Notch signalling regulates asymmetric division and inter-conversion between lgr5 and bmi1 expressing intestinal stem cells

Rapidly cycling LGR5+ intestinal stem cells (ISCs) located at the base of crypts are the primary driver of regeneration. Additionally, BMI1 expression is correlated with a slow cycling pool of ISCs located at +4 position. While previous reports have shown interconversion between these two populations following tissue injury, we provide evidence that NOTCH signaling regulates the balance between these two populations and promotes asymmetric division as a mechanism for interconversion in the mouse intestine. In both in vitro and in vivo models, NOTCH suppression reduces the ratio of BMI1+/LGR5+ ISCs while NOTCH stimulation increases this ratio. Furthermore, NOTCH signaling can activate asymmetric division after intestinal inflammation. Overall, these data provide insights into ISC plasticity, demonstrating a direct interconversion mechanism between slow- and fast-cycling ISCs.


Animal experiments for Mouse Intestinal Stem Cell Studies
LGR5-EGFP and LGR5-EGFP-creER/POFUT-1 flox/flox mice on a mixed 129/C57BL/6 background were a generous gift from Dr. Leonard Augenlicht's laboratory. Mice were genotyped to confirm allelic identity using PCR analysis with LGR5 common forward primer:  FlowJo software was used to analyze data and to gate populations according to 7-AAD viability, and forward and side scattering. Cutoff thresholds were provided by using unstained cells as a negative control.
Single LGR5-EGFP+ ISCs were embedded in 50ul Matrigel (BD Biosciences) seeded on pre-warmed 24-well plates at a concentration of 1000 cells or crypts/mL. Following Matrigel polymerization, complete ISC media was added to each well. The formulation for ISC media is

In vitro Mouse Intestinal Stem Cell Studies
For in vitro studies, LGR5-EGFP organoids were seeded on chamber slides and treated with one of the following: 10uM DAPT (EMD Millipore) added to the media for 48 hours [2], or embedded in Matrigel containing 1uM JAG-1 (AnaSpec) for 48 hours [3].
LGR5-EGFP-creER/POFUT-1 flox/flox ISCs were treated with 500nM 4-hydroxytamoxifen (Sigma) added to the media for 48 hours to induce Cre recombinase and generate LGR5-EGFP-creER/POFUT-1 -/-ISCs [4]. For FACS analysis, harvested organoids were mechanically dissociated and resuspended in 2 mls PBS and passed through a 40um filter to collect single cells. After cell counting, the suspension was incubated with 7-AAD dye to assess viability and a BMI1 primary antibody that was subsequently labeled with PE.
LGR5-GFP expression was detected by FITC emission. FlowJo software was used to analyze data and to gate populations according to 7-AAD viability, and forward and side scattering. Cutoff thresholds were provided by using unstained cells as a negative control. LGR5-EGFP-CreERT2 x Rosa26-YFP-NICD knock-in mouse strain was treated daily with 75 mg/kg Tamoxifen for 8 consecutive days to induce Cre enzyme activity and NICD-OE phenotype. All murine frozen tissue specimens were fixed with 4% PFA for coimmunofluorescence using BMI1, LGR5 (detected by GFP), or ASCL2 expression according to the protocol described below. DAPI (Invitrogen) was used as a counterstain for IF on the inverted fluorescent microscope. Murine frozen tissue sections were also used for staining with Hematoxylin and Eosin according to standard protocol.

Immunofluorescence of Murine Crypts or in vitro organoids
LGR5 and BMI1 expression was quantified within individual crypts of n = 5 mice/condition with n=500 crypts/mouse measured.
Intestinal organoids embedded in Matrigel that were treated under stated conditions were either fixed with 4% PFA for 15 minutes at room temperature for co-immunofluorescence according to the established protocol below or harvested for protein analysis via Western Blotting using methods previously described [6]. Visualization of organoids was performed using BMI1 and LGR5 (detected by GFP) expression and ToPro-3 as a nuclear counterstain on a Zeiss LSM 510 laser scanning confocal microscope using an Apo 40× 1.40 oil objective. Images were collected and analyzed with confocal software (LSM 510 Meta; Zeiss). The experiment was performed in triplicate and LGR5 and BMI1 expression was quantified within individual organoids using n=500 organoids/replicate.

Immunofluorescence
Following fixation, cells were permeabilized with 0.2% Triton X-100 and incubated in a blocking solution (5% BSA or normal serum (goat, rabbit or horse) and 0.1% Triton-X in PBS) for 1 hour. For single or co-immunofluorescence staining, primary antibodies diluted in blocking solution were added overnight at 4°C overnight. To ensure specificity, a no primary antibody control staining was performed. The slides were then washed in PBS and incubated with the appropriate secondary antibody for 1 hour at room temperature and counterstained/mounted with Vectashield containing DAPI (Vector Laboratories). For ASCL2 staining, antigen retrieval was performed as previously described [7] and incorporated into the protocol for IF.