ROLE OF TYPE O-GLYCOSYLATION IN THE CYTOTOXIC ACTIVITY OF VARIOUS CHEMOTHERAPEUTIC AGENTS AS A FUNCTION OF

Transcription

1 ROLE OF TYPE O-GLYCOSYLATION IN THE CYTOTOXIC ACTIVITY OF VARIOUS CHEMOTHERAPEUTIC AGENTS AS A FUNCTION OF THEIR PHYSICOCHEMICAL PROPERTIES: EVALUATION OF MUCINOUS CANCER CELL LINES Master s Thesis Defense by Shweta A Raina Advisor: Dr. Robert B. Campbell, PhD. The Bouve Graduate School of Health Sciences In Partial Fulfillment of the Requirements for the Degree of Master of Science in Pharmaceutical Sciences with a specialization in Pharmaceutics and Drug Delivery NORTHEASTERN UNIVERSITY BOSTON, MASSACHUSETTS 20 th January, 2009.

2 PROJECT OVERVIEW Mucins are cell surface glycoproteins with protective functions. They are over-expressed in a variety of cancerous conditions. In addition to protection, lubrication and signal transduction functions they have long been implicated in the invasion, metastasis, poor prognosis and extremely poor survival of cancer patients. This exaggerated expression of mucin isoforms in tumor cells is accompanied by their aberrant O-glycosylation. Although O-glycosylation inhibitors have previously been predominantly used to study mucin structure, function and biosynthesis, it has recently been shown that pre-exposing human pancreatic cancer cells with O-glycosylation inhibitors significantly enhances the cytotoxic effects of 5-FU in pancreatic cancer cells. This has spurred speculation on whether the deviant O-glycosylation in tandem with the overexpression of mucins could play a role in tumor refractoriness to chemotherapy. In this study we explore the atypical mucin O-glycan in tumor cells as an emerging barrier to various chemotherapeutic agents varying as a function of their physicochemical properties. Chemotherapeutic agents belonging to several different classes of anti-neoplastics and possessing different physicochemical properties, such as molecular weight and hydrophilic and lipophilic character, have been employed. Cytotoxicity studies were carried out with these agents in the presence and absence of benzyl-α-galnac, an O-glycosylation inhibitor. A significant benefit in cytotoxic drug effect was observed when each one of these chemotherapeutic agents was used in conjunction with a non-toxic concentration of benzyl-α-galnac. Neither the physicochemical properties of the agents employed nor the differences in concentrations of agents used influence the extent to which chemotherapeutic agents benefited from inhibition of mucin O-glycosylation. It now appears that mucinous carcinomas may develop resistance to chemotherapy on account of the rigid aberrantly O-glycosylated mucin barrier which limits drug uptake into the target cell. This project looks also to determine whether the benefits in cytotoxicity following inhibition of mucin O- 2

3 glycosylation in pancreatic cancer cells extends to include mucin expressing cell lines derived from other organs and tissues. 3

4 20 th January pm. 344 CSC 4

5 TABLE OF CONTENTS Page Number TITLE PAGE AND DESCRIPTION 1 PROJECT OVERVIEW 2 TABLE OF CONTENTS 5 List of Tables 8 List of Figures 9 1. BACKGROUND AND SIGNIFICANCE Barriers to Chemotherapeutic Delivery in Tumors Pancreatic Cancer and Mucinous Adenocarcinomas Mucins and O-Glycosylation: An Emerging Barrier Mucins in Normal Physiology Mucin Function Mucins in Cancer Mucins in Cancer Therapy Benzyl-α-GalNAc: O-Glycosylation Inhibition and Mechanism Chemotherapeutic Agents: Mechanism and Physicochemical properties OBJECTIVES AND SPECIFIC AIMS Statement of the Problem Objective Hypothesis. 41 5

6 2.4. Specific Aims MATERIALS AND METHODS In-vitro Adenocarcinoma Models ad Cell Culture Materials Preparation of Benzyl-α-GalNAc and Drug Solutions Sulforhodamine-B Assay Determination of Cytotoxicity of Chemotherapeutic Agents Determination of Maximum Non-Toxic Concentration of Benzyl-α-GalNAc in in-vitro Adenocarcinoma Models Evaluation of Percent Benefit in Cytotoxic Activity Fluorescence Activated Cell Sorting Validation of O-glycosylation Inhibition Using FACS Analysis Statistical Analysis RESULTS AND CONCLUSIONS Light Microscopy of Mucinous Adenocarcinoma Cell Lines Determination of Cytotoxicity of Chemotherapeutic Agents in Capan Evaluation of Percent Benefit in Cytotoxic Activity of Chemotherapeutic Agents in the Presence and Absence of Benzyl-α-GalNAc in Capan Determination of Maximum Non-Toxic Concentration of O-Glycosylation Inhibitor in Breast and Colon in-vitro Adenocarcinoma Models Validation of O-glycosylation Inhibition in Breast and Colon in-vitro 6

7 Adenocarcinoma Models Determination of Cytotoxicity of 5-FU in Breast and Colon Adenocarcinoma in-vitro Models in the Presence and Absence of Benzyl-α-GalNAc DISCUSSION REFERENCES. 72 7

8 LIST OF TABLES Page Number TABLE 1 Hydrophilic Agents: Structure and Mechanism of Action 37 TABLE 2 Lipophilic Agents: Structure and Mechanism of Action 38 TABLE 3 Chemotherapeutic Agents and their Physicochemical 39 Properties TABLE 4 Cytotoxicity Values of Hydrophilic Drugs in the Presence 52 and Absence of Benzyl-α-GalNAc. TABLE 5 Cytotoxicity Values of Lipophilic Drugs in the Presence 55 and Absence of Benzyl-α-GalNAc. TABLE 6 Maximum Non-Toxic Concentration of Benzyl-α-GalNAc 59 used for Different Mucinous Adenocarcinoma Cell Lines. TABLE 7 Percent Benefit Observed in Different Mucinous 66 Adenocarcinoma Cell Lines 8

9 LIST OF FIGURES Page Number FIGURE1 Tumor Regions and Vasculature 12 FIGURE 2 Relative mrna Expression Levels of Different 15 Adenocarcinoma Cell Lines. FIGURE 3 Relative mrna Expression levels of Different Surgically 16 Resected Adenocarcinoma and Normal Tissues. FIGURE 4 Relative MUC1 Expression Levels of Different Cancer Cell 17 Lines. FIGURE 5 Biosynthesis and Secretion of Mucin 20 FIGURE 6 Major Core Structures in O-Glycans 21 FIGURE 7 Schematic of a Typical Mucin Glycoprotein. 22 FIGURE 8 Mucin Expression Profile in Different Organs 23 FIGURE 9 MUC Domains in Membrane Bound and 25 Secreted Mucin FIGURE 10 Schematic Representation of O-Glycosylation Inhibition 34 FIGURE 11 Schematic of Steps Involved in Sulforhodamine-B 45 Assay 9

10 FIGURE 12 Light Microscopy of Mucinous Adenocarcinoma Cell 49 Lines FIGURE 13 Plot of Percent Viability vs. Concentration of Various 53 Hydrophilic Chemotherapeutic Agents in the Presence and Absence of Benzyl-α- GalNAc FIGURE 14 Plot of Percent Viability vs. Concentration of Various 56 Lipophilic Chemotherapeutic Agents in the Presence and Absence of Benzyl-α- GalNAc. FIGURE 15 Determination of Maximum Non-toxic Concentration 58 of Benzyl-α-GalNAc FIGURE 16 Quantitative Validation of O-Glycosylation 61 FIGURE 17 Qualitative Validation of O-Glycosylation 64 FIGURE 18 Chemotherapeutic Benefit in Different Mucinous 67 Adenocarcinoma Cell Lines 10

11 1. BACKGOUND AND SIGNIFICANCE 1.1. Barriers to Chemotherapeutic Delivery in Tumors Overcoming barriers to delivery of chemotherapeutic agents to solid tumors is an important concern. The tumor physiology poses a plethora of barriers. Heterogeneous vasculature, tortuous blood flow, leakiness of vasculature and extravasation of large molecules from extravascular space all limit the drug delivery of cytotoxic drugs to tumor cells [1-4]. In addition, elevated IFP (interstitial fluid pressure) and the presence of several regions within a tumor mass such as semi-necrotic and necrotic regions also retard uptake of chemotherapeutic agents into tumor cells. These tumor cells are immortal and divide uncontrollably [1-4]. In order to sustain itself, the tumor mass initially relies on the passive diffusion of O 2 and nutrients for survival from surrounding normal vasculature and tissues. Eventually, on account of the high amounts of VEGF (vascular endothelial growth factors) and vasoactive factors such as bradykinins and nitric oxide produced by the tumor, blood vessels are recruited to the primary tumor site [1]. This is known as neo-vascularization. Neovascularization is required to maintain adequate tumor growth. These neo-vasculatures are highly permeable on account of vasodilation. As a result large proteins (and even drug molecules) can extravasate into the tumors. As the tumor mass continues to grow, several regions are formed within the tumor namely the avascular necrotic, seminecrotic, stabilized and advance tumor regions [1]. Various regions present in a tumor are shown in Figure 1. The necrotic regions have little or no vasculature and are highly hypoxic where as the advance front and stabilized regions have the highest density of vascularization. The necrotic regions greatly limit the access of cytotoxic agents to the tumor interiors. Thus even if the cytotoxic agents severely compromised the outer vascular tumor regions, the inner hypoxic regions remain unaffected and may contribute to sustaining the development of the tumor mass [4]. 11

12 FIGURE1: Tumor Regions and Vasculature Other contributors to the tumor barrier effect are EPR (enhanced permeability and retention) and IFP (Interstitial Fluid Pressure). The neo-vascular structures formed in order to support tumor cells are leaky and highly permeable. Moreover, the reduced lymphatic drainage in the tumor mass results in enhanced retention. One would presume that this would aid in retaining large drug molecules within the tumor. However, the net interstitial pressure within the tumor is in the outward direction (from the center to tumor periphery) and is responsible for the poor penetration of large drug molecules into the tumor[1-4]. Additional barriers to drug delivery are routes of administration and the disintegration and dissolution of the dosage form. Most anti-cancer drugs are administered systemically and directly encounter tumor physiological and cellular membrane and efflux proteins [4]. An emerging barrier to delivery is the aberrantly glycosylated mesh of secreted and membrane bound mucin molecules which form a rigid glycation barrier around tumor cells. The glycation mesh has been shown to prevent chemotherapeutic uptake [5]. These aberrant over-expressed mucin molecules have altered function in tumor cells. They enhance metastasis, aggressiveness and migration 12

13 to distant sites [6]. In addition, the mucin protein core which extends over 2 μm beyond the cell surface is suitably positioned to capture large macromolecules and charged moieties by binding them to charged groups on their oligosaccharide chains. The gel forming mucins create an intense aqueous environment just above the highly lipophilic cell membrane [6-9]. This may retard the entry of lipophilic molecules into the target cells. The hydrophilic moieties that do reach the cell membrane through this aqueous layer are then unable to permeate through the lipid membrane. Several strategies in cancer therapeutics are aimed at overcoming each of these barriers. The scope of this thesis lies is overcoming the mucin barrier and improving cytotoxicity of cancer chemotherapeutics in-vitro in a reduced mucin environment. We are specifically interested in discovering whether the physicochemical properties of chemotherapeutic agents will ultimately determine the extent to which the cells will benefit from the inhibition of mucin O-glycosylation. 13

14 1.2. Pancreatic Cancer and Mucinous Adenocarcinomas: The mucin mesh may pose a significant barrier to tumors which are highly mucinous [5]. The majority tumors which fall in this purview are mucinous adenocarcinomas. Adenocarcinomas are tumors of epithelial cells which line ducts or tubules of organs. Such epithelial cells are mucinous in normal physiology, but in cancerous conditions they exhibit altered mucin expression profile. This altered mucin expression profile is characterized by the up-regulation of some mucin isoforms and downregulation of others. The overall expression of mucins however is exaggerated [10]. Pancreatic adenocarcinomas account for about 95% of all pancreatic cancer cases and are regarded as highly mucinous [10,11]. Overall pancreatic cancer has the poorest 5 year survival rates amongst all cancers. According to the American Cancer Society 2008 statistics, pancreatic cancer diagnosed at advanced stages showed a 5-year survival rate of less than 1.7% for the period [11]. Available treatment options such as resection, radiation and chemotherapy may prolong survival slightly but seldom provide a cure. Several reasons can be attributed for this resistance to treatment. Firstly, symptoms of pancreatic cancer such as abdominal pain, jaundice etc. are commonly observed in GI disturbances and hepatitis conditions. Diagnosis therefore is made at advanced stages only. Advanced stage diagnosis offers no scope for resection as the tumors have metastasized to distant sites. Radiation and chemotherapy form the mainstay of treatment. Radiotherapy poses numerous side effects such as nausea, weakness, skin and eye infections and has been observed to even compromise the immune system [10]. Chemotherapy on the other hand encounters several limitations pertaining to drug delivery in tumors as discussed above in addition to the emerging mucin barrier [10]. Recent research has revealed that inhibiting mucin O-glycosylation in mucinous pancreatic adenocarcinomas enhances the chemotherapeutic action of 5-FU in-vitro [5]. A similar benefit may be observed for other highly mucinous adenocarcinomas such as breast, lung, colon and rectal adenocarcinomas. These mucinous cancer cells of the breast, colon and lung show differential patterns 14

15 of expression in terms of mucin iso-forms as well as levels of individual iso-forms expressed [12-15]. Dahiya et. al.,. have shown that the breast adenocarcinoma cell line ZR-75-1 exhibits higher concentrations of MUC1 mrna as compared to Capan-2 (pancreatic adenocarcinoma cells), Chago-K- 1(Lung Adenocarcinoma cells) and MGC-803 (gastric carcinoma cells). Additionally, all of the 4 cell lines showed moderate MUC2 mrna levels and none showed any MUC3 mrna levels as is evident from Figure 2 [12]. FIGURE 2: Relative mrna Expression levels of different Adenocarcinoma cell Lines [12]. Gastric Adenocarcinoma Pancreatic Adenocarcinoma Lung Adenocarcinoma Breast Adenocarcinoma In another study by Dahiya and Samuel et. al., surgically resected adenocarcinoma tissue belonging to different organs such as breast, lung, colon and stomach were evaluated for MUC 1, 2 and 3 mrna expressions levels and compared to expression levels in normal 15

16 FIGURE 3: Relative mrna Expression levels of different surgically resected Adenocarcinoma and Normal Tissues [13]. 16

17 Tissue [13]. It was found that all 4 adenocarcinoma tissues belonging to breast, lung, colon and stomach exhibited an exaggerated expression of MUC1 mrna levels as compared to normal tissue. MUC 2 and 3 mrna levels were exaggerated in stomach adenocarcinomas but only the colloid colon adenocarcinoma showed exaggerated expression of MUC2 and MUC3 when compared to normal tissue [13]. In the paper by Moore et. al., relative levels of MUC1 expression have been evaluated using a multi-modal MUC1 probe. The authors observed that both the breast cancer cell lines ZR-75-1 and BT- 20 show the highest levels of MUC1 expression followed by Capan-2, a pancreatic adenocarcinoma cell line. The colorectal and lung adenocarcinoma cell lines namely LS-174-T and Chago-K-11 respectively showed a more moderate MUC1 expression whereas brain glioblastoma cell line U-87MG shows the most negligible levels of MUC1 expression as is evident from Figure 4. [16] FIGURE 4: Relative MUC1 Expression Levels of Different Cancer Cell Lines [16]. Thus different adenocarcinomas exhibit differential pattern of mucin expression depending on their organ of origin when compared to normal cells or tissue. 17

18 1.3. Mucins and O-Glycosylation: An Emerging Barrier Mucins are cell surface glycoproteins which extend to over 2 microns from the cell surface. They are primarily protective in nature but also exhibit signal transduction functions. Mucin isoforms exhibit deviant O-glycosylation and differential expression profile varying as a function of tumor type. This shift in expression and O-glycosylation is responsible for changes in its inherent functions, and has been associated with an increase in metastatic growth potential, poor tumor response to chemotherapy and poor patient survival [6-9] Mucins in Normal Physiology: Structure, Synthesis and Secretion Mucins are extracellular glycoproteins. Secreted from epithelial cell surfaces, mucins are generally found in harsh environments such as acidic the PH in stomach, aerodigestive tract and so on. Mucins are complex oligosaccharides and have a relatively large molecular mass of greater than 10 6 daltons. This large molecular mass is on account of their ability to form highly ordered structures on polymerization [6-9, 17-19]. In general, mucin molecules have the following characteristic features: a) VNTR regions and b) Glycosylation a) VNTR Regions: VNTR or variable number tandem repeat regions are a hallmark of all mucins. These tandem repeat regions comprise of similar sequences rich in serine, threonine and proline residues. Serine and threonine residues hold the sites of O-glycosylation. The VNTR number and the specific sequence of repeats are different for different mucins within the same species and also vary between the different species. These repetitive sequences create the highly ordered structures characteristic of mucins [6-9,17-19]. 18

19 b) Glycosylation: Glycosylation is the chemical linkage formed between the hydroxyl group of serine and threonine residues and the mucin monomer. Mucin moieties show the presence of mainly 3 types of glycosylation namely O-glycosylation, N-glycosylation and S-glycosylation. O-glycosylation predominates with 50-80% oligosaccharides having O-linkages. O-glycosylation is enzyme dependent and requires glycosyltransferases, sulfotransferases, fucosyltransferases etc. O-glycosylation is crucial to mucin structure and function. A variety of chemical structures and compositions in mucins can be achieved by variations in O-glycosylation. Post-translational modifications such as sulfation in the O- linkages add further variety. N- and S- forms of glycosylation account for 20-30% and less than 10% of respectively. Thus glycosylation adds to the complex biochemical composition of mucins by linking mucin monomers and large oligosaccharide chains [6-9]. O-Glycosylation forms a major step in mucin biosynthesis, occurring from a handful to several hundred O-glycosylation sites in the VNTR region. The overall amount of glycosylation that occurs depens on the number of available serine and threonine residues. As shown in Figure 2, the activation of transcription factors upregulates the expression of MUC genes present in the nucleus. MUC mrna is then formed on the ribosomes and inserted into ER (endoplasmic reticulum). The nascent mucin polypeptide then traverses the Golgi. Here it is O-glycosylated in the presence of enzymes such as N- acetylgalactosyl transferases, galactosyl transferases, glucosyltransferases and fucosyl transferases. The primary moiety to attach to serine or threonine residue is the N-acetylgalactosamine moiety [20]. 19

20 FIGURE 5: Biosynthesis and Secretion of Mucin [20] The glycan chain is elongated by the addition of glucosamine, fucose, N-acetylneuraminic acid and N-acetylglucosamine in the presence of suitable enzymes. Four major core structures present in mucin O-glycans are shown in Figure 4. The O-glycan can be terminated by sialic acid, fucose, sulphates or even certain blood group determinants. Overall, the addition of N-acetylgalactosamines to mucin backbone confers linearity to the molecule and facilitates further addition of other sugar moieties [20]. 20

21 FIGURE 6: Major Core Structures in O-Glycans [20] Fully glycosylated mucins are regarded as mature mucins and are packaged into secretory vesicles and released at the cell surface in response to signals from the mucin secretagogue [20]. 21

22 FIGURE 7: Schematic of a Typical Mucin Glycoprotein [20]. The above figure is a schematic of a typical mucin glycoprotein with its protein backbone and attached sugar moieties from serine and threonine resides. It has two end terminals namely the NH 2 terminal shown in blue and the COOH terminal shown in green [20]. Mucins can be either membrane-bound or belong to the secreted variety. Soluble mucins arise either from cell secretion or from proteolytic cleavage of membrane-bound mucin. Apart from the proteases which may cleave and release extracellular portions of membrane-bound mucin, changes in ionic concentrations, PH and harsh conditions may also lead to cleavage, forming soluble mucins. In addition, membrane bound mucins undergo a typical post-translational modification in the golgi, forming two subunits which remain associated with each other. One subunit remains bound to the membrane, while the other subunit, called the juxtamembrane moeity, is expressed on the cell surface [6,8,9]. Depending on the composition of the VNTR regions, mucins can be classified into several isoforms. Each organ/ tissue in the body exhibits a unique mucin isoform expression profile as is apparent from Figure 8 [21]. 22

23 FIGURE 8: Mucin Expression profile in Different Organs [21] 23

24 Mucin Function: Mucins play a plethora of functions in normal physiology. Mucin molecules have a variety of domains and each domain has a specific function Functions of Specific Domains: Mucins show the presence of several different domains. Some domains are present on all mucins, while others are present in certain mucins depending on whether it is membrane-bound or secreted. The cysteine rich/ knot domain is rich in cysteine, is not heavily O-glycosylated and mediates dimerization of mucin molecules. The D domains are responsible for trimerization and oligomerization of secreted mucin core proteins. The cytoplasmic tail is present on the cytoplasmic side of the cell surface membrane and mediates association of phosphorylation sites present on it with cytoskeletal elements. The transmembrane domain spans the membrane and forms the integral membrane protein. Several other domains also exist such as Epidermal Growth Factor (EGF) domain, tandem repeat region and proteolytic cleavage site [6-9]. 24

25 FIGURE 9: MUC Domains in Membrane Bound and Secreted Mucin [21] 25

26 Functions of Mucins in Normal Cells: As previously noted, mucins are primarily protective in nature. The secreted mucins are in contact with the membrane bound mucins through covalent and non-covalent interactions. There also exists inter- and intra-molecular crosslinking. Thus the outer layer of mucin forms a physical barrier against the external environment and harsh conditions and maintains a controlled internal environment with respect to PH and ionic composition. The oligosaccharide composition on account of the tandem repeats confers stoichiometric power. Each tandem repeat is composed of a sequence of amino acids and the higher number of repeats, higher is the array of similar molecules limited to a local area. This confers stoichiometric power. Thus charged or neutral oligosaccharide chains can bind to certain ligands and molecules to a greater extent [6,7]. Sulfation and sialylation of oligosaccharide chains creates negatively charged chains which can bind to positively charged molecules leading to ion-exchange effects. Secreted mucins are highly hygroscopic and can significantly influence the degree of hydration above the cell. Mucins may also sequester cytokines and other factors on account of charge and stoichiometric power and maintain high local concentrations of these molecules within the glycation mesh. Secreted mucin forms a gel like mesh around the cells. It was previously proposed by our group that the mucin mesh creates a steric barrier limiting the entry of drug molecules into the cell. In support of our findings, it has been shown elsewhere that the transport of smaller molecules such as glucose and ethanol is also affected by mucin gel networks [46]. Mucins are also known to exhibit signal transduction effects probably by altering either the mucin layer or changes in ligand status of their extracellular domains in response to external stimuli [8,9]. Mucin molecules have been known to exhibit both adhesive and anti-adhesive properties. This is important in maintaining cell-cell adhesion [8,9]. 26

27 Mucins in Cancer: Mucins show an aberrant pattern of expression, O-glycosylation in cancer. This is on account of the dysreglation of the mucin core protein and enzymes responsible for glycosylation. This dysregulation of the mucin structure in cancer alters the role it normally plays [6,7,17, 22] Invasion and Metastasis: On account of the over expression and aberrant glycosylation of mucins in cancer, there is a loss in apical localization and an increase in the steric hindrance around cells. The aberrant glycosylation further adds to the exposure new epitopes. This exposes the core mucin peptides present on the tandem repeats. Both these irregularities in addition to the anti-adhesive properties of the negatively charged glycosidic chains cause tumor cells to detach. Decreased cell-cell interactions promote metastasis via invasion to nearby and distant sites [6,17, 22] Control of Local Microenvironment: Cancer cells, on account of anti-adhesive mucin properties, may migrate to distant sites. Once at distant sites, tumor cells may now use protective functions of mucin molecules to manipulate the local microenvironment and thus survive and proliferate in relatively harsh environments. Moreover, as previously stated, mucin molecules may capture and bind molecules such as interleukins, growth and regulatory factors secreted by cells. This prevents these molecules from being freely secreted. The capture of growth factors may aid tumor cells in growth whereas capture of interleukin or other factors may aid interaction between immune, stromal or inflammatory cells with tumor cells [6, 17, 22]. 27

28 Tumor Suppression: While altered properties of mucins in cancer aid tumor cell survival and metastasis, certain mucins may also play key roles in tumor suppression [17, 22]. In a study carried out in which MUC2 gene was inactivated, spontaneous GI tract tumors developed in greater than 65% of the mice [23] Mucins in Cancer Therapy: The altered mucin functions and properties form the basis of many immunotherapeutic and diagnostic applications. Diagnostic applications of mucins in cancer as well as therapeutic methods have yet to attain clinical success Immunotherapeutic Applications: Majority of the approaches in immunotherapeutic applications have been directed towards MUC1 and MUC4 molecules [26]. a) Serum Diagnostic Assay: An increase in serum levels of circulating mucins is seen in cancer. This is on account of the overexpression of mucin in tumor cells. Thus clinically elevated levels of mucin can be correlated with the presence of an underlying cancerous condition. Altered levels of MUC1, episialin, H23Ag, ETA (epithelial Tumor Antigen), PEM (Polymorphic epithelial Mucin), EMA (Epithelial Membrane Antigen), CA15-3 and MCA (Mammary Carcinoma Antigen) have not only been associated with poor patient survival but has been used to assess tumor burden in response to therapy [24-25]. Several other cancer markers that recognize epitopes or mucin type oligosaccharides are CA125, CA19-9, DuPan-2, SPAn-1, CA50 and CA242, CA195 and CAM43. These markers detect epitopes or oligosaccharide structures which may also be present on certain glycolipids. Thus none of 28

29 the assays have any proven use in early detection of cancer. As these markers do not take into consideration the core proteins to which these epitopes are attached [26]. Efforts need to be made to detect oligosaccharide epitopes as well as the core proteins on which these epitopes are present. This will pave the way to early detection and diagnosis of cancer. b) Immunotherapy and tumor vaccines: The basis of all immunotherapeutic applications is the fact that cancer patients have been shown to possess circulating antibodies against mucin. MUC reactive T helper cells have also been isolated from patients with breast, ovarian and pancreatic cancer. Thus this has presented the cancer biologist with the option of designing effective immunotherapeutic strategies [26,27,29]. A plethora of efforts have been made to enhance cellular immune response against MUC proteins for example conjugation of MUC1 with mannan, fusion of MUC1 positive cells with APC (Antigen Presenting Cells), and peptide based vaccines. Dendritic cell approach is also being used wherein a dendritic cell is fused with a tumor cell which is genetically modified to express a tumor antigen. The VNTR region of mucins is yet another target. The specific repeat sequences are highly immunogenic and have been used to develop a series of monoclonal antibodies. Thus the VNTR regions in mucin molecules can act as target sites [30]. The immunoconjugate approach has also been attempted. An immunoconjugate combines an antibody against mucin fused with either a chemotherapeutic agent or a radioisotope such as I 131. Bispecific antibodies have also been used which recognize both tumor antigens and cytotoxic T-cell antigens[17,22,26]. These immunoconjugates have also been used in diagnosis. The immunoconjugate approach is however hindered by the relatively large size of antibodies used. Poor tumor penetration of such large 29

30 conjugates can be overcome by reducing the size of the antibody. This can be achieved by only using portions of the antibody responsible for specificity. For example, proteolytic cleavage of the antibody to obtain Fab and Fab fragments are all that is required for specificity and targeting. However tumor cells act as poor antigen presenting sites and also contribute to an inhospitable tumor microenvironment. Both these factors limit the approach or delivery of immunotherapeutics into tumor cells [17, 22, 26] Gene Therapy: Antisense therapy has been used to decrease the expression of mucins in cancer. In a study carried out by Adachi et. al., antisense cdna was of N-acetyl-galactosaminyltransferase was transfected into an in-vitro model of human gastric cancer. This enzyme is responsible for the addition of N-acetyl-galactosamine, the first step in O-glycosylation of mucins. On transfection, they found that the gastric cancer cells were more susceptible to natural killer cells [28]. Other antisense approaches have also been attempted with regard to fucosyltransferases, glucosyltransferases etc. in an attempt to inhibit O-glycosylation [29]. MUC antisense cdna have also been used [30] Emerging Barrier to Drug Delivery: Mucin is an established barrier and is known to be protective in nature against the adverse environment present in the body. Mucin lines all most all cell surfaces, epithelial cells and tissue layers [9]. It is also believed to act as a barrier to infection in the GI tract. In cancerous conditions, on account of its upregulation and over expression, its protective role is enhanced. Under these conditions mucin forms a barrier against any adverse conditions thereby protecting cancer cells [34]. This was demonstrated by the study conducted by Kalra and Campbell in which the inhibition of mucin O-glycosylation disruption significantly increased the sensitivity of cancer cells to the cytotoxic effects 30

31 of chemotherapeutic agents [5]. It has been hypothesized that in addition to being a barrier against infection [34], cell surface mucins are also barriers limiting the uptake of chemotherapeutic drugs [5] Mucins and Role in Resistance to Drug Delivery Mucins have long been implicated in drug resistance in cancer. Although their barrier effect does seem to affect sensitivity to chemotherapeutics in cancer, mucins are also believed to play a role in resistance though no direct correlation has as yet been established between mucin expression and Pgp activity. P-glycoproteins (Pgp) and Multi-drug resistant proteins (MRP) are believed to be drug efflux pumps. These proteins actively efflux drug out of the cell and promote drug resistance in a variety of cancers. In a study carried out by Lampidis et. al., human melanoma cells which were MUC4+ showed increased resistance to doxorubicin, taxol and vinblastine and underwent apoptosis less frequently when compared to MUC4- cells. All of these drugs are recognized by Pgp and MRP-1. However, MUC4+ cells showed a decreased expression of Pgp and MRP-1 when compared to MUC4- cells [35]. Another study carried out by Siragusa et. al.,, has shown that the down regulation of MUC1 expression in thyroid cancer cells (by direct targeting with sirna) enhanced the sensitivity of cancer cells to the effects of chemotherapeutic agents [36]. Although some believe mucins play a role in drug resistance linked to Pgp, it may be that the drug resistance seen is simply on account of the aberrant mucin glycation mesh formed in cancerous conditions. 31

32 1.4.Benzyl-α-GalNAc : O-Glycosylation Inhibition and Mechanism Glycosylation is an enzyme dependent post-translational modification which links oligosaccharide moieties with proteins via O, -N and S linkages. Enzymes in the O-glycosylation process are galactosyltransferases, glucosyltransferases, N-acetylgalactosyl transferases, fucosyl transferases, sialyltransferases and N-acetylglucosyl transferases. Typically, O-glycosylation inhibitors competitively inhibit these enzymes and compromise the glycosylation process [37-40]. N-acetylgalactosamine is the first oligosaccharide moiety to attach to the mucin protein core via serine or threonine residues. Benzyl-α-GalNAc is an analog of N-acetylgalactosamine. The transferase family of enzymes mentioned above O-glycosylate of N-acetylgalactosamine to the serine and threonine residues on the mucin core protein. Benzyl-α-GalNAc inhibits O-glycosylation by acting as a substrate for the above mentioned transferases and competitively inhibits these enzymes from latching on oligosaccharide moieties the naked MUC core protein in the golgi. Furthermore, the oligosaccharides that are attached to serine and threonine residues may show decreased branching. Often this is associated with inhibition of terminal sialylation of the oligosaccharide chains [20]. 32

33 FIGURE 10: Schematic Representation of O-Glycosylation Inhibition Benzyl-α-GalNAc MUC core protein Oligosaccharide chains Serine residues Threonine residues Cell membrane The O-glycosylation inhibition process is concentration-dependent [38]. The efficiency with which galactosyl and glucosyltransferases carry out O-glycosylation may drop down to one third of its original depending on the concentration of benzyl-α-galnac used. Figure 4 above demonstrates the effects of benzyl-α-galnac on O-glycosylation in mucins. In a study by Sylavianne et. al Ht-29 cells treated with benzyl-α-galnac exhibited a 2.6 fold decrease in sialic acid content and a 1.5 fold decrease in galactosamine content. This further reinforces the fact that benzyl-α-galnac reduces oligosaccharide content of mucins formed [39]. In another study carried out by Gouyer et. al., after exposure of HT-29 cells with benzyl-α-galnac for a week, morphological changes such as swelling of cells, and accumulation of numerous cysts and 33

34 lyzosomes was documented [38]. These cysts were believed to be filled with incorrectly glycosylated mucins. Negligible or no such changes were observed in CaCo-2 cells. This observations point out to the cell specificity and differences in the way cellular machinery may respond to benzyl-α-galnac [37-40]. 34

35 1.5. Chemotherapeutic Agents: Mechanism and Physicochemical Properties Chemotherapeutic agents used to study the enhancement in cytotoxic drug action, in the presence and absence of benzyl-α-galnac, are agents commonly used in the clinical treatment of a wide variety of cancers. These agents were selected on the basis of their variability in terms of molecular weights, solubilities and mechanisms of action [41]. Following are the different mechanisms of actions to which chemotherapeutic agents belong: Anti-metabolites: Anti-metabolites mimic required body metabolites or NTP s (nucleotide triphostphates) and inhibit vital enzymes in DNA synthesis. DNA Intercalator: These agents intercalate with DNA and destabilize the DNA backbone. This leads to disruption in DNA strands and disrupts DNA replication eventually leading to RNA and protein synthesis inhibition. Alkylating Agents: Alkylating agents alkylate DNA and disrupts its structure leading to DNA replication and RNA and protein synthesis inhibition. Topoisomerase Inhibitors: Topoisomerases are enzymes responsible for unfolding DNA during replication. Thus inhibition of these enzymes inhibits DNA replication DNA synthesis. Microtubule Inhibitors: Microtubules are part of the spindle apparatus required during chromosomal separation. Inhibition of microtubule formation leads to inhibition in DNA replication DNA cell death. Table 1 and 2: Show the different chemotherapeutic agents, their molecular weights and classes. 35

36 [41, 42] TABLE 1: Hydrophilic Agents: Structure and Mechanism of Action Molecular Weight Name Structure Mechanism of Action FU Anti-metabolite 298 Cisplatin Alkylating Agents 300 Gemcitabine Anti-metabolite 371 Carboplatin Alkylating Agents 454 Methotrexate Anti-metabolite 580 Doxorubicin Hydrochloride Antibiotic Intercalator 909 Vinblastine Sulfate Microtubule Inhibitor 36

37 [41, 42] TABLE 2: Lipophilic Agents: Structure and Mechanism of Action Molecular Weight Name Structure Mechanism of Action 233 Lomustine Alkylating Agent 246 Busulfan Alkylating Agents 348 Camptothecin Topoisomerase Inhibitor 589 Etoposide Topoisomerase Inhibitor 657 Teniposide Topoisomerase Inhibitor 854 Paclitaxel Microtubule Inhibitors 37

38 Table 3: Chemotherapeutic Agents and their Physicochemical Properties [42-44] Chemotherap eutic Agents Molecular Weight Lipophilic Character Solubility Hydrophilic- Elecetronegative atoms Log P 5-FU 130 Hydrophilic 10g/L -ve (F) Lomustine 233 Lipophilic 0.01g/L - 3 Busulfan 246 Lipophilic 0.1g/L Cisplatin 298 Hydrophilic 1g/L -ve (Cl) Gemcitabine 300 Hydrophilic 15g/L -ve (Cl) -1.4 Camptothecin 348 Lipophilic 0.005g/L Carboplatin 371 Hydrophilic 14g/ml Methotrexate 454 Hydrophilic <0.1g/100ml +ve (Na) -2.2 Doxorubicin 580 Hydrophilic 10g/ L -ve (Cl) -0.5 Etoposide 589 Lipophilic 0.01g/L - 1 Teniposide 657 Lipophilic 0.1g/L Paclitaxel 854 Lipophilic g/l - 3 Vinblastine 909 Hydrophilic 10g/L -ve (SO 4 ) 3.9 Log P: Partition Co-efficicent octanol/water -ve/ +ve: Charge on the ionic moiety present in the salt form of the drug 38

39 2. OBJECTIVES AND SPECIFIC AIMS 2.1. Statement of the Problem Mucinous pancreatic adenocarcinomas have poor survival rates and poor response to chemotherapy. The over-expressed and aberrantly O-glycosylated mucins have been associated with high metastatic growth and progression of disease in pancreatic cancer. Inhibiting mucin O-glycosylation has been shown to benefit 5-FU uptake in pancreatic adenocarcinoma cells namely Capan-1 and HPAF-II. It has been hypothesized by Kalra and Campbell that the dense O-glycan mesh formed around mucinous cancer cells poses a significant barrier to intracellular drug uptake. Reducing this mucin mesh may improve chemotherapeutic uptake of several other clinically used anti-neoplastic agents in pancreatic cancer. In addition to pancreatic adenocarcinomas, several other carcinomas are also highly mucinous namely breast, colorectal and lung. Other mucinous cancer cells may also elicit a similar chemotherapeutic gain on inhibiting the mucin O-glycan mesh Objective The objective of this project is to evaluate the role of extracellular mucin in the growth inhibitory properties of twelve different chemotherapeutic agents varying in terms of their physicochemical properties using mucinous adenocarcinoma cells in vitro. 39

40 2.3. Hypothesis Irrespective of their inherent physicochemical properties, all chemotherapeutic agents will show a significant benefit in their cytotoxic drug effects against mucinous carcinoma cells following the inhibition of mucin O-Glycosylation Specific Aims Specific Aim 1# To evaluate the extent of benefit in cytotoxic drug activity of several different hydrophilic and lipophilic anti-cancer agents following inhibition O- glycosylation. Twelve different anti-cancer agents will be used and cytotoxic drug effects will be measured using the SRB assay in the presence and absence of the maximum non-toxic concentration of benzyl-α-galnac using Capan-1 cells (in-vitro pancreatic adenocarcinoma model). Percent benefit in cytotoxicity in Capan-1 cells was calculated for each chemotherapeutic agent as explained in the Material and Methods section to assess he relationship between physicochemical properties. Specific Aim #2 To determine the in-vitro maximum non-toxic concentration of benzyl- α- GalNAc which can be used in breast, lung, thyroid and colorectal cancer cells. In order to evaluate the benefit of inhibiting mucin O-glycosylation in other adenocarcinoma models, the maximum non-toxic concentration of benzyl- α- GalNAc will first be determined. 40

41 Specific Aim #3 To validate inhibition of O- glycosylation at the maximum non-toxic concentration of benzyl-α-galnac in lung, breast, thyroid and colorectal invitro adenocarcinoma models. The maximum non-toxic concentration of benzyl- α- GalNAc previously determined will be used to confirm inhibition of mucin O-glycosylation in-vitro for lung, colon and breast adenocarcinoma models using CD227 (anti-muc1) antibody for FACS analysis. Specific Aim#4 To estimate the benefit in cytotoxic activity following O- glycosylation inhibition in-vitro using 5-FU for treatment of lung, colon, thyroid and breast adenocarcinoma cells. 5-Fluorouracil was used to estimate if a similar benefit in cytotoxicity will be observed in other in-vitro adenocarcinoma models when benzyl-α-galnac is used in conjunction with chemotherapeutic agents. SRB assay was used to evaluate cytotoxicity. 41

42 3. MATERIALS AND METHODS 3.1. In-vitro Adenocarcinoma Models and Cell Culture The three main mucinous human in-vitro adenocarcinoma models were used in this study namely Capan-1 (pancreas), BT-20 and ZR-75-1 (breast) and COLO-205 and LS-174T (colorectal) (ATCC, Manassas, VA). Capan-1 cells were cultured and maintained in IMDM (Iscove s modified Dulbeco s medium), COLO-205 and ZR-75-1 cells in RPMI-1640 and LS-175T and BT-20 in EMEM (Eagle s minimum essential medium). All cell culture media was supplemented with 10% FBS (Fetal bovine serum) and cells were grown in a humidified 5% CO2 atmosphere Materials Cell culture media viz. RPMI, IMDM and EMEM and trypsin-ethylenediaminetetraacetic acid was purchased from ATCC (Manassas, VA). Sulforhodamine B, Benzyl-α-GalNAc and the following chemotherapeutic agents namely 5-FU, Cisplatin, Carboplatin, Gemcitabine, Doxorubicin hydrochloride, Methotrexate, Vinblastine sulphate, Lomustine, Busulfan, Camptothecin, Etoposide, Teniposide, Paclitaxel were purchased from Sigma Aldrich (St. Louis, MO). TCA (Trichloroacetic acid) and 1% acetic acid was purchased from Fischer Scientific (Fairlawn, NJ). FITC (fluorescein isothiocyanate) conjugated CD227 monoclonal antibody was purchased from BD Pharmingen (San Jose, CA). FLICA apoptosis kit was purchase from Immunohistochemistry Technologies (Bloomington, MN) Preparation of Benzyl-α-GalNAc and Drug Solutions Appropriate concentrations of Benzyl-α-GalNAc and various drugs were solubilized in suitable media before addition to cells seeded in well plates. To facilitate solubilization of lipophilic drugs, sonication and vortexing was used in addition to incubation in water-bath at 42 0 C 42

43 3.4. Sulforhodamine-B Assay Cytotoxicity in cells treated with Benzyl-α-GalNAc or drug was determined using the SRB Assay. Sulforhodamine-B is a dye mainly employed to determine cytotoxicity and proliferation in cell-based assays. The dye stains proteins in a PH dependent manner. Cells are first fixed with tri-chloroacetic acid. The dye binds to basic amino acids of proteins in cells. Washing with acetic acid provides mild acidic conditions under which excess dye can be washed off and the dye bound to proteins is more firmly bound. In order to quantify the intensity of bound dye, PBS is added which provides mild basic conditions. The dye is extracted in these mild basic conditions and solubilizes in PBS. The color intensity in ach well can then be quantified using a fluorescence intensity plate reader at 545/590 wavelength [45]. 43

44 FIGURE 11: Schematic of Steps involved in Sulforhodamine B Assay Seed plate + 24 hrs Incubation Add 0.4 mg/ml benzyl-α-galnac + Incubate 48 hrs PBS Wash + Add drug + Incubate 24 hrs PBS Wash ul TCA + 1 hr at 40 C (PBS wash removes dead cells and cell debris. Tri-carboxylic acid fixes cells) Distilled Water Wash ul 04.% w/v SRB + 30 mins dark (Sulforhodamine-B dye stains basic amino acids of cell protein) 1% w/v Acetic Acid wash + Air Dry plate (Mild acid condition of acetic acid helps dye bind to cell proteins) Add 1ml PBS (Provides mild basic condition to solubilize dye for quantification) Fluorescence Microplate Reader at 540/590 nm 44

45 Determination of Maximum Non-Toxic Concentration of Benzyl-α-GalNAc in in-vitro Adenocarcinoma Models Approximately 10,000 cells were seeded in a 48-well plate following an incubation period of 24hrs at 37 0 C. Cells were then treated with a range of concentrations of Benzyl-α-GalNAc and further incubated for a period of 48 hrs. Post 48 hrs of incubation cells were washed with PBS and fixed using TCA. Fixed cells were washed with distilled water and stained with Sulforhodamine B. Cells were then washed with acetic acid to remove excess dye and air dried. Dye taken up by cells was then solubilized by adding PBS to all wells and fluorescence was measured using a fluorescence microplate reader (Bio- Tek Instruments Inc., VT). Percent cell viability was determined using the following formula: Percent Cell Viability = Fluorescence Intensity of Treated cells X 100 Fluorescence Intensity of Untreated cells (Control) The maximum non-toxic concentration of benzyl-α-galnac for a given cell line was the concentration of benzyl-α-galnac at which 95% cell viability was observed Determination of Cytotoxicity of Chemotherapeutic Agents in the Presence and Absence of Benzyl-α-GalNAc in Capan-1 cells Cells were seeded in a 48-well plate at a concentration of 10 4 cells. The seeded plate was then incubated for a period of 24hrs at 37 0 C. Cells were then treated with 0.4 mg/ml of benzyl-α-galnac and further incubated for a period of 48 hrs. Post 48 hr incubation; cells were washed with PBS and treated with appropriate molar concentration of drug solution. After 24 hrs of incubation cells were 45

46 washed with PBS, fixed using TCA and SRB assay was carried out as discussed above. Cytoxicity was determined for 12 drugs (2 concentrations each) in the presence and absence of benzyl-α-galnac Evaluation of Percent Benefit in Cytotoxic Activity Percent benefit in cytotoxicity is the advantage seen when cells are treated with drug in the presence of benzyl-α-galnac versus drug alone. The percent benefit in cytotoxic drug activity was calculated using the difference in percent viability values for each concentration of drug in the presence and absence of benzyl-α-galnac was taken. To determine percent benefit in cytotoxicity for any drug the average of the percent benefit values of its two concentrations was considered Fluorescence Activated Cell Sorting (FACS) FACS analysis was used to determine O-Glycosylation inhibition and apoptosis if any in the presence and absence of benzyl-α-galnac Validation of O-glycosylation Inhibition Using FACS Analysis and Fluorescence Microscopy Approximately 2 X 10 4 cells were seeded in a 24 well plate and incubated for 24 hrs at 37 0 C. Cells were then treated with the appropriate maximum non-toxic concentration of benzyl-α-galnac and incubated for 48 hrs. Cells were then washed with PBS and treated with 4 ul of CD227 monoclonal antibody followed by a 24 hr incubation period. Cells were washed to remove unbound antibody, trypsinized, centrifuged to form a pellet and resuspended in PBS for FACS analysis. FACS was carried out using BD FACSCalibur (San Jose, CA). For fluorescence microscopy, cells were seeded in 24 well-plates with sterile cover slips, pre-treated with benzyl-α-galnac for the 48 hour incubation period, incubated for 24 hours with 4 ul of CD227 46

47 antibody. Coverslips were then washed with PBS and placed face down on glass slides with mounting media containing DAPI nuclear stain Statistical Analysis Statistical significance between experimental groups was evaluated with either parametric t-test or ANOVA where applicable using SPSS statistical package

48 4. RESULTS AND CONCLUSIONS 4.1. Light Microscopy of Mucinous Adenocarcinoma Cell Lines. Light Microscopy of the following cell lines was carried out as described in the Materials and Methods section at 20X and 40X magnifications: Capan-1, BT-20, ZR-75-1, Colo-205, LS-174T and LLC. Figure 12: Light Microscopic Images of Mucinous Adenocarcinoma Cells Capan-1 20X 40X 48

49 BT-20 20X 40X ZR X 40X 49

50 LLC 20X 40X 20X COLO X 50

51 4.2. Determination of Cytotoxicity of Chemotherapeutic Agents in Capan-1 Cytotoxicities of chemotherapeutic agents were measured in the presence and absence of benzyl-α-galnac and statistical analysis carried to determine the overall benefit of inhibiting mucin O- glycosylation in terms of cytotoxicity. Two concentrations were picked for each drug to evaluate concentration dependent effects if any. Following table lists hydrophilic chemotherapeutic agents along with their concentrations and average cytoxicities seen (calculated from arbitrary fluorescence units). TABLE 4: Cytotoxicity Values of Hydrophilic Drugs in the Presence and Absence of Benzyl-α-GalNAc. Chemotherapeutic Agent Molecular Weight Concentration x/ ml Cells + Free Drug Cells + Benzyl-α- GalNAc + Free Drug p Percent Benefit % 5-FU umol 85+/ /-2.1 * 22 Cisplatin nmol 78+/ /-2.2 ** nmol 59+/ /-2.3 ** 10 Gemcitabine nmol 76+/ /-7.1 * nmol 45+/-4 39+/-1.4 * 14 Carboplatin nmol 93+/ /-2.2 *** 10 1 umol 67+/ /-6.9 *** 32 Methotrexate nmol 98+/ /-1.44 * umol 94+/ /-1.42 * 5 Doxorubicin nmol 80+/ /-2.7 *** nmol 80+/ /-4.8 ** 17 Vinblastine nmol 96 +/ /-1.1 ** nmol 73+/ /-1.8 * 13 t- test * p < 0.05 ** p< 0.01 *** p<0.001 FIGURE 13 : Plot of Percent Viability vs. Concentration of various Hydrophilic 51

52 Chemotherapeutic Agents in the Presence and Absence of Benzylα-GalNAc. Cisplatin Gemcitabine Percent Viability * 50 nmol 250 nmol * Percent Viability * 50 nmol 250 nmol * Carboplatin Methotrexate Percent Viability * 250 nmol 1 umol * Percent Viability * * 350 nmol 1.5 umol Free Drug Free Drug + benzyl-α-galnac t- test * p < 0.05 ** p< 0.01 *** p<

53 Doxorubicin Vinblastine * Percent Viability * 50 nmol 350 nmol * Percent Viability nmol 250 nmol * For each of the hydrophilic drug molecules, a significant advantage was observed in cytotoxicity when cells were pre-treated with benzyl-α-galnac followed by treatment with drug. This significant advantage in cytotoxicity observed held true for the two different concentrations used for each of the drugs evaluated. 53

54 Following table lists lipophilic chemotherapeutic agents alongwith their concentrations and average cytotoxicities seen (calculated from arbitrary fluorescence units). TABLE 5: Cytotoxicity Values of Lipophilic Drugs in the Presence and Absence of Benzyl-α-GalNAc. Chemotherapeutic Agent Molecular Weight Concentration x/ ml Cells + Free Drug Cells + Benzyl-α- GalNAc + Free Drug p Percent Benefit % Lomustine nmol 86+/ /-2.0 * 13 1 umol 75+/ /-6.9 ** 22 Busulfan nmol 102+/ /-5.5 * 11 1 umol 85+/ /-6.5 ** 18 Camptothecin nmol 90+/ /-4.0 *** 33 1 umol 77+/ /-2.6 * 19 Etoposide nmol 94+/ /-2.8 * nmol 86+/ /-4.4 ** 11 Teniposide nmol 94+/ /-5.5 ** 11 1 umol 89+/-3 74+/-5.1 *** 17 Paclitaxel nmol 94+/ /-3.8 * 10 1 umol 80+/ /-4.7 *** 15 t- test * p < 0.05 ** p< 0.01 *** p<

55 FIGURE 14: Plot of Percent Viability vs. Concentration of various Lipophilic Chemotherapeutic Agents in the Presence and Absence of Benzyl-α- GalNAc. Lomustine Busulfan Percent Viability * 250 nmol 1 umol * Percent Viability * 250 nmol 1 umol * Camptothecin Etoposide Percent Viability * 350 nmol 1 umol * Percent Viability * 50 nmol 250 nmol * Free Drug Free Drug + benzyl-α-galnac t- test * p < 0.05 ** p< 0.01 *** p<

56 Teniposide Paclitaxel Percent Viability * 250 nmol 1 umol * Percent Viability * 50 nmol 1 umol * For each of the lipophilic drug molecules, a significant advantage was observed in terms of cytotoxicity when cells were pre-treated with benzyl-α-galnac followed by treatment with drug. This significant advantage in cytotoxicity was similar for the two different concentrations used for each of the drugs. 56

57 4.3. Determination of Maximum Non-Toxic Concentration of O- Glycosylation Inhibitor in Breast and Colorectal in-vitro Adenocarcinoma Models In order to evaluate if the benefits of inhibiting mucin O-glycosylation can extend to include cell lines other than pancreatic cancer, the maximum non-toxic concentrations of benzyl-α-galnac were determined in two breast cancer cell lines namely BT-20 and ZR-75-1, and two colorectal cancer cell line COLO-205 and LS-174T, a lung cancer cell line LLC and a thyroid cancer cell line CAL62. FIGURE 15: Determination of Maximum Non-toxic Concentration of Benzyl-α- GalNAc Percent Viability mg/ml 2mg/ml 1mg/ml 0.8mg/m l 0.6mg/m l 0.4mg/m l 0.2mg/m l 0.1mg/m l mg/ml ZR-75-1 LS-174T BT-20 LLC CAL62 From the above plot of percent cell viability we selected the maximum non-toxic concentrations of benzyl-α-galnac for each cell line. Maximum non-toxic concentration of benzyl-α-galnac was not toxic to > 95% of the cells. 57

58 TABLE 6: Maximum Non-Toxic Concentration of Benzyl-α-GalNAc used for Different Mucinous Adenocarcinoma Cell Lines. Cell line Organ Origin Concentration of Benzyl-α-GalNAc Percent Cell Viability BT-20 Breast Human 4 mg/ml 98.3+/-3.3 ZR-75-1 Breast Human 4 mg/ml /- 3.1 COLO-205 Colorectal Human 0.05 mg/ml Previously determined LS-174T Colorectal Human 2 mg/ml /- 3.7 LLC Lung Murine 0.8mg/ml / CAL62 Thyroid Human 1 mg/ml /

59 4.4. Quantitative Validation of O-Glycosylation Inhibition in Pancreatic, Breast and Colon in-vitro Adenocarcinoma Models In order to confirm or detreminee whether the maximum non-toxic concentration of benzyl-α- GalNAc previously determined for various adenocarcinoma models inhibits O-glycosylation, FACS analysis using CD227 antibody was carried out. CD227 is an anti-muc1 FITC labeled antibody against the core protein in MUC1. Thus in-vitro adenocarcinoma cells expressing MUC should stain positive with CD227, and demonstrate a shift to the right as compared to a control of untreated cells in the FACS analysis plot. Cells treated with benzyl-α- GalNAc followed by exposure to CD227 antibody should show a shift to the right relative to both the control of untreated cells as well as cells treated with CD227 alone. The figures below demonstrate FACS analysis of various Adenocarcinoma cells using the following scheme: Control#1 Control#2 Sample#1 Sample32 Untreated Cells Cells + benzyl-α-galnac Cells + CD227 Cells + benzyl-α-galnac + CD227 FACS analysis for BT-20 cells revealed a shift to the right for cells treated with CD227 and an even greater shift to the right for cells pre-treated with 4 mg/ml (maximum-non-toxic concentration) of benzyl-α-galnac followed by CD227 treatment. This confirms O-glycosylation inhibition using 4mg/ml of benzyl-α-galnac. Similar shifts were observed in ZR-75-1 cells using 4mg/ml (maximum non-toxic concentration) of benzyl-α-galnac. Thus the 4mg/ml concentration of benzyl-α-galnac inhibits O-glycosylation in ZR cells. 59

60 FIGURE 16.1: Validation of O-Glycosylation in BT-20 cells Breast Adenocarcinoma Model: BT-20 Cells FIGURE 16.2: Validation of O-Glycosylation in ZR-75-1 cells Breast Adenocarcinoma Model: ZR-75-1 Cells 60

61 FACS analysis for COLO-205 and LS-174T did not reveal a shift to the right for either group i.e. cells treated with CD227 or cells pre-treated with benzyl-α-galnac (0.05 mg/ml and 2 mg/ml respectively) followed by CD227 treatment. This can lead to two conclusions: a) The maximum nontoxic concentration used did not inhibit O-glycosylation or b) Since colon carcinoma cells predominantly express MUC2 versus MUC1, the anti- MUC1 antibody CD227 may not have bound to any surface mucin. The CD227 antibody is against the core mucin protein in MUC1. MUC protein cores are the VNTR regions which are highly specific for different isoforms and thus vary from one mucin isofrom to the next. It is therefore possible probable that CD227 showed no shift in COLO-205 and LS-174T cell lines on account of negligible expression levels of MUC1. Furthermore, cytotoxicity studies using and the maximum non-toxic concentrations of benzyl-α-galnac and 5-FU against the growth of LS-174T and COLO-205 cell lines revealed an advantage in chemotherapeutic activity (results outlined in section 4.5). Exposing LS-174T and COLO-205 to the o-glycosylation inhibitor probably altered the O-glycosylation profile of other mucin varieties (eg. MUC2) thus resulting in the improved cytotoxic drug effect (section 4.5). [19] 61

62 FIGURE 16.3: Validation of O-Glycosylation in COLO-205 cells Colorectal Adenocarcinoma Model: COLO-205 Cells FIGURE 16.4: Validation of O-Glycosylation in LS-174T cells Colorectal Adenocarcinoma Model: LS-174T Cells 62

63 4.5. Qualitative Validation of O-Glycosylation Inhibition in Breast and Colon in-vitro To qualitatively evaluate the inhibition of O-Glycosylation in adenocarcinoma models, respective maximum non-toxic concentrations of benzyl-α-galnac were used with FITC-labeled CD-227 antibody. DAPI was used for nuclear staining. Cells were exposed to DAPI and CD-227 in the presence and absence of benzyl-α-galnac. Fluorescence microscopy was carried out to visualize fluorescence of various cells in the blue and green channel. FIGURE 17: Qualitative Validation of O-Glycosylation 63