NEW YORK(GenomeWeb) – Researchers at New York University's Langone Medica lCenter have developed a mass spec workflow capable of using laser-capture microdissected formalin-fixed, paraffin embedded samples.
Detailed in a paper published this week in Nature Scientific Reports, the work demonstrates for the first time the feasibility of mass spec-based proteomics on FFPE samples as small as several thousand cells, NYU researcher Thomas Wisniewski, senior author on the study, told GenomeWeb.
The work simultaneously tackles a pair of sampling issues that have traditionally proved a challenge for mass spec – use of FFPE tissue and use of tissue isolated by LCM.
FFPE tissue is one of the most common type of clinical samples, but, due to the large amount of protein cross-linking caused by the fixing process, proteomics researchers have had trouble using such samples.
This has changed in recent years as a number of research groups and companies have developed workflows for extracting proteins from FFPE samples. Last week, for instance, a team led by researchers at the European Institute of Oncology presented in Molecular & Cellular Proteomics an approach for mass spec-based analysis of histone post-translational modifications in FFPE samples.
A separate issue has been combining LCM with mass spec analysis. LCM uses a laser coupled to a microscope to allow pathologists to isolate particular portions of a sample of interest with high precision. The method enables researchers to, for instance, collect specific cell types for analysis, or, in the case of cancer research, potentially improve proteomic analyses by allowing for the removal of stromal cells that could mask the protein content of tumor cells.
Traditionally, however, the technique leaves sample sizes too small to perform biomarker discovery on mass spec, and so it has been more commonly used to prepare samples for antibody-based analyses like reverse phase protein arrays.
Here, too, though, researchers have made advances recently. For example, in a study published last year in the Journal of the National Cancer Institute, Erasmus University scientists used LCM coupled to mass spec analysis for a breast cancer biomarker study in which they identified an 11-protein signature that appeared to distinguish between more and less aggressive forms of triple-negative disease.
This work, however, used frozen tissue, which is generally harder to come by and trickier to handle than FFPE samples. The novelty of the NYU effort, Wisniewski said, is that it managed to combine FFPE samples with LCM followed by mass spec, opening the large amount of FFPE material around the world to such an analysis.
Wisniewski and his colleagues worked to determine the optimal sample prep and extraction conditions that would allow for the identification of the maximum number of proteins via mass spec. The main contributing factor to their success, though, he said, was simply improvements in mass spec technology.
"I tried a very similar project five years ago, and it failed miserably," he said. "The LCM technology is a little bit better now, as well, but the key thing is the mass spec and the mass spec software that is available to deconvolute the results."
In the Scientific Reports paper, the researchers used a Thermo Fisher Scientific Q Exactive to look at microdissected neurons taken from FFPE samples of the temporal cortex of patients with Alzheimer's disease. In all, they were able to identify more than 400 proteins from these cells, 78 percent of which were known neuronal proteins, demonstrating the approach's ability to extract relatively uniform populations of cells.
The success of the technique, Wisniewski said, opens up a variety of research possibilities, both in his field of neurodegenerative disease and other areas like cancer.
Currently, he is using the method to look at how the proteomic profiles of neurons, amyloid plaques, and microglia in patients with very rapidly progressing Alzheimer's disease compare to those in patients with more normally progressing Alzheimer's. Preliminary results from this work indicate that the neuronal proteins and amyloid plaques are different in the two types of disease, he said.
Wisniewski and his colleagues are also using the technique to look at proteomic profiles specifically of cells in the medial entorhinal cortex, where the effects of Alzheimer's disease are thought to first manifest themselves.
"There are selected cell populations that are affected first, and we are looking to see what are the proteomic profile changes that are way upstream from the [ultimate] amyloid and tau pathology," he said.
They plan to then follow patients sequentially through the various stages of the disease looking at how the proteomic profiles of affected brain regions change.
"So, I think it will allow us to do proteomic profiling of the very earliest stages of Alzheimer's disease, which I think might give us some interesting clues as to what is really driving the ultimate amyloid/tau pathology and perhaps provide us with some novel therapeutic targets," he said.
The method will also allow for proteomic follow-up to recent GWAS efforts that have demonstrated that macrophage and microglial function play a key role in the development of Alzheimer's disease. "We can look at the proteomic profiles of each of these cell populations differentially at the different stages of the disease," he said.
Looking beyond neurodegenerative disease, Wisniewski noted significant interest in the technique on the part of his colleagues in cancer research. One application, he noted, might be using it to look at the proteomic profiles of specific populations of cells in, for example, glioblastoma, with different oncogenes either turned off or on.
"It is a technical advance that allows many important questions to be easily addressed using small amounts of archival tissue that people have a ton of," he said. "So it makes life much easier for investigators in any field."