Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Single tube genotyping of sickle cell anaemia using PCR-based SNP analysis

Single tube genotyping of sickle cell anaemia using PCR-based SNP analysis © 2001 Oxford University Press Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 Single tube genotyping of sickle cell anaemia using PCR-based SNP analysis Christy M. Waterfall* and Benjamin D. Cobb Molecular Sensing plc, Unit 3, Challeymead Business Park, Bradford Road, Melksham, Wiltshire SN12 8LH, UK Received August 31, 2001; Revised and Accepted October 9, 2001 ABSTRACT energy transfer, mass spectrometry or direct sequencing of PCR products have also been shown to be effective tools for Allele-specific amplification (ASA) is a generally SNP detection (2). Implementation of these methods can applicable technique for the detection of known require specific reagent synthesis and downstream purification, single nucleotide polymorphisms (SNPs), deletions, making them complicated and expensive. insertions and other sequence variations. Convention- However, assays based on PCR for SNP diagnosis have ally, two reactions are required to determine the broad potential in clinical diagnostics because of their inherent zygosity of DNA in a two-allele system, along with simplicity and potential low cost. Established PCR-based significant upstream optimisation to define the methodologies include certain ligation assays (3,4), genetic bit analysis (5), arrayed primer extension (6) and restriction specific test conditions. Here, we combine single tube fragment length polymorphisms (7), and use of sequence- bi-directional ASA with a ‘matrix-based’ optimisation specific fluorescent probes (8). strategy, speeding up the whole process in a reduced PCR has also been adapted for the detection of well-characterised reaction set. We use sickle cell anaemia as our model SNPs using allele-specific amplification (ASA); oligonucleo- SNP system, a genetic disease that is currently screened tides complementary to a given DNA sequence except for a using ASA methods. Discriminatory conditions were mismatch at their 3′-hydroxyl residue will not function as rapidly optimised enabling the unambiguous identi- primers in the PCR under appropriate conditions. A typical fication of DNA from homozygous sickle cell patients ASA test consists of two complementary reactions, each (HbS/S), heterozygous carriers (HbA/S) or normal containing a common primer, an allele-specific primer and Taq DNA in a single tube. Simple downstream mathematical DNA polymerase lacking 3′→5′ proofreading activity. The analyses based on product yield across the optimisation first reaction contains a primer specific for the normal (or wild- set allow an insight into the important aspects of type) DNA sequence and refractory to amplification from mutant DNA at a given locus. Similarly, the second reaction priming competition and component interactions in contains a mutant-specific primer unable to amplify wild-type this competitive PCR. This strategy can be applied to DNA. Molecular conformation is achieved by analysis of the any polymorphism, defining specific conditions resulting PCR amplicon profiles. A normal individual will using a multifactorial approach. The inherent generate product in the first reaction only; a heterozygote simplicity and low cost of this PCR-based method amplifies products in both reactions; and a homozygous validates bi-directional ASA as an effective tool in mutant individual does so only in the mutant-specific reaction. future clinical screening and pharmacogenomic Internal control amplification is necessary, providing a research where more expensive fluorescence-based positive control for the PCR test (9). approaches may not be desirable. More recently, single tube adaptations have been introduced for known SNP detection in a two-allele system including competitive oligonucleotide priming (10), multi-coloured INTRODUCTION fluorescent oligo labelling systems (2), tetra-primer PCR (11) and overlapping PCR strategies (12,13). The latter both use bi- The successful implementation of single nucleotide polymorphism directional primer arrangements in which common ‘outer’ (SNP) screening for clinical research and the administration of primers define the size of each allele-specific fragment medicine tailored to an individual’s genotype, so-called allowing simple identification using electrophoretic methods. ‘personalised medicine’, lies in the ability to deliver highly reliable, accurate and inexpensive assays that can discriminate Maintaining the integrity of all ASA techniques, particularly between key DNA polymorphisms. Biotechnology and for diagnostic application, requires the optimisation of many pharmaceutical companies have addressed this challenge with experimental parameters. The literature suggests that a magnesium the development of methodologies including single-strand and oligonucleotide titration is sufficient to obtain discriminatory conformation polymorphism analysis, heteroduplex analysis conditions. Generally, further mismatches are incorporated at and denaturing gel electrophoresis. Each of these techniques the 3′-end of the allele-specific primer to weaken hydrogen relies on conformation-induced gel mobility changes (1). bonding between the primer and template, increasing the like- Simple homogeneous assays based on fluorescence resonance lihood of discrimination (14). However, these conventional *To whom correspondence should be addressed. Tel: +44 1225 706245; Fax: +44 1225 700056; Email: christy.waterfall@molecular-sensing.com e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 2 OF 8 approaches are simplistic considering the complex component Bi-directional allele-specific PCR amplifications interactions in a PCR reaction and prove time consuming and WT-AS (5′-ATG GTG CAC CTG ACT CCT GA-3′)and expensive. This has limited the applicability of PCR-based WT-CP517 (5′-CCC CTT CCT ATG ACA TGA ACT-3′) SNP diagnosis for routine clinical application. In particular, bi- were designed for amplification of a 517 bp fragment from the directional systems require significant optimisation due the normal β-globin gene (wild-type primer set). MUT-AS (5′-CAG effects of primer competition caused by multiple primer sets in TAA CGG CAG ACT TCT CCA-3′)and MUT-CP267(5′-GGG the PCR amplification. TTT GAA GTC CAA CTC CTA-3′) were designed for ampli- Here, we modify current systems at a number of levels with fication of a 267 bp fragment from homozygous mutant DNA the application of a multifactorial optimisation procedure to (HbS/S) conferring sickle cell disease (mutant primer set). All single tube bi-directional ASA. Our strategy enables the rapid reactions were performed in a volume of 25 µ l containing identification of component ratios favouring ASA in a greatly 10 ng total human genomic DNA template, 10× PCR reaction reduced experiment set using unmodified oligonucleotides. –1 buffer, 1 U µ l Taq DNA polymerase and water. The wild-type This low cost optimisation strategy combined with a single primer set, mutant primer set, MgCl and dNTP concentrations tube genotyping system introduces improvements to PCR-based are shown in Table 1 and were aliquoted according to an SNP screening. orthogonal array (Table 2) (17). Amplification was performed using an MJ Thermal Cycler (GRI, UK) through an initial SNP model denaturation at 95°C for 2 min, followedby30cyclesofdena- We demonstrate our system by detecting the single base pair turation at 95°C for 30 s, primer annealing at 60°C for 30 s and mutation that causes the autosomal recessive disease sickle cell extension at 72°C for 35 s. A small blind trial was performed anaemia (GenBank accession no. M34058). A single A→T in which 14 human genomic DNA samples of mixed zygosity transversion in the sequence encoding codon 6 of the human were used as substrate DNA under discriminatory conditions β-globin gene causes the substitution of amino acid glutamine that were considered optimal (reaction 9, Tables 1 and 2). The by valine, forming a mutant globin chain termed HbS. Haemo- exact template concentration of each sample was not specified, globin S is freely soluble when fully oxygenated, but under –1 varying between 5 and 25 ng µ l . conditions of low oxygen tension the erythrocytes assume an irregular ‘sickle’ shape, leading to aggregation and haemo- Analysis of PCR products lysis. The sickled erythrocytes become trapped in the micro- An aliquot (10 µ l) of each PCR product was loaded with 6× circulation, depriving organs of essential oxygen, causing pain loading buffer (Promega) onto a 2% agarose gel and run in and chronic anaemia (15). Homozygous HbS is a serious 0.5× TBE buffer at 30 mA constant current. Amplification haemoglobinopathy found almost exclusively in the Black products were visualised by UV transillumination (305 nM) population. About 8% of American Blacks are carriers and ∼0.2% are affected. In the UK sickle cell disease affects ∼10 000 people (actionresearch.co.uk/camp/sick.html). Hetero- Table 1. Concentration levels of components used in a multifactorial array for zygotes (sickle cell carriers) are clinically normal, although the identification of discriminatory reaction conditions for bi-directional ASA their red cells will sickle when subjected to very low oxygen pressure in vitro (genelink.com/technical/sickle.html). PCR-based Level A B C techniques for prenatal diagnosis of sickle cell disease were WT-AS/WT-CP517 (µ M) 0.2 0.4 0.8 introduced in 1989 following the invention of the amplification MUT-AS/MUT-CP267 (µ M) 0.2 0.4 0.8 refractory mutation system (9) and are still the method of MgCl (mM) 0.5 1 2 choice in clinical research laboratories (16). A single tube 2 adaptation of conventional ASA would add significant benefits dNTPs (mM) 80 120 200 to routine testing for sickle cell disease. Table 2. An orthogonal array for testing four reaction parameters, each at MATERIALS AND METHODS three different levels Reagents Reaction WT-AS/ MUT-AS/ MgCl dNTPs PCR reagents (Taq DNA polymerase, HS TaQUANT-OFF, WT-CP517 MUT-CP267 10× Mg-free Taq buffer, 25 mM MgCl ) were obtained from 1 AAAA Q-Biogene (CA), except for dNTPs (2 mM each dATP, dCTP, 2 A BBB dGTP and dTTP), which were from Promega (UK). Primer selection was accomplished using the PrimerCalc software 3 A CCC package (Q-Biogene). For standard gel electrophoresis, molecular 4B A B C biology grade agarose was used (Q-Biogene) in 0.5× TBE 5 BBCA buffer (45 mM Tris, 45 mM borate, 1 mM Na EDTA, pH 8.3). 6 BCA B Double deionised water (18 MΩ ) was used for all buffers and solutions for electrophoresis and PCR amplifications. 7C A C B Homozygous HbS human genomic DNA and heterozygous 8 CBA C HbA/S DNA were a generous gift from the National Haemo- 9 CCBA globinopathies Reference Service (Oxford, UK). PAGE 3 OF 8 Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 Figure 1. Schematic representation of bi-directional allele-specific PCR. The wild-type primer set is shown in green and the mutant primer set in pink. Three possible fragments can be formed depending upon substrate DNA zygosity, as well as enzyme system and cycling parameters. The green amplicon defines the allele-specific product formed when only the wild-type primers are extended and the pink amplicon defines the allele-specific product formed when only the mutant primers are extended. A heterozygote is identified by the presence of both PCR products. The target region amplified by non-allele-specific outer primers is shown in pale grey. Our selected enzyme system was inefficient at amplifying longer sequences and in the presence of competition from a smaller fragment insufficient target region was amplified for visualisation on a stained gel. after staining with SYBR Gold (Molecular Probes, UK) for WT-AS has a single A nucleotide at the 3′-end, matching the T 30 min. The agarose gels were photographed and analysed by nucleotide on the non-coding strand of the normal β-globin gel imaging software to obtain maximum band optical density gene. Primer MUT-AS has a single A nucleotide at the 3′-end, measurements as a representation or ‘score’ of specific product matching the T base on the coding strand of the sickle cell yield. gene. The wild-type outer primer (WT-CP517) was positioned 517 bp downstream on the opposite strand of WT-AS. The mutant outer primer (MUT-CP267) was placed 267 bp RESULTS upstream of MUT-AS. Unlike normal ASA, this assay has an inherent PCR control, as at least one allele-specific fragment Experimental design should be yielded per reaction under optimised conditions. A schematic of bi-directional ASA is shown in Figure 1. A This negates the need for a distant pair of primers to direct the wild-type allele-specific primer is designed as complementary amplification of an internal control fragment. to the non-coding strand of the wild-type allele terminating at Optimisation strategy for discrimination between the the known polymorphic site. This primer will be extended in normal and sickle cell alleles the presence of wild-type DNA only under appropriate PCR conditions. A further allele-specific primer is included in the Inclusion of multiple allele-specific primer sets in a single reaction, designed as complementary to the coding strand of reaction tube requires the consideration of many experimental the known mutant allele, again terminating at the SNP site. factors. For typical multiplex reactions, the key parameter for This primer will be extended in the presence of the mutant optimisation is primer concentration. At equimolar concentrations, allele only, as it is refractory to extension of wild-type DNA amplification by the two primer sets may be weak or strong at due to a single base pair mismatch at the 3′-end. Two ‘outer a given locus, an event dependent upon primer stabilities, primers’ are placed on the opposite strand at a pre-defined binding efficiencies, priming competition, product size and number of bases from their allele-specific pair in order that concentration of other reaction components (18). Normally, molecular confirmation can be achieved by analysis of the primer ratios are individually optimised for each reaction, resulting amplicon profiles using gel-based methods. A differ- however, an adjustment in primer concentration has associated ence of at least 100 bp is preferable to increase electrophoretic implications considering the complex interactions between separation. other components in a PCR. Unique to this assay, each primer Taq DNA polymerase will extend both outer primers in all is complementary to a different part of the DNA sequence, reactions as they are non-allele-specific, but the yield of the introducing additional variability to the reaction. large amplicon spanning the whole target region is subject to MgCl and dNTPs have been shown to affect the efficiency variation according to the enzyme system and cycling of priming and extension by altering the kinetics of association parameters employed. The HotStart enzyme system used in and dissociation of primer–template duplexes at annealing and this study was less efficient at amplifying larger fragments and extension temperatures. These components also alter the in the presence of competition from a smaller fragment did not efficiency with which the polymerase recognises and extends amplify sufficient DNA for the larger band spanning the whole such duplexes. Of particular relevance to allele-specific PCR target region to be visualised on an agarose gel. methods, the concentration of MgCl and dNTPs required for We designed two 20/21mer allele-specific primers, WT-AS optimal amplification depends largely on the target and primer and MUT-AS, complementary to the normal β-globin or sickle sequences. The presence of excess magnesium in a reaction cell genes, respectively, arranged in a bi-directional orientation may result in the accumulation of non-specific amplification such that both primers terminate at the mutation site. Primer products and insufficient concentrations reduce product yield. e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 4 OF 8 Figure 3. Gel image data from a small blind trial. Fourteen human DNA samples of varying zygosity were selected and amplified under conditions specified by reaction 9 (Table 2). Diagnosed template zygosity is shown on the gel image. Three normal samples were identified by amplification of the 517 bp product only. Six heterozygous samples were identified by amplification of both the 267 and 517 bp fragments. For heterozygotes, the 267 bp band was slightly more intense than the 517 bp band, likely due to staining disparity between optimisation sets A and B. Five homozygous sickle cell samples were identified by amplification of the 267 bp amplicon only. Figure 2. Gel image data of PCR amplifications. Outer lanes contain 100 bp ladders. Reactions 1–9 according to the matrix (Table 2) are labelled accordingly. Set A represents reactions in which wild-type template DNA was the substrate. amplified in all cases using our enzyme system and cycling Set B represents reactions in which HbS/S template DNA was included. The 517 and 267 bp allele-specific fragments were generated in each set, respectively. parameters. Product yield varied across each reaction set. Table 3 represents maximal Diagnostic conditions were selected by qualitative analysis optical density (OD) of the amplicon bands, used as a representation of product of PCR yields from set A (WT DNA) and set B (HbS/S DNA). yield/reaction score. Optimal conditions for zygosity determination are those that amplify each allele-specific amplicon at similar efficiencies, as measured by band intensity from the gel. In this study, the conditions and interactions of reaction 9 (Table 2) satisfied It has also been demonstrated that dNTPs quantitatively bind these criteria, producing 517 and 267 bp bands with maximum 2+ Mg , so that any modification of dNTP concentration will OD values of 160.67 and 165.93 units, respectively. The optimal require a compensatory adjustment of MgCl (17). Additionally, nature of these conditions was confirmed unambiguously in a 2+ Taq polymerase requires free Mg as a cofactor so any excess small blind trial of 14 DNA samples of different zygosity of dNTPS can have a detrimental effect on product yield by (Fig. 3). Reactions generating the 517 bp fragment identified a chelating cofactor ions (19). Standard approaches to multiplex normal genotype. Amplification of the 267 bp fragment or allele-specific PCR optimisation are unable to account for confirmed a homozygous mutation and sickle cell carriers these multiple interactions, as they are largely based on trial- (HbA/S) were identified by the amplification of both the 517 and and-error strategies. 267 bp bands, diagnostic of heterozygous inheritance. The We applied a multifactorial optimisation strategy in which robustness of the optimised reaction conditions was demon- optimal component concentrations for allele discrimination strated by the ability to amplify from a range of unknown template were identified in a single trial with few reactions. The wild- concentrations without compromising the allele specificity of the type primer set, mutant primer set, MgCl and dNTPs were system. selected as four critical parameters, the interactions of which Effect of each parameter on specificity are most likely to affect the outcome of bi-directional PCR. The concentrations of these components were varied according A quantitative analysis was performed in which the actual product yield for each reaction was used to estimate the effects to an orthogonal array and differences in product yield across of individual components on amplification in bi-directional the reaction set were used for two purposes, qualitatively in the PCR. Signal-to-noise (SNL) ratios were calculated using the identification of diagnostic reaction conditions and quantitatively maximal optical density (max OD) of the allele-specific PCR to study the individual effect of each parameter on the assay. products (Table 3) to represent product yield (y) in the quadratic Optimal conditions for zygosity determination loss function, n 2 SNL = –10 log(1/n 1/y ) Figure 2 shows gel image data from reaction sets A and B. Both i= 1 i sets are identical in component ratios and cycling conditions, assuming that variable y denotes the performance of the system containing normal and HbS/S DNA templates, respectively. measured for a particular combination of factor settings in a Although product yield varied considerably across the reaction given experiment (total of n repeated measurements per sets, only the appropriate allele-specific amplicon was generated experiment) (17). Reactions giving non-allele-specific fragments according to template zygosity. Set A amplified the 517 bp or no amplification at all were given a score of 1. The effects of amplicon and set B generated the 267 bp fragment only. The reaction components on the SNL ratio were determined by larger fragment spanning the whole target region was not regression analysis for both normal and HbS/S DNA templates PAGE 5 OF 8 Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 Table 3. Maximal optical density (max OD) measurements of band intensities from amplification sets A and B in Figure 2 Reaction 1234 567 89 Set A, max OD wild-type 5.13 90.83 80.30 63.40 63.70 1 132.27 1 160.67 Set B, max OD mutant 113.6 172.53 169.20 108.73 161.13 183.67 144.80 1 165.93 All values are in arbitrary units. Figure 4. Normal DNA. Effects of reaction components on the SNL ratios for the amplification of a wild-type 517 bp amplicon from normal DNA. Increased SNL value represents a larger effect on product yield than a low SNL. (Figs 4 and 5, respectively), using calculated SNL values The regression profiles of primer sets generated by normal and mutant DNA samples were quite different, demonstrating (Table 4). differences in primer stabilities, efficiencies and competition. We were able to demonstrate the significance of priming For normal DNA the aim is to direct amplification of the competition and component interaction on the outcome of 517 bp fragment only. The lower concentration of the wild- bi-directional ASA reactions. The SNL analysis revealed that type primer set tested gave the largest SNL value relating to the optimal magnesium requirement for the generation of product yield. Intermediate concentrations decreased the SNL maximal product yield without compromising specificity was and the upper concentration showed an incremental increase in 1.5 mM. This value did not vary according to template SNL. An identical profile was observed with the mutant primer zygosity, which is in accordance with the multiple roles of set, optimally redundant in the presence of wild-type DNA. It magnesium in PCR. Likewise, the optimal concentration of is possible that at lower concentrations of both primer sets, dNTPs in sets A and B were similar, with lower levels being i.e. conditions of minimal competition, the wild-type pair is able favoured. The quantitative binding of magnesium ions by to direct allele-specific amplification of the 517 bp fragment. dNTPs may account for this observation, as excess dNTPs may When the mutant allele is present, the aim is to direct ampli- chelate the magnesium cofactor ions of Taq, decreasing fication of the 267 bp amplicon only. The optimal concentration processivity or simply reducing the yield and specificity. of the mutant primer set was predictably at the upper level. e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 6 OF 8 Figure 5. HbS/S DNA. Effects of reaction components on the SNL ratios for the amplification of a mutant 267 bp amplicon from HbS/S DNA. An increased SNL value represents a larger effect on product yield than a low SNL. Likewise, the inclusion of lower concentrations of the wild- a report by Kwok et al. (20) in which only A:G, G:A and C:C type primer allowed maximal optimal amplification of the mismatches reduced overall PCR product yield significantly. mutant fragment, as the wild-type set thereby exhibited least Other combinations required the incorporation of additional competition with the mutant primer set. We have observed mismatches near the 3′-end to increase discrimination. This similar trends using other SNP models, confirming that primer difference may be due to improvements in the process of oligo- interactions have a significant role in determining the outcome nucleotide synthesis and the use of recombinant Taq DNA of bi-directional amplifications. polymerases. In particular, the use of hot start systems was seen to improve the specificity of our allele-specific amplifications. The optimisation system used and the increased priming competition DISCUSSION in our bi-directional assay will have positively contributed to the absolute discrimination. We have successfully demonstrated the use of bi-directional The implementation of an orthogonal array in our optimisation ASA for single tube genotyping of the SNP responsible for system generated product yield data from two small reaction sickle cell anaemia. Our simple matrix optimisation strategy sets that served a dual purpose in our experiments. Firstly, we offers advantages over conventional methods, allowing the determination of specific test conditions in a greatly reduced were able to use product yield data qualitatively to identify reaction set. Combined with single tube bi-directional allele- conditions under which both allele-specific fragments were specific PCR, rapid diagnosis of normal, homozygous and amplified to similarly maximal efficiencies according to heterozygous template DNA can be achieved simply and template zygosity. Secondly, reaction ‘scores’ were used applied to the screening of any known SNP. Bi-directional quantitatively in a simple regression analysis, allowing the ASA is based on the recent adaptation of ASA methods, using effects of ‘control factors’ influencing reaction performance to a single tube for the diagnosis of a two-allele system, differing be assessed simultaneously. This provided an insight into the from other protocols described in which modified oligonucleo- component interactions and priming competition that dominate tides or cycling parameters are generally required to maintain bi-directional allele-specific amplifications. specificity. The conditions selected for the blind trial were 0.8 µ Mfor We found that the A:A or T:T mismatches designated by the each primer from the wild-type and mutant sets, 1 mM magnesium sickle cell polymorphism were completely refractory to ampli- ions and 80 µ M dNTPs. This is unlikely to be in accordance fication under the optimised conditions. This is discordant with with optimal component levels calculated by SNL values PAGE 7 OF 8 Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 because of the different objectives of each enquiry. The theoretical Table 4. Signal-to-noise ratios (SNL) for reaction components used for the amplification of (A) the 517 bp wild-type allele-specific fragment and (B)the target of the SNL analysis is to define conditions producing the 267 bp mutant allele-specific fragment, calculated using product yields from maximum yield of each product. This differs from the practical Figure 2 target of specifying diagnostic conditions under which each allele must be amplified specifically and at equal efficiencies Level enabling a heterozygous mutation to be identified without a SNL A B C gross over-amplification of one allele over another. Thus, to (A) Normal DNA template achieve this objective, sub-optimal conditions for the amplifica- tion of one allele may be combined with optimal conditions for WT-AS/WT-CP517 18.94 4.77 4.77 the amplification of another by a less efficient primer set, as in MUT-AS/MUT-CP267 18.95 4.77 4.77 this study. 2+ Mg 1.68 39.95 38.12 Two further variables that have an effect on reaction dNTP 18.94 4.77 4.77 outcome are the enzyme system and instrumentation used for (B) HbS/S DNA template amplification. The HotStart Taq DNA polymerase used in this study was extremely accurate, amplifying allele-specific products WT-AS/WT-CP517 43.12 42.93 4.77 with only minimal spurious products. Additionally, the MUT-AS/MUT-CP267 41.01 4.77 44.73 HotStart system was inefficient at amplifying larger fragments 2+ Mg 4.77 42.88 43.94 (>600 bp) when in competition with a smaller fragment, thus dNTP 42.94 44.32 4.77 the non-allele-specific fragment spanning the whole target region was not amplified in this case. In similar studies, this larger fragment has been designated an internal control amplification, avoiding the incorporation of a distant primer pair conventionally required for diagnostic tests (12). ACKNOWLEDGEMENTS However, any successful bi-directional ASA reaction will yield a fragment serving as an internal positive control for the We would like to thank Drs John Old and Chris Fisher at the PCR. Thus we favour our system where the processivity of Taq National Haemoglobinopathy Reference Service, Oxford, UK, is expended generating allele-specific products only. We tested for the kind gift of DNA from affected sickle cell individuals and sickle cell carriers. This work was supported by grants a conventional enzyme system in similar experiments and found that the larger fragment was generally produced at the from the University of Bath and Molecular Sensing plc. expense of the allele-specific fragment, with decreased specificity (data not shown). Application to other thermal cyclers may REFERENCES require further optimisation, since different reaction vessels 1. Brookes,A.J. (1999) The essence of SNPs. Gene, 234, 177–186. and heating/cooling methods can dramatically change the 2. Myakishev,M.V., Khripin,Y., Hu,S. and Hamer,D.H. (2001) High-throughput dynamics of a PCR. SNP genotyping by allele-specific PCR with universal energy-transfer Here, genotyping was simply based on size discrimination of labelled primers. Genome Res., 11, 163–169. 3. Lee,H.H. (1996) Ligase chain reaction. Biologicals, 24, 197–199. the resulting allele-specific PCR products, visualised by agarose 4. Bi,W. and Stambrock,P.J. (1997) CCR: a rapid and simple approach for gel electrophoresis. With the emergence of qPCR technologies, mutation detection. Nucleic Acids Res., 25, 2949–2951. offering rapid thermal cycling and on-line product detection, 5. Nikiforov,T.T., Rendle,R.B., Goelet,P., Rogers,Y.H., Kotewicz,M.L., our system can be adapted to detect PCR products based on Anderson,S., Trainor,G.L. and Knapp,M.R. (1994) Genetic Bit Analysis: a solid phase method for typing single nucleotide polymorphisms. their specific melting temperature using non-specific DNA-binding Nucleic Acids Res., 22, 4167–4175. dyes. This would address high throughput issues whilst 6. Tonisson,N., Kurg,A., Kaasik,K., Lohmussar,E. and Metspalu,A. (2000) maintaining the advantages of low cost and the use of simple Unravelling genetic data by arrayed primer extension. Clin. Chem. Lab. chemistries associated with the gel-based assay. Med., 38, 165–170. 7. Gu,W., Aguirre,G.D. and Ray,K. (1998) Detection of single-nucleotide The technique described could be modified to detect more polymorphisms. Biotechniques , 24, 836–837. than one SNP simultaneously with the combination of our 8. Kuklin,A., Munson,K., Gjerde,D., Haefele,R. and Taylor,P. (1997/98) optimisation strategy and accurate primer design. It has Detection of single-nucleotide polymorphisms with the WAVE™ DNA recently been suggested that the grouping and interaction of fragment analysis system. Genet. Test., 1, 201–205. 9. Newton,C.R., Graham,A., Heptinstall,L.E., Powell,S.J., Summers,C., several SNPs in haplotypes may be more important than single Kalsheker,N., Smith,J.C. and Markham,A.F. (1989) Analysis of any point SNPs in the identification of drug response and disease mutation in DNA. The amplification refractory mutation system (ARMS). susceptibility of individuals (21). The commercial applications Nucleic Acids Res., 17, 2503–2516. of this are widespread, as the development of reliable, cost- 10. Gibbs,R.A., Nguyen,P. and Caskey,C.T. (1989) Detection of single DNA base differences by competitive oligonucleotide priming. Nucleic Acids effective tests will be considerably more complex if haplotypes Res., 17, 2437–2448. need to be determined rather than single SNPs. 11. Ye,S., Humphries,S., and Green,F. (1992) Allele specific amplification by In conclusion, we have described various factors that have a tetra-primer PCR. Nucleic Acids Res., 20, 1152. combined influence on the outcome of bi-directional PCR 12. Liu,Q., Thorland,E.C., Heit,J.A. and Sommer,S.S. (1997) Overlapping PCR for bidirectional PCR amplification of specific alleles: a rapid amplifications. We have demonstrated the technique by identifying one-tube method for simultaneously differentiating homozygotes and accurate conditions for determining the zygosity of sickle cell heterozygotes. Genome Res., 7, 389–398. patients and carriers, a technique that can be simply applied to 13. Sasvari-Szekely,M., Gerstner,A., Ronai,Z., Staub,M. and Guttman,A. the detection of any known genetic mutation. (2000) Rapid genotyping of factor V Leiden mutation using single-tube e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 8 OF 8 bi-directional allele-specific amplification and automated ultrathin-layer 18. Henegariu,O., Heerema,N.A., Dlouhy,S.R., Vance,G.H. and Vogt,P.H. agarose gel electrophoresis. Electrophoresis, 21, 816–821. (1997) Multiplex PCR: critical parameters and step-by-step protocol. 14. Sommer,S.S., Groszbach,A.R. and Bottema,C.D.K. (1992) PCR Biotechniques, 23, 504–511. amplification of specific alleles (PASA) is a general method for rapidly 19. Linz,U., Delling,U. and Rubsamen-Waigmann,H. (1990) Systematic detecting known single-base changes. Biotechniques, 12, 1152. studies on parameters influencing the performance of the polymerase 15. Wu,Y., Ugozzoli,L., Pal,B.K. and Wallace,R.B. (1989) Allele-specific chain reaction. J. Clin.Chem. Clin.Biochem., 28, 5–13. enzymatic amplification of β-globin genomic DNA for diagnosis of sickle 20. Kwok,S., Kellogg,D.E., McKinney,N., Spasic,D., Goda,L., Levenson,C. cell anemia. Proc. Natl Acad. Sci. USA, 86, 2757–2760. and Sninsky,J.J. (1990) Effects of primer-template mismatches on the 16. Old,J., Petrou,M., Varnavides,L., Layton,M. and Modell,B. (2000) polymerase chain reaction: human immunodeficiency virus type 1 model Accuracy of prenatal diagnosis for haemoglobin disorders in the UK: studies. Nucleic Acids Res., 18, 999–1005. 25 years’ experience. Prenat. Diagn. 20, 986–991. 21. Davidson,S. (2000) Research suggests importance of haplotypes over 17. Cobb,B.D. and Clarkson,J.M. (1994) A simple procedure for optimising SNPs. Nature Biotechnol., 18, 1134–1135. the polymerase chain reaction (PCR) using modified Taguchi methods. Nucleic Acids Res., 22, 3801–3805. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nucleic Acids Research Oxford University Press

Single tube genotyping of sickle cell anaemia using PCR-based SNP analysis

Loading next page...
 
/lp/oxford-university-press/single-tube-genotyping-of-sickle-cell-anaemia-using-pcr-based-snp-dMm8ogEd7d

References (22)

Publisher
Oxford University Press
ISSN
0305-1048
eISSN
1362-4962
DOI
10.1093/nar/29.23.e119
Publisher site
See Article on Publisher Site

Abstract

© 2001 Oxford University Press Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 Single tube genotyping of sickle cell anaemia using PCR-based SNP analysis Christy M. Waterfall* and Benjamin D. Cobb Molecular Sensing plc, Unit 3, Challeymead Business Park, Bradford Road, Melksham, Wiltshire SN12 8LH, UK Received August 31, 2001; Revised and Accepted October 9, 2001 ABSTRACT energy transfer, mass spectrometry or direct sequencing of PCR products have also been shown to be effective tools for Allele-specific amplification (ASA) is a generally SNP detection (2). Implementation of these methods can applicable technique for the detection of known require specific reagent synthesis and downstream purification, single nucleotide polymorphisms (SNPs), deletions, making them complicated and expensive. insertions and other sequence variations. Convention- However, assays based on PCR for SNP diagnosis have ally, two reactions are required to determine the broad potential in clinical diagnostics because of their inherent zygosity of DNA in a two-allele system, along with simplicity and potential low cost. Established PCR-based significant upstream optimisation to define the methodologies include certain ligation assays (3,4), genetic bit analysis (5), arrayed primer extension (6) and restriction specific test conditions. Here, we combine single tube fragment length polymorphisms (7), and use of sequence- bi-directional ASA with a ‘matrix-based’ optimisation specific fluorescent probes (8). strategy, speeding up the whole process in a reduced PCR has also been adapted for the detection of well-characterised reaction set. We use sickle cell anaemia as our model SNPs using allele-specific amplification (ASA); oligonucleo- SNP system, a genetic disease that is currently screened tides complementary to a given DNA sequence except for a using ASA methods. Discriminatory conditions were mismatch at their 3′-hydroxyl residue will not function as rapidly optimised enabling the unambiguous identi- primers in the PCR under appropriate conditions. A typical fication of DNA from homozygous sickle cell patients ASA test consists of two complementary reactions, each (HbS/S), heterozygous carriers (HbA/S) or normal containing a common primer, an allele-specific primer and Taq DNA in a single tube. Simple downstream mathematical DNA polymerase lacking 3′→5′ proofreading activity. The analyses based on product yield across the optimisation first reaction contains a primer specific for the normal (or wild- set allow an insight into the important aspects of type) DNA sequence and refractory to amplification from mutant DNA at a given locus. Similarly, the second reaction priming competition and component interactions in contains a mutant-specific primer unable to amplify wild-type this competitive PCR. This strategy can be applied to DNA. Molecular conformation is achieved by analysis of the any polymorphism, defining specific conditions resulting PCR amplicon profiles. A normal individual will using a multifactorial approach. The inherent generate product in the first reaction only; a heterozygote simplicity and low cost of this PCR-based method amplifies products in both reactions; and a homozygous validates bi-directional ASA as an effective tool in mutant individual does so only in the mutant-specific reaction. future clinical screening and pharmacogenomic Internal control amplification is necessary, providing a research where more expensive fluorescence-based positive control for the PCR test (9). approaches may not be desirable. More recently, single tube adaptations have been introduced for known SNP detection in a two-allele system including competitive oligonucleotide priming (10), multi-coloured INTRODUCTION fluorescent oligo labelling systems (2), tetra-primer PCR (11) and overlapping PCR strategies (12,13). The latter both use bi- The successful implementation of single nucleotide polymorphism directional primer arrangements in which common ‘outer’ (SNP) screening for clinical research and the administration of primers define the size of each allele-specific fragment medicine tailored to an individual’s genotype, so-called allowing simple identification using electrophoretic methods. ‘personalised medicine’, lies in the ability to deliver highly reliable, accurate and inexpensive assays that can discriminate Maintaining the integrity of all ASA techniques, particularly between key DNA polymorphisms. Biotechnology and for diagnostic application, requires the optimisation of many pharmaceutical companies have addressed this challenge with experimental parameters. The literature suggests that a magnesium the development of methodologies including single-strand and oligonucleotide titration is sufficient to obtain discriminatory conformation polymorphism analysis, heteroduplex analysis conditions. Generally, further mismatches are incorporated at and denaturing gel electrophoresis. Each of these techniques the 3′-end of the allele-specific primer to weaken hydrogen relies on conformation-induced gel mobility changes (1). bonding between the primer and template, increasing the like- Simple homogeneous assays based on fluorescence resonance lihood of discrimination (14). However, these conventional *To whom correspondence should be addressed. Tel: +44 1225 706245; Fax: +44 1225 700056; Email: christy.waterfall@molecular-sensing.com e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 2 OF 8 approaches are simplistic considering the complex component Bi-directional allele-specific PCR amplifications interactions in a PCR reaction and prove time consuming and WT-AS (5′-ATG GTG CAC CTG ACT CCT GA-3′)and expensive. This has limited the applicability of PCR-based WT-CP517 (5′-CCC CTT CCT ATG ACA TGA ACT-3′) SNP diagnosis for routine clinical application. In particular, bi- were designed for amplification of a 517 bp fragment from the directional systems require significant optimisation due the normal β-globin gene (wild-type primer set). MUT-AS (5′-CAG effects of primer competition caused by multiple primer sets in TAA CGG CAG ACT TCT CCA-3′)and MUT-CP267(5′-GGG the PCR amplification. TTT GAA GTC CAA CTC CTA-3′) were designed for ampli- Here, we modify current systems at a number of levels with fication of a 267 bp fragment from homozygous mutant DNA the application of a multifactorial optimisation procedure to (HbS/S) conferring sickle cell disease (mutant primer set). All single tube bi-directional ASA. Our strategy enables the rapid reactions were performed in a volume of 25 µ l containing identification of component ratios favouring ASA in a greatly 10 ng total human genomic DNA template, 10× PCR reaction reduced experiment set using unmodified oligonucleotides. –1 buffer, 1 U µ l Taq DNA polymerase and water. The wild-type This low cost optimisation strategy combined with a single primer set, mutant primer set, MgCl and dNTP concentrations tube genotyping system introduces improvements to PCR-based are shown in Table 1 and were aliquoted according to an SNP screening. orthogonal array (Table 2) (17). Amplification was performed using an MJ Thermal Cycler (GRI, UK) through an initial SNP model denaturation at 95°C for 2 min, followedby30cyclesofdena- We demonstrate our system by detecting the single base pair turation at 95°C for 30 s, primer annealing at 60°C for 30 s and mutation that causes the autosomal recessive disease sickle cell extension at 72°C for 35 s. A small blind trial was performed anaemia (GenBank accession no. M34058). A single A→T in which 14 human genomic DNA samples of mixed zygosity transversion in the sequence encoding codon 6 of the human were used as substrate DNA under discriminatory conditions β-globin gene causes the substitution of amino acid glutamine that were considered optimal (reaction 9, Tables 1 and 2). The by valine, forming a mutant globin chain termed HbS. Haemo- exact template concentration of each sample was not specified, globin S is freely soluble when fully oxygenated, but under –1 varying between 5 and 25 ng µ l . conditions of low oxygen tension the erythrocytes assume an irregular ‘sickle’ shape, leading to aggregation and haemo- Analysis of PCR products lysis. The sickled erythrocytes become trapped in the micro- An aliquot (10 µ l) of each PCR product was loaded with 6× circulation, depriving organs of essential oxygen, causing pain loading buffer (Promega) onto a 2% agarose gel and run in and chronic anaemia (15). Homozygous HbS is a serious 0.5× TBE buffer at 30 mA constant current. Amplification haemoglobinopathy found almost exclusively in the Black products were visualised by UV transillumination (305 nM) population. About 8% of American Blacks are carriers and ∼0.2% are affected. In the UK sickle cell disease affects ∼10 000 people (actionresearch.co.uk/camp/sick.html). Hetero- Table 1. Concentration levels of components used in a multifactorial array for zygotes (sickle cell carriers) are clinically normal, although the identification of discriminatory reaction conditions for bi-directional ASA their red cells will sickle when subjected to very low oxygen pressure in vitro (genelink.com/technical/sickle.html). PCR-based Level A B C techniques for prenatal diagnosis of sickle cell disease were WT-AS/WT-CP517 (µ M) 0.2 0.4 0.8 introduced in 1989 following the invention of the amplification MUT-AS/MUT-CP267 (µ M) 0.2 0.4 0.8 refractory mutation system (9) and are still the method of MgCl (mM) 0.5 1 2 choice in clinical research laboratories (16). A single tube 2 adaptation of conventional ASA would add significant benefits dNTPs (mM) 80 120 200 to routine testing for sickle cell disease. Table 2. An orthogonal array for testing four reaction parameters, each at MATERIALS AND METHODS three different levels Reagents Reaction WT-AS/ MUT-AS/ MgCl dNTPs PCR reagents (Taq DNA polymerase, HS TaQUANT-OFF, WT-CP517 MUT-CP267 10× Mg-free Taq buffer, 25 mM MgCl ) were obtained from 1 AAAA Q-Biogene (CA), except for dNTPs (2 mM each dATP, dCTP, 2 A BBB dGTP and dTTP), which were from Promega (UK). Primer selection was accomplished using the PrimerCalc software 3 A CCC package (Q-Biogene). For standard gel electrophoresis, molecular 4B A B C biology grade agarose was used (Q-Biogene) in 0.5× TBE 5 BBCA buffer (45 mM Tris, 45 mM borate, 1 mM Na EDTA, pH 8.3). 6 BCA B Double deionised water (18 MΩ ) was used for all buffers and solutions for electrophoresis and PCR amplifications. 7C A C B Homozygous HbS human genomic DNA and heterozygous 8 CBA C HbA/S DNA were a generous gift from the National Haemo- 9 CCBA globinopathies Reference Service (Oxford, UK). PAGE 3 OF 8 Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 Figure 1. Schematic representation of bi-directional allele-specific PCR. The wild-type primer set is shown in green and the mutant primer set in pink. Three possible fragments can be formed depending upon substrate DNA zygosity, as well as enzyme system and cycling parameters. The green amplicon defines the allele-specific product formed when only the wild-type primers are extended and the pink amplicon defines the allele-specific product formed when only the mutant primers are extended. A heterozygote is identified by the presence of both PCR products. The target region amplified by non-allele-specific outer primers is shown in pale grey. Our selected enzyme system was inefficient at amplifying longer sequences and in the presence of competition from a smaller fragment insufficient target region was amplified for visualisation on a stained gel. after staining with SYBR Gold (Molecular Probes, UK) for WT-AS has a single A nucleotide at the 3′-end, matching the T 30 min. The agarose gels were photographed and analysed by nucleotide on the non-coding strand of the normal β-globin gel imaging software to obtain maximum band optical density gene. Primer MUT-AS has a single A nucleotide at the 3′-end, measurements as a representation or ‘score’ of specific product matching the T base on the coding strand of the sickle cell yield. gene. The wild-type outer primer (WT-CP517) was positioned 517 bp downstream on the opposite strand of WT-AS. The mutant outer primer (MUT-CP267) was placed 267 bp RESULTS upstream of MUT-AS. Unlike normal ASA, this assay has an inherent PCR control, as at least one allele-specific fragment Experimental design should be yielded per reaction under optimised conditions. A schematic of bi-directional ASA is shown in Figure 1. A This negates the need for a distant pair of primers to direct the wild-type allele-specific primer is designed as complementary amplification of an internal control fragment. to the non-coding strand of the wild-type allele terminating at Optimisation strategy for discrimination between the the known polymorphic site. This primer will be extended in normal and sickle cell alleles the presence of wild-type DNA only under appropriate PCR conditions. A further allele-specific primer is included in the Inclusion of multiple allele-specific primer sets in a single reaction, designed as complementary to the coding strand of reaction tube requires the consideration of many experimental the known mutant allele, again terminating at the SNP site. factors. For typical multiplex reactions, the key parameter for This primer will be extended in the presence of the mutant optimisation is primer concentration. At equimolar concentrations, allele only, as it is refractory to extension of wild-type DNA amplification by the two primer sets may be weak or strong at due to a single base pair mismatch at the 3′-end. Two ‘outer a given locus, an event dependent upon primer stabilities, primers’ are placed on the opposite strand at a pre-defined binding efficiencies, priming competition, product size and number of bases from their allele-specific pair in order that concentration of other reaction components (18). Normally, molecular confirmation can be achieved by analysis of the primer ratios are individually optimised for each reaction, resulting amplicon profiles using gel-based methods. A differ- however, an adjustment in primer concentration has associated ence of at least 100 bp is preferable to increase electrophoretic implications considering the complex interactions between separation. other components in a PCR. Unique to this assay, each primer Taq DNA polymerase will extend both outer primers in all is complementary to a different part of the DNA sequence, reactions as they are non-allele-specific, but the yield of the introducing additional variability to the reaction. large amplicon spanning the whole target region is subject to MgCl and dNTPs have been shown to affect the efficiency variation according to the enzyme system and cycling of priming and extension by altering the kinetics of association parameters employed. The HotStart enzyme system used in and dissociation of primer–template duplexes at annealing and this study was less efficient at amplifying larger fragments and extension temperatures. These components also alter the in the presence of competition from a smaller fragment did not efficiency with which the polymerase recognises and extends amplify sufficient DNA for the larger band spanning the whole such duplexes. Of particular relevance to allele-specific PCR target region to be visualised on an agarose gel. methods, the concentration of MgCl and dNTPs required for We designed two 20/21mer allele-specific primers, WT-AS optimal amplification depends largely on the target and primer and MUT-AS, complementary to the normal β-globin or sickle sequences. The presence of excess magnesium in a reaction cell genes, respectively, arranged in a bi-directional orientation may result in the accumulation of non-specific amplification such that both primers terminate at the mutation site. Primer products and insufficient concentrations reduce product yield. e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 4 OF 8 Figure 3. Gel image data from a small blind trial. Fourteen human DNA samples of varying zygosity were selected and amplified under conditions specified by reaction 9 (Table 2). Diagnosed template zygosity is shown on the gel image. Three normal samples were identified by amplification of the 517 bp product only. Six heterozygous samples were identified by amplification of both the 267 and 517 bp fragments. For heterozygotes, the 267 bp band was slightly more intense than the 517 bp band, likely due to staining disparity between optimisation sets A and B. Five homozygous sickle cell samples were identified by amplification of the 267 bp amplicon only. Figure 2. Gel image data of PCR amplifications. Outer lanes contain 100 bp ladders. Reactions 1–9 according to the matrix (Table 2) are labelled accordingly. Set A represents reactions in which wild-type template DNA was the substrate. amplified in all cases using our enzyme system and cycling Set B represents reactions in which HbS/S template DNA was included. The 517 and 267 bp allele-specific fragments were generated in each set, respectively. parameters. Product yield varied across each reaction set. Table 3 represents maximal Diagnostic conditions were selected by qualitative analysis optical density (OD) of the amplicon bands, used as a representation of product of PCR yields from set A (WT DNA) and set B (HbS/S DNA). yield/reaction score. Optimal conditions for zygosity determination are those that amplify each allele-specific amplicon at similar efficiencies, as measured by band intensity from the gel. In this study, the conditions and interactions of reaction 9 (Table 2) satisfied It has also been demonstrated that dNTPs quantitatively bind these criteria, producing 517 and 267 bp bands with maximum 2+ Mg , so that any modification of dNTP concentration will OD values of 160.67 and 165.93 units, respectively. The optimal require a compensatory adjustment of MgCl (17). Additionally, nature of these conditions was confirmed unambiguously in a 2+ Taq polymerase requires free Mg as a cofactor so any excess small blind trial of 14 DNA samples of different zygosity of dNTPS can have a detrimental effect on product yield by (Fig. 3). Reactions generating the 517 bp fragment identified a chelating cofactor ions (19). Standard approaches to multiplex normal genotype. Amplification of the 267 bp fragment or allele-specific PCR optimisation are unable to account for confirmed a homozygous mutation and sickle cell carriers these multiple interactions, as they are largely based on trial- (HbA/S) were identified by the amplification of both the 517 and and-error strategies. 267 bp bands, diagnostic of heterozygous inheritance. The We applied a multifactorial optimisation strategy in which robustness of the optimised reaction conditions was demon- optimal component concentrations for allele discrimination strated by the ability to amplify from a range of unknown template were identified in a single trial with few reactions. The wild- concentrations without compromising the allele specificity of the type primer set, mutant primer set, MgCl and dNTPs were system. selected as four critical parameters, the interactions of which Effect of each parameter on specificity are most likely to affect the outcome of bi-directional PCR. The concentrations of these components were varied according A quantitative analysis was performed in which the actual product yield for each reaction was used to estimate the effects to an orthogonal array and differences in product yield across of individual components on amplification in bi-directional the reaction set were used for two purposes, qualitatively in the PCR. Signal-to-noise (SNL) ratios were calculated using the identification of diagnostic reaction conditions and quantitatively maximal optical density (max OD) of the allele-specific PCR to study the individual effect of each parameter on the assay. products (Table 3) to represent product yield (y) in the quadratic Optimal conditions for zygosity determination loss function, n 2 SNL = –10 log(1/n 1/y ) Figure 2 shows gel image data from reaction sets A and B. Both i= 1 i sets are identical in component ratios and cycling conditions, assuming that variable y denotes the performance of the system containing normal and HbS/S DNA templates, respectively. measured for a particular combination of factor settings in a Although product yield varied considerably across the reaction given experiment (total of n repeated measurements per sets, only the appropriate allele-specific amplicon was generated experiment) (17). Reactions giving non-allele-specific fragments according to template zygosity. Set A amplified the 517 bp or no amplification at all were given a score of 1. The effects of amplicon and set B generated the 267 bp fragment only. The reaction components on the SNL ratio were determined by larger fragment spanning the whole target region was not regression analysis for both normal and HbS/S DNA templates PAGE 5 OF 8 Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 Table 3. Maximal optical density (max OD) measurements of band intensities from amplification sets A and B in Figure 2 Reaction 1234 567 89 Set A, max OD wild-type 5.13 90.83 80.30 63.40 63.70 1 132.27 1 160.67 Set B, max OD mutant 113.6 172.53 169.20 108.73 161.13 183.67 144.80 1 165.93 All values are in arbitrary units. Figure 4. Normal DNA. Effects of reaction components on the SNL ratios for the amplification of a wild-type 517 bp amplicon from normal DNA. Increased SNL value represents a larger effect on product yield than a low SNL. (Figs 4 and 5, respectively), using calculated SNL values The regression profiles of primer sets generated by normal and mutant DNA samples were quite different, demonstrating (Table 4). differences in primer stabilities, efficiencies and competition. We were able to demonstrate the significance of priming For normal DNA the aim is to direct amplification of the competition and component interaction on the outcome of 517 bp fragment only. The lower concentration of the wild- bi-directional ASA reactions. The SNL analysis revealed that type primer set tested gave the largest SNL value relating to the optimal magnesium requirement for the generation of product yield. Intermediate concentrations decreased the SNL maximal product yield without compromising specificity was and the upper concentration showed an incremental increase in 1.5 mM. This value did not vary according to template SNL. An identical profile was observed with the mutant primer zygosity, which is in accordance with the multiple roles of set, optimally redundant in the presence of wild-type DNA. It magnesium in PCR. Likewise, the optimal concentration of is possible that at lower concentrations of both primer sets, dNTPs in sets A and B were similar, with lower levels being i.e. conditions of minimal competition, the wild-type pair is able favoured. The quantitative binding of magnesium ions by to direct allele-specific amplification of the 517 bp fragment. dNTPs may account for this observation, as excess dNTPs may When the mutant allele is present, the aim is to direct ampli- chelate the magnesium cofactor ions of Taq, decreasing fication of the 267 bp amplicon only. The optimal concentration processivity or simply reducing the yield and specificity. of the mutant primer set was predictably at the upper level. e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 6 OF 8 Figure 5. HbS/S DNA. Effects of reaction components on the SNL ratios for the amplification of a mutant 267 bp amplicon from HbS/S DNA. An increased SNL value represents a larger effect on product yield than a low SNL. Likewise, the inclusion of lower concentrations of the wild- a report by Kwok et al. (20) in which only A:G, G:A and C:C type primer allowed maximal optimal amplification of the mismatches reduced overall PCR product yield significantly. mutant fragment, as the wild-type set thereby exhibited least Other combinations required the incorporation of additional competition with the mutant primer set. We have observed mismatches near the 3′-end to increase discrimination. This similar trends using other SNP models, confirming that primer difference may be due to improvements in the process of oligo- interactions have a significant role in determining the outcome nucleotide synthesis and the use of recombinant Taq DNA of bi-directional amplifications. polymerases. In particular, the use of hot start systems was seen to improve the specificity of our allele-specific amplifications. The optimisation system used and the increased priming competition DISCUSSION in our bi-directional assay will have positively contributed to the absolute discrimination. We have successfully demonstrated the use of bi-directional The implementation of an orthogonal array in our optimisation ASA for single tube genotyping of the SNP responsible for system generated product yield data from two small reaction sickle cell anaemia. Our simple matrix optimisation strategy sets that served a dual purpose in our experiments. Firstly, we offers advantages over conventional methods, allowing the determination of specific test conditions in a greatly reduced were able to use product yield data qualitatively to identify reaction set. Combined with single tube bi-directional allele- conditions under which both allele-specific fragments were specific PCR, rapid diagnosis of normal, homozygous and amplified to similarly maximal efficiencies according to heterozygous template DNA can be achieved simply and template zygosity. Secondly, reaction ‘scores’ were used applied to the screening of any known SNP. Bi-directional quantitatively in a simple regression analysis, allowing the ASA is based on the recent adaptation of ASA methods, using effects of ‘control factors’ influencing reaction performance to a single tube for the diagnosis of a two-allele system, differing be assessed simultaneously. This provided an insight into the from other protocols described in which modified oligonucleo- component interactions and priming competition that dominate tides or cycling parameters are generally required to maintain bi-directional allele-specific amplifications. specificity. The conditions selected for the blind trial were 0.8 µ Mfor We found that the A:A or T:T mismatches designated by the each primer from the wild-type and mutant sets, 1 mM magnesium sickle cell polymorphism were completely refractory to ampli- ions and 80 µ M dNTPs. This is unlikely to be in accordance fication under the optimised conditions. This is discordant with with optimal component levels calculated by SNL values PAGE 7 OF 8 Nucleic Acids Research, 2001, Vol. 29, No. 23 e119 because of the different objectives of each enquiry. The theoretical Table 4. Signal-to-noise ratios (SNL) for reaction components used for the amplification of (A) the 517 bp wild-type allele-specific fragment and (B)the target of the SNL analysis is to define conditions producing the 267 bp mutant allele-specific fragment, calculated using product yields from maximum yield of each product. This differs from the practical Figure 2 target of specifying diagnostic conditions under which each allele must be amplified specifically and at equal efficiencies Level enabling a heterozygous mutation to be identified without a SNL A B C gross over-amplification of one allele over another. Thus, to (A) Normal DNA template achieve this objective, sub-optimal conditions for the amplifica- tion of one allele may be combined with optimal conditions for WT-AS/WT-CP517 18.94 4.77 4.77 the amplification of another by a less efficient primer set, as in MUT-AS/MUT-CP267 18.95 4.77 4.77 this study. 2+ Mg 1.68 39.95 38.12 Two further variables that have an effect on reaction dNTP 18.94 4.77 4.77 outcome are the enzyme system and instrumentation used for (B) HbS/S DNA template amplification. The HotStart Taq DNA polymerase used in this study was extremely accurate, amplifying allele-specific products WT-AS/WT-CP517 43.12 42.93 4.77 with only minimal spurious products. Additionally, the MUT-AS/MUT-CP267 41.01 4.77 44.73 HotStart system was inefficient at amplifying larger fragments 2+ Mg 4.77 42.88 43.94 (>600 bp) when in competition with a smaller fragment, thus dNTP 42.94 44.32 4.77 the non-allele-specific fragment spanning the whole target region was not amplified in this case. In similar studies, this larger fragment has been designated an internal control amplification, avoiding the incorporation of a distant primer pair conventionally required for diagnostic tests (12). ACKNOWLEDGEMENTS However, any successful bi-directional ASA reaction will yield a fragment serving as an internal positive control for the We would like to thank Drs John Old and Chris Fisher at the PCR. Thus we favour our system where the processivity of Taq National Haemoglobinopathy Reference Service, Oxford, UK, is expended generating allele-specific products only. We tested for the kind gift of DNA from affected sickle cell individuals and sickle cell carriers. This work was supported by grants a conventional enzyme system in similar experiments and found that the larger fragment was generally produced at the from the University of Bath and Molecular Sensing plc. expense of the allele-specific fragment, with decreased specificity (data not shown). Application to other thermal cyclers may REFERENCES require further optimisation, since different reaction vessels 1. Brookes,A.J. (1999) The essence of SNPs. Gene, 234, 177–186. and heating/cooling methods can dramatically change the 2. Myakishev,M.V., Khripin,Y., Hu,S. and Hamer,D.H. (2001) High-throughput dynamics of a PCR. SNP genotyping by allele-specific PCR with universal energy-transfer Here, genotyping was simply based on size discrimination of labelled primers. Genome Res., 11, 163–169. 3. Lee,H.H. (1996) Ligase chain reaction. Biologicals, 24, 197–199. the resulting allele-specific PCR products, visualised by agarose 4. Bi,W. and Stambrock,P.J. (1997) CCR: a rapid and simple approach for gel electrophoresis. With the emergence of qPCR technologies, mutation detection. Nucleic Acids Res., 25, 2949–2951. offering rapid thermal cycling and on-line product detection, 5. Nikiforov,T.T., Rendle,R.B., Goelet,P., Rogers,Y.H., Kotewicz,M.L., our system can be adapted to detect PCR products based on Anderson,S., Trainor,G.L. and Knapp,M.R. (1994) Genetic Bit Analysis: a solid phase method for typing single nucleotide polymorphisms. their specific melting temperature using non-specific DNA-binding Nucleic Acids Res., 22, 4167–4175. dyes. This would address high throughput issues whilst 6. Tonisson,N., Kurg,A., Kaasik,K., Lohmussar,E. and Metspalu,A. (2000) maintaining the advantages of low cost and the use of simple Unravelling genetic data by arrayed primer extension. Clin. Chem. Lab. chemistries associated with the gel-based assay. Med., 38, 165–170. 7. Gu,W., Aguirre,G.D. and Ray,K. (1998) Detection of single-nucleotide The technique described could be modified to detect more polymorphisms. Biotechniques , 24, 836–837. than one SNP simultaneously with the combination of our 8. Kuklin,A., Munson,K., Gjerde,D., Haefele,R. and Taylor,P. (1997/98) optimisation strategy and accurate primer design. It has Detection of single-nucleotide polymorphisms with the WAVE™ DNA recently been suggested that the grouping and interaction of fragment analysis system. Genet. Test., 1, 201–205. 9. Newton,C.R., Graham,A., Heptinstall,L.E., Powell,S.J., Summers,C., several SNPs in haplotypes may be more important than single Kalsheker,N., Smith,J.C. and Markham,A.F. (1989) Analysis of any point SNPs in the identification of drug response and disease mutation in DNA. The amplification refractory mutation system (ARMS). susceptibility of individuals (21). The commercial applications Nucleic Acids Res., 17, 2503–2516. of this are widespread, as the development of reliable, cost- 10. Gibbs,R.A., Nguyen,P. and Caskey,C.T. (1989) Detection of single DNA base differences by competitive oligonucleotide priming. Nucleic Acids effective tests will be considerably more complex if haplotypes Res., 17, 2437–2448. need to be determined rather than single SNPs. 11. Ye,S., Humphries,S., and Green,F. (1992) Allele specific amplification by In conclusion, we have described various factors that have a tetra-primer PCR. Nucleic Acids Res., 20, 1152. combined influence on the outcome of bi-directional PCR 12. Liu,Q., Thorland,E.C., Heit,J.A. and Sommer,S.S. (1997) Overlapping PCR for bidirectional PCR amplification of specific alleles: a rapid amplifications. We have demonstrated the technique by identifying one-tube method for simultaneously differentiating homozygotes and accurate conditions for determining the zygosity of sickle cell heterozygotes. Genome Res., 7, 389–398. patients and carriers, a technique that can be simply applied to 13. Sasvari-Szekely,M., Gerstner,A., Ronai,Z., Staub,M. and Guttman,A. the detection of any known genetic mutation. (2000) Rapid genotyping of factor V Leiden mutation using single-tube e119 Nucleic Acids Research, 2001, Vol. 29, No. 23 PAGE 8 OF 8 bi-directional allele-specific amplification and automated ultrathin-layer 18. Henegariu,O., Heerema,N.A., Dlouhy,S.R., Vance,G.H. and Vogt,P.H. agarose gel electrophoresis. Electrophoresis, 21, 816–821. (1997) Multiplex PCR: critical parameters and step-by-step protocol. 14. Sommer,S.S., Groszbach,A.R. and Bottema,C.D.K. (1992) PCR Biotechniques, 23, 504–511. amplification of specific alleles (PASA) is a general method for rapidly 19. Linz,U., Delling,U. and Rubsamen-Waigmann,H. (1990) Systematic detecting known single-base changes. Biotechniques, 12, 1152. studies on parameters influencing the performance of the polymerase 15. Wu,Y., Ugozzoli,L., Pal,B.K. and Wallace,R.B. (1989) Allele-specific chain reaction. J. Clin.Chem. Clin.Biochem., 28, 5–13. enzymatic amplification of β-globin genomic DNA for diagnosis of sickle 20. Kwok,S., Kellogg,D.E., McKinney,N., Spasic,D., Goda,L., Levenson,C. cell anemia. Proc. Natl Acad. Sci. USA, 86, 2757–2760. and Sninsky,J.J. (1990) Effects of primer-template mismatches on the 16. Old,J., Petrou,M., Varnavides,L., Layton,M. and Modell,B. (2000) polymerase chain reaction: human immunodeficiency virus type 1 model Accuracy of prenatal diagnosis for haemoglobin disorders in the UK: studies. Nucleic Acids Res., 18, 999–1005. 25 years’ experience. Prenat. Diagn. 20, 986–991. 21. Davidson,S. (2000) Research suggests importance of haplotypes over 17. Cobb,B.D. and Clarkson,J.M. (1994) A simple procedure for optimising SNPs. Nature Biotechnol., 18, 1134–1135. the polymerase chain reaction (PCR) using modified Taguchi methods. Nucleic Acids Res., 22, 3801–3805.

Journal

Nucleic Acids ResearchOxford University Press

Published: Dec 1, 2001

There are no references for this article.