By Ben Butkus
Canon US Life Sciences researchers have developed a rapid assay based on high-resolution melting analysis to accurately genotype 23 mutations in the cystic fibrosis transmembrane conductance regulator gene responsible for CF and related disorders.
According to the investigators, the study serves as proof of concept for the use of high-resolution melting, or HRM, analysis as a clinical genotyping tool, and highlights the method's advantages over competing techniques such as real-time PCR, multiplex end-point PCR, and next-generation sequencing.
In addition, the work provides some insight into the types of assays that CULS has planned for the microfluidics-based PCR and HRM molecular diagnostics platform it is currently developing and hopes to commercialize by the end of 2013.
Due to the high incidence of CF in non-Hispanic white populations and other ethnic populations, the American College of Obstetricians and Gynecologists and the American College of Medical Genetics have recommended routine carrier and prenatal screening for the disease, and to that end have recommended a panel of 23 mutations in the CFTR gene.
According to the Canon researchers, current commercially available tests can take up to eight hours to perform and require several tube transfers, making them less than ideal for rapid carrier and prenatal screening.
Examples of commercially available molecular CF tests include Quest Diagnostics' CF Complete, based on automated, full-length sequencing of the CF gene; Ambry Genetics' CF 102 Screening panel, based on PCR amplification and Illumina sequencing; and Luminex's xTag Cystic Fibrosis assay, which uses that company's xMAP microsphere technology.
In addition, the CFTR gene is complex and can contain multiple mutations that are very close to one another, as well as benign mutations that differ very little from disease-associated mutations.
According to Ling Xu, a senior scientist at CULS in Rockville, Md., the most promising recent work to develop new molecular CF screening tests has been based on next-generation sequencing, which can discern these myriad mutations but is still not cheap or fast enough for practical clinical implementation — although that is quickly changing.
Nevertheless, Xu and colleagues set out to develop a simpler and faster assay, and chose HRM analysis as the testing method for several reasons.
"After evaluating all 23 mutations we decided to choose HRM, the reason being this method is very simple, fast, accurate, and cost-efficient," Xu told PCR Insider. "Also, the specificities are really high; and it's also really sensitive."
The HRM process involves heating DNA until its two strands separate at a specific temperature based on its composition. Samples from two normal DNA samples yield the same melt curve; however, if one of the samples has a mutation, the melting temperature is different and results in a different melt curve.
Differences in these various melting temperatures can be detected using a fluorescent DNA-binding dye, and if a detection platform is engineered appropriately, it can discern dozens of different melt curve differences in one assay in real time — a much higher level of multiplexing than is possible using real-time PCR with separate fluorescent probes.
"The beauty of this method is that this dye can go in with the PCR master mix … during the PCR reaction … and at the end of 40 cycles we have enough dye to bind to the double-stranded amplicons," Xu said. "Then we just look at the fluorescence signal transition to define whether or not the mutation is there. This method is accurate enough to detect a one-base-pair difference. It’s really sensitive and very specific to the targeted mutation."
Xu presented details of the assay's development and performance in a scientific poster last month at the Association for Molecular Pathology annual meeting in Grapevine, Texas. The group tested 33 DNA samples acquired from the Coriell Institute and one DNA sample from the University of Utah for each of the 23 mutation targets in the ACOG/ACMG guidelines.
They designed small amplicon assays to genotype 18 of the 23 mutations using HRM analysis; and developed an unlabeled probe method to identify the remaining five mutations and benign variants. In addition, as a control, the group tested 16 synthetic plasmid DNA constructs specifically designed for the targeted regions.
The group amplified the DNA samples using the Roche LightCycler 480 system and Idaho Technologies' LC Green Plus dye; and analyzed the melt curves using either in-house software or the MeltWizard software developed by Carl Wittwer, Idaho Tech co-founder and professor of pathology at the University of Utah.
Using their method, Xu and colleagues were able to genotype all 33 genomic DNAs and 16 synthetic DNAs with 100 percent accuracy. In addition, they were able to distinguish between clinically relevant mutations spaced just a few base pairs apart; as well as between clinically relevant mutations and benign mutations.
Differentiating between these different mutation types is particularly important, Xu said, because prior methods have made mistakes in this area, leading to misdiagnosis of CF.
"More and more researchers and clinical labs are starting to adopt HRM analysis," Xu said. "This method is actually still quite young compared to qPCR, but it's gaining a reputation."
In fact, CULS, a wholly owned subsidiary of electronics giant Canon, has in recent years put more R&D into the molecular diagnostics area, and has begun using HRM as its assay development method of choice for most of its molecular diagnostics.
The 23-mutation CTFR panel is a particularly good example of the method's power, Xu said. "I personally believe that for successful [mutation] panel design … good primer design plus a good method will give us a very robust result, which we have shown here," she added.
Although the CULS group used a Roche LightCycler in its study, it has also vetted its method on other HRM-capable platforms – including on its own, a prototype molecular diagnostics instrument that the company has been developing for the past few years using microfluidics IP licensed from Caliper Life Sciences, as well as consumer electronics-inspired detection technology developed at Canon.
Renee Howell, director of R&D at CULS and co-author of the AMP poster, told PCR Insider that the company has a prototype of its platform developed and is at the moment transferring the CFTR and other assays to the instrument.
"We are working on all of these assays … to help us prove concept," Howell said. "They're not necessarily what we would intend to do commercially … [although] it's likely that our first commercial products will likely be geared toward those kinds of uses."
The group is working on another assay for medium-chain acyl-CoA dehydrogenase, or MCAD, deficiency, a genetic metabolism disorder similar to CF in that "you have to screen many exons of a gene to be able to give a good diagnostic," Howell said.
"What we can do with HRM — and in particular with a technique called scanning — is to look at a whole gene, scan exons, and look for heterozygotes," she added. "That's our real focus in using HRM. It's that particular ability of the technique to quickly screen large genetic sequences."
In July, CULS and the University of Maryland launched a research project to develop automated, cartridge-based molecular testing technology for rapid infectious disease diagnosis (PCR Insider, 7/21/11), but that work represents the next generation of the molecular diagnostics platform that Canon plans to commercialize, and as such is much farther down the road.
According to Howell, the next step for the group is to begin submitting some of its data related to HRM assays for peer-reviewed publication; and then continue building out its molecular diagnostics platform for commercialization, ideally by the end of 2013.
"We have some clinical testing data with, for instance, the MCAD [assay], and we're looking at just publishing that as a method," she said. "We did a field trial a year ago with our first version of the prototype. So we've basically spent this first year ramping up and making changes to that first prototype to improve performance. We're about to go through another series of field trials, and after that we anticipate that we'll be ready to build a commercial version of the instrument."
Have topics you'd like to see covered in PCR Insider? Contact the editor at bbutkus [at] genomeweb [.] com.