8:30-8:40 Welcoming Remarks from Conference Director
Cambridge Healthtech Institute

Advances in FT-ICR-MS for Biomarker Discovery

8:40-8:45 Chairperson’s Opening Remarks

8:45-9:10 High-Resolution Differential Mass Spectrometry (dMS) Reveals Markers of Alzheimer’s Disease in Humans
Ronald C. Hendrickson, Ph.D., Director, MP Proteomics, Merck Research Labs

9:10-9:35 Plasma Protein Biomarker Discovery for COPD Progression Using FTICR-MS Accurate Mass and Time Tag
Joel Pounds, Ph.D., Senior Staff Scientist, Cell Biology & Biochemistry, Biological Sciences Division; Science Advisor, Environmental Biomarkers Initiative, Pacific Northwest National Laboratory 
Chronic Obstructive Pulmonary Disease is the fourth leading cause of death in the United States, and the annual cost to the nation is over 30 billion dollars. The objective of this study was to identify plasma protein biomarkers that are predictive for, or correlate with, progression of COPD. Two hundred COPD subjects from the Utah Lung Health Study were stratified according to the rate of loss of pulmonary function. Plasma from the first and fifth quintile was depleted of abundant proteins using GenWay Seppro12 spin columns and protein from the depleted samples digested with trypsin and analyzed in duplicate using capillary LC-Fourier Transform Ion Cyclotron Resonance mass spectrometry. Partial Least Squares – discriminant analysis was used to identify peptides >250 peptides and the >75 proteins discriminating the FEV1 phenotype and may be useful as pharmacodynamic indicators for use in clinical drug development. 

9:35-10:00 Can Novel Cancer Serological Biomarkers Be Discovered Using Proteomics, and Why Have Past Attempts Failed?
David W. Speicher, Ph.D., Professor & Chair, Systems Biology Division; Director, Proteomics Laboratory, The Wistar Institute
Discovery of novel cancer biomarkers in plasma or serum using proteomics could have a dramatic positive impact on early cancer diagnosis as well as clinical management of disease after diagnosis. Despite this great potential, there has been little success during the first decade of the proteomics era because the complexity of human plasma proteomes greatly exceeds protein profiling capacities of conventional methods such as 2-D PAGE and LC/LC-MS/MS. Two of the most important challenges are the wide range of plasma protein abundances and the highly heterogeneous nature of plasma protein profiles in the human population. Due to the presence of a handful of very abundant, heterogeneous plasma proteins, conventional protein profiling methods can detect only a few proteins below the microgram per mL level. However, the most specific cancer biomarkers are expected to be present at low ng per mL levels or less. Hence, it is not surprising that no novel specific biomarkers have been discovered by conventional protein profiling methods. To further expand the detection of low abundance plasma proteins, additional separation steps must be incorporated into novel higher dimensional protein profiling strategies. A 4-D protein/peptide separation strategy recently developed in our laboratory is particularly promising because it detects many proteins known to be present in plasma at the low ng/mL level, and even detects some proteins at the pg/mL level. Of course, an inevitable consequence of increasing sample fractionation and depth of analysis is further reductions in the number of proteomes that can be analyzed. Therefore, these more comprehensive methods are best suited for projects that require analysis of only a few samples, such as mouse xenograft models of primary and metastatic cancer. After primary and/or metastatic tumors have developed, the plasma is analyzed using the 4-D method together with a high mass accuracy LTQ FT- ICR mass spectrometer to identify human proteins in the mouse plasma. Concentrations in normal human plasma of human proteins specifically shed by the tumors are then estimated using a comprehensive human plasma proteome database. Using this strategy, at least 100 low abundance human plasma proteins can be identified per experiment. The most promising candidate for cancer biomarkers must then be validated in plasma from cancer and control patients. Current efforts are focused on developing medium-throughput methods capable of testing a large number of biomarker candidates in modest sized patient cohorts.

10:00-10:45 Coffee Break with Exhibit and Poster Viewing

Biomarker Identification from Complex Mixtures

10:45-11:10 Targeted Proteomics and Biomarker Quantitation Using Multiple Reaction Monitoring Tandem Mass Spectrometry
Jeffrey S. Patrick, Ph.D., Integrative Biology, Biomarkers, Eli Lilly & Co. 
The development of LC/MS/MS/MRM (MRM = multiple reaction monitoring) methods for the quantitative analysis of specific proteins in complex biological samples will be discussed. The transition from a global LC/MS/MS (scouting) proteomics method to the targeted MS/MS(MRM) method will be pivotal. Key considerations in the choice of peptides and transitions chosen for the MRM will be discussed. Target fluids include cerebrospinal fluid, serum, and plasma as well as tissue samples. Different approaches to quantitation will be discussed including relative, isotopic peptides and isotopic proteins. The quantitation of phosphorylation will be discussed using MRMs. Methods utilizing enrichment techniques including immunopreciptation and those without, will be compared and contrasted. The ability to quantify proteins through their peptide surrogate at the ng/mL level in undepleted serum or in the pg/mL concentration using enrichment techniques will be demonstrated and discussed. Finally, approaches to the validation of the methods, including examples, will be provided.

11:10-11:35 Proteomics: Finding Biomarkers for Ischemic Myocardium
Jennifer Van Eyk, Ph.D., Director, The Hopkins NHLBI Proteomics Center, Director, Bayview Proteomics Group, Associate Professor Medicine, Division of Cardiology, Biological Chemistry and Biomedical Engineering, Johns Hopkins University
There is currently a debate centered on the feasibility of biomarker discovery directly from serum versus indirectly based on tissue analysis coupled to a second serum validation step. The type of biomarker and the target disease dictates the usefulness of each approach. With myocardial ischemia, there is a need for a diagnostic marker(s) capable of early detection of ischemia in chest pain patients presenting to the emergency department. The traditional cellular necrosis markers, cTnI and cTnT, which originate from the heart, are detected only late during the evolving ischemic episode. Even though these cellular-based markers undergo specific ischemic-induced modifications that can be used for risk stratification, specific markers for ischemic are still desired. To address this clinical need, we have carried out in-depth proteomic analysis of individuals undergoing cardiac catherization and/or angioplasty in which balloon inflation induces a timed ischemic event. Serial serum samples were obtained under strict collection protocols from 7 patients (3 time points) diagnosed with myocardial infarction at baseline, ischemic and at peak cell necrosis (based on cTnI) and compared to time matched samples from 7 patients with stable angina (disease control) and healthy controls (pooled and 30 individuals) using multiple protein separation technologies prior to MS to increase proteome coverage and depth.

11:35-12:00 In-Depth Proteomic Analysis of Human Plasma Using Depletion of Abundant Proteins and Multi-Lectin Affinity Chromatography (M-LAC) for Biomarker Discovery
Marina Hincapie, Ph.D., Principal Research Investigator, The Barnett Institute of Chemical and Biological Sciences, Northeastern University
We report on the development of a robust, reproducible and high-throughput platform suitable for in-depth proteomic analysis of human plasma and biomarker discovery. The method consists of automated in-line depletion of abundant plasma proteins using immunoaffinity columns and further fractionation of plasma into a non-glycosylated (unretained) and glycosylated (retained) protein fractions using multi-lectin affinity chromatography (M-LAC). The unretained and retained fractions are digested with trypsin; the peptides are separated and analyzed by LC-MS/MS. All steps in the method are monitored at multiple quality control points. When applied to the analysis of human plasma, this method detected proteins that are present in plasma at concentrations of 10-100 ng/mL. At this level of detection we identify numerous tissue leakage proteins representing different protein families such as: transcription factors, protein kinases and cell adhesion proteins. When the method was used in an autoimmune disease biomarker discovery effort, eleven proteins were found to have a change in differential abundance in comparison with matched controls. Potential candidate biomarkers were quantitatively verified by ELISA measurements.

12:00-1:30 Lunch on Your Own

Protein Microarrays for Biomarker Discovery

1:30-1:35 Chairperson’s Opening Remarks

1:35-2:00 “Reverse Capture” Autoantibody Microarray for Biomarker Discovery
Brian Liu, Ph.D., Assistant Professor of Urology, Director of Translational Research in Urology, Brigham and Women’s Hospital, Harvard Medical School
Diagnosing cancers based on serum profiling is a particularly attractive concept. However, the technical challenge to identify serum biomarkers is the dynamic range of protein amounts. However, the patients’ sera contain antibodies that react with a unique group of autologous cellular antigens. Proteins not present in normal cells may elicit a host immune response, which affords a dramatic amplification of signal in the form of antibodies relative to the amount of the corresponding antigens. To date, studies of autoantibody reactivity using protein microarray technology have relied on recombinant proteins and/or synthetic peptides as arrayed features. A major drawback of this approach is that recombinant proteins may not contain native protein conformations and important disease-related post-translational modifications (PTMs). We have recently developed a “reverse capture” microarray platform, which allows the detection of autoantibodies against 500 unique antigens that are immobilized in their native configuration onto an array surface, and facilitates effective comparison of autoantibody profiles between different patient cohorts, including antigens with unique disease-associated PTMs. We will present its use for antigen-autoantibody profiling with prostate cancer as a case study.

2:00-2:25 Protein Microarray Profiling of the Immune System Identifies Diagnostic, Prognostic and Tissue Biomarkers
Michael Tainsky, Ph.D., Professor, Molecular Biology, Karmanos Cancer Institute
Global profiling of the immune system can be exploited as a biosensor to characterize a disease condition. We use an unbiased high-throughput identification of disease-related antigens by employing patients’ immunoglobulin-G molecules, both as the bait for cloning the biomarkers, as well as for discriminating cancer patients from healthy subjects in a two-color fluorescence system using protein microarrays. The essential feature of the approach is the acknowledgment of the heterogeneous nature of disease and that the immune system provides an exquisitely sensitive sensor. We employ specialized data informatics analysis using machine learning to identify attributes of disease and validate the systemic response to disease. We demonstrated that the protein antigens were overexpressed in ovarian epithelial tumors for three out of three antigens chosen. This confirms our hypothesis that overexpression of proteins in tumors leads to the induction of serum antibodies in ovarian cancer patients.

Technology Showcase

2:25-2:40 Protein Sample Preparation Lisa Bradbury, Director, Research and Development, Pall Corporation

2:40-2:55 Label Free Quantification and High Performance Capillary Liquid Chromatography Nanospray Mass Spectrometry for the Characterization of Cerebrospinal Fluid
Tony Tegeler, PhD, INCAPS
Typical proteomics experiments using cerebrospinal fluid (CSF) require at least 500 mL of each sample to be analyzed. We have developed a nanospray approach that can be carried out with as little as 150 mL of CSF per sample. With this approach, more proteins are identified with high confidence, and we maintain a very high quality chromatographic alignment making this method suitable for label-free quantification. The smaller volume requirement for our approach enables investigators to maximize the information that can be generated with the available human samples. It also opens the possibility to conduct experiments using rat CSF.

2:55-3:10 Discovery to Verification of Candidate Biomarkers for Metastatic Cancer  John Hevko, Senior Field Applications Specialist, Applied Biosystems

3:10-4:00 Refreshment Break with Exhibit and Poster Viewing

4:00-4:25 Applications of Proteomics Technologies to Cancer Diagnostics: Case Study Highlighting Two Clinical Trials Accruing Patients for Biomarker Discovery
Elise C. Kohn, M.D., Head, Molecular Signaling Section, Principal Investigator, Laboratory of Pathology, NCI

4:25-4:50 Role of Tissue Biomarker Proteomics in Cancer Drug Discovery
Jing Wei, Ph.D., Senior Scientist, Biological Mass Spec, Biogen Idec
Proteomics technologies have been applied throughout the drug discovery and development process to support biomarker discovery and provide a global perspective of the integrated proteome network. Proteomic analysis of primary tumor tissue is particularly relevant to the development of cancer therapeutics since tumor tissue samples are potentially enriched with disease-related and pharmacodynamic protein biomarkers. Insight into the tumor proteome thus has the potential to provide a better understanding of disease progression, drug response, mechanism of action for lead candidates in disease models, patient stratification in clinical design, and assessment of efficacy biomarkers in the clinic using validated biomarkers. However, tumor tissue presents a particular challenge for mass spectrometry-based proteomic analysis. We have developed a suite of methods allowing us to profile tissue proteomes directly from various sample sources including fresh frozen xenograft tumors and formalin-fixed paraffin-embedded (FFPE) clinical specimens. Protocols for sample preparation have been established to achieve highly efficient and reproducible protein extractions and proteolytic digestions. To address the highly complex nature of these tumor profiles, we have developed a fully automated tandem multidimensional separation system coupled with ESI-tandem MS–on-line (LC/LC/LC-MS/MS/MS) to enable high-resolution global proteome profiling of tumor tissues. Bioinformatic software packages have been customized for optimal analysis of these large (multi-GB) datasets. The system is highly flexible, and can analyze sample amounts from as low as tens of micrograms up to milligram quantities of sample. From a typical sample load (a few hundred micrograms), each single analysis yields over 10,000 high-confidence protein identifications from IPI database searches with a false positive rate of less than 2% at the protein level. Detailed and reproducible proteome maps constructed from various tumor samples can be utilized as a foundation for quantitative tissue biomarker discovery. To illustrate Biogen Idec’s quantitative proteomics process, study results from tumor xenograft models treated with Carboplatin will be discussed.

4:50-5:15 Organelle Proteomics - Closer Look into Liver and Hepatocellular Carcinoma Plasma Membranes
Djuro Josic, Ph.D., Professor, Medicine (Research), Brown Medical School, Director, Proteomics Core, COBRE Center for Cancer Research Development, Rhode Island Hospital
To find potential biomarkers for hepatocellular carcinomas, rat liver and hepatocellular carcinoma Morris hepatoma 7777 were compared. To separate plasma membranes from contaminating organelles, monoclonal antibodies against specific integral plasma membrane proteins immobilized on magnetic beads or on highly porous monolithic supports were used. After sample preparation, proteins were identified by 1 and 2D LC-MS/MS. The role of differently expressed proteins as possible cancer biomarkers is discussed. Membrane proteins play a key role in malignant modification, cell-cell and cell-matrix interaction, and recognition of cancer cells by immune system. Some of them such as transferrin receptor are candidates as potential biomarkers. However, because of their hydrophobicity and post-translational modifications, identification of these proteins by MS/MS is still a problem that will be discussed in the presentation.

5:30-6:30 Roundtable Discussions Discussion Topics Include: Advances in Mass Spectrometry Multi-Dimensional Separations and Sample Enrichment Protein Microarrays Diagnostic Development Cancer Biomarkers Post-Translational Modifications Predicting Response to Therapy