By Monica Heger
Techniques such as ChIP-seq are thought to hold potential in uncovering the epigenetic modifications that are relevant for cancer and other diseases, but ChIP-seq studies on banked samples have been hindered to date because chromatin is difficult to extract from formalin-fixed, paraffin-embedded tissue.
Now, researchers from the University of Urbino in Italy and the European Institute of Oncology have devised a protocol, dubbed pathology tissue ChIP, or PAT-ChIP, in which they have essentially reversed the paraffin embedding process and optimized the chromatin extraction and preparation steps to enable ChIP-seq or other analyses from FFPE samples. In a proof-of-principle paper published last month in PNAS, they demonstrated their technique on spleen tissues from a mouse model of leukemia.
Saverio Minucci, a professor of pathology at the University of Milan and senior author of the study, said his team is now using the technique on human breast cancer samples to see if it can find biomarkers indicative of how a patient responded to treatment.
"This is just one way we think this approach will be useful," Minucci told In Sequence. "Up until now, it's been difficult to study the epigenetic profile of a patient" in FFPE samples.
FFPE samples are valuable resources because they usually come with a complete clinical history of the patient. While large collections of these samples exist, it hasn't been possible to do ChIP-seq analyses on them due to a lack of methods for extracting and preparing the chromatin in a way that would provide sufficient intact chromatin, Minucci added.
"There has been a lot of interest in whether you could take these pathology samples and get information about histone modifications," said Michael Stitzel, a postdoctoral fellow at the National Human Genome Research Institute who has used ChIP-seq in combination with GWAS to identify regulatory elements associated with type 2 diabetes (IS 11/9/2010). "There are lots and lots of banked paraffin embedded tissue, and being able to look at the epigenetic changes is very exciting."
The problem with paraffin-embedded tissue, Stitzel said, is that fixing the tissues requires the use of a high concentration of formaldehyde, which causes crosslinking among the transcription factors and chromatin. The crosslinking can make it difficult to distinguish between technical artifacts and true histone modifications.
"The paraffin embedding presents an additional hurdle compared to fresh cells that are doing their thing and have never been crosslinked," Stitzel said.
Minucci and his team were able to essentially reverse the paraffin-embedding process by first incubating the tissue in a hystolemon solution, which removes the paraffin wax, and then rehydrating it by soaking it in decreasing concentrations of ethanol.
To test the approach, they used a mouse model of leukemia and extracted two sets of cells from the spleen — one of which they first fixed and embedded in paraffin, the second of which they did not. On the FFPE sample, they tested their PAT-ChIP method, and on the isolated cells, they performed the standard ChIP procedure. They found that both yielded comparable results.
They then tested PAT-ChIP on human tissue — from an FFPE testicular germ cell tumor and six FFPE breast tumor samples. Chromatin from the samples was immunoprecipitated against known histone modifications and then analyzed by qPCR.
"Taken together, these results demonstrate that PAT-ChIP can be used to explore the epigenome of patient-derived, formaldehyde-embedded tissue samples stored for several years in pathology archives," the authors wrote.
Next, they tested their PAT-ChIP method in conjunction with high-throughput sequencing on the Illumina Genome Analyzer. Using spleen tissue from the mouse leukemia model, they performed two experiments each of standard ChIP-seq on the isolated cells, and PAT-ChIP-seq on the FFPE sample. There was a high correlation between the two different methods, comparable to what was seen from comparing the two results from the same method. Namely, there was a 90 percent peak overlap within a 2.5 kilobase window between the two different methods.
Stitzel said that the method appears to be relatively simple to use, and could be an important way to study human FFPE tissue samples. However, he said that further tests are required to determine whether the method is compatible with all types of tissue, or whether additional modifications will have to be made.
"It will be interesting to see if this is one-size-fits-all, or if you need to do modifications to prepare samples [from different tissues]. Sample prep is a big deal for any of these deep sequencing techniques," he said, so if the method did require extensive modifications, that would likely hinder its adaptation. However, he added that it "looks from the protocol that it's pretty reasonable to do."
"There's a lot of paraffin-embedded archived material out there that could be used for these purposes," he said.
Researchers at the University of Münster in Germany recently began testing the technique on FFPE lung tumor samples. "For the first time we can really look in detail at patient samples," said Carsten Müller-Tidow, a professor of medicine and clinician at the university.
The team has around 400 samples of paraffin-embedded lung tumors, 100 of which they are planning to study using the PAT-ChIP-seq technique on Life Technologies' SOLiD. The team is in the very early stages of the experiment and still in the process of establishing the technique, and Müller-Tidow said that it didn't seem like many adjustments would need to be made from the published protocol. "We'll be using the method in the near future," he added.
Have topics you'd like to see covered in In Sequence? Email the editor at mheger [at] genomeweb [.] com.