Expression Analysis (EA) now offers microarray processing and analysis services for whole-genome expression and copy number for FFPE samples. EA has formed collaborations with Agencourt Biosciences and NuGEN Technologies for nucleic acid isolation and target preparation for Affymetrix U133 expression analysis and 500K copy number services.

EA has offered array-based profiling for copy number analysis since 2005. The addition of whole-genome expression profiling services for FFPE samples is the result of a new collaboration with NuGEN Technologies, providers of the target preparation reagents and chemistry.

“We’re able to get about 500 nanograms of RNA per 10-micron section,” said Thomas Goralski, Ph.D., Vice President of Laboratory Operations, EA. “The RNA works quite well with the NuGEN amplification and labeling system. We see up to 50 percent of transcripts present on the Affymetrix Human Genome U133A 2.0 Array, which is an improvement of several orders of magnitude when compared to other amplification and labeling systems.”

The NuGEN WT-Ovation FFPE System is a whole transcriptome amplification assay optimized for degraded and small amounts of input RNA. EA typically starts with about 50 nanograms of RNA when using the NuGEN kit for amplification and labeling. RNA extracted from FFPE samples is typically severely degraded, and suffers from chemical modifications that make it less amenable to reverse transcription.

Goralski recently sat down with User Forum

editor Tommy Broudy to discuss EA’s microarray profiling services for RNA and copy number analysis for DNA isolated from FFPE samples. Topics covered include:

• Performance of FFPE samples for whole-genome expression analysis
• Collecting expression and copy number data from the same sample
• Range of services offered by EA

RNA PROFILING PERFORMANCE
Broudy: What type of performance do your clients expect for expression profiling of from FFPE samples?

Goralski: Our recent results have been very encouraging. When we analyze RNA from blocks that are three years old, the distribution on Agilent Bioanalyzer looks like the RNA is totally degraded. But when we hybridize to the array, we can see up to 50 percent Present on the Affymetrix U133A 2.0 Array, which is about 10 times higher than we had ever seen before.

Using the Agencourt Biosciences FormaPure isolation kit, we’re able to get between 100 and 1,000 nanograms of RNA per 10-micron section, which was quite a surprise. Currently, there is a little concern about the A260/280 ratios being a little lower than we’d like and we are continuing to work with Agencourt on this, but the RNA works quite well with the NuGEN amplification and labeling system. The first time we got to the 50 percent Present mark with the array data we were just blown away.

You know, there are a number of small PCR-based

tests for paraffin-embedded samples that look at one or two dozen transcripts. But the ability to look genome-wide adds a lot of value to people’s ability to make use of these tissues.

FFPE vs. FRESH FROZEN TISSUE
Broudy: How comparable is the array data between fresh frozen and FFPE samples?

Goralski: One of the things that people have been caught up with is demonstrating equivalency by looking at scatter plots and R2 values and such—I think that’s going to be a challenge and not necessarily productive.

Our clients are much more interested in determining whether or not they can identify a previously known pattern, to the extent that they can continue to use that pattern as a biomarker. To look at those types of comparisons, we recently ordered a “quad set” where we have the fresh frozen and the paraffin-embedded for both the tumor and the normal adjacent tissue, and we ordered two more tissue sets from our vendor for colon and for lung. Those experiments are currently in progress.