A vast reserve of proteins exists within tissues that have been fixed in formalin and embedded in paraffin, but these proteins have been too difficult to extract for most large-scale proteomics experiments. Two new technologies have recently been developed, however, that could change the way proteomics researchers look at preserved tissues. The methods, from the University of Utah School of Medicine and the company Expression Pathology, enable proteins to be extracted and solubulized in a large-scale manner, and then identified by mass spectrometry.
A team of researchers led by Kojo Elenitoba-Johnson, the director of molecular hematopathology and proteomics at the University of Utah School of Medicine, described one technique for extracting and analyzing proteins from formalin-fixed paraffin-embedded (FFPE) material in a paper published Sept. 5 in Laboratory Investigation.
Elenitoba-Johnson and his team first removed paraffin from the preserved sample by passing it through a number of solvents. They then used enzyme digestion to liberate proteins that had been crosslinked to DNA through formalin fixation. The liberated peptides were analyzed by mass spectrometry.
"We used a bottom-up approach," Elenitoba-Johnson explained. "We decided to do a protein digestion, then analyze the peptides and track them back to the databases. In terms of large-scale identification of proteins extracted from paraffin, I think ours would be considered one of the more comprehensive search approaches."
"This unlocks a large number of archived specimens that prior to now have been inaccessible to this kind of large-scale proteomic analysis."
Elenitoba-Johnson's team used their bottom-up approach to identify 324 proteins from a three-year-old FFPE cell block of a human lymphoma cell line. There was "significant" overlap between proteins identified from the preserved sample and proteins identified from a fresh sample of human lymphoma cells.
"I think our paper highlights how this can be done using relatively simple approaches," said Elenitoba-Johnson. "FFPE is the most widely used form of tissue archiving, so I'm sure a lot of people are going to take advantage of this [new method]. Anybody studying those kinds of samples with the intention of finding biomarkers will probably be looking at this because it facilitates the extraction of proteins."
Immunohistochemistry and Western blotting are traditionally used for analyzing FFPE samples, Elenitoba-Johnson noted. With those methods, it requires more work and time to identify the proteins inside the FFPE samples, so researchers can't perform proteomic-scale analysis of those preserved specimens.
"This unlocks a large number of archived specimens that prior to now have been inaccessible to this kind of large-scale proteomic analysis," said Elenitoba-Johnson.
Aside from Elenitoba-Johnson's method, a commercialized kit has also been developed to extract and analyze FFPE proteins in a relatively high-throughput manner.
The kit, developed by Expression Pathology, extracts proteins and nucleic acids from formalin-fixed tissue using the company's patented Liquid Tissue Protein formula. Next, cells are heated, causing proteins to be "shot out" of the sample. The resulting solution can then be analyzed using mass spec or other techniques.
"The method is very simple and you can use standard equipment. It's a relatively easy technique to learn," said David Krizman, the chief scientific officer of Expression Pathology. "Assuming that you have the sample in the tube, from that point, it takes about two hours of incubation, then an overnight incubation after that."
The Expression Pathology kit for FFPE protein extraction costs about $835 for a four-prep kit.
Krizman noted that pathologists have always been happy using preserved tissue for analysis, but the general community had not always been that keen. With the development of a relatively simple kit for FFPE protein extraction, research biologists as well as pathologists have access to a large collection of tissue bearing previously inaccessible information.
"In terms of applications, with toxicity studies, you can treat animals with a drug, and if it comes down with a toxic response you can determine why. And you can better study animal models of disease," said Krizman. "If you add those two things together alone, there [are] huge amounts of archives out there with clinical and research data that are really tightly knit to those tissues."
In a study published in the Aug. 9 issue of Molecular & Cellular Proteomics, Krizman and his colleagues identified about 1,200 proteins from a preserved prostate cancer sample and 800 proteins from a preserved benign prostate hyperplasia sample. Among the identified proteins were biomarkers for prostate cancer that had previously been found before by other researchers.
"The study was a kind of proof of principle of the application of this technology for mass-scale proteomic analysis," said Krizman. "We chose prostate cancer because a lot of biomarkers for prostate cancer are known."
One particularly interesting biomarker found during the study was a protein called GDF-15, Krizman said, which was overexpressed in prostate cancer cells, but not in benign prostate hyperplasia cells. That could make the protein useful for differentiating between the two conditions.
"We're not going to pursue [GDF-15] as a clinical test, but someone else might pick it up and do so," said Krizman.
The next step after the prostate cancer study will be to look for biomarkers for other types of cancers using FFPE tissue, Krizman said.
In addition, Expression Pathology is looking into forming a co-marketing agreement with a "bigger name" company so that more people will find out about the FFPE protein extraction technology, Krizman added.
"Right now we're not marketing our technology very aggressively," he said. "But we're starting to look into co-marketing with others."
Krizman declined to compare Elenitoba-Johnson's method with Expression Pathology's technology.
"I don't really want to say if it's better or worse or comparable," he said. "I think it's great work, but I don't really want to compare it."
— Tien-Shun Lee ([email protected])