Jaroslaw P. Maciejewski and Bartlomiej Przychodzen
BACKGROUND
The application of molecular biology and genetic techniques has greatly contributed to recent advances in hematology. Many new technologies have found utility in the clinical routine. This chapter illustrates the application of molecular techniques in the diagnosis of hematologic diseases and explains the principles and details of the most commonly used tests. The individual techniques are described in the context of specific applications; many of the methods are applied in a variety of diseases described in specific chapters of this handbook.
The most essential technologies of the next decades will include:
Polymerase chain reaction (PCR) and its modification
Sanger sequencing
DNA microarray technologies
Next generation sequencing (NGS)
In particular, these technologies are of fundamental importance for many derived diagnostic tests, and we provide here a short overview of their general principles, while subsequent paragraphs show specific diagnostic areas and modifications. Of note is that many indirect methods of sequence variant determination, such as melting curve analysis, restriction fragment polymorphism, PCR amplification with sequence-specific primers (SSP) and hybridization with a sequence-specific oligonucleotide probe (SSOP), will be increasingly replaced by direct sequencing.
Polymerase Chain Reaction
PCR revolutionized molecular diagnostics in hematology; various modifications of this technique exist. Both DNA and RNA reverse transcribed into cDNA can be used as a template. In the presence of forward and reverse DNA primers that bind to the sequence-specific regions of the target DNA, Taq polymerase extends both strands of the DNA. Repeated cycles of annealing, extension, and denaturation lead to the exponential amplification of the targeted DNA sequence with the specificity provided by the DNA primers (Fig. 29.1). PCR primers can be designed to distinguish polymorphic sequences; primers can be labeled to facilitate detection or quantitation. Various modifications of the basic PCR technologies have been described below in conjunction with specific applications.
FIGURE 29.1 Principle of polymerase chain reaction.Template consists of either DNA or cDNA generated by reverse transcription of mRNA. In the presence of forward and reverse DNA primers that bind to the sequence-specific region of the target DNA,Taq polymerase extends both strands of the DNA such that repeated cycles of annealing, primer extension, and denaturation lead to amplification and accumulation of the targeted newly synthesized DNA sequence with the specificity provided by DNA primers.
Traditional Sequencing
The most popular method uses modified chain-termination method and was introduced by Sanger in the early 1970s. Sanger sequencing relies on the use of color-labeled, dideoxynucleotide chain terminators (ddNTPs). Compared to a regular reaction mixture, it consists four, regular deoxynucleotides (dNTPs) mixed with color-labeled ddNTPs. In principle, when the ddNTP is added at the end of the fragment it restricts further elongation of the DNA chain. Since the addition of the dNTPs and ddNTPs is random, a whole array of fragments of different lengths will be produced. Each fragment will be end-labeled with only one dye that represents one of four different nucleotides. Because all the fragments have different sizes, they can be separated and visualized by capillary electrophoresis. In capillary electrophoresis, each color-labeled fragment migrates according to its length, shorter fragments migrating faster. At the end of the capillary, each fragment is analyzed by a laser beam and fluorescence detector. Since each fragment reaches the detector sequentially, nucleotide sequence is reconstructed from a wavelength chromatogram; in most cases, only a small portion of the gene of interest (usually 1 exon per reaction) is sequenced. Consequently, even single-gene sequencing is extremely labor and cost intensive, requiring multiple PCR reactions and site-specific sets of primers.
New Generation Sequencing
NGS overcomes the limitation of Sanger sequencing by using a massive, parallel sequencing approach (Fig. 29.2). DNA is randomly fragmented and millions of fragmented DNA molecules are physically arrayed and attached to universal adaptors on a solid surface. A set of universal primers is then used to massively amplify and ultimately sequence millions of fragments in parallel. Elongation of each of the small fragments is similar to conventional DNA sequencing. Four different color-labeled terminator nucleotides are added sequentially, elongating the DNA chain, unincorporated nucleotides being washed off. After one terminator nucleotide is successfully added, it is able to emit one of the four distinct colors when the laser beam is applied. With the advent of ultra-sensitive charged coupled device (CCD) cameras, added nucleotides can be distinguished. Since the process is done in parallel, “reading” of each of millions of sequenced fragments is accomplished. Finally, the cycle is finished with removal of reversible terminator. By repeating this cycle multiple times (until the template fragment is fully extended), the complete sequence of the fragment, one nucleotide at the time, is completed. The lengths of the fragments are determined by the number of cycles that are needed to completely sequence a given DNA fragment. The process produces millions of short fragments (reads) that need to be organized into meaningful, organized, continuous sequences corresponding to specific target genes or chromosomal regions or transcripts (if mRNA is the starting material). Computer algorithms and sequence aligners assemble the sequences for analysis.
FIGURE 29.2 Principle of next generation sequencing.The first step usually consists of DNA/RNA fragmentation. It is done either enzymatically (restriction enzyme) or mechanically (sonication). Subsequently all the fragments are arrayed by hybridization to universal adapters attached to a solid surface. Successful attachment along with proper dilution guarantees only one fragment per location.This enables each fragment to be sequenced in parallel. Usually fragments are being copied prior to sequencing. Sequencing proceeds in four steps. The first step is the addition of a reversible terminator that results in elongation by one nucleotide. Secondly, all the unincorporated nucleotides are removed. Subsequently the whole reaction surface with all the fragments is scanned using the laser beam and CCD camera.The last step involves modification of the termination moiety, and the cycle can start over.The number of cycles represents the length of millions of fragments sequenced.
DETECTION OF INDIVIDUAL GERMLINE MUTATIONS/ POLYMORPHISMS
Precise diagnosis of many hematologic diseases or detection of susceptibility to develop complications depends on the identification of mutated genes. Clinically applicable methods mostly involve detection of defined mutations occurring at specific sites within the genes. Currently, most protocols use PCR to amplify the involved gene fragments. For the identification of the presence of individual mutations, various methods can be used (Fig. 29.3). For example, they are applied in routine diagnosis of genetic hematologic diseases including thalassemia and other hemoglobinopathies, hereditary familial hemochromatosis (HFE) gene mutations such as C282Y and H63D, factor V Leiden, prothrombin gene mutations G20210A, and thermolabile C677T 5,10-methylenetetrahydrofolate reductase.1,2 Similar methods can be applied for the detection of other clinically relevant mutations or polymorphisms.
FIGURE 29.3 Application of PCR technology for the detection of a gene mutation.Various techniques based on PCR can be used for specific applications in hematology. Fluorescent primers can be used for determining the small differences in the size of the amplified product, a technique called genotyping. Fluorescent probes can also be selected to hybridize between the primer sequences of the template allowing design of real-time PCR. DNA amplicons generated in the process of PCR reactions can be used for restriction endonuclease digestion. If restriction sites for specific enzymes within the amplicon contain a mutation, restriction endonuclease digestion of the PCR product will result in fragments of different sizes, which can be resolved on either capillary or agarose gel electrophoresis. Finally, using specific fluorescent probes that hybridize to the amplified sequences, melting curves can be recorded to distinguish individual alleles.The presence of sequence differences between the probe and template results in different melting curves; these curves are recorded based on the emission of light induced by melting off the fluorescent probes from the template.
Restriction Fragment Length Polymorphism Analysis
Prior to the advent of PCR technology, traditional Southern blotting of genomic DNA followed by probe hybridization was used to detect changes in the endonuclease restriction patterns. Currently, restriction fragment length polymorphism (RFLP) analysis is used in conjunction with PCR amplification. Restriction digests can be performed either prior to or after amplification. If a mutation affects the restriction endonuclease digestion patterns, its presence can be easily demonstrated using RFLP analysis. After PCR amplification of a relevant gene fragment that carries a specific mutation, the resulting amplicons are subjected to restriction endonuclease cleavage. Using gel electrophoresis, changes in the fragment size can be demonstrated. Through comparison with a wild-type form, heterozygote and homozygote patterns can be easily distinguished. When a fluorochrome-labeled primer is used, capillary gel electrophoresis can be applied allowing for high sensitivity and throughput. Detection of HFE mutations by RFLP of PCR products serves as an example for this technique.
Melting Curve Analysis of Polymerase Chain Reaction Products
More recently, a light-cycler PCR method combined with melting curve analysis has been used for the detection of gene mutations allowing for a reduction in the workload and enabling automation. Melting curve analysis exploits the fact that even a single-nucleotide mismatch between the labeled probe and the targeted sequence significantly reduces the melting temperature. Consequently, amplicon/probe mismatches will melt off at lower temperatures different than that of matched target DNA. PCR amplification of specific gene fragments is performed in the presence of a fluorescent DNA probe or probes (anchor and sensor probe) that release light on hybridization to the internal portion of the amplicon containing the potential mutation. After completion of the reaction, the hybridized fragments are denatured; the release of the probe decreases the amount of the emitted fluorescence, a process recorded in the form of a melting curve. The shape of the melting curves identifies the presence of two normal alleles (singular curve), or heterozygotes (two peaks). For mutation homozygotes the curve is shifted, producing a singular characteristic peak. If multiple mutations are present in a gene, specific probes and primers must be applied to detect heterozygotes, homozygotes, and compound heterozygotes.
Allele-Specific Polymerase Chain Reaction
The tetra-primer amplification refractory mutation system (ARMS) PCR is one of the variants of allele-specific PCR; it allows for detection of single nucleotide polymorphisms (SNPs) as well as single-gene mutations. Two different allele-specific amplicons and a larger (non-allele-specific) control amplicon are generated by a pair of two common (outer) primers and by two allele-specific (inner) primers that have opposite orientation (allele 1–specific primer, antisense, and allele 2–specific primer, sense). Because the common primers are designed so that the mutation is located nearer one of them, the two allele-specific amplicons will have different lengths and thus easily separated by gel electrophoresis: the wild-type genotype generates two bands on gel electrophoresis, homozygous mutation generates two bands, and heterozygous mutation generates all three bands.
Direct Polymerase Chain Reaction–Amplified Product Sequencing
Alternative methods of mutation analysis include direct sequencing of PCR products. Both alleles can be easily identified, and the direct sequencing method has the advantage that it does not target a specific mutation and that all possible sequence differences can be detected within the sequenced gene region.
MOLECULAR DIAGNOSIS OF HEMOGLOBINOPATHIES
Hemoglobinopathies constitute a large group of inherited autosomal recessive hematologic disorders. While routine laboratory tests and clinical presentation are often sufficient for a proper diagnosis, molecular analysis is mandatory for the confirmation of the defect and precise characterization of the abnormal hemoglobin.3 For example, combinations of specific mutations may greatly affect the phenotype expected in the progeny. Thus, molecular diagnosis may have significant consequences for counseling affected patients, asymptomatic carriers, and prenatal diagnostics.
Traditionally, Southern blotting was used, but, recently, PCR-based methods are preferred. Allele-specific oligonucleotide (ASO) hybridization and allele-specific priming are the most commonly applied techniques. The first method relies on hybridization of ASO probes (wild type and mutant) to PCRamplified genomic DNA. In the dot-blot assay, ASO is labeled, while the reverse dot-blot technique utilizes labeled amplified DNA, allowing for simultaneous screening of multiple mutations. Allele-specific priming is based on the principle that a perfectly matched primer pair amplifies target DNA more efficiently than a mismatched pair. In the ARMS, genomic DNA is challenged with both wild-type and mutant primer sets. Multiple mutations may be simultaneously screened in a multiplex PCR assay using fluorescently labeled ARMS primers, producing products of different length that can be detected using an automated DNA analyzer. Large deletions of both the α and β globin gene may be screened using the gap-PCR, with primers complementary to the breakpoint sequences. However, for some deletion mutants, Southern blotting is still standard.
Combining all these approaches in the context of the ethnic- and region-specific distribution of globin mutations, successful molecular identification is possible in more than 90% of cases. Mutations remaining unknown after standard molecular screening may be investigated further by denaturing gradient gel electrophoresis or heteroduplex analysis; nevertheless, complete sequencing of the globin gene represents the best option to identify rare or unknown mutations.
CYTOGENETIC DIAGNOSTICS
Metaphase Karyotyping
Traditional cytogenetics, utilizing banding techniques, is performed on chromosomal metaphase spreads. Because mitotic activity is required, metaphase karyotyping is performed after cell culture in the presence of mitogens. For myeloid disorders, either lymphocyte-conditioned media or hematopoietic growth factors are most commonly used, while, for lymphoid malignancies, lectins are added. Various banding methods have been utilized for chromosome identification and resolution of individual chromosomal fragments, but G-banding is usual in clinical diagnostics. Characteristic bands result from the biochemical properties of chromatin such as AT and GC content.4-8
Cellularity and mitotic activity affect the diagnostic yield of the procedure, and the proportion of noninformative spreads varies from disease to disease. In myelofibrosis, marrow is often not aspirable. In aplastic anemia and myelodysplasia, noninformative results are frequent due to the lack of progenitor cells. In such cases, cytogenetic analysis may be also performed on blood specimens.
Approximately 330 chromosomal bands can be distinguished by routine karyotyping, and each band may contain as much as 107 base pairs (bp) and a multitude of genes. Classic karyotyping can identify defects of approximately 5 Mb; thus, smaller defects and their locations may remain undetected (resolution). The sensitivity level depends on the number of analyzed cells; routinely 20 cells are counted with a detailed analysis of at least 2 cells. Analysis may be more complicated if several clones, each harboring a distinct defect, are present. Depending on the nature of the identified defect, the sensitivity limit is approximately 10% (i.e., identification of 2 abnormal cells) in 20 cells tested.
Both balanced and unbalanced translocations can be identified, but some defects may require a more intricate analysis. Some of the balanced translocations are highly diagnostic; examples are t(9;22) in chronic myelogenous leukemia (CML), t(15;17), inv 16 and t(8:21) in acute myeloid lymphoma (AML), t(15:17) in acute promyelocytic leukemia (APL), t(9:22) and t(12:21) in acute lymphoblastic leukemia (ALL), as well as t(14:18), t(11:14), t(11:18) in lymphomas. Once a specific defect is identified, metaphase karyotyping can be used for monitoring of therapy response (cytogenetic remission); however, the sensitivity of this method is limited.5-8
Fluorescence In Situ Hybridization
For the targeted detection of specific abnormalities, fluorescence in situ hybridization (FISH) is the most commonly applied method particularly helpful in the characterization of structural chromosomal abnormalities and identification of chromosomes of uncertain origin. However, FISH is not suitable for screening for unknown defects unless a high clinical suspicion exists. FISH does not require cell division and consequently cell culture, and is more sensitive than traditional cytogenetics. FISH provides a more accurate measure for the true frequency of abnormal cells and can be used for the monitoring of minimal residual disease (MRD). Identification of the donor versus recipient origin of the blood cell production following hematopoietic stem cell transplantation is another application of this technology (see below). The technique can be applied to blood, marrow, body fluids, tissue touch preparations as well as to paraffin-embedded tissues.5,9
In FISH, specific fluorescently labeled single-stranded DNA probes are hybridized to the nuclei of metaphase or interphase cells attached to glass slides. The use of probes labeled with different dyes allows for multicolor FISH on a single slide. Probes can also be designed to identify a specific chromosomal structure, hybridize to multiple chromosomal sequences, and to identify unique DNA sequences. Probes recognizing α-satellite sequences are chromosome specific; in diploid cells both chromosomes are labeled. Chromosome painting probes are derived from whole chromosomes (see also spectral karyotyping [SKY], discussed next). Probes can be derived from unique sequences cloned from specific regions of the genome. Finally, telomeric probes can be used to determine the telomere length based on the intensity of the hybridization.
For balanced translocations, probes spanning individual breakpoints are used. Dual-color/dual-fusion probes or single-fusion/dual-color FISH probes target sequences located at opposite ends of two breakpoints. In addition, two-color break-apart probes, recognizing DNA sequences from the 3´ and 5´ends of a single gene, can be applied. These probes yield combined yellow signal in the normal germline configuration while two colors are seen when target sequences are separated because of translocation. FISH is more reliable for the detection of duplication of chromosome fragments than deletions. In general, FISH is less sensitive than PCR, with detection limits of 1 of 100 cells. As a result of the false-positive rate, it is not clear whether sensitivity can be increased through routine counting of a higher number of cells.
FISH techniques have been widely applied for the detection of lymphoma-specific translocations, in the diagnosis of CML, myelodysplastic syndrome (MDS), and T-cell acute lymphoblastic leukemia (T-ALL) and B-cell acute lymphoblastic leukemia (B-ALL) (Table 29.1).6,8-10 In addition, FISH is frequently used for intracellular detection of Epstein-Barr virus (EBV) in certain non-Hodgkin’s lymphomas, Hodgkin’s disease, and aggressive natural killer (NK) cell lymphomas (see later).
Spectral Karyotyping
SKY allows for the visualization of all 24 chromosomes and analysis of their structure based on hybridization with multicolor painting probes.11 These probes are derived from individual chromosomes using degenerate primer-based PCR and differentially labeled with fluorochromes. After hybridization to metaphase spreads, a digital camera is used to record the complete emission spectra. As a result, each chromosome-specific probe is distinctively labeled and easily identified. SKY has a much higher precision than traditional cytogenetics and allows for identification of new, previously unidentified reciprocal translocations and defects that cannot be resolved by traditional banding. In one study, SKY allowed detection of new translocations in 35% of cases and resulted in confirmation of the previously known defects and refinement of 35% of diagnoses.
Comparative Genomic Hybridization and Single Nucleotide Polymorphism Arrays as Karyotyping Platforms
Array-Based Comparative Genomic Hybridization
Array-based comparative genomic hybridization (A-CGH) allows for cytogenetic analysis of chromosomal defects that can be applied in many clonal diseases. In this technique, genomic DNA from malignant cells and normal reference DNA are fragmented, differentially labeled with fluorescent dyes, and cohybridized to immobilized DNA probes. Chromosomal imbalances across the genome in tumor DNA can be detected, quantitated, and the position defined by analysis of the fluorescence intensity of the two different colors. Initially, this analysis was performed on metaphase chromosome preparations (M-CGH). However, the resolution of CGH as applied to metaphase spreads is limited by the standard cytogenetic resolution of approximately 5 Mb, and considerable cytogenetic expertise is required to accomplish such an analysis. Therefore, M-CGH has never become a widely utilized technique, and remained limited to specialist research applications.
Bacterial artificial chromosome was originally used to produce CGH arrays. The advent of oligoarray technology made CGH applicable to the study of genomic alterations in human disease. Sixty-mer oligonucleotide probes corresponding to individual SNPs cover the whole genome and include both coding and noncoding regions; thus an ordered array of DNA segments at high genomic resolution can be generated, circumventing the limitations associated with the use of metaphase preparations as the hybridization template. The fluorescence ratio of the two colors can be compared between different spots representing different genomic regions, providing a genome-wide molecular profile of the sample with respect to regions of the genome that are deleted or amplified (Fig. 29.4A). Resolution is dependent on a combination of the number, size, and map positions of the DNA elements within the array.
Single Nucleotide Polymorphism Array–Based Karyotyping
Recently introduced SNP arrays to study the genetic predisposition of diseases also can be applied for analysis of copy number variation and loss of heterozygosity (LOH). This technique utilizes arrays containing oligonucleotide probes corresponding to SNPs present throughout the human genome.36 Test DNA is fragmented, ligated to universal linkers, amplified with primers corresponding to the linkers, and labeled (Fig. 29.4B). Unlike in CGH, there is no requirement for normal reference DNA. Following hybridization, fluorescence intensity is measured for each spot on the array. After bioinformatic analysis, genotyping calls are possible for each SNP in order to determine homozygosity or heterozygosity for each SNP. In addition to genotyping, the copy number for each locus (tagged by a specific SNP probe) can be deduced based on the fluorescence intensity that corresponds to hyperploid or hypoploid gene copy number. A whole variety of array designs (bead hybridization or array liquid phase hybridization) platforms are currently available with a density of >1 million markers covering 22 autosomes and the X chromosome. The average intermarker distance is approximately 10 Kb, resulting in a superior resolution. Initially designed for genotyping and whole genome association studies, SNP-A also can be used for karyotyping. As a karyotyping tool, SNP-A allows for a very high resolution (depending on the density of the SNP probes), does not require cell division (no culture and proliferation of cells is needed) but, due to the nature of the technology allows for detection of only balanced translocations. When compared to metaphase cytogenetics, SNP-A-based karyotyping allows for detection of clonal unbalanced chromosomal defects in a higher percentage of patients with hematologic malignancies including MDS, Multiple Myeloma (MM), AML, and chronic leukemia/lymphoma (CLL). The additional advantage over metaphase karyotyping and A-CGH is the ability to detect copy number neutral LOH (uniparental disomy), present in many solid tumors and myeloid malignancies. The sensitivity of SNP-A karyotyping is relatively low (as it depends on the proportion of clonal cells in the sample), comparable to metaphase karyotyping. In addition to whole genome arrays with various probe densities, customized SNP arrays are available or can be designed for specific applications: targeting non-synonymous SNPs, particular set of SNPs or specific regions of the genome (as for the human leukocyte antigen, HLA, locus).
Because of the high resolution of SNP-A and the convenient microarray format, these methods will likely be introduced into the clinical routine, especially for diseases in which unbalanced translocations are expected.12
FIGURE 29.4 A. Array-based comparative genomic hybridization (A-CGH). A-CGH consists of hybridization of tester and reference DNA samples that have been differentially labeled with fluorochromes to arrays of sequences corresponding to specific portions of chromosomes. Unbalanced translocations, such as deletion or duplication of individual genes (or portions of chromosomes) can be detected based on the disparity between fluorescent spectra emitted by the tester vs. reference DNA. Consequently, depending on the number of probes, a very intricate analysis of chromosomes can be performed with regard to the presence or absence of specific DNA segments. B. SNP array–based karyotyping. The principle of the method relies on amplification and labeling of fragmented genomic PCR products with subsequent hybridization to arrays containing oligonucleotide probes homologous to allelic variants of SNPs. Based on the density of the arrays and choice/location of SNPs to be detected, various resolution levels can be achieved. Bioinformatic analysis of hybridization results allows for detection of copy number changes (fluorescence intensity) for individual loci as well as loss of heterozygosity that can be a result of deletion or segmental uniparental disomy due to mitotic recombination.
QUANTITATION OF SOMATIC MUTATIONS AND TRANSLOCATIONS
Polymerase Chain Reaction–Based Analysis of Translocations
PCR has found a wide application in the diagnosis of malignant disorders associated with specific translocations of genetic material.5,7,8 The main advantage of PCR is the high specificity and sensitivity, but there is need for immaculate technique to avoid contamination. DNA primers are designed to flank the specific translocated region, producing a PCR product with a characteristic size, while in the absence of the specific translocation the amplification product is not generated. Appropriate controls can be incorporated into the reaction. Because of the higher template copy number per cell, reverse transcriptase (RT) real-time PCR may be a more sensitive modification of this technique. In real-time PCR, mRNA of the abnormal transcript is reverse transcribed and cDNA serves as template for the amplification reaction (Fig. 29.5). Sensitivity and specificity can be further improved by an additional round of amplification with a pair of internal primers, termed nested PCR. The sensitivity of this method approaches 1 malignant cell per 106 normal cells, allowing for the assessment of MRD in various conditions.5
FIGURE 29.5 Real-time quantitative PCR. Either cDNA generated from mRNA or DNA can be used as a template.The principle of real-time PCR consists of target sequence amplification in the presence of fluorochrome-labeled probes. Such probes are designed to target the sequences between the forward and reverse primer that hybridize to the PCR product accumulated during amplification.The probes are usually labeled with a reporter fluorochrome at the 5´ and a quencher fluorochrome at the 3´ end. Probes are designed to have a higher melting temperature than the extension primer. As long as both fluorochromes are connected through the DNA sequence, no light is emitted. However, the 5´ to 3´ exonuclease activity of Taq polymerase degrades the probe and releases the fluorochromes. Consequently, progression of the reaction can be monitored by the detection of the fluorescent signal generated during the exponential phase of the reaction and the cycle number at which the reporter dye intensity rises above the background noise.This value, also referred to as the threshold number, is inversely related to the copy number of the target template.
Recently, introduction of the real-time light-cycler PCR assay (Fig. 29.5) has led to quantification of the numbers of cells up to one malignant cell per 105 normal cells. The reference standard includes a single-copy gene. The principle of real-time PCR consists of target sequence amplification in the presence of fluorochrome-labeled probes. Such probes are designed to target the sequences between the forward and reverse primers. The probe is labeled at the 5´ end with a reporter fluorochrome (6-FAM) and a quencher fluorochrome (6-carboxytetramethyl-rhodamine [TAMRA]) at the 3´ end and are designed to have a higher melting temperature than the extension primers. As long as both fluorochromes are connected through the DNA sequence, no light is emitted. However, the 5´ to 3´ exonuclease activity of Taq polymerase degrades the probe and releases the fluorochromes. Consequently, progression of the reaction can be monitored by the detection of the fluorescent signal generated during the exponential phase of the reaction, and the cycle number at which the reporter dye intensity rises above the background noise. This threshold number is inversely related to the copy number of the target template. Measurement of the frequency of abnormal cells is based on standard curves and is generated with dilution of control cells or DNA/cDNA containing the targeted mutation. The results can be expressed as copy numbers of fusion transcripts per microgram of RNA or as the frequency of abnormal cells. Ubiquitously expressed housekeeping transcripts/genes are commonly incorporated, and the PCR cycle threshold number of the fusion gene is normalized to the value of the housekeeping gene.5
In clinical practice, PCR technology, including quantitative PCR, has proven predictive of therapy response and relapse. For certain hematologic malignancies with translocations coding for specific targets, cytogenetic and molecular remissions have been defined as specific end points of therapy. Risks of relapse associated with molecular versus cytogenetic remission have been determined. The most common applications of PCR in the detection of disease-defining translocations are for CML (bcr/abl), APL (PML/RARA), mantle cell and follicular lymphomas (cycD1/IgH and IgH/bcl2, respectively), and B-cell lymphoblastic leukemia (bcr/abl, rearranged IgH) (Table 29.2).
Clearly, separate screening for all described abnormalities is too expensive and time consuming to be applied as a global diagnostic battery. Multiple attempts have been made to adapt PCR technology for precise molecular diagnosis associated with individual leukemias and lymphomas. In multiplex PCR, multiple primer pairs are combined to enable detection of several translocations in a single PCR reaction. For example, primer mixtures have been designed to allow for the detection of 28 different translocations, including 80 breakpoints and splice variants, using a limited number of PCR reactions. The introduction of such technologies in clinical practice is hampered by the need for positive controls and limited sensitivity. However, the potential value of more comprehensive screening techniques has been demonstrated in cases that showed PCR positivity for PML/RARA transcripts despite M3 morphology or the cytogenetic presence of t(15:17), and cryptic translocations such as t(12:21) or t(4:11) in ALL and in AML.
Detection of Somatic Gene Mutations
Acquired mutations of individual genes can lead to the acquisition of a malignant phenotype and their presence or absence may have clinical significance.5 The principle of PCR-based detection of such mutations is similar to that used for germline mutations (see above). Methods include direct DNA PCR, if the mutation results in a length difference between the amplified fragments (e.g., internal tandem duplication of the FTL-3 gene13) wild-type and mutated amplicons will be clearly identified by electrophoresis. When a labeled primer is used for amplification, even small size differences can be resolved by capillary gel electrophoresis. Single-nucleotide changes can also be identified when the mutation produces a new restriction site or abolishes an existing one. Consequently, after amplification, PCR products are digested with appropriate restriction endonucleases and subjected to electrophoresis to identify the presence of the mutation (see above RFLP). Allele-specific PCR with primers designed to amplify either the mutated or wild-type allele is currently the most commonly employed method. Finally, exons affected by recurrent mutations can be amplified and directly sequenced.
Apart from canonical mutations or genes affected by a few invariant mutations, mutational testing may be very difficult for larger genes (multiple exons) affected by diverse mutations located in various exons. PCR-based site-targeted methods, including direct sequencing of the amplified PCR product may be difficult to implement for such genes due to the labor intensity involved. NGS sequencing may alleviate some of these problems.
In contrast to the detection of germline mutation, using PCR technologies, the proportion of mutated cells in the sample may affect the results of the test: even if 100% of cells carry the mutation, only 50% of the alleles will be affected. The current detection limits of PCR-based techniques can reach 10% of cells carrying the mutation.
Of note is that the choice of cellular material used for the detection of single-gene mutations may influence the results with regard to the status of homozygous mutations, such as Jak2V617F. In most methods beginning with unfractionated cells, the distinction between heterozygous and homozygous may not be possible due to contamination with wild-type cells.
Most recently, NGS technologies have been adopted for detection of somatic lesions. The diagnostic strategies can include unbiased approaches, in which whole genome, exome or individual chromosomes are sequenced and either known invariant mutations are identified and reported or a comparison is made with the parallel sequenced germline sample, all somatic mutations are identified. Various algorithms have been developed to distinguish false-positive results due to technologic errors and germline-encoded sequence alterations from true somatic events. Another NGS approach is targeted, deep sequencing, which can include enrichment steps for exons of selected or single genes. Because of the smaller overall size of the amplicon to be sequenced, a greater sequencing depth can be achieved (“deep sequencing”), which allows for assessment of mutational burden/clonal size. Such technologies may be used in the future to monitor the therapy effectiveness or MRD.
CLONALITY STUDIES
Where acquired defects (point mutations, translocations) have been identified, they can serve as a suitable marker. However, in many clinical situations such a marker is not available. Consequently, a number of clonality assays have been developed that allow for the diagnosis of oligoclonal or skewed hematopoietic function.
X Chromosome Inactivation Pattern Analysis
X chromosome inactivation pattern (XCIP) analysis is particularly useful in the analysis of disorders without a disease-specific clonality marker.14 XCIP clonality analysis can be informative in a large proportion of female patients. Clearly, clonal/oligoclonal XCIP does not secure the diagnosis of malignancy but may be complementary with other clinical signs and laboratory results.
XCIP is based on the inactivation of one X chromosome in female mammalian cells.14 The inactivation pattern is random, laid down early in embryogenesis and stably inherited by all daughter cells. The inactivation mechanism includes methylation of certain portions of the DNA. Based on the single-cell theory of malignant disorders, XCIP representation should change in the affected tissue, most significantly in blood. Consequently, the distinction between the two chromosomes made by using a polymorphic marker gene located on X chromosome and differentiation between the active and inactive X chromosomes underlie XCIP clonality analysis. While a disease-specific marker is not required, the pathologic change can be extrapolated from the expected pattern. The most commonly used polymorphic markers include human androgen receptors (HUMARA), phosphoglycerate kinase, or the fragile X (FRM1) genes. The heterozygosity rate for HUMARA is approximately 90%. Modern XCIP analysis techniques utilize PCR technology.
In the HUMARA assay, DNA is digested with a methylation-sensitive restriction endonuclease, amplified, and the electrophoretic pattern of the amplicons compared to the undigested amplicons. HpaII digests the unmethylated allele and leaves the methylated allele available for amplification. Comparison of the intensity of the bands allows extrapolation as to the skewing of the normally equally distributed amplicons of the methylated and unmethylated gene fragment.
In other common assays, single-base or short-tandem-repeat (STR) polymorphisms in the coding sequences of X chromosomal genes distinguish between the usage of the inactive or active chromosome by RNA-based techniques. RT-PCR is followed by RFLP of the amplicons. Interpretation of the results of the XCIP analysis must account for age-rated skewing of inactivation and also for a nonrandom pattern among normal cells that may be encountered in up to 25% of women.
ANALYSIS OF T-CELL RECEPTOR AND IMMUNOGLOBULIN REARRANGEMENT
During T- and B-cell ontogeny, rearrangement of VDJ genes of heavy (H) and light (L) Ig chains and α (A) and β (B) or δ (D) and γ (G) chains of T-cell receptor (TCR), respectively, provides the molecular basis for the heterogeneity of B- and T-cell recognition repertoire.15-17
For immunoglobulin (Ig) H-chain, there are at least 40 functional variable (VH) gene fragments, 27 diversity (DH) fragments, 6 functional junctional (JH) fragments, and several constant (CH) gene segments. The Igκ gene complex consists of 35 Vκ gene segments, 5 Jκ, and a single Cκ gene. The Igλ gene is generated through recombination of 30 Vλ segments and 4 Cλ segments, all preceded by a Jλ region.
The TCR is a homodimer consisting either of TCR α and β or TCR δ and γ chains. Similar to Ig genes, α and δ chains are encoded by recombined VDJ fragments (65 VB segments, 7 JB fragments, and 2 CB segments, each preceded by a DB region for β chain and 8 VD, 3 DD, and 4 JD gene segments, as well as a single CD region). TCR α and γ chain are generated through recombination of V and J regions. TCR α-chain gene complex consists of more than 50 VA, 61JA, and 1 CA gene segments. The TCR γ gene complex consists of 6 VG and 2 CG segments, each preceded by 2 or 3 JG1 or JG2 gene segments. Ig and TCR rearrangements are not lineage-restricted: B and T cells may contain complete or incomplete cross-lineage rearrangements that can be used for assessment of clonality. Clonal expansion results in overrepresentation of the specifically rearranged Ig or TCR configuration, a property of lymphocytes exploited for the diagnosis of T- and B-cell malignancies.
Immunoglobulin and T-Cell Receptor Junctional Region Polymerase Chain Reaction
PCR analysis of Ig and TCR gene segments is based on selective amplification of junctional regions. Amplification is only possible when the Ig or TCR gene are juxtaposed through rearrangement, because the distance between these gene segments in the normal germline configuration is too large for PCR amplification. In contrast to the Southern blot,18 PCR relies on both the combinatorial and the junctional diversity of the appropriate rearrangements. PCR can be easily applied to blood and tissue specimens including lymph nodes and skin (particularly helpful in the diagnosis of cutaneous T-cell lymphoma). For both IgH and TCR amplification, sets of consensus primers covering the full spectrum of the V regions combined with sets of JB primers are used in multiplex PCR. For B-cell populations, analysis of the IgH rearrangement is most informative, as the IgH locus rearranges first; a complete VH-JH rearrangement is usually investigated. However, as somatic hypermutation within the IgV gene hampers annealing and amplification, analysis of incomplete DH-JH rearrangement may also be helpful in identifying non-productive partial rearrangement in some immature B cells. Additionally, comprehensive Ig clonality assessment may include analysis of IgL chain, especially the Igκ locus. As discussed above, given the hierarchical rearrangement of IgL, all mature B cells have either a productive or a non-productive Igκ rearrangement (generally involving the κde element) if cells are Igλ+. Further analysis of the IgL genes may be added to increase the sensitivity of clonality assessment.
For T-cell clonality, analysis of the γ TCR locus has been the paradigm; the main advantage is that the rearrangement of the γ chain occurs early and is present in both α/β and δ/γ cells, and in some B cells. Furthermore, the number of primers required for all the possible combinations is small, and junctional diversity is limited compared to other TCR loci. Rearrangement of β genes is a very powerful tool for detection of α/β clonality; even considering the extreme combinatorial diversity, amplification of almost the entire repertoire may be obtained using a relatively limited number of reactions containing appropriate sets of primers. Both complete VB-JB and incomplete DB-JB rearrangements may be studied, increasing the sensitivity of the method; the extreme junctional diversity of the B locus gives high sensitivity to β TCR clonality, even if sophisticated analysis of the PCR product may be required (see below). Analysis of the D locus is relatively easy and may add information on immature T cells and G/D populations, but is usually not informative for A/B cells. In contrast, α genes are not too helpful given the extreme complexity of gene segments, and concomitant rearrangement of β locus.
Two modifications of PCR are currently used for the detection of clonality: heteroduplex electrophoresis19 and genescan analysis. In both methods, PCR amplicons are analyzed for fine discrimination and identification of identical products, utilizing different biologic properties of DNA.
Genotyping of rearranged TCR or Ig genes relies on the amplification of the gene fragments with primers that are labeled with a fluorochrome and detection of the labeled products using capillary gel electrophoresis.20 Most capillary electrophoresis-based automated gene sequencers can be adopted for this technique, which allows for resolution of PCR products that usually vary in size by multiples of 3 bp. Under normal circumstances, the amplification products are polyclonal, showing a large number of distinct peaks on capillary electrophoresis tracing. Best results are seen with IgH and TCR B loci, which both have a high frequency of in-frame V-J rearrangements, while for other loci the spacing between peaks is not preserved because of out-of-frame (nonproductive) or incomplete rearrangements. If a monoclonal population is present, a distinct singular peak is present that corresponds to the immunodominant clone. An equivalent technique can be used for the identification of Ig rearrangement. The sensitivity of genotyping is around 5% of malignant (clonal) cells in a cell mixture. As with heteroduplex analysis, frequently a biallelic rearrangement can be detected.
As discussed previously, Ig and TCR clonality can be assessed regardless of the lineage restriction of a putatively clonal population; however, a classic genescan pattern with triplet spacing of peaks is limited to functional in-frame rearrangements. In this case, representation of productive rearrangements may be also obtained by RT-PCR starting from RNA.
Rearranged dominant junctional CDR3 sequences can be sequenced and used for the design of clonotypic PCR primers that uniquely amplify only the malignant clone. Using nested PCR and variable and J primers, further rounds of amplification with a clonotype specific primer (such as an internal J primer) can be added. Such an approach may increase the sensitivity and specificity of detection.
The application of TCR and Ig rearrangement analysis include determinations of B- and T-cell clonality in blood, marrow, lymph nodes, and skin lesions for the diagnosis of T-ALL, B-ALL, multiple myeloma, lymphomas, large granular lymphocyte (LGL) leukemia, and CLL. In CLL, Ig rearrangement analysis is a major prognostic marker (see also Chapter 14). Detailed sequencing of clonal IgH may reveal homology to the germline IgV gene. Somatic hypermutations of the IgV gene usually occur after antigen priming and, consequently, the IgV mutation status allows for distinction between pre- and postgerminal center CLL, entities with different clinical behavior.
MOLECULAR DIAGNOSIS OF INFECTIONS IN HEMATOLOGY
Molecular methods are increasingly supplementing serologic and histochemical methods in microbiology.21 Precise detection, localization, and quantitation of the nucleic acids of pathogenic viruses allows for better distinction of individual disease entities and are often used for therapeutic decisions. For instance, detection of EBV activation is essential for early diagnosis and therapy of posttransplant lymphoproliferative disorder. For DNA viruses such as herpes viruses, detection of mRNA transcripts by means of RT-PCR allows diagnosis of active infections, while DNA PCR is positive even when only latent virus in present. Molecular techniques are most often used for the diagnosis to EBV, cytomegalovirus (CMV), human herpes virus-6 (HHV-6), as well as retroviruses such as human T-cell leukemia virus (HTLV-1; Table 29.3).22,23
Epstein-Barr Virus
Both PCR and FISH can detect EBV in lymphoid neoplasms (Table 29.3). FISH probes complementary to EBV localize the virus to the malignant cells. Because of its sensitivity and lack of quantitation, DNA PCR may provide positive results without clinical relevance due to the frequency of latent EBV infection in the community. In contrast, quantitative light-cycler PCR assays provide highly precise copy numbers, but they are less sensitive methods. Both Taqman probes (single-stranded DNA probes that release light on degradation by Taq polymerase; see previous discussion) and beacon probes (DNA probes that release light on conformational change after hybridization to the internal segment of the amplicons) are used. TaqMan PCR is performed as described above for the detection of translocations. Sensitivity levels can be as low as 1 viral copy per 1 to 2 × 105 cells.24 Viruses are always cell-associated and DNA for analysis is extracted from blood leukocytes.
FISH performed with probes detecting EBV-encoded RNA (EBER) is the most commonly applied test for the detection of virus in malignant cells. The sensitivity of this method is related to the high copy number of these transcripts and the ability to detect EBV genome in its latent state (EBER transcription is not dependent on induction of a productive rival life cycle).
Cytomegalovius
Molecular methods have a wide application in the diagnosis of CMV disease and compete with traditional culture and histochemistry-based techniques. DNA PCR can be used for the detection of the CMV genome, but because of its high sensitivity, results may not be informative in seropositive individuals. CMV is strictly cell-associated and blood leukocytes are used as a source for DNA. Quantitative PCR utilizing the light-cycler technology is the usual method for detection of CMV viremia and titer quantitation. Both beacon and Taqman probes have been developed (as described for EBV). Based on standard curves with calibrated positive controls, the number of viral genome copies can be precisely calculated.25 CMV can be detected at levels as low as one and as high as 5 × 105 copies per milliliter, correlating well with the antigenemia measured per 2 × 105 leukocytes.
B19 Parvovirus
Serologic methods are informative only in a minority of circumstances. Similar to CMV, B19 PCR may provide a high positivity rate that does not reflect clinically relevant viremia. Because of the extremely high copy number of virions during active infection, B19 can be detected and quantitated in serum using DNA hybridization methods without amplification. Dot-blot hybridization is most suitable to provide quantitation by serial dilution of positive sera with defined B19 genome copy numbers. Viral titer can be determined by comparison to the dilution standards.21
Other Viruses
In theory, all viruses of known nucleic acid sequence can be detected using PCR. In some clinical circumstances, viral detection may have diagnostic consequences (Table 29.3). For example, herpes virus-6 may be identified in primary effusion lymphomas; presence of adenovirus (type 11) and polyomavirus (BK, JC) DNA may be helpful for diagnosis of hemorrhagic cystitis after bone marrow transplantation.
MOLECULAR DETECTION OF DONOR/RECIPIENT CHIMERISM
Determination of the contribution of host versus recipient blood cell production after allogeneic stem cell transplantation has become a standard laboratory test with clinically relevant implications. Several techniques have been devised, including STR analysis, RFLP or FISH for X or Y chromosomes. Selection of various cell types for analysis allows for the separate determination of chimerism in individual hematopoietic lineages.
Polymerase Chain Reaction–Based Short-Tandem Repeat Analysis
Highly polymorphic microsatellite STRs exist for a large number of human loci. STR sequences show great variability in the number of repeats, their length is inherited, and their pattern may be individually specific.26-29 Examples of such loci include FGA, VWA, TH01, F13A1, and D21S11. A large number of primer pairs allow for amplification of STR from several loci that show biallelic deletion/insertion polymorphisms. Their size patterns enable forensic identity determination and paternity testing.
For bone marrow transplantation, STR for several loci can be amplified in a donor and recipient. Informative loci can be selected for the highest resolution of size differences between the donor and recipient pair. After transplantation, blood is sampled and DNA extracted. Following amplification of informative STR, gel electrophoresis is used to determine the presence of donor and/or recipient bands. When fluorochrome-labeled primers are used, PCR amplicons can be genotyped using capillary gel electrophoresis to precisely resolve the amplified products by size. The areas under the informative peaks are measured and the percentage of donor chimerism is calculated by the division of donor’s peak area by the sum of the areas of the recipient and donor. This calculation can be performed for several informative loci and averaged. The sensitivity of this type of STR analysis permits the detection of 5% of donor/recipient cells for all loci and around 1% for all patients and selected loci. Combining this method with techniques of cell separation allows determination of chimerism within lymphoid or myeloid cells compartments, adding useful information in the setting of nonmyeloablative transplantation.
Short-Tandem-Repeat Analysis by Real-Time (Quantitative) Polymerase Chain Reaction
Real-time PCR STR analysis is a more sensitive and highly quantitative method.30-32 Two primer pairs, each specific for the donor and recipient STR alleles, are selected. STRs suitable for this analysis have biallelic polymorphism, with both alleles varying by at least two consecutive bases and showing a high level of heterozygosity. The sensitivity of this method can be as low as 0.1%, but there is need for a large selection of labeled primers and probes for identification of the most informative loci.
Restriction Fragment Length Polymorphism Analysis
Many loci within the human genome show significant allelic polymorphism, resulting in changes in endonuclease restriction sites. After enzyme digestion, DNA is electrophoresed. The resulting Southern blot is hybridized with tagged DNA probes derived from the polymorphic loci, resulting in the appearance of donor- and recipient-specific bands if there are allelic differences at the locus.
Fluorescence In Situ hybridization Analysis of Sex Chromosomes
Centromeric X and Y chromosome probes can be used for the detection and quantitation of donor and recipient cells. The assay is informative only in sex-mismatched transplants.
MOLECULAR HUMAN LEUKOCYTE ANTIGEN TYPING
Traditional serologic testing is increasingly replaced by molecular testing that allows for a higher precision and resolution of HLA alleles and polymorphisms.33,34 Molecular analysis of HLA loci resulted in the identification of a large number of new alleles, with new polymorphisms still being added. PCR-based methods and primers have been developed for intermediate resolution (IR) and high-resolution (HR) level typing. Serologic testing continues to be performed in many institutions for class I and II alleles, but elsewhere serologic testing has been abandoned for class II alleles.35 Serologic testing retains some role, especially for the confirmation of null alleles after molecular testing.
Polymerase Chain Reaction–Based Human Leukocyte Antigen Testing
Two methods have dominated modern HLA testing technology: SSP and SSOP. Allele- or group-level typing is commonly performed using SSP. Group- and locus-specific primers can be used in the first stage of testing followed by allele-specific primers. Standardized primer sets have been developed for both IR and HR testing.36 However, direct sequencing technologies will likely dominate the future of HLA testing.
SSOP testing can be used for the identification of individual alleles or for the detection of SNPs. In general, SSOP analysis includes differential hybridization of amplicons to specific probes that are designed to match nucleotide sequences at all polymorphic sites of exons 2 and 3. Standardized probe sets have been developed. Several modifications of SSOP hybridization are possible, including membrane- and bead-based fluorometric techniques. For flow cytometric approaches, such as using luminex technology, PCR amplification is performed in the presence of a fluorescently labeled primer pair. The probes are immobilized to polystyrene beads tagged with fluorochromes that can be characterized by a flow cytometric technology using their orange-red emission profile combined with the counterfluorescence of the tagged amplicons. The presence of specific alleles can be identified based on a specific double-fluorescence profile of beads that carry a probe complementary to the amplicon. This assay can be multiplexed to allow a wide screening for many alleles.
SNPs within individual HLA alleles can also be detected using sets of primers designed to investigate polymorphism. This method, also referred to as single nucleotide extension, can also be multiplexed, similar to SSOP.
Finally, for greatest resolution and detection of previously unknown or new polymorphisms, individual region of HLA genes can be PCR amplified and directly sequenced.
References