The Ingenuity Behind Impactful Assays

Developing an assay that measures DNA methylation, an independent predictor of leukemia

An interview with Gerald BW Wertheim

Gerald B.W. Wertheim, MD, PhD
Assistant Professor,
Children’s Hospital of Philadelphia


Acute myelogenous leukemia (AML) is generally a disease of older people, although it is also the second most common blood cancer in children. In AML and other leukemias, alterations in DNA methylation are well recognized, and the aggressiveness of AML tumors can be determined by studying methylation. Mutations in the genes that regulate DNA methylation have been shown to cause different forms of leukemia. Studies have also demonstrated that the prognosis of patients with AML can be predicted by this altered DNA methylation pattern with as few as 17 loci. However, tests that directly measure multiple-locus DNA methylation are typically expensive and technically challenging, making them difficult to perform in a routine setting.

Gerald Wertheim and his colleagues have developed a novel approach to simultaneously analyze DNA methylation patterns at the 17 target loci by integrating multiplexed branched DNA technology from Thermo Fisher Scientific and fluorescent microsphere technology from Luminex. The method uses techniques that are inexpensive and can be easily performed in a routine setting. The technique is called expedited microsphere HpaII tiny fragment enrichment by ligation-mediated PCR, or xMELP. It accurately reflects the methylation levels at each analyzed locus and enables segregation of individuals with acute myeloid leukemia into prognostic subgroups. Dr. Wertheim and his colleagues have published papers in The Journal of Molecular Diagnostics and the journal Clinical Chemistry (1,2).

Tell us about your studies of DNA methylation.

My primary interest has been on developing methods to better diagnose leukemia and lymphoma and to assess their prognosis in patients. Ultimately, we hope that clinicians can use our findings to help guide their therapy choices. We have known for some time that mutations are present in leukemia tumor cells, but we have not had the information to risk-stratify patients optimally. So I have been focusing on DNA methylation, which is thought to control transcription in certain numbers of genes. There are a number of pieces of evidence demonstrating that differential DNA methylation is important for AML prognosis. Many researchers have shown that genes that regulate DNA methylation are mutated in different types of leukemias, and some have directly shown that looking at DNA methylation alone can predict the outcome in patients with AML. Our goals are to find a set of genetic loci that vary in their DNA methylation patterns among patients, which can predict who will do well and who will do poorly with therapy, and to then develop a clinically relevant assay to detect levels of methylation at these loci.

How did you develop your analysis method?

With an ultimate goal of taking our assay to the clinic, we wanted to find a technique that was relatively inexpensive and could be multiplexed. We wanted to be able to look at multiple methylated regions to improve robustness, and we needed an assay that was highly reproducible. Because we use microsphere technology in the clinical lab, I initially tried a microsphere assay without branched DNA, and it did not have the appropriate analytical sensitivity. One of my collaborators was aware of the Invitrogen QuantiGene Plex Assay from Thermo Fisher Scientific and suggested that we look at it for our work. The QuantiGene Plex Assay uses branched DNA (bDNA) to amplify the signal rather than amplifying the target, and it sufficiently improved our sensitivity. Having the branched DNA amplification signal has been critical. We found that this QuantiGene Plex Assay was remarkably quantitative, with the results obtained by the QuantiGene Plex Assay and qPCR being virtually identical. Thus, this looked like the robust, multiplex assay program that could ultimately be used in high-throughput clinical lab settings.

Evaluation of methylation is done through the measurement of microspheres and does not require custom-made, solid-phase oligonucleotide microarrays or high-throughput sequencing. The examination of methylation levels is performed by flow cytometric analysis of fluorescent microspheres, alleviating the need for high-throughput sequencing. This assay format can be easily expanded for the evaluation of 80 loci without an increase in reactions; the simultaneous assessment offers advantages, including a streamlined workflow and relatively fast turnaround time.

The technical staff from Thermo Fisher taught us how to set up the instruments and run the assay. They were readily available by phone to help when I had questions. We were also able to have Thermo Fisher run some of our samples so we could compare runs and rapidly troubleshoot problems.

I think that going forward, the diagnostic field is going to have to look at some of these other aspects, such as RNA expression analysis or epigenetic events. Hopefully, our assay will give clinicians a more accurate indication of prognosis and will help guide their decisions.


  1. Wertheim GB et al. (2015) Validation of DNA methylation to predict outcome in acute myeloid leukemia by use of xMELP. Clinical Chemistry 61(1):249–258. 
  2. Wertheim GB et al. (2014) Microsphere-based multiplex analysis of DNA methylation in acute myeloid leukemia. Journal of Molecular Diagnostics 16(2):207–215. 

High sensitivity immunoassay innovation

Many tools have been developed to advance our understanding of cellular mechanisms. However, a simple, scalable, and affordable method for low-level quantitation in limited serum samples has seemed out of reach. Our latest innovation—Invitrogen ProQuantum immunoassays—is a set of ready-to-use kits for quantifying low-abundance proteins in small sample volumes. The assays utilize matched antibody pairs for specificity and Applied Biosystems TaqMan qPCR technology for amplification and detection. Cloud-enabled ProQuantum software is available for the quantitative analysis of protein concentrations interpolated from a standard curve.

Powered byThermo Fisher Cloud