Modeling inflammation and oxidative stress in gastrointestinal disease development using novel organotypic culture systems

Gastroesophageal reflux disease (GERD), Barrett's esophagus (BE), graft-versus-host disease (GVHD), and inflammatory bowel diseases such as ulcerative colitis and Crohn's disease are common human gastrointestinal diseases that share inflammation as a key driver for their development. A general outcome resulting from these chronic inflammatory conditions is increased oxidative stress. Oxidative stress is caused by the generation of reactive oxygen and nitrogen species that are part of the normal inflammatory response, but are also capable of damaging cellular DNA, protein, and organelles. Damage to DNA can include DNA strand breaks, point mutations due to DNA adducts, as well as alterations in methylation patterns leading to activation of oncogenes or inactivation of tumor suppressors. There are a number of significant long-term consequences associated with chronic oxidative stress, most notably cancer. Infiltrating immune cells and stromal components of tissue including fibroblasts contribute to dynamic changes occurring in tissue related to disease development. Immune cells can potentiate oxidative stress, and fibroblasts have the capacity to contribute to advanced growth and proliferation of the epithelium and any resultant cancers. Disease models for GERD, BE, GVHD, and ulcerative colitis based on three-dimensional human cell and tissue culture systems that recapitulate in vivo growth and differentiation in inflammatory-associated microphysiological environments would enhance our understanding of disease progression and improve our ability to test for disease-prevention strategies. The development of physiologically relevant, human cell-based culture systems is therefore a major focus of our research. These novel models will be of enormous value, allowing us to test hypotheses and advance our understanding of these disorders, and will have a translational impact allowing us to more rapidly develop therapeutic and chemopreventive agents. In summary, this work to develop advanced human cell-based models of inflammatory conditions will greatly improve our ability to study, prevent, and treat GERD, BE, GVHD, and inflammatory bowel disease. The work will also foster the development of novel therapeutic and preventive strategies that will improve patient care for these important clinical conditions.

colitis and Crohn's disease). Induction of GERD and BE requires gastric acid and bile refl ux to induce tissue injury, which results in the release of proinfl ammatory cytokines and subsequent recruitment of infl ammatory cells [4][5][6], while GVHD results from donor-derived T lymphocytes damaging host tissue in response to MHC disparities [7], and ulcerative colitis and Crohn's disease are caused, in part, by dysregulation of mucosal immunity [8]. Th ese diseases are important US healthcare concerns associated with signifi cant morbidity and mortality [9][10][11][12].

Knowledge gaps
Mouse transgenic and surgical models have been the primary means of mechanistic studies for investigating infl ammation-based GI diseases. While these models have been useful, they have obvious limitations because mice do not normally develop BE or ulcerative colitis [13]. Improved models based on human cells and tissues are urgently needed to understand pathogenesis as well as to explore novel therapeutic strategies. Th e lack of robust models has been a major impediment to these fi elds.
Cell culture has been used in several forms for the study of GI disease [14]. Th e use of human cells in tissue culture systems, such as proposed here, allows for the ability to mimic the in vivo interplay between the epithelium, immune cells, and the underlying stroma, providing a microphysiologically relevant environment [14] (Figure 1a). Immune cells activated in response to tissue injury or other mechanisms are signifi cant producers of ROS, including hydrogen peroxide , and secrete cytokines that further amplify endogenous and exogenous ROS (Figure 1b,c). In the esophagus and intestine, esophagitis and acute IBD elicit the expression of Thelper type 1 proinfl ammatory cytokines such as IL-1β, IL-8, and IFNγ, while BE and the later stages of IBD present with a predominantly T-helper type 2 humoral response, with signifi cantly increased levels of IL-4 [15,16]. However, the mechanism by which this infl ammatory process results in metaplasia, dysplasia, and cancer is not known.
Current models suggest ROS generated by immune cell responses induce DNA single and double-stranded breaks or modify DNA bases by forming adducts (Figure 1a). Th e major forms of oxidative DNA damage are nonbulky lesions such as 8-oxo-2'-deoxyguanosine and thymine glycol [17]. Th ese modifi cations can lead to base mispairing and point mutations that, if not corrected prior to DNA replication, can cause mutations that activate oncogenes or inactivate tumor suppressors [18]. Furthermore, epigenetic changes in gene expression can also enhance disease progression [19]. Using artifi cial approaches (addition of hydrogen peroxide to cell cultures), oxidative stress was recently established to profoundly alter epigenetic patterning in vitro [17]. Using physiologically rele vant conditions, however, it remains unproven whether the oxidative stress associated with infl ammatory responses is the primary driver of the pathogenesis of BE, GVHD, and IBD, and the dysplasia and cancer associated with BE and IBD.
Esophageal three-dimensional (3D) organotypic culture (OTC) methods utilize epithelial cells grown on top of collagen/Matrigel matrices containing human fetal esophageal fi broblasts [20]. OTC methods provide a highly physiological relevant model for studying esopha geal squamous epithelial growth and diff erentiation. However, the full potential of this culture system to model human disease conditions such as BE has not been fully explored. Fibroblasts are an essential compo nent of the 3D OTC system; they remodel the matrix and secrete factors that support squamous epithelial cell proliferation and diff erentiation [20]. BE epithelial cells probably require diff erent factors to sustain their growth. Several studies have identifi ed some of these factors; in vivo they include bone morphogenetic protein 4 and wingless int-1 protein-2 [21,22], and in vitro they include Noggin, wingless int-1 protein-3a, and R-spondin-1 [23]. In addition, the activity of the oncogene cyclooxygenase-2 is induced in fi broblasts as well as in the epithelium in GERD and BE [24]. Our laboratory previously demon strated that endogenous overexpression of cyclooxy genase-2 in epithelial cells in 3D OTC results in the development of intestinal mucin-fi lled epithelial cysts, consistent with the progression of BE development [25]. Nevertheless, the extent that factors produced from fi broblasts infl uence BE disease development and/or progression is largely unknown.
In 2011, four groups independently published descriptions of methods by which human intestinal and colonic stem cells can be maintained and expanded in culture to mimic in vivo microenvironments [23,[26][27][28] (Figure 2a). Th ese techniques built upon and extended a slightly older, more established method of culturing mouse small intestine stem cells [29]. In this culture method, human intestinal and colonic crypts or purifi ed stem cells are embedded in Matrigel. Over several days these cells form spherical or elongated oval structures with a crypt-like lumen, referred to as human enterospheres [30] ( Figure 2b). Small intestine and colon spheres can expand into multilobulated enteroids (Figure 2c) and colonoids that mimic the ordered structure of the epithelium complete with crypts containing multipotent columnar base stem cells and Paneth cells. Th e remaining cell types, including enterocytes, goblet cells, and enteroendocrine cells [23,30], can be observed in larger cysts away from the stem cell compartment (Figure 2a). While the majority of studies conducted with human enteroids and colonoids thus far have focused on stem cell charac terization and regenerative medicine, there is immense potential for this culture system in GI research, particularly to model GVHD and IBD. However, they have not as yet been widely adopted to model these and other human disease conditions.

Future needs and research directions
Our work seeks to develop novel multicellular in vitro human tissue engineered models that physiologically recapitulate the acute and chronic infl ammatory microenvironment in order to model the eff ects of oxidative injury on the epithelial cells of the esophagus and intestine. Th ese studies are anticipated to yield well-charac terized culture systems that are representative of the pathogenesis of GERD, BE, GVHD, and IBD, thereby providing useful laboratory models for these conditions. Experimental models for these conditions utilizing human tissues and cells are an unmet research imperative.
In addition, we anticipate more specifi c products of our eff orts. Th e development of a dynamic multicellular system to physiologically model infl ammation will permit a more mechanistic understanding of the contributions of infi ltrating immune cells and oxidative stress to DNA damage and epigenetic changes, particularly regarding their contribution to the progression to dysplasia and cancer. Moreover, these novel human cell tissue models can place human mucosal immunology research back in its appropriate context; that is, human epithelial systems. Pursuit of this end will require modifi cations of our currently established 3D organotypic and intestinal enteroid/colonoid culture systems to refl ect the infl ammatory processes active during human diseases such as GERD, BE, GVHD, and IBD. Lastly, these human cell-based models of chronic infl ammatory conditions can serve as a platform for testing novel therapeutic drugs and chemopreventive agents.
We are presently pursuing several important research directions. Th e fi rst is to better characterize the contributions of fi broblasts to the growth and diff erentiation of the squamous epithelium in the OTC system. Specifi cally, we are exploring whether fi broblasts function primarily to remodel the underlying matrix by altering physical properties such as stiff ness, or instead secrete required growth and maintenance factors, or some combination of both. Moreover, we are investigating whether we can engineer the fi broblasts to secrete other growth factors that support columnar epithelial growth, to determine whether this can promote epithelial metaplasia as observed in BE. We are actively exploring modifi cations to both the esophageal OTC and intestinal enteroid/organoid culture systems that will permit the inclusion of viable purifi ed human immune cells (Figures 1b,c and 2b). We plan to further refi ne the human OTC and intestinal enteroid/ colonoid cultures to include additional microenvironmental cell types and factors to better and more physiologically model human disease. For instance, eliciting a polarized response of purifi ed human immune cells against the mismatched human leukocyte antigen of the growing human enteroid would provide an excellent model for GVHD. Once optimized, these co-culture systems could evaluate the effi cacies of novel immunosuppressant drugs and strategies to treat GVHD.

Conclusion
Our research eff orts are focused on developing physiologically relevant human cell-based tissue culture models for chronic infl ammatory conditions of the GI tract. With these systems we will mechanistically explore the eff ect of the infl ammatory response on epithelial oxidative stress and DNA damage. Ultimately, these novel model systems have the potential to permit the rapid testing of hypotheses and advance our understanding of chronic infl ammatory disorders. Moreover, this research will foster the development of novel therapeutic and preventive strategies that will improve the lives and wellbeing of patients affl icted with these important clinical conditions.