Diagnostic imaging advances in murine models of colitis

Abstract

Inflammatory bowel diseases (IBD) such as Crohn’s disease and ulcerative colitis are chronic-remittent inflammatory disorders of the gastrointestinal tract still evoking challenging clinical diagnostic and therapeutic situations. Murine models of experimental colitis are a vital component of research into human IBD concerning questions of its complex pathogenesis or the evaluation of potential new drugs. To monitor the course of colitis, to the present day, classical parameters like histological tissue alterations or analysis of mucosal cytokine/chemokine expression often require euthanasia of animals. Recent advances mean revolutionary non-invasive imaging techniques for in vivo murine colitis diagnostics are increasingly available. These novel and emerging imaging techniques not only allow direct visualization of intestinal inflammation, but also enable molecular imaging and targeting of specific alterations of the inflamed murine mucosa. For the first time, in vivo imaging techniques allow for longitudinal examinations and evaluation of intra-individual therapeutic response. This review discusses the latest developments in the different fields of ultrasound, molecularly targeted contrast agent ultrasound, fluorescence endoscopy, confocal laser endomicroscopy as well as tomographic imaging with magnetic resonance imaging, computed tomography and fluorescence-mediated tomography, discussing their individual limitations and potential future diagnostic applications in the management of human patients with IBD.

Key Words: Confocal laser endomicroscopy; Contrast enhanced ultrasound; Dextran Sodium Sulphate colitis; Experimental colitis; Fluorescence imaging; Endoscopy; Imaging; Inflammatory bowel disease; Tomography

Core tip: Murine models of experimental colitis are a vital component of research into human inflammatory bowel disease (IBD). Recent advances mean revolutionary non-invasive imaging techniques for in vivo murine colitis diagnostics are increasingly available. These techniques not only allow direct visualization of intestinal inflammation and enable molecular imaging of the inflamed mucosa but also allow for longitudinal evaluation of intra-individual therapeutic response. This review discusses the latest developments in the different fields of (molecularly targeted) contrast agent ultrasound, fluorescence endoscopy, confocal laser endomicroscopy as well as tomographic imaging with fluorescence-mediated tomography, discussing their potential future diagnostic applications in human IBD.

INTRODUCTION

Inflammatory bowel diseases (IBD) such as Crohn’s disease (CD) and ulcerative colitis (UC) are chronic-remittent inflammatory disorders of the gastrointestinal tract characterized by symptoms such as diarrhoea, abdominal pain or anaemia[1-3]. The course of IBD can be complicated and relapsing, involving challenging clinical diagnostic and therapeutic situations, which may often lead to hospitalization or surgery[4,5]. As the pathogenesis of IBD is still incompletely understood, current therapeutic regimes often are aimed at unspecific suppression of the adaptive immune system[6,7]. The therapeutic armamentarium of IBD was significantly advanced by the advent of antibodies directed towards tumor necrosis factor-α (TNF-α)[8-10]. Subsequently, various novel molecular targets have been identified and antibodies including anti-integrins and anti-interleukins have been tested for the treatment of IBD patients in clinical trials[11-13]. However, despite scientific advances in medical treatment, the success of such anti-inflammatory drugs remains hampered by potentially serious side effects including the increased risk of opportunistic infections and bone marrow suppression as well as limited long-term efficacy[14]. Furthermore, over 50% of CD patients still need surgery after 10 years of disease[15,16]. Therefore, novel therapeutic approaches are needed.

Evaluation of potential new drugs is usually performed in experimental models of colitis since sophisticated models are needed to mimic the complex pathogenesis of human IBD. Due to the abundant genetic and genomic information known and its availability in transgenic and knockout strains, the laboratory mouse mus musculus is the preferred animal for colitis research models[17]. These models are frequently used for evaluation of drug candidates as well as characterization of novel preclinical diagnostic or therapeutic approaches[18,19].

Daily monitoring of the murine body weight or determination of disease activity by appropriate indices[20,21] are classical indirect indicators for the severity of colitis and therapeutic response in vivo. As measurement of faecal and serum markers of inflammation[22-24] may not exactly reflect the course of the colitis, in humans or mice, histological analysis of the inflamed colon still remains the most valid approach for determining the severity of colitis[25]. Unfortunately, in mice, histological analysis of the colon commonly requires post mortem analyses, requiring death of the animal, and therefore do not allow for longitudinal observations at repetitive time points. Non-invasive and imaging-based methods to assess the course of experimental colitis are a promising approach to overcome this limitation.

This review summarises current crucial advances of promising imaging techniques to monitor disease activity in murine models of colitis in vivo. Recent developments in ultrasound- and endoscopy-based modalities as well as cross-sectional imaging are presented here. We also highlight the potential of these advances in diagnostic imaging in murine models of colitis to improve diagnosis in human IBD patients whilst also discussing their limitations.

ULTRASOUND

The quality of small animal high-frequency ultrasound and transducer techniques has dramatically improved with emerging research concerning interdependent sonographic criteria resolution, frequency and penetration[26-28]. Advanced linear transducer array technology with around 15 to over 50 MHz allowing B-mode, Doppler-mode, sophisticated contrast-agent and 3D ultrasound was broadly available since 2009[29]. As this technique facilitates imaging of small anatomical structures of about 50 μm[26], it enables the visualisation and examination of murine intestine, meeting criteria of guideline-directed human bowel ultrasound[30]. Key sonographic features of human IBD detected in B-mode are increased bowel wall thickness and loss of bowel wall stratification[31,32]. These features are also detected with ultrasound of the colon of mice suffering from chemically induced Dextran Sodium Sulphate (DSS)-colitis and correlate strongly with established clinical parameters like weight loss[33]. Standardised examination protocols concerning comparable positioning of the animal, the transducer, anatomic landmarks and preparation of the colon with enema, help to reliably assess the course of murine colitis in vivo both intra-animal by longitudinal time course and inter-animal[34].

High-frequency power Doppler provides an additional in vivo ultrasonographic examination method, enabling examination of the murine vascular system. Moreover, high-frequency power Doppler in native small-animal ultrasound has already been used in inflammatory disorders[35] and can detect colorectal intraluminal tumours[36], quantify tumour volume and proportion of vascularisation[37,38] and may furthermore assess blood flow velocity[39].

As an advancement of classical B-mode technique, nonlinear contrast enhanced ultrasound (CEUS)[27,40] has further improved the sensitivity and specificity of small animal ultrasound, for instance in the detection and quantification of pathological vascular growth in tumour angiogenesis[41]. Ultrasound contrast agents are gas-liquid emulsions consisting of a biocompatible protein or lipidic shell and are filled with gas (perfluorocarbon, sulfur hexafluoride, or nitrogen)[42,43] which further enhance ultrasonic signal strength by displaying different acoustic properties than tissue or blood. With their micron size (usual diameter between 1-4 μm) and well-engineered technical features, microbubbles (MBs) can pass through microcapillaries like the pulmonary capillary bed and transiently persist in the blood stream[44,45]. Sophisticated ultrasound-based molecular imaging can therefore be used as a targeted modality, for example with the application of small molecules such as peptides or antibodies bound to the outer shell of MBs used as molecular contrast agents[46]. By using special software programmes, the ultrasound signals of contrast agent targeting molecular structures such as endothelial cells or intravascular blood cells of interest can be selectively enhanced in tissues or cells expressing the target molecules[42,47].

Regarding ultrasound assessment of experimental murine colitis at a molecular level, various endothelial cell adhesion molecules critically involved in leukocyte homing in intestinal inflammation have been evaluated as promising targets for contrast agents. Deshpande et al[48] showed in vivo the ultrasound signal in mice suffering from 2,4,6-trinitrobenzene sulphonic acid (TNBA)-induced colitis to be significantly higher using anti-P-selectin-targeted MBs than using control MBs, which was corroborated by measurement of mucosal P-selectin expression levels ex vivo. Additionally, expression of tissue- and disease-specific adhesion molecule mucosal addressin cellular adhesion molecule-1 (MAdCAM-1) used as a target molecule[49] is massively increased in the inflamed small bowel and colon of human IBD patients[50] and in murine experimental colitis[51,52]. Our group has also demonstrated earlier CEUS with anti-MAdCAM-1-labelled MBs to accurately measure the course and severity of DSS-colitis over time. Furthermore, MAdCAM-1 expression as assessed by CEUS correlated strongly with clinical parameters of colitis such as weight loss[53].

A benefit of ultrasound-based molecular imaging is that MBs directed against molecular targets are delivered direct to highly localized anatomical regions using small molecules or plasmids[54]. Different methods of attaching drugs to MBs have been reviewed in detail elsewhere[55]. With the method of sonoporation for example, high-pressure ultrasound can lead to destruction of MBs with local release of encapsulated drugs into the microcirculation, however on the other hand, ultrasound itself increases cell permeability enhancing transmembrane transport of therapeutic molecules[56,57]. In this context, Tlaxca et al[58] report about successful specific luciferase plasmid delivery to mesenteric vasculature via MBs targeting MAdCAM-1 and vascular cell adhesion protein-1 in a murine model of CD.

For the future, the evaluation of optoacoustic imaging or multispectral optoacoustic imaging (MSOT) as a novel imaging tool to diagnose murine colitis is desirable. This technique is based on the absorption of ultra short light pulses that lead to the generation of acoustic waves, which in MSOT have extended the spectral identification of chromophoric molecules in tissues, and shows good results for depth resolution, accuracy and visualisation of biochemical processes[59]. Other applications reported are the detection of pancreatic lesions via upregulated epidermal growth factor[60] or imaging of pharmacokinetics in multiple different organs[61].

While ultrasound allows visualization and serial follow-up examinations in animals with colitis, it cannot confirm the complete absence of mucosal inflammation. However, so-called “mucosal healing” is shown to be predictive of a beneficial disease course in human IBD patients[62,63] and has emerged as a goal of treatment[64]. Therefore, endoscopy-based approaches, which may facilitate a direct assessment of the murine mucosa, represent an adjunct imaging modality.

ENDOSCOPY-BASED APPROACHES

In vivo endoscopic imaging in murine models of colitis has the potential to merge both the pathogenic understanding of human IBD by continuous monitoring of the course of disease and the specific diagnostic imaging of the mucosa. White-light colonoscopy allows repeated examination, including biopsy sampling, of one mouse following the course of disease. Becker et al[65,66] first described a high-resolution chromoendoscopic system, which enables monitoring of DSS-colitis and also colitis-associated cancerogenesis. An objective score has also been established to describe intestinal inflammation in murine models of IBD - the MEICS-Score[67]. Parameters included in this score, are: thickening of the intestinal wall, vascular pattern, presence of fibrin, granularity of the mucosal surface and also stool consistency[67].

By allowing repeated examinations and intra-individual follow up, white light endoscopy allows examination of the effect of therapeutic substances on colitis activity and may lead to further clinical studies in humans[68-71].

Additionally, white-light endoscopy in murine colitis may also be used to evaluate intestinal wound healing, administer diagnostic or therapeutic agents and obtain mucosal biopsies[72]. There has been great progress in the development of novel endoscopic techniques[73]. Video-endoscopy including magnification and high-resolution endoscopy[74], chromo-endoscopy[75], narrow-band imaging[76], auto-fluorescence imaging[77], confocal laser endomicroscopy (CLE)[78,79] and optical coherence tomography[80] are some of the recent advances that have shown promise in surveillance examinations. Regarding small animal imaging and its translation from basic science into clinical application, fluorescence imaging, fluorescence endoscopy (FE) and CLE are the most promising candidates. The latter, CLE, has been established as a new imaging technique and virtual histopathology in vivo has also now been realized[81-83]. In addition to morphological characterization[84,85], CLE also allows immunohistochemistry to be performed in vivo. Foersch et al[86] have demonstrated the feasibility of specifically targeting vascular endothelial growth factor expression in gastrointestinal tumors in vivo with labeled antibodies and CLE. Similarly, Goetz et al[87] were able to use fluorescently labeled epidermal growth factor receptor (EGFR) antibodies to examine EGFR expression patterns and consequently diagnose colorectal cancer and predict response to targeted therapy. Visualization of EGFR expression in vivo may also help to identify promising drug candidates for anti-angiogenic treatment. In experimental colitis, CLE has been successfully applied to identify intra-mucosal bacteria in vivo[88], to monitor the course of disease[89] and to determine the presence of intestinal cancer stem cells in colitis associated tumorigenesis[90]. Moreover, nanoparticles such as avidin-nucleic acid nano-assembly have also been used to define chronically inflamed intestinal mucosa and to characterize murine models of IBD in vivo. Buda et al[91] characterized the mucosal surface with CLE and nanoparticles demonstrating them to be only observable in the inflamed and not healthy mucosa. Confirming the significant potential of CLE for characterizing intestinal inflammation in human IBD patients, Atreya et al[92] recently used fluorescently-labeled antibodies to quantify membrane-bound TNF-α positive intestinal immune cells and were able to predict treatment response towards an anti-TNF-α therapy in CD patients.

FE of specific molecular targets with near-infrared light allows in vivo molecular imaging with simple photographic methods[93,94]. This method employs the topical application of fluorescein-labeled monoclonal antibodies. For example, by the use of antibodies directed against carcinoembryonic antigen on the mucosal surface, colorectal carcinoma detection during conventional colonoscopy was possible with a specificity of 100% in human patients[95]. Confirming it’s high diagnostic potential, local sensitization with hexaminolevulinate, a selective photosensitive precursor, FE enables early detection of colonic premalignant conditions during conventional colonoscopy in patients with a reported specificity of 72%[96]. In a murine model of colon carcinogenesis, fluorescently-labeled biomarkers were used to observe the tumor microenvironment by FE over time[97]. Furthermore, a cypate-labeled cyclic peptide sequence detected flat colorectal lesions by FE in colitis-associated cancer[98]. Other groups have developed fluorescently labeled peptides to mark colitis-induced adenoma[99,100].

As a “red flag technique” FE may allow screening of the large mucosal surface for early colorectal cancer, which may be especially helpful in UC patients who are at an increased risk of colitis-associated cancer[101]. An overview of different antibodies used in FE and CLE, their mode of application and target murine model is provided in Table 1.

Table 1 Tracer-antibodies used in fluorescence endoscopy and confocal laser endomicroscopy.

Publication	Tracer	Application	Method	Mouse model
Charanya et al[98], 2014	Cypate- cyclic peptide sequence, D-Cys-Gly-Arg-Asp-Ser-Pro-Cys-Lys [c(DCGRDSPC)K]	Topical	FE	AOM-CRC, mice
Gounaris et al[157], 2013	ProSense 680	Intravenous	FE and Fluorescence	IL-10-/- mice, mice
			Reflectance Imaging
Joshi et al[99], 2012	Cy5.5-AKPGYLS peptide multimer	Topical	FE	CPC; Apc mice
Miller et al[100], 2011	FITC-Ahx-QPIHPNNM peptide	Topical	FE	CPC; Apc mice
Urano et al[158], 2011	gGlu-HMRG	Topical	Laparoscopy + FE	diss. Ovarial-Ca, mice
Foersch et al[86], 2010	VEGF-Ab	Topical	Colonoscopy + CLE	APCmin mice, Stomach cancer
Goetz et al[87], 2010	EGFR-Ab	Intravenous	CLE	Balb/c nu/nu mice, CRC-cells implanted
Funovics et al[159], 2006	Cy5.5 nano particle u.	Intravenous	Colonoscopy + FE	APCmin mice, CT26 (CRC) in nu/nu
	AF750-Beacon

FE: Fluorescence endoscopy; AOM: Azoxymethane; CRC: Colorectal cancer; IL-10: Interleukin 10; Cy5.5: Cyanine 5.5; CPC: Cdx2-Cre transgene; Apc: Adenomatous polyposis coli; FITC: Fluoresceinisothiocyanat; Ahx: Aminohexanoic acid, HMRG: Hydroxymethylrhodamine green; VEGF: Vascular endothelial growth factor; CLE: Confocal laser endomicroscopy; EGFR: Epidermal growth factor receptor; Balb: Albino laboratory-bred mouse strain; AF750: Alexa fluor 750; CT26: Cancer cell line.

Nevertheless, since the diagnostic capacity of endoscopy is restricted to endoluminal findings and assessment of single molecular targets of the colonic wall, assessment of colon wall thickness or even extra-intestinal findings, cross-sectional imaging techniques including magnetic resonance imaging (MRI) and computed tomography (CT) can provide additional information.

MRI AND CT

In human patients with IBD, colonoscopy represents the gold standard to assess luminal colonic inflammation. However, MRI has been established as a non-invasive imaging technique[30]. While possessing a limited sensitivity and specificity for the detection of colonic inflammation[102,103], it can detect extramural lesions as well as IBD-related complications (e.g., stenosis abscesses and fistula). In contrast to endoscopic procedures, MRI imaging does not confer a risk of endoscopy-associated complications such as infection, perforation or hemorrhage[104-106].

In experimental colitis, MRI was first described to be as useful to assess disease activity in TNBA-treated rabbits[107]. Subsequently, Larsson et al[108] established MRI to investigate experimental colitis in the DSS-colitis model, using colonic wall thickness, contrast media-enhanced T1-weighted (T1_W)[109] as well as T2-weighted (T2_W) signals to detect intestinal inflammation in accordance to the diagnostic approach used in human patients with IBD[110-113]. Strong correlations between MRI measurements and clinical and histological disease activity were found. Michael et al[114] proposed to quantify colonic inflammation by assessment of colon wall thickness and tissue density factor. However, to achieve this, the 3D MRI with respiratory triggering required administration of spasmolytic substances to inhibit breathing and intestinal motility which was a complex and time-consuming procedure, resulting in a complete MRI examination requiring 20-30 min per animal[108]. Consequently, Melgar et al[115] established a 2D rapid T2_w MRI protocol to provide characterization of the intestinal inflammation in under seven minutes which also avoided the need for respiratory triggering or antispasmodics. Whilst this protocol accurately detected inflammation in colonic segments below the bladder, in proximal colonic segments the MRI findings and inflammatory parameters only correlated weakly, limiting the applicability. An additional limitation of 2D acquired data sets is their prerequisite for perfect alignment of the colon along the coronal plane[115].

However, recently there is a growing body of evidence supporting the utility of MRI for the assessment of intestinal inflammation, using colon wall thickness, contrast enhancement and T2_w sequences[78,116-121]. Novel parameters like dynamic contrast enhancement[117,118] are also being included and advanced criteria including target sign patterns and intramural hemorrhage detection have been introduced to characterize experimental colonic inflammation[119]. Pohlmann et al[78] were able to follow the course of TNBS-induced colitis in rats and the therapeutic response after anti-inflammatory treatment with sulfasalazine with repeated measurements of colonic wall thickness and T2_w sequences. Colitis is characterized by inflammation-mediated edema occurring by immune cell infiltration, vasodilatation and increased capillary permeability, which eventually result in thickening of the colonic wall[20,108,122]. Therefore, MRI measurements of colon wall thickness and T2_w signal intensity reflecting edema can be very sensitive for murine colon inflammation, however, these parameters are also unspecific biomarkers of inflammation. To improve the specificity, Frericks et al[123] administered ultra-small superparamagnetic iron oxide particles intravenously which accumulate in the reticuloendothelial system, differentiating IBD from unspecific causes of edema.

Bowel wall thickening and increased T2_w signal intensity are also core features of intestinal fibrosis and partly correlate with the degree of fibrosis[124]. Therefore, MRI may also be used to detect fibrosis in murine models as well as inflammation. However, it remains challenging to accurately distinguish between persisting inflammation and fibrosis, a problem exacerbated by the fact that both are frequently simultaneously observed[125-127]. Adler et al[126] proposed a possible solution to this involving the application of magnetization transfer ratios in MRI not only to detect intestinal fibrosis, but also to show fibrotic progression over time. These results have been similarly reproduced in a different rat model of colitis[128], and the technique has also been used to monitor therapeutic response of anti-fibrotic therapy in rats[129]. Furthermore, Breynaert et al[127] found a regression in MRI T₂ relaxometry over the course of several DSS-cycles that might be used to differentiate early inflammation from fibrosis.

According to current guidelines, CT has high accuracy for the assessment of specific clinical questions, for example, to confirm the presence of suspected penetrating complications or abscesses in human patients with IBD[30]. Interestingly, only a few studies solely evaluate CT for the assessment of intestinal inflammation in experimental colitis. Recently, Fredin et al[130] demonstrated the diagnostic potential of micro-CT (μCT) to detect and monitor the course of inflammation and also therapeutic response by assessing colonic wall thickness. Moreover, based on the CT findings, the authors were able to differentiate between responders and non-responders of anti-inflammatory treatment in early acute colitis. However, since the thickness of the colonic wall remains unchanged during the healing phase of colitis, the applicability of this read-out is restricted to the early stages of colitis. Furthermore, to obtain an image of sufficiently good quality during small animal CT-examinations, 10-fold higher dosages are required for animals compared to human CT-studies[131], which could potentially interfere with the inflammatory model being examined[132], especially when imaging is performed longitudinally for disease monitoring. Several clinical studies in human patients have confirmed CT enterography for detecting inflammation in the small bowel during CD is highly accurate[133-136]. Bodily et al[137] showed a strong correlation of mural enhancement and wall thickness indicating active small-bowel inflammation, which could be confirmed by histological analysis. However, to date no studies have been performed in experimental colitis[138-140].

Recently, to utilise the favourable spatial resolution of CT with a functional imaging modality to more specifically detect intestinal inflammation, μCT with positron emission tomography (PET) was introduced. Activated leukocytes accumulate ¹⁸F-fluorodeoxyglucose (FDG), which can be traced by PET[141]. ¹⁸F-FDG PET/CT has shown promising results for clinical assessment of IBD patients in preliminary studies[142-145]. Since acute DSS-colitis is neutrophil-mediated[146], our group have employed ¹⁸F-FDG PET/CT for detecting and monitoring DSS-induced intestinal inflammation which had a strong correlation with established inflammatory parameters, endoscopic findings and histological analysis[147,148]. Furthermore, Brewer et al[149,150] demonstrated increased CD4+ T cell uptake of either ¹⁸F-FDG or ¹⁸F-1-(2’-deoxy-2’-arabinofuranosyl)cytosine (D-FAC) in chronic intestinal inflammation, a method which could be utilized for the detection of colitis. In line with those findings, another transfer colitis model was successfully used for monitoring intestinal inflammation[151]. Most recently, PET/CT was performed on a large scale in 70 mice and the findings compared to those using contrast agent-guided ultrasound for detection of intestinal inflammation[152]. One limitation of imaging colitis by PET/μCT is an obscured pelvic region because of tracer accumulation in the bladder. However, these artefacts can be minimised by emptying the bladder of the animal manually using gentle pressure[149,150] or with use of continuous bladder flushing[153]. Further studies are needed to evaluate the possible influence of radiation during the monitoring of inflammatory disease[132] and to improve limited detection rates of minor mucosal alteration[142,148].

In addition to MRI and CT, FMT can be used to characterize biological processes in experimental colitis. With the use of labelled antibodies that specifically target mucosal colitis markers or distinct cell populations on a molecular level, use of FMT to quantify protein expression and localization in vivo and to continuously monitor disease progression and therapeutic effects has been promising[154,155]. We could show that FMT targeting mucosal myeloid-related protein-14 expression, allows specifically detection of intestinal inflammation at early stages of DSS-colitis[156]. Tomography with fluorescent light currently yields images with a resolution of 1-2 mm. Of particular importance for potential future use in human IBD is that FMT does not require radioactive labelling. Future studies are warranted to further elucidate the full diagnostic potential of FMT.

CONCLUSION

Ultrasound diagnostics in murine colitis requires a profound and analogous preparation of the examination to reliably assess classical sonographic parameters of inflammation of the colon wall. Notably, ultrasound provides high spatial resolution, allows serial real-time examination and with the use of molecularly targeted contrast agent, tissue- and disease-specific molecular imaging can be assessed.

White light colonoscopy is an indispensable tool for direct visualization of murine mucosa in colitis models. Sophisticated advancement such as CLE and FE allow further characterization of colorectal lesion on a molecular level, may improve the understanding of the pathogenesis of colorectal disease in general and intestinal carcinogenesis in particular and may facilitate early detection of colitis-associated cancer.

Finally, MRI using T2_W sequences or contrast enhancement has been successfully established for intestinal inflammation detection and monitoring of disease course and may additionally be used to evaluate intestinal fibrosis. Due to its low specificity, μCT alone remains insufficient for the diagnostic in experimental colitis; however, μCT combined with PET becomes a valuable tool for the assessment and monitoring of intestinal inflammation.