The etiopathogenesis of ulcerative colitis (UC) remains obscure, due to lack of an ideal animal model[1-8]. With the improvement of theory and methodology in the last 30 years, people used to adopt chemicals (acetic acid, ethanol, carrageen, etc.), immunotechniques (humoral or cellular immunity, immune complex) and substance derived from UC patients to set up various kinds of UC animal model, which mimic the pathologic changes of human UC, so far these remain far from reality[6-10]. A possible exception is the spontaneous colitis developed in the cotton-top tamarin when captured, but this animal is rarely available and expensive preventing its usage[11]. Therefore, establishing an ideal animal model becomes the focus and key of the research study of UC. Nowadays, the models produced by 2,4-dinitrochlorobenzene (DNCB) and acetic acid (AA) came into use because of their simplicity and pathologic changes simulating those of human UC[12-21], however, the characteristics of short course of DNCB method and absence of immunoreactivity in AA method, these two models are not ideal either. In the present study, we established a new rat UC model produced by combination of DNCB and AA, and observed the changes of general condition, the disease course, the pathology, ultrastructure, apoptosis, immunoreaction and intracolonic pressure, in order to develop a more ideal UC animal model.

General condition and disease course Anorexia, bloody diarrhea, mucus in stools were seen in all A, B and C groups after 1 to 2 weeks, which lasted for 16 weeks in group A with weight loss and 8 weeks in group C, while in group B, the bloody diarrhea and mucus in stools decreased gradually and recovered 2 weeks later. No symptoms were observed in group D.

Pathological findings Diffuse hemorrhage, edema, congestion, superficial ulceration in the mucosa with infiltration of lymphocytes, plasma cells and polymorphonuclear cells, cryptitis, crypt abscess could be observed in all A, B and C groups (Figure 2), and these characteristics lasted for 16 weeks in group A and 8 weeks in group C while only 2 weeks in group B. In group D, the bowel wall was normal by gross and microscopic examinations.

Figure 2 Pathological features of ulcers and crypt abscess. HE × 25

Ultrastructural changes There were decreased number of cells, shortened microvilli, swollen mitochondria with depleted ridge, maldevelopment of goblet cell and increased number of lysosomes during the acute phase (Figure 3), which recovered to normal gradually (Figure 4).

Figure 3 Swelling of mitochondria with ridge depletion in goblet cell. × 4000

Figure 4 In healing phase, the decreased number of the cell count and swelling of mitochondria recovered to normal gradually. × 10 K

Apoptosis The apoptosis indexes, 9.9 ± 3.8 in group A, 8.6 ± 3.5 in group B, 8.1 ± 2.9 in group C, were significantly higher than those in group D (4.0 ± 2.1,P < 0.05). Under scanning electronmicroscope, shrinkage of cells, condensation and margination of chromatin could be seen (Figure 5).

Figure 5 Apoptosis of epithelia cell with condensation and margination of chromatin. × 12 K

Immunoreactivity The changes of CD₄⁺CD₂₉⁺ in ulcerative colitis models were shown in Table 1.

CD₄⁺CD₂₉⁺ increased significantly (P < 0.01) in group A and B but not in group C and D 1 week after set up of model (P > 0.05).

Colonic pressure and motility One week after set up of the model in group A, the basal intracolonic pressure was apparently lower than that of the pre-model (proximal pressure: 0.78 kPa ± 0.13 kPa vs 0.88 kPa ± 0.14 kPa, distal pressure 0.76 kPa ± 0.11kPa vs 0.89 kPa ± 0.15 kPa, P < 0.05). The frequency of migrating motor complex waves in vivo were significantly faster 1 week after set up of the model as compared with that prior to (1.59/min ± 0.27/min vs 0.60/min ± 0.12/min, P < 0.05) in 7 rats of group A, this belonged to the pathologic colon of high dysrhythmia. The amplitude of migrating motor complex waves decreased markedly in 10 rats of group A after 1 week of post-model than that of pre-model (proximal pressure: 0.64 kPa ± 0.24 kPa vs 1.98 kPa ± 0.38 kPa, distal pressure: 0.92 kPa ± 0.37 kPa vs 2.45 kPa ± 0.63 kPa, P < 0.01), which belonged to asthenia colon. In the other 3 rats, while the proximal amplitude of migrating motor complex waves were lowered significantly (0.96 kPa ± 0.31 kPa), the distal amplitude of migrating motor complex waves remained still higher (2.35 kPa ± 0.50 kPa).

Success rate The ulceration pattern was present in all 20 rats of group A, the success rate therefore, could be considered 100%.

So far, the precise etiopathogenesis of UC remained unelucidated[1,25-40], though experimental colitis had been produced by various methods mimiking human UC somehow, up to the present there is not one ideal animal model which conforms with human UC in pathogenesis, pathology and biologic behavior[6-21,41,42]. An ideal animal model should fullfil the following requirements[15,24]: ① it should reflect the histological characteristics of the diseases; ② it should be autoimmune in nature; ③ it should have the similar clinical manifestations as in human UC; ④ it should be simple and reproducible. In this study, the UC model produced by AA method was related to its chemical stimulation to the colonic mucosa[12-14], which led to increment of vascular permeability and activate inflammation mediators, resulting in bloody diarrhea, mucus in stools and histological features as diffuse edema, congestion and ulceration, crypt abscess and mucosal infiltration of inflammation cells. It can be used to study the inflammatory mechanism and anti-inflammatory drugs[43-45], but it lacked the immune response which is a dsawback.

DNCB is a hapten, when bound with tissue proteins will be able to elicit immunologic response and induce colitis[18-21]. The clinical symptoms and histological features, in particular CD₄⁺CD₂₉⁺ cells are similar to those of human UC, but the self-limited course of 2 weeks, is too short to be utilized as an ideal model.

In order to overcome these shortcomings we establish a new chronic UC model by using DNCB and AA in combination which has the following advantages[15-17]. ① Clinically it manifests mucus in stools, bloody diarrhea and weight loss, just like those occur in human UC. ② It can reflect the pathologic characteristics of UC, such as continuous superficial colonic inflammation. Microscopically there exist mucosal edema and congestion, infiltration of lymphocytes, plasma cells and polymorphonuclear cells, crypt abscresses and ulceration. ③ It is an immune response model, immunology is well studied in UC[1,46-49]. One of the important immunoregulatory abnormality in UC is related to the T cell response[23,50]. CD₄ interacts with HLA class II molecules, CD₄ positive T cells can be divided into Th1 and Th2 cells whreas CD₂₉ reacts with 130 KD integrin β₁ subunit which is expressed as a heterodimeric complex with one of six α subunits, forming the very late activation antigen (VLA) subfamily of adhesion receptors. The β1 subunit has a broad distribution, and is expressed on lymphocytes, monocytes but weakly on granulocytes. These receptors are involved in a variety of cell-cell and cell- matrix interactions. Co-expression of CD₄⁺ and CD₂₉⁺ can be used as a marker to identify the Th1 cells subgroup, whose main function is to help B lymphocyte to induce antibody production and cell-mediated dissolution[15,23]. The increase of CD₄⁺CD₂₉⁺ can lead to highly activated B lymphocyte and immunoregulatory abnormality. In this test, CD₄⁺CD₂₉⁺ cells are significantly higher after the set up of models by combination of DNCB and AA, Which is in accord with the requirement of immune response similar to that in human UC. ④ It has a long disease course, if can last at least for 16 weeks with chronic damage of the bowel hence, eligible for the assessment of the drug effects. ⑤ The ultrastructural features are similar to those of human UC. The decreased number of epithelial cells and shortened microvilli, swelling of mitochondria with depletion of ridges can lead to impairment of water absorption of colon resulting in diarrhea. The maldevelopment of goblet cell, may lead to mucus in stools. ⑥ The above findings are not only consistent with the changes in human UC[51-59], but also concordant with other reports[60,61]. The model produced by combination of DNCB and AA is a more ideal animal model of UC. Apoptosis was first described by Kerr[62] and is referred to as programmed cell death, which is genetically controlled. In active UC, injury is common. Did apoptotic cells increase in colonic epithelia which led to ulceration This hypothesis had been studied since 1996[63-67]. Apoptosis of the normal colon was localized in the superficial epithelium, far less than one apoptotic body per crypt. In active UC, the loss of epithelial cells occurred mainly by apoptosis in crypts of involved and uninvolved adjacent areas, resulting in impairment of protective mucosal barrier. The mediators of apoptosis are partly related to the Fas/Fas-L interaction and/or changes in Bcl-2 expression. Our previous study demonstrated the number of apoptotic cells of colonic mucosal epithelial cells in human UC increased as shown by flow cytometry[68], This led to the damage of the barrier resulting in ulceration. Under scanning electron microscope, shrinkage of cells, condensation and margination of chromatin could all be seen.

The symptoms of diarrhea and abdominal pain are partly related with colonic motility disturbance[69-75]. In this study, we first used PC Polygraf HR multichannel recording system to measure the intracolonic pressure and motility in physiological and pathological conditions from multiple sites over prolonged periods. The resultant signals were digitized, analyzed and displayed in a readily interpretable manner, and could be easily subjected to a variety of statistical manipulations, as the colon was situated within the abdominal cavity, the temperature, humidity and pH were maintained by the rats themselves, and not interfered by operative maneuver. In this study, UC rats are characterized by decreased intracolonic basal pressure and disturbance of frequency and amplitude of migrating motor complex which may lead to the symptoms of urgency, diarrhea and abdominal pain. In conclusion, the new chronic UC rat model produced by combination of DNCB and AA is similar to human UC in clinical manifestations, histology, ultrastructural changes, immune response, apoptosis and colonic motility, and it is simple, inexpensive and reproducible with high successful rate.