Send to printer »

Feature Articles : Jun 1, 2011 (Vol. 31, No. 11)

Streamlining Nucleic Acid Sample Prep

Advances in Long-Read Technology, Target Enrichment, and Collection Reshape Operations
  • Josh P. Roberts

As budgets tighten, throughputs increase, and downstream protocols become more exacting, serious efforts are being made to make the most of nucleic acid sample preparation. Whether it’s different devices to gather clinical or potential bioterror samples, innovative methods to enrich and purify samples, better ways to mash them up, or efficient ways to move them about, researchers at Knowledge Foundation’s “Sample Prep” meeting, held recently in San Diego, had much to talk about.

A lot has been made lately of next- (second) generation sequencing, characterized as being faster, cheaper, and more massively parallel than Sanger sequencing. Yet next-gen sequencing generally suffers from relatively short DNA read lengths.

SCODA Technology

FLIR Systems is gearing up for next-next (third) generation sequencing. Many of these newer sequencers will require DNA inputs that are in the tens, if not hundreds, of kilobases in length, said senior laboratory scientist Milena Iacobelli Martinez.

Working under a grant from the Defense Threat Reduction Agency (DTRA) to develop technology for detecting biothreats, FLIR has created an automated sample-preparation device that isolates 20–50 kb DNA. The prototype uses a single disposable cartridge to input up to 1 mL samples, and it lyses spores, vegetative cells, and viruses by a combination of chemical and mechanical means. It’s “really a compromise” between using enough strength to break open the spores and yet not overshear the DNA, she pointed out.

DNA in the lysate then gets concentrated using Boreal Genomics’ SCODA technology that is integrated into the prototype. “Their technology allows high molecular weight DNA to be isolated and low molecular weight DNA to be rejected,” Martinez explained. “In the end you get the concentrated DNA sample in a low volume (about 50 microliters), free of PCR inhibitors that you find in environmental samples and in clinical samples as well.”

Shorter DNA is likely to compete with and perhaps saturate out the longer strands, she said. It’s critical to remove the low molecular weight DNA before it gets to the sequencer, otherwise “you’re really not utilizing the full potential of the long-read technology.”

By switching run parameters on the SCODA, the instrument can also be used to concentrate all the nucleic acids, says Martinez, “so that it can be used for next-gen sequencing as well.”

SCODA—which stands for synchronous coefficient of drag alteration—takes advantage of the fact that nucleic acids undergo a dramatic, nonlinear, physical change under an electric field. “This distinguishes them physically from other molecules: proteins, humic acid in soil, polysaccharides—whatever you might have that could be inhibitors of downstream analysis,” said Andre Marziali, Boreal Genomics’ president and CSO.

Instead of relying on the more traditional solid-phase extraction—binding to a surface, washing, and eluting—SCODA “is taking DNA in a gel and focusing it to a spot at the center.” Meanwhile inhibitors, even those with similar chemical properties, either pass out of the gel or don’t enter it at all.

In his talk, Dr. Marziali, who is also director of engineering physics at the University of British Columbia, focused on a new Boreal development: sequence-specific extraction. By encoding a specific sequence in the gel itself, the instruments can enrich for that target. Single-stranded DNA is run on a gel at the melting temperature of the target probe duplex—the target is slowed down, but does not stick for very long. “What we’re really doing is we’re driving hundreds of thousands of hybridizations that cumulatively cause a focusing force on the target.”

Among other opportunities, Dr. Marziali is looking toward medical diagnostics. A pathogen may be present in a few copies per milliliter of a clinical blood sample containing 100 µg of human DNA in the background, and “we believe that we can enrich for the pathogen DNA, leaving the human DNA behind,” Dr. Marziali said. Similarly, disease markers can be enriched in cancer patients’ blood, as can fetal DNA taken from a maternal sample—situations “where ultimately the inhibitors are not now blood components, they’re really human DNA.

“If we put a pool of DNA into our instrument that has, say, two sequences that differ by a single nucleotide, we can enrich our targeting against a single nucleotide mismatch by about 10,000 fold. In fact, we’re so specific that we can distinguish two sequences that differ by a single methylation site.”

Micro-Size Me

Research into de novo pathogen detection at the Sandia National Laboratory utilizes more conventional molecular biological principles, but a less-than-conventional platform, to enrich nucleic acids.

Unlike PCR- and other probe-based technologies, next-gen (and next-next-gen) sequencing doesn’t require advanced knowledge of the target sequence. You read the entire genetic code and then use phylogenetics and the like to categorize and classify. Yet with human DNA and background flora and fauna DNA in the mix, finding pathogen DNA in matrices such as nasopharyngeal swabs is like looking for “a needle in a haystack,” said senior scientist Kamlesh (Ken) Patel.

Patel’s team uses a two-pronged approach to eliminate much of the background DNA (like housekeeping genes) that take up sequencer bandwidth, he explained. Exome capture beads “pull out all the human things we don’t care about and remove them from the supernatant.”

They also use a normalization process to get rid of high-abundance DNA: DNA is denatured and allowed to re-anneal for a set time. Higher abundance sequences will find their complement faster than the rare fractions—similar to a C0t analysis. The double-stranded DNA is then either digested or separated out from the remaining single strands (among which the rare pathogen DNA would lie) using hydroxyapatite.

All this is done on an automated miniaturized platform. Digital microfluidics manipulate single-digit microliter volume droplets by altering the voltage at each square of the platform, allowing reagents to be brought to the sample, and the sample itself to move to different squares. The device is connected with various lab-on-a-chip modules, such as a capillary tube-based heated microreactor, by means of a syringe pump.

The platform manages reactions at a scale and concentration that can directly interface with an Illumina flow cell. “We’re technically sequencer-agnostic,” Patel said. The firm has focused on Illumina first, he explained, because of its familiarity and also its collaborators.

Because the interfaces are flexible, “we can change pretty quickly if we need to do a different type of protocol.”

Respiratory Diseases

Sometimes gathering nucleic acids for identification of pathogens is done in more mundane ways. Current practice for tuberculosis and other lung ailment diagnostics requires sputum, which is difficult to collect and often contaminated by saliva. A nebulizer may be needed to aid patients unable to produce sputum normally, while in some cases a bronchoalveolar lavage is required.

Deton engineered a small device to collect intact bacterial aerosols that are coming out of a cough. “The sample can be extracted from the device and sent to the lab, just the way it’s being sent right now,” said principal Patrick Sislian. “But instead of sputum it would be in a buffered solution: more concentrated than sputum, and without having those negatives that the sputum samples would.” The instrument looks similar to CPAP masks used for sleep apnea, connected to a palm-sized box with a liquid-filled chamber where the samples are collected. It is designed to be extremely simple to use, completely disposable, and powered by a lab vacuum. With a projected price point of about $250—less than that of a nebulizer treatment—Sislian thinks the Deton device “would have the economic and the practical advantage” over current practice.

He expects the prototype to undergo testing on a few patients within several months, with a larger patient study further down the road.

As part of his chemical engineering doctoral training at UCLA, Sislian studied the use of aerodynamic shock to lyse both bacterial spores and vegetative cells, and created a prototype device that he ultimately hopes to integrate into the cough analyzer. “We’re also looking for people to partner with on the biosensor side,” to integrate nucleic-acid testing into the device as well.

Mash It Up

Bertin Technologies’ way of breaking down cells is “not really innovative,” confessed Quitterie Brossard-Desjonqueres, the company’s lab equipment marketing manager. “The shock between tubes, sample, and beads will grind the sample. It’s quite a simple technology. There are a lot of other systems working with bead beating.”

What’s different about Bertin’s Precellys line of biological sample grinders, she said, is the specific movement of the tubes. In addition to moving up and down they also move on the lateral aspect, making the beads form a figure 8 inside the tube, which allows for more efficient grinding.

To homogenize tissue, spores, noncultivatable microorganisms, and even viruses, samples are placed in a tube with buffer and beads. Different samples demand different sizes and different material beads—glass, ceramic, or metallic. The 2,000 or so protocols on the Bertin website will also indicate the optimal speed and duration. “If you go too strong or too long, your sample will be damaged and your results will be bad,” Brossard-Desjonqueres noted.

Mortar and pestle are still used in many R&D labs that don’t have a lot of samples, or that frequently process different samples. “The problem is that it’s not a standardized method, it’s always different from one operator to another one.”

Earlier this year Bertin launched Minilys, a smaller bead beater-homogenizer, based on the same technology. Minilys can accommodate either three small (0.5 or 2 mL) tubes, or a single 7 mL tube.

Other models, designed for higher throughput, can handle up to 24 tubes simultaneously. These are available as stand-alone units—which give R&D-type labs flexibility in terms of sample type and size—or integrated into a global solution.