July 1, 2007 (Vol. 27, No. 13)

Elizabeth Lipp

Attention to Detail and Quality Control Are Essential

With drug pipelines needing a jump-start, many scientists are turning to long-forgotten samples in cold storage. At Cambridge Healthtech’s recent “7th Annual G.O.T. Summit” in Boston, the spotlight was on whole genome amplification as presenters examined some of the challenges in genomic sample preparation. Topics covered included whole genome amplification and the impact of tissue fixation time on array comparative genomic hybridization (CGH) profiles; novel methods of genomic sample preparation; gene-expression profiling, and targeted nucleic acid sample-preparation methods for expression analysis.

Michael Brownstein, director of functional genomics at the J. Craig Venter Institute, posited that investigators who are new to the field of functional genomics are put off by the apparent difficulties inherent in labeling probes, hybridizing and washing arrays, and analyzing results. “This is perfectly appropriate, but assuming that preparing nucleic acid templates for one’s studies is trivial, would be incorrect. Attention to detail and quality control are essential, and different applications demand different extraction methods. Template purificaton is definitely not sexy, but it’s very important,” said Brownstein.

“Access to a virtually unlimited amount of tissue and being able to process it yourself is the best case scenario,” Brownstein said. “Often one has to deal with small samples and rely on collaborators to collect and process them. This creates problems.”

“Once upon a time, there weren’t kits,” Dr. Brownstein noted. Scientists had to decide for themselves how best to solve problems like template extraction. “Now you can order solutions to your problems online, but it is worth giving some thought to the strengths and limitations of the reagents available,” Brownstein said. “Spend as much time with your collaborators as you can to make sure you all understand what you want to discover and how best to accomplish your goals.”

Sample Preparation

Todd Peterson, vp cloning and protein expression R&D, Invitrogen (www.invitrogen.com), discussed sample preparation, and Invitrogen’s broad product portfolio that touches on gene-expression analysis. “Sample prep is very important. It doesn’t matter how good the labeling and detection is, it’s all about input sample quality, and streamlining and standardizing methods as much as possible.”

Peterson said that isolation of intact nucleic acids with high purity and yield are critical steps for gene-expression analysis. Due to the complexity of the various sample types and RNA populations used in current research, targeted methods have been developed to assure optimal input and output for any given expression system.

“Applications where abundant, uninformative transcripts, such as ribosomal RNAs in total RNA samples or known abundant transcripts such as globin mRNAs in blood, benefit from enhanced purity to prevent downstream inhibition and interference for targeted gene-expression analysis. Examples include whole-genome transcript analysis and size fractionation during preparation to improve miRNA analysis,” Peterson said.

Archived samples used for retrospective clinical studies, routinely preserved in an FFPE format, can be troublesome in processing efficient yields and obtaining intact nucleic acids. Several targeted commercial methods for the purification of high-quality enriched RNA, miRNA, and total RNA from FFPE samples were described. At the conference data showing targeted nucleic acid sample preparation for specific and improved gene-expression analysis across various downstream applications including microarrays, qRT-PCR, miRNA qRT-PCR, aCGH, and cDNA library construction, were discussed.

“Many sample preparation technologies for gene-expression analysis fit nicely in the overall workflow for FFPE,” said Peterson. “All these sample-preparation products are optimized in integrated kits and workflows for gene expression and genomic profiling with labeling and detection for a wide range of sample types and for low- to high-throughput applications.

“There is no standard for fixation protocols, tissues themselves are highly variable, and sample quality generally declines with the age of the fixed tissue. So compared to fresh tissue, FFPE is never going to be as good, but these samples can provide valuable, useful differential expression data with optimization and proper experimental controls.”

“Access to biologically and clinically relevant samples for the discovery, refinement, and validation of gene-expression signatures reflecting clinical phenotypes has been hindered by the inability to generate high-quality gene expression results from the most readily available tissue sources, FFPE,” said Gianfranco DeFeo, senior director of customer solutions, NuGEN (www.nugenic.com). “We have developed a linear, isothermal, and robust amplification approach to perform whole-transcript amplification from small and degraded RNA samples.”

This technology will not only allow researchers to perform splice-variant analysis but will also allow for the use of degraded RNA samples, such as RNA isolated from FFPE sources, in expression experiments, DeFeo explained. Data demonstrating the performance of the technology, including sensitivity, linearity, dynamic range, and differential expression accuracy will be shown. In addition, data demonstrating the performance of the approach to clinically relevant samples such as RNA isolated from FFPE tissue sources will also be shown.

According to De Feo, there are a large number of stored samples with unique information associated with them. “Most companies have a goldmine of information in their biobanks that can be used for biomarker discovery and validation work that could previously only be done by performing costly and time-consuming prospective clinical studies. Researchers have samples from patients with known clinical outcomes making them ideal for this type of gene-expression based biomarker research. Companies need to deepen their drug pipelines, and the most efficient and effective means of doing that is making the most of the samples they’ve already gotten.”

Therein lies the challenge. A lot of these samples have not been treated so well and have been stored for going on a couple decades, now, DeFeo says, “One of the key advances we’ve made is to use a whole transcript-based amplification system that utilizes both oligo dT and random hexamer primers allowing the successful amplification of these highly degraded RNA sources.”

Do gene-expression results from these samples capture real biology? “Anything that is useful to scientists has to be reflected in their ability to capture real-world data,” said DeFeo. “And ours do. We have correlated findings between fresh frozen samples and FFPE samples in a colon cancer study. We have many ways to assess the similarities; for example the linear correlation of differential expression values between the fresh frozen samples and the FFPE samples were around .91, which is high given the degradation associated with the FFPE RNA. The other advantage is that our technology is focused on using tiny amounts of degraded RNA. It’s optimized for the small amount of RNA, we are recommending 50 ng of FFPE-isolated RNA.”

Tissue Fixation Time

Susan J. Done, scientist, division of applied molecular oncology, Ontario Cancer Institute (OCI), discussed the impact of tissue-fixation time and whole-genome amplification on array comparative genomic hybridization profiles. She focused on molecular pathology, translational research working with small samples of patient tissues. “This has great potential to identify novel diagnostic and prognostic markers for human disease. We use a small amount of starting material, WGA, then array comparative genomic hybridization. It is essentially a whole genome screening technique that can be performed on virtually any clinical sample.”

In studying patient samples, often the only material available is archived FFPE samples. “We have performed studies on the effect of different periods of formalin fixation on array profiles. We found that periods of fixation similar to those employed clinically have only a minor impact on DNA quality.”

Previous articleResearch Shows Expanded Role of ncRNAs
Next articleAldagen Establishes Manufacturing Facility