DNA Extraction Kits

We are currently using two different types of DNA extraction kit, depending essentially on the quantity of DNA required. The first (The QIAamp system, Qiagen, Crawley, West Sussex) is a robust method for obtaining ≈5 μg of DNA from 200 μl of whole blood. This method revolves around a high-affinity DNA binding matrix in a spin column, which can be used in a microcentrifuge or in an automated set-up. The DNA obtained is usually of high quality and sufficient for most routine analysis.

When larger amounts of DNA are required, perhaps for storage and more extensive analysis of important samples, we use from 3 to 10 ml blood in the Gentra Puregene Blood kit (Qiagen), which can yield upwards of 100 μg of DNA. This method depends on salting-out of proteins after sequential red cell and then white cell lysis, followed by isopropanol precipitation of the DNA.

Protocols are of course supplied with any kit method and will therefore not be given in detail here. For preparation of reagents, and a protocol for the manual extraction of DNA from blood using organic solvents, the reader is referred to previous editions of this book.

Principle

A DNA polymerase will synthesize the complementary strand of a DNA template in vitro. A stretch of double-stranded DNA is required for the synthesis to be initiated. This double-stranded sequence can be generated by annealing an oligonucleotide (oligo), which is a short, single-stranded DNA molecule usually between 17 and 22 bases in length, to a single-stranded DNA template. These oligos, which are synthesized in vitro, will prime the DNA synthesis and are therefore referred to as primers.

In the PCR, at least two oligos are used. One primes the synthesis of DNA in the forward direction or along the coding strand of the DNA, whereas the other primes DNA synthesis in the reverse direction or along the non-coding strand. The other components of the reaction are the DNA template from which the DNA fragment will be amplified, the four deoxynucleotide triphosphates (dATP, dTTP, dCTP and dGTP) required as the building blocks of the newly synthesized DNA, a salt buffer containing MgCl₂ and the thermostable DNA polymerase (Taq polymerase).

The first step of the reaction is to denature the DNA, generating single-stranded templates, by heating the reaction mixture to 95°C. The reaction is then cooled to a temperature, usually between 50°C and 68°C, that permits the annealing of the primers to the DNA template but only at their specific complementary sequences. The temperature is then raised to 72°C, at which temperature the Taq polymerase efficiently synthesizes DNA, extending from the primers in a 5′ to 3′ direction. Cyclical repetition of the denaturing, annealing and extension steps, by simply changing the temperature of the reaction in an automated heating block, results in exponential amplification of the DNA that lies between the two primers (Fig. 8.1).

Figure 8.1 The polymerase chain reaction. Cyclical repetition of three temperatures for denaturing, annealing and extending DNA, respectively, gives rise to an exponential amplification of a DNA fragment between two primer sequences directing DNA synthesis on opposite strands of the DNA template.

The specificity of the DNA fragment that is amplified is therefore determined by the sequences of the primers used. A sequence of 17–22 base pairs (bp) is statistically likely to be unique in the human genome and so primers of this length and longer will anneal at only one specific place on a template of genomic DNA. One general requirement of the PCR therefore is some knowledge of the DNA sequence of the gene that is to be amplified. The relative positioning of the two primers is another important consideration. They must anneal to complementary strands and must prime DNA synthesis in opposite directions pointing towards one another. There is also an upper limit to the distance apart that the oligos can be placed; fragments of several kilobase pairs (kb) in length can be amplified, but the process is most efficient for fragments of several hundred bp.

Reagents

Method

Optimal conditions for the reaction have to be derived empirically, with the magnesium concentration and annealing temperature being the most important parameters.⁷ The choice of buffer depends on the enzyme being used and the company will usually supply the most appropriate one. For genes with a high GC content, buffer II in combination with 10% DMSO may give better amplification. In most cases, a 25 μl reaction volume suffices. A blank control should always be included (i.e. a reaction without any template) to control for contamination. If the blank control yields a product, the analysis is invalid. A DNA sample that is known to amplify can also be included and this sample may then be used as a normal or positive control.

The risk of contamination cannot be overemphasized. It can be minimized by using plugged tips and having dedicated micropipettes and areas for each step of the analysis. The optimum cycling conditions need to be determined for each thermocycler. Specificity is often improved by ‘hot start’ PCR. This is achieved by setting up all the PCR tests on wet ice and transferring the tubes to the thermocycler once it reaches 95°C or by using an enzyme that only becomes activated when heated at 95°C for several minutes. In preparing a group of reactions, a premix solution is prepared that can be dispensed into microcentrifuge tubes, tube strips or PCR reaction 96-well plates to which the template DNA is added. When a particular PCR is to be performed repetitively over a period of time, it is helpful to prepare a large volume (e.g. 10 ml) of the reaction mixture (without DNA or Taq polymerase), aliquot it and store it at –20°C.

Modifications and Developments

The procedure described earlier is a guideline for setting up and checking a standard PCR amplification. As the test dictates, modifications can be used, such as the following:

Problems and Interpretation

If the amplification has been successful, a discrete fragment of the expected size is seen in an SYBR Safe-stained agarose gel in all samples, except where a blank control (in which the DNA template is replaced by water) is loaded. If a product is seen in the blank control, then one of the solutions has been contaminated. In this case, the experiment and all the working solutions must be discarded and the micropipettes must be cleaned. Cleaning micropipettes prior to the start of each experiment is highly recommended. To avoid contamination, setting up a master mix of all reagents is recommended before DNA samples are added to each tube.

Conversely, the absence of a fragment in all tracks indicates that the PCR has failed. This could occur for a number of reasons, the most obvious being the poor quality or omission of one of the essential reagents. The reaction may also fail if the magnesium concentration is too low (standard concentration 1.5 mM) or if the annealing temperature is too high. DNA quality is often one of the major reasons for failure. If one particular DNA sample repeatedly fails to amplify, then the sample should be re-extracted using Proteinase K (Sigma, Poole) and phenol and chloroform and reprecipitated in one-tenth volume of 5 mol/l ammonium acetate or 3M sodium acetate pH 4.8 and 2.5 volumes of ethanol. We have also found that for samples prepared using the Gentra method, passing the DNA through a Qiagen column substantially improves DNA quality and PCR efficiency. Another problem is the presence of non-specific fragments or just a smear of amplified product. This can occur if the magnesium concentration is too high or if the annealing temperature is too low.

Size of the PCR Product

Restriction Enzyme Digestion

Principle

Restriction enzymes (RE) cleave DNA at short specific sequences. Because many RE are available, it is not uncommon for a single point mutation to coincidentally create or destroy an RE recognition sequence. If this is the case, digestion of the appropriate PCR product prior to agarose gel electrophoresis enables the mutation to be identified. A difference in the size of the restriction fragments seen in normal and mutant samples can be predicted from a restriction map of the amplified fragment and the site of the mutation that changes a restriction site. The observed fragments should be consistent with either the mutant or the normal pattern. An example is shown in Figure 8.2 on p. 147 in the diagnosis of the sickle cell mutation.

Figure 8.2 Detection of the sickle cell mutation. (A) A sketch of the β globin gene shows the position of the primers used to amplify a 768 bp fragment in a polymerase chain reaction (PCR). The sequence of β I-5′ is 5′TAAGCCAGTGCCAGAAAGAGCC3′ and that of β II-3′ is 5′CATTCGTCTGTTTCCCATTCTA3′. Maps of the Bsu36 I restriction sites and the fragment sizes from β^A and β^S genes are shown below. (B) An ethidium-bromide-stained mini-gel illustrates the fragment sizes generated by Bsu36 I digestion of the PCR product from normal (A/A), sickle cell trait (A/S) and sickle cell anaemia (S/S) individuals, along with the undigested amplified fragment (U) and the molecular size marker (M).

Principle

Under appropriate conditions, short oligonucleotide probes will hybridize to their exact complementary sequence but not to a sequence in which there is even a single base mismatch.¹⁰ A pair of oligos is therefore used to test for the presence of a point mutation: a mutant oligo complementary to the mutant sequence and a normal oligo complementary to the normal sequence, with the sequence difference placed near the centre of each oligo.

The stability of the duplex formed between the oligo and the target DNA being tested (the product of a PCR reaction) depends on the temperature, the base composition and length of the oligo and the ionic strength of the washing solution. For allele-specific oligonucleotide hybridization (ASOH) studies, an empirical formula has been derived for the dissociation temperature (Td), the temperature at which half of the duplexes are dissociated: for hybridization of oligonucleotides of 14–20 bases in length. The Td can be estimated as 2°C for each dA:dT pair plus 4°C for each dG:dC base pair. This value is used as a guideline; the exact temperature at which only perfect base pairing is maintained is usually determined by trial and error.

This methodology has been widely applied for the detection of point mutations using fluorescently labelled TaqMan probes that distinguish the two alleles. Two short allele-specific probes are used, one of which will hybridize only to the wild-type allele and one of which will hybridize only to the mutant allele. Each probe is labelled with a different fluorescent colour, which is quenched while the probe remains intact, but is released if and when the probe hybridizes to its perfectly complementary sequence during the PCR reaction, as it will then be broken up by the exonuclease activity of the Taq polymerase. An example of this analysis is the detection of the factor V Leiden mutation in Figure 8.5 on p. 150.

Figure 8.5 Allelic discrimination of factor V Leiden using the TaqMan assay system. Each spot on the graph represents one well of the plate in which one DNA template is under test. In each well we either see no signal (bottom left) – NTC; only ‘red’ signals, coming from the wild-type probe (bottom right) – homozygous wild-type; only blue signals coming from the FVL probe (top left) – homozygous FVL; or a mixture of signals to give green spots coming from both the wild-type and FVL probes (top right) – heterozygous for FVL.

Interpretation

The oligos will hybridize to their perfectly complementary DNA sequence, such that the mutant oligo gives a signal only when the mutant allele is present, with the same happening for the wild-type allele. When this is the case, the interpretation of the result is straightforward; a positive signal from a particular oligo indicates the presence of that allele in the test sample. Heterozygotes and homozygotes are distinguished by using the mutant and normal oligos in tandem. With the two fluorescent colours of the Taqman probes, the heterozygote is identified as the sum of the two fluorochromes.

Other non-radioactive probes, with detection systems involving horseradish peroxidase, have also been quite widely used in this procedure.¹¹ The technique has also been modified such that the allele-specific oligonucleotides are immobilized onto nylon membranes and the patient-specific PCR product is used as the probe–the reverse dot blot procedure.¹² This allows for several different mutations to be analysed simultaneously and has proved particularly useful in the diagnosis of β thalassaemia mutations.¹³

Principle

The Sanger chain termination method for direct DNA sequencing has become a standard diagnostic tool. In many laboratories, this procedure has superseded targeted mutation detection as it provides a robust and relatively rapid method to identify all sequence changes that may be present in a particular DNA fragment. This approach is particularly relevant where multiple different mutations may underlie a particular disorder. This is the case for β thalassaemia, glucose-6-phosphate dehydrogenase (G6PD) deficiency and the red cell membrane disorders among others and so it is not surprising that DNA sequencing has often become the method of choice for the molecular diagnosis of these diseases.

In outline, the method revolves around the de novo synthesis of DNA strands in one direction from a PCR-derived template DNA fragment. The chain is lengthened by a thermostable DNA polymerase using deoxynucleotide triphosphates in the normal way; however, included in the reaction mixture is a small proportion of labelled dideoxynucleotide triphosphates (ddNTPs), which when incorporated will prevent any further extension of the chain. This process happens millions of times along a relatively short piece of DNA (usually up to 1000 bases), which means that chaintermination will occur many times at each position along the fragment. In the Applied Biosystems BigDye system (Warrington), each of the ddNTPs is labelled with a different fluorochrome and so the products of the sequencing reaction will consist of single-stranded DNA fragments, each differing in size by one base pair and each labelled with a different colour. These fragments can then be separated by capillary electrophoresis and the order with which the different colours exit the capillary will correspond to the sequence of the DNA template.

Reagents

Method

Interpretation

Reading the DNA sequence from a good trace – known as an electrophoretogram – is completely straightforward: As are called as green peaks; Ts as red; Cs as blue; and Gs as black. Free software packages, such as Chromas (at: http://chromas-lite-version.fyxm.net/), are available for viewing these traces and will call the DNA sequence in the file. Simple alignment of this sequence to the GenBank reference sequence can be performed at the National Center for Bioinformatics (NCBI) using the Blast program (at: http://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=BlastHome) and will identify any sequence changes. Heterozygous point mutations will be seen as double peaks, with two colours overlaid. Small heterozygous insertions or deletions (indels) are harder to decipher, as the sequence 3′ of the mutation will be a double sequence, with the normal and indel allele superimposed on one another: the extent of the indel can be defined by subtracting from the expected normal sequence.

Method

A variety of different methods have been used to detect these mutations and we have recently adopted a TaqMan-based assay. The reagents are provided by Applied Biosystems and the protocol provided has been adapted as described here. Differentially labelled, but quenched, TaqMan probes hybridize specifically to the normal and mutant alleles in a given patient’s sample. Detection of the light that is emitted as the probes are released during the PCR amplification is used to discriminate between the presence of normal and mutant alleles in the sample.

Reagents

All are obtained from Applied Biosystems (Warrington): 96-well optical PCR plate, optical adhesive lid, pipettes and tips, TaqMan Genotyping master mix, FVL genotyping assay (C_11975250_10), PTM genotyping assay (C_8726802_20), ddH2O, Applied Biosystems 7900HT Fast or 7500 Real-Time PCR apparatus.

Method

Defrost the genotyping assay mixes on ice.

In each experiment, include a normal, heterozygous and homozygous control for both the factor V Leiden and the prothrombin mutations; also include three blanks to which no template DNA is added. In the remaining wells, aliquot 2 μl of test DNA into the bottom of the appropriate well.

Prepare a PCR mix for each test using 0.5 μl of the relevant genotyping assay, 10 μl of water and 12.5 μl of the TaqMan Genotyping master mix per sample to be analysed.

Aliquot 23 μl of this PCR mix to each well, seal the plate with the optical adhesive lid and spin briefly.

Use the ‘Allele Discrepancy Assay’ on the computer associated with the Applied Biosystems 7900HT Fast or 7500 Real-Time PCR System. Thermocycling conditions are a denaturation at 95°C, for 10 min for 1 cycle followed by denaturation at 92°C for 15 s and annealing/extension at 60°C for 60 s, repeated for 40 cycles.

Interpretation

The results can be evaluated using the allelic discrimination plots for the individual genotyping assays. Each data point represents one sample and they can be seen to fall into four clusters according to the genotype of the sample or the absence of the signal in the non-template control (NTC): see Figure 8.5. If the assay has worked, these clusters should be clearly distinct. An automatic genotype annotation is then based on this clustering. Some points will not be automatically called by the software: these data points can be annotated manually by assigning them to the appropriate cluster.

Principle and Interpretation

The BCR–ABL1 analysis is performed by two-stage RT-PCR. The RNA extracted from nucleated cells is reverse transcribed by RT to generate coding or cDNA using random primers 6 bp long (hexamers). Following the RT step, the samples are subjected to multiplex PCR, to test for the presence or absence of BCR–ABL1.²⁹ Multiplex PCR is similar to conventional PCR but includes more than one pair of primers in a single PCR test. This strategy enables the detection of the vast majority of the BCR–ABL1 transcripts. The most commonly observed transcripts are b3a2 (e14a2), b2a2 (e13a2) and e1a2, giving rise to 385, 310 and 481 bp amplicons, respectively (Fig. 8.8). However, the identification of fragments of different sizes may be indicative of the presence of a ‘variant fusion gene’ such as e6a2 (1125 bp), e6a3 (951 bp), e8a2 (1319 bp), e8a3 (1145 bp) or e19a2 (908 bp).

In addition to BCR–ABL1, the normal BCR gene is co-amplified, yielding an 808 bp amplicon. The co-amplification of BCR is an indication of the quality of RNA and the efficiency of cDNA synthesis. Absence of any fragments indicates failure of the procedure. The latter is often the result of an aged sample (i.e. more than 72 h old). For RT-PCR analysis, the sample should be processed to lysate stage (see Nuclear lysate preparation, below) within 48 h of collection. In addition, the BCR fragment is often not observed in diagnostic samples where the BCR–ABL1 is preferentially amplified.

If BCR–ABL1 is undetectable by multiplex PCR in follow-up samples from patients undergoing therapy, the cDNA is tested at higher level of sensitivity by nested PCR.³⁰ Nested PCR enables the detection of one leukaemic cell in a background of 10⁵ to 10⁶ normal cells. The choice of primers for nested PCR is dependent on the type of transcript detected by multiplex PCR at presentation. The primers for e14a2 and e13a2 are the same; however, for e1a2 transcript a different set of primers is used. The nested PCR yields fragments of 385 bp (e14a2), 310 bp (e13a2) or 481 bp (e1a2) in length. Nested PCR is not indicated for testing cDNA from a patient suspected of having CML. Diagnosis is made by expression of BCR–ABL1 by multiplex PCR. Furthermore, in post-therapy samples this will indicate molecular relapse.

Reagents

Solutions for the Complementary DNA Mix

The compositions for the multiplex and nested PCR mixes are given in Table 8.3, including the optimum MgCl₂ concentration for each mix as well as the primers used in the different PCR reactions.

Methods

Principle

qRT-PCR permits quantification of number of transcripts of gene of interest at high levels of sensitivity. This is achieved by developing a technology that permits the detection of PCR products as they accumulate. Furthermore, the rate of accumulation is proportional to the number of mRNA molecules of the target gene in the starting material during the exponential phase of the PCR. The accumulation of amplicons is detected by including a sequence-specific probe labelled with fluorochromes in addition to the primers as in a conventional PCR (Fig. 8.9). Since the advent of qRT-PCR several types of probes have been developed, although all are dependent on the fluorescence resonance energy transfer (FRET) principle. The two commonly used systems involve hybridization or hydrolysis of the probe. A widely used methodology is TaqMan, which involves hydrolysis of the probe.³³ This technology is based on the 5′ exonuclease assay, which exploits the inherent 5′ to 3′ exonuclease activity of the Taq DNA polymerase.³⁴ The Taq DNA polymerase cleaves a dual-labelled probe annealed to the target sequence during PCR amplification (Fig. 8.10A). Briefly, the cDNA synthesized from total RNA is added to the PCR reaction containing standard PCR components plus a probe that anneals to the template between the two primers as per conventional PCR. The probe has a fluorescent reporter dye, FAM, at the 5′-end (6′carboxyfluorescein; emission λ_max = 518 nm) and quencher dye, TAMRA, at the 3′-end (6-carboxy-tetramethyl-rhodamine; emission λ_max = 582 nm). While the probe is intact, the proximity of the quencher greatly reduces the fluorescence emitted by the reporter dye by Förster resonance energy transfer (FRET).³⁵ Adequate quenching is observed for probes with the reporter dye at the 5′ end and the quencher at the 3′ end.

Figure 8.9 Probes and primers. The position of probes and primers used for quantification of BCR–ABL1 and ABL1 transcripts are shown. The BCR–ABL1 fusion gene is shown in yellow- and purple-filled boxes, with the yellow representing the exons derived from the BCR gene and purple those originating from the ABL1 gene. The dual-labelled probes are shown as red lines and primers are shown as arrows. Probes and primers were designed using the Primer Express software to detect e13a2 and e14a2 junctions in a single reaction by qRT-PCR. The BCR–ABL1 (FAM-cccttcagcggccagtagcatctga-TAMRA) and ABL1 (FAM-tgcttctgatggcaagctctacgtctcct-TAMRA) probes are dual labelled with 6-carboxyfluorescein (FAM) and 6-carboxy-tetramethyl-rhodamine (TAMRA). BCR–ABL1 forward primer: 5′-tccgctgaccatcaayaagga-3′; BCR–ABL1 reverse primer: 5′-cactcagaccctgaggctcaa-3′; ABL1 forward primer: 5′-gatacgaagggagggtgtacca-3′; ABL1 reverse primer: 5′-ctcggccagggtgttgaa-3′. The BCR–ABL probe and primer set was designed in association with Europe Against Cancer Collaborative study.

Figure 8.10 Real-time quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR). (A) Schematic representation of the principle of real-time PCR. The forward and reverse primers (blue lines) are extended by the Taq DNA polymerase as per conventional PCR. The quencher (red symbol) suppresses the fluorescence from the reporter (green symbol) dye while the two are attached to the probe. As the forward primer is extended, the probe is displaced and is cleaved by the Taq DNA polymerase 5′–3′ exonuclease activity. Fluorescence is detected once the reporter dissociates from the quencher on probe being degraded and the synthesis of the new DNA continues to completion. Therefore the fluorescence increases with increment in number of amplicons. (B) Amplification plot. The real-time PCR amplification plot of the standards in log increments is shown. A 0.5 log standard between 10 and 100 BCR–ABL1 copies is included. The number of cycles for the fluorescence to cross the threshold is reported as Ct (cycle threshold). The greater the number of copies, the lower the Ct value (i.e. it is inversely proportional). (C) A standard curve. A standard curve drawn from qPCR data from serial dilutions of plasmid containing BCR–ABL1 (black symbols) is shown. The BCR–ABL1 transcripts for the unknown samples (red symbols) are read off the standard curve.

Thus, while TAMRA and FAM are closely attached to the probe, fluorescence from the reporter dye is quenched by TAMRA. During PCR, as the Taq DNA polymerase replicates the DNA strand to which the TaqMan probe is annealed, the probe is degraded by the intrinsic 5′–3′ exonuclease activity of the polymerase. The effect is to dissociate FAM from TAMRA; therefore FRET is no longer applicable and fluorescence from FAM can be detected by a laser integrated in the sequence detector (TaqMan ABI 7500 Real-Time PCR system, Applied Biosystems, Warrington). Fluorescence increases in each cycle, proportional to the rate of probe degradation. The number of cycles taken for the fluorescence to cross a threshold value of 10× the standard deviation of baseline emission is used for quantitative measurement. The threshold is set significantly above the baseline and can be adjusted manually. The upper baseline limit can be altered by the operator. The fluorescence is plotted as ΔRN against number of cycles (Fig. 8.10B), where ΔRN = (Rn⁺)–(Rn⁻). The Passive reference, ROX (6-carboxy-X-rhodamine), is included in the qRT-PCR Universal master mix. The formulae used to calculate the Rn⁺ and Rn⁻ follow:

From PCR with template :

and from PCR without template or early cycles of PCR:

The number of cycles taken to pass the defined threshold is called the cycle threshold (Ct) and it is inversely proportional to the quantity of target starting material. The number of transcripts of the target is read off a standard curve. The quantity of target gene and endogenous control gene transcripts are measured off their respective standard curves, generated from dilutions of the appropriate plasmids and run on each plate. The level of expression of the target gene is then reported as a percentage ratio of the target gene to the control gene to obtain a normalized value for the gene of interest independent of the integrity of the RNA and efficiency of the reverse transcription reaction.

Interpretation

Patients who achieve complete cytogenetic remission may still harbour up to 1 × 10¹⁰ leukaemic cells. The kinetics of the leukaemic load in these patients and their response to therapeutic agents can only be monitored by measuring MRD.³⁶ However, it should be noted that patients in whom qRT-PCR fails to detect BCR–ABL1 may still harbour up to 1 × 10⁶ leukaemic cells. Of those who are in complete cytogenetic remission (BCR–ABL1/ABL1) × 100 is invariably <2.0%. Investigators previously reported that patients who achieve (BCR–ABL1/ABL1) × 100 <0.045% while receiving α-interferon therapy have a considerably reduced risk of cytogenetic relapse.³⁷ More recently, qRT-PCR has been used to monitor patients being treated with imatinib. An international multicentre study showed that a 3 log reduction in BCR–ABL1 copies was consistent with a good prognosis.³⁸ Quantification of BCR–ABL1 also helps to identify patients at risk of relapse and therefore provides a window for early clinical intervention with the aim of reversing disease progression. For patients who have undergone stem cell transplantation (SCT) this generally means infusion of lymphocytes isolated from the original donor (i.e. adoptive immunotherapy). Although there is some debate as to the precise criteria for molecular relapse, there is little doubt that a confirmed 1 log increase in BCR–ABL1 transcripts is clinically significant (i.e. from 0.002% to 0.02%). Patients who achieve a 0.02% BCR–ABL1/ABL1 ratio on three consecutive occasions are said to be in molecular relapse.

Advantages of qRT-PCR

The major advantage of qRT-PCR is the ability to detect accumulation of amplicons during the exponential phase of PCR. This permits quantification of the DNA of interest in the starting material. This is not possible with conventional PCR because samples are analysed at the end of the PCR run and therefore any differences in the copy number between samples in the starting material are generally not discernible. This is illustrated by the amplification plot shown in Figure 8.10B, where all the samples at the end of 50 cycles have the same level of fluorescence, despite having varying target copy numbers in the starting material as seen in the exponential phase of the PCR. Post-PCR handling is eliminated. On completion of an assay, the sealed microtitre plate is discarded, thereby minimizing risk of contamination. This also eliminates the need to handle stained gels on completion of the qRT-PCR assay. The qRT-PCR offers a higher level of specificity because, in addition to primers annealing to DNA sequence of interest, a third oligonucleotide (the probe) anneals to the region between the primers at a higher temperature. To achieve a level of sensitivity that is similar to qRT-PCR, nested conventional PCR is required, but this is associated with a greater risk of contamination. Conventional PCR is also less amenable to automation. More importantly, because qRT-PCR can be automated, inter-laboratory standardization becomes feasible when measuring patients’ responses to therapy. This permits rapid evaluation of new therapeutic modalities because methodology and protocols can be standardized, permitting international inter-laboratory studies.

With the aim of administrating an optimum dose it is critical to ensure negligible inter-laboratory variation as patients move from one place to another. This would also assist in simplifying multicentre therapeutic agent clinical trials as best illustrated by International Normalized Ratio (INR) for prothrombin time, used to monitor patients whose coagulation status is controlled with the anticoagulant warfarin. There have been efforts to establish a single internationally recognized scale for monitoring CML patients and this may be soon achieved by the establishment of a calibrated and accredited primary reference reagent for BCR–ABL1 q-PCR assays.³⁹

Reagents

×2 Universal master mix. The ×2 Universal master mix contains dATP, dCTP, dGTP and dTTP at 200 μmol/l each, 5.5 mmol/1 MgCl₂ and 0.025 μmol/l AmpliTaq-Gold. It also contains the passive background reference dye, ROX. The Universal master mix can be purchased with or without uracil DNA glycosylase, which degrades any PCR contaminating products prior to starting the PCR by heating the plate to 95°C for 10 min (see later). It is possible to assemble master mixes by purchasing various components, such as dNTP, buffer, MgCl₂ and hot start Taq DNA polymerase. However, in-house preparation of master mixes is not recommended to minimize intra-assay and inter-assay variation and contamination, essential for monitoring patient response to therapy.

Probe–Primer mix. For convenience and to minimize inter-assay variation a bulk preparation of the qRT-PCR assay mix, containing the probes and primers at required concentration, minus the master mix, is recommended. The mixture is stored at –20°C or –70°C. This also avoids repeated freezing and thawing of probes and primers because this may affect the probe–primer integrity. Furthermore, the probe should not be left exposed for prolonged periods to direct sunlight because this leads to degradation. In general 300 nmol/l of each primer and 200 nmol/l of probe permits optimum qRT-PCR sensitivity; however, this should be determined for each assay by titrating one primer against the other. The optimum concentration of the primers is one that gives the lowest Ct. Similarly; the optimum probe concentration is determined by varying the quantity of the probe. The quantity yielding the lowest Ct is the optimum probe concentration. The probe–primer mixture is then prepared using the determined optimum concentrations. The mixture can then be aliquoted into microcentrifuge tubes for a required number of samples, allowing for standards and positive and negative controls. Furthermore, for MRD studies it is advisable to measure the target gene in replicates of three to minimize sampling error at low copy number values. Because the endogenous control gene copy number is expected, assays in duplicate will suffice. However, standards for the endogenous control should be performed in triplicate.

Designing probe and primers. The probe and primers are designed using Primer Express Software (Applied Biosystems, Foster City, CA). Optimum design of probes and primers is critical to the sensitivity of qRT-PCR. The probe is designed such that it has higher T_m than the primers and works optimally at default PCR conditions settings using the universal master mix. The probe should not have a guanine base at the 3′ end and the number of guanine bases should be fewer than the number of cytosine bases. Furthermore, there should not be more than four guanine bases in tandem. Similar rules apply to design of primers. The annealing temperature of the probe should be 10°C greater than that of the primers. It is essential to design probes and primers such that the assay is RNA specific. This is achieved by positioning the forward and reverse primers in separate exons or by placing either of the primers or the probe across a splice site.

Standard curve. It is acceptable to report data as Ct, with an increase of 3.3 being clinically significant, because this represents 1 log increase in BCR–ABL1 copies; it takes 3.3 cycles for every log increase in amplicons. For clinical samples, however, it is essential to generate a standard curve from which the unknown test samples can be calculated. The standard curve can be generated using serially diluted cDNA derived from a cell line expressing the target gene at high levels (e.g. K562 for BCR–ABL1). As an alternative to using cDNA, a plasmid can be used to create this curve, which provides stability over time. The method of preparation of the plasmid is beyond the scope of this book. Serially diluted plasmids for commonly occurring fusion genes and endogenous control genes are commercially available. In the absence of a standard curve the MRD values are reported as a delta-delta Ct (ΔΔCt). This is calculated by first normalizing the fusion gene Ct (CtFG) to the control gene (CtCG) to obtain a ΔCt for the follow-up samples – that is, the ΔCt(follow-up) = Ct_FG – Ct_CG. The same calculation is performed for the Ct value for the sample taken at diagnosis to obtain a ΔCt(diagnosis). The ΔCt(diagnosis) is then subtracted from the ΔCt(follow-up) to obtain a ΔΔCt. From this the MRD value is calculated as 10^ΔΔCt/3.3. To apply the ΔΔCt method of reporting the slope and intercept values for the control and fusion genes must be similar. More precisely it is recommended that the slope values for the fusion and control gene should not differ by more than 0.01 (i.e. the PCR efficiency for CG and FG are similar). The major advantage of using ΔΔCt is it obviates the need for a standard curve. Therefore, eliminating the need for a plasmid- or RNA-based standard curve reduces the risk of contamination further and frees microtitre plate wells for patient samples.

PCR cycling conditions. It is convenient to design the probes and primers so that they are able to work efficiently using standard qRT-PCR conditions to amplify the cDNA, which are 2 min at 50°C (to allow uracil DNA glycosylase-mediated elimination of exogenous PCR product contamination), an enzyme heat-activation step of 10 min at 95°C, followed by 50 cycles of 15 s at 95°C for denaturation and 1 min at 60°C for annealing and extension.

Method

The qRT-PCR assay is normally performed in 96-well microtitre plates in a 25 μl final reaction volume containing universal master mix. The composition of a 25 μl qRT-PCR reaction is shown in Table 8.4.

The instrument sets the threshold at ×10 the standard deviation of baseline emission; however, this can be reset manually within the exponential phase of the PCR to avoid any background fluorescent interference. Alternatively, the threshold can be set at the same value for each assay, for instance, at 0.05, assuming this is within the exponential phase of PCR, thus avoiding operator variation. The data for any samples with a Ct >38 are considered unreliable.

The baseline limits are set by the instrument; however, this should be adjusted so that the upper limit of the baseline is 4 cycles less than the lowest Ct value for a sample. For example, if the lowest Ct is 20, then the upper baseline limit is set at 16, thus giving a clear margin between the baseline and the samples. The lower limit set by instrument rarely requires adjusting.

The standard curve is generated and accepted if the slope value is between –3.3 and –3.6. A slope value of –3.3 represents 100% PCR efficiency because it takes 3.3 cycles for every log increase in PCR products. Ideally the curve correlation coefficient should not be <0.98. If the standard curve is acceptable, then the copy number for the samples can be recorded. The qRT-PCR for the endogenous control gene is performed similarly using the target-specific probes and primers. On completion of qRT-PCR for target and endogenous control, the data are reported as a percentage (i.e. (BCR–ABL1/ABL1) × 100).

Notes

To minimize sampling error, the target gene and the standards are assayed in triplicate. It is essential to include positive controls with a known number of transcripts. To minimize risk of contamination the patients’ samples and standards should be handled in geographically separate locations. Because qRT-PCR assays are BCR–ABL1 transcript-type specific it is essential to perform a multiplex PCR on a presentation sample to assign the transcript type expressed by the patient. This ensures that correct qRT-PCR assay is used. A qRT-PCR assay designed for b2a2 or b3a2, now referred to as e13a2 and e14a2, respectively, would give rise to false-negative data in patients expressing e13a3 and/or e14a3. A single qRT-PCR for e13a2 and e14a2 is recommended rather than two separate assays for e13a2 and e14a2. This is achieved by designing the assay such that the probe and reverse primer map to the second exon (i.e. a2) of the ABL1 gene and the forward primer maps to the e13 exon (i.e. b2) of the BCR gene (Fig. 8.9). Apart from being cost-effective and efficient, it increases the accuracy of the assay because a significant minority of patients with a single nucleotide polymorphism in exon 13 of the BCR gene have the potential to express both e13a2 and e14a2.

Principle

This analysis is possible because the IGH@ and TCR genes undergo a rearrangement during the normal differentiation of B and T lymphocytes, respectively, but not during differentiation of other cells. This rearrangement results in a unique fusion of variable, diversity and joining (VDJ) segments, interdigitated by random nucleotide (N region) insertion or deletion. The sequence and length of the DNA at these sites of recombination are therefore characteristic of any particular lymphocyte clone.

For many years, Southern blot analysis was the gold standard for the detection of rearranged IGH@ and TCR genes. For details on how these were performed and interpreted, we refer the reader to previous editions of this book. More recently, because of its simplicity, the small amount of DNA required and potential sensitivity, PCR has been used to detect rearrangement of the IGH@ and TCR genes. Because of the N region diversity, a polyclonal population of cells will give rise to a ladder of various fragment sizes. However, if one clone becomes abnormally large, a discrete fragment size will begin to dominate the products of the PCR – the basis of the so-called ‘fingerprinting’ method for the diagnosis of lymphoproliferative disorders.⁴¹ This analysis can be refined using heteroduplex analysis or SSCP gels in which the sequence as well as the size of the amplified product determines its mobility. To gain further sensitivity in following disease, the product of a ‘clonal’ amplification can be sequenced to derive a clone-specific sequence at the site of rearrangement. This sequence can then be used for the design of clone-specific oligonucleotide probes or primers that can be used in ASOH, ARMS or qRT-PCR. This methodology has been used to monitor MRD in lymphoproliferative disorders.^42,43

A comprehensive report has been published on the design and standardization of PCR primers and protocols for the detection of IGH@ and TCR gene rearrangements.⁴⁰ The detection rate of clonal rearrangements is very high, but the comprehensive nature of the test requires 107 primer pairs in 18 multiplex PCR tubes, which are now commercially available. The methods described here are more restricted but more widely applicable.

Method: Immunoglobulin Heavy Chain Gene Rearrangement

To study immunoglobulin gene rearrangement, the locus of choice is the heavy chain gene. A single primer can be used, which will anneal to a consensus sequence shared by all joining (JH) segments. The choice of variable (VH) segment primers is more difficult and for the more comprehensive analysis, primers from all three framework regions for each of the six or seven VH families are used.³⁹ A reasonable starting point, however, is to use the JH primer in conjunction with a different primer derived from a consensus sequence of the framework 1 region for each VH family. These primers are as follows:

PCR buffer I (see above) is used in these reactions, with an annealing temperature of 60°C. The products of the reactions are in the order of 310–350 bp and can be visualized as either a smear or discrete band in high-percentage agarose gels (1.5–2%). However, a better resolution is obtained when the products are resolved by capillary electrophoresis on an automated fragment analyser (e.g. the ABI 3130xl DNA Analyser). Fragments are visualized by attaching a fluorescent label (HEX or FAM or VIC) to the VH primers and reading the peaks of fluorescence as an electrophoretogram or ‘genescan’ Alternatively, the PCR products can be subjected to heteroduplex analysis, denaturing them at 95°C for 5 min and annealing at 50°C for 1 h, prior to electrophoresis in a non-denaturing 6% polyacrylamide gel.

Interpretation

Because of the variable number of nucleotides either removed or added at the point of joining of VDJ segments of the immunoglobulin heavy-chain gene, the distance between V segment and J segment primers will alter accordingly. For the gene to be functional, the reading frame must be maintained and therefore variations in length must be in multiples of 3 bp, although this may not always be the case in acute leukaemias where non-productive rearrangements may occur. The polyclonal population of B cells therefore gives rise to a characteristic ‘ladder’ or Gaussian size distribution, with maximal intensity observed at the median length at its centre. However, if one B-cell clone is abnormally large, it will give rise to a disproportionately intense peak at the size (and using a V primer from the appropriate family) corresponding to the length of VDJ fragment derived from that clone. At presentation of a B-cell malignancy, this band may be the only one visible, confirming the presence of a dominant abnormal B-cell clone. Subsequently, an abnormal intensity of this fragment size in the background of a ladder can be used to monitor the disease.

In heteroduplex analysis, the polyclonal population of B cells will give rise to fragments with many different sequences, which, on denaturing and reannealing, will generate heteroduplexes that will appear as a smear spreading across the gel. If one large B-cell clone is present, however, homoduplexes will form and these will migrate as a discrete fragment with a migration consistent with its size alone.

There are two main problems that can be encountered in this analysis. The first is that the consensus V primers may not amplify all V segments because of mutations in sequence of the region recognized by the primers (either J, V or both). To overcome this problem, primers for the leader region of the V gene and for the constant region of the heavy chain identified by immunofluorescence on the surface of the B-cell clones may be used on cDNA. The distance between the leader primers and the constant regions may be too large for DNA amplification. It is therefore important that in patients with chronic lymphocytic leukaemia (CLL), B-cell lymphomas and myeloma, RNA is also stored from diagnostic samples in case this approach is required. However, it is also important that appropriate diagnostic material is referred for analysis. For instance, lymph nodes in case of lymphomas, bone marrow (BM) in case of myeloma and peripheral blood (PB) in CLL. A BM or PB sample in a case of lymphoma may be totally ‘empty’ of disease and therefore not suitable for this type of investigation.

Another problem encountered, especially in ALL, is the emergence of a subclone or clones during the course of the disease, which may result in a change in the fragment size and family.

Method: T-Cell Receptor Gene Rearrangement

The first choice of locus for PCR analysis of the TCR gene rearrangement is the TCR gamma locus as it is rearranged in the vast majority of T-cell clones and does not have the complexity of the TCR beta locus, which has 24 different V segment families. Amplification of DNA around the joining region of the TCR gamma gene is performed in a similar way to that described earlier for the IGH@ locus, using a consensus primer for the joining segment (JγC) and one for each of the four variable segment families (Vγ I–IV). The primer sequences are as follows:

Standard PCR amplification conditions can be employed using these primer combinations and analysis is again performed on acrylamide gels or an automated fluorescence analyser.

Interpretation

As in the analysis of the IGH@ gene, discrete bands of amplified product are obtained from clonal T-cell populations, whereas ladders are obtained from polyclonal populations by gene scanning or heteroduplex analysis. The main problem with the interpretation of this kind of data is that it does not distinguish reactive T-cell clones from malignant T-cell clones. Three different patterns are clearly distinguished: a large malignant clonal population, an oligoclonal population and a polyclonal population. However, because of the sensitivity of this method, problems can arise in trying to detect a small malignant clone against a polyclonal population compared to a clonal population stimulated by antigen, particularly when a patient is lymphopenic.

Immunoglobulin Heavy Chain and TCR Gene Rearrangements as Targets for Minimal Residual Disease Analysis

The value of detecting residual disease with greater sensitivity than light microscopy, using molecular or immunological techniques, has been extensively evaluated in childhood and adult ALL. Minimal residual disease is defined as the ‘lowest level of disease detectable in patients in CR (i.e. complete remission) by the methods available’ and the newly developed real-time quantitative PCR techniques have revolutionized MRD investigation and patient management. The methodology makes use of the information derived from the cloning and sequencing of the rearranged clonal leukemic clone and the generation of an allele-specific primer able to identify the unique rearrangement belonging to the leukaemic cell with a sensitivity of 1 in 100 000 normal cells. Guidelines for the interpretation of real-time quantitative PCR data in the analysis of minimal residual disease by IGH@/TCR gene rearrangement have been generated by a European study group. These are currently being followed and applied in a large study at international level and we refer the reader to such extensive publications⁴³ for further information.

In brief, these studies in children have concluded that conversion to MRD negativity shortly after induction therapy and maintenance of MRD negativity are prerequisites for long-term disease-free survival and that MRD positivity often precedes clinical relapse.⁴⁴ In adults also, MRD analysis carried out during induction therapy provides a strong predictor of outcome and should also be applied to harvested BM to exclude patients with detectable leukaemia burden of more than 1:10³ normal cells. Finally, predictions of outcome based on MRD analysis in children and adults are more accurate than predictions based on other prognostic indicators, such as age, gender, immunophenotype, presenting white cell count, karyotype and time taken to achieve first CR.^45–47

JAK2 Mutation Analysis

In 2005 several groups independently identified a unique mutation in the JAK homology 2 (JH2) pseudokinase domain of the Janus kinase 2 (JAK2) gene in patients with a variety of myeloproliferative neoplasms (MPNs), including polycythemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF).^48,49 The mutation was reported in up to 90%, 40% and 25% of patients in the three diseases, respectively, but not in chronic myelogenous leukaemia or in 700 normal controls. It consisted of a G to T substitution at nucleotide 1849 of exon 14, leading to a valine to phenylalanine substitution at codon 617 (V617F). This mutation caused loss of JAK2 auto-inhibition and constitutive activation of the cytoplasmic JAK2 kinase and has become recognized for definitive molecular diagnosis of PV. Until this mutation was identified the distinction between primary and secondary polycythemia was sometimes difficult to establish.

A variety of molecular methods have been employed to establish the presence of this mutation, including direct DNA sequencing, allele-specific PCR, amplification refractory mutation system, real-time quantitative PCR and melting curve analysis as a semi-quantitative method, restriction fragment length polymorphism (RFLP) analysis and pyrosequencing. There are other methods which could be applicable to this type of mutation analysis, including SSCP or denaturing high-performance liquid chromatography (DH-PLC), both of which are technically challenging and labour intensive.

More recently, some relatively rare mutations in the exon 12 of the JAK2 gene have been identified in V617F-negative patients with polycythaemia vera or unexplained erythrocytosis (thus redefined as polycythaemia vera) using PCR and direct DNA sequencing as well as allele-specific PCR. We are currently screening for these mutations by denaturing HPLC analysis, but a full description of this test lies beyond the scope of this chapter and readers are referred to reference 52 for information. For these investigations, granulocytes represent the preferred source for DNA or RNA preparation, but total white blood cells are also commonly used. DNA and RNA are prepared according to the protocols described in other sections of this chapter.

Principle: JAK2 Exon 14 Mutation Analysis

Mutation of exon 14 of the JAK2 gene in MPNs is characterized by a single nucleotide mutation which is identical in all patients carrying this defect. PCR amplification can be designed to exploit the specific annealing of a primer at the 3′ end in single or multiplex conditions. By designing a primer that matches the codon 617 wild-type (G) sequence, the primer will bind and amplify only the wild-type DNA. Conversely, the primer designed to match the mutant sequence (T) will amplify the mutant allele and not the wild type. After the PCR has been carried out, the products can be visualized in several ways. One can simply run the product on a 2% agarose gel and visualize the products by incorporating SYBR Safe dye in the gel and using a UV transilluminator. This methodology is marred by problems, such as possible lack of specificity and false-positive results.

A more successful approach in our hands is the use of primers designed to include a fluorescent label that differs between the wild-type and the mutant primer. The PCR product can then be run on a fragment analyser (the ABI 3130 for instance, Applied Biosystems, Warrington) with the wild-type and mutant allele appearing as two distinct fragments of the expected size in separate channels (Fig. 8.11).

Figure 8.11 JAK2 genotype analysis by genescanning. Primers specific for the wild-type JAK2 sequence (green fluorescent signal) and for the mutant allele (blue fluorescent signal) have been applied to two samples and run on a gene-scanning ABI apparatus. (A) The presence of a single signal in the wild-type amplification indicates that this sample is wild type for the G>T mutation. (B) The presence of a signal on both PCR amplifications indicates that this sample is heterozygous for the mutation.

This represents a rapid and cost-effective method for the analysis of large cohort of patients: 100–200 ng of total DNA or 1/20th of a cDNA reaction (normally containing 2 μg of total RNA in a 50 μl final volume reaction) is sufficient for this test, which is described in more detail below.

A different approach, pyrosequencing, is also extensively used in our laboratory for detecting JAK2 mutations. This approach involves the amplification of a fragment containing the nucleotide mutation by PCR using a biotinylated primer. The amplified product is captured on streptavidin-coated beads and is then used as a substrate for sequencing by synthesis at the mutated nucleotide/s. The four nucleotides are added stepwise and – when they are incorporated – a pyrophosphate (PPi) molecule is released. This is subsequently converted to ATP by ATP sulphurylase, which in turn enables a luciferin molecule to become oxidized in a luciferase reaction, producing light in amounts that are proportional to the amount of ATP. Unincorporated nucleotides and ATP are degraded by the apyrase and the reaction can restart with another nucleotide. Therefore the incorporation of the mutant and/or wild-type nucleotide will be captured and converted into a light emission signal, which can be measured and quantified. By comparing the relative intensity of the signal obtained after incorporation of the mutant and wild-type nucleotides, their relative abundance in the target sequence can be established. In this respect, pyrosequencing is a quantitative method and therefore offers a great advantage compared to other methodologies.

Interpretation

All tests are carried out including a positive control (DNA from HEL cell line, grown in vitro or any other DNA from cell lines or patients carrying a homozygous mutant allele), a negative control (normal individual DNA) and a non-template control, containing all reagents and water in place of DNA. Samples are tested in duplicate for the pyrosequencing and evaluated for the presence of a peak in the amplification well containing primers for the mutant allele. The amplitude of the luminescence signal provides quantification of wild-type and mutant allele amplicons as percentages of total. If using agarose gels, the presence of a band in the wild-type combination but not in the mutant will identify a wild-type patient; the presence of a band only in the mutant reaction will identify a homozygous patient, while the heterozygous cases will show amplification in both wild-type and mutant wells. The same principle applies when using a fragment analyser (Fig. 8.11). Fragment analysis can be semi-quantitative if the PCR cycle number is kept low, with the peak height reflecting the amount of target sequence present in the starting material. Several quantitative PCR tests have recently being developed and we refer to the wider literature for information on such tests.⁵⁰

Testing for Other Common Fusion Gene Products

In addition to the BCR–ABL1 fusion gene, a variety of fusion genes have been identified and characterized at the molecular level in acute and chronic leukaemias, all suitable for PCR amplification and use as markers at diagnosis and for application for minimal residual disease detection. The identification of these at the time of diagnosis may carry prognostic significance and subsequently influence treatment modalities. There are a great number of potential fusion genes, but the PML–RARA, ETV6–RUNX1 and MLL–MLLT2 are among the most frequent targets of investigation in acute myeloid leukaemia (AML) and acute lymphoblastic leukaemia (ALL). They are the result of the t(15;17), t(12;21) and t(4;11) translocations, respectively. The former two chromosomal aberrations are commonly associated with good response to therapy and patients rarely require transplantation. However, the t(4;11) is associated with a less favourable response to therapy. For this group even the option of transplant offers limited advantages for long-term survival and this is especially in adults. Each of these translocations will be briefly discussed in the following sections.

t(15;17)(q22;q21); PML-RARA Fusion Gene

The t(15;17)(q22;q21) translocation, associated with acute promyelocytic leukaemia (APL)(French–American–British – FAB – M3), fuses the PML gene, on 15q22, to the RARA gene on 17q21. The breakpoint cluster region on chromosome 17 localizes within RARA intron 2. Within PML there are three breakpoint cluster regions: breakpoint cluster region 3 (bcr3) in intron 3, bcr2 in exon 6 (rarely in intron 5) and bcr1 in intron 6. At the messenger RNA level, bcr1, bcr2 and bcr3 are also known as the long (L), variant (V) and short (S) isoforms, respectively. Alternative splicing within PML transcripts and the alternative use of two RARA polyadenylation sites are responsible for the production of additional PML–RARA transcripts of different sizes.

There are at least four variant translocations involving the RARA gene, with the same breakpoint, associated with the APL or similar phenotype. The fusion partners are NPM1 at 5q35, ZNF145 (previously PLZF) at 11q23, NUMA1 at 11q13 and STAT5B at 11q11.

As a consequence of the t(15;17)(q22;q21) translocation, both PML–RARA and RARA–PML fusions can be transcribed. While PML–RARA is consistently found in APL patients, RARA–PML is detectable in only 70% of cases. Thus, PML–RARA is most widely used for diagnosis and MRD monitoring in APL. This transcript can be detected by RT-PCR with a sensitivity of 1:10⁴.

The PML–RARA transcript can be detected with increased sensitivity by qRT-PCR compared to semi-quantitative nested/two-round PCR techniques. Accurate quantification can identify patients with an increasing number of transcripts preceding haematological relapse. Patients treated at the time of molecular relapse have 2-year event-free survival (EFS) rates superior to those treated at the time of haematological relapse.

11q23 Abnormalities

Rearrangements involving the MLL (mixed leukaemia lymphoma) gene (previously also called ALL1, HRX or Htrx1) on chromosome 11q23 and multiple partner genes are found in precursor B-ALL, T-ALL, AML, MDS and therapy-related leukaemia. The presence of MLL rearrangements is usually associated with a poor prognosis.

In ALL, the most common translocation partner of MLL is the MLLT2 gene on chromosome 4q21. Between 50% and 70% of infant ALL cases and approximately 5% of paediatric and adult ALL cases are MLL–MLLT2 positive and are associated with a pro-B-ALL (‘null’) phenotype (CD19+, CD34+, terminal deoxynucleotidyl transferase+, cytoplasmic CD79a+, CD10−). There is also frequent expression of myeloid antigens (CD15 and/or CD65).

The MLL and MLLT2 genes are composed of 37 and 20 exons, respectively, and at least 10 different fusion transcripts have been identified due to translocation breakpoints occurring in different introns of both genes. Breakpoints downstream of the MLL exon 9 in adult and paediatric ALL, but downstream of exon 11 in infant ALL and upstream of exon 4 of the MLLT2 gene, are commonly detected by PCR. Differential splicing is a common finding, leading to more than one fusion transcript in some patients. All t(4;11)-positive cases transcribe the MLL–MLLT2 fusion gene, while only 70% of cases transcribe the reciprocal, MLLT2–MLL, product. Interestingly, low levels of the MLL–MLLT2 transcript have been detected in some ALL cases without cytogenetically detectable t(4;11) and in haemopoietic tissues of healthy individuals.

Using a nested PCR strategy, the various MLLT2–MLL transcripts can be identified with a detection limit of 1:10⁴–10⁵. MRD studies have shown that early conversion and persisting MRD negativity is consistently associated with complete cytogenetic remission.

t(12;21)(p13;q22); ETV6-RUNX1 Fusion Gene

The t(12;21)(p13;q22) is a cryptic translocation not readily observed by conventional cytogenetics and the ETV6(TEL)–RUNX1 (AML1 or CBFA2) fusion transcript can be detected in patients with and without a cytogenetically visible chromosome 12 and/or 21 abnormality. FISH and PCR are therefore important techniques employed for its detection.

The t(12;21) is usually associated with a common ALL (cALL) and more rarely pre-B ALL. It is rarely seen in adult ALL, although it is the commonest translocation in childhood ALL, with a peak incidence in the 2–5-year age group. The breakpoint region usually lies between exons 5 and 6 of the ETV6 gene and between exon 1 and 3 of RUNX1. Fusion transcripts resulting from splicing of ETV6 exon 5 to other RUNX1 exons can also be formed. The majority of positive patients have the ETV6 gene deleted on the non-translocated allele and it has been suggested that the translocation occurs during fetal development. The ETV6–RUNX1 fusion transcript can be identified with a sensitivity of 1:10^-4–10^-5 using a nested RT-PCR approach. Accurate quantification of ETV6–RUNX1 has shown that MRD measurements are a good prognostic indicator.

Method and Interpretation

The above three translocations are determined and monitored by amplification of the mRNA, resulting from the fusion gene, as this is leukaemia-specific in the same way as the BCR–ABL1 testing for the t(9;22) (see p. 155 for the preparation of cDNA and conditions of PCR amplification). However, for primers specific for amplification at diagnosis and for qRT-PCR for follow-up samples we follow the protocols and guidelines for interpretation of results provided by the Europe Against Cancer Consortium.^51,52

It is worth noting that due to the multiplicity of the MLL–MLLT2 fusion products and the limitations of hydrolysis probe amplicon size, a genomic DNA-based assay may be employed for following affected patients as previously described.⁵³ This involves the characterization of the individual breakpoint by direct sequencing and designing an ASO in a manner similar to that for Ig/TCR-based MRD.⁵⁴ As controls for amplification we use the Ipsogen plasmids (www.ipsogen.com ). For interpretation of data, we refer to the Europe Against Cancer guidelines.^50,51

FLT3 PCR-based Mutation Analysis

In the past 10 years, genetic changes identified in the leukaemic cells at the time of diagnosis have greatly influenced treatment and management of AML. In addition to chromosomal abnormalities, mutations in specific genes have also emerged as important prognostic indicators. Among the most common are mutations in tyrosine kinase genes such as FLT3 and NPM1.^55–57

Principle

The fms-like tyrosine kinase gene 3, FLT3, shows an internal tandem duplication (ITD) in approximately 15–30% of patients with AML. The mutations are restricted to the leukaemic cells and patients with FLT3-ITD have a poorer outcome and may require transplant or more aggressive chemotherapy. Therefore, the rapid identification of these leukaemic-specific changes is important and analysis should be carried out in pre-treatment samples of all AML patients.

FLT3-ITD occurs with higher frequency in older adults than younger patients. Mutations map to the negative regulatory juxta-membrane (JM) domain and change the amino acid sequence, which subsequently interrupts inhibition and constitutively activates the protein.

Between 8% and 12% of AML patients have other types of FLT3 mutation that map to the activation loop, most frequently involving aspartic acid 835 or the immediately adjacent isoleucine 836. Neither mutation appears to carry prognostic significance and therefore they are less often tested for than FLT3-ITD.

Methods

To investigate the presence of an ITD in the FLT3 gene, the method of choice is PCR amplification using primers flanking the region containing the mutation and this can be carried out on DNA or cDNA.

For DNA amplification:

For cDNA amplification:

See Figure 8.12 for PCR strategy and amplicon size.

Figure 8.12 FLT3-ITD mutation analysis. Primers specific for the DNA amplification (G11–G12) and RNA amplification (R5 and R6) can be used on DNA and RNA, respectively. The amplicon size in an unmutated/wild-type allele is given in bp. In the presence of an internal tandem duplication (ITD), the size will increase according to the size of the insertion and, consequently, this may vary from individual to individual and will be restricted to the leukaemic cell population.

PCR cycling conditions are: 95°C for 5 min initial denaturation and then 30 cycles using 95°C for 30 s denaturation step, 56°C for 45 s annealing step and 72°C for 30 s extension with a final 72°C for 7 min extension step. Analysis of amplicon size is carried out using Genescanning on a fragment analyser (Applied Biosystems, Warrington).

Interpretation

A 328 bp amplicon using cDNA (R5F–R6R primers; Fig. 8.12) and a 365 bp amplicon when using DNA (and G11F and G12R primers; Fig. 8.12) are detected in the presence of a wild-type allele. The mutant allele gives rise to a larger amplicon of variable size depending on the size of the internal tandem repeat, different in each individual patient. Because the wild-type and mutant amplicons are very similar in size the PCR product is better visualized using a fragment analyser. This offers a more accurate separation of similar-size products but requires one of the primers used in the reaction to be fluorescently labelled. The samples can be reported as ‘wild-type’ or ‘mutated’ depending on the results obtained and the size of the mutant allele should be noted in the report (in bp).

The other common FLT3 mutation (D835) results in the introduction of an EcoRV restriction site absent in the wild-type allele. This facilitates the identification of this mutation by digestion of the PCR product containing this site.

NPM1 PCR-based Mutation Analysis

The nucleophosmin-1 (NPM1) gene maps to chromosome 5q35. The cDNA has a coding sequence equivalent to a protein of 294 amino acids. Nucleophosmin is an abundant nucleolar phosphoprotein constantly shuffling between the nucleus/nucleolus and cytoplasm. NPM1 mutations represent the most frequent gene alteration in AML. More than 26 different NPM1 mutations have been identified, at breakpoint positions from 956 to 971, characterized by simple 1- or 2-tetranucleotide insertions, a 4-base pair (bp) or 5-bp deletion combined with a 9-bp insertion or a 9-bp deletion combined with a 14-bp insertion. NPM1-mutated AML occurs in all FAB categories except French–American–British (FAB) M3. However, the frequency among the FAB subgroups varies: it is lower in M2 (about 20%) and higher in M4 (acute myelomonocytic leukemia, 40–50%), M5a (acute monoblastic leukaemia, 40–50%) and especially M5b (acute monocytic leukemia, up to 90%). In the World Health Organization 2008 classification, NPM1-mutated AML comprises a specific provisional category. The most common NPM1 mutation type, accounting for 75–80% of cases, is referred to as mutation A (NPM1-mutA). NPM1 gene mutations are common in AML with a normal karyotype, occurring in 50–60% of cases. There is a high frequency of FLT3 gene mutations in NPM1-mutated AML, mostly of the internal tandem duplication type. However, no association between NPM1 gene mutations and TP53, NRAS, CEBPA (CCAAT/enhancer binding protein-α) and MLL gene mutations or recurrent genetic abnormalities has been found. These NPM1 alterations have been shown to possess prognostic significance because they appear to identify patients who will respond well to chemotherapy and may not require transplantation.

Reagents and Methods

It is important to note that since analysis for NPM1 mutations is carried out on cDNA samples, every effort to reduce DNA contamination should be made to avoid cross-amplification of DNA regions which contain pseudogenes. These may amplify using the primers commonly used for this test. To this end, samples are treated with DNAse prior to RNA extraction using Qiagen DNAse Free RNA extraction kit (Cat. No. 79294).

Primer sequences for the test and PCR amplification strategy on cDNA are as follows:

PCR amplification conditions are standard using a 20 μl final volume and 1.5 mM MgCl₂ concentration. Amplification is carried out using the following cycling conditions: initial extended denaturation at 95°C for 5 min; then 30 cycles using a denaturation at 95°C for 30 s, annealing at 56°C for 45 s; extension at 72°C for 30 s; and a final 72°C for 7 min extension.

Interpretation

The NPM1 amplification should result in the production of one peak corresponding to 348 bp wild-type amplicon and another 352 bp mutant amplicon, when the mutant version with an insertion of 4 bp is present. The samples can be reported as ‘wild-type’ or ‘mutated’ and the size of the mutant allele should be noted in the report (in bp).

The test is analysed on a gene-scanning apparatus and frequently run as a duplex test in combination with FLT3. This is possible because NPM1 and FLT3 primer pairs have been designed to yield fragments of different sizes.

Principle

Provided the amplification cycle number is kept reasonably low (25 cycles), the PCR reaction is semi-quantitative and the amount of product will reflect the amount of starting material. Therefore, once an informative difference is established, the amount of PCR product of the different host and donor alleles will reflect the proportions of host and donor DNA in a sample. Fluorescent-labelled primers are used in multiplex PCR reactions that are run on a capillary-based genetic analyser.

Method

This method has been modified from Mann et al.⁵⁹ by Griffiths and Mason (pers. comm.). Five primer pairs are used, as listed in Table 8.5. PCR reactions are carried out using buffer III with 1.5 mM MgCl₂. Amplification conditions are 94°C for 5 min, then 25 cycles of 94°C for 30 s, 57°C for 1 min and 72°C for 2 min followed by an extension at 72°C for 5 min. Products are then analysed by genotyping as follows. One μl of the PCR product, which may need to be diluted from 1:4 to 1:10 in water, is added to 10 μl of formamide containing the size marker Rox 500 (Applied Biosystems, Warrington) diluted 7.5 μl per 500 μl of formamide, aliquoted into an optical 96-well reaction plate and run on the ABI 3700 DNA analyser (or equivalent). Peaks representing the DNA fragments are visualized with the Genotyper software (Applied Biosystems, Warrington).

Interpretation

Comparing the host (pre-transplant) and donor DNA samples, an informative difference is sought at one or more of the STR loci such that host- and/or donor-specific alleles are identified. Correction factors are established by comparing the relative peak heights of the two host or donor alleles at these STRs. Post-transplant, the area under each informative peak is recorded and used to assess the relative proportion of host and donor DNA. Examples of full-engraftment and mixed chimerism are given in Figure 8.13.

Figure 8.13 Short tandem repeat (STR) analysis of bone marrow engraftment. Capillary electrophoresis of the polymerase chain reaction (PCR) products of STR amplification as viewed by the Genotype software. H, host DNA; D, donor DNA; P, post-transplant DNA. (A) The host and donor share an allele at 352 bp, whereas a host-specific allele is seen at 376 bp with the marker D18S386; the latter is not detected in the post-transplant sample. (B) The host and donor share an allele at 266 bp but have specific alleles at 286 and 282 bp, respectively. All three alleles are seen in the post-transplant DNA, indicating that the sample is chimeric. From the relative area under the host- and donor-specific peaks, the proportion of donor DNA is estimated at 39%.