Early Bmi1⁺ lineage detection. (a–h) Bmi1^CreER/+;Rosa26^LacZ/+ mice were injected with tamoxifen, and after 20, 22, 24 or 30 h, the duodenum and the jejunum were stained in whole mount for LacZ and then longitudinally sectioned. Crypts cut along their major axis containing only single cell labeled were analyzed. a–c,e,g,h show representative pictures of crypts with a single LacZ⁺ cell present at position +4, +5 (yellow arrowheads). c,d,f show representative pictures of crypts in which more than one labeled cell is present. In c, the tip of the second cell is barely visible. Black arrowheads in d,f indicate cells that have already started migrating upward. (i–k) In situ hybridization of the small intestine with a control probe (i) labeling the Paneth cells and with a Bmi1 probe (j). Bmi1mRNA is present only at the bottom of the crypts in cells located above Paneth cells (red arrowhead). Digital magnification (2.5×) (k) of the crypt highlighted in j shows Bmi1 in a cell (black dashed line) located above a Paneth cell (red dashed line); next to the Paneth cell a crypt base columnar cell is visible (yellow dashed line). All photographs (except k) were captured at the same magnification. Scale bar, 50 µm.

Repopulation kinetics of the Bmi1⁺ lineage. (a–l) Small intestines were harvested from Bmi1^CreER/+;Rosa26^LacZ/+ mice at different time points from day 2 to day 30. We analyzed and followed crypts on serial adjacent sections and observed that, even after 19, 24 and 30 d, many crypts were half filled by the Bmi1⁺ lineage. (m) To evaluate the repopulation kinetics, fully labeled crypts were counted at different time points after tamoxifen injection. All the pictures were captured at the same magnification. Scale bar, 50 µm.

Bmi1⁺ lineage analysis in the small intestine. (a–e) After a single injection of tamoxifen, the Bmi1⁺ lineage is present from day 5 up to 9 months. The Bmi1⁺ lineage in the whole intestine is more abundant in the duodenum (top left corner) and in the first part of the jejunum, with almost no staining in the ileum (bottom right corner). After 1 month, there is an apparent modest progressive reduction in the number of labeled crypts. (f) The intestine of a 4-month-old Bmi1^CreER/+;Rosa26^LacZ/+ mouse does not show any LacZ staining without tamoxifen induction. (g,h) Two Bmi1^CreER/+; Rosa26^LacZ/+ mice received tamoxifen injection for 3 consecutive d (g), whereas two other mice with the same genotype were treated three times at 5-d intervals (h). Five days after the last injection, the intestines were harvested and stained for LacZ. (i) Intestinal mucosa of a Bmi1^CreER/+; Rosa26^LacZ/+ mouse 4 months after a single tamoxifen injection. The Bmi1⁺ lineage is visible on villi as a ‘line’ of cells coming from the crypt. White scale bar, 5 mm; black scale bar, 1 mm.

Assessment of Bmi1⁺ lineage after 5 d and after 12 months to evaluate the colocalization with differentiated-cell markers. (a–c) Five days after tamoxifen injection, the Bmi1⁺ lineage (YFP; green arrowheads) is present at the bottom of a crypt. The antibody to lysozyme (Lys) identifies Paneth cells (a). DBA stains Goblet cells (b). (c) Intestinal section showing staining with YFP, indicating the Bmi1⁺ lineage; chromogranin A (ChGA), labeling enteroendocrine cells (red arrowhead); and lysozyme antibody, indicating Paneth cells (yellow arrowhead). (d–l) After 12 months, the Bmi1⁺ lineage is present in all differentiated intestinal cell types. In d, the Bmi1 lineage is expanding from a crypt all the way to the top of an adjacent villus; several goblet cells (DBA⁺) are present in the Bmi1 lineage. e is a 2×magnification of d. The white arrowhead indicates a goblet cell. Another Bmi1⁺ goblet cell is shown in f (white arrowhead). g–i show three examples of enteroendocrine cells (ChGA⁺; white arrowheads) within the Bmi1 lineage. Two transverse (j,l) and one longitudinal (k) section of Bmi1⁺ crypts, showing that Paneth cells (Lys⁺; white arrowheads) are present in the Bmi1 lineage. Scale bars, 25 µm.

Bmi1^CreER/+;Ctnnb1^Ex3LoxP/+ cross. Forty-five days after tamoxifen injection, the duodenum and the jejunum contain multiple adenomas. (a,b) Low (a) and high (b) magnifications. (c) The adenomas were BrdU positive. (d,e) Low (d) and (e) high magnifications. At day 30, only a few adenomas are visible. In many areas with or without adenomas, many crypts were in fission (red arrowheads). (f,g) Many crypts showed numerous periodic acid–Schiff (PAS)-positive cells (Paneth and goblet cells) present (f), indicating an expansion of the secretory compartment compared to the wild-type intestine (g). (h,i) Low (h) and (i) high magnification. At day 22 only rare adenomas were visible. (j) After staining with Ki67, many crypts showed increased number of proliferating cells crowded at the position where ISCs are located (white arrowheads). The yellow arrowhead indicates an adjacent crypt showing the normal Ki67 staining. (k) β-catenin staining, showing ubiquitous expression. Scale bars: 25 µm (b,e,i); 50 µm (j,k).

Intestinal stem cell ablation. After one injection of tamoxifen, Bmi1^CreER/+;Rosa26^DTA/+ mice were killed at different time points and their small intestine analyzed. (a–i) Representative pictures at 6, 11 and 22 d showing patches of intestinal mucosa disorganized and in disarray, without crypts. The ablation of the Bmi1⁺ ISC lineage is responsible for the crypt loss. Scale bar, 50 µm.