Different Mechanism and Ecacy of Dextran-Sodium Sulfate and 2,4,6-Trinitrobenzene Sulphonic Acid in Modeling Ulcerative Colitis in C57BL/6 Mice

3 6 Background and Aims: Dextran-sodium sulfate and 2,4,6-trinitrobenzene sulphonic acid are 3 7 common modeling methods in studying ulcerative colitis. Little attention has been paid to the 3 8 mechanism differences between the two approaches. Here, we aim to compare the mechanisms and 3 9 efficacy of these two models and wish to provide fundamental proves for choosing ideal ulcerative 4 0 colitis models. Methods: Dextran-sodium sulfate and 2,4,6-trinitrobenzene sulphonic acid were applied to induce 4 2 the colitis in C57BL/6 mice for seven days. Body weight and disease activity index were assessed. 4 3 Hematology was detected by routine blood test. Histopathology was analyzed by hematoxylin-eosin 4 4 staining section. Enzyme-linked immunosorbent assay, Western blot and quantitative real-time PCR 4 5 were used to detect the cytokines protein levels and mRNA levels. Flow cytometry were used to 4 6 detect the cycles and subsets of splenic cells. 4 7 Results: Dextran-sodium sulfate induced colitis in C57BL/6 mice showed higher acute immune 4 8 activities, while 2,4,6-trinitrobenzene sulphonic acid induced colitis showed chronic immune 4 9 activities with high platelet amounts and activation. Dextran-sodium sulfate is more suitable for 5 0 modeling acute ulcerative colitis. On the contrary, 2,4,6-trinitrobenzene sulphonic acid is more 5 1 appropriate for modeling chronic ulcerative colitis. 5 Conclusions: Dextran-sodium sulfate treatment within 7 days in C57BL/6 mice is a suitable 5 3 experimental model for studying human acute ulcerative colitis with immune response, fecal blood 5 4 and acute pathogenic damage. Conversely, 2,4,6-trinitrobenzene sulphonic acid treatment within 7 5 5 days is more appropriate for studying human chronic ulcerative colitis with hypercoagulable state, 5 6 IL-2 over-expression state and chronic pathogenic damage. To further investigate the peripheral blood immune cell subgroups of the two UC model mice, flow cytometry analysis was performed. The results showed that the percentage of total leukocytes of UC patients under the microscope showed that the lesions were limited to the mucosa and 3 2 0 submucosa layer, sometimes deeper. For patients with chronic diseases, the lesions show distortion 3 2 1 of the colonic crypt structure and may have reduced crypt branching and number, often accompanied 3 2 2 by the increased distance between the base of the crypt and the mucosal muscle layer. Some 3 2 3 individuals have basal plasma cell and lymphocyte aggregation, and may have mucosal vascular 3 2 4 congestion, accompanied by edema and focal hemorrhage, and infiltration of inflammatory cells 3 2 5 such as neutrophils, lymphocytes, and macrophages. The present results showed that DSS-induced 3 2 6 severe colonic mucosal injury was characterized by massive inflammatory cell infiltration through 3 2 7 mucosal, submucosal, and muscular layers and severe crypt damage. DSS-induced damage in the 3 2 8 intestinal barrier further allowed the entry of bacteria and antigens from the intestinal lumen into 3 2 9 the intestinal mucosal layer, causing local inflammation throughout the intestinal wall, which was 3 3 0 in line with the lesions in human patients with fulminant or severe UC. [5,6,17] The TNBS induced 3 3 1 epithelial damage in the mucosal layer, accompanied by loss of necrotic cell and partial crypt 3 3 2 damage. A small amount of inflammatory cell infiltration was seen in the lamina propria, 3 3 3 manifesting as mucosal erosion and localized congestion in the lamina propria, which is inseparable 3 3 4 from the cascade immune response caused by TNBS. The TNBS entered the mucosa and bound to 3 3 5 the lysine ε-amino group to form histone protein, which together formed a complete antigen to cause 3 3 6 an immune response and further the cascade immune response in the mucosa. From the above 3 3 7 intestinal mucosal injury, TNBS injury is more like human UC patients in the chronic or remission 3 3 lesions.


12
To further investigate the peripheral blood immune cell subgroups of the two UC model mice, 159 flow cytometry analysis was performed. The results showed that the percentage of total leukocytes 160 was 161 was mildly to moderately congested, with necrotic areas visible in some samples. Next, the CMDI 185 14 score was calculated: colonic weight and CMDI scores were significantly higher in both of DSS 186 group and TNBS group; CMDI scores in DSS group were significantly higher than those in TNBS 187 group, while colonic weight in DSS group was significantly lower than that in TNBS group. The 188 length of the colon was also measured. The colonic length was significantly shorter in DSS group. 189 Nevertheless, no significant change of colonic length appeared in TBNS group. Next, the colon of 190 mice in pathological sections was evaluated, and the results showed that the colon tissues of mice 191 in DSS group all showed damage to the mucosal layer, submucosal layer and muscular layer; local 192 loss of epithelium in the mucosal layer; some of the mucosal layer epithelium was necrotic and 193 hemorrhagic; the structure of the lamina propria was damaged; a large number of inflammatory cells 194 15 or new granulation tissue was seen in all layers; the crypts were lost. 195 by different methods, RT-qPCR was used to explore the mRNA expression of colon cytokines. As 219 shown in Figure 5, the IL-1β mRNA and IL-10 mRNA were significantly upregulated in DSS group. The mRNA expression of splenic IL-8; L. The mRNA expression of splenic IL-10; Date are means 235 ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P＜0.0001. 236

237
As is shown in Figure 6, IL-1β: Compared with control group and TNBS group, DSS group shows 238 a significant increase, while no significant change happened in Control group and TNBS group; IL-239 10 also shows the same condition as IL-1β. However, IL-2 shows an upregulating trend in TNBS 240 groups when compared with control group and DSS group. 241

Differences in the effects of DSS and TNBS staging on the spleen 251
Spleen, as a center of systematic immune, is reported to be influenced after colitis induced in animal 252 models. [19] To better understand the difference of systematic immune activity between these two 253 modeling methods, we also investigated the splenic immune cells subsets by flow cytometry, and 254 the expression levels of several cytokines by Western blot and RT-qPCR. As shown in Figure 5, the 255 mRNA expression of IL-2 and IL-10 were significantly upregulated in TNBS group. The mRNA 256 expression of IL-10 was significantly upregulated in DSS group. Also, we detected the expression 257 of IL-1β, IL-2 and IL-10 in splenic issues through western blot. As showen in Figure 6, IL-10 258 significant upregulation in DSS group can be seen, and significant increase of IL-2 also showed in 259 TNBS group, which provide the best agreement with the situation of mRNA expression. 260 Interestingly, a significant downregulation of IL-1β was observed in DSS group. Besides, no 261 significant change of IL-10 expression was detected in TNBS group, which means IL-10 262 transcription was upregulated but failed to finish its generation. Sequencely, we investigated the 263 splenic immune cells subsets. As showen in Figure 7 Eisen HN first proposed the TNBS modeling method, which has since been used in various 294 experimental animals. [26] Some studies have initially explored the differences between the two 295 models of DSS and TNBS through endoscopy, pathological diagnosis, or a class of bio-signaling 296 molecules alone. Hitherto, some scholars believe that TNBS is more like human Crohn's disease in 297 23 some pathological mechanisms. However, there is no study to explore the effect of these two models 298 on the efficacy of human UC disease model from the perspective of local colon and overall immune 299 response. In our study, morphological damage, colonic immune response, and overall immune 300 response of C57BL/6 mice were investigated by modeling with DSS and TNBS, respectively. 301 Importantly, we found significant differences in morphology, local and overall immune response 302 between the two models, which directly guide the different applicability of the two models. The 303 ordinary circumstances of DSS group mice showed a progressive aggravation, which were similar 304 to the period of human IBD activity. And the circumstances of TNBS group mice were more like 305 the remission period of human IBD. 306 307 Colonic injury is a typical lesion common to IBD. The common features of the colon in patients 308 with ulcerative colitis are erythema and fine-grained surface seen in mild inflammation, mucosal 309 bleeding, edema, and ulceration in severe lesions. [5] In this experiment, the colons of DSS group 310 mice were significantly wrinkled, with obvious macroscopic mucosal hemorrhage, edema and 311 ulceration. In contrast, there was no significant change in colon length in TNBS group. The colon 312 weight increased in both models, which was caused by colonic edema and congestion. Both models 313 resulted in increased CMDI scores, among which the degree of colonic mucosal damage was more 314 severe in the mice of the DSS group. Meanwhile, differences in colorectal pathogenic development 315 direction also exist between the two models. DSS induced colitis started from the rectum and 316 gradually extended to the proximal end, which was consistent with the direction of human UC.

cells (dc) and effector T cells, produce increased amounts of soluble and membrane-bound TNF, 343
thereby promoting the secretion of IL-1 and TNF-α by macrophage. [28] Therefore, in order to better 344 understand the difference between these two modeling methods, we also compared the immune cells the mechanism by which DSS-induced colitis causes peripheral macrophages to migrate and 367 accumulate in the colon. It's reported that UC is considered to have a Th2 profile. [36] IL-2 is also 368 mentioned to be one of the specific product of Th2. In this study, colonic IL-2 upregulating tendency 369 was observed in TNBS group mice, while IL-2 expression in peripheral blood was increased. 370 Conversely, colonic and serum IL-2 of DSS group shows no significant change, which is 371 interestingly different with human UC. It's also reported that IL-2 can stimulate the proliferation 372 and activation of NK1.1+ cells. [35,37] Significant increases in both peripheral blood T cell subsets 373 and splenic NK1.1+ T cells happened in TNBS group, while peripheral blood NK cells did not show 374 a significant increase, which to some extent indicates that peripheral blood IL-2 has an effect on 375 splenic NK1.1+ T cells, but had little effect on peripheral blood NK1.1+ T cells. In addition, IL-1β 376 and TNF-α did not show significant increases, which might led by the chronic inflammatory activity 377 induced by TNBS. It has been reported that in human IBD, TNF-α promotes IL-8 release by 378 stimulating monocytes [38,39] , thus recruiting granulocytes to infiltrate colon tissue, but IL-8 levels 379 in DSS and TNBS groups were not significantly increased. Interestingly, the expression of IL-10 380 was altered in DSS-induced colitis mice, which may be a compensatory increase after the onset of 381 DSS-induced colitis. IL-10 is a cytokine that inhibits the development of inflammation by targeting 382 a variety of white blood cells, attenuates excessive immune responses, and protects the epithelium 383 from inflammation-induced damage. [40,41,42] It can prevent mitochondrial dysfunction by inhibiting 384 mTOR, thus avoiding the inflammatory expansion triggered by mitochondrial abnormalities. [42,43]  be specific to CD patients rather than UC patients, and IL-8 secreted by B cells is one of the causes 407 28 of aggravation of CD. [45] In contrast, there is no significant upregulation of IL-8 in peripheral blood 408 and colonic localization but upregulation of IL-10 in colonic localization, which indicates that the 409 B cells that responded in the DSS group of mice may be regulatory B cells. Regulatory B cells can 410 secrete IL-10 to regulate inflammatory activities. [46] Although IL-10 upregulation was shown in 411 mucosal T cells of patients with active ulcerative colitis, human UC shows a decline of regulatory 412 B cells, which is totally different from that of DSS group. [47,48,49,50] 413 414 According to the pathogenesis of human UC, the blood routine often shows an anemic, 415 hypercoagulable state with reduced hemoglobin and increased platelet count, especially in active 416 stage. Our blood routine results showed a decrease in hemoglobin and platelet macrophage ratio in 417 the DSS group, and no significant changes in platelets. This suggested that DSS led to continuous 418 bleeding and anemia in mice, similar to the spontaneous bleeding seen clinically in human UC. In 419 TNBS group, hemoglobin slightly increased and platelets significantly increased. This was similar 420 to the hypercoagulable state observed in patients with active UC, suggesting that TNBS-induced 421 colitis did not resolve its hypercoagulable condition despite a trend of remission on the 7 th day of 422 modeling. The hypercoagulable state of blood in UC is often one reason for ulcer formation and 423 extension. Upregulation and activation of platelets can cause blood hypercoagulation in UC patients. 424 Also, platelets activation has been proved to be involved in inflammatory response. Therefore, 425 further studies are still needed to explore the role of the platelet activation factors in UC. 426

427
In conclusion, DSS-induced colitis is a suitable experimental method for the study of human acute 428 ulcerative colitis. This recruitment of macrophages and granulocytes in the model is more significant, 429 29 showing an approached phenomenon related to human ulcerative colitis. The TNBS-induced colitis 430 more appropriately restore chronic UC in humans. Moreover, the TNBS-induced colitis in C57 mice 431 replicates the mechanism of IL-2 elevation in human UC more effectively. TNBS is more suitable 432 for the analysis of platelet activation and hypercoagulable state associated with colitis than DSS. In 433 human acute and chronic colitis, there are often undifferentiated colitis and indistinguishable colitis. 434 In this study, both DSS and TNBS have their applicable areas. Therefore, the selection of animal 435 models of ulcerative colitis needs to be based on the potential mechanism of model and the research 436 direction accordingly. condition under a natural light-dark cycle. Mice were allowed free access to laboratory animal chow 444 (Dashuo Co., Chengdu, Sichuan, China) and drinking water for acclimation. During the 445 experimental period, each mouse was controlled the daily feeding of 8g±2.75g. After feeding 446 adaptively for 7 days, mice were divided randomly into three groups, i.e. control group, DSS group, 447 and TNBS group. The weight differences of the mice from each group are not significant (P > 0.05). 448 449