Anand Mehta - US grants

DESCRIPTION (provided by applicant): MHC I mediated presentation of peptides derived from many glycoproteins requires translocation from the endoplasmic reticulum (ER), following synthesis, to cytosolic proteasomes for degradation. Immune mediated clearance of, and recovery from, hepatitis B virus (HBV) is thought to involve recognition of MHC I presented HBV glycoproteins by T lymphocytes. Failure to resolve HBV infection resulting in chronic disease is associated with an inadequate MHC I restricted lymphocyte response. This proposal will determine if the MHC I presentation of HBV envelope glycoproteins can be enhanced by production of variants that are degraded with enhanced efficiency by cytosolic proteasomes. Briefly, plasmids specifying mutant envelope proteins unable to enter the ER, and misfold, will be constructed and tested for enhanced degradation and MHC I presentation to human and woodchuck CTLs, in tissue culture. Promising constructs will then be introduced into naive and, if indicated, hepadnavirus chronically infected woodchucks and tested for MHC I presentation. The effect of delivery of plasmids into infected versus uninfected cells upon clearance of virus will be compared. This work will thus help in our understanding of the role proteasome degradation in MHC I presentation of HBV, as well as test the possibility that manipulation of the immune system by expression of mutant polypeptides is beneficial.

DESCRIPTION (provided by applicant): The hypothesis of the original grant was that fucosylation increases with the development of hepatocellular carcinoma (HCC) and that fucosylated glycoprotein(s) will make sensitive and specific markers of HCC. This hypothesis has been confirmed and in our analysis of over 1000 patient samples, we have clearly shown that fucosylated glycoproteins can make sensitive and specific markers of HCC, either alone or in combination with other markers. However, in our analysis, we have determined that in addition to core fucosylation there are many other changes that occur with liver disease. Some of these changes are cancer specific and can be used to complement our existing markers, while others can occur with just liver disease (inflammation) and lead to false positives. Thus in aim 1, we will develop novel and unique reagents that will dramatically improve our assays and continue our discovery efforts to find biomarkers that can be used clinically for the management of HCC. In aim 2 we will utilize our new lectins and continue our discovery efforts in an effort to find new biomarkers in biomarker negative populations that can complement our existing markers and lead to 100% sensitivity and 100% specificity. Finally in aim 3, we will test our lead markers in a NCI sponsored study comprising of over 350 cases of HCC and which we pre-qualified for. At the end of this 5 year period we will have validated our biomarkers and definitively proved our hypothesis that fucosylated proteins can make sensitive and specific markers of HC.

DESCRIPTION (provided by applicant): We, along with others, have shown increased levels of fucosylation with the development of hepatocellular carcinoma (HCC). In an effort to determine the origin of this increased fucosylation, we have performed N-linked glycan analysis of HCC tissue, the surrounding non tumor tissue, and compared this to tissue from a non-diseased adult liver. Surprisingly, no difference in the level of fucosylation was observed from the three donor groups, suggesting that the increased levels of fucosylation observed in the serum of those with HCC is not the result of increased synthesis of fucosylated proteins in the cancer tissue. In addition, the level of fucosylation observed in tissue was much higher than that observed on liver derived serum glycoproteins. On the other hand, increased levels of a tetra-antennary glycan were observed in the HCC tissue as compared to the surrounding tissue or to the non-diseased livers. Recent work by our collaborator has suggested that fucosylation controls the polarized secretion of glycoproteins and directs their secretion apically into the bil. Indeed, bile glycoforms are more heavily fucosylated than in serum and closely resemble those observed in the serum of HCC patients. It is our hypothesis that the over-expression of ¿ 1,6-N-acetylglucosaminyltransferase V (MGAT-5), which is responsible for the increased branching we have observed, leads to a loss of a tight blood-biliary barrier and the release of fucosylated proteins into the blood. This hypothesis, along with the use of this knowledge to find new biomarkers of HCC, will be tested in 3 aims. First we will examine the role of MGAT-5 in enforcing the polarized secretion of core fucosylated proteins. It is hypothesized that the overexpression of MGAT-5 that we observe in HCC tissue leads to a defect in the blood- biliary barrier and the aberrant release of fucosylated proteins into the blood. In specific aim 2 we will perform glycoproteomic analysis of bile for markers of HCC. The logic here is that all hepatocytes have the ability to secrete into the blood stream or into the bile capillaries, which eventually forms the bile ducts. Loss of hepatocyte polarity will lead to the aberrant appearance of fucosylated proteins in the blood. In specific Aim 3 we will examine the ability of our identifid biomarkers to differentiate patients with HCC from patients with liver cirrhosis in a cohort of 1500 patients. The ability of these markers to distinguish HCC from cirrhosis as compared to AFP and other potential biomarkers of HCC will be determined.

DESCRIPTION (provided by applicant): The hypothesis of the original grant was that fucosylation increases with the development of hepatocellular carcinoma (HCC) and that fucosylated glycoprotein(s) will make sensitive and specific markers of HCC. This hypothesis has been confirmed and in our analysis of over 1000 patient samples, we have clearly shown that fucosylated glycoproteins can make sensitive and specific markers of HCC, either alone or in combination with other markers. However, in our analysis, we have determined that in addition to core fucosylation there are many other changes that occur with liver disease. Some of these changes are cancer specific and can be used to complement our existing markers, while others can occur with just liver disease (inflammation) and lead to false positives. Thus in aim 1, we will develop novel and unique reagents that will dramatically improve our assays and continue our discovery efforts to find biomarkers that can be used clinically for the management of HCC. In aim 2 we will utilize our new lectins and continue our discovery efforts in an effort to find new biomarkers in biomarker negative populations that can complement our existing markers and lead to 100% sensitivity and 100% specificity. Finally in aim 3, we will test our lead markers in a NCI sponsored study comprising of over 350 cases of HCC and which we pre-qualified for. At the end of this 5 year period we will have validated our biomarkers and definitively proved our hypothesis that fucosylated proteins can make sensitive and specific markers of HC.

Abstract: Alterations in glycosylation have been associated with a number of diseases. Generally, these changes are observed through glycan analysis of complex protein mixtures or through the analysis of a few specific individual proteins. Indeed, in most cases, glycan analysis is done on released glycan pools without any connection to the protein carrier. When analysis is done with the protein carrier included, this generally requires large amounts of protein and is only feasible for a small number of proteins, thus structural glycan information is often limited. In this application we propose a novel method that will allow for the glycan analysis of 100?s if not 1000?s of individual proteins found in complex protein mixtures such as serum. This method will involve three steps to achieve this goal: (i) the capture of specific proteins using an antibody array, (ii) treatment of captured protein with highly active recombinant PNGase F, and (iii) the glycan analysis of the specific captured proteins on a spot by spot basis by MALDI Imaging Mass Spectrometry. In this R21 application, we will first optimize the detection of N-linked glycan on immuno-captured proteins using specific well-known proteins. Subsequently, we will perform N-linked glycan analysis of over 30 serum proteins that we believe can act as biomarkers of liver cancer in both control patient and patients with liver cancer. This number is also chosen as it will allow us to create a microarray of four 32 spot quadrants, which will be used for both analysis and controls for signal and protein information. In the R33 application we will expand this platform to include over 100 capture antibodies and us this method to determine glycan changes in the serum of patients with hepatocellular carcinoma (HCC). At this end of this R21 application, we will have validated the basic methodology that will allow for the complete structure N-linked glycan information to be obtained from 100?s to 1000?s of glycoproteins in a complex solution at one time.

The Mehta laboratory has identified a number of potential biomarkers of HCC. These biomarkers, which are fucosylated serum glycoproteins, can be used alone or in combination with other biomarkers and clinical factors in a diagnostic algorithm. The lead biomarker is fucosylated kininogen, which combines with alpha feto protein (AFP), age, gender, alkaline phosphatase (ALK) and Alanine transaminase (ALT) to greatly increase the detection of HCC as compared to the currently used markers. In collaboration with Dr. Goldman we have recently performed glycan analysis of human kininogen from patients with cirrhosis or patients with HCC in the background of cirrhosis. Increased levels of fucosylation were observed in patients with HCC as compared to patients with just cirrhosis. Through glycoproteomics we identified the specific sites of glycan modification in HCC as well. Surprisingly, the level of fucosylation on one specific site had substantially better discriminatory ability than lectin analysis of the whole protein or the application of this fucosylated kininogen algorithm. When comparing the methods on the exact same samples, the lectin-ELISA for fucosylated kininogen resulted in an AUROC of 0.65, and the diagnostic algorithm containing AFP, ALT, AST, age and gender increased this to 0.91. Amazingly, the fucosylation from one specific glycan site resulted in an AUROC of 0.99 on these exact sample samples. Clearly, our results highlight the benefit of looking at one specific site of glycan modification as opposed to the whole protein and suggest a new avenue for the development of better biomarkers of HCC. Currently the mass spectrometry based glycopeptide method is the only way to examine the glycosylation of one specific site of N-linked glycan. However, it is not a clinically or commercially viable method for use in large scale studies. Thus, we propose to create a new method for the analysis of glycopeptides. We refer to this method as a Lectin-Tryptic-ELISA. In this method, an antibody is made to a specific peptide (non glycosylated) that is formed from the tryptic cleavage of kininogen. This peptide is synthesized and used as an standard immunogen. The key point is that in human kininogen, the selected peptide contains the altered N-linked glycan of interest. For the Lectin-Tryptic-ELISA, the serum is treated with trypsin and allowed to go to completion. Next, the tryptic peptide (in serum this would be a glycopeptide) of interest is captured through a simple antibody-antigen reaction and the level of fucosylation detected using a lectin. A key requirement is knowledge of the site of fucosylation and the development of the antibody. In principle, the assay is very simple and will depend upon the ability to develop peptide specific antibodies. Currently, commercial antibody development utilizes peptides for antibody generation and thus, this aspect of the proposal can be achieved easily through outside vendors. The key thing will be to identify the specific peptides to target and refinement of the conditions that allow for peptide generation and assay performance. ! !

Abstract: In the Annual Report to the Nation on the Status of Cancer, mortality from Hepatocellular carcinoma (HCC) increased at an annual rate of 2.8% in men and 2.2% in women, making it the cancer with the greatest increase in mortality in the United States (USA) over the last 10 years. The occurrence of liver cancer is predicted to continue rising in the USA and will exceed 50,000 cases by 2021. The majority of HCC arises in the background of liver fibrosis and cirrhosis, usually associated with chronic viral infection (hepatitis B and hepatitis C virus) or nonalcoholic fatty liver disease/nonalcoholic steatohepatitis (NAFLD/NASH) associated with obesity. Our laboratory has shown alterations in both core and outer-arm fucosylation in HCC. These glycan modifications have promise as biomarkers of cancer and are actively being commercialized by a number of groups (including us). We have a developed a diagnostic panel that is comprised of clinical data along with one additional novel biomarker, fucosylated kininogen, that dramatically improves upon the detection of HCC, and in particular, the identification of those with early stage HCC. In an effort to identify the source of fucosylated serum glycoprotein, we have developed a novel method for tissue-based glycan analysis. In an analysis of 145 HCC tissue and 112 adjacent control or cirrhotic tissue control samples, we have identified two major changes in the N-linked glycan family that are associated with HCC. The first change observed was increased levels of fucosylation, a modification also often observed in serum of patients with HCC. However, ~50% of the tissue samples analyzed had no increase in fucose. Often these tumors without increased fucosylation had increases in tetra-antennary glycan. We hypothesize that the genetic heterogeneity of the tumor might have an impact upon the glycan heterogeneity in the tissue and serum. Consequently, the glycans may not only be used as biomarkers for the early detection of cancer but offer information into the specific genetics of the cancer. In this application, we will link the glycomic changes observed in both tissue and serum with the underlying genetic changes. As we will have matching tissue and serum, we will be able to determine if our current biomarker panel is capable of identifying fucose negative HCC. Lastly, we will utilize several proteomic and glycomic methods to identify biomarkers for cancer without increased levels of fucosylation.

Abstract The technologies available to rapidly and efficiently manipulate the expression of hundreds of glycan biosynthetic genes to study the functional role of complex glycans in growing cells have eclipsed the analytical capabilities to evaluate each cell phenotype in a comparative or quantitative manner. Current glycome profiling approaches require specialized plate-based sample handling resources, extensive processing and purification prior to analysis, and are expensive in regards to processing time and enzyme. These are barriers to both non-glycomic and glycomic researchers adopting large scale glycan analysis workflows applied to biofluids and cells. Our collaborative group has recently developed a streamlined antibody capture slide array approach to directly profile N-glycans of captured serum glycoproteins like IgG, a method requiring a few microliters of sample and simplified processing workflows that require no purification or sugar modifications prior to analysis. N-glycans are released from captured glycoproteins and directly analyzed by MALDI-TOF mass spectrometry. We propose to expand and adapt our slide array-based immune capture workflows to isolate immune cells directly from biofluids, and provide rapid analysis workflows of cultured cells. The goal in these assays is to develop rapid isolation workflows with minimal processing and direct glycan analysis, as described in three Specific Aims: SA1. Development of an on slide method for glycan analysis of immunoglobulin subtypes: SA 2. Development of Glyco-Cell Typer as applied to immune cell sub-types. SA 3. Analysis of cultured cells on slides for direct glycan measurements. Optimization of slide chemistry and processing will be emphasized, as well as conditions for metabolic isotope labeling and quantitative glycan analysis. The goal will be to provide validated workflows such that any research or core laboratory with a MALDI mass spectrometer will be able to perform routine glycan analysis on the most common types of samples.

This application proposes to translate our identification of a biomarker panel for hepatocellular carcinoma to commercial and clinical use. We have identified a specific post-translation modification that is altered in HCC. This alteration is a shift in the composition of N-linked glycans found on proteins made in and secreted by HCC cells. Many of the proteins containing this glycan alteration have been identified and patented. Our strong preliminary evidence has shown that some of these proteins, when combined with existing biomarkers and clinical factors, have excellent discriminatory ability between those with HCC and those with cirrhosis. Importantly, our biomarker efforts have focused on the detection of early stage lesions using our current biomarker panel, which is a simple logistic regression algorithm composed of fucosylated kininogen, alpha- fetoprotein (AFP), alkaline phosphatase (ALK), aspartate aminotransferase (AST), age and gender. Called the kininogen panel, it had an area under the receiver operator curve (AUROC) of 0.97 for the differentiation of early stage HCC from cirrhosis. In an independent external validation of this panel, in three independent cohorts this panel had an AUROC of 0.92 to 0.97 in the detection of early stage HCC. Thus our enthusiasm for this biomarker panel is high. To fully understand the clinical and commercial benefits of any biomarker, a large scale clinical validation study that is appropriate powered is essential. Hence, the two goals of this application are to validate the kininogen panel for the detection of early stage HCC at the time of HCC diagnosis (Aim 1) and determine when in time an individual becomes biomarker positive prior to developing HCC (Aim 2). At the completion of these two aims, Glycotest will have validated this biomarker panel and have confidence in its performance. Glycotest's strategy is focused on commercialization of its tests as Laboratory Developed Test (LDT) service products. These tests will be marketed to and ordered by specialists who care for the patient populations at risk for liver cancers and fibrosis-cirrhosis. LDTs are tests developed and run in a single CLIA lab regulated by the Center for Medicare & Medicaid Services (CMS) through the Clinical Laboratory Improvement Amendments (CLIA). The LDT business model has been used successfully by many other US companies that have commercialized tests as service products in multiple disease areas. The Company will establish an internal selling and marketing capability focused initially on major influential medical centers and key opinion leader sites as well as the west coast and east coast geographical areas, and will pursue a strategy for establishing Medicare reimbursement that it has developed in collaboration with external experts.