pubmed.ncbi.nlm.nih.gov

Melt analysis of mismatch amplification mutation assays (Melt-MAMA): a functional study of a cost-effective SNP genotyping assay in bacterial models - PubMed

Melt analysis of mismatch amplification mutation assays (Melt-MAMA): a functional study of a cost-effective SNP genotyping assay in bacterial models

Dawn N Birdsell et al. PLoS One. 2012.

Abstract

Single nucleotide polymorphisms (SNPs) are abundant in genomes of all species and biologically informative markers extensively used across broad scientific disciplines. Newly identified SNP markers are publicly available at an ever-increasing rate due to advancements in sequencing technologies. Efficient, cost-effective SNP genotyping methods to screen sample populations are in great demand in well-equipped laboratories, but also in developing world situations. Dual Probe TaqMan assays are robust but can be cost-prohibitive and require specialized equipment. The Mismatch Amplification Mutation Assay, coupled with melt analysis (Melt-MAMA), is flexible, efficient and cost-effective. However, Melt-MAMA traditionally suffers from high rates of assay design failures and knowledge gaps on assay robustness and sensitivity. In this study, we identified strategies that improved the success of Melt-MAMA. We examined the performance of 185 Melt-MAMAs across eight different pathogens using various optimization parameters. We evaluated the effects of genome size and %GC content on assay development. When used collectively, specific strategies markedly improved the rate of successful assays at the first design attempt from ~50% to ~80%. We observed that Melt-MAMA accurately genotypes across a broad DNA range (~100 ng to ~0.1 pg). Genomic size and %GC content influence the rate of successful assay design in an independent manner. Finally, we demonstrated the versatility of these assays by the creation of a duplex Melt-MAMA real-time PCR (two SNPs) and conversion to a size-based genotyping system, which uses agarose gel electrophoresis. Melt-MAMA is comparable to Dual Probe TaqMan assays in terms of design success rate and accuracy. Although sensitivity is less robust than Dual Probe TaqMan assays, Melt-MAMA is superior in terms of cost-effectiveness, speed of development and versatility. We detail the parameters most important for the successful application of Melt-MAMA, which should prove useful to the wider scientific community.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1
Figure 1. Melt-MAMA approach to SNP genotyping.

A) Allele-specific (AS) primers are designed to detect the two alleles at a SNP locus on DNA template. These primers work in concert with a third common primer to generate a PCR amplicon. Depending on the template, only one of the AS primers will generate the amplicon in concert with the common primer. The ability to differentiate amplicon derived from each allele-specific primer is accomplished by the GC-clamp. The success of one versus the other is evaluated by melt-dissociation analysis, which is facilitated by the GC-clamp on the 5′ end of the one AS primer. B) Allele-specificity of primers is achieved by the incorporation of two independent design features that work nearly synergistically. First, the primer is designed in which the 3′ ultimate nucleotide directly overlaps the SNP locus in perfect complement. Second, allele-specificity is strengthened by engineering a mismatch at the antepenultimate position of the primer (MAMA primer) that uniquely differs in base composition to the template and alternate AS primer. This difference creates two mismatched nucleotides in the 3′ region of the primer for the non-allelic template, but only one difference in the correct allele template. This dramatically improves specificity by destabilizing the 3′ end of the non-allelic primer-template complex. Once a single round of PCR amplification has occurred, the antepenultimate difference is incorporated into the template for efficient subsequent rounds of amplification. Hence, the initial priming specificity is critical to accurate genotyping. Because the AS primers are added together in one reaction, they compete for priming, which also increases genotype fidelity over two independent reactions.

Figure 2
Figure 2. The principle of the Melt-MAMA PCR reaction.

Four different scenarios involving two alternate SNP allele templates (I & II vs. III & IV) and the interaction of Allele-Specific (AS) PCR amplification using MAMA primers. The annealing of AS-MAMA primers to their allelic templates is shown with one primer labeled with a 5′ GC-clamp (Ia) whereas the other is not (IVa). (Ib and IVb) Taq Polymerase extends from the 3′ matched AS-MAMA primer despite the antepenultimate destabilizing nucleotide. (Ic and IVc) The second PCR cycle replicates from a newly synthesized DNA template made in the previous step (Ib and IVb). With the synthesized DNA serving as the template, a perfect primer-template complex is formed eliminating the antepenultimate destabilizing mismatch observed in Iab and IVab. At PCR endpoint (Id and IVd), the amplicons generated from the 3′ matched AS-MAMA primer greatly outnumbers the amplicons generated by the mismatched AS-MAMA primer. Temperature-dissociation curve plots (Ie and IVe) of each AS-PCR product (Iabcd, IIab and IIIab, IVabcd), showing the fluorescent intensity and the rate of fluorescent intensity change (derivative) as a function of temperature. For each allelic template reaction (I & II vs. III & IV), the melt profiles (Ie and IVe) show only a single change in fluorescent intensity. This indicates the amplification of the perfect-matched amplicon and little to no amplification of the mismatched amplicon. The GC –clamp “labeled” amplicons dissociate at higher temperatures (∼3°C to 5°C) than non-GC amplicons. Nonproductive primer annealing is shown for an AS-MAMA primer (IIa) and a GC-clamp AS-MAMA primer (IIIa) binding with their respective corresponding mismatched templates. The lack of Watson-Crick base pairing at two 3′ positions (the antepenultimate nucleotide at the 3′ end) of the AS primer introduces instability at this region (IIb and IIIb). This prevents efficient extension by the polymerase, which retards or prevents product amplification (Ie and IVe).

Figure 3
Figure 3. Real-time PCR amplification and dissociation (melt) curve plots.

B. anthracis Melt-MAMA SYBR® Green assay targeting the A.Br.004 genetic clade. (A & C) The amplification of two alleles are illustrated for haploid template (Bacillus anthracis) possessing an ‘A’ polymorphic SNP-state or ‘G’ state. Each amplification plot represents a single PCR reaction containing a reverse “common” primer and two allele-specific MAMA primers. The AS-MAMA primers anneal to the same template target and then compete for extension across the SNP position. The polymerase-mediated extension rate of the 3′match AS-MAMA primer (perfect primer-template complex) exceeds that of the 3′mismatched MAMA primer (mismatched primer-template complex), thus the perfect match primer-template complex outcompetes the mismatched primer-template complex and dominates the PCR amplification. (B & D) Plots of the temperature-dissociation (melt) curve of the final PCR products for the two allele templates are shown next to their respective amplification plots (green arrows). Allele-specific PCR products are easily differentiated through temperature-dissociation (melt) curve analysis, which is conferred by the GC-clamp engineered on one of the AS-MAMA primer.

Figure 4
Figure 4. Assay optimization by altering the primer ratios.

Primer design does not always result in perfectly matched combinations and additional improvement can be achieved by altering primer ratios. The derivative plots of the temperature-dissociation (melt) curves of two PCR products amplified in a F. tularensis assay are shown. Each product was amplified from genomic template with one of the two allele SNP-states (A or G). A) Under equal primer concentrations, this Melt-MAMA mis-genotyped the ‘G’ allele gDNA template because the G specific primer was more efficient than the A specific primer. In this case, the mismatched primer for the ‘G’ SNP allele state outcompeted the perfect matched primer, resulting in the amplification of the incorrect allele-specific PCR product. B) Primer ratios were then altered so that the matched primer for the “G’ SNP state was four times more concentrated than the respective mismatched primer. Under these unequal primer concentrations, the ‘G’ allele gDNA template accurately genotyped without the disruption of the accurate genotyping functions of the ‘A’ allele gDNA template.

Figure 5
Figure 5. Melt-MAMA validation work flow.

This figure shows the sequential steps involved in validation of Melt-MAMA assays. After SNP selection (step I), Melt-MAMA are designed so that the amplicon is <100 bp in length (step II). Assays are screened across ancestral and derived DNA templates under 3 primer ratio conditions where 1∶1 represents equal primer ratio, 4∶1 represents ancestral primer 4x and derived primer 1x, and 1∶4 represents ancestral primer 1x and derived primer 4x (step III). Five outcomes are indicated (step III a–e). Based on the performance of B. anthracis, F. tularensis, and Y. pestis assays, 70–80% Melt-MAMAs accurately genotyped at one of the tested primer ratio condition (step IIIa). These successful assays were immediately screened on a diversity panel of DNA samples (step IV). The remaining assays (20–30%) resulted in one of the other four outcomes (step III b–e). Each outcome required additional specific validation steps to determine the optimal PCR conditions or the need to abandon the SNP altogether. Our overall design success rate increased from 46% to 87%.

Figure 6
Figure 6. TaqMan assay performance at a broad range of DNA amounts.

A B. anthracis TaqMan assay was used to screen the polymorphic ‘G’ or ‘A’ DNA templates (ancestral and derived, respectively) used in the B. anthracis Melt-MAMAs. (A & B) The respective amplification plots of genomic DNA of ‘G’ allele and ‘A’ SNP allele templates show the amplification curves of templates titrated in ten-fold serial dilutions and in replicates of eight. The number assigned to each amplification curve (1–9) denotes the DNA amount for the starting template. (C) Both genomic template types were of equal amounts. The consistency of amplification dropped with lower amounts of initial template, but the dilution levels containing less than a single copy (B. anthracis single copy ∼6 fg) was still detectable in some reactions. Detection of low-level DNA template by TaqMan assays is subject to stochastic sampling effects, which is predictable using a Poisson distribution .

Figure 7
Figure 7. Genotyping over a broad range of DNA amounts.

Melt-MAMA genotyping accuracy is not diminished at lower amounts of DNA, even at near-single copy for some assays. The sensitivity of individual melt-MAMAs varies greatly. This B. anthracis melt-MAMA (A.Br.003 clade) accurately genotyped DNA regardless of starting amounts as long as it was sufficient to support amplification. (A & B) The respective amplification plots of genomic DNA of ‘G’ allele and ‘A’ SNP allele templates show the amplification curves of templates titrated in ten-fold serial dilutions and in replicates of eight. The number assigned to each amplification curve (1–8) denotes the DNA amount for the starting template. (C) The temperature-dissociation (melt) curve derivatives for all initial template amounts are shown (numbers denote DNA amount shown). This panel illustrates that genotyping accuracy was not affected by DNA amounts, even at near-single copy levels. Similar to TaqMan assays, the detection of low levels of DNA template by Melt-MAMA is also subject to stochastic sampling effects (B. anthracis single copy ∼6 fg), which is predictable using a Poisson distribution .

Figure 8
Figure 8. Competition between specific and non-specific amplification at extremely low level DNA amounts.

With extremely low-level DNA amounts (<∼2 copies), stochastic, spurious, non-specific amplification could outcompete allele-specific amplification. The Bacillus anthracis melt-MAMA targeting the A.Br.003 clade stochastically amplified allele-specific product and non-specific spurious products at amounts of less than a single copy (∼0.19 copies). (A and B) The respective amplification plots of genomic DNA of ‘G’ allele and ‘A’ SNP allele templates show the amplification curves of templates at 1.15 ng and at two low level ten-fold dilution series (near-single copy and less than a single copy) in replicates of eight. The number assigned to each amplification curve (3, 8–9) denotes the DNA amount for the starting template. (C & D) The temperature-dissociation (melt) curve derivatives for the 1.15 ng and lowest template amounts are shown. This panel illustrates that genotyping accuracy was not compromised at DNA template amounts near single copy level, but spurious amplification was observed in dilution points below this level. This spurious amplification had a unique melt-profile that did not match the profile of either allele types (red arrow).

Figure 9
Figure 9. Genotyping over a broad range of DNA amounts.

Melt-MAMA sensitivity to low level DNA amounts varies greatly among different assays. B. anthracis melt-MAMA targeting the A.Br.006 clade accurately genotyped DNA at amount ∼19 copies. (A & B) The respective amplification plots of genomic DNA of ‘G’ and ‘A’ SNP allele templates show the amplification curves of templates titrated in ten-fold serial dilutions and in replicates of eight. The number assigned to each amplification curve (1–7) denotes the DNA amount for the starting template. (C & D) The temperature-dissociation (melt) curve derivatives for all initial template amounts are shown (numbers denote DNA amount shown). Panels C and D illustrates that genotyping accuracy is obtained across a broad range of DNA template amounts of ∼115 ng to 115 fg. Assay sensitivity to template is limited to ∼19 copies and above. An inherent characteristic of this assay is the occurrence of spurious amplification at extended cycle times (>35) in the absence of template as indicated by the NTCs. Melt-MAMAs detecting low level DNA are subject to stochastic sampling effects (B. anthracis single genome copy ∼6 fg), which is predictable using a Poisson distribution .

Figure 10
Figure 10. Competition between specific and non-specific amplification at low level DNA amounts.

The B. anthracis Melt-MAMA targeting the A.Br.006 clade stochastically amplified allele-specific product and non-specific spurious products at amounts of less than ∼19 copies. (A & B) The respective amplification plots of genomic DNA of ‘G’ allele and ‘A’ SNP allele templates show the amplification curves of templates at 1.15 ng and at two low level ten-fold dilution series (∼19 copies and near a single copy) in replicates of eight. The number assigned to each amplification curve (3, 7–9) denotes the DNA amount for the starting template. (C & D) The temperature-dissociation (melt) curves derivatives for the 1.15 ng and lowest template amounts are shown. This panel illustrates that assay sensitivity to template is limited to ∼19 copies and above. Below this template amount, spurious amplification is possible and difficult to differentiate from template-specific amplification.

Figure 11
Figure 11. Two-Locus (duplexed) Melt-MAMA development.

(A) A phylogenetic topology of the three subspecies of F. tularensis rooted with F. novicida. The SNP-signatures specific for the two pathogenic subspecies of F. tularensis (indicated by black bars) were incorporated into Melt-MAMAs. The table (right) indicates expected allele states (derived and ancestral) for strains from each F. tularensis subspecies represented on the topology; F. novicida would have the same allelic states as F. tularensis subsp. mediasiatica. (Bi–iv) Temperature-dissociation (melt) curves (derivative) of allele-specific PCR products from F. tularensis strains amplified in the duplexed assay (Type A and Type B). Each profile show two melt-curve peaks generated from a single F. tularensis strain. Each peak corresponds to the allele-specific PCR product for a single SNP-locus in the duplexed assay.

Figure 12
Figure 12. Allele-Specific MAMAs Based Upon Agarose Sizing.

Allele-specific SNP MAMAs from two pathogen species were transferred from a real-time PCR instrument (A & C; Melt-MAMA) to conventional PCR coupled with agarose gel electrophoresis (B & D; Agarose-MAMA) under identical assay conditions. Genotypes from Melt-MAMAs are determined by temperature-dissociation (melt) curves, whereas genotypes from Agarose-MAMAs are determined by the amplicon size differences.

Similar articles

Cited by

References

    1. Emahazion T, Feuk L, Jobs M, Sawyer SL, Fredman D, et al. SNP association studies in Alzheimer's disease highlight problems for complex disease analysis. Trends Genet. 2001;17:407–413. - PubMed
    1. Schork NJ, Fallin D, Lanchbury JS. Single nucleotide polymorphisms and the future of genetic epidemiology. Clin Genet. 2000;58:250–264. - PubMed
    1. Pirmohamed M, Park BK. Genetic susceptibility to adverse drug reactions. Trends Pharmacol Sci. 2001;22:298–305. - PubMed
    1. Van Ert MN, Easterday WR, Huynh LY, Okinaka RT, Hugh-Jones ME, et al. Global genetic population structure of Bacillus anthracis. PLoS One. 2007;2:e461. - PMC - PubMed
    1. Foster JT, Okinaka RT, Svensson R, Shaw K, De BK, et al. Real-time PCR assays of single-nucleotide polymorphisms defining the major Brucella clades. J Clin Microbiol. 2008;46:296–301. - PMC - PubMed

Publication types

MeSH terms

Substances