Figure 1.
Schematic of the scale lengths and analyses used for analyzing crypt production in the colon.
The mouse colon is considered in 3 regions between the caecum and the anus: Proximal, middle and distal. The epithelial surface is separated from the submucosa to produce isolated crypts. The morphology and cellular distribution along the crypt lumen axis are analyzed and the nuclear, cytoplasmic and membrane distribution of E-cadherin and β-catenin were determined. An understanding of the differences between regions of the colon, the morphology of the isolated crypts, the cell packing of the crypts and the subcellular compartments enables the structure and functional relationships of the underlying crypt formation and cell production to be determined.
Figure 2.
Confocal images of isolated proximal, middle and distal crypts from mouse colons of different ages.
(A) Selected 2D confocal images of isolated colonic crypts from the different regions of the colon (proximal, middle and distal) obtained from mice of varying ages. The size and shape of crypts differ with the age of the mice and regions of the colon. Crypts isolated from very young mice (< 2 weeks) are fragile and easily broken. Images are of the same magnification with scale bar of 50µm. (B) Screen capture of a Fiji/ImageJ window with a typical image stack showing how the crypts are labeled, counted and measured. In this instance, a zoomed-in image of the crypt being measured is shown (middle panel) with a green segmented line marking the length of the crypt and a black segmented line marking the basal width of the crypt (drawn at about 20% of the crypt length from the crypt base). (C) Length measurements are made (in ImageJ as described in B) and tabulated for crypts isolated from mice aged 2.3, 4, 7 and 40 weeks. The proximal crypts are significantly shorter than the crypts from the middle or distal regions of the colon (p<0.01). Note: Images are 2D sections of image stacks consisting of a phase contrast image and DAPI staining in blue.
Figure 3.
Stages of colon crypt development and budding rates for mice of different ages.
(A) 3D confocal image of colonic crypts at the four stages of crypt development (defined as stages 0 to 3) as proposed in this study. An asymmetrical crypt production concept is proposed (budding), contrary to the symmetrical concept (fission). Schematic representations of the two concepts: red representing the development of a new bud; green sections depict the continuation of the original crypt; black dotted line represents the developing front of the crypt(s). Stages are defined based on the position of the new crypt base and the point of division of the crypts as described below the representative confocal image of each stage. Stage 0 is the only non-budding phase; Stage 1 represents the onset of crypt bud formation; Stage 2, in the midst of budding and Stage 3, late completion stage of budding. Stage 3 crypts transits into independent stage 0 crypts upon detachment from each other. (B) 2D images from image stacks showing the advantage provided by confocal 3D imaging of isolated crypts in capturing previously undetected buds of colonic crypts that are in stage 1 (white arrows). Onset of budding is difficult to detect using micro-dissection [17,32,33] or serial tissue sectioning [34]. Asymmetrical budding is also shown to occur in stage 1. P: Parental Crypt, D1: Daughter crypt 1, D2: Daughter crypt 2.
Figure 4.
Measurements and analysis of colon crypt production.
(A) Mature crypts (Stage 0 + 3) are similar in length while the daughter crypt (Stage 3) is significantly shorter than the mature crypt in different colon regions and animal age (* p<0.05, **p<0.01, #daughter crypt count < 10, S0-Stage 0, S3-Stage 3, M- mature crypts, D- daughter crypts). (B) The average basal width of isolated crypts (Stage 0+3) for the different regions of the colon showing crypts from the middle regions are wider than the other regions. (C) Using an estimated rectangular area as a measure of crypts size, the average crypt widths and lengths (stage 0+3) from the proximal, middle and distal regions for mice aged 2.3, 4, 7 and 40 weeks are multiplied. The crypts from the middle regions are the largest with the proximal crypts the smallest for each age group. (D) Derivations for asymmetry index (mean % deviation, AI) describing the degree of crypts budding asymmetrically is shown. 2D DAPI-labeled images of crypts at different level of asymmetry shown with corresponding AI labeled above the image. Using 1% AI as the cut-off to approximate symmetry during budding, a histogram of the AI frequency of crypts (4 and 7 weeks of age) indicates ~ 90% (89.9% and 90.4%) of the crypts has AI greater than 1%. (E) “Stage 2” and “Stage 1+2” crypt budding rates, showing a significant decrease with mice age. The “Stage 2” analysis was used as a comparison with previously reported numbers of crypt budding using micro-dissection [17,32,33] or serial sectioning [34] (where Stage 1 buds are difficult to detect). Crypt budding rates for the proximal, middle and distal regions of the colon are compared to that reported by Bjerknes et al [35]. (F) Crypts scored in Stages 0, 1 and 3 of the study are tabulated.
Figure 5.
β-catenin and E-cadherin expression in small intestine and colonic crypts.
Immunofluorescent 2D images extracted from 3D image stacks of one small intestinal crypt (A) and two isolated colonic crypts (B-C) with the DAPI, β-catenin, E-cadherin, overlayed RGB images and 3D view shown (see supporting information for movies and data sets). (A) Validation of the immuno-staining was conducted by applying the same isolation and analysis protocol on small intestinal crypts with panel A showing the corresponding 2D images of DAPI, β-catenin, E-cadherin and overlaid RGB images from one 3D image stack of a small intestinal crypt. The high levels of β-catenin in the nucleus of a Paneth cell (white arrow) are typical of the small intestinal crypts [31]. (B) For 60% of the colon crypts (heterogeneous population, HE crypts) scored, β-catenin expression is mainly present at the membrane along the whole length of the crypt. E-cadherin exhibit: “bands” or “rings” of E-cadherin low expressing cells (white arrows) at the lower half of the crypt leading to distinct clusters of high E-cadherin expressing cells at the bottom surrounded by the low expression clusters (yellow arrow). Zoomed images of β-catenin and E-cadherin at the crypt bottom; β-catenin localizes to the lateral (red arrows) membrane with some apical or basal (green arrows) localization observed. E-cadherin localizes to lateral-apical (red arrows) and basal membrane (blue arrows). (C) For the remaining 40% of the colon crypts (homogenous population, HO crypts) scored, β-catenin and E-cadherin expression appears to be present homogenously at the membrane along the whole length of the crypt. Note: Crypt compartments shown in C, RGB panels (5 evenly divided regions of interest); Inverted black color used in place of individual color for clarity. Images digitally scaled for display purposes using Fiji/ImageJ [68].
Figure 6.
Subcellular distribution of β-catenin and E-cadherin the colon crypt compartments.
Isolated crypts images were divided into five evenly divided regions (from the top of the crypt to the base, Figure 5B) with levels of β-catenin and E-cadherin quantified. The relative integrated intensities of β-catenin (A) and E-cadherin (B) in the nuclear, cytosol and membrane compartments were quantified for each region and crypt population. The E-cadherin integrated intensities were background subtracted with the nuclear intensities (no E-cadherin is expected in the nucleus) of the respective regions. Nuclear β-catenin is higher at the bottom of the crypt while the membrane compartments appear to be lower at the middle regions of the crypt. An increasing E-cadherin level was observed towards the top of the crypt for both the cytosol and membrane compartments. Note: 32 crypts were quantified (n=32, 20 HE-crypts and 12 HO-crypts). The error bars depict the RMS standard error of the mean (SEM).
Figure 7.
β-catenin and E-cadherin levels in budding colon crypts.
Immunofluorescent 2D montage images extracted from 3D image stacks of isolated colonic crypts undergoing budding in (A) stage 1, (B) stage 2 and (C) stage 3 respectively, with the overlaid RGB, DAPI, β-catenin, E-cadherin images and 3D reconstruction shown. The image sequence indicates the daughter crypt extending from the bottom half of the parent crypt (stage 1) and extends upwards (stage 2) before gradually growing into mature crypts (stage 3). Non-uniformity in the β-catenin and E-cadherin staining can be observed (HE-crypt population) at localized regions near the base of the crypts. Crypt formation process is asymmetrical with the daughter crypt developing as a bud from the lower half of the parental crypt. A movie of the 3D budding crypts (stage 1) is available in the supporting information Video S1. Note: 3D images and movies were produced using Fiji/ImageJ 3D Viewer[68]. Intensity calibration microspheres are images together with the crypts for calibration and intensity standardization within and between images. Each image of the crypt is made up from more than 200 2D confocal images sections to generate the 3D reconstruction. Section numbers of 2D confocal images displayed in the individual montage are shown below each figure.
Figure 8.
Schematic of hypothesized aberrant crypt budding.
Perturbation of the normal crypt production sequence is hypothesized to lead to deregulated crypt budding where the parental crypt starts budding before the original daughter crypt has finished its crypt production sequence. This leads to multiple daughter crypts on one parent crypt consistent with the crypt budding seen on adenomas (see Figure 1E and 1G in ref [23]).