Member Information

Publications by

Karen Abbott


Principal Investigator: Karen L. Abbott, Ph.D.
Institution: University of Oklahoma Health Sciences Center, Oklahoma City, OK

Our project, "Glycomics Laboratory for the Development of Ovarian Cancer Biomarkers", is a joint collaboration between our lab at the University of Oklahoma Health Sciences Center (OUHSC), the Complex Carbohydrate Research Center (CCRC) at the University of Georgia, and the Harvard Medical School Center for Glycoscience at Harvard University. The investigators at OUHSC specialize in biomarker discovery using glycomic methods and functional studies of glycosyltransferase enzymes in ovarian cancer progression, the co-investigators at the CCRC specialize in the mass spectrometry identification of O-linked glycans present on the Notch receptor and the functional analysis of these glycans in Notch activation, and the co-investigators at Harvard specialize in the development of antibodies binding to tumor-specific glycans and the development of glycopeptide arrays. The aims of this application are (1) To develop and validate N-linked and O-linked glycoforms for the detection of ovarian cancer and (2) To discover and develop glycosylphosphatidylinositol (GPI) anchored protein glycoforms and GPI anchor binding proteins for the detection of ovarian cancer.

We have developed an innovative method to select antibodies toward cancer-specific glycans. We have recently used this method to select and purify a human scFv antibody that binds to the bisecting glycan structures found in ovarian cancer and glioblastoma. This antibody recognizes tumor-specific bisecting glycans present on numerous glycoproteins on the cancer cell surface that are also shed into circulation. We will be working to develop this humanized scFv antibody into novel therapeutics and diagnostics. In addition to this innovative biomarker development and validation, we have made the novel discovery that GnT-III expression and the presence of these unusual bisecting glycan structures are regulating the expansion of the side population or cancer stem cell population through the activation of the Notch signaling pathway. In this proposal we will be expanding this research to include collaborations with Dr. Robert Haltiwanger who has developed a novel mass spectrometry-based methods to map O-fucose, O-glucose, O-xylose and O-GlcNAc modifications on the Notch extracellular domain (ECD) to discover new glycan biomarkers for ovarian cancer. Studies to characterize the structures and functions of Notch glycosylation in ovarian cancer have never been done making these studies novel and of high impact. This proposal will also make significant innovations for the study of another class of glycoprotein biomarkers known as glycosylphosphatidylinositol (GPI) anchored proteins. We will utilize new mass spectrometry methods and instrumentation to fully characterize the structures of the GPI anchor glycan region for GPI anchored proteins identified in ovarian cancer. We will collaborate with Dr. Richard Cummings to produce a peptide GPI anchor array, a completely novel type of array never generated before. Covalently linking peptides carrying a GPI anchor to an array will allow the identification of cancer-specific carbohydrate binding proteins that bind to the GPI anchor.

Synopsis of Research and Network Resources

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

The study of glycosylation changes in ovarian cancer has advanced considerably from the 1960s when researchers found Schiff staining of mucin polysaccharides could distinguish non-malignant and malignant ovarian cancer tissues. We can now use high scan speed, high accuracy, mass spectrometry methods to fully characterize glycan structures and identify glycoproteins that are specific for ovarian cancer. Improved early detection of epithelial ovarian cancer (EOC), both primary and recurrent disease, is a high priority for improving the survival of patients. 

Laboratory-specific Studies to Meet the Research Needs

Our glycomics laboratory, located at the University of Arkansas for Medical Sciences, in collaboration with the Complex Carbohydrate Research Center (CCRC) and SRI International, is focused on the discovery, validation, and functional studies of novel glycan structures and glycoproteins for EOC.  We have discovered that ovarian cancer subtypes, endometrioid and serous, share similar changes in two forms of glycosylation, an N-linked glycosylation known as bisecting,  and elevated levels of glycosylphosphatidylinositol (GPI)-anchored proteins.  Therefore, the two most common subtypes of EOC, comprising about 90% of all cases, have glycosylation changes that can be used for detection and therapeutic targeting.

We have characterized the N-linked bisecting glycan structures from membrane proteins isolated from ovarian cancer tissues in collaboration with Dr. Michael Tiemeyer and Dr. Kazihuro Aoki at the CCRC.  Our data has identified a unique abnormal bisecting structure not found in normal cells.  Furthermore, this structure is not similar to any bisecting glycan structures reported in published studies from cells of human origin. This finding may allow for the development of antibodies that can target this structure for use in diagnostics and therapeutics.  We are collaborating with Dr. Nathalie Scholler at SRI International to produce humanized scFv probes to target this unusual glycan structure. 

The GPI transamidase (GPIT) is a multiple subunit enzyme that adds a  GPI anchor to the C-terminus of proteins in the membrane of the endoplasmic reticulum.  The mRNA levels of GPIT subunits are amplified in EOC suggesting that GPI anchored proteins are elevated.  We have developed a method to capture GPI-anchored proteins using a toxin purified from Clostridium septicum.  We are currently using this toxin to capture GPI-anchored proteins from the tissues and plasma of patients with EOC.  GPI-anchored proteins are localized on the cell surface and can be cleaved into circulation by the action of proteases and GPI-specific phosphatases.  We have examined the levels of GPI-anchored proteins in human plasma collected from patients with various malignant diseases and healthy non-diseased patients.  Our results indicate that there are very low levels of GPI-anchored proteins in plasma from patients with no malignant disease with higher levels in the plasma of patients with various cancers.  Furthermore, our mass spectrometry identification of GPI anchored proteins from breast, ovarian, and prostate cancer indicate that different GPI anchored proteins are expressed in each cancer type.  The unique expression of GPI anchored proteins for different cancer types will improve the specificity of cancer detection assays that may be developed to detect these cancers.

Resources and Reagents for Sharing

The translation of these glycan and glycoprotein biomarkers into sensitive detection assays is a high priority.  Currently, there are no methods with adequate sensitivity for population wide screening for ovarian cancer.  The antibodies that Dr. Scholler is developing will be a major advance to enable sensitive, specific detection of EOC.  These antibodies will be capable of in vivo and in vitro detection methods.  Cancer stem cells are known to play a role in the dissemination and recurrence of EOC.  The use of antibodies that can detect abnormal glycan structures on specific glycoproteins for EOC may allow for the early detection of disease recurrence.  We are actively pursuing studies to examine if the glycan structures that are detected on primary tumors also exist in recurrent disease. 

Our laboratory has developed the use of alpha toxin from C. septicum for the identification of GPI anchored proteins..   This affinity reagent is not currently available from commercial sources but we are able to routinely produce it for research purposes.

Public Health Implications and Advancing the Field of Glycobiology

Our studies are advancing the knowledge of glycosylation found in healthy and cancerous human tissues.  The identification of a very unique glycan structure in ovarian cancer raises interesting questions that we hope to answer such as (i) How is this structure formed in ovarian cancer?  The structure is under galactosylated and minimally branched indicating a possible disrupted processing in the Golgi.  Ovarian cancer cells may allow future studies to identify novel changes in Golgi transport or processing. (ii) What are the functional impacts of this glycan structure?  If we knockdown key enzymes participating in formation of this structure how does this affect the development and progression of ovarian cancer?  The prevalence of this glycan structure for the two most common histological subtypes of EOC suggests an importance of this glycan for tumor cell survival.   If we disrupt the formation of this structure will the efficacy of cytotoxic therapies be improved?

Cancer-Specific Relevance: Detection, Prevention and Treatment

As mentioned earlier, improved detection of EOC is needed for improved survival. We are investigating the novel glycan structures and glycoproteins for the two most prevalent types of EOC, which can be used for detection and therapeutic targeting. The subclasses of biomarkers we are investigating are novel in the sense that no other laboratories have identified these candidates in association with ovarian cancer.

Opportunities for Collaboration

Our laboratory is actively collaborating with several investigators on the identification of GPI anchored proteins for various human cancers using alpha toxin.   We welcome collaborations to exploit the use of alpha toxin and the glycoproteomic methods we have developed for the advancement of studies in cancer biology and cancer detection strategies.


  • GPI-specific PI-PLC enzyme for release of GPI anchored proteins
  • Glycoproteomic extraction and analysis methods
  • Thermo Ultimate 3000 UHPLC purification for glycopeptides
  • Lentiviral constructs for stable expression of GFP fusion GPI anchored proteins
  • Radical Fringe CRSPR/Cas9 KO ovarian cancer cells (and CRISPR plasmids that target the RFNG gene)
    • scFvC9 nanobody (recognizes tumor bisecting structures found in ovarian, glioblastoma, and lung cancer)
    • scFvC9-Fc monobody (a human scFv antibody fused to the human Fc domain for detection of tumor bisecting structures using ELISA or Western blot procedures)