ESOPHAGEAL CANCERS AND MODEL SYSTEMS

Abstract

Gastrointestinal cancers are largely epithelial in nature and arise from the esophagus, stomach, pancreas, colorectum and liver. In aggregate, these cancers are the most common malignancies in the United States and worldwide, but suffer from poor outcomes in late stages. Our overall work aims to elucidate the following: 1) how normal epithelial cells become metaplastic and dysplastic; 2) how tumor cells invade and interact with activated fibroblasts and immune cells; and 3) how tumor cells disseminate into the circulation and colonize distant organs (metastatic organotropism). We develop three-dimensional cell culture models and genetically engineered mouse models to decipher mechanisms. Our overarching desire is to translate preclinical models to clinical trials that impact upon outcomes in patients with metastatic gastrointestinal cancers. We will frame these principles and approaches in the context of esophageal cancers and three-dimensional models.

INTRODUCTION

The preponderance of gastrointestinal (GI) cancers is acquired or sporadic with advancing age, and involves a complex interplay between environmental, dietary, and lifestyle factors with genomic/genetic/transcriptomic factors. Studies in human tissues, based upon genotyping and with the explosion of deep DNA sequencing (e.g., genomic and exomic), underscore the importance of genetic alterations. To that end, somatic mutations may vary in frequency within a given GI cancer. Key alterations are found in oncogenes (epidermal growth factor receptor [EGFR], Kras, cyclin D1, c-MYC), tumor suppressor genes (TP53, E-cadherin, p120-catenin, SMAD4), cell cycle regulatory genes (CDK4/6), and transcriptional factors. It is important to highlight the importance of epigenetic regulation in GI cancers through chromatin modification. Additionally, alterations may be present in coding and noncoding RNA as well as proteins revealed through transcriptomic and proteomic analyses. As a result, one can generally construct a multistep model of GI cancers where genetic and biochemical alterations are associated with specific stages of the transition from normal to precancerous to frank primary cancer to metastatic cancer.

A key illustration of principles and approaches in GI cancers is in the context of esophageal cancers. There are two main subtypes of esophageal cancer: esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC). ESCC is more common worldwide, whereas EAC is more common in Western Europe and the United States (1). There is a continuum of normal esophagus to squamous epithelial dysplasia to ESCC; and from normal esophagus to esophageal intestinal metaplasia (also referred to as Barrett's esophagus [BE]) to EAC (2). The ability to model these stages is new and exciting, and paves the way for consideration of diagnostic and therapeutic modalities to improve upon the 5-year survival rate of approximately 15% for each cancer.

MATERIALS AND METHODS

Three-Dimensional Organotypic Culture Model in Tissue Culture

A hybrid matrix of collagen and Matrigel (BD Biosciences, San Jose, California) is mixed with fibroblasts that facilitate stiffening within 1 to 2 days submerged within media (transwell chambers). Then, esophageal squamous epithelial cells are layered on top of this matrix (3). Media is exchanged every 2 to 3 days. After 12 to 13 days, an air-liquid interface is created, triggering epithelial stratification and differentiation. After approximately 15 days from inception, the transwell can be processed for histology, immunohistochemistry (IHC), immunofluorescence (IF), imaging, and laser capture microdissection (3).

Three-Dimensional Organoid Model in Tissue Culture

Tissues can be obtained from endoscopic biopsy specimens or surgical specimens related to BE, ESCC, and EAC. Tissues are enzymatically digested and single cell-derived three-dimensional (3D) organoids in Matrigel and media of the distinct histology can be initiated and passaged. These organoids form rapidly in Matrigel, typically within 14 days. They can be processed for histology, IHC, IF, nucleic acid and protein extraction, and imaging as well (4).

Genetically Engineered Mouse Models

The Epstein-Barr virus L2 promoter (EBV-L2) can be used to target genes to the mouse esophageal squamous epithelium (along with the related squamous oral cavity and squamous forestomach tissues) (5). The EBV-L2 promoter can be fused to Cre recombinase and crossed with the p120catenin (p120ctn) loxP/loxP to generate L2Cre;p120loxP/loxP mice, which were aged and analyzed for gross lesions, histology, and IHC (6). Likewise, EBV-L2 promoter was crossed with the interleukin 1 beta (IL-1 beta) gene to generated L2-IL-1beta mice that were aged and analyzed for gross lesions, histology, and IHC (7).

RESULTS

We have generated genetically engineered models that phenocopy human ESCC and human EAC. For human ESCC, we have focused upon the adherens junctions. The adherens junctions is maintained to a large extent though the interaction between the transmembrane E-cadherin (the prototype of cadherins) and p120catenin (p120ctn). There are other interactions between E-cadherin and beta-catenin and alpha-catenin. P120ctn stabilizes E-cadherin at the cell membrane. If p120ctn is deleted or mislocalized, then E-cadherin is destabilized with compromise of cell-cell interactions. We reasoned that disruption of p120ctn in the mouse esophageal epithelium might foster progression from a normal state to a cancerous state. We have generated L2Cre;p120loxP/loxP mice that show progression to oral-esophageal squamous dysplasia to oral squamous cell carcinoma and ESCC with involvement of lymph nodes after approximately 6 to 9 months (6). Furthermore, there is evidence of the recruitment of myeloid-derived suppressor cells (MDSCs) and stromal desmoplasia (6). For human EAC, we have generated L2-IL-1beta transgenic mice. Once aged for up to 12 months, and beyond, these mice develop intestinal metaplasia at the squamocolumnar junction (verified by histology and electron microscopy) and culminating in EAC (7). The BE and EAC can be accelerated by administering bile acid in the drinking water or exposure to nitrosamines or inhibited by IL-6 deficiency (7).

These novel mouse models of ESCC and EAC allow for elucidation of mechanisms and potential approaches for diagnostics and therapeutics. In a complementary fashion, we have generated 3D tissue culture models to facilitate translational aspects. The first is referred to as a 3D organotypic culture (3D OTC) model in tissue culture. Herein, the interplay between human esophageal epithelial cells and a matrix of collagen, Matrigel, and fibroblasts recapitulates the human esophageal epithelium and some aspects of the stromal microenvironment (3). We have used the 3D OTC model to recapitulate dysplasia and ESCC (8), as well as BE (9). Additionally, we have generated 3D organoid models in tissue culture. Single-cell suspensions from tissue biopsy specimens or surgical specimens allow us to phenocopy the cardinal features of ESCC (4), BE, and EAC. These 3D model systems are rapid, feasible, and allow for the potential for personalized medicine.

DISCUSSION

The vast majority of GI cancers do not have a good prognosis, reflecting the high frequency with which patients present at late stages of disease which makes effective therapeutics challenging. Another aspect of this challenge is that GI cancers undergo rapid dissemination to the blood vessel and lymphatic vessel systems, allowing for local (lymph nodes) and distant metastasis (often to liver, lung, peritoneum, and bone marrow). Although early-stage GI cancers are amenable to either surgery followed by possible adjuvant chemotherapy (e.g., pancreatic cancer and colon cancer) or neoadjuvant chemotherapy/radiation therapy followed by surgery and adjuvant chemotherapy (e.g., esophageal cancer and rectal cancer), there is much need to enhance therapy with targeted biologics (e.g., kinase inhibitors) and immunotherapy.

A key illustration of GI cancers is in esophageal cancer, which has two main subtypes —ESCC and EAC. These esophageal cancers highlight the progress and the challenges that confront GI cancers in general. We have contributed to significant progress in these esophageal cancers through the development and characterization of genetically engineered mouse models and 3D culture models. Such models have led to the delineation of molecular mechanisms involved in the pathogenesis of these cancers as well as translational applications, including the possibility of drug testing and personalized medicine in the 3D culture models. As we look forward, we outline the following in esophageal and other GI cancers:

Unmet Needs and Future Directions

For improvements to occur globally in GI cancers, the following directions are imperative with development and integration of new platforms:

1. Early detection with biomarkers: New serological markers, and circulating tumor materials (exosomes, nucleic acids, cells, proteins).

2. Early detection with imaging: Enhancements in contrast agents, molecular probes coupled to endoscopic procedures and cross-sectional imaging.

3. Risk stratification of patients based upon factors identified in precancerous and cancerous lesions, allowing for variations in screening, surveillance, and therapy.

4. Therapeutics: Development and implementation of new therapeutics, as individual and combinatorial agents, coupled with conventional therapeutic approaches.

5. Prevention: Ultimately, modifications in environmental, dietary, and lifestyle factors, while not as attractive and requiring societal investment as well as education and self-discipline, will have a positive impact.

DISCUSSION

Mackowiak, Baltimore: I have a question about microbiome. When these gastric acid inhibitors were developed, one of the initial fears was that you would get a gastric overgrowth of complex bacterial populations and that pre-carcinogens would be converted to carcinogens in the stomach and this would lead to a rise in gastric cancers. As far as I know that hasn't happened. I'd like to hear your thoughts on that.

Rustgi, Philadelphia: This is complex and largely borne out of epidemiological studies. H. pylori induces a hypochlorhydric state. Conversely, H. pylori eradication can lead to an increase in acid reflux which might be associated with an increase in esophageal adenocarcinoma. With the widespread use of proton pump inhibitors it is not clear if that's leading to an increase in gastric adenocarcinoma. I would say there is definitively a hypochlorhydric state with H. pylori infection that then leads to lymphocytic recruitment into the submucosa and gastric intestinal metaplasia. Dr. Marty Blaser is looking at the microbiome in great detail of the commensal and pathogenic organisms that coexist with H. pylori and modifying the microenvironment that may promote gastric adenocarcinoma.