A guide to histomorphological evaluation of intestinal inflammation in mouse models

Abstract

Histomorphology remains a powerful routine evaluating intestinal inflammation in animal models. Emphasizing the focus of a given animal study, histopathology can overstate differences between established models. We aimed to systematize histopathological evaluation of intestinal inflammation in mouse models facilitating inter-study comparisons. Samples of all parts of the intestinal tract from well-established mouse models of intestinal inflammation were evaluated from hematoxylin/eosin-stained sections and specific observations confirmed by subsequent immunohistochemistry. Three main categories sufficiently reflected the severity of histopathology independent of the localization and the overall extent of an inflammation: (i) quality and dimension of inflammatory cell infiltrates, (ii) epithelial changes and (iii) overall mucosal architecture. Scoring schemata were defined along specified criteria for each of the three categories. The direction of the initial hit proved crucial for the comparability of histological changes. Chemical noxes, infection with intestinal parasites or other models where the barrier was disturbed from outside, the luminal side, showed high levels of similarity and distinct differences to changes in the intestinal balance resulting from inside events like altered cytokine responses or disruption of the immune cell homeostasis. With a high degree of generalisation and maximum scores from 4-8 suitable scoring schemata accounted specific histopathological hallmarks. Truly integrating demands and experiences of gastroenterologists, mouse researchers, microbiologists and pathologists we provide an easy-to-use guideline evaluating histomorphology in mouse models of intestinal inflammation. Standard criteria and definitions facilitate classification and rating of new relevant models, allow comparison in animal studies and transfer of functional findings to comparable histopathologies in human disease.

Introduction

During the past five decades animal models of intestinal inflammation substantially contributed to our understanding of the pathogenesis of human chronic intestinal inflammatory disorders. Even if none of the individual models represents all aspects and stages i.e. of ulcerative colitis mostly restricted to the colon mucosa or of Crohn’s disease affecting the whole gastrointestinal tract with transmural inflammation [1], they became indispensable to rationalise and evaluate therapeutic strategies for inflammatory bowel diseases (IBD) [2,3]. As for systematic attempts structuring the diversity, animal models are grouped by the mode of induction of the intestinal inflammation like administration of exogenous agents, genetic models and immunological strategies [4], by the mode of generating intestinal inflammation and components contributing to an IBD-like course [5,6], or by immunological mechanisms driving the inflammatory processes [3,7].

Histomorphology of the gastrointestinal tract reliably assesses the tissue architecture and the cellular composition even from small samples significantly supporting clinical diagnosis in human disease [8]. Experimental disease courses and therapeutic effects in animal models are also typically monitored by a multitude of methods at cell or molecular levels complemented by histopathology [3,7]. Acknowledging the compartmentalisation of immune responses underlines the importance of local events in the context of systemic diseases.

Despite a consent about major aspects to be evaluated that derive from experiences in human pathology, histomorphological scoring schemata used to assess the degrees of intestinal inflammation in established IBD models are as numerous as these models and as numerous as the research groups employing them. As for example of maximum scores, colitis induced by transfer of naϊve CD4⁺ T cells from wild-type mice into a syngeneic immunodeficient recipient rates 5 according to the group of Fiona Powrie [9] but 17 according to the group of Matthew B. Grisham [10]. These differences in numbers certainly do not reflect the level of sophistication of the analysis but rather the degree of comprehensible sub-categorization within the individual approaches. In a blinded fashion, we evaluated IBD models for many research groups dealing with T cell-mediated immune reactions in the gastrointestinal tract or with cellular approaches to suppress unwanted immune reactions and systematically adapted scoring systems. To this end, we defined categories and standard criteria to evaluate any established and new IBD model. Exemplified by standardized scoring schemata for colonic inflammation induced by erosive chemicals, by luminal antigens, by T-cell transfer or in mice with defined cytokine defects as well as for inflammation of the small intestine in genetic IBD models, upon infections with helminths or parasitic protozoans as widely used IBD models, we here provide an approved guideline facilitating decisions in the histomorphological evaluation of relevant mouse models of intestinal inflammation.

Materials and methods

Animals

C57BL/6 wild-type mice were obtained from the Bundesinstitut für Risikobewertung (Berlin, Germany). B6.129S7-Rag1^tm1Mom/J (Rag1-ko) mice were obtained from The Jackson Laboratory (Maine, USA). Mice deficient for interleukin-10 on the 129/SvJ background (IL-10ko) were generously provided by R. Balfour Sartor (University of North Carolina, Chapel Hill, USA). Mice deficient for IL-2 on the B6.129P2 background (IL-2ko) were generously provided by Ivan Horak (Leibniz Institute for Molecular Pharmacology, Berlin, Germany). C57BL/6 mice with a heterozygous mutation in the adenosine-uracil-rich element of the tumor necrosis factor gene (TNF^ΔARE/+) were generously provided by George Kollias (Biomedical Sciences Research Center Alexander Fleming, Vari, Greece) and conventionally raised at the animal facility of the Technical University of Munich-Weihenstephan. All other mice were bred, housed and treated under specific pathogenic free conditions at the Forschungseinrichtungen für Experimentelle Medizin of the Charité-Universitätsmedizin Berlin. Protocols for animal experiments were approved by the regional animal study committee of Berlin and Munich.

Colitis models

All colitis models were performed as previously described. For acute DSS-induced colitis, wild-type mice received 3% DSS (36,000-50,000 Da; ICN, Berlin, Germany) via the drinking water ad libitum [11,12]. In order to induce 2,4,6-trinitrobenzene sulfonic acid (TNBS) colitis, wild-type mice were sensitized with 2.5% TNBS (Sigma-Aldrich, Munich, Germany) applied to the shaved abdomen and one week later, mice received 2% TNBS intrarectally [13]. As for 4-ethoxymethylene-2-phenyl-2-oxazolin-5-one (oxazolone; Sigma-Aldrich) colitis, again wild-type mice were sensitized with 3% oxazolone and intrarectally challenged with 1% oxazolone after 5 days [14]. Intraperitoneal transfer of 4×10⁵ CD4⁺CD45RB^high T cells into syngeneic Rag1-ko mice induced intestinal inflammation within 4-8 weeks [15]. Colitis in IL-10ko was assessed starting at 4 months of age [16], in IL-2ko mice starting at 4 weeks of age [17] and in TNF^ΔARE/+ mice starting at 1-2 weeks of age [18,19]. Formalin-fixed colon tissue from a model of ovalbumin (OVA)-induced colitis 6 weeks after application of 40 mg OVA and simultaneous intravenous transfer of 6×10⁶ naïve OT-I T cells into wild-type mice were generously provided by Eckart Schott (Charité-Universitätsmedizin Berlin, Germany) [20].

Models of small intestinal inflammation

Wild-type mice were orally infected with 100 cysts of Toxoplasma gondii (T. gondii) strain ME49 as previously described [21] and ileum samples were obtained after 7 days. Wild-type mice were orally infected with Heligmosomoides polygyrus (H. polygyrus) and Nippostrongylus brasiliensis (N. brasiliensis), respectively. Formalin-fixed samples of the duodenum and jejunum were obtained 6 and 7 days after infection and generously provided by Susanne Hartmann (Freie Universität Berlin, Germany) [22].

(Immuno)Histochemistry

Intestines were thoroughly rinsed before representative samples were fixed in 10% neutral-buffered formalin (Sigma-Aldrich) for 12-24 hours and embedded in paraffin. Paraffin sections (1-2 μm) were cut and de-waxed prior to (immuno)histochemical staining. For scoring, sections were stained with hematoxylin/eosin (H&E; Merck, Darmstadt, Germany). For immunohistochemistry, sections were subjected to a heat-induced epitope retrieval step prior to incubation with polyclonal rabbit anti-mouse inducible nitric oxide synthase (iNOS) antibody (Abcam, Cambridge, UK) followed by detection employing the EnVision™ Detection System Peroxidase/DAB Rabbit (Dako, Glostrup, Denmark). For detection of endothelial cells, sections were incubated with rat anti-mouse CD31 primary antibody (clone SZ31; Dianova, Hamburg, Germany) followed by biotinylated rabbit anti-rat secondary antibody (Dianova) prior to detection using the LSAB Dako REAL^TM Detection System (Dako). Primary antibodies were omitted for negative controls and nuclei were counterstained with hematoxylin.

Histopathological evaluation

Tissue sections were evaluated and images taken by standard light microscopy using an AxioImager Z1 microscope (Carl Zeiss MicroImaging, Jena, Germany). Tissue samples were well-oriented with longitudinally cut crypts to precisely assess alterations in the overall intestinal tissue architecture. As for samples from the small intestine, at least 4-5 consecutive villi well-distended from base to tip were rated. If not stated otherwise, representative images within the figures show longitudinal sections.

Results

Categories and general criteria evaluating histopathology in animal models of intestinal inflammation

Based on established scoring systems for chemically induced [12,23] and transfer colitis [9,10], inflammatory cell infiltrates, epithelial changes and mucosal architecture were defined as the three main categories with respective subordinate specific criteria (Table 1).

Table 1.

Definition of categories and general criteria in histomorphological scores for intestinal inflammation in mouse models

Category	Criterion	Definition	Score Value1
I, Inflammatory cell infiltrate	Severity	Leukocyte density of lamina propria area infiltrated in evaluated hpf:
		Minimal: <10%	1
		Mild: 10-25%; scattered neutrophils	2
		Moderate: 26-50%	3
		Marked: >51%; dense infiltrate	4
	Extent	Expansion of leukocyte infiltration:
		Mucosal	1
		Mucosal and submucosal	2
		Mucosal, submucosal and transmural	3
II, Epithelial changes2	Hyperplasia	Increase in epithelial cell numbers in longitudinal crypts relative to baseline epithelial cell numbers per crypt; visible as crypt elongation:
		Minimal: <25%	1
		Mild: 25-35%	2 or 3
		Moderate: 36-50%; mitoses in middle/upper third of crypt epithelium, distant from crypt base	3 or 4
		Marked: >51%; mitoses in upper third of crypt epithelium, distant from crypt base	4 or 5
	Goblet cell loss	Reduction of goblet cell numbers relative to baseline goblet cell numbers per crypt:
		Minimal: <20%	1 or 2
		Mild: 21-35%	2 or 3
		Moderate: 36-50%	3 or 4
		Marked: >50%	4
	Cryptitis	Neutrophils between crypt epithelial cells	2 or 3
	Crypt abscesses	Neutrophils in crypt lumen	3 to 5
	Erosion	Loss of surface epithelium	1 to 4
III, Mucosal architecture2	Ulceration	Epithelial defect reaching beyond muscularis mucosae	3 to 5
	Granulation tissue	Connective tissue repair with new capillaries, surrounded by spindle-shaped fibroblasts, myofibroblasts, macrophages, neutrophils and mononuclear cells as well as cell debris; pseudopolyps: villiformous, hypertrophied areas projecting into the lumen	4 or 5
	Irregular crypts	Non-parallel crypts, variable crypt diameters, bifurcation and branched crypts	4 or 5
	Crypt loss	Mucosa devoid of crypts	4 or 5
	Villous blunting	Mild: villous-to-crypt-length ratio of 2:1 to 3:1	1 to 3
		Moderate: villous-to-crypt-length ratio of 1:1 to 2:1	2 to 4
		Villous atrophy	3 to 5

¹Values given for single criteria are not additive but represent the final score value that include all relevant criteria;

²From well-oriented samples/biopsies with longitudinally cut crypts only, at least 5 comparable crypts.

Leukocytes infiltrating the lamina propria sometimes extending to subjacent layers were the common feature in all models and in all parts of the inflamed intestine. Although infiltrates comprised neutrophils, eosinophils, monocytes, plasma cells and lymphocytes at differing ratios, evaluating the relative cellularity of mixed leukocytes within tissues per high-power field (hpf; 0.237 mm²), here summarized as inflammatory cells, sufficiently reflected inflammation-related alterations (Table 1, Figure 1A). Distinguishing from the healthy lamina propria containing only few immune cells, four grades estimated percentages of leucocytes per representative hpf. Scattered neutrophils were a typical feature in mild infiltrates (Figure 1B). Besides numbering inflammatory cells, the extent of the infiltrate completed this category. Starting from mainly focal mucosal localization of inflammatory cells (Figure 1C), increased severity first included the submucosa (Figure 1D), subsequently extended to the muscularis propria and eventually led to transmural infiltrates (Figure 1E). Major changes that originally apply to epithelial cell layers included crypt epithelial cell hyperplasia, the loss of goblet cells, cryptitis and crypt abscesses and erosions (Table 1). Proliferation of epithelial cells as a sign of increased regeneration was a general feature of intestinal inflammation in the small intestine as well as in the colon. Four stages of increase in crypt length in relation to normal intestinal tissue appearance adequately graded the severity of intestinal inflammation (Figure 2). Minimal and mild hyperplasia essentially presented itself as increased cell numbers in the epithelial lining. Moderate and marked hyperplasia also revealed mitoses in regions distant from the crypt base that were easily acknowledged as mitotic nuclei from H&E staining. Goblet cell loss from the epithelial cell layers of the small and large intestine clearly marked the severity of intestinal inflammation and was rated in relation to baseline goblet cell numbers in the given region of the intestine (Figure 2).

Figure 1.

Representative images of H&E-stained colon sections illustrate specifics of inflammatory cell infiltrates. (A) Infiltrating mixed leukocytes in a cross section (×400, scale bar 20 μm). (B) Scattered neutrophils (arrows; ×400, scale bar 20 μm). (C) Multifocal mucosal infiltration of mixed inflammatory cells (arrows; ×100, scale bar 100 μm). (D) Mucosal and submucosal (bracket) infiltrate of inflammatory cells (×100, scale bar 100 μm) and (E) transmural inflammatory cells (×200; scale bar 50 μm).

Figure 2.

Representative images of H&E-stained colon sections illustrate the grades of epithelial hyperplasia and of goblet cell loss. Hyperplasia appears as elongated crypts due to increase in epithelial cells (upper row; ×100) and loss of goblet cells that display clear mucus droplets (lower row; ×100).

Neutrophils interspersing in the epithelial cell layer of the crypt in terms of a cryptitis (Figure 3A) intensified to crypt abscesses where neutrophils accumulate in the crypt lumen eventually disrupting the crypt epithelium (Figure 3B). Erosions, i.e. the loss of surface epithelium with underlying inflammation where the epithelial defect reached the basement membrane presented itself mostly focal (Figure 3C). Scoring this feature especially depended upon the model used as indicated by the wide range of final score values (Table 1). With a “first hit” from the outside erosion was even abundant at a value of 1 (Table 2); erosion resulting from severe mucosal inflammation due to an inside event accompanied values of 4 (Table 4).

Figure 3.

Representative H&E- and immunohistochemically stained colon sections illustrate epithelial changes and altered mucosa architecture. (A) Cryptitis appears as neutrophils (arrows) between crypt epithelial cells in a cross section (×400, scale bar 20 μm). (B) Crypt abscesses with neutrophils in the lumen and nearly intact epithelium (white arrowhead) or damaged epithelium (black arrowhead) and complete crypt loss (arrow) in cross section (×400, scale bar 20 μm). (C) Loss of surface epithelium marking erosion (arrowheads; ×100, insert ×400). (D) Ulceration (×100, scale bar 100 μm). (E) Ulceration covered by an exudate rich in iNOS⁺ neutrophils (brown; ×100, scale bar 100 μm) and (F) granulocyte-fibrin eschar (×100, insert ×400, scale bar 20 μm). Consecutive sections of (G) granulation tissue (×100, insert ×400, scale bar 20 μm) comprising (H) CD31⁺ endothelial cells (red; ×100, insert ×400, scale bar 20 μm) and (I) apical iNOS⁺ cells (brown; ×400, scale bar 20 μm).

Table 2.

Classification of IBD mouse models evaluated within this study

Organ	“First hit”	Induction	Example	Sampling1	Scheme
Colon	Outside	Induced by erosive chemicals	Oral DSS; single cycle	Distal colon, day 2-42	1 (Table 3, Figure 6)
			Oral DSS; repeated cycles	Distal colon, day 142
			Rectal TNBS; single dose	Distal colon, day 2-4
			Rectal oxazolone; single dose	Distal colon, day 2-4
		Mediated by luminal antigen	Rectal ovalbumin, transfer of OVA-specific CD8+ T cells	Colon3, week 6-8	2 (Table 4, Figure 7)
	Inside	Mediated by disturbed immune cell homeostasis	Transfer of naïve CD4+ T cells into syngeneic immunodeficient mice	Colon3, week 4-8	3 (Table 5, Figure 8)
		Mediated by genetically altered cytokine balance	Homozygous IL-10ko	Ascending and transverse colon, week 34	4 (Table 6, Figure 9)
			Homozygous IL-2ko	Colon3, week 6-15
Small intestine	Inside	Mediated by genetically altered cytokine balance	TNFΔARE/+	Terminal ileum, week 4-8	5 (Table 7, Figure 10)
	Outside	Mediated by luminal antigen	Oral infection with H. polygyrus	Duodenum, day 6	6 (Table 8, Figure 11)
			Oral infection with N. brasiliensis	Jejunum, day 7
			Oral infection with T. gondii	Ileum, day 7-8	(Table 9, Figure 12)

¹Most affected organ and time point of most severe inflammation; can vary according to the experimental setup;

²After DSS removal;

³Severity increases from proximal to distal colon;

⁴Development of chronic colitis and adenocarcinoma ≥3 months of age.

Table 4.

Scoring scheme 2 for colonic inflammation mediated by luminal antigens

Inflammatory cell infiltrate		Epithelial changes	Mucosal architecture	Score
Severity	Extent
Minimal	Mucosa	Minimal hyperplasia		1
Mild	Mucosa and submucosa	Mild hyperplasia, minimal goblet cell loss		2
Moderate	Mucosa and submucosa, sometimes transmural	Moderate hyperplasia, mild goblet cell loss		3
Marked	Mucosa and submucosa, often transmural	Marked hyperplasia with moderate to marked goblet cell loss	Ulcerations, crypt loss	4

All changes reaching deeper than the surface epithelial cell layer were considered alterations in the overall mucosal architecture and graded from the presence of ulcerations, granulation tissue, irregular crypts or of crypt loss. Specifics like the ratio of villous length to crypt depth accounted for general anatomical differences along the whole small intestine (Table 1). Ulcerations i.e. erosions exceeding to the submucosa (Figure 3D) were occasionally covered by an exudate rich in iNOS⁺ neutrophils (Figure 3E) or by granulocyte-fibrin eschar (Figure 3F). Connective tissue repair was the hallmark of granulation tissue underlying ulcerations (Figure 3G). It comprised new capillaries lined by CD31⁺ endothelial cells (Figure 3H), myofibroblasts, iNOS⁺ macrophages and neutrophils (Figure 3I) as well as lymphocytes and cell debris. Generally, crypt appearance throughout the intestine and the villous morphology within the small intestine defined a major level of mucosal architecture. Irregular crypts with variable diameters along the depth of single crypts or dilated crypts (Figure 4A), crypts that did not run parallel (Figure 4B), bifurcation at the base of the crypt (Figure 4C, 4D) that might extend to branched crypts (Figure 4E), to a mucosa completely devoid of crypts (Figure 4F, 4G) or crypts herniated to the submucosa (Figure 4H) marked mucosal histopathology. In the duodenum and jejunum, the villi are long with a villous-to-crypt length ratio on the order of 3:1 to 5:1, while in the ileum the villi are typically shorter (Figure 5). Thus, progressive broadening and blunting of the villi up to complete loss of villous structures clearly indicated pathological changes. We found two degrees of villous blunting (Figure 5) and villous atrophy sufficient to grade the severity.

Figure 4.

Representative H&E-stained colon sections illustrate altered crypts. (A) Variable diameters in adjacent crypts in a cross section with dilated crypts containing mucus (arrows; ×100, scale bar 100 μm). (B) Slight crypt distortion with crypt abscess (arrow; ×100, scale bar 100 μm). Bifurcated crypts (arrow) in (C) longitudinal section (×100, scale bar 100 μm) and (D) cross section (×400, scale bar 20 μm). (E) Branched crypts (×100, scale bar 100 μm). Mucosa containing only remains of crypts (arrowheads) in (F) longitudinal section (×100, scale bar 100 μm) and (G) cross section (×400, scale bar 20 μm). (H) Herniated crypts (×100, scale bar 100 μm).

Figure 5.

Representative H&E-stained sections of the small intestine illustrate villous blunting with normal and altered villi from the duodenum and ileum (×100, scale bar 100 μm).

As for epithelial changes and the mucosal architecture, each model system had different key aspects regarding degree, quality, extent and depth that made overall generalization in terms of a single generalized scoring system using all criteria for each model not feasible. Our schemata propose maximum severity rated 4-5 from combined changes in all three categories. Characteristic discontinuities in inflammatory cell infiltrates and intestinal architecture were best reflected by additive scores with a maximum of 6 or 8, relating to the well-established evaluation of DSS-induced colitis [11].

We methodically inspected widely used relevant murine IBD models according to the defined categories and criteria and suggested scoring schemata. Evaluating histomorphology from H&E stained sections grouped our models according to the initial step weakening the barrier rather than the exact nature of immune interactions in the whole processes of acute or chronic intestinal inflammation. “First hits” from the outside, the luminal side, induced histopathological specifics that were comparable among each other but distinguishable from those models resulting from inside events, i.e. from imbalanced immune dispositions (Table 2, Supplementary Figure 1).

Additionally, in line with representative tissue samples excised from the most affected organ, the most affected section should be scored. This goes especially for discontinuous inflammations. If the extent of the inflammation shall be taken into account, a score index can be calculated as the product of the score and the relative tissue area affected.

Scheme 1: chemically induced colonic inflammation

All models of chemically induced colonic inflammation start from a disrupted epithelial barrier enabling an immunoreactive agent to interact with immune cells within the lamina propria [24]. Acute colitis was induced by rectal application of TNBS or oxazolone to sensitized mice and oral administration of DSS, respectively. Repeated cycles of low-dose DSS gave rise to chronic intestinal inflammation. The main finding in all models was the discontinuity of mucosal changes as extensive infiltrates were observed without epithelial defects and vice versa that was reflected in an additive score separately evaluating inflammatory cell infiltrates and intestinal architecture with maximum values of 3 each (Table 3). Severe colitis presented barely preserved mucosal architecture showing extended ulcerations, marked, often transmural inflammatory cell infiltrates and destroyed crypts. Colon sections of DSS-induced colitis illustrate this scoring system in Figure 6.

Table 3.

Scoring scheme 1 for chemically-induced colonic inflammation

Inflammatory cell infiltrate:		Score 1	Intestinal architecture:		Score 2
Severity	Extent		Epithelial changes	Mucosal architecture
Mild	Mucosa	1	Focal erosions		1
Moderate	Mucosa and submucosa	2	Erosions	± Focal ulcerations	2
Marked	Transmural	3		Extended ulcerations ± granulation tissue ± pseudopolyps	3
			Sum of scores 1 and 2:		0-6

Figure 6.

Histomorphology of colon tissue in DSS-induced colitis in C57BL/6 wild-type mice at day 2 after DSS removal representing scores according to scheme 1 referring to Table 3. A. Sum score 1: mild mucosal inflammatory cell infiltrates (score 1: 1) with intact epithelium (score 2: 0); B. Sum score 2: inflammatory cell infiltrates into mucosa and submucosa (score 1: 2) with undamaged epithelium (score 2: 0); C. Sum score 3: mucosal infiltrates (score 1: 1) with focal ulceration (score 2: 2); D. Sum score 4: inflammatory cell infiltrates in mucosa and submucosa (score 1: 2) and focal ulceration (score 2: 2); E. Sum score 5: moderate inflammatory cell infiltration into mucosa and submucosa (score 1: 2) with extensive ulcerations (score 2: 3); F. Sum score 6: transmural inflammation (score 1: 3) and extensive ulceration (score 2: 3). Original magnification ×100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid) and submucosa (dotted); white arrowhead-ulceration; bracket–transmural inflammation.

Scheme 2: colonic inflammation mediated by luminal antigens

Luminal antigens include antigens from food components or from the intestinal microbiota. The scheme for antigen-specific models acknowledged the severity of intestinal inflammation with a maximum rated 4 (Table 4). While early stages were mainly characterised by gradually increasing inflammatory cell infiltrates, late stages typically showed ulcerations. This group is exemplified by a model where rectally applied OVA induced colitis in fully immune-competent wild-type mice that simultaneously received OVA-specific CD8⁺ T cells for readout issues (Figure 7).

Figure 7.

Histomorphology of colon tissue 6 weeks after intrarectal application of ovalbumin and simultaneous intravenous transfer of ovalbumin-specific T cells into wild-type mice representing scores according to scheme 2 referring to Table 4. A. Score 1: intact epithelium with minimal focal inflammatory cell infiltrates in the mucosa; B. Score 2: scattered inflammatory cell infiltrates in mucosa and submucosa; C. Score 3: diffuse mucosal and submucosal inflammatory cell infiltrates; D. Score 4: moderate inflammatory cell infiltrates in the submucosa which is edematous, focal ulceration. Original magnification ×100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid) and submucosa (dotted); white arrowhead-ulceration.

Scheme 3: colonic inflammation mediated by a disturbed immune cell homeostasis

Adoptive transfer of naïve CD4⁺ T cells into syngeneic immunodeficient recipients results in continuous inflammation of the colon and the small intestine as well as in associated hepatitis [25] and pancreatitis (data not shown). It did not matter whether transferred T-cell populations were defined as CD4⁺CD45RB^hi, CD4⁺CD62L⁺ or CD4⁺CD25^- T cells, given that increasing severity of colonic inflammation was found as concurrent increase in inflammatory cell infiltrates and epithelial changes with striking hyperplasia (see also Figure 2). In late stages, these were accompanied by ulcerations that were not as severe as in the model of antigen-mediated colitis above and by altered or lost crypts. A non-additive scoring system with a maximum of 5 reproducibly reflected all models of transfer colitis (Table 5) and is exemplified by transfer of CD4⁺CD45RB^hi T cells into Rag1-ko mice (Figure 8). For the evaluation of the accompanying small bowel inflammation in the transfer colitis model, we used the scheme 6 (Table 8).

Table 5.

Scoring scheme 3 for colonic inflammation mediated by disturbed immune cell homeostasis

Inflammatory cell infiltrate		Epithelial changes	Mucosal architecture	Score
Severity	Extent
Minimal	Mucosa	Minimal hyperplasia		1
Mild	Mucosa, sometimes extending into submucosa	Mild hyperplasia, minimal goblet cell loss ± erosions		2
Moderate	Mucosa and submucosa	Moderate hyperplasia ± few crypt abscesses, moderate goblet cell loss ± erosions		3
Marked	Mucosa and submucosa	Marked hyperplasia ± several crypt abscesses and/or erosions	± Irregular crypts or crypt loss ± ulcerations	4
Marked	Transmural	Marked hyperplasia ± multiple crypt abscesses	± Irregular crypts or crypt loss ± ulcerations	5

Figure 8.

Histomorphology of colon tissue after transfer of CD4⁺CD45RB^hi T cells into Rag1-ko mice representing scores according to scheme 3 referring to Table 5. A. Score 0: normal colon mucosa with intact epithelium; B. Score 1: scattered inflammatory cell infiltrates in the mucosa; C. Score 2: diffuse mucosal infiltrates without submucosal spreading and intact epithelial layer; D. Score 3: moderate infiltration of inflammatory cells into mucosa and submucosa with epithelial hyperplasia and goblet cell loss; E. Score 4: marked inflammatory cell infiltrates in mucosa and submucosa accompanied by crypt abscesses and loss of goblet cells and crypts; F. Score 5: marked inflammatory cell infiltrates within the mucosa spreading to the submucosa going along with crypt loss and hemorrhage. Original magnification ×100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid) and submucosa (dotted); black arrowhead-crypt abscess, blue arrowhead-goblet cell loss; yellow arrowhead-crypt loss; red arrowhead-hemorrhage.

Table 8.

Scoring scheme 6 for inflammation of the small intestine mediated by luminal antigens

Inflammatory cell infiltrate:		Score 1	Intestinal architecture:		Score 2
Severity	Extent		Epithelial changes	Mucosal architecture
Mild	Mucosa and sometimes submucosal	1		Mild villous blunting	1
Moderate	Mucosa and sometimes submucosal	2	Mild hyperplasia	Moderate villous blunting	2
Marked	Mucosa and submucosa, sometimes transmural	3	Moderate hyperplasia and goblet cell loss	Moderate villous blunting and broadening, sometimes villous atrophy	3
Marked	Transmural	4	Marked hyperplasia and goblet cell loss	Villous atrophy, ulcerations	4
			Sum of scores 1 and 2:		0-8

Scheme 4: colonic inflammation in genetically altered mice with defined cytokine defects

Models with recombinant transgenic expression or the deficiency to produce a defined cytokine are often referred to as genetic IBD models [1,6,7]. Due to the systemic cause, intestinal inflammation often affects the colon as well as the small intestine [3]. Colonic inflammation was very similar to transfer colitis. Hallmarks were massive leukocyte infiltration of the lamina propria, hyperplasia of the crypt epithelium and crypt abscesses. These findings were integrated to a non-additive scoring scheme that due to the overall lower severity compared to the transfer colitis rated a maximum of 4 (Table 6). It is illustrated by IL-10ko mice that developed a chronic enterocolitis at the age of 3 weeks (Figure 9) and was also applicable for the colitis within homozygous IL 2ko mice (not shown).

Table 6.

Scoring scheme 4 for colonic inflammation in consequence of cytokine imbalance

Inflammatory cell infiltrate		Epithelial changes	Mucosal architecture	Score
Severity	Extent
Minimal	Mucosa	Minimal hyperplasia		1
Mild	Mucosa, sometimes submucosa	Mild hyperplasia, mild goblet cell loss ± cryptitis ± erosions		2
Moderate	Mucosa and submucosa	Moderate hyperplasia, moderate goblet cell loss, cryptitis ± crypt abscesses	Ulcerations	3
Marked	Mucosa and submucosa, often transmural	Marked hyperplasia, marked goblet cell loss, multiple crypt abscesses	Extended ulcerations ± pseudopolyps	4

Figure 9.

Histomorphology of colon tissue in IL-10 ko mice at 4 months of age representing scores according to scheme 4 referring to Table 6. A. Score 1: minimal inflammatory cell infiltrates in the mucosa with intact epithelium; B. Score 2: mild inflammatory cell infiltrates in the mucosa with mild hyperplasia and mild goblet cell loss; C. Score 3: moderate inflammatory cell infiltrates in mucosa and submucosa with moderate goblet cell loss; D. Score 4: marked inflammatory cell infiltration into mucosa and submucosa with marked hyperplasia and marked goblet cell loss, multiple crypt abscesses and crypt loss. Original magnification ×100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid) and submucosa (dotted); black arrowhead-crypt abscess; yellow arrowhead-crypt loss; blue arrowhead-goblet cell loss.

Scheme 5: inflammation of the small intestine in genetically altered mice with defined cytokine defects

IBD models with involvement of the small bowel are rare and characteristics of small intestinal inflammation did not differ significantly from inflammation of the colon. Our scoring system considered specifics in architecture and also graded the alterations of the villous structure caused by (sub)mucosal inflammatory cell infiltrates and epithelial hyperplasia (Table 7). Despite the different locus and main features, the late inflammation was as severe as seen in transfer colitis and we also rated the maximum as 5. Tissues from the terminal ileum of TNF^ΔARE/+ mice displayed typical features of small intestinal inflammation (Figure 10).

Table 7.

Scoring scheme 5 for small intestinal inflammation in consequence of cytokine imbalance

Inflammatory cell infiltrate		Epithelial changes	Mucosal architecture	Score
Severity	Extent
MiId	Mucosa			1
Mild	Mucosa and submucosa			2
Moderate	Mucosa and submucosa	Mild hyperplasia	Mild villous blunting	3
Marked with submucosal lymphoid aggregates and granuloma	Mucosa and submucosa, sometimes transmural	Moderate hyperplasia	Moderate villous blunting	4
Marked with submucosal granuloma	Mucosa, submucosa and transmural	Marked hyperplasia	Villous atrophy, branched crypts	5

Figure 10.

Histomorphology of tissue from the terminal ileum in TNF^ΔARE mice starting at 2 weeks of age representing scores according to scheme 5 referring to Table 7. A. Score 0: normal ileum with intact epithelium and short, finger-like villi; B. Score 1: mild mucosal inflammatory cell infiltrate; C. Score 2: mild diffuse inflammatory cell infiltrate in mucosa and submucosa; D. Score 3: moderate inflammatory cell infiltrates in mucosa and submucosa with villous blunting; E. Score 4: marked mucosal, submucosal and transmural inflammatory cell infiltration with lymphoid aggregates predominantely in the submucosa accompanied by villous broadening; F. Score 5: marked transmural inflammatory cell infiltration and villous atrophy (insert: submucosal granuloma; ×400). Original magnification ×100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid), submucosa (dotted), and transmural (white); black arrowhead-crypt abscess; circle-lymphoid aggregates.

Scheme 6: inflammation of the small intestine induced by luminal antigens

Luminal endoparasites like helminths directly trigger a host immune response by attachment, migration and feeding activities or indirectly from opportunistic infections [26]. The complex sequence of events finally encompassed a weakened intestinal epithelial barrier. Like for chemically-induced colitis, we found an additive scoring system suitable that independently described the level of leukocyte infiltration and the intestinal architecture and acknowledged the severity with maximum combined rates of 8 (Table 8). The use of this scheme is exemplified by duodenal tissue from the model of oral infection with Heligmosomoides polygyrus (Figure 11) but it also worked for the jejunum upon infection with Nippostrongylus brasiliensis (not shown). The obligate intracellular parasite Toxoplasma gondii causes an acute small bowel inflammation due to T helper (Th)-1 cell-mediated mechanisms [21]. Rapid development of an intestinal pathology with very early severe mucosal necroses and tissue destruction necessitated a separate scoring system that we adapted to our general outline with a maximum of 4 (Table 9, Figure 12).

Figure 11.

Histomorphology of duodenal tissue in mice infected with Heligmosomoides polygyrus representing scores according to scheme 6 referring to Table 8. A. Sum score 0: normal duodenum; B. Sum score 2: mild mucosal and submucosal inflammation (score 1: 1) with normal crypts and villi (score 2: 0); C. Sum score 4: moderate infiltration of mucosa (score 1: 2) with mild villous blunting and mild hyperplasia (score 2: 2); D. Sum score 5: transmural inflammation (score 1: 3) with moderate villous blunting and mild hyperplasia (score 2: 2); E. Sum score 6: mucosal inflammatory cell infiltration (score 1: 2) with moderate villous blunting and marked hyperplasia (score 2: 4); F. Sum score 7: marked transmural inflammation (score 1: 4) with moderate blunting and moderate hyperplasia (score 1: 3). Original magnification ×100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid), submucosa (dotted), and transmural (white).

Table 9.

Score for small intestinal inflammation upon infection with Toxoplasma gondii

Inflammatory cell infiltrate		Epithelial changes	Mucosal architecture	Score
Severity	Extent
Minimal, focal	Mucosa	Intact epithelium		1
Mild, focal to diffuse	Mucosa and submucosa	Sometimes associated with erosions		2
Moderate, diffuse	Mucosa and submucosa	Erosions and/or ulcerations; cryptitis and/or crypt abscesses	Distortion of villous structure	3
Severe, diffuse	Mucosa and submucosa, sometimes transmural	Ulcerations associated with necroses and fibroses	Distortion of villous structure	4

Figure 12.

Histomorphology of ileal tissue in mice infected with Toxoplasma gondii at day 7 representing scores referring to Table 9. A. Score 1: minimal and focal inflammatory cell infiltrates in the mucosa with an intact epithelial layer; B. Score 2: mild, diffuse inflammatory cell infiltrates in mucosa and submucosa with flattened epithelium and villous blunting; C. Score 3: moderate, diffuse inflammatory cell infiltrates in mucosa and submucosa with erosions and distorted villlous structure; D. Score 4: necrotic and fibrotic mucosa with distorted villi. Original magnification x100; scale bars 100 μm; arrows-inflammatory cell infiltrates within mucosa (solid) and submucosa (dotted).

Discussion

Comparing intestinal histomorphology as assessed from H&E stained tissue sections from different mouse models or from the same model in the hands of different researchers requires consistent criteria. Evaluating inflammatory cell infiltrates, epithelial changes and the mucosal architecture as main categories along our definitions grouped the animal models according to the initial step weakening the epithelial barrier (‘first hit’). “First hits” from the outside, the luminal side, induced histopathological specifics that were comparable among each other but distinguishable from those models resulting from imbalanced immune dispositions (inside events).

Infiltrating leukocytes i.e. granulocytes, monocytes and lymphocytes were common in all models and parts of the inflamed intestine. So all models fitted the definitions evaluating inflammatory cell infiltrates in category I.

Epithelial cell proliferation was also a general feature of intestinal inflammation in all models and is a criterion in category II comprising epithelial changes. Except for the crypt base dominated by intestinal stem cells and in the small intestine by Paneth cells, crypts are lined by a single layer of epithelial cells with interspersed goblet cells and occasional enteroendocrine cells [27]. Overall increase of epithelial cell numbers visible as crypt elongation in longitudinal crypts is also referred to as crypt hyperplasia [28], mucosal hyperplasia [29], thickening of colon wall [30], crypt elongation [31,32], or epithelial hyperplasia [9,33]. Epithelial hyperplasia is typical for transfer colitis, the best characterized model of colitis induced by disruption of the T-cell homeostasis [34]. A scoring system was initially developed for Rag2-ko mice or C.B-17 SCID mice that were injected intraperitoneally with CD4⁺CD45RB^high T cells from syngeneic wild-type mice [9]. In the meantime the definition for naϊve CD4⁺ T cells to be transferred and the recipient host were broadened [35]. Transfer colitis in SCID mice is usually more severe than in Rag1-ko mice and the presence of B cells normally attenuates the disease [10,36]. Four stages of epithelial hyperplasia adequately graded the severity of intestinal inflammation. Additionally, detection of proliferating Ki67⁺ epithelial cells by immunohistochemistry may be used for confirmation of hyperproliferation giving rise to crypt hyperplasia [37]. Hyperplasia should be assessed by quantifying the epithelial cells lining the crypts. A minimum of 10 crypts should be evaluated. Technically challenging, the crypt length could be given in micrometers from exactly oriented longitudinal sections as well as the percentage of Ki67⁺ proliferating epithelial cells could be determined. Another criterion in category II is the loss of goblet cells. Especially in UC, goblet cells are reduced in number and size and it is still unknown whether the alterations in mucin production contribute to the initiation or result from the inflammation [38]. Referring to the product secreted by the goblet cells this is also described as mucin depletion [22] or mucin depletion from goblet cells [9]. Normally, the proportion of goblet cells among epithelial cells increases along the parts of the intestine from 4% within the duodenum to 16% within the distal colon [39]. Histochemical staining of proteoglycans with Alcian blue and/or periodic acid-Schiff stain underlines the presence mucins [40]. Here again, goblet cell loss should be assessed by counting of goblet cells per crypt or per villus. Here again, a minimum of 10 crypts and villi, respectively, should be analysed. Further criteria for epithelial changes are neutrophils interspersing among epithelial cells (cryptitis) aggravating to crypt abscesses [41] and erosions. We define erosions as the loss of surface epithelium with underlying inflammation where the epithelial defect crossed the basement membrane and presented itself mostly focal [42].

Although the maximum contributions of changes in the epithelium and the mucosal architecture for the intestinal inflammation were included in our definitions, each model had slightly different key aspects with respect to these categories. All changes reaching deeper than the epithelial cell layer were considered alterations in the overall mucosal architecture and summarized in category III. Ulcerations are erosions exceeding to the submucosa with underlying regenerating tissue [41]. Extended inflammatory processes, especially in acute models comprised large areas of so-called granulation tissue. The hallmark was connective tissue repair comprising new capillaries surrounded by (myo) fibroblasts, macrophages, neutrophils, lymphocytes and cell debris [43,44]. Hypertrophic granulation tissue resulted in pseudopolyps in the colon i.e. villiform structures projecting into the lumen [45] and are reported among others in DSS-induced colitis [46]. Generally, crypt appearance throughout the intestine and villous morphology within the small intestine defined the main level of mucosal architecture. In severe or long-standing inflammation, crypt architecture becomes distorted with irregularly shaped crypts and even crypt loss. In the small intestine, the mucosal architecture is also defined by the appearance of the villi. Hence, the villous architecture was taken into account in the category III. Although the villous structure changes along the small intestine, two degrees of villous blunting and villous atrophy sufficiently graded the severity of histomorphological changes. Widely used mouse models for intestinal helminth infection are Heligmosomoides polygyrus and Nippostrongylus brasiliensis [47,48]. Despite strong Th-2 cell responses in various tissues, the pathology is most prominent in the small intestine [49]. Other IBD models involving the small intestine are rare like SAMP/Yit, TNF^ΔARE/+ and Roquinko mice [19,50,51]. Histopathology of small intestinal inflammation as described by the categories I and II appeared similar to colonic inflammation within the respective models. While changes in the mucosal architecture occurred in moderate to marked inflammation of the colon, villous architecture already changed in mild inflammation. One exception in our schemes is the acute small bowel inflammation after infection with Toxoplasma gondii. The rapid development of pathology with very early severe mucosal necroses and tissue destruction required a specific score that is not applicable to other models of small intestinal inflammation.

Integrating the special experiences of surgical pathologists, mouse researchers, microbiologists and gastroenterologists, we here provide a guideline for scoring of histomorphological changes in mouse models of intestinal inflammation. Our schemata are easy and fast to apply as we focussed on main categories. If need be, the scores can be broadened by determining additional parameters employing (immuno)histochemistry e.g. composition of the infiltrate, degree of vascularisation, apoptosis rate. Employing the three categories according to the “first hit” might help all researchers working with mouse models of intestinal inflammation to classify their model into our schemata and finally compare their experimental findings unbiased.

Disclosure of conflict of interest

None.