Background
Inflammatory bowel disease (IBD), encompassing Crohn’s disease (CD) and ulcerative colitis (UC), is recognized as a widespread, debilitating condition with increasing incidence in Western societies in both children and adults [
1‐
3]. The natural history of IBD is characterized by relapse and remission, with several factors known to trigger relapses including infection, ingestion of non-steroidal anti-inflammatory drugs and changes in smoking habits [
4]. The aetiology of IBD is still not fully understood, despite decades of extensive research. It is believed that the imbalance of pro-inflammatory and anti-inflammatory cytokines contributes to the development of colitis [
5‐
7].
Interleukin-1β, primarily secreted by monocytes and macrophages upon activation is one of the main drivers of inflammation. Macrophages are recruited and activated from peripheral blood into the inflamed colon [
8,
9]. IL-1β stimulates the production of inflammatory eicosanoids that subsequently induce neutrophil - chemoattractant and neutrophil-stimulating [
10]. Released mature IL-1β protein resulting from inflammatory stimulus at the injured tissue, together with other cytokines and mediators (e.g. oxygen radicals) cause a cascade of inflammatory responses and tissue damage [
11,
12]. The binding between IL-1 and IL-1 receptor activates the NF-κB signal-transduction pathway [
13], resulting in the up-regulation of other pro-inflammatory mediators such as TNF, IL-6 and IL-12 [
14]. IL-1β is one of the key mediators of intestinal inflammation in IBD with a role in amplifying mucosal inflammation [
15,
16], consistent with the finding that IL-1β is up-regulated in IBD patients [
17] and animal models [
18,
19]. On the other hand, IL-1β receptor antagonist reduces inflammation in an animal model of colitis [
18,
19].
IL-1β in inflamed intestine is mainly produced by infiltrating lamina properia monocytes including macrophages in the IBD mucosa [
16]. However, IL-1β can also be produced by intestinal smooth muscle cells and fibroblasts [
20]. The precise source of IL-1β producing cells in our animal model will be investigated in our future experiment.
Animal models of experimental colitis have been useful in confirmation of these clinical observations [
11,
21,
22]. Furthermore, developing a method to monitor real time IL-1β activity
in vivo would provide a unique opportunity to assess the precise progression of intestinal inflammation, using a DSS induced colitis model.
In this paper, colitis was induced using dextran sodium sulfate (DSS) in a cHS4I-hIL-1βP-Luc transgenic mouse, in which the expression of luciferase reporter gene was under the control of the human IL-1β gene promoter [
23,
24]. A “biophotonic” imaging system equipped with a highly light-sensitive camera allows non-invasive study of the transcriptional activity of IL-1β gene promoter in real time during the development of IBD, which could be used to evaluate the effects of anti-inflammatory compounds on IL-1β gene induction
in vivo.
Methods
Genotyping of cHS4I-hIL-1βP-Luc transgene in mice
cHS4I-hIL-1βP-Luc transgenic mice, generated in the C57/B6 × CBA background [
23,
24], were backcrossed to C57/B6 for 3 generations before the experiment. Transgenic founders and their offsprings were identified by PCR using the forward-luc (5
′ TTCCGCCCTTCTTGGCCTTTATGA 3
′) and reverse-luc (5
′ CAGCTATTCTGATTACACCCGAGG 3
′) primers specific for the luciferase gene. All animals were housed under conventional laboratory conditions with food and water
ad libitum. Experiments adhered to the guidelines of the local institutional animal care and use committee.
Induction of colitis
Adult (10 week old, male) cHS4I-hIL-1βP-Luc transgenic mice were given
ad libitum 3% w/v dextran sulphate sodium (DSS, MW 36 000–44 000; MP Biomedicals, CA, USA) dissolved in tap water for four consecutive days, as described [
11,
12,
22], while control groups received tap water only. On day five, the DSS solution was replaced with water, allowing some degree of colonic epithelial recovery. To confirm that the luciferin activity was inflammation specific, the mice were challenged with 3% DSS in drinking water and also dexamethasone (St. Louis, MO, USA,1.5 mg/mg) i.p. daily for five days. The luciferase signal was imaged and compared with that of the control group, which was injected with saline.
Assessment of the extent of experimental colitis
To confirm the severity of the colitis model, DSS-induced colitis was evaluated by body-weight and stool score daily. Weight loss on each day was calculated as the percentage of the baseline of bodyweight. Blood loss and stool consistency were scored, based on our previous description [
11,
12,
22]. Scores were defined as follows: Blood loss: 0 = negative, 2 = positive, 4 = gross bleeding. Stool consistency: 0 = normal, 2 = loose stools, 4 = diarrhea.
Histopathologic analysis
For histopathologic analysis at day 6 (2 days after the end of DSS challenge), transverse colon was collected and fixed in 10% buffered formalin phosphate, embedded in sucrose, frozen in dry ice using optimal cutting temperature (OCT) compound and cryosectioned. Cross sections were stained with hematoxylin/eosin (H&E, Lerner, New Haven, CT). Histopathological scores were used to quantify the intestinal inflammation, as described previously [
11,
12,
22].
In vivo and ex vivo imaging
In vivo bioluminescent imaging was performed using an IVIS imaging system (Bio-Real, QuickView3000, Austria). At the selected time points, the mice were anesthetized with isoflurane/oxygen, then were injected i.p. with substrate luciferin (Biosyth, Basel, Switzerland) dissolved in PBS (15 mg/ml) at a dose of 150 mg/kg. After 12 min of luciferin injection, images were taken on the imaging stage for 1–5 min. Photons emitted from specific regions were quantified using a LivingImage software (Bio-Real, QuickView3000, Austria). In vivo luciferase activity was presented in photons emitted per second.
For ex vivo imaging, the experimental mice were scarified at the days 4, 6 and 8 post DSS challenge. The colon, MLN and spleen were collected and were imaged with the same settings used for the in vivo studies on a heated stage in the IVIS system.
Statistics
All data are expressed as means ± SE. The data were analysed by one-way ANOVA. A P value of < 0.05 was considered significant.
Discussion
Our DSS-induced model of colonic inflammation displays pathological and clinical similarities to human colitis and has been widely used for pharmacological and pathophysiological studies [
11,
21,
25,
26]. DSS induced colitis results from the influx of bacteria into the lamina propria, due to an alteration of the colonic inner mucus layer [
11]. The inflammatory cytokine, IL-1β, plays a major role in intestinal inflammation development and increased dramatically in the colonic mucosa during disease [
27,
28]. Our results are in agreement with these findings, showing an increase in luciferase expression driven by IL-1β gene promoter in the inflamed colon. The luciferase activity was significantly higher in the intestine, and relatively higher expression levels were also seen at the position of the mesenteric lymph node. These
ex vivo data were consistent with previous studies.
Conventional methods for monitoring IL-1β gene expression rely on either measuring circulating levels of IL-1β in the serum or mRNA expression in tissues. Compared with these methods, the approach reported in this study is convenient and sensitive, while allowing less animals to be used. Moreover, this approach offers kinetic quantification and information on the anatomical distribution of IL-1β gene expression.
IL-1β, scarcely distributed throughout the abdominal region prior to challenge, was up-regulated and broadly distributed, but most densely observed near the centre/left lower quadrant of the abdomen on day 1, which supports that DSS gradually induced intestinal inflammation. The increased IL-1β appeared close to plateau by day 6 with the distribution seemingly shifted to the proximal colon, which is line with induced colitis present after 5 days of DSS challenge. Condensed but strong distribution was observed on days 7 to 9, but then slowly decreased on days 11 to 13, accompanied by more diffuse distribution (Figure
2). Our
ex-vivo data showed that IL-1β was highest in the colon at day 6, but declined at day 8 (Figure
3), supporting that the main source of induced IL-1β is in the inflamed colon
in vivo. A similar pattern of IL-1β in MLN
ex vivo suggests IL-1β producing leucocytes migrated into the draining lymph nodes. No significant increase present in the spleen suggests that DSS induced acute colitis focuses on the gut rather than systemic inflammation.
Despite DSS challenge ceasing on day 5, IL-1β activity continued to increase, peaking at day 9 following DSS challenge, consistent with faecal and blood scores over this time frame. This data suggests that IL-1β activity correlates with the severity of colitis, confirmed with histopathological findings, making the bioluminescence model a reliable method for monitoring inflammation.
Despite the robustness of our bioluminescence model, we acknowledge that there remain a number of limitations. Currently, this model has limited resolution and cannot pinpoint the exact source of IL-1β at the cellular level. Furthermore, it doesn’t provide information about IL-1β translation and whether this correlates with observed transcriptional changes. In future experiments tissues will be collected at different time points for detection of IL-1β protein, using Western blot and immunohistochemistry to confirm the relationship between transcription and translation.
The observed dexamethasone dependant reduction of IL-1β expression suggests that this model can be used to evaluate the efficacy of medical therapy by showing decreased expression of inflammatory mediators. The bioluminescence/IL-1β model could be also used to study experimental therapies in IBD [
29] such as faecal transplantation [
30], which is an alternative therapy in treatment of IBD with promising outcomes.
Acknowledgements
We thank Geneway International Trading Co. Limited for technical support. This work was supported by grants from the National High Technology Research and Development Program (2008AA02Z126), National Key Project (2010CB945501), Science and Technology Commission of Shanghai Municipality (10410703800, 10DZ2251500, 09ZR1422600) E-Institutes of Shanghai Municipal Education Commission (E03003).
Competing interests
All the authors declare that there is non-financial competing interest.
Authors’ contributions
LL and ZL performed the major part of the experiment. XY and HY performed partial live imaging experiments. SB contributed intellectual input to the experiments and wrote manuscript. JF designed the experiments and coordinated the project. All authors read and approved the final manuscript.