March 1, 2015 (Vol. 35, No. 5)
Neil McKenna, Ph.D.
Making Molecular Analysis Count in Diagnostics, Drug Selection, and Monitoring
The revolution in nucleic acid detection and quantification heralded by the invention of polymerase chain reaction (PCR) in the late 20th century greatly advanced the field of molecular diagnostics. Classical PCR is used to qualitatively or quantitatively interrogate a target molecule in a given sample in the context of all other molecules in the sample, with the final signal averaged over the total sample volume.
In digital PCR (dPCR), by contrast, the sample is divided into a large number of distinct, small volume reactions in which single molecules are amplified in an all-or-none manner. By counting the total number of target molecules in a digital manner, dPCR theoretically offers greatly increased sensitivity relative to, for example, quantitative PCR, as well as greater applicability when sample sizes are limited.
dPCR applications were recently showcased at two CHI events, the Digital PCR Conference in La Jolla and the Molecular Medicine Tri-Conference in San Francisco. Both events served to highlight the growing importance of this platform in clinical diagnostics.
“About 60% of all genes are alternatively spliced,” said Dianna Maar, Ph.D., senior scientist, applications, Digital Biology Center, Bio-Rad Laboratories. Alternative splicing refers to the process whereby nascent pre-mRNA transcripts are edited into a variety of downstream species, each of which can be translated into a protein, often with functional properties distinct from one another.
“The T-cell receptor CD45 exists in a number of different isoforms that are characteristic of different types of T cell,” Dr. Maar added, pointing out that the long form of CD45 is predominantly expressed in naïve T cells, whereas the short form is primarily found in activated T cells.
A substantial body of evidence indicates that aberrant splicing events, arising either from mutations in splicing targets or in RNA processing factors, contribute to the etiology of a variety of diseases.
“One of our areas of interests is identifying alternative splice forms that are associated with a specific disease, and which can provide diagnostic or prognostic information in the treatment of that disease,” Dr. Maar pointed out. Bio-Rad’s Droplet Digital PCR (ddPCR) platform, said Dr. Maar, provides for increased quantification and isoform discrimination over gel-based methods and quantitative PCR.
Duchenne’s muscular dystrophy (DMD), a congenital disorder characterized by progressive weakness and wasting of skeletal muscles, has been linked to aberrant splicing events in DMD, the gene encoding dystrophin. These events result in complete absence of the mature protein. “Antisense oligonucleotide (AON)-based therapies for DMD are designed to induce skipping of specific DMD exons during pre-mRNA splicing that will partially restore function by splicing into an exon with the correct reading frame,” explained Dr. Maar.
Dr. Maar pointed out that the high sensitivity of ddPCR can be leveraged to monitor the relative levels of incorrect and correctly spliced transcripts to reflect the progress of AON therapy in DMD.
Cell-free DNA (cfDNA) analysis refers to the quantification of DNA molecules outside their cells of origin, typically in circulating blood or other body fluids. “Most cfDNA analysis is currently being performed as part of clinical research studies,” said Alison Devonshire, Ph.D., science leader, nucleic acid metrology, LGC. “Examples include colorectal cancer screening, noninvasive liquid biopsy of tumors to help refine therapeutic options, monitoring organ transplant rejection, and prenatal diagnostics,” she added.
Dr. Devonshire pointed out that by increasing the ratio of the target (e.g., tumor) DNA to wild-type (e.g., nontumor) DNA in each partition, dPCR allows for sensitive but potentially informative genetic measurements to be detected. Citing an important advantage of dPCR over analog real-time PCR, Dr. Devonshire observed “dPCR allows reproducible absolute quantification of target DNA copy number without the need for a calibration curve, which can be an additional source of measurement bias.”
She cautioned that standardization and the development of common reference points between laboratories are critical parameters if cfDNA-based diagnostics are to realize wider routine application in the clinic.
“Given its high precision and sensitivity, dPCR is potentially suited to quantifying reference materials and samples for quality control and proficiency testing schemes,” Dr. Devonshire noted. “Reference materials can be used to assign values to in-house controls so that tests using various calibration materials can be traced to a common reference point.”
Such reference materials include ERM® AD-623, a plasmid that contains the BCR-ABL1 fusion gene transcript. In research facilitated by LGC and other expert laboratories, the ERM AD-623 reference material was tracked by means of dPCR to monitor disease load in chronic myeloid leukemia.
“The P450 class of genes encodes one of the most important groups of genes involved in the metabolism of xenobiotics,” said Iain Russell, Ph.D., senior product manager, genetic analysis, Thermo Fisher Scientific. “Approximately 85% of clinically relevant drugs are metabolized through the cytochrome P450 oxidase system.”
Across the cytochrome P450 gene family a broad collection of polymorphisms, such as single nucleotide polymorphisms (SNPs), insertions or deletions (InDels), gene copy number variations (CNVs), and gene conversion events, have been shown to impact pathway function and, as a consequence, the rate at which drugs are metabolized.
“Individual responses to drug treatment appear to be linked to individual differences in the functionality of the cytochrome P450 system,” Dr. Russell noted. “Accordingly, it is becoming commonplace to use genotype information to determine an individual’s metabolizing phenotype prior to administering certain classes of drugs (e.g., codeine) so as to reduce or avoid an adverse drug reaction.”
Dr. Russell cited the example of the CYP2D6 gene, which encodes an enzyme involved in the metabolism of approximately 20% of clinically used drugs, and in which more than 80 mutations in both coding and regulatory regions have been identified.
“Based upon genotype information, CYP2D6 alleles are classified based upon the presence of a fully functional gene, partially functional gene, or nonfunctional gene,” Dr. Russell continued. “Together, it is the inherited combination of these alleles that determines an individual’s drug metabolizer phenotype.”
Dr. Russell observed that although current CNV analysis tools and approaches enable the detection of gene duplication or loss events, they are unable to elucidate the particular allele that was duplicated. In situations where an individual carries fully functional and partially function alleles at a particular cytochrome P450 locus, accurately discerning these alleles is critical for assigning the individual’s appropriate drug metabolizer phenotype.
“By allowing us to partition the standard PCR reaction mix into thousands of independent reactions, each counting a specific allele of interest, dPCR delivers highly quantitative and precise information about the numbers of each individual allele,” Dr. Russell explained.
Biorepository Tissue Samples
The Rutgers University Cell and DNA Repository (RUCDR) is a biorepository program led by Andrew Brooks, Ph.D., associate professor, department of genetics, Rutgers University. “RUCDR supports clinical trials management currently for five NIH institutes, facilitating sharing and distribution of tissue, nucleic acid, and study results,” said Dr. Brooks. “RUCDR currently holds nearly 12 million nucleic acid samples and nearly 8 million cell lines.”
Dr. Brooks pointed out that given the finite amount of sample available, and the multiple downstream analytical end points that might exist in different laboratories, it is important to develop assays that use the smallest amount of repository tissue possible. He also stressed the importance of increased sensitivity in quality control.
“Single molecule detection is becoming increasingly important in our quality control workflows,” Dr. Brooks added. “[It helps] ensure that sample quality is appropriate for what investigators will use them for, such as deep sequencing.
Dr. Brooks described the use of the RainDrop dPCR system (RainDance Technologies) for target validation. “Digital quantitation of target molecules using dPCR enables us to very sensitively determine the absolute concentrations of those molecules in the sample,” he remarked.
Dr. Brooks highlighted the incorporation of dPCR in the analysis of fresh-frozen and formalin-fixed, paraffin-embedded colorectal cancer samples with respect to validation of mutations in the BRAF, KRAS, and TP53 genes. “dPCR enabled us to integrate sample quality control, targeted molecular profiling, and target validation,” he asserted, [all] in a single platform with nanogram quantities of starting material.”