Goldman - Georgetown University

Principal Investigator: Radoslav (Rado) Goldman, PhD
Institution: Georgetown University, Washington, DC

Subcontract Principal Investigator: Raja Mazumder, PhD
Institution: George Washington University

Treatment of cancer diseases improves with early detection. We are investigating site-specific protein glycoforms as a molecular signature of hepatocellular carcinoma (HCC), the most common form of liver cancer. The incidence of HCC increases in the United States primarily due to hepatitis C viral infection. HCV infection triggers changes in glycosylation of liver secreted proteins that can be followed along the progression of liver disease to HCC. Alpha fetoprotein is an example of a liver secreted glycoprotein used clinically to detect hepatocellular carcinoma. This protein detects approximately 60% of HCC patients and its glycosylated form has an improved specificity for the detection of HCC. In this study, we are looking for novel glycoforms of proteins that would improve detection of liver diseases and HCC.

Detailed characterization of the liver secreted glycoproteins is challenging due to the heterogeneity of glycoforms at each glycosylation site. We are optimizing analytical methods for glycoprotein characterization with special interest in quantification of the site specific protein glycoforms. We are developing mass spectrometric methods for characterization and quantification of glycoforms and, in collaboration with Dr. Nathan Edwards developing software tools for glycopeptide analysis. These methods are applied to the characterization of patient materials with the goal to establish a clinically applicable assay based on the quantification of disease-specific glycoforms.

In addition, we are examining the connection of genomic variants affecting protein glycosylation with the development of hepatocellular carcinoma. In collaboration with Dr. Raja Mazumder, George Washington University, we have established infrastructure for identification of the variant glycoproteins and propose to use it in the study of factors affecting the development of HCC.

Synopsis of Research and Network Resources

Introduction: A Brief Summary of the State of the Science and Research Needs

Better treatment outcomes (e.g. reduction in cause-specific mortality) can be achieved when cancers are detected early. We are investigating site-specific protein glycoforms as a molecular signature of hepatocellular carcinoma (HCC), the most common form of liver cancer. The incidence of HCC has been increasing in the United States primarily due to hepatitis C (HCV) infection and increased incidence of non-alcoholic steatohepatitis (NASH). HCV infection triggers changes in glycosylation of proteins secreted by the liver that can be followed with the progression of liver disease to HCC. Alpha-fetoprotein (AFP) is an example of a liver-secreted glycoprotein used clinically to detect HCC. This protein detects approximately 60% of HCC patients and its glycosylated form has an improved specificity for the detection of HCC. In our Alliance study, we are looking for novel glycoforms of liver-secreted proteins that will improve detection of liver diseases and HCC and develop clinically applicable LC-MS methods for their quantification.

Laboratory-specific Studies to Meet the Research Needs

We combine liquid chromatography-mass spectrometry (LC-MS)-based glycoproteomic profiling of patient sera with informatics approaches for identifying liver-secreted glycoprotein biomarker candidates specific for liver diseases, in particular, HCC. Detailed characterization of liver-secreted glycoproteins is challenging due to glycan heterogeneity at each glycosylation site. We have designed glycosidase-assisted LC-MS methods for the selection of biomarker candidates. These methods allow detection of disease-related changes at the glycopeptide level and are able to resolve linkage positions of relevant glycoforms. Informatics analyses, including development of software tools for interpretation of glycopeptide LC-MS datasets, further refine identification of biomarker candidates. A new software for the analysis of glycopeptide MS-MS datasets was developed in collaboration with Dr. Nathan Edwards (Georgetown University). Our glycosidase-assisted LC-MS discovery method provides us with information on chromatographic and mass spectrometric behavior of the glycopeptide biomarker candidates. We expect that these methods will be compatible with clinical testing of patient samples to allow quantification of disease-specific glycoforms of glycoprotein biomarkers.

In addition, we have developed tools for analysis of potential N-linked glycan attachment sites based on the NXS/T sequon. In collaboration with Dr. Mazumder, integrated databases of genomic variants that affect protein sequences in a way that might alter the N-glycosylation state of proteins have been developed.. We suspect that such protein variants are associated with disease progression and will expand identification of disease biomarkers. Site-specific glycoforms of the selected protein candidates are examined directly in patient samples in the appropriate disease context.

Our discovery studies consistently detect increased fucosylation in connection with liver disease progression. We examine whether the fucosylation occurs in the core or outer arm, and whether any class of fucosylation, is specific for HCC development. The candidates are further evaluated in quantitative verification studies.

Resources and Reagents for Sharing

The glycosidase-assisted LC-MS and LC-MS-MRM (multiple reaction monitoring) methods provide an important resource for the entire research community. An advantage of the assays is their ability to quantify site-specific glycopeptide changes. This is a highly specific and quantitatively accurate detection method for this class of protein modifications, which opens new opportunities for molecular disease classification.

The informatics analyses and software tools represent another set of resources that we share. The glycopeptide MS data analysis software is well received by the research community and is freely available at https://edwardslab.bmcb.georgetown.edu/trac/GlycoPeptideSearch/. Additional informatics resources are available at https://hive.biochemistry.gwu.edu. This includes all non-synonymous single-nucleotide variations (nsSNVs) mapped to the human proteome, which allows researchers to assess how polymorphisms result in gain or loss of N-linked glycosylation sites. Structure Feature Analysis Tool (SFAT), is a tool for structure-based comparative analysis of N-linked glycosylation sites, that was developed for comparative analysis of N-glycosylation in five species ( H. sapiens, M. musculus, D. melanogaster, A. thaliana, and S. cerevisiae). Dr. Mazumder also integrated data into a software (SNVDis) that his group developed for a proteome-wide analysis of SNV distribution. This analysis tool integrates major variation data and sequence feature annotation from major bioinformatics databases, including UniProtKB, dbSNP, COSMIC, Pfam, RefSeq, CDD, and PANTHER, and then further maps the integrated data to the human reference proteome. SNVDis provides a detailed overview of nsSNVs distributed over the complete human proteome, as well as integrated information on active sites, pathways, and binding sites/domains. This tool and integrated data forms the basis for BioMuta, which is a database of cancer mutations being prepared for the Early Detection Research Network (EDRN) community.

Public Health Implications and Advancing the Field of Glycobiology

Identification of biomarker candidates and their accurate quantification by glycosidase-assisted LC-MS-MRM methods has profound implications in the field of glycobiology and public health. The glycosidase-assisted LC-MS methods have the ability to resolve linkage isoforms at site-specific locations. This allows an accurate quantitative assessment of glycosylation changes at specific proteins and in specific sequons of a given protein. This also allows the desired resolution of the site-specific glycoforms, which is important for molecular disease classification. These methods have applicability in molecular classification of a broad spectrum of diseases associated with site-specific differences in protein glycosylation.

Cancer-Specific Relevance: Detection, Prevention and Treatment

As mentioned earlier, early detection of cancer improves treatment outcomes. We are investigating site-specific protein glycoforms as a molecular signature of HCC because improved non-invasive monitoring of liver disease and accurate serologic methods for early detection of HCC are expected to have immediate impact on patient health.

Opportunities for Collaboration

We have already started to provide informatics analysis of select proteins of interest to the members of the Alliance of Glycobiologists. We collaborate with Dr. Abbott (University of Arkansas)for the analysis of glycosylphosphatidylinisotol  ( GPI)-anchored proteins in the context of liver disease. We are also collaborating with the Alliance laboratories at University of California, San Francisco in measuring sulfatase levels in various biological samples in association with liver or head & neck cancers. Our glycosidase-assisted LC-MS and LC-MS-MRM methods provide outstanding opportunities for collaboration with investigators interested in targeted molecular classification of diseases.