation by direct allele-specific hybridization of PCR amplicons tophotoimmobilized locked nucleic acids. Clin Chem 1999;45: 1898-1905

50. Jacobsen N, Fenger M, Bentzen J, Rasmussen SL, Jakob-sen MH, Fenstholt J, Skouv J. Genotyping of theapolipoprotein B R3500Q mutation using immobilizedlocked nucleic acid capture probes. Clin Chem 2002; 48:657-660.

51. Jacobsen N, Bentzen J, Meldgaard M, Jakobsen MH,Fenger M, Kauppinen S, Skouv J. LNA-enhanced detec-tion of single nucleotide polymorphisms in theapolipoprotein E. Nucleic Acids Res 2002; 30: e100.

52. Guerasimova A, Nyarsik L, Liu JP, Schwartz R, Lange M,Lehrach H, Janitz M. Liquid-based hybridization assay withreal-time detection in miniaturized array platforms. Bio-mol Eng 2006; 23: 35-40.

53. Mein CA, Barratt BJ, Dunn MG, Siegmund T, Smith AN,Esposito L, Nutland S, Stevens HE, Wilson AJ, Phillips MS,Jarvis N, Law S, de Arruda M, Todd JA. Evaluation of sin-gle nucleotide polymorphism typing with invader on PCRamplicons and its automation. Genome Res 2000; 10:330-343.

54. Kaiser MW, Lyamicheva N, Ma W, Miller C, Neri B, Fors L,Lyamichev VI. A comparison of eubacterial and archaealstructure-specific 5'-exonucleases. J Biol Chem 1999; 274:21387-21394.

55. Ito N, Eto M, Nakamura E, Takahashi A, Tsukamoto T,Toma H, Nakazawa H, Hirao Y, Uemura H, Kagawa S,Kanayama H, Nose Y, Kinukawa N, Nakamura T, Jinnai N,Seki T, Takamatsu M, Masui Y, Naito S, Ogawa O. STAT3polymorphism predicts interferon-alfa response in patientswith metastatic renal cell carcinoma. J Clin Oncol 2007;25: 2785-2791.

56. Eto M, Kamba T, Miyake H, Fujisawa M, Kamai T, UemuraH, Tsukamoto T, Azuma H, Matsubara A, Nishimura K,Nakamura T, Ogawa O, Naito S; Japan ImmunotherapySNPs-Study Group for Kidney Cancer. STAT3 polymor-phism can predict the response to interferon-� therapy inpatients with metastatic renal cell carcinoma. Eur Urol2013; 63: 745-752.

57. Faruqi AF, Hosono S, Driscoll MD, Dean FB, Alsmadi O,Bandaru R, Kumar G, Grimwade B, Zong Q, Sun Z, Du Y,Kingsmore S, Knott T, Lasken RS. High-throughput geno-typing of single nucleotide polymorphisms with rolling cir-cle amplification. BMC Genomics 2001; 2: 4.

58. Alderborn A, Kristofferson A, Hammerling U. Determina-tion of single-nucleotide polymorphisms by real-time py-rophosphate DNA sequencing. Genome Res 2000; 10:1249-1258.

59. Fakhrai-Rad H, Pourmand N, Ronaghi M. Pyrosequencing:an accurate detection platform for single nucleotide poly-morphisms. Hum Mutat 2002; 19: 479-485.

60. Sivertsson A, Platz A, Hansson J, Lundeberg J. Pyrose-quencing as an alternative to single-strand conformationpolymorphism analysis for detection of N-ras mutationsin human melanoma metastases. Clin Chem 2002; 48:2164-2170.

61. Wittwer CT, Reed GH, Gundry CN, Vandersteen JG, PryorRJ. High-resolution genotyping by amplicon meltinganalysis using LCGreen. Clin Chem 2003; 49: 853-860.

62. Vaughn CP, Elenitoba-Johnson KS. High-resolution melt-ing analysis for detection of internal tandem duplications.J Mol Diagn 2004; 6: 211-216.

63. Willmore C, Holden JA, Zhou L, Tripp S, Wittwer CT, Lay-field LJ. Detection of c-kit-activating mutations in gas-trointestinal stromal tumors by high-resolution ampliconmelting analysis. Am J Clin Pathol 2004; 122: 206-216.

64. Willmore-Payne C, Holden JA, Layfield LJ. Detection ofEGFR- and HER2-activating mutations in squamous cellcarcinoma involving the head and neck. Mod Pathol2006; 19: 634-640.

65. Willmore-Payne C, Holden JA, Tripp S, Layfield LJ. Humanmalignant melanoma: detection of BRAF- and c-kit-activating mutations by high-resolution amplicon meltinganalysis. Hum Pathol 2005; 36: 486-493.

66. Willmore-Payne C, Volmar KE, Huening MA, Holden JA,Layfield LJ. Molecular diagnostic testing as an adjunct tomorphologic evaluation of pancreatic ductal systembrushings: potential augmentation for diagnostic sensi-tivity. Diagn Cytopathol 2007; 35: 218-224.

67. Wojdacz TK, Dobrovic A. Methylation-sensitive high reso-lution melting (MS-HRM): a new approach for sensitiveand high-throughput assessment of methylation. NucleicAcids Res 2007; 35: e41.

68. Kyger EM, Krevolin MD, Powell MJ. Detection of thehereditary hemochromatosis gene mutation by real-timefluorescence polymerase chain reaction and peptide nu-cleic acid clamping. Anal Biochem 1998; 260: 142-148.

69. Schatz P, Distler J, Berlin K, Schuster M. Novel method forhigh throughput DNA methylation marker evaluationusing PNA-probe library hybridization and MALDI-TOF de-tection. Nucleic Acids Res 2006; 34: e59.

70. Di Francia R, Berretta M, Catapano O, Canzoniero LM,Formisano L. Molecular diagnostics for pharmacogenomictesting of fluoropyrimidine based-therapy: costs, methodsand applications. Clin Chem Lab Med 2011; 49: 1105-1111.

71. Ma Q, Lu AY. Pharmacogenetics, pharmacogenomics, andindividualized medicine. Pharmacol Rev 2011; 63: 437-459.

72. Berretta M, Di Francia R, Tirelli U. Editorial – The new on-cologic challenges in the 3rd millennium WCRJ 2014; 1:e133.table

14

Documents

RATIONALSELECTIONOF PCR-BASEDPLATFORMSFOR … · 2016. 6. 7. · ResolutionMelting(HRM). 1 RATIONALSELECTIONOF PCR-BASEDPLATFORMSFOR PHARMACOGENOMICTESTING A.DEMONACO 1,A.D’ORTA