Animal Care and Use Committee of The Bagcilar Training and Research Hospital of Health Science University approved all experimental procedures in this study (protocol number: 2016-10; approval date: 3th March 2016).

Thirty-two female Sprague-Dawley rats (250 ± 30 g, with ages 4–8 weeks), supplied by the Bagcilar Training and Research Hospital Animal Center, Istanbul, Turkey, were maintained under laboratory conditions, housed in a controlled room with 12-hour light/dark cycles at 22℃, and fed with standard pellet chow including 21% protein and daily fresh water. All subjects were kept in separate cages each including 8 rats. All rats were divided into 4 random groups (n = 8) as; group 1, sham group, received equal amount of isotonic solution; group 2, control (radiation) group, received radiation (RT) + equal amounts of isotonic solution; group 3, prophylaxis group, received 3 mg/kg/day ruscogenin for 14 days and then RT; group 4. treatment group, received first RT, then 3 mg/kg/day ruscogenin for 14 days [9].

Under ketamine anesthesia by 5-mg/kg xylazine (Rompun, Bayer, Leverkusen, Germany) and 30-mg/kg ketamine (Ketalar, Pfizer, Istanbul, Turkey), rats of groups 2–4 were fixed on an acrylic plate in the supine position by strapping the tails and legs (Fig. 1). All parts of the body and extremities were excluded from the radiation area except the lower pelvic region in 3- × 4-cm dimensions including the 2-cm-long rectum. The dimensions of radiation area (Elekta, Model Precise Treatment System, 6-mev electron energy, 0.5-bolus material) were 20 × 20 cm² and a single fraction of 2,500 cGy was given [12].

A single 3-mg/kg/day dose of ruscogenin (Chengdu Biopurify Phytochemicals Ltd., Chengdu, Sichuan, China) was injected daily by oral gavage to group 3 for 2 weeks before radiation; and to group 4 for 14 days after radiation. Group 1 was injected daily with a single dose of isotonic solution at the same volume by oral gavage for 14 days without performing radiation. Group 2 was injected daily with a single dose of isotonic solution at the same volume by oral gavage for 2 weeks after radiation. None of the rats died during the experiment.

Laparotomy was performed on every subject at the second week after radiotherapy under ketamine anesthesia (10-mg/kg xyilazine and 60-mg/kg ketamine) (Fig. 1); the rats were sacrificied by intracardiact puncture.

Rectums were dissected and washed with isotonic solution for histopathological and immunohistochemical evaluations. Under observation of a pathologist, the tissues were cut into 3 equal horizontal pieces and put into relevant fixatives [12].

All relevant pieces of rectal segments at 1 cm³ were fixed by 4% neutral formalin and embedded into the paraffin blocks. Five-micrometer sections from every paraffin block were stained by hematoxylin and eosin. Slides were evaluated by the same blinded pathologist twice under a light microscope (Nikon Eclipse E600, Melville, NY, USA).

Lamina propria (inflammation), invasion of leucocytes through the cryptic epithelium (cryptitis), deterioration of crypts (cryptic distortion), regenerative/restorative atypia of epithelial cells (reactive atypia), and inflammatory infiltration were investigated. Lesion severity was evaluated as 1, mild lesion; 2, moderate lesion; and 3, severe lesion; and 0 was accepted as none [12].

Cryptitis abscess was noted as positive or negative for each slide. Infiltration depth was measured by an image analysis system of a video attached microscope (Zeiss Axioplan 2, Scientific Imaging Corp., Campbell, CA, USA).

All relevant pieces of rectal segments were fixed by 4% neutral formalin, routine procedures were applied for tissue preparation and embedded into the paraffin blocks. Sections were stained immunohistochemically by using Streptavidin-Biotin-Peroxidase method with monoclonal and polyclonal antibodies tagged to indicate cell and tissue antigens, according to a previously described method [13]. Primary antibodies against inflammatory markers for TNF-α (Clontech Laboratories Inc., Mountain View, CA, USA), hypoxia-inducible factor 1α (HIF-1α: Novus Bio Inc., Littleton, CO, USA), IL-1β (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), FGF-2 (Santa Cruz Biotechnology Inc.), and angiogenesis marker of vascular endothelial growth factor (VEGF: RayBiotech Inc., Norcross, GA, USA) were used. All antibodies were diluted as 1:100 with fresh phosphate buffer saline. Positively stained regions with relevant antigens were analyzed semiquantitatively in terms of staining intensity (modified H-SCORE analysis) given values between 0–300 and percentages for 5 different regions of each slide [13]. Thereby, rectal localization of these proteins and variance of the staining intensity and regional differences were determined.

Rectal segments at 1 mm³ were fixed by 2.5% glutaraldehyde and prepared for ultrastructural analysis as described previously [13]. Sections were evaluated under a Jeol Jem 1011 transmission electron microscope. The images were transferred to the Soft Imaging System Analysis programme by Megaview III digital camera, and pictures were taken. Each group was analyzed by 2 researchers and observational positive results were given as histopathogical changes in ultrastructures of the glands, glandular epithelium, surface epithelium and its thickness and nuclei, intercellular spaces, Goblet cells, mucus secretion, microvilli, tight junctions, and the basal membrane.

Semiquantitative analysis of cytokines immunohistochemically marked in tissue of all groups was performed by modified H-SCORE, given values between 0–300 for five different regions of each slide [13]. Statistical analysis was performed by GraphPadInstat ver. 3.06 (Graphpad Software Inc., La Jolla, CA, USA). Mean values of continuous variables are compared by one-way analysis of variance and variations between groups were compared by Tukey-Kramer Multiple Comparison Test. Categorical variables were evaluated by chi-square test. P < 0.05, P < 0.01, and P < 0.001 were accepted as statistically significant for different values of the groups.

Light micrographs with H&E staining are given in Fig. 2. As a result of light microscopical evaluations, it was observed that normal structure of the mucosa and submucosa, cryptic glands and other epithelial components of the sham group were not preserved in the control group, as expected. The control group drew attention by damaged mucosal structure, a decrease in the amount of cryptic glands, high levels of epithelial degenerations. Moreover, the thickness of mucosa increased locally due to cryptitis and cryptic apsis, compared to the sham group; leucocyte infiltration and fibrosis were also observed in the submucosa of the control group. On the other hand, the prophylaxis group had similar images of normal mucosal and submucosal components as in the sham group, except for the lesion severity and cryptic distortion. The treatment group showed a decrease in cryptic distortion, as well as in other histopathological parameters except for lesion severity (Fig. 2).

Inflammation, cryptitis, cryptic distortion in lamina propria, reactive atypia of epithelial cells, and inflammatory infiltration were evaluated in H&E stained tissues and lesion severity were graded as 0–3 points. Lesion severity of sham group was found as 0 (no lesion). In control group, its mean was estimated as 2.375 (moderate-severe lesion); 2 (moderate) for treatment group and 1.75 (mild-moderate) for prophylaxis group; it was observed to decrease compared to the control group. However, the difference between the control group and prophylaxis and treatment groups was not statistically significant (Table 1).

There were statistically significant increases in cryptic apsis, cryptitis, and reactive atypia in the control group, compared to the sham group (P < 0.05) while there wree significant decreases in the treatment and prophylaxis groups (P < 0.05). Cryptic distortion showed a significant decrease in the treatment group, compared to the control group (P < 0.05) while there was no significant change in the prophylaxis group (Table 1).

A significant increase was detected for infiltration depth in mucosa of the control group measured by light microscopy, compared to the other groups. It was due to massive lymphocyte infiltration (P < 0.001). There was no significant difference between the sham, treatment, and prophylaxis groups (Table 1), thus, ruscogenin treatment ameliorated the effect of acute proctitis.

H-SCORE analysis of immunoreactivities for inflammation markers of TNF-α, HIF-1α, IL-1β, FGF-2 and angiogenesis marker of VEGF are given in Table 2. Light micrographs of the tissues stained with relevant markers are shown in Fig. 3. Immunoreactivities on rectal tissues were evaluated for three regions: epithelium, crypts, and stroma.

Immunoreactivity of the first cytokine, IL-1β, showed a significant increase in epithelium and stroma of the control group, with regard to other groups (P < 0.05) and treatment and prophylaxis groups indicated significant decreases (Table 2), compared to the control group. But, these differences were not detected for FGF-2 cytokine in the crypts or stroma while the epithelium showed increased FGF-2 immunoreactivity in the control group (Table 2). The prohylaxis and treatment groups indicated significant decreased stainings for FGF-2 cytokine in the epithelium, compared to the control group, but not in other parts of the rectum. Light micrographs also proved these results in Fig. 3.

Immunoreactivity of the other cytokine TNF-α increased significantly only in the epithelium of the control group, compared to the other groups (P < 0.01), and decreased significantly in the treatment and prophylaxis groups (Table 2). Stroma and crypts had no such significant differences of staining (Table 2).

There was a significant increase in immunoreactivity of HIF-1α in epithelium, crypts, and stromal regions of the control group, comparing the other groups (P < 0.001 for epithelium, P < 0.01 for crypts and P < 0.05 for stroma), and significant decreases in the treatment and prophylaxis groups (Table 2). These decreases were also observed in the light micrographs (Fig. 3).

Comparing the number of vessels stained for angiogenesis marker, VEGF in the stroma of the mucosa and submucosa, the control group showed significant increase (P < 0.01) and other groups had significant decreases (Table 2). In other words, the sham, treatment, and prophylaxis groups had no significant differences for VEGF immunoreactivity (Table 2). These findings were also observed in light micrographs (Fig. 3).

Electron microscopical results were parallel to the light microscopy as there were histopathogical changes in the control group related to the acute radiation (Fig. 4). With atrophy and distortion on the glands in the control group, the glandular epithelium was diminished and corroded (Fig. 4B), compared to the sham group (Fig. 4A); migrated nuclei of the epithelium to the lumen side and dilatation of intercellular spaces were observed. The number of goblet cells decreased, mucus secretion diminished, epithelial cells shrank, and there were a few damaged cryptic stem cells. Decreased numbers of microvilli on the epithelium and diminished length, as well as apical vacuolization, were remarkable on the epithelium. On some regions, microvilli were totally deleted, nuclei of the epithelium were enlarged, chromatin modelling differed, and even some gained apoptotic appearance (Fig. 4B).

In the prophylaxis and treatment groups, the epithelium of the rectum was at normal thickness, number and length, as in the sham group; regular crypts were also observed at normal depth (Fig. 4C, D). However, the treatment group had some apoptotic epithelial cells locally. Intercellular spaces were observed at normal distance but with local mild dilatations in both groups. The treatment group had more of these dilatations than the other groups (Fig. 4D2). Number and height of microvilli projecting to the lumen was preserved with their normal structure in these groups (Fig. 4C1, D1), as in the sham group (Fig. 4A1). Many tight junctions protected the integrity of the epithelium. Unlike the control group, prophylaxis group preserved the continuity and homogenicity of the basal membrane and had normal digitations in the intercellular spaces; however, the treatment group lost continuity and homogenicity and had abnormal digitations (Fig. 4D2).