SSCP Analysis Using an Automated Electrophoresis System

                        SSCP analysis can be performed in an automated electrophoresis system, typically an automated DNA sequencer, which monitors the mobility of the fluorescently labeled DNA fragments during electrophoresis. PCR products can be fluorescently labeled in a number of ways. Target DNA regions can be amplified by forward and reverse primers labeled with the same single fluorescent dye or two dyes of different colors. 

                           Post-PCR labeling can also be done easily and is more cost-effective because unlabeled primers are  used. Internal labeling of PCR products is another option. The automated DNA sequencer can be gel-based or, more commonly now, capillary-based . Throughput can be increased by analysis in the same lane or capillary of multiple PCR fragments labeled with multiple colors. Mobility of DNA fragments can be standardized by inclusion in the test sample of an internal mobility standard that contains DNA fragments of known size and labeled with a color different from those of the fragments being analyzed. 

                          Electrophoresis can be performed at ambient temperature or above because most automated DNA sequencers are equipped with a built-in heating device, but not a cooling unit. One exciting development is microchip electrophoresis, i.e., performance of electrophoresis in microchannels. Microchip SSCP analysis can be finished within a few  minutes, and thus greatly reduces the analysis times by more than 100-fold when compared with conventional methods. This may revolutionize molecular genetic testing in diagnostic laboratories in the future. 

Gel Composition, Buffer System, and Electrophoresis System

                         Polyacrylamide gels are defined by two parameters: %T and %C. The %T refers to the total amount in grams of acrylamide and N,N'-methylene-bisacrylamide (a common crosslinker, usually abbreviated as bis) in 100 mL solution. The %C refers to the proportion of the total monomers (acrylamide plus bis) that is the crosslinker (bis). The %C can also be expressed in another format as the ratio of acrylamide to bis; for example, 2%C is equivalent to an acrylamide:bis ratio of 49:1. 

                         The probability of detecting sequence variations in a PCR product is higher if the %C is lower (9). Low levels of crosslinking produce large pores in the gels, and thus allow efficient separation of bulky single strand conformers. We use a nondenaturing gel of 10%T/1%C as a starting point in conjunction with a conventional vertical electrophoresis system (e.g., SE600 from Hoefer) and a medium-sized gel (e.g., 16 × 14 cm). 

                         Large-sized gels of 5%T and 1−2%C are also commonly used together with conventional electrophoresis systems for manual sequencing. Another commercially available gel matrix called Mutation Detection Enhancement (MDE) Gel is also widely used for SSCP analysis. It is a polyacrylamide-like matrix and is claimed to be very sensitive to DNA conformational changes. The buffer most commonly used in SSCP analysis is Tris-borate-EDTA buffer with an alkaline pH. 

                           However, low pH buffer (e.g., Tris-MES-EDTA, pH 6.3) can still maintain very high sensitivity of detecting sequence variations for PCR fragments up to 800 bp in length.

Size of PCR Products

                          The ability of detecting sequence variations in a PCR product decreases with increasing length of the products. It is recommended that the size of PCR products be less than 300 bp to increase the chance of picking up sequence variations in the fragments although sequence variations can sometimes be detected in fragments larger than 1,000 bp. 

                            The throughput of SSCP analysis can be increased by running in the same lane two or more small products of different sizes, or several small products generated by restriction digestion of a long PCR product. The gel must of course be large enough for the required separation.

Interpretation of SSCP Results

                         The interpretation of SSCP results is simple. What one is looking for is a variation in the SSCP banding patterns among different samples. The variation can be a band shift or some additional bands. Depending on such factors as gel composition and electrophoresis temperature, the variation can be very conspicuous in some cases, but subtle in other cases. 

                          Subtle changes in banding patterns may be difficult to recognize if the bands are very broad as in the case of sample overloading, which should thus be avoided. Clean and specific PCR products are also a prerequisite for easy interpretation of banding patterns. Representative samples with distinctive SSCP patterns can then be sequenced to define the underlying sequence variations. In this way, initially unknown sequence variations can be identified and defined. 

                          Alternatively, known sequence variations can be correlated with specific banding patterns if SSCP is used as a diagnostic method. If the DNA samples are from diploid organisms like human beings, then the banding patterns are additive. In other words, if one type of homozygote (e.g., A/A) has a particular pattern and another type of homozygote (e.g., G/G) has another pattern, then the heterozygote (A/G) has a banding pattern that is the summation of these two patterns. 

                            However, this additive property is not applicable to the analysis of the X chromosome in human males, DNA from mitochondria or chloroplasts (except in heteroplasmy), and DNA from bacteria (haploid organisms).

Outline of PCR-SSCP Analysis

                        PCR is used to amplify a DNA region to be analyzed. The PCR products are diluted in a loading solution that contains formamide and indicator dyes (e.g., bromophenol blue and xylene cyanol FF). The diluted PCR products are heated to over 90°C for a few minutes to denature the products into single strands and then cooled immediately in ice water. 

                        High concentration of formamide (a chemical denaturant) in the loading solution and immediate cooling are required to keep a sizable proportion of the products in single strands. The denatured PCR products are then loaded onto a nondenaturing polyacrylamide gel (i.e., without chemical denaturants). Samples diluted in formamide are denser than the buffer and will sink to the bottom of the wells. Separation of the single strands is achieved by electrophoresis. 

                         The duration of electrophoresis depends on the gel composition, voltage applied, buffer/gel temperature, and the size and base composition of the PCR products. After electrophoresis, the DNA bands are visualized by silver staining or, less commonly, SYBR Green II. Though less popular now, PCR products can also be radioactively labeled and the bands detected by autoradiography. Banding patterns of samples are compared. 

                          The presence of different banding patterns among samples indicates that sequence variations exist in the DNA sequence amplified by the two PCR primers. Silver-stained gels can be dried in a gel dryer and the dried gels kept for permanent records if so desired.

Principle of SSCP Analysis

                         DNA molecules have two strands intertwined with each other in antiparallel direction with the base-pairing of A with T and C with G. The two strands are complementary to each other, but not identical. The two strands of a PCRamplified product can be separated into single strands by heat. Each single strand can coil around itself to form a 3-dimensional structure (or conformation) through intramolecular (or  intrastrand) hydrogen bonds. 

                          This conformation is dependent on the length of the strand and its base composition. The two complementary strands may form different conformers because they are not identical. If the DNA fragment contains a single base change, there are four different single strands on denaturation, which may form four different conformers. 

                          These conformers may have different 3-dimensional sizes and shapes, and hence migrate at different speeds in a polyacrylamide gel. Thus, different samples may give different banding patterns because of the presence of a base change in the amplified DNA region.

Types of Methods for Detecting Sequence Variations

                        There are two broad groups of methods for identifying DNA sequence variations: screening (or scanning) methods and diagnostic methods. From a technical point of view, sequence variations may or may not be known to exist in a stretch of DNA in a given sample before the search is begun. Screening methods are used to search unknown sequence variations; for example, a sample from a patient with a certain genetic disease is screened for diseasecausing mutations in the different exons of a putative disease gene. 

                        Diagnostic methods are used to determine the genetic make-up (or genotype) of a sample for a known sequence variation at a known location; for instance, a pregnant woman is genotyped for her Rh(D) status at the RHD locus. Single strand conformation polymorphism (SSCP) analysis of DNA fragments amplified by polymerase chain reaction (PCR) can be used as a screening method, a diagnostic method or both in any single electrophoretic run, depending on the purpose of the experiments and the region of DNA sequences being examined. 

                          On the  basis of the size of the DNA sequences involved, sequence variations can be of small or large scale. Small-scale sequence variations involve a few basepairs (bp), such as base substitutions and small insertions/deletions. Large-scale sequence variations involve a large stretch of DNA sequences, and are exemplified by large insertions/deletions and gross gene arrangements. 

                             There are no definitive cut-offs between small-scale and large-scale sequence variations. However, small-scale and large-scale sequence variations, either known or unknown, do require different methods for detection.

Single Strand Conformation Polymorphism

                        Variations in DNA sequences underlie the differences among different members of the same species and also between different species. DNA sequence variations are usually known as polymorphisms if  the commonest allele is less than 0.99 in a given population. 

                         DNA polymorphisms are widespread in many different species, particularly in humans. Examples include single nucleotide polymorphisms (SNPs), microsatellites, minisatellites, small insertions/ deletions, and large insertions/deletions. DNA polymorphisms may not haveany phenotypic effect at the protein level or at the level of the whole organism. 

                         On the other hand, they are usually called disease-causing or pathogenic mutations if they cause a change in the phenotype and results in a disease status. The frequencies of individual mutations are usually not high because of selection pressure against such less favorable base changes. It is thus important to study DNA sequence variations in various branches of biological sciences. 

Applications of DGGE

                        Denaturing gradient gel electrophoresis is a method to identify small mutations (e.g., point mutations). The definition of a point mutation is the transition or transversion of one nucleotide into another. However, there are more types of small mutation such as deletions or insertions of one or more nucleotides that can be identified by DGGE as well. 

                         In fact, these mutations will cause a large difference in melting temperatures in both the homoduplexes and the heteroduplexes and can therefore be seen quite clearly on the gels. As mentioned earlier, DGGE products typically range from 200 to 400 bp, making DGGE well suited for analyzing exons in genomic DNA, although DGGE can be applied to RNA screening as well. However, RNA is more vulnerable to degradation than DNA and requires conversion into complementary (cDNA). Scanning for mutations in genes involved in hereditary disorders is therefore often done on genomic DNA for both diagnostic purposes and for research. 

                          For instance, DGGE is widely applied in the analysis of the various genes involved in hereditary colorectal cancer such as APC, MSH2, MSH6, MLH1, and so forth  Presymptomatic diagnosis is particularly important with a potentially lethal disease such as colorectal cancer that can be treated. Mutation analysis has revealed that in families with colorectal cancer, the mutation is often unique. Obviously, screening a family for an unknown mutation requires a technique, such as DGGE, that is tried and proven, particularly when the stakes are very high. 

                            However, for research purposes, reliability is important as well. Investigation into types of mutation requires that the screening will reveal almost all point mutations so that an unbiased analysis of the mutation spectrum can be made. 

Visualization of Mutations

                        To separate the homoduplexes and heteroduplexes, DGGE fragments are run on acrylamide gels. Gels with 9% acrylamide ensure sharp bands and are easy to handle. To pour these gradient gels, two types of stock solution are used: 9% 37.5:1 acrylamide/bisacrylamide in 0.5X TAE and the same stock solution with 7M urea and 40% formamide. The former is called 0% denaturant and the latter is called 100% denaturant.  

                        Gradients from 100% to 0% are rarely used because, in such a broad gradient, the denaturing points of the homoduplexes and heteroduplexes would probably be very close to each other; therefore, a range of 30% for the gradients is recommended. To select a urea/formamide gradient, the predicted melting temperature (Tm) of a product as obtained from the computer analysis is used in the empirical formula Tm × 3.2−182.4 = % denaturant. 

                        The melting point is positioned approximately in the center of the gel by simply adding and subtracting 15% from this calculated urea/formamide concentration to obtain the desired 30% gradient. Acrylamide gels with gradients of urea and formamide are poured using a simple gradient mixer that consists of two reservoirs that are connected at the base with a short tube. Such a system is widely used for pouring all types of gradient. 

Designing the PCR Products

                        The melting characteristics of PCR products screened for point mutations are crucial for DGGE. It is important that the fragment, when it reaches the critical point in the gel, denatures immediately, instead of slowly denaturing at one end and progressing with this process as the product runs deeper in the gel. Such a slow-melting process will result in fuzzy bands or smears, rendering mutation detection impossible. 

                         Because the melting characteristics are vital for success, primers to amplify the target should be chosen with great care. With this aim, special software that analyzes the melting curves of possible PCR products is used for primer selection. A number of programs are available for various platforms, either commercially or for free. There are websites where a sequence can be analyzed online as well. The experimenter will usually see a rather irregular melting curve when analyzing a target sequence. Attachment of a GC clamp of 40–60 nucleotides most often flattens this curve dramatically. 

                        The curve should be flat within a range of 1°C. Of course, the melting temperature around the GC clamp is very high. If attaching a GC clamp at one side does not flatten the curve, one should try attaching it to the other side, because for DGGE, it does not matter whether the GC clamp is attached to the forward or to the reverse primer. 

                        The selection process involves trying various combinations of forward and reverse primers to find products with a good flat curve and primers that will  work well together in a PCR. DGGE products typically range from 200 to 400 bp. It is difficult to  find  the  correct  melting  curve  for products  longer  than 400 bp. 

Denaturing Gradient Gel Electrophoresis

                        Denaturing gradient gel electrophoresis (DGGE) is a robust method for point mutation detection that has been widely used for many years. It is a polymerase chain reaction (PCR)-based method, the principle being the altered denaturing temperature of a PCR product with a mutation compared to the wild-type product. PCR performed on DNA of an individual with a point mutation in one of two genes will lead to a mixture of different products. 

                        PCR products from both the wild-type gene and the mutated gene will be formed. These are known as the homoduplex products. The difference in melting temperature between these two products, however, is subtle. Another type of product, heteroduplexes, consisting of a wild-type strand combined with a mutant strand of DNA, will also be formed during the last cycles of the reaction. The real strength of DGGE lies in the fact that the heteroduplex PCR products will have much lower melting temperatures compared to the homoduplex PCR products, because the heteroduplexes have a mismatch. 

                        To visualize the different melting temperatures of these homoduplexes and heteroduplexes, the products should be run on an acrylamide gel with a gradient of denaturing agents: urea and formamide. These denaturing agents alone are not sufficient. In addition, the gel should be run at a high temperature, usually 60°C. During electrophoresis, the PCR products will run through the gel as double-stranded DNA until they reach the point where they start to denature. Once denatured, the PCR products could continue running through the gel as single-stranded DNA, but the fragments have to remain precisely where they denatured. To achieve this, a so-called GC clamp is attached, to prevent complete denaturing. 

                     This GC clamp is a string of 40–60 nucleotides composed only of guanine and cytosine and is attached to one of the PCR primers. PCR with a GC clamp results in a product with one end having a very high denaturing temperature. A PCR product running through a DGGE gel will, therefore, denature partially. The GC clamp remains double stranded. The fragment will form a Y-shaped piece of DNA that will stick firmly at its position on the gel.

Forensic Applications

                        In forensic investigations, mitochondrial DNA (mtDNA) is used to obtain genetic information from forensic samples. The great abundance and stability of mtDNA facilitate the successful investigation of limited quantity of samples obtained from crime scenes. Hypervariable regions 1 and 2 (HV1/ HV2) of the displacement loop are usually examined for mtDNA analysis. 

                          Apart from DNA sequencing, the analysis can also be done by immobilized sequence-specific oligonucleotide probes, DGGE, SSCP, microarray, and mass  spectrometry. However, some limitations of these methods would influence the accuracy of the results. DHPLC has recently been used to screen the HV1 and HV2 regions of the mtDNA displacement loop. The  screening method is used to separate mixtures of DNA molecules obtained from body fluid mixtures. The target regions of mtDNA from DNA mixtures are amplified by PCR. 

                             Under the partially denaturing HPLC conditions, the homo- and hetero-duplexes are evaluated, and hence the mtDNA species from forensic sample mixtures are resolved and separated. DHPLC provides a rapid, accurate and cost-effective platform for forensic investigations.  

DNA Methylation Analysis

                        DNA methylation is the modification of DNA by the addition of a methyl group to the 5-position of cytosines. The process is implicated in gene regulation, genomic imprinting, embryonic development, and cell growth and differentiation. Alteration of DNA methylation may lead to diseases including cancer. 

                          As a result, it is important to investigate patterns of DNA methylation status. Different traditional techniques have been used in methylation studies, including sequencing of bisulfite-treated DNA, methylation-sensitive restriction enzymes and Southern blotting, methylation-sensitive enzymes and PCR amplification. However, these methods are labor-intensive and time-consuming. DHPLC provides an efficient alternative for rapid and reliable methylation detection. 

                           DHPLC analyzes the methylation-specific PCR products under partially denaturing HPLC condition or the PE products under completely denaturing HPLC condition. The DHPLC method is capable of distinguishing overall methylation profiles of differentially methylated regions of imprinted genes. The technique also allows the quantification of relative amounts of methylated and unmethylated molecules. 

Purification and Isolation of Nucleic Acids

                         Conventionally, nucleic acids are analyzed by gel electrophoresis for the purposes of separation, identification, and purification. However, the process of the gel-based analysis involves labor-intensive steps such as sample and gel preparations, sample loading, gel staining, and photographic processing. 

                        The DHPLC system has high resolving power and thus allows the automatic purification and isolation of nucleic acids. It has been demonstrated that dsDNA, ssDNA, and RNA can be separated, quantified, and then recovered by the fragment collector of the DHPLC system. Under nondenaturing conditions, dsDNA molecules such as PCR products and restriction fragments are separated. The isolated dsDNA fragments can then be collected for downstream applications such as sequencing and cloning. In purification and isolation of ssDNA, DHPLC can directly separate ssDNA from dsDNA under fully denaturing conditions (75°C). 

                         The purification is facilitated by using a tagged primer, which has a hydrophobic moiety such as a biotin group or fluorescein, in PCR. The hydrophobicity of the ssDNA generated by the tagged primer is increased and leads to the increased retention time in DHPLC analysis. As a result, the two ssDNA species from the dsDNA PCR products can be separated. DHPLC is a simpler and faster method of purifying ssDNA than other techniques involving a variety of analytical molecular biology procedures. Moreover, the fully denaturing conditions can be applied to the purification and quantification of mRNA from total RNA. 

                           In DHPLC analysis, RNA degradation and spurious transcription can also be detected, and hence the quality and integrity can be determined. DHPLC greatly improves the analysis and purification of RNA as compared to the conventional methods by simplifying the lengthy experimental procedures.

Quantification of Gene Expression

                         The DHPLC technology can be employed for accurate, absolute quantification of gene expression, which is estimated by the competitive reverse transcription (RT) PCR. Competitive RT-PCR is based on competitive coamplification of a dilution series of known concentrations of internal standard RNA (competitor) together with a constant amount of total RNA (target) in one reaction tube. 

                         The amplified RT-PCR products are verified by restriction fragment analysis. Fragments of expected sizes are then quantified by DHPLC analysis. The use of DHPLC in the quantification process offers a rapid, accurate, and automated measurement of gene expression. It is superior to the time-consuming and labor-intensive gel electrophoresis with the use of radiolabeled or fluorescent components. 

                           In a recent breast cancer study, the quantitative method identified that nine candidate genes were over-expressed in breast tumor cells. The method was also used for studying the differences in quantitative gene expression of α and γ sodium pump subunits of nephron segments from hypertensive rats. 

Microbial Analysis

                         In addition to the applications in human genetic studies, DHPLC has been used in the investigation of microbes because the method is cost-effective and time-saving. One application is the characterization of drug resistance in different bacterial pathogens. DHPLC was first applied to mutation screening in Staphylococcus aureus. 

                          The study showed that DHPLC provided a rapid detection platform of identifying the quinolone resistance alleles of gyrA, gyrB, grlA, and grlB genes. Similarly, mutations in the quinolone resistancedetermining regions of Salmonella enterica (gyrA, gyrB, parC, and parE) and Yersinia pestis (gyrA) could also be detected by DHPLC. Multiplex PCR together with DHPLC analysis has also been developed for the detection of plasmid-mediated ampC β-lactamase gene mutations in Gramnegative bacteria. 

                        For the investigation of the drug-resistance genes of Mycobacterium tuberculosis, mutations were detected by DHPLC in six genes including katG, rpoB, emB, gyrA, pncA, and rpsL, which are responsible for isoniazid, rifampicin, ethambutol, fluoroquinolone, pyrazinamide, and streptomycin resistance respectively. Besides, the DHPLC analysis system was also introduced to high-throughput bacterial identification. 

                         Through analysis of the prokaryotic 16S rRNA gene, different species could be differentiated by the corresponding distinctive DHPLC peak profiles. This application showed an overall specificity of 100% and a sensitivity of 91.7%. An enhanced version of the WAVE  HPLC system known as the WAVE Microbial Analysis System (Transgenomic) has been developed specifically for the purpose of microbial analysis. One recent application is the highthroughput typing of M. tuberculosis strains based on twelve loci of variable number of tandem repeat present in mycobacteria. 

                         Typing results based on nondenaturing HPLC showed 100% concordance with those generated by agarose gel electrophoresis. It should be noted that such applications do not fully utilize the benefits of nondenaturing HPLC because the amounts of PCR products are not measured. DHPLC can also be used for the detection and identification of fungal species in blood culture and fecal samples. This approach provides a simpler and quicker method than the culture-based approach. 

Quantitative Genotyping

                         In completely denaturing HPLC analysis, the PE products can be quantified by their absorbance at 260 nm. If the starting test sample is a mixture prepared by pooling many DNA samples in equal amounts (a process known as DNA pooling), the relative allele frequencies of a SNP in the DNA pool can be estimated by measuring the relative signal intensities of the two extension products of the DNA pool with reference to those of a heterozygote sample. 

                        Note that a SNP only has two alleles and their relative allele frequencies sum up to 1. This provides a very cost-effective method for estimating the relative allele frequencies of a large number of different samples. Conventionally, estimation of allele frequencies is achieved by genotyping all samples individually, and this approach is of course very expensive and time-consuming if the number of samples is very large (in the range of several hundreds, or more preferably over 1,000 in genetic association studies, see the following). 

Individual Genotyping

                         The PE reaction in combination with completely denaturing HPLC analysis offers a simple, robust and automatic genotyping platform because a single analytic condition can be used. The accuracy and sensitivity can be increased by using fluorescent-labeling method. Equipped with the DNASep-High Throughput (HT) column and the High Sensitive Detector, the 4500HT-HS model of the WAVE System (Transgenomic) is developed for high-throughput genotyping. 

                          The PE/DHPLC genotyping platform has been used in the genotyping of the mutations of the hemochromatosis gene and the β-globin gene for the diagnosis of hereditary hemochromatosis and β-thalassemia respectively. Two mutation sites (C282Y and H63D) of the hemochromatosis gene were simultaneously genotyped in a multiplex format including the PCR, PE reactions and DHPLC detection. 

                            For β-thalassemia, two different studies demonstrated the successful simultaneous genotyping of five common mutations within the β-globin gene. Amplicons containing the five common mutations were amplified, and the mutations distinguished by multiplex PE reactions followed by DHPLC analysis. 

                              This approach would help the development of diagnostic mutation panels for the diagnosis of β-thalassemia and other genetic diseases showing extensive allelic heterogeneity.

 

Direct Detection of Deletions and Duplications

                          Nondenaturing HPLC can be used to quantify PCR products. To allow for quantification of PCR products, the number of PCR cycles has to be around 25 so that the amount of product amplified is proportional to the initial gene copy number. 

                           This type of analysis is very useful for detecting gene rearrangements such as deletions and insertions. Applications are exemplified by the detection of large deletions of the X-linked dystrophin gene in Duchene muscular dystrophy and the demonstration of exon deletions and duplications in the RB1 tumor suppressor gene. 

                            With a wild type or normal chromatogram for comparison, homozygous deletion is indicated by the absence of a particular elution peak, and heterozygous deletion (or a carrier) by a peak of half height.

Mutation Detection

                        The most significant function of DHPLC is mutation detection. Partially denaturing HPLC is used to detect unknown sequence variations. Heteroduplexes are generated before analysis. This step is essential for the detection of X-linked mutations in males and homozygous mutations. The screening  throughput can be increased by mixing several test samples with 1 reference sample. 

                          As has been mentioned above, the ideal size of PCR products is 150–450 bp for detection of unknown sequence variations. Long DNA fragments tend to have more than 1 melting domain and require several column temperatures for complete screening of the fragment. 

                          There are occasions in which several sequence variations are found within a small DNA region in different chromosomes (i.e., in different individuals). These are well illustrated by the diverse mutations in the CFTR and HBB genes. Mutations in CFTR result in cystic fibrosis whereas mutations in HBB give rise to β-thalassemia or sickle cell disease. Distinct mutations in a PCR product usually give consistent distinct chromatograms. 

                           Therefore, partially denaturing HPLC can also be used to genotype known sequence variations, particularly known mutations, once their corresponding distinct chromatograms have been established. However, it is still possible that different mutations may share indistinguishable chromatograms. 

Data Analysis

                        After sample injection, the signal intensity of DHPLC profile is shown real-time on the computer screen. A DNA fragment appears as a peak in the chromatogram with the corresponding intensity and retention time. Interpretation of DHPLC data is based on the comparison between sample and reference chromatograms. 

                        For sizing of a DNA fragment, the sample peak is compared with a series of DNA fragments of known size or commercial size standards. Because the retention time increases with increasing DNA fragment size, the retention time becomes a measure of DNA fragment size. Accordingly, the size of a DNA fragment is determined by comparing the retention time obtained and the retention times of the size standards. In mutation detection, the elution profile of a test sample is compared with that of the wild type sample. 

                         Any altered elution profile such as shoulder peak and additional peak(s) implies the presence of heteroduplexes and hence mutation(s) in the DNA fragment. The DHPLC profile of a test sample showing any aberrant elution profile is then confirmed by direct sequencing. In allelic discrimination, the alleles are distinguished by the retention times of allele-specific PE products. 

                           Different sequence compositions of PE products give different retention times, which can then be verified by DNA samples of known genotypes. The genotype of a test sample can be determined on the basis of the number of elution peaks and their corresponding retention times. In addition, the detected DNA samples can be quantified by the peak height or area. Accordingly, the amount of DNA samples and the allele frequencies of DNA pools can be estimated.

DHPLC Analysis

                        Regardless of the operation mode, the DNA samples are placed in the cooled 96-well autosampler compartment of the DHPLC system. It is controlled by the software from the manufacturer of the system (e.g. the WAVEMAKER software for the WAVE DNA Fragment Analysis System from Transgenomic). 

                         All the analytical parameters are entered into the program, including the sample information (sample identity and volume to be injected), application types, column temperature (mode of HPLC), and the linear gradient profile (the percentage of acetonitrile in the mobile phase). A linear gradient of 1.8–2.0% per  minute at a flow rate of 0.9 ml/min is usually used. In general, the start- and end-points of the gradient are adjusted according to the size of the DNA fragments. 

                          For the detection of unknown mutation, the optimal column temperature and the linear gradient profile are calculated by the algorithm of the software. The gradient profile for each injection includes the column regeneration and equilibrium steps prior to the next injection. The eluted DNA fragments are shown as peaks in the chromatogram. The retention time and the intensity of the peaks (height or area) are determined by the software. 

Primer Extension Reactions

                         For allelic discrimination, the amplicons containing the target polymorphic site serve as templates for PE reactions. Prior to the PE reactions, the amplicons are purified by treatment at 37°C with exonuclease I and shrimp alkaline phosphatase in order to remove the unincorporated single strand oligo primers and deoxynuclotides respectively.

                         The enzymes are then inactivated at 80°C. After purification, the PCR products are mixed with Thermo Sequenase (GE Healthcare) and appropriate ddNTPs. The PE reactions are carried out in a thermal cycler with a universal thermal cycling condition that includes denaturation at 96°C, annealing at 43°C followed by extension at 60°C.

                         The single strand extended products are analyzed by DHPLC under completely denaturing condition.

Heteroduplex Formation

                         For mutation detection, the PCR product from a test sample is mixed with a homozygous reference PCR product in a 1:1 (v/v) ratio.

                         The mixed DNA fragments are denatured at 95°C and then cooled slowly for reannealing of the DNA strands by gradually lowering the temperature at a rate of 1°C per 20 seconds from 95°C to 25°C.

                          A single base pair difference between the test and reference fragments will produce two heteroduplexes and two homoduplexes. The mixture of DNA samples are then analyzed by DHPLC under partially denaturing conditions.

DNA Sample Preparation

                         Before DHPLC analysis, the target DNA fragments of interest are first amplified by PCR. Three DNA polymerases are suggested for use with DHPLC analysis, including Optimase® Polymerase, Maximase™ Polymerase and T-Taq™ Polymerase (Transgenomic). 

                          The functional integrity of the column might be adversely affected if other DNA polymerases are used. It is also important to note that under no circumstances should PCR additives such as mineral oil, dimethyl sulfoxide, and formamide be used. 

                           Otherwise, the column would be damaged. The amplified PCR products can be injected directly into the separation column if they are to be analyzed under the nondenaturing condition. Otherwise, they are further processed as described in the following sections on the basis of the purpose of the analysis. 

The Hardware of DHPLC

                         The hardware of DHPLC consists of components similar to those in conventional HPLC. Together with a gradient system, a pressure pump delivers buffers or solvents from buffer reservoirs in appropriate proportions to a column for the separation of DNA molecules. The buffer is the mobile phase while the column is the stationary phase. 

                          DNA samples are placed in an autosampler plate and injected into the column through an injection unit. The column is housed in a temperature-controlled oven. Under appropriate conditions, DNA molecules are separated in the column into individual components and then eluted from the column. The eluted components are monitored by a detector and the data collected in a computer system. A  UV detector is the most common option for measuring DNA molecules at a wavelength of 260 nm although a fluorescence detector can also be installed. 

                        The results are displayed as chromatograms or elution profiles showing the amounts and elution times for various components separated by the column. An optional fragment collector can be connected to the detector to collect the separated components into vials for further analysis and processing. The heart of the system lies in the column—the stationary phase. The most widely used column is DNASep (Transgenomic). 

                          DNA separating columns (e.g., Eclipse and Helix columns) from other manufacturers use a different type of stationary phase, and are less popular when compared with DNASep.  

Completely Denaturing HPLC

                        The mode of operation is completely denaturing if the column temperature is maintained between 70°C and 80°C. Under such a high temperature, DNA molecules are completely denatured and become single-stranded. Completely denaturing HPLC can differentiate single-stranded (ss) DNA (and RNA) molecules with the separation depending on both the length and the base composition of the single-stranded nucleic acid molecules. 

                         It can be used to analyze and isolate synthetic oligonucleotides and RNA molecules. More commonly, it is used to analyze the products from primer extension (PE) reactions (also known as minisequencing). In PE reaction, an extension primer is annealed immediately upstream of the polymorphic site. 

                          In the presence of a modified DNA polymerase (e.g., Thermo Sequenase from GE Healthcare) and appropriate unlabeled dideoxynucleotides (ddNTPs), the primer is then extended in a template-dependent manner. The allele-specific extension products with different sequence compositions are then detected and discriminated by completely denaturing HPLC. 

                          As such, completely denaturing HPLC provides a robust platform of medium throughput for genotyping known mutations or SNPs. The throughput can be increased by multiplexing several primer extensions in a single reaction.  

Partially Denaturing HPLC

                        The major application of DHPLC is to screen for unknown mutations and putative SNPs. To achieve this, the partially denaturing HPLC mode is used and the column temperature is maintained above 50°C, but below 70°C. The column temperatures vary with different DNA fragments to be analyzed and depend on the melting domains in the fragments. Before the DHPLC analysis, the DNA fragments are amplified by PCR. 

                        The PCR product from a test sample is mixed with a homozygous reference PCR product in equal volume. The mixed DNA fragments are denatured and allowed to reanneal by gradually lowering the temperature. With this process of heteroduplex formation, two dsDNA molecules that differ by a single base pair (e.g., A-T vs G-C) will give two heteroduplexes and two homoduplexes. Stability of the DNA duplexes determines the order of elution from the column: the more stable the duplexes, the longer the elution time. Heteroduplexes with mismatches are less stable than and are thus eluted before homoduplexes. 

                        The partially denaturing HPLC allows separation of homoduplexes and heteroduplexes produced as a result of even a single base difference between two otherwise identical dsDNA molecules at an optimized column temperature. With reference to a homozygous wild type control, any difference in the elution profile is indicative of the presence of a sequence variation. In fact, a 4-peak pattern is not frequently seen. The test DNA samples are then sequenced to confirm the presence and characterize the nature of the mutations. 

                        Nevertheless, false positive results can sometimes be obtained and an altered elution pattern is demonstrated for DNA fragments without sequence variations. The ideal size of PCR products is 150–450 bp for detection of unknown sequence variations although mutations have been detected in fragments as large as 1500 bp.  Long DNA fragments tend to have more than 1 melting domain and hence require several column temperatures for complete screening of the fragment. 

Nondenaturing HPLC

                         Nondenaturing HPLC is used for the sizedependent separation of dsDNA molecules, which depends on the length of the molecules, but not the base composition. After polymerase chain reactions (PCR), the amplified DNA fragments are directly injected into the DHPLC analysis system. The column temperature is maintained at 50°C and DNA molecules remain double-stranded. The concentration of the eluent (acetonitrile) is increased with time. 

                        The shorter DNA fragments will be eluted and detected first, followed by the longer fragments. The eluted DNA fragments are detected by the ultraviolet (UV) detector and the chromatographic peaks representing the corresponding DNA fragments are demonstrated on the computer screen. Similar to conventional gel electrophoresis, nondenaturing HPLC can accurately determine the size of the amplicons. Unlabeled products can be used for analysis even with a UV detector if the injection volumes are large. 

                        Use of unlabeled products reduces the cost of analysis. With small injection volumes, reliable quantification can also be achieved by adding dsDNA intercalation dye such as SYBR Green I and measuring the green fluorescence with a fluorescence detector. The dye is mixed with the DNA samples after elution from the column (postcolumn addition). The throughput of such analyses can further be increased by multiplex PCR in which several fragments of different sizes are amplified in the same tube. 

Modes of DHPLC

                        Three modes of operation are available for chromatographic analysis of nucleic acids, depending on the temperature of the column. They are nondenaturing, partially denaturing, and completely denaturing modes. 

                         Each mode of operation serves a different purpose in nucleic acid analysis. In brief, the nondenaturing condition is applied to the size-dependent separation of double strand (ds) DNA molecules. The partially denaturing condition is used for screening of putative SNPs or detection of unknown mutations. 

                          The third operation mode performed under completely denaturing condition is used for the analysis of short DNA fragments such as products of primer extensions and synthetic oligonucleotides, as well as RNA. 

                            Nucleic acid chromatography is most widely used under partially or completely denaturing conditions, and hence is frequently called denaturing HPLC (DHPLC). 

The Principle of DHPLC

                       DHPLC is a chromatographic technique for the separation and analysis of DNA fragments with different length and/or base composition. This technique can be applied for mutation detection in DNA fragments of 200–1000 bp in length with a high sensitivity (> 96%) and specificity (> 99%) (3). In addition, the high resolving power of DHPLC allows the distinction of short nucleic acid fragments such as primer extension products for allelic discrimination. 

                        DHPLC is based on a reversed phase system in which the stationary phase is nonpolar and the mobile phase polar (4,5). The hydrophobic stationary phase, DNASep column marketed by the company Transgenomic, is made up of alkylated nonporous poly(styrene-divinylbenzene) particles 2–3 μm in diameter. The polar mobile phase is acetonitrile (CH3-CN). However, DNA molecules are large anions because of the negative charges on the phosphate groups in the phosphate-sugar backbones of the DNA strands. 

                       Organic cations are required to allow interaction between DNA anions and the nonpolar stationary phase. The organic cation carries a positively charged portion to interact with the negative charge of DNA molecules on the one hand, and also a hydrophobic portion to interact with the nonpolar stationary phase on the other hand. The most commonly used organic cation is triethylammonium, (CH3CH2)3N+, in the form of triethylammonium acetate (TEAA). Thus, TEAA is used as an ion pairing reagent. The triethylammonium cations bind to the phosphate groups of DNA molecules and hence effectively coat the DNA molecules with a hydrophobic layer (the triethyl portion). 

                       The number of TEAA molecules coating the DNA molecules is proportional to the length of the DNA molecules and in turn determines the degree of interaction between the DNA molecules and the stationary phase. DNA molecules are eluted from the column in an increasing gradient of acetonitrile, which weakens the interaction between coated DNA molecules and the stationary phase. In other words, coated DNA molecules bind onto the stationary phase and will be released from the stationary phase when acetonitrile in the mobile phase reaches  a specific concentration. 

                         Thus, shorter DNA molecules are eluted earlier from the column than and hence separated from longer DNA molecules under the same buffer condition. In summary, the separation of DNA molecules is based on the principle of ion-pair reversed phase liquid chromatography.

DNA Sequence Variants

                       Since the completion of the Human Genome Project, it is clear that the human genome carries about 30,000 genes occupying less than 5% of the 3 billion base pairs (bp) of DNA sequence. Meanwhile, the human genome was also found to carry a very large number of sequence variations. On average, there are about 3 million sequence differences (0.1% of the whole genome) between any two unrelated individuals from a population. 

                        The analysis of DNA sequence variations is very important in genetic studies. Two broad types of DNA sequence variations are classified: polymorphisms and diseasecausing mutations. Polymorphisms refer to those sequence variations that are found in normal individuals and do not result in diseased phenotypes. They include single nucleotide polymorphisms (SNPs), microsatellites, and minisatellites. 

                        A SNP (pronounced as snip) is a sequence variation owing to change in a single nucleotide. Microsatellites and minisatellites are caused by variations in the number of repeat units that are themselves a short stretch of DNA sequence. They are very useful in research for locating the position of genes in our chromosomes, a process known as gene mapping. 

                        On the other hand, disease-causing mutations are those sequence variations that result in diseased phenotypes because they adversely affect the functions of the proteins, either qualitatively or quantitatively. The identification of mutations is important for the diagnosis of genetic diseases in clinical medicine.