A Novel Mouse Model to Study Image-Guided, Radiation-Induced Intestinal Injury and Preclinical Screening of Radioprotectors Free

Radiotherapy is an important modality in cancer treatment with approximately half of all cancer patients receiving radiotherapy during the course of their treatment (1). As many patients receiving radiotherapy for abdominal or pelvic malignancies experience some form of radiation-induced gastrointestinal injury, there is great interest in understanding the mechanisms behind this pathology and developing effective treatments for it (2). In addition, there is increased concern regarding intentional or accidental exposure of humans to large doses of ionizing radiation due to the onset of acute radiation syndrome (ARS). ARS is comprised of three phases: the hematopoietic phase, gastrointestinal phase, and neurovascular phase in a dose-dependent order (3). Therapies currently exist to manage the hematopoietic phase, although no significant countermeasures exist to address the gastrointestinal or neurovascular phases at this time.

At present, animal models for targeted irradiation (IR) of the small intestine, which mimic clinical patient care, are scarce. This is due to the fact that the prevailing models require either whole body or abdominal irradiation, or require a surgical exteriorization of the small intestine. A more physiologically relevant model, which does not require repeated surgeries, was developed by Hauer-Jensen and colleagues involving transposition of a loop of the ileum to the animal's scrotum, allowing local irradiation of that segment (4). However, this model may only be used on segments of the gastrointestinal tract in close proximity to the scrotum. The model described in this article allows for targeted in situ irradiation of small intestinal segments. Moreover, the model only requires a single, minimally invasive surgery, which is used to place radiopaque markers for image-guided radiation. In addition, multiple fractions of conformally delivered irradiation may be administered without additional surgical intervention. This model is essential for understanding the pathophysiology of radiation injury, and testing potential novel radioprotectors and mitigators to improve cancer radiotherapy and be able to respond to radiologic incidents.

One potential radioprotective agent for the small intestine is curcumin, the principal curcuminoid of turmeric (Curcuma longa). Curcumin has a long history of use in traditional medicine. It is known to have oxygen radical–scavenging abilities as well as anti-inflammatory actions during a damage response (5). We and other groups have shown that it acts as a radiosensitizer in cancerous tissue while acting as a radioprotector in normal tissue (5, 6). Furthermore, it has a low toxicity profile with humans being able to tolerate at least 12 g per day of oral curcumin (7). Although curcumin exhibits low bioavailability in plasma, it shows the highest concentration in the gastrointestinal tract after oral gavage treatment (8). Along these lines, curcumin has been shown to be a radioprotector in the ileum when administered orally (9, 10). Therefore, we tested the ability of curcumin to protect against cellular and tissue-level changes associated with fibrosis, a clinically relevant outcome in radiotherapy of the GI tract.

Our model allows for image-guided, targeted in situ irradiation of short intestinal segments. It constitutes a new tool to rapidly evaluate multiple biological markers of DNA damage in the mouse intestine and we believe it will be valuable in future screens for radiation response modifiers and the study of molecular aspects of IR damage.

Eight to 10-week-old female C57BL/6 mice (The Jackson Laboratory) were maintained in the University of Pennsylvania animal facilities. All experimental procedures were conducted in accordance with the guidelines and protocols approved by the Institutional Animal Care and Use Committee. Mice were subjected to image-guided radiation using the Small-Animal Radiation Research Platform (SARRP) and euthanized to collect small intestine segments for immunofluorescence and histology staining. A cohort of animals received 100 mg/kg of curcumin (Sigma Aldrich, C1386) three days before irradiation and 2 months after irradiation via oral gavage. All mice were injected intraperitoneally (i.p.) with 200–400 μg of 5-ethynyl-2′-deoxyuridine (EdU; Thermo Fisher Scientific C10337) in PBS 3–4 hours before euthanasia. Three or more mice were used in each group. More details are found in Supplementary Material.

Animals were anesthetized using inhaled isoflurane at a concentration of 2%–2.5%. A 1–2 cm midline abdominal incision was made and the jejunum of the mouse was exposed in situ. One side of the radiopaque marker was coated in Medbond Tissue Glue and then rapidly applied to the middle of the jejunum. The peritoneum was sutured using 5-0 PGA-PCL absorbable sutures (McKesson) and the skin was sutured using 5-0 Nylon nonabsorbable sutures (Ethicon). All surgical procedures were conducted aseptically. Mice were administered with 10 mg/kg of meloxicam subcutaneously every 24 hours for 72 hours and with 1 mg/kg of sustained-release buprenorphine at the time of surgery. Animal weights and survival were monitored from the time of surgery until euthanasia.

For these studies, all mice were imaged and irradiated while immobilized with 2.5% isoflurane anesthesia with medical air as the carrier gas (VetEquip). Following 5 minutes in an induction chamber, the fully anesthetized mouse was placed on a customized platform in the ventral recumbent position. The head of the mouse is placed in a face mask that allows the gas to be scavenged (Xerotec Inc.). The platform was placed off the stage of the SARRP (Xstrahl Life Sciences). Using the SARRP Control Interface, a Cone-Beam CT (CBCT) image was initiated with the X-ray tube operating at 65 kV, 0.5 mA with aluminum filtration. The images were reconstructed with Xstrahl's MuriSlice Software. Dosimetry was performed by measuring EBT2 gafchromic film exposure, at depth using solid-water phantom material, with a Microtek Artixcam M1 camera. The dose-exposure calibration was done using a cesium irradiator with calibration dose rate. Utilizing the software, the surgically implanted radiopaque marker was visualized and an isocenter was selected for targeted radiotherapy (Supplementary Fig. S1A). Once the isocenter was determined, the robotic stage moved the animal to the proper location. Final adjustments of the beam delivery angle to avoid the spine were made if necessary. Delivery of the 12 or 18 Gy single dose was made using a 5 × 5 mm collimated beam operating at 175 kV, 15 mA with copper filtration (Supplementary Fig. S1A). The dose rate was 1.65 Gy/minute. All imaging and X-ray irradiations in this study were approved and in compliance with the Institutional Animal Care and Use Committee (IACUC) at the University of Pennsylvania (Philadelphia, PA).

Samples were processed and stained as described previously (11–13). Details are given in Supplementary Material.

Intestinal tissue homogenates and plasma samples were analyzed for cytokine/chemokine levels as described previously (14). Details are given in Supplementary Material.

Whole blood was collected into heparinized collecting tubes (Fisher Scientific, 02-668-10) also containing EDTA, from the tail vein of the mice pre- and 4 days postirradiation. Plasma was obtained by centrifugation for 10 minutes at 1,000–2,000 × g using a refrigerated centrifuge. Plasma samples were aliquoted and stored at −80°C.

Statistical analysis was carried out using GraphPad Prism V software. Data were presented as means ± SEM. Statistical significance was calculated with unpaired Student t test. P < 0.05 was considered to be significant.

In the current study, we aimed to develop a mouse model that would be more pathophysiologically relevant to the intestinal injury sustained during a clinical course of radiotherapy to the abdomen. Unlike current animal models that mostly irradiate the whole abdomen, in clinical practice, patients receive high radiation doses to relatively small volumes of bowel and the radiation injury is focal. One week after the implantation of radiopaque markers, mice were placed under isoflurane anesthesia for CBCT scan and then irradiated with a single dose of 12 or 18 Gy using the SARRP (Fig. 1A). The implantation of the radiopaque marker on the jejunum of the C57BL/6 mice was well tolerated. Of 120 mice that were used in this study, 5 mice (4.2%) died because of intestinal infection and sepsis, while 12 (10%) died of jejunum obstruction. No mortality was associated with anesthesia (mice were generally awoke 2–3 minutes after the isoflurane was removed). To minimize mortality, a small, 1-cm incision in the abdominal area was made followed by suturing both the peritoneum (with absorbable sutures) and epidermis (regular sutures; Fig. 1B). The whole procedure took place under an aseptic environment. Postoperation, the mice lost 2%–9% of their initial body weight, but their weights fully recovered by the time of the CBCT scan/IR (Fig. 1C), 7 days later. Moreover, there was a minor weight loss of the irradiated mice, which did not differ significantly from the unirradiated mice. A few days after irradiation, the mice resumed normal growth (Fig. 1C).

To locate the radiopaque marker for targeted radiotherapy, a CBCT image was obtained using the XStrahl's MuiriSlice Software and reconstructing the images and visualizing the marker on the mouse intestine (Fig. 1D and E; Supplementary Video S1). On the basis of the determined 3D location of the radiopaque marker, we are able to deliver the beam, with its angle adjusted to avoid the spine in the irradiated field (Supplementary Fig. S1B and S1C). The radiopaque marker was shown to be well tolerated by the mice without causing side effects that can impact their survival. For evaluation of the intestinal damage caused by radiation, mice were euthanized at various time points (Fig. 1A) and intestinal segments of 1.5 cm were harvested with the marker located in the middle of the tissue (Supplementary Fig. S2A and S2B). Then, the marker was removed and the intestine segment was placed in either optimal cutting temperature (OCT) medium or formalin for further processing (Supplementary Fig. S2B).

The main objective of this is study was to test whether we could induce localized intestinal damage by radiation. We used several molecular methods to assay direct DNA damage or the biological consequences of DNA damage (e.g., inhibition of proliferation, apoptosis). One of the early subcellular effects after exposure to radiation is the phosphorylation of histone H2AX at serine 139 (γH2AX). Its detection is considered a sensitive and quantitative method for DNA double-strand breaks. Results from the immunofluorescence staining for γH2AX of the frozen tissues harvested at 1 hour post-IR, showed that the DNA damage was localized primarily at the site of the implanted marker (Fig. 2A). The size of the damaged area was 5–6 mm, consistent with the size of the collimator used in this study (5 × 5 mm; Fig. 2A and Supplementary Fig. S3A). It is noteworthy that there was no visible DNA damage on either side of the irradiated area (IR area) and that the γ-H2AX intensity decreased abruptly from the irradiated to the nonirradiated area (Supplementary Fig. S3B and S3C). The late-phase γH2AX foci (≥12 hours) are indicative of chromatin remodeling associated with unrepaired DSBs. To follow temporally the damage of the irradiated area, we performed a time-dependent analysis of the H2AX phosphorylation up to 24 hours post-IR. We found that the intensity of the γH2AX staining in all compartments of the crypt-villus unit was gradually decreased by 6 hours post-IR, which coincides with dsDNA repair, while the signal was almost undetectable by 24 hours post-IR, indicating that there was no persistent DNA damage at the irradiated area at these high doses (Fig. 2B).

We also performed a time course analysis for EdU incorporation from 6 hours up to 4 weeks post-IR. The untreated tissue (control) showed a robust EdU incorporation (Fig. 3A and B). The intensity of the EdU immunofluorescence staining was significantly decreased in the irradiated tissues at 6 hours and 1 day post-IR (Fig. 3A). Representative images of single crypts show that the damaged area is characterized by fewer positive EdU cells as well as with lower intensity of the EdU immunofluorescence signal (Fig. 3B). On the basis of the staining at the latter time points, the induced damage can be characterized as acute because the proliferation of the crypt compartments was completely restored, starting from 4 days post-IR (Fig. 3A). This was also confirmed by the high number of the positive EdU cells per crypt as well as by the percentage of the regenerated crypts (Fig. 3C and D). Also, the damage was restricted mainly at the marker area without affecting the nonirradiated areas (Fig. 3E and F). Results from both γ-H2AX and EdU immunofluorescence staining indicate that the damage caused by radiation was highly restricted to the marker area without causing any significant damage to the surrounding tissues.

Although radiation-induced damage can be repaired by DNA repair mechanisms, apoptosis can occur if the DNA damage above a certain threshold is unresolved. The extent of apoptosis as well as the type of cells damaged in the crypt dictates whether the damage is reversible or can lead to loss of cells and subsequently barrier function and death. Results from the immunofluorescence staining of the intestinal tissues using the TUNEL assay showed that the irradiated areas (on marker) presented a significant increase in apoptotic cells at 6 hours postradiation that remained high at 1-day time point, while apoptosis was barely detectable 4 days post-IR (Fig. 4A and B). This was presumably due to clearance of apoptotic bodies by immune infiltrating cells. Staining of the nonirradiated tissues revealed little, if any, apoptosis and no apparent loss of barrier function (Fig. 4C and D). These results are consistent with those from the EdU incorporation assay showing that radiation induced acute damage only at the marker site at early time points. This damage is temporary and the irradiated area fully recovered as characterized by the lack of apoptotic cells and the high levels of proliferative crypt compartments.

High doses of localized radiation can cause a permanent damage to the epithelial barrier and can lead to compromise of barrier function. In this process, there is activation of the immune system as well as induction of inflammatory responses accompanied by changes in the expression level of cytokines, chemokines, and transcription factors. To investigate the levels of inflammation following the two IR doses, we used a mouse-specific 11-plex cytokine microbead assay to assess the levels of inflammatory response at 4 days and 2 weeks postirradiation. We used whole tissue homogenates of the two different intestinal segments, on marker (OM) and distal to the marker (DM) from irradiated and mock-irradiated mice (Supplementary Fig. S4). The analysis of the OM tissue showed that the levels of 7 of 11 cytokines/chemokines were significantly induced at 4 days post-IR. Among them, IL6 and MIG/CXCL9 showed a 4.3- and 2.7-fold increase (P < 0.01), respectively, followed by a 1.7- to 2-fold increase of the IL10 (P < 0.05), IL12 (p70; P < 0.05), and TNFα (P < 0.01; Fig. 5A). However, the levels of all cytokines subsided to pre-RT levels at 14 days post-IR. The cytokine/chemokine levels of the DM tissue showed that there was no significant inflammation on that area compared with that of the OM tissue at both time points (Fig. 5B).

It is known that late radiation damage in most of the normal tissues is characterized by the loss of parenchymal cells, collagen deposition and excessive formation of fibrous tissue (15–17). Thus, in addition to the acute effects of radiation on inflammatory response, we evaluated its long-term effects on the two intestinal segments at 4 months after 12-Gy irradiation. Our findings revealed focal ulcers with subtending submucosal and serosal fibrosis, crypt hyperplasia, as well as lymphocyte and plasma cell infiltration (Fig. 5C). The lesions were restricted to the marker site and the surrounding tissue was relatively spared, while the DM section was histopathologically similar as the nonirradiated tissue (Fig. 5D). These results demonstrate that both acute inflammatory responses as well as the delayed effects of radiation are restricted to the irradiated (marker site), without affecting the proximal and distal to marker tissues.

We showed that focal radiation caused a significant elevation of proinflammatory and anti-inflammatory cytokines with most prominently induced cytokine at the irradiated area to be IL6. Collagen deposition and fibrosis are the characteristics of the delayed damage around the marker area 4 months postradiation. Curcumin has been found to possess antioxidant, anti-inflammatory, and antitumor activities in a variety of animal models of human diseases. Others and we previously showed that curcumin acts as a radiosensitizer in cancerous tissue while acting as a radioprotector in normal tissue. To check whether curcumin can confer radioprotection to the irradiated intestinal area in our model, mice were treated with 100 mg/kg of curcumin by oral gavage 3 days preirradiation and 2 months post-IR (Fig. 6A). Although, mice treated with curcumin did not present any difference in the EdU⁺ cells/crypt (Fig. 6B), there was a significant decrease in the TUNEL⁺ cells per crypt compared with the mice treated with vehicle (Fig. 6C). Results from the cytokine profile analysis (3-plex assay) on the intestinal lysates (OM only) 4 days post 18 Gy irradiation showed that IL6 levels increased a 6-fold while there was a 4.3-fold increase with 12 Gy. The levels of other tested cytokines were not significantly elevated. However, mice treated with curcumin and irradiation presented a significant decrease (P < 0.05) of the levels of IL6 with no significant variations in the levels of the other 2 cytokines compared with the irradiation-only cohort (Fig. 6D). In addition to the intestinal lysates, we also assessed the effects of curcumin at the systemic level. Blood plasma analysis showed that irradiation (18 Gy) caused an almost 4-fold increase on the IL6 levels at 4 days post-IR, which were significantly reduced after curcumin treatment (P < 0.05; Fig. 6E). It is noteworthy that curcumin treatment also decreased the basal levels of IL6, before IR.

Moreover, we also investigated the role of curcumin on the long-term effects post-IR. The single dose of 18 Gy caused a focal serosal fibrosis at the marker site (OM) compared with the tissue from mock-irradiated mice, as it is indicated by both H&E and Trichrome staining (Fig. 7A and B). Within the wall of this area, there was a focus of granulation tissue (proliferating fibroblasts along with neovascularization) and a mixture of inflammatory cells, lymphocytes, plasma cells and neutrophils. Sections from the mice treated with curcumin contained minimal multifocal areas of serosal fibrosis and there were few (mild) aggregates of lymphocytes, plasma cells, and neutrophils within the lamina propria throughout the section (Fig. 7A–C). These results indicate that administration of curcumin significantly reduces the levels of inflammatory markers and substantially reduced the fibrosis formation as a long-term effect post-IR, suggesting it may be an effective radioprotector in the clinic.

In the current study, we have developed and tested a novel mouse model of image-guided radiation-induced intestinal damage that can be used to assess the efficacy of novel radioprotectors. Using a radiopaque marker to guide the IR beam, we were able to cause focal intestinal damage similar to that encountered in the clinical practice of abdominal radiotherapy. We have also been able to demonstrate cytokine changes and molecular evidence of focal radiation injury in this model, as well as demonstrate a radioprotective effect of curcumin. Radiation-induced intestinal injury is the most important toxicity in clinical abdominal radiotherapy. In a review of 1,160 patients treated with hypofractionated radiotherapy for pancreatic cancer, serious gastrointestinal toxicities consisted of ulcers, hemorrhages, obstructions, structures, and perforations (18). The median late grade 3 or greater GI toxicity was 7.4% (but up to 32.7% in one report). The most frequently reported toxicity was non-hemorrhagic, nonperforated ulcer with a median rate of 9.1 % (range: 3.9%–20%). In another study of Stereotactic Body Radiotherapy (SBRT) for pancreatic cancer, the 12-month rate of G2 or higher GI toxicity was 29% (19). In patients who received 15 Gy or more to greater than 9.1 cm³, the rate increased to 52%. Severe gastrointestinal toxicity was also the dose-limiting toxicity in a phase I/II clinical trial of fractionated chemoradiotherapy for unresectable pancreatic cancer (20).

This model builds upon other existing models of intestinal injury but offers several unique advantages. Prior models to study radiation-induced gastrointestinal injury involve an extensive surgical procedure to exteriorize a segment of the small intestine, placing the animal in a shielded container with only the exteriorized segment exposed, and irradiating the animal (4). While multiple surgical procedures may be conducted in dose fractionation studies with low mortality rates, this requires more resources and has a greater impact on the overall health of the animal when compared with single surgical procedures (21). Another method immobilizes a segment of the small intestine in a sleeve of abdominal muscle. One may then insert temporary sutures and apply tension to these sutures to pull the abdominal wall and attached jejunum through an opening in a lead box, which the animal is shielded in to allow for irradiation targeted to this segment (22). In addition, DeBoth and colleagues developed a rat model involving a Thiry-Vella fistula with both stomas placed in the scrotum (23). While the fistula formation preserves the vascular and mesenteric connections, the isolated loop no longer contains normal intestinal contents, which reduces the physiologic relevance of the model. A more physiologically relevant model that does not require repeated surgeries was developed by Hauer-Jensen and colleagues involving transposition of a loop of the ileum to the rat's scrotum, allowing local irradiation of that segment (4). However, this model may only be used on segments of the gastrointestinal tract in close proximity to the scrotum. In addition, only relatively small lengths of intestine may be translocated to the scrotum. Another limitation of rat models compared with the mouse models is that the latter can be genetically modified to study the molecular mechanisms of radiation-induced damage. Finally, the lower cost of using mouse models, renders them a more feasible model for this type of studies.

The model described in this article allows for targeted in situ irradiation of small intestinal segments utilizing the capabilities of the SARRP. The model only requires a single short surgery, characterized by low mortality rates, which may be used to place multiple markers. Although not tested in this study, our model also allows for multiple fractions of irradiation administered without additional surgical intervention, rendering it a more physiologically relevant mouse model of intestinal injury. The mechanism of acute intestinal injury involves the barrier of the intestinal epithelium becoming disrupted following ionizing radiation exposure. This is mainly attributed to apoptosis of the intestinal stem cells, being unable to repopulate the epithelium (24). Furthermore, late effects of radiation exposure to the bowel include an upregulation of cytokines (25) in the submucosa and the induction of fibrosis in the connective tissue (26).

Another important feature of the model we report herein is the ability to study radiation volume effects. In clinical radiotherapy, it has been appreciated for a long time that intestinal toxicity is influenced not just by the maximal dose delivered to the intestine, but by the volume of intestine irradiated as well (27). However, the exact relationship between dose and volume of irradiated intestine and toxicity, and the mechanisms underpinning this observation remain mostly unknown. We believe our model will be very useful in elucidating these details, which, in turn, would be a very significant contribution to the clinical care of cancer patients. It is interesting to note in this context that in our model, we do not see any mortality when we induce a focal intestinal injury with high radiation doses (i.e., up to 24 Gy). This is very different that the mortality observed after whole abdominal radiation to much lower doses (typically LD50s are around 14 Gy).

Although radiosensitization is important to enhance tumor control, equally important is normal tissue toxicity minimization. Clinically approved radioprotectants are currently limited, with amifostine being the most potent and widely used clinical agent for head and neck cancer by scavenging the free radicals (28, 29). Recently, curcumin has been shown to confer radiosensitizing and radioprotective activities (10, 30), but the detailed mechanisms of curcumin-mediated radioprotection have not been clearly elucidated. In our study, we used curcumin treatment for 3 days pre-IR and 2 months post-IR. Mice treated with curcumin presented significant decrease in apoptotic cells and changes in both levels of inflammation as well as in long-term effect markers. Specifically, the levels of IL6 in both intestinal lysates and blood samples were significantly reduced after curcumin treatment. H&E staining of intestinal sections revealed that mice treated with curcumin presented a significantly lower number of areas of serosal fibrosis compared with the control group that showed extensive fibrotic areas on the marker site along with aggregates of lymphocytes, plasma cells, and neutrophils. So far, administration of curcumin has been shown to significantly reduce the levels of inflammatory markers and substantially reduced the fibrosis formation as a long-term effect post-IR. Taken together, these data along with the fact that curcumin is characterized by low toxicity profile, as it has been shown in phase I and II clinical trials (31), make it a potent radioprotector.

In conclusion, we developed a novel mouse model as a new tool for precise and quantitative evaluation of multiple biological markers of radiation-induced DNA damage in short segments of the intestine. Results from curcumin treatment indicated significant radioprotection in both acute and long-term effects. Thus, this model can also be used for future screens of radiation response modifiers.

No potential conflicts of interest were disclosed.

Conception and design: I.I. Verginadis, B. Bell, E. Ben-Josef, C. Koumenis

Development of methodology: I.I. Verginadis, R. Kanade, B. Bell, E. Ben-Josef, C. Koumenis

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): I.I. Verginadis, R. Kanade, B. Bell, S. Koduri, C. Koumenis

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): I.I. Verginadis, R. Kanade, B. Bell, E. Ben-Josef, C. Koumenis

Writing, review, and/or revision of the manuscript: I.I. Verginadis, E. Ben-Josef, C. Koumenis

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): I.I. Verginadis, E. Ben-Josef, C. Koumenis

Study supervision: C. Koumenis

This work was supported by Development funds from the Department of the Radiation Oncology, Perelman School of Medicine, University of Pennsylvania (Philadelphia, PA). I.I. Verginadis is supported by a postdoctoral fellowship from the Bodossaki Foundation (Greece).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.