| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Research Articles: Therapeutics, Targets, and Development
The glycotope-specific RAV12 monoclonal antibody induces oncosis in vitro and has antitumor activity against gastrointestinal adenocarcinoma tumor xenografts in vivo
Raven Biotechnologies, Inc., South San Francisco, California
Requests for reprints: Deryk Loo, Raven Biotechnologies, Inc., 1140 Veterans Boulevard, South San Francisco, CA 94080. Phone: 650-624-2636; Fax: 650-553-9169. E-mail: dloo{at}ravenbio.com
Abstract
RAV12 is a chimeric antibody that recognizes an N-linked carbohydrate antigen (RAAG12) strongly expressed on multiple solid organ cancers. More than 90% of tumors of colorectal, gastric, and pancreatic origin express RAAG12, and a majority of these tumors exhibit uniform RAAG12 expression. RAV12 exhibits potent cytotoxic activity in vitro against COLO 205 colon tumor cells via an oncotic cell death mechanism. RAV12-treated COLO 205 cells undergo morphologic changes consistent with oncosis, including cytoskeletal rearrangement, rapid plasma membrane swelling, and cell lysis. RAV12 inhibited the growth of RAAG12-expressing gastrointestinal tumor xenografts in athymic mice. In the case of SNU-16 tumor cells, twice weekly treatment of established s.c. tumors with 10 mg/kg RAV12 caused a 
70% suppression of tumor growth at the end of the study. This preclinical data has led to the initiation of a phase I/IIA clinical study of RAV12 in patients with metastatic or recurrent adenocarcinoma. [Mol Cancer Ther 2007;6(3):856–65]
Introduction
The clinical success of monoclonal antibody therapeutics, such as rituximab, trastuzumab, cetuximab, and bevacizumab, validates the use of monoclonal antibodies as antitumor agents. Antibody therapeutics can confer antitumor effects via multiple mechanisms. Antibodies can deliver chemotherapeutic agents; they can target and activate host cell effector functions; and in some instances, they can bind and elicit antiproliferative or cytotoxic activities directly upon target cells. Because of the genetic variability of tumors, antibodies with multiple effector mechanisms may be needed to achieve maximal antitumor effects. Although eight monoclonal antibodies are marketed for use in treating cancer, there continues to be a great need for potent tumor selective therapeutic antibodies with unique antitumor mechanisms.
We are using a cell-based immunization approach to generate therapeutic antibodies against surface expressed targets found on human fetal progenitor stem cell lines. The progenitor stem cell lines have the capacity to provide a rich source of novel oncofetal antigens (1–3). Many of these developmentally regulated oncofetal antigens may be re-expressed on rapidly growing cancer cells. The stem cell lines have been used as immunogens to generate panels of monoclonal antibodies directed against cell surface proteins expressed during specific stages of fetal development and overexpressed by human cancers. Using this approach, we identified a monoclonal antibody (KID3) as a candidate therapeutic agent. Initial studies showed KID3 bound to the cell surface of multiple solid organ adenocarcinomas, possessed cytotoxic activity in vitro, and exhibited antitumor activity in tumor xenograft models. Based on results from these studies, the chimeric monoclonal antibody RAV12 (IgG1 subclass) was generated for preclinical/clinical development.
Herein, we describe the biological properties of RAV12 and the parental murine KID3 monoclonal antibody. RAV12 recognizes a primate-restricted N-linked carbohydrate epitope, designated RAAG12, expressed on a number of human carcinomas. RAV12 exhibits cytotoxic activity in vitro against the RAAG12-expressing human colon tumor cell line COLO 205. Cells undergoing RAV12-induced cell death exhibit characteristics consistent with oncotic cell death (also called programmed cell necrosis; ref. 4), including sodium-dependent rapid cellular swelling and lysis. RAV12 exhibits potent antitumor activity against human colon, gastric, and pancreatic tumor xenografts in athymic mice. A phase I/IIA clinical study of RAV12 in patients with metastatic or recurrent adenocarcinoma has been initiated.
Materials and Methods
Cell Lines
The tumor cell lines COLO 201, COLO 205, SNU-16, A549, and SU.86.86 were obtained from the American Type Culture Collection (Manassas, VA). All cell lines were maintained in F12/DMEM (Invitrogen, Carlsbad, CA), supplemented with 10% fetal bovine serum (FBS; Hyclone, Logan, UT) in a humidified atmosphere of 95% air and 5% CO2 at 37°C.
Animals
Female (for s.c. xenograft studies) or male (for subrenal capsule xenograft studies) athymic mice (CRL.nu/nu), 6 to 8 weeks of age, were purchased from Charles River Laboratories (Wilmington, MA). Mice were housed in specific pathogen-free conditions in accordance with U.S. Department of Agriculture guidelines. All studies were carried out under approved institutional experimental animal care and use protocols. Animals were provided pelleted food and water ad libitum.
Antibodies
The anti-Lewisa (Lea) antibody PR4D2 was obtained from Chemicon International (Temecula, CA). Antibodies to Leb, Lex, and Ley were obtained from Calbiochem (San Diego, CA). Anti-Ca19-9 was obtained from Vector Laboratories (Burlingame, CA). Anti-Ca50 and anti-CA242 were obtained from Lab Vision (Ferment, CA). Mouse IgG was obtained from Jackson Immunochemicals (West Grove, PA).
Glyco-Analogues
The glyco-analogues and bovine serum albumin (BSA)–conjugated glyco-analogues were obtained from Dextra Labs (Reading, United Kingdom).
Generation of Murine KID3 Monoclonal Antibody and Chimeric RAV12 Antibody
The murine KID3 monoclonal antibody was generated by immunizing BALB/c mice in the foot pad with a mixture of intact, viable, human fetal kidney progenitor cells and Ribi adjuvant (Corixa, Hamilton, MT) over a period of 3 months. Antibody titer was monitored by flow cytometry using the human fetal kidney progenitor cells. Lymph nodes from mice showing sufficient titers were harvested, and isolated lymphocytes were fused with the mouse myeloma line X63-Ag8.653 (American Type Culture Collection) using standard methodologies (5).
To generate chimeric RAV12, cDNAs encoding the variable heavy and light domains of KID3 were cloned using standard recombinant DNA methodologies (6) and ligated into expression vectors containing human IgG1 and 
constant domains. RAV12 was expressed in a genetically engineered Chinese hamster ovary cell line (CRL-9096; American Type Culture Collection) using a proprietary serum-free culture medium. The protein was purified by multiple chromatography steps and available as a clear, sterile liquid.
Purification of RAAG12-Bearing Proteins from COLO 205 Cells
RAV12-reactive antigens were purified from COLO 205 cell lysate by passing the cell lysate over a KID3-conjugated Sepharose affinity resin column. KID3 was conjugated to N-hydroxysuccinimide–activated Sepharose 4 fast flow (Amersham Biosciences, Piscataway, NJ) according to the manufacturer's protocol. Econo-prep columns (Bio-Rad, Hercules, CA) packed with 1 mL KID3-conjugated Sepharose were pre-eluted with 0.1 mol/L glycine HCl (pH 2.8; Sigma-Aldrich, St. Louis, MO) followed by 0.1 mol/L ethanolamine (pH 11.2; Sigma-Aldrich). Cell lysates were passed over the column; the column was washed extensively with lysis buffer and eluted with 0.1 mol/L ethanolamine (pH 11.2) followed by 0.1 mol/L glycine HCl (pH 2.8). Fractions were collected and neutralized with 1 mol/L Tris (pH 7.6; final concentration of 0.1 mol/L). Purified samples were precipitated with methanol/chloroform and resuspended in buffer for subsequent analysis.
Deglycosylation of RAAG12-Bearing Proteins
Purified protein samples were deglycosylated using the Enzymatic Deglycosylation kit and Deglycosylation Plus kit (ProZyme, San Raphael, CA) according to the manufacturer's protocol. Standard digestion period was 3 h at 37°C. Additional exoglycosidases or endoglycosidases were obtained from ProZyme. Deglycosylated samples were analyzed by SDS-PAGE and Western blotting.
Western Blot Analysis
RAAG12-bearing proteins purified from COLO 205 cell lysates were reduced with 2 mmol/L DTT (Pierce Biotechnology, Inc., Rockford, IL) and alkylated with 10 mmol/L iodoacetamide (Sigma-Aldrich), subjected to SDS-PAGE using Invitrogen/NOVEX 4% to 20% Tris-Glycine gels (Invitrogen, Carlsbad, CA), and transferred to nitrocellulose membranes (Invitrogen) according to manufacturer's protocol. Membranes were blocked with 5% nonfat dry milk in Tween-TBS [0.1% Tween 20 in 100 mmol/L Tris-Cl (pH 7.5), 0.9% NaCl] for 30 min, probed with KID3 at 2 µg/mL in blocking buffer for 1 h, detected using an alkaline phosphatase–conjugated anti-mouse secondary antibody for 30 min (Jackson ImmunoResearch, West Grove, PA), and visualized using 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium substrate according to the manufacturer's protocol (Pierce Biotechnology). KID3 was used for the Western blot studies due to reagent availability. Subsequent studies with chimeric RAV12 showed equivalence to KID3.
Binding/Competition ELISA
The BSA-conjugated glyco-analogue lacto-N-fucopentose II (LNFPII; 2 µg/mL) was immobilized onto NUNC Maxisorb 96-well plates (Rochester, NY) for 2 h at room temperature in HBSS. Plates were blocked for 30 min at room temperature with blocking buffer (5% BSA in HBSS). For direct antibody binding assays, 2 µg/mL of antibody was incubated on the plate for 1 h at room temperature. For glyco-analogue competition assays, 2 µg/mL KID3 or PR4D2 antibody was incubated with 100 µg/mL glyco-analogues at 1:10 final volume for 30 min at 37°C, diluted to final volume, and incubated on the plate for 1 h at room temperature. Plates were washed with HBSS; secondary horseradish peroxidase–conjugated antibody was added and incubated for 30 min at room temperature (horseradish peroxidase–conjugated donkey anti-mouse IgG H+L from Jackson ImmunoResearch was used at 1:1,000 dilution). Plates were washed with HBSS and developed with 3, 3', 5, 5'-tetramethylbenzidine substrate according to the manufacturer's protocol (KPL Laboratories, Inc., Gaithersburg, MD). KID3 was used for these ELISA studies due to technical constraints. Subsequent studies with chimeric RAV12 showed equivalence to KID3.
Tissue Immunohistochemistry
For the cancer and limited normal tissue surveys, sections were cut from frozen tumors (8 µm), formalin-fixed, paraffin-embedded (FFPE) tumors or normal tissues (4–5 µm), or obtained as precut FFPE cancer tissue microarray slides. Frozen tumors were obtained from Ardais (Lexington, MA), Phenopath (Seattle, WA), Cureline (South San Francisco, CA), Oncotech (Tustin, CA), and D. Mangham (University of Birmingham, Birmingham, England). Normal tissues were obtained from Asterand (Detroit, WI). FFPE tumor specimens were obtained from Cureline, Asterand, and D. Mangham. FFPE cancer tissue microarray slides were obtained from Biogenex (San Ramon, CA.), Imgenex (San Diego, CA), Petagen (Seoul, South Korea), and Asterand. Tumor tissue sections were stained using the VectaStain Elite ABC kit (Vector Laboratories) with 5 µg/mL biotinylated RAV12 according to the manufacturer's protocol with the following modifications. Antigen retrieval of FFPE sections was done by incubating slides in citric buffer (pH 6.0) in the DC2002 Decloaking Chamber Pro (Biocare Medical, Walnut Creek, CA) at 125°C for 30 s. Endogenous biotin was blocked by incubation for 30 min in 5% goat serum containing 0.1 mg/mL avadin followed by incubation in 30 µg/mL d-biotin blocking solution for 30 min. Endogenous peroxidase was quenched by incubation with 3% hydrogen peroxide/methanol solution for 10 min. Frozen sections were fixed with a 50:50 mixture of acetone/ethanol before staining.
For the expanded normal tissue Good Laboratory Practice survey (Pathology Associates Division, Charles River Laboratory), biotinylated RAV12 (2–50 µg/mL) was applied to frozen human sections from autopsy or biopsy samples. Samples were stained as described above for the tumor tissue sections with the following modifications. Endogenous peroxidase was quenched by incubation of the slides with glucose oxidase (1 unit/mL) and sodium azide (1 mmol/L) for 1 h, and nonspecific protein binding was reduced by incubation with protein blocking solution [PBS (pH 7.2) containing 5% human 
-globulins, 5 mg/mL salmon sperm DNA, 0.5% casein, 1% BSA, and 1 mg/mL heat aggregated human IgG] for 1 h.
In vitro Alamar Blue Assay
COLO 205 cells were plated at 500 per well in 96-well plates in F12/DMEM containing 10% FBS. An equal volume of test antibody (diluted in F12/DMEM containing 10% FBS) was added to the wells, then the plates were incubated at 37°C for 4 days. Cell viability was determined using the Alamar Blue assay according to the manufacturer's protocol (Trek Diagnostics, Cleveland, OH).
Video Microscopy
COLO 205 cells were released from flasks with 10 mmol/L EDTA in PBS. A T25 flask (BD Falcon, Franklin Lakes, NJ) was seeded with 10,000 cells in F12/DMEM containing 10% FBS. Antibodies were added at a final concentration of 20 µg/mL immediately following plating, and the flask was placed on the stage of a Nikon Diaphot phase-contrast microscope housed within a 95% air/5% CO2/37°C controlled chamber. Images were captured with an Optronics DEI470 CCD video camera (Goleta, CA) and recorded on tape with a Panasonic High Density Time Lapse Video Cassette Recorder AG6790 (Secaucus, NJ).
Cell Immunostaining
COLO 205 cells cultured in chamber slides in the presence of RAV12 antibody were washed with HBSS and fixed for 10 min at 4°C with 4% (w/v) paraformaldehyde/PBS (Sigma, St. Louis, MO). The fixed cells were washed with HBSS and blocked/permeabilized for 10 min at 4°C with HBSS containing 1% BSA and 0.5% Triton X-100. The permeabilized cells were then washed with blocking buffer (HBSS containing 1% BSA) and blocked for 1 h at 4°C. Blocked cells were stained with antibodies (2 µg/mL) for 1 h at room temperature, washed with blocking buffer, then incubated with Alexa Flour–conjugated secondary antibodies (Invitrogen) for 30 min at room temperature in the dark. Stained cells were washed with blocking buffer followed by HBSS. Coverslips were mounted using Vectashield Hardmount mounting media with propidium iodide (Vector Laboratories). The slides were visualized using a Nikon Eclipse E-800 microscope.
Subrenal Capsule Xenograft Protocol
Human tumor-derived cells cultured in vitro were harvested with 10 mmol/L EDTA in PBS and resuspended in F12/DMEM containing 10% FBS. Cells (500,000 per implant) were resuspended in a 50-µL button of rat tail type I collagen and cultured in vitro overnight. The tumor cell–containing collagen buttons were placed under the kidney capsule of male athymic mice through a paralumbar surgical approach as previously described (7). Following 2 to 8 days of tumor cell establishment, monoclonal antibodies were administered i.p. thrice weekly for up to 2 weeks. Two days following the final dose, animals were sacrificed, and human tumors were recovered from the kidney capsules and photographed, then DNA was extracted and human DNA was quantified by the Taqman quantitative human DNA assay.
Taqman Quantitative Human DNA Assay
Human tumors were recovered from mouse kidney capsules, and total genomic DNA was isolated using the Promega Wizard SV Genomic DNA Purification kit according to the manufacturer's protocol (Promega, Madison, WI). The amount of human tumor DNA extracted from the subrenal capsule tumors was quantified by quantitative PCR using human-specific RPL19 gene primers and probe (ref. 8; forward primer, 5'-CTTTCCTTTCGCTGCTGC; reverse primer, 5'-TAACCAGACCTTCTTCTT; probe, 5'-CCGCAGCCATGAGTATGCTC; Bioresearch Technologies, Inc., Novato, CA). Human DNA was amplified in a reaction containing 4 mmol/L MgCl2, 200 µmol/L deoxynucleotide triphosphates, 250 nmol/L primers, 150 nmol/L probe, 5 units BstXI, and 1.25 units Taq Gold in a Applied Biosystems SDS 7000 machine according to the manufacturer's protocol (Applied Biosystems, Foster City, CA). The Ct values from experimental conditions were compared with the generated standard curve (human DNA input versus Ct value) to determine the quantity of human DNA present in each sample. The R2 values for the standard curve were >0.95; all experimental Ct values were within the linear portion of the standard curve. Statistical analysis by ANOVA and Student's two-tailed t test were used to compare treatment groups.
Subcutaneous Xenograft Protocol
Tumor cells grown in vitro were harvested with Trypsin-EDTA (Invitrogen), washed 1x with PBS, and resuspended in Matrigel (BD Biosciences, Bedford, MA) to a final concentration of 107 per mL. Six- to 8-week-old female athymic mice were injected s.c. in the suprascapular region with 0.5 mL of the cell suspension. Animals were dosed twice weekly with antibodies diluted in PBS via i.p. or i.v. injection beginning on the day of implant or following a period of tumor establishment. Tumor volume was measured twice weekly using the formula: length x width x height x 0.5. The percent reduction in tumor growth was calculated using the formula: [(control group average volume – treatment group average volume) / control group average volume] x 100, at each time point reported. Statistical analysis by ANOVA and Student's two-tailed t test were used to compare treatment groups.
Results
RAAG12 Is a N-Linked Carbohydrate Epitope
Western blot analysis showed the RAAG12 epitope was present on multiple membrane-associated proteins over a molecular weight range from 25 to >250 kDa (data not shown). Digestion of RAAG12-bearing proteins purified from COLO 205 cells with a range of enzymes showed KID3 reactivity decreased following treatment with N-glycanase (PNGaseF; Fig. 1A
). Treatment with N-glycanase and fucosidase (
1-3,4 fucosidase or 1-2,4,6 fucosidase) further decreased KID3 reactivity. Treatment with endo-O-glycanase or sialidase (sialidase A/NANase III) did not alter KID3 reactivity. Additionally, treatment with Pronase did not alter KID3 reactivity (data not shown). These studies suggested that RAV12 bound exclusively to an N-linked carbohydrate moiety and prompted examination of whether RAV12 recognized known tumor-associated carbohydrate antigens.
|
RAAG12 Is Highly Expressed by Gastrointestinal Adenocarcinoma Tumors
RAAG12 is expressed predominantly by adenocarcinomas. Colorectal, gastric, and pancreatic adenocarcinomas express RAAG12 most frequently and with the most uniform expression. Photomicrographs of representative primary human tumors are shown in Fig. 2A
. As shown in Table 1A
, more than 90% of tumors of colorectal, gastric, and pancreatic origin expressed RAAG12, and a majority of these tumors (64–75%) exhibited uniform RAAG12 expression (staining of >75% of tumor cells present). A number of other adenocarcinomas, including breast, prostate, lung, kidney, and ovarian, showed overall frequencies of RAAG12 expression between 30% and 65%.
|
|
Immunohistochemical analysis showed limited RAAG12 expression on normal human tissues. No expression of RAAG12 was seen on connective tissue or tissues from the cardiovascular, endocrine, hematolymphatic, neuromuscular, and central nervous systems. RAAG12 expression was observed in exocrine epithelia in multiple tissues, which in general are the tissues from which the corresponding RAAG12-positive adenocarcinomas arise. The most frequent and intense RAAG12 expression was seen in mucosal epithelium of the gastrointestinal tract, biliary epithelium, mucosal stratified squamous epithelium, pancreatic acinar/ductal epithelium, skin sweat glands/ducts, and glandular/ductal breast epithelium. The majority of RAAG12 in cells from these normal epithelial tissues was localized to the cytoplasm. Membrane expression in gastrointestinal and ductal epithelium was predominantly restricted to the apical surface in normal tissue samples, whereas RAAG12 expression in adenocarcinomas tended to be throughout the membrane and accessible on the surface (Fig. 2B).
RAV12 Is Directly Cytotoxic to COLO 205 Cells In vitro via Oncosis
Initial studies showed that the murine KID3 antibody was cytotoxic to the RAAG12-expressing COLO 205 cell line, and that the bioactivity was retained by the RAV12 chimeric antibody. As shown in Fig. 3A
, RAV12 induced a concentration-dependent cytotoxicity on COLO 205 cells with an ED50 of 7 µg/mL. RAV12-mediated cytotoxicity seems to require cross-linking of cellular components: Fab fragments of RAV12 did not induce cytotoxicity, whereas cross-linking of the Fab fragments with a secondary antibody restored the cytotoxic activity to the level of the intact antibody (data not shown). Following RAV12 treatment, COLO 205 cells in vitro seemed to undergo a type of regulated cell death termed oncosis. The hallmarks of oncosis are cell and organelle swelling, vacuolization, and increased membrane permeability, followed by cell lysis (4, 9, 10). Microscopic examination revealed that COLO 205 cells underwent rapid swelling within 1 h following RAV12 treatment (Fig. 3B). Double staining for RAV12 and filamentous-actin showed that RAV12 was associated with the plasma membrane, and that treatment with RAV12 caused disruption of the actin cytoskeleton. Time-lapse microscopy using KID3 revealed that cellular swelling led to membrane leakage and/or rupture following treatment (Fig. 3C). Elevated lactate dehydrogenase was detected in the medium within 30 min of antibody addition, indicative of membrane rupture (data not shown).
|
RAV12 Inhibits Tumor Growth in Multiple Gastrointestinal Xenograft Models
Studies with KID3 in the subrenal capsule model established that tumors derived from tumor cell lines with high uniform expression of RAAG12 (COLO 201, COLO 205, SNU-16, and SU.86.86) were sensitive to treatment with KID3, whereas cell lines expressing low levels of RAAG12 (<10% of cells staining; A549, HT-29, and CaKi-2) were insensitive. Based on these results, a series of subrenal capsule xenograft studies were conducted to assess the in vivo bioactivity of RAV12. A bridging study comparing KID3 and RAV12 showed comparable activity toward COLO 201 xenografts in the subrenal capsule model (data not shown). Dose response studies with RAV12 showed that treatment decreased tumor size by 60% relative to control at 1 mg/kg, the lowest dose examined, whereas doses of 25 mg/kg decreased tumor size by more than 92%, and doses of 50 mg/kg eliminated detectable COLO 201 tumors in the majority of treated animals (Fig. 4A
). RAV12 treatment caused a similar dose-dependent reduction of tumor size in subrenal capsule xenografts derived from SU.86.86 tumor cells (Fig. 4B).
|
Antitumor activity was also observed with RAV12 in s.c. xenograft models. As shown in Fig. 5A , administration of 50 mg/kg RAV12 i.p. twice weekly beginning on the day of SNU-16 gastric tumor implantation suppressed tumor growth by 69% relative to control at day 25 after implantation. Similarly, animals given 50 mg/kg RAV12 i.p. twice weekly beginning on the day of COLO 205 implantation showed a 74% decrease in tumor growth at the end of study (Fig. 5B). In contrast, A549 tumors, which express low, heterogeneous levels of RAAG12, were not affected by RAV12 treatment at 50 mg/kg (Fig. 5C). The dose response of RAV12 was evaluated using SNU-16 s.c. xenografts. RAV12 was administered at 3, 10, 30, and 100 mg/kg i.v. on the day of tumor implantation. As shown in Fig. 5D, administration of 3 mg/kg RAV12 resulted in a 71% reduction in tumor growth of SNU-16 xenografts on day 22. The 10 mg/kg dose level resulted in an 80% reduction in tumor growth on day 22, which was not significantly improved by higher dose levels nor was it significantly different from the 3 mg/kg dose. The i.v. dosing study with SNU-16 xenografts was repeated with doses of 1, 3, and 10 mg/kg of RAV12. The 3 mg/kg dose of RAV12 caused a 82% reduction in tumor growth, which was not statistically different from the 10 mg/kg dose group (data not shown). Based on these results, RAV12 antitumor activity was examined with SNU-16 s.c. xenografts in an established tumor mode. Twice weekly i.p. treatment of established SNU-16 xenografts (150–200 mm3 tumor volume) with RAV12 at 10 mg/kg completely suppressed tumor growth over the initial 15 days of treatment and resulted in an overall decrease in tumor growth of 66% at the end of the study (Fig. 5E). Treatment of the established SNU-16 xenografts with RAV12 at 50 mg/kg i.p. led to a similar pattern of reduced tumor volume with an overall reduction of 58% at the end of the study.
|
Many therapeutic antibodies have been developed for the treatment of cancer. However, only a small subset of these antibodies has been shown to elicit direct cytotoxic activity. The data presented here show that RAV12 is able to exert a direct cytotoxic effect in vitro against human colorectal adenocarcinoma cells. COLO 205 cells exposed to RAV12 in vitro undergo rapid membrane swelling and lysis, hallmarks of oncosis. RAV12-induced cytotoxicity in vitro occurs in both serum-containing and serum-free cell culture, supporting the conclusion that RAV12 had direct cytotoxic activity on tumor cells (data not shown). The cytotoxic activity observed for RAV12 in vitro was reflected in the potent antitumor activity observed in multiple gastrointestinal xenografts in vivo, including colon, gastric, and pancreatic tumors. Significant antitumor activity was observed with RAV12 at the 1 mg/kg dose in the subrenal capsule model, where treatment started 2 days following tumor implantation, and increased to a 92% reduction in tumor size at the 25 mg/kg dose. Treatment with 50 mg/kg or greater of RAV12 reduced tumors to undetectable levels in many animals. Consistent with the activity observed in the subrenal capsule studies, potent antitumor activity was also observed in both preventative and established s.c. xenograft studies. For SNU-16 tumors, the 3 mg/kg dose of RAV12 initiated on the day of tumor implantation resulted in a 71% decrease in tumor growth at the end of study, whereas a 10 mg/kg dose of RAV12 initiated after tumors had established to 150 to 200 mm3 led to a 68% decrease in tumor growth at the end of study. We are expanding our xenograft studies to investigate the potential for additive or synergistic effects of approved standard of care therapies.
The term oncosis was initially proposed to describe ischemic death of osteocytes with swelling (9). Oncosis has been recently adapted by Majno et al. to describe a mechanism of cell death distinct from the more extensively studied cell death mechanism of apoptosis (10). Although apoptosis is characterized by cell shrinkage, chromatin condensation, and formation of apoptotic bodies, oncosis is characterized by rapid cell swelling, organelle swelling, membrane permeability, and cell lysis. Cell death by apoptosis leads to necrosis with karyorrhexis, whereas cell death by oncosis may result in necrosis with karyolysis (9, 10). Thus, postmortem examination of both apoptotic and oncotic cells may reveal cellular and/or molecular characteristics of necrosis. An emerging view is that cell death in many systems previously described as necrosis may result from activation of nonapoptotic cell death mechanisms such as oncosis. Two examples of antibodies that mediate cell death resembling oncosis have been reported. The RE2 antibody is cytotoxic to lymphocytes by a mechanism distinct from apoptosis (13). Although the molecular target of the RE2 antibody remains unknown, RE2-mediated cell death was accompanied by rapid formation of plasma membrane pores. Similar to our observations with RAV12, RE2-mediated cell death occurred in the absence of complement and required cross-linking of cell surface components. A second antibody directed against porimin, a 110-kDa cell surface receptor, induces complement-independent cell death with features of oncosis. Anti-porimin–mediated cell death was preceded by rapid cell aggregation and pore formation on the plasma membrane, cell swelling, and membrane blebbing (14, 15). Defining the molecular entities mediating oncosis in these systems may aid in understanding the relationship between these oncotic cell death signals and whether they share common molecular pathways.
Oncosis occurs in numerous human diseases, including acute renal and liver failure, myocardial infarction, and stroke (16). In some cases, cell death is accompanied by inflammation due to the release of intracellular contents of oncotic cells. Although the molecular mechanisms of oncosis are not fully understood, the nature and sequence of events in cells undergoing oncosis is beginning to emerge. In the case of hepatocytes exposed to menadione or KCN, Carini et al. (17) observed that the intracellular concentrations of sodium and ATP can play a key role in determining cellular fate. Treatment of hepatocytes with menadione or KCN caused a drop in intracellular ATP. In the presence of sodium in the bathing medium, most cells died following treatment; however, when sodium was removed from the bathing medium, the cells did not swell or lose viability. In this system, the combination of low intracellular ATP and high intracellular sodium were necessary for oncosis. Consistent with this observed role of sodium influx in oncosis, we found that RAV12-mediated lysis could be attenuated by reducing the extracellular sodium concentration.
Immunohistochemical analysis showed that adenocarcinomas, especially colorectal, gastric, and pancreatic adenocarcinomas, have a high prevalence of strong, uniform RAAG12 expression. Additional studies have shown that both primary and metastatic lesions in colon cancer have similar frequencies of strong, uniform RAAG12 expression, suggesting that the RAAG12 glycotope persists in metastatic disease. In gastrointestinal adenocarcinomas, RAAG12 membrane localization progresses from apical and increased basolateral expression in well to moderately differentiated tumors to expression across the entire membrane surface in poorly differentiated and diffuse tumors. Conversely, RAAG12 expression on normal polarized epithelia is predominantly localized to the cytoplasm or apical membrane surfaces. We hypothesize that the differential localization of RAAG12 in tumor versus normal epithelium may translate to preferential access of tumors to i.v. administered RAV12.
Aberrant glycosylation has been observed in human cancers and is thought to play a role in oncogenic transformation, tumor invasion, and metastasis (18–20). Many of the carbohydrate epitopes resulting from aberrant glycosylation have the potential to be tumor-associated antigens that are either tumor specific or overexpressed on tumors and may serve as promising targets for cancer therapy. The Lewis blood group family of carbohydrate antigens represents a prototypical tumor-associated antigen. Expression of Lewis blood group antigens on normal adult tissues is restricted to granulocytes and differentiated epithelial cells but is overexpressed on carcinomas. In the case of Ley, abundant expression has been reported on carcinomas of the colon, breast, ovary, lung, pancreas, and prostate (21), and this expression has been correlated with increased invasion and metastasis and decreased survival rates in patients (22, 23). Our studies using glyco-analogues showed that KID3 is reactive to the Lea-related analogue LNFPII. Competition studies showed that the Lea triose was insufficient to support KID3 binding, whereas lacto-N-tetraose completely inhibited KID3/LNFPII interaction. Lactose (Galß1-4Glc) alone was also not supportive of KID3 binding as all glyco-analogues tested were lactose-based derivatives. Taken together, our results support the conclusion that the RAAG12 glycotope is a N-linked type 1 glycan chain consisting minimally of Galß1-3GlcNAcß1-3Gal and represents a potentially novel tumor-associated carbohydrate antigen.
The protein molecular entities that confer cellular sensitivity to RAV12 remain unidentified. We hypothesize that cell sensitivity to RAV12 may be mediated by a subset of RAAG12-bearing proteins. The requirement for cross-linking of RAV12 Fab fragments for cytotoxicity is consistent with the hypothesis that RAV12 cross-links RAAG12 epitopes, leading to antibody-receptor complex internalization and initiation of pro-oncotic signals. A recent report by Klinger et al. (24) reported that ErbB1 and ErbB2 are modified with Ley on human breast and epidermoid tumor cells, and monoclonal anti-Ley antibodies block epidermal growth factor– and heregulin-stimulated phosphorylation of mitogen-activated protein kinase in these cells. The inhibition of ErbB-induced downstream signaling was noncompetitive, and anti-Ley antibody–mediated rerouting of receptor trafficking has been proposed as a mechanism for the inhibitory activity. We are initiating studies to survey RAAG12-bearing cell surface proteins on responsive and nonresponsive cell lines in an attempt to identify differentially expressed molecular targets that contribute to RAV12 sensitivity.
In conclusion, we have generated a chimeric monoclonal antibody (RAV12) that induces oncotic cell death in vitro and exhibits potent antitumor activity in vivo against multiple human tumor cell lines of gastrointestinal origin. Based on the biological properties of RAV12, the expression pattern of the RAAG12 glycotope antigen, and the safety profile observed in a repeat-dose primate toxicology study, a phase I/IIA clinical trial was designed to examine safety, pharmacokinetics, and preliminary efficacy in patients with relapsed, refractory adenocarcinoma of gastrointestinal (gastroesophageal, pancreatic, and colorectal) origin or other origin if proven to express RAAG12 by immunohistochemistry. This trial is currently accruing patients.
Acknowledgments
We thank Drs. Gordon Vehar and Stanford Steward for helpful discussions and critical review of the article, the Raven Biotechnologies scientific staff for expert technical assistance, and Linda Davis for assistance with preparation of the article.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Current address for N. Pryer: Novartis Institute for Biomedical Research, 4560 Horton Street, Emeryville, CA 94608.
1 P. Young and J.P. Mather, in preparation. 
Received 9/18/06; revised 12/11/06; accepted 1/25/07.
References
This article has been cited by other articles:
|
|
 |
|
  J. H. Overmeyer, A. Kaul, E. E. Johnson, and W. A. Maltese Active Ras Triggers Death in Glioblastoma Cells through Hyperstimulation of Macropinocytosis Mol. Cancer Res., June 1, 2008; 6(6): 965 - 977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chi-Hang Li and R. Li RAV12 Accelerates the Desensitization of Akt/PKB Pathway of Insulin-like Growth Factor I Receptor Signaling in COLO205 Cancer Res., September 15, 2007; 67(18): 8856 - 8864. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Article on RAV12 antibody, oncosis, and antitumor activity Mol. Cancer Ther., May 1, 2007; 6(5): 1682 - 1682. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, Gal; 
, GlcNAc; 
, Glc; 
, GalNAc; 
, Fuc. Representative of four independent experiments. LNnT, lacto-N-neotetraose; LNT, lacto-N-tetraose; LNnH, lacto-N-neohexaose; LNH, lacto-N-hexaose.
