Critical considerations and relevance of pathological findings of the Dextran Sulfate Sodium model

Inflammatory Bowel Disease (IBD) is comprised of two major clinical phenotypes: Ulcerative Colitis (UC) and Crohn’s Disease (CD). Few pharmacological options are available to patients. While UC mainly affects the mucosal lining of colon and rectum, CD can affect any part of the gastro-intestinal tract with pathological changes involving deeper layers of the intestinal wall. Interplay between host immune system and gut microbiota has been recognized as one of the critical factors in IBD etiology and prognosis. Genetic studies have shown that an abnormal immune response to gut microbiota might be the initial trigger of the disease. Thus, in addition to genetic predisposition, history of previous gut infections and use of antibiotics as well as diet and age may have important roles in the onset of IBD, and pathogenesis of these chronic, relapsing conditions are important considerations when establishing and utilizing rodent models of IBD.

The current therapeutic landscape for IBD is considered mostly unmet due to limited efficacy of most of the pharmacological agents, including biologics. Aminosalicylates such as mesalazine, sulfasalazine, immunosuppressants such as cyclosporine A, and biologics like anti-TNF-alpha monoclonal antibodies are the current mainstays of therapy in IBD. Microbiome intervention has recently gotten attention in IBD drug discovery based on the understanding of the importance of gut microbiota in disease etiology.

Modeling Strategies for IBD

Animal models are widely used in IBD research in order to explore key scientific questions, as well as to screen novel mechanisms and compounds. The three most common approaches to recapitulate IBD in animal models for drug discovery research are: chemically-induced models, adoptive T cell transfer and interleukin 10 (IL-10) knockout mice models. To create a chemically-induced model, compounds known to damage gut epithelial integrity are administered orally or rectally, thereby instituting changes within the colon. In adoptive cell transfer models, transfer of specific subsets of naïve T cells to immunodeficient mice, which lack T and B cells, cause colitis following activation of the T cells by gut antigens. Eventually, mice with the targeted deletion of the IL-10 gene spontaneously develop colitis. While the adoptive cell transfer and genetically modified mouse models might be more related to the mechanistic etiology of human IBD, the chemically-induced models offer better controls in terms of onset, duration and disease severity. In addition, many of the immunological and histopathological features of chemically-induced IBD models resemble human IBD including mucosal inflammation and host immune response to microbiome due to compromised epithelial barrier integrity.

 Among the various chemically-induced IBD models, the Dextran Sulfate Sodium (DSS) model is the most commonly used and primarily mimics features of UC. Although the model has been widely used, there are important factors that greatly impact testing outcome. This model of colitis is induced by feeding animals ad libitum, a particular grade of DSS polymer in drinking water. The molecular weight of the polymer and its concentration are the major determining factors for disease severity. Because of the wide range of molecular weights in the material and possible batch variations, the colitis induction-capability of each batch must be assessed before initiating screening studies. The strain, gender, supplier and housing of the animals are also critical for optimal induction of colitis. In light of the possible contribution of gut microbiota in disease etiology, it is important to obtain animals from the same barrier room and perform animal handling activities under similar laboratory conditions. Among other factors to consider are ensuring the study is appropriately powered (i.e. proper group size) and that pre-selection and randomization of animals is based on pre-study clinical evaluation of symptoms.  

The DSS model in action  

Although we don’t fully understand exactly how DSS causes damage to the colon wall, it has been reported that DSS first induces an acute toxicity to colonic epithelial cells and thereby compromises epithelial barrier integrity, which then allows further entry of DSS and other luminal contents. Initial inflammatory reactions due to the acute toxic effects on epithelial cells start within a day of DSS-exposure, while significant increases in membrane permeability are observed at around three days, resulting in an inflammatory response including infiltration of immune cells (see Figure below). 

Figure: H&E-stained whole colon Swiss-roll sections from control water-treated mice (Control), 3% DSS treated acute colitis model mice (Acute DSS) and 2% DSS treated chronic colitis model mice (Chronic DSS). Acute DSS colitis are characterized by edema, ulceration of epithelial layer, crypt loss and predominantly neutrophilic inflammation in mucosal layer that often extends to the submucosa and muscularis. In the chronic DSS colitis model, in addition to surface ulceration and crypt loss, inflammation with predominantly mononuclear plasmacytoid infiltration and squamous metaplasia of the colonic epithelium and crypts are also observed.

This is reflected in the in-life colitis scores that are visible after 3-4 days of DSS exposure. Typical in-life clinical signs of DSS-induced colitis are weight loss, diarrhea and occult blood in stool. Based on the duration of exposure, DSS can induce acute colitis, chronic colitis or colitis-associated colorectal cancer. Acute colitis is usually induced by 7-9 days of continuous DSS exposure, mimicking the effects of a UC patient experiencing a flare up of symptoms. Induction of chronic colitis usually involves several cyclical exposures of several days duration with a water washout regimen in between that recapitulates relapse-remission of inflammation in UC patients. Typical colonic histopathological features of DSS-induced colitis are edema, ulceration and infiltration of inflammatory cells in gut mucosal layer. Thus, many of the clinical and histopathological features of the DSS-induced colitis model are similar to human ulcerative colitis and therefore this model can be viewed as appropriate for screening therapeutic agents targeting UC.

The hope is that this model, which more closely resembles people with UC, will lead us to better therapies for this debilitating condition.