Send to printer »

Feature Articles : Mar 1, 2012 (Vol. 32, No. 5)

NGS Advances Spawn Novel Challenges

  • Greg Crowther, Ph.D.

One measure of the dramatic progress in next-generation sequencing is that scientists no longer want to talk about advances in instrumentation. They would prefer to talk about other issues, including sample prep and analysis.These timely topics and more will be discussed at the upcoming “Genomic Research” conference, organized by Select Biosciences.

“The sequencing instrument companies have done a very good job speeding up the sequencing machines,” says John Langmore, Ph.D., chief scientific officer of Rubicon Genomics. “And now it’s up to everyone to speed up the sample preparation, speed up the analysis, and also focus the sample preparations on specific clinical samples that are of utility, and focus the informatics on getting the right information from the samples out quickly.”

Scott Diehl, Ph.D., agrees that the collection of raw sequence data is no longer the rate-limiting step of NGS research. Dr. Diehl—director of the Center for Pharmacogenomics and Complex Disease Research at the University of Medicine & Dentistry of New Jersey—believes that the biggest upcoming challenges of clinical research will be (1) gaining the cooperation of the huge numbers of test subjects necessary to understand complex multifactorial diseases and (2) balancing these subjects’ right to privacy with the need to analyze their genomes and medical histories in great detail.

“It’s really a cultural challenge as much as a technological challenge in order to gain the fruits of these amazing advanced technologies,” he says.

Dr. Diehl points to the decades-old needle-in-a-haystack problem of finding single disease-causing mutations amidst an entire genome. Beginning in the 1980s, researchers successfully uncovered the mutations responsible for such disorders as cystic fibrosis, muscular dystrophy, neurofibromatosis, and Huntington disease.

“We said, ‘OK, this works great for simple single-gene diseases; let’s apply it to more complex conditions,’” Dr. Diehl says.

“And to make a long story short, initially using family-based linkage studies and then later using population-based case-control studies of various diseases, we found lots of genes, but...even adding them all up, they only explain a relatively modest portion of the inherited component of the diseases. And so this leads to the conclusion that there must be hundreds of needles we need to find to fully explain the causes of complex diseases.”

To understand how dozens or hundreds of genes interact with each other and the environment in conditions such as diabetes, cancer, heart disease, and mental illness, Dr. Diehl says, “We ultimately need to build up registries of millions of people who are tracked over decades in order to tackle the biostatistical analysis.” The recent genotyping study of 100,000 Kaiser Permanente members is a step in this direction, he says.

Engaging the public on such a large scale will require a massive education and outreach campaign, including frank disclosure of possible threats to privacy. “We need to build safety and privacy protection mechanisms, but sooner or later major genetic databases are likely to get hacked and put out onto the web,” Dr. Diehl concedes.

While Dr. Diehl hopes to analyze large populations of subjects, Marek Malecki, M.D., Ph.D., associate professor at Western University of Health Sciences, is working on populations of cells within an individual—cancer cells, in particular.

A basic problem with cancer is that different tumor cells develop different mutations, leading to distinct subpopulations within the tumor. “Tumor growth is propelled by many different clones with different properties,” Dr. Malecki notes, “and when you deal with a heterogeneous tumor, then basically it’s very hard to apply a one-size-fits-all therapy or diagnosis.”

But what if you could capture, sequence, and treat each specific clonal subpopulation? That is one of Dr. Malecki’s long-term goals. To sort out different populations of normal cells and tumor cells from clinical samples, he and his team have created a panel of BioTags and OncoTags—synthetic antibodies that recognize specific structural variants of cell-surface proteins (e.g., EGFR) often mutated in cancer cells.

Dr. Malecki’s vision is that once the tumor cells are sorted into different clonal populations, they could all be sequenced via NGS and subjected to test treatments. Treatments that appear effective against isolated cancer cell populations could then be administered to the patient as targeted personalized medicine.

One advantage of an OncoTag-based approach, Dr. Malecki says, is that it preserves cells’ viability better than fluorescence-activated cell sorting (FACS) and other existing methods, thus facilitating clonal expansion for profiling of the genome, microRNA, and transcriptome, as well as for testing of targeted therapies. Regardless, the sequencing of the many subpopulations of a tumor is only feasible because of high-throughput NGS.

Improving Sample Prep

Once samples are collected—whether from a clinic or a basic research lab—they cannot be loaded directly into a sequencing machine, of course. One issue is that the DNA of interest needs to be relatively free of irrelevant DNA and other impurities. A second issue is that the samples must be rendered ready for a specific sequencing machine.

“The instrument companies utilize varied adapters and different ligation strategies, making it difficult for users to prepare a one size fits all sample,” explains Masoud Toloue, Ph.D., director of scientific research at Bioo Scientific. “Unfortunately, I don’t see that changing in the future.”

Rubicon Genomics currently offers kits to prepare clinical samples for sequencing via the Illumina platform and is working toward compatibility with the Life Technologies platform.

For a review of a paper published about Life Technologies' Ion Torrent sequencing technology, click here.

Dr. Langmore notes that sample-preparation procedures acceptable for research samples are usually inadequate for clinical samples.

Researchers starting with microgram amounts of homogeneous DNA can afford to lose 99% of it during preparation, but that is not an option for people aiming to sequence sub-nanogram amounts of fetal DNA or tumor DNA mixed with nanogram amounts of maternal or nontumor DNA.

Furthermore, short segments of free-floating DNA found in plasma samples are best handled differently than formalin-fixed cells prepared for pathology analysis, in which the DNA is preserved in much larger pieces but is damaged in the fixing process.

Rubicon’s ThruPLEX-FD kit is designed to retain >90% of target DNA from plasma and formalin samples in a one-tube, two-hour, three-step process compatible with the high-throughput format of Illumina sequencers. Dr. Langmore believes that forthcoming kits targeting individual sample types—for example, a plasma-specific kit—can further simplify the process to two steps and less than one hour.

At Bioo Scientific, Dr. Toloue is tackling the related challenge of developing genome-scale methods for profiling DNA methylation. 5´ methylation of cytosine bases is a general mechanism for suppressing gene expression and is thus of great interest to both basic researchers and clinicians.

However, among other problems, the DNA polymerases commonly used in amplification and sequencing reactions do not recognize uracils introduced via bisulfite conversion of unmethylated cytosines, which allows these cytosines to be distinguished from methylated ones.

Dr. Toloue will discuss advances in methods to capture whole-genome methylation data and will compare the methylomes of differentiated and undifferentiated cells. “At this stage we are examining research samples,” he says. “The results of this study, however, expand into development of induced pluripotent stem cells as clinical replacements to human embryonic stem cells.”

Translating Data into Clinical Advice

Many researchers consider valid and affordable clinical applications of whole-genome analysis to be at least a couple of years away. Existence Genetics, however, already offers assessment of clients’ genetic predisposition toward hundreds of diseases.

While whole-genome sequencing is not yet a cost-effective way of gathering clinically relevant data, Existence’s Nexus DNA Chip tests genetic loci linked to hundreds of diseases. Brandon Colby, M.D., the firm’s founder and CEO, describes it as “an aggregation of all known loci associated with a phenotype—an increased or decreased risk of disease—as discovered by third-party research studies over the last three decades.”

The current version of the Nexus chip tests about 10,000 loci, though this number will increase as additional disease-linked genetic variants are discovered.

In addition to partnering with medical clinics, Existence offers its services through outlets such as the Equinox Fitness chain of health clubs, which offer to estimate clients’ relative risk of developing such conditions as arthritis and osteoporosis.

“If somebody has a very high risk of knee arthritis, then trainers tailor a program, such as cardiovascular exercise, that has low impact on the knees,” Dr. Colby explains. “Instead of running on the treadmill, they’re focusing on the elliptical, or on the bike, or on swimming.”

Similarly, women whose results indicate a predisposition toward osteoporosis are given extra motivation to strengthen their bones via resistance training.

For his part, Dr. Diehl worries that the promise of NGS advances will be distorted by the “hype cycle” of media organizations and profit-driven companies. Yet he remains optimistic that the public will ultimately embrace the ambitious long-term studies needed to shed further light on the genetic basis of complex diseases.

He cites a recent conversation with Danish researchers who have had good success recruiting subjects for pain studies.

“Who wants to volunteer for a study of pain?” Dr. Diehl wonders. “Good grief! But it sounds like the culture in Denmark is that people feel a responsibility that, if I and my family are going to benefit from biomedical research, I should be responsible for participating in this research. It’s not fair for me to just sit on the sidelines…and then gain the benefits of the new medicines.”