How Dextran Sodium Sulfate Induces Colitis in Research

Dextran sodium sulfate is a water-soluble sulfated polysaccharide that has become the most widely used chemical for inducing intestinal inflammation in laboratory animals, primarily mice. First shown to cause colitis in hamsters in 1985 and then adapted to mice in 1990, the compound reliably produces gut inflammation that shares features with human inflammatory bowel disease when simply dissolved in drinking water.¹ Its ease of use and flexibility have made it the backbone of preclinical IBD research, though the model has quirks and limitations that are worth understanding.

How DSS Breaks Through the Gut’s Defenses

The colon is lined with a two-layered mucus barrier. The outer layer is loose and hosts resident bacteria; the inner layer is dense and normally keeps those bacteria away from the cells underneath. When DSS reaches the colon, it disrupts the inner mucus layer with startling speed. In explant experiments, exposing mucus to DSS thinned the inner layer and made it permeable to particles roughly the size of bacteria within about 15 minutes. Unsulfated dextran, which is chemically similar but lacks the sulfate groups, had no effect on mucus thickness or permeability, confirming that the sulfate component is what does the damage.²

Once bacteria slip through the compromised mucus, they reach the epithelial cells lining the colon. But DSS is not finished. It also attacks the tight junctions that glue those epithelial cells together. In cell-culture studies, DSS disrupted tight junctions, adherens junctions, and the internal scaffolding of the cells, leading to barrier breakdown.³ One key protein lost in this process is ZO-1, a structural component of tight junctions. In mice given DSS, ZO-1 began disappearing after just one day, and by day three the colon had become measurably more permeable.⁴ The sequence matters: mucus failure comes first, then bacterial contact with the epithelium, then tight junction loss, and finally the inflammatory response. Bacteria were seen penetrating the inner mucus layer as early as 12 hours after DSS exposure, well before any immune cells arrived at the scene.⁵

The Inflammatory Response That Follows

Once bacteria breach the barrier, the immune system reacts aggressively. A central player is a molecular sensor called the NLRP3 inflammasome, which acts as an alarm inside immune cells called macrophages. When macrophages were incubated with DSS in the lab, they secreted large amounts of the inflammatory signal IL-1β. That secretion depended on a chain involving NLRP3, an adaptor protein called ASC, and the enzyme caspase-1. Mice genetically lacking NLRP3 developed significantly milder colitis after drinking DSS, and pharmacologically blocking caspase-1 achieved a comparable degree of protection.⁶

The inflammatory picture is not just about one pathway, though. During acute DSS exposure, the colon floods with neutrophils, and levels of TNF-α, IL-17, and other pro-inflammatory signals climb rapidly.⁷ The timing of these signals is staggered in interesting ways. Most plasma inflammatory markers begin rising after about a week of DSS treatment, but IL-17 does not spike until after DSS is withdrawn. IL-1β and IL-17 can continue climbing during recovery even as visible signs of colitis fade, suggesting that the inflammatory process has its own momentum once triggered.⁸ During recovery, anti-inflammatory signals like IL-10, IL-4, and TGF-β rise to help resolve the inflammation, while NF-κB activity and regulatory T-cell markers gradually increase, reflecting the immune system’s attempt to restore balance.⁹

Why Molecular Weight Is So Important

Not all DSS is created equal. The compound is sold in different molecular weight ranges, and the molecular weight dramatically affects how much damage it does. In one foundational study, BALB/c mice were given 5% DSS at three different molecular weights in their drinking water for seven days. The 40-kDa form caused the most severe colitis, concentrated in the lower colon. The 5-kDa form caused milder disease, mostly in the cecum and upper colon. The 500-kDa form caused no colon lesions at all.¹⁰ The 40-to-50-kDa range has become the standard for research because it reproducibly produces severe colitis.¹¹

This sensitivity to molecular weight is a practical headache for researchers. Different suppliers and even different lots from the same supplier can vary, and switching to a new lot sometimes means recalibrating the dose. Published protocols typically specify 2% to 5% DSS (weight-to-volume) in the drinking water for five to seven days, with the 36-to-50-kDa range being the most commonly cited.¹² Chronic colitis models use repeated cycles. One approach cycles mice through four days of DSS and four days of plain water over two rounds, which is enough to push the inflammation from acute into a chronic phase.¹³

The Surprising Role of Gut Bacteria

You might assume that because DSS-induced colitis involves bacteria breaching the gut wall, animals without any gut bacteria would be protected. The reality is more complicated and somewhat counterintuitive. Germ-free mice given DSS show minimal inflammation since there are no bacteria to trigger an immune response, yet they suffer worse epithelial damage, severe hemorrhage, and higher mortality than conventional mice. Their intestinal barrier is inherently weaker, with lower levels of protective proteins and mucus components.¹⁴

In one striking experiment, germ-free mice given the standard 5% DSS dose developed massive intestinal bleeding by day one and died by day three, before colitis even had time to develop. Conventional mice given the same dose survived the full seven-day experiment. Even at a mild 1% dose, which causes only slight crypt damage in normal mice, germ-free mice developed severe colitis and eventually died.¹⁵ The lesson is that gut bacteria serve a double role. They are needed to trigger the inflammatory component of DSS colitis, but they also help maintain the structural integrity of the gut lining. Without them, the chemical damage from DSS is essentially unopposed.

Specific bacteria can tip the balance toward protection. Colonizing germ-free mice with a single species, Bacteroides fragilis, increased survival by about 40% after DSS challenge, reduced immune cell infiltration, and boosted anti-inflammatory signaling.¹⁶ Meanwhile, DSS itself reshapes the microbial community in mice that have normal flora, reducing overall diversity and shifting the proportions of major bacterial groups.¹⁷ DSS-treated mice showed increased abundance of Firmicutes and decreased Bacteroidetes, along with rises in potentially harmful bacteria like Klebsiella.¹⁸ This dysbiosis mirrors patterns seen in human IBD patients, which is one reason the model remains so popular.

Mouse Strain and Sex Make a Big Difference

The same dose of DSS can produce wildly different outcomes depending on which strain of mouse receives it. A systematic comparison of inbred strains found that C3H/HeJ mice were highly susceptible to DSS colitis, while the closely related NON/LtJ strain was quite resistant. C57BL/6J, one of the most commonly used laboratory strains, showed intermediate susceptibility that varied by anatomical site along the colon.¹⁹ The genetic factors behind these differences are complex. The major histocompatibility complex, which controls much of the immune response, turned out not to be the main driver, since transferring one strain’s immune-gene region onto a resistant background did not confer susceptibility.²⁰

Even between the two most popular strains, BALB/c and C57BL/6, the disease looks different. C57BL/6 mice tend toward early and severe systemic damage with elevated inflammatory markers across multiple organs, increased intestinal permeability, and significant brain-related effects at higher DSS doses. BALB/c mice, in contrast, show more prominent local colonic damage with greater goblet cell loss and neutrophil infiltration, alongside liver and pancreatic involvement.²¹ These differences mean that a treatment showing benefit in one strain might fail or behave differently in another, a variable that many published studies do not adequately account for.²²

Sex adds another layer. Female mice are partially protected against DSS-induced colitis compared with males. In direct comparisons, female mice had less weight loss, fewer inflammatory cell infiltrates, less crypt damage, and lower TNF-α in the colon. Removing the ovaries eliminated this protection, and supplementing ovariectomized mice with estradiol restored it, pointing to estrogen as the protective factor.²³ This sex difference even extends to specific immune pathways. Male mice lacking a particular nicotinic receptor (the α7 subtype) developed more severe colitis than their normal counterparts, and drugs targeting that receptor reduced disease severity in males but had no effect in females.²⁴ Despite all this, a large fraction of DSS studies historically used only male mice or failed to analyze results by sex.

Modeling Colitis-Associated Cancer

One of the more powerful extensions of the DSS model combines it with azoxymethane (AOM), a chemical carcinogen. A single injection of a low dose of AOM followed by one or more cycles of DSS in the drinking water creates a two-hit model of colitis-associated cancer. The AOM initiates genetic mutations, and the chronic inflammation from DSS promotes tumor growth. In male ICR mice, this combination produced colon adenocarcinomas in 100% of animals by week 20, with an average of about 5.6 tumors per mouse.²⁵ The model has become a standard tool for studying how chronic intestinal inflammation drives cancer development and for testing potential preventive agents.²⁶

Strain susceptibility matters here too. Different mouse strains develop tumors at different rates and in different numbers when given the AOM/DSS combination, which means investigators have to choose their strain carefully depending on the question they are asking.²⁷

Effects Beyond the Gut

Although DSS is delivered orally and the colitis it causes is centered in the colon, the consequences do not stay there. In hamsters, DSS exposure at higher concentrations caused a systemic inflammatory response that included changes in liver function: plasma triglycerides rose, and the liver showed altered expression of genes involved in fat metabolism.²⁸ In mice, spatial mapping of molecular changes across multiple tissues after DSS treatment revealed significant metabolic and immune shifts at sites far from the colon, including some molecular changes not previously linked to intestinal inflammation.²⁹

The strain comparison mentioned earlier found that C57BL/6 mice given higher DSS doses developed measurable brain-related damage, with increased expression of a key immune receptor in both colon and brain tissue, while BALB/c mice showed more liver and pancreatic damage.³⁰ These extraintestinal effects are relevant because human IBD is also associated with problems outside the gut, including joint inflammation, liver disease, and neurological symptoms. The fact that the DSS model recapitulates some of these systemic effects makes it more useful than its reputation as a purely colonic model suggests.

A Hidden Problem for Lab Work

There is a practical trap in DSS research that has probably skewed more published results than anyone would like to admit. DSS that gets absorbed into colon tissue during the experiment contaminates RNA extracted from that tissue, and even trace amounts powerfully inhibit the enzymes used in quantitative PCR, one of the most common techniques for measuring gene expression. Complete inhibition of PCR amplification occurred at DSS concentrations of just 10 nanomolar in the reaction mix.³¹

Follow-up work showed that DSS interferes with both the reverse transcriptase enzyme (which converts RNA to DNA for analysis) and the polymerase enzyme (which amplifies the DNA signal), meaning the contamination can distort results at two stages of the process.³² The problem has been documented in mice and confirmed in a piglet model as well.³³ Without specific purification steps to remove residual DSS before running the assay, gene expression data from DSS-treated tissues can be inaccurate in ways that are not obvious. A gene might look like it has been downregulated when in reality the PCR reaction was simply inhibited by chemical contamination. Researchers who are unaware of this artifact and skip the extra purification step may be reporting misleading data. Both poly-A purification and lithium chloride precipitation have been validated as workarounds.

What Happens When DSS Is Removed

One of the features that makes the DSS model so flexible is that the inflammation is largely reversible. When DSS is withdrawn from the drinking water, mice enter a recovery phase that researchers use to study mucosal healing. Even while the inflammatory flare is at its worst, the tissue has already begun switching on repair programs. Gene expression of key healing signals, including TGF-β, IL-15, IL-22, and IL-33, ramps up early during the acute phase, preparing the epithelium for regeneration once the insult stops.³⁴

IL-22 plays a particularly active role in driving mucosal repair. In studies where IL-22 signaling was disrupted, intestinal healing after DSS withdrawal slowed significantly.³⁵ This recovery phase is increasingly used to test therapies aimed at promoting healing rather than just suppressing inflammation, a distinction that matters because human IBD treatment is shifting toward mucosal healing as a treatment goal.

The Model’s Strengths and Honest Limitations

The DSS model is popular for good reasons. It is technically simple, requiring only that researchers add a chemical to drinking water. It is flexible, producing acute, chronic, or cancer-associated disease depending on the protocol. It generates inflammation that shares important features with human ulcerative colitis: barrier disruption, immune cell infiltration, crypt loss, and dysbiosis. And it is used for drug testing, where it remains the standard first step in evaluating oral formulations intended for IBD patients.³⁶ Drug candidates have been identified using the model, including compounds that work by suppressing specific immune cell types involved in colitis.³⁷

But the model has real limitations that are sometimes glossed over. DSS colitis is driven by chemical injury to the barrier, while human IBD involves an immune system that attacks the gut on its own. The initial trigger is fundamentally different. Results obtained in the DSS model do not always translate to humans, and some findings are model-specific artifacts rather than insights into human disease.³⁸ Compared with other mouse colitis models such as TNBS-induced colitis, the DSS model shows a different time course: TNBS produces rapid weight loss and mortality starting around day one, while DSS colitis takes about five days to become clinically apparent.³⁹ Each model captures different aspects of human IBD, and no single model captures them all. The best preclinical studies confirm findings across multiple models before moving toward clinical trials.