Quantitative peptide analysis by mass spectrometry (MS) has become indispensable for interrogation of proteins in complex biological samples. This complexity requires multiple approaches, so a thorough understanding of these options helps to select the most appropriate protocol for the given question.
The high diagnostic potential of peptides and proteins from cells, tissues, and biological fluids is unquestionable. But have the MS-based strategies matured enough to unlock this potential and to allow the translation from specialized applications into diagnostics and therapeutics?
Pancreatic adenocarcinoma, the most lethal type of cancer in the United States, often begins in the duct connecting the pancreas to the duodenum. The silent progression conceals the diagnosis at the early stages. Thus, little is known about early biomarkers of pancreatic cancer.
“Our team was the first to conduct an in-depth molecular comparison of cancer and normal tissues for the presence of cancer stem cells,” says David M. Lubman, Ph.D., professor of surgery at the University of Michigan, Ann Arbor. “Our method overcomes a common problem of heterogeneity of cancer tissues. Most surgically removed tissue samples contain a low percentage of tumor cells embedded in a dense stroma. Identification of crucial cancer biomarkers requires analysis of a pure and homogeneous cell population.”
“However, usefulness of microdissection is often limited by the difficulties in identifying desired cell types by morphology alone.”
Dr. Lubman’s team used a combination of immunostaining with laser capture microdissection (immuno-LCM).
“This method allowed us to zero in on a small population of cancer stem cells expressing CD24, a specific cell surface biomarker shown to correlate with the development and metastasis of pancreatic cancer,” continues Dr. Lubman. “The excised material contained only 10,000–15,000 cells.”
Analysis of immunophenotypically defined cell populations markedly increased specificity of liquid chromatography-tandem mass spectroscopy (LC-MS/MS). Over 375 differentially expressed proteins were analyzed by Ingenuity Pathway Analysis (Ingenuity Systems). Not surprisingly, the major groups of differentially expressed proteins were involved in tumor migration, invasion, and progression.
Three selected candidate biomarkers (CD59, CD70, and CD74) mediate the evasion of the immune response. Overexpression detected by MS was further confirmed by tissue staining and microarrays. CD74 specifically correlated with early cancer development and was critically positioned in the NFκB signaling pathway.
“The early appearance of CD74 means that this candidate biomarker could be a promising tool in therapeutic drug discovery. Our next step is to search for these biomarkers in serum,” asserts Dr. Lubman.
Improved Sensitivity of MS Analysis
According to Yuri van der Burgt, Ph.D., at the Leiden University Medical Center in the Netherlands, sample workup is an essential part of any proteomics workflow.
“Biomarker discovery requires both reduction of sample complexity and retention of coverage,” he explains. “At the same time, such sample preparation for the pipeline must be cost-efficient and robust.”
His team from the Center for Proteomics and Metabolomics, headed by Prof. André Deelder, developed two solid-phase peptide extraction protocols that precede an MS-based analysis. The first is based on magnetic beads functionalized with chromatographic properties, and the second deploys disposable column-like cartridges. Both approaches consistently generate high-quality starting material for identification of peptide signatures in biological fluids.
To minimize deviations stemming from preanalytical conditions and to focus on biological variations of the specimen fluids, the team standardized operating procedures for each step of the process.
“We spent a lot of effort on optimizing our protocol for high-throughput analysis,” continues Dr. van der Burgt. Using this method, samples from patients with diagnosed pancreatic cancer were compared to the healthy controls. An automated solid-phase extraction protocol utilized weak cation exchange magnetic beads. The protocol allowed for cleanup and enrichment of certain subsets of peptides and improved sensitivity of the MS assay.
“Early detection is paramount for patient survival,” says Dr. van der Burgt. ”Imaging technologies provide good tools for this purpose, but can still benefit from additional information such as blood-based serum biomarkers that may pick up the small changes of cellular function.”
Out of several hundred peptides reproducibly identified by this profiling, a set of seven was found to be sufficient to differentiate pancreatic cancer from normal samples with a sensitivity of 78% and a specificity of 89%. The validation efforts continue with acquisition of additional samples from other clinical labs.
In the future, Dr. van der Burgt sees tremendous diagnostic value for individuals with an increased risk of developing pancreatic cancer, either genetically or due to smoking or diabetes. A similar approach was used to detect breast cancer signatures. To capture a larger percentage of the proteome, researchers used two types of functionalized beads and combined the resulting data. Meanwhile, the research continues to improve the method to make it ready for clinical use.
Reliable Proteomic Analysis
The holy grail of protein profiling is to determine comprehensively the identity, abundance, and isoforms of proteins in a large number of clinical samples, maintains Jun Qu, Ph.D, associate professor at the University of Buffalo, SUNY.
“Biomedical discovery necessitates comparison of multiple samples to overcome known problems with individual variability. However, accurate and reliable profiling of multiple samples remains challenging,” he explains.
“We developed a highly reproducible and quantitative method that can detect proteomic signatures despite a highly variable biological background.”
The “precipitation/on-pellet-digestion” method relies on a unique detergent cocktail to achieve a near complete extraction, including that of membrane proteins. Addition of organic solvents, like acetone, denatures and aggregates the pellet, which is notoriously difficult to re-solubilize.
“Our sample prep philosophy is the simpler, the better,” continues Dr. Qu. “We let the enzymes digest the proteins right on the pellet. This method produces a clean peptide mixture suitable for LC-MS analysis.”
The team designed a unique separation column that features low void volume, large loading capacity, and high separation efficiencies. Peptides are separated on a heated, 100-cm-long column during a seven-hour run. The team determined that a label-free MS approach was the most cost-efficient method. It provided excellent accuracy for relative protein quantitation when well controlled.
The process was previously applied to several clinically important questions, such as finding underlying causes of hibernating myocardium and chronic obstructive pulmonary disease. The most recent research focused on HIV.
“Five percent of HIV carriers never progress to a disease state,” says Xiaomeng Shen, a Ph.D. candidate in Dr Qu’s lab. “Proteomic comparison of normal progressors vs. long-term nonprogressors (LTNPs) provides highly valuable insights for directing therapeutic efforts.”
Comparison of 10 individuals from each cohort resulted in the identification of 78 differentially expressed proteins. Subsequent annotation by Ingenuity Pathway Analysis software pointed to several signaling pathways that may account for viral control mechanisms in LTNPs.
“We were excited to find the SAMHD1 protein in the interferon pathway, which was recently implicated as a critical player in viral restriction,” remarks Dr. Qu, who notes that it would take many years to generate the same amount of data by the traditional methods.