The scope of real-time quantitative (qPCR) assays for gene-expression analysis in research applications and for sequence-specific nucleic acid detection for diagnostic purposes continues to grow. Driving this growth are increasingly robust, automated, and high-throughput technologies and multiplexed analytical strategies.
At Intelligent Enterprise Solutions’ “qPCR Symposium” held recently in Millbrae, CA, presenters explored advances providing improved qPCR throughput and quantitative resolution, discussed the challenges in preanalytical sample processing and the selection of a qPCR assay with high specificity, and described emerging application areas such as microRNA (miRNA) profiling in biological samples.
Melissa Kelley, Ph.D., R&D research scientist at Thermo Fisher Scientific, discussed the challenges in distinguishing between closely related members of a gene family when using qPCR in an interference RNA (RNAi) study to assess the phenotypic effects of gene knockdown.
As an example, she reviewed the use of Thermo Scientific’s Solaris qPCR gene-expression assays to study the regulation of a FOXO transcription factor (FOXO1), which is regulated via phosphorylation by AKT enzymes, a family of serine/threonine protein kinases. The AKT family comprises at least three highly homologous members that share 70–80% sequence identity. Silencing of AKT activity results in FOXO dephosphorylation, translocation of the transcription factor to the nucleus, and up-regulated gene expression. AKT dysfunction has been associated with various disease states, including cancer and diabetes.
In the experiment described by Dr. Kelley, small interfering RNAs (siRNA) were designed to knockdown the expression of each individual AKT family member. qPCR was then used to measure target gene knockdown, followed by high-content analysis to assess the biological phenotype. Dr. Kelley focused on the question of how to select a qPCR assay when the gene of interest has multiple splice variants, as do two members of the AKT family.
Solaris qPCR assays include probe/primer sets designed using a tier-based algorithm. Each assay detects all known splice variants of a gene and allows the researcher to discriminate between closely related members of a gene family. The assays are designed to target a consensus sequence created from a region common to all of the transcript splice variants. The assay kit provides the researcher with the probe and primer sequences to facilitate compliance with evolving MIQE guidelines.
To increase the sequence design space and to raise the melting temperature of bound complexes to allow for use of shorter, more specific probes, the design algorithm incorporates minor groove binder and Superbase technologies from Epoch Biosciences, now part of Nanogen. This strategy allows for greater than 97% coverage of both the human and mouse genomes. It also “minimizes amplification of contaminating genomic DNA and eliminates off targeting,” said Dr. Kelley. She presented relative gene-expression data generated using the Solaris qPCR assay demonstrating specific siRNA-mediated knockdown of a targeted AKT enzyme and expression levels of the other two AKT family members comparable to those measured in an untreated, control sample.
Analysis of the relative expression of AKT gene family members following RNAi-mediated knockdown, combined with biological assays to evaluate the cellular localization of the FOXO1 transcription factor, enabled the determination of AKT isoform-specific regulation of FOXO1 and illustrated the phenotypic effects of inhibiting the AKT pathway.
The experiment showed that silencing of individual AKT enzymes did not result in a significant change in the localization of FOXO1, supporting the theory that there is redundancy built into the AKT signaling pathway. In contrast, knockdown of combinations of two AKT family members, and simultaneous downregulation of all three enzymes led to increasing FOXO1 translocation from the cytoplasm to the nucleus.
Resolution and Throughput
Whether qPCR can achieve better than twofold discrimination and whether greater quantitative resolution is feasible in a cost-, time-, and resource-efficient manner were the questions put forth by Ken Livak, senior scientific fellow at Fluidigm. He described a study using qPCR-based copy number variation (CNV) determination as a paradigm for analyzing the parameters of quantitative resolution. The study was designed to determine how many experimental replicates would be needed to be able to distinguish between one, two, three, four, or five copies of an X-linked gene. The ability to differentiate four copies from five copies would correspond to a 25% difference, or 1.25 discrimination.
Livak performed the study using Fluidigm’s 96.96 Dynamic Array Integrated Fluidic Circuit (IFC), a microfluidic-based biochip that utilizes a matrix design to carry out 9,216 real-time qPCR experiments in parallel. The 96.96 Dynamic Array contains inlets for up to 96 different samples and 96 different assays, which are combined pair-wise.
In the CNV experiment, Livak evaluated five different samples—each contained one, two, three, four, or five copies of the X chromosome—and each sample was pipetted into 19 sample inlets. Each qPCR assay mixture was pipetted into 24 assay inlets, resulting in 19 x 24 or 456 replicates for each sample. A diploid gene—present in two copies/cell—was used as a reference calibrator.
Fluidigm’s other qPCR platform, the Digital Array IFC, yields a digital output and can perform endpoint PCR. A sample/reagent mixture introduced onto a chip is divided into hundreds or thousands of different chambers, with an identical volume in each chamber. Following PCR, a positive result will indicate a chamber containing one or more copies of the gene of interest, whereas a negative chamber will have zero copies of the gene. The system determines the number of target molecules per microliter of sample by counting the number of positive chambers, applying a Poisson distribution correction, and dividing the resulting number by the total input volume.
The company recently introduced the 48.770 Digital Array IFC, which can partition up to 48 different sample/reagent mixtures into 770 chambers per reaction. Livak described a CNV experiment performed using endpoint qPCR, in which a sample containing four copies of a target gene was compared to a sample with five copies of the target gene; the samples also contained a reference gene. Each sample was loaded into 24 panels of a 48.770 Digital Array and dispersed into 770 chambers.
The results demonstrated that 1.25-fold discrimination could be achieved using real-time PCR on the Dynamic Array with 18 replicates, while 86 replicates could distinguish 1.1-fold differences. In comparison, with digital PCR, a 1.25-fold difference could be discriminated using approximately 1,200 chambers, with 8,000 chambers needed to achieve a 1.1-fold discrimination.
Whereas, real-time qPCR offers advantages in throughput, digital PCR is easier to perform, with less chance for error, and it can be run as an endpoint assay, concluded Livak. “Quantitative resolution of qPCR can be enhanced to 1.25-fold, or even 1.1-fold, by running a large number of reactions,” which is feasible and practical on a high-throughput microfluidic platform.
Fluidigm is exploring the use of the Digital Array IFC for noninvasive prenatal analysis of fetal DNA. The goal is to develop a qPCR system that can reliably and cost-effectively amplify the small amounts of fetal DNA present in a mother’s blood, accurately quantify gene copy number, and distinguish, for example, the three copies of chromosome 21 that are present in a sample from a fetus with trisomy 21 (Down syndrome) from the two copies of chromosome 21 in the cells of a fetus without the genetic disorder.
High throughput real-time PCR is moving in two main directions, noted Matthias Hinzpeter, Ph.D., head of program management qPCR and NAP systems at Roche Applied Science: a high-density microplate-based array format or a microfluidics-based chip format. Roche’s newest LightCycler® PCR system, the LightCycler 1536, is a microplate-based qPCR platform that processes 1,536 reactions in less than 50 minutes. Reaction volumes range from 0.5 to 2.0 microliters.
“The major advantage of this system is the high flexibility with regard to number of targets and number of samples,” said Dr. Hinzpeter. “The technology breakthrough was the development of a 1,536-well microplate with rapid and homogeneous temperature distribution across the 1,536 wells with small reaction volumes.”
Roche’s Universal ProbeLibrary qPCR assays are based on 165 prevalidated probes suitable for studies of samples from a variety of organisms. The probes incorporate locked nucleic acid (LNA) technology for increased specificity and sensitivity and to allow for higher melting temperatures. Universal ProbeLibrary Reference Gene Assays allow for multiplex assays to quantify expression levels of human, mouse, or rat genes.