Fig 1.
Colonospheres and organoids with crypts (budding organoids) at day 6 of culture.
Colon crypts were isolated from mouse and cultured embedded with Matrigel in growth factors containing DMEM/F12 media. Two distinct morphologies of cultures can be identified, namely: (A) Colonospheres, which are spherical or cystic in shape without crypt-like projections. (B) Colonoids, which are oblate ellipsoids shaped structures with crypt-like buds extending outwards into the Matrigel.
Fig 2.
WEHI-YH2 feeder layer co-culture stimulates stem cells in colon crypt cultures.
Epithelial cells were purified using EpCAM-PE/Cy7, CD44-PE and CD24-APC antibodies. (A) The CD44+CD24+ cells in quadrant 2 in the bivariate plot were gated. A further gate on SSCloCD44hi subset followed by CD24sub-lo/lo/med subsets were used to purify colon epithelial cells into 3 further subsets: EpCAM+SSCloCD44hiCD24sub-lo, EpCAM+SSCloCD44hiCD24lo and EpCAM+SSCloCD44hiCD24med. (B) Representative images of colonoids formed after the 3 subsets of colon cells were embedded in Matrigel and co-cultured with a monolayer of WEHI-YH2 cells for 8 days. Scale bar = 50 μm. (C) The formation efficiency (%) of colonospheres and colonoids in day 8 culture. Error bars: SEM n≥3, 1-way ANOVA, *p<0.05.
Fig 3.
Concentrated medium collected from WEHI-YH2 cells (cYH2CM) stimulates colon crypt formation.
Conditioned Medium (CM) from WEHI-YH2 cells (YH2CM) was collected from confluent cultures incubated for different times (24, 48 and 72 hours) in DMEM/F12 with the N2 and B27 supplements. The CM was concentrated 10-fold using a Centriprep YM3 filter (with a 3kDa molecular weight cut-off) (cYH2CM). Different amounts of YH2CM and cYH2CM were used to grow colon crypts in culture and the results are compared with the WEHI-YH2 colon crypt co-cultures. Images of the cultures were taken on Day 0 and Day 6, processed and scored for either colonospheres or organoids with crypts: (A) colonosphere- and (B) colon crypt–forming efficiency on Day 6 (the number of colonospheres or colonoids on Day 6 divided by the number of crypts plated on Day 0). The dotted lines show the colonosphere and colon colonoid formation efficiency in co-cultures with WEHI-YH2 cells. Error bars: SEM, n = 4 from 2 independent experiments of duplicates, 2-way ANOVA multiple comparisons Tukey’s test.
Fig 4.
cWEHI-YH2 CM stimulates colon crypt formation maximally.
Different proportions (2.5% to 50%, v/v) of ten-fold concentrated YH2CM conditioned media (cYH2CM, using the Vivaflow 50R) was used to grow colon crypts in culture. Corresponding co-cultures with WEHI-YH2 cells and without cYH2CM were used as positive and negative control respectively. Whole well image stacks of the cultures were acquired on days 0, 2, 4 and 6. Images were quantitated for average counts, sizes, crypts per organoid as well as the efficiency for formation of Colonosphere (A and C) and colonoid (B, D-F). For colonosphere formation, 12.5% of cYH2CM is sufficient to match the effect of YH2-CM co-culture in maintaining spheroid counts (A) and promoting increase in spheroid size (C). For forming colonoids, 25% cYH2CM was needed to match the effect of the WEHI-YH2 feeder layers in terms of counts (B) and proliferation (D). (E) When used at ≥12.5% (v/v) the cYH2CM improves the average crypt budding rate per colonoid significantly (compared to the co-cultures). (F) The colonosphere and colonoid formation efficiencies at Day 6 (compared to Day 0) using ≥25% cYH2CM (v/v) is comparable to the WEHI-YH2 co-cultures. Error bars: SEM, n = 3 experiments of duplicates, 2-way ANOVA multiple comparisons (A, B) Dunnett test, (C,D,E) Tukey’s test.
Fig 5.
Factors simulating colon crypt cultures appear to be proteins.
To determine the biochemical characteristics of the factors that simulate colon crypt formation in vitro, thermal or enzyme treatments were applied to concentrated YH2CM-DMF conditioned media (cYH2CM). The treated cYH2CM were used to stimulate colon crypts in culture at 50% v/v and the results compared with that of the untreated cYH2CM. Whole well image stacks of the cultures were acquired on days 0 and 6. Images were quantitated and the formation efficiency of colonospheres and colonoids (% of Day 0) was calculated. (A) Heat treatment of cYH2CM at 68°C for 1 hour (denatured) obliterates the ability to grow colon crypt cultures. (B) Representative images showing the colonospheres formed in the presence of in cYH2CM, compared to no colonospheres in the cultures treated with the heat denatured cYH2CM. (C) cYH2CM was pre-treated with 100 μg/ml of trypsin (T) or denatured trypsin (Th, heated at 95°C for 10 minutes) at 37°C for 6 hours before deactivating the trypsin with 400 μg/ml of trypsin inhibitor (I). Trypsin treatment completely obliterates colonoid forming capability and reduces the colonosphere forming activity of cYH2CM. (D) Representative images of colon crypt cultures showing the differences between trypsin-treated cYH2CM cultures and cultures treated with inactivated trypsin-treated cYH2CM. Scale/Error bars and analysis: (A) ±SEM n = 4, Tukey's multiple comparisons test. (C) Scale bar: 50 μm, ±SEM n = 3, 1-way ANOVA multiple comparisons, Uncorrected Fisher's LSD test.
Fig 6.
The cYH2CM does not stimulate canonical Wnt and Notch signaling pathways.
The ability of the YH2CM to stimulate the Wnt or Notch signaling pathway was tested in cell lines by its ability to stabilize the β-catenin protein (for Wnt pathway) or Notch Intra-Cellular Domain protein (NICD, for Notch pathway). (A) L-cells were incubated with partially purified Wnt3a (0.2% and 0.5% v/v) or cYH2CM (10% and 50% v/v) for 4 hours before Western blotting for β-catenin and β-tubulin. The normalized β-catenin intensities were plotted indicating that cYH2CM does not stabilize β-catenin. (B) Western blot detecting β-catenin using mouse anti-human/mouse β-catenin monoclonal antibody (Clone 14/ β-catenin, BD Transduction Laboratories #610153) to probe the L-cell cell lysate. (C) HCT116 cells were incubated either with 50% (v/v) of DMEM (control) or cYH2CM for 1h, 3h and 72h. For EDTA treatment, the cells were incubated with EDTA/PBS (10 mM) for 15 min. Western blotting for NICD and β-tubulin was conducted on the cell lysate with the normalized NICD intensities plotted. cYH2CM does not stimulate NICD protein in β-catenin. (D) Western blot detecting NICD using rabbit anti-mouse cleaved Notch 1 monoclonal antibody (Clone D3B8, Cell Signaling #4147) to probe the HCT116 cell lysate. Error bars: SEM, n = 3, 2-way ANOVA Tukey’s test.
Fig 7.
Bone morphogenetic protein 4 (Bmp4) inhibits colonoid formation.
Recombinant human Bmp4 (50 and 100 ng/ml) and noggin (100 ng/ml) was added to the colon crypt cultures on Day 0, 2 and 4 and assayed in duplicates. (A) Representative bright field Extended Depth of Field (EDF) images of the cultures on Day 6. (B) The numbers of colonospheres and colonoids were scored on Day 6 and the colony formation efficiencies were determined and tabulated. Error bars and analysis: Mean ± SEM, * p< 0.05, n = 3, scale bar = 100 μm, 2-way ANOVA multiple comparisons, Tukey’s test.
Fig 8.
Inhibitors of TGF-β signaling enhanced colonoid formation.
(A) A83-01, inhibitor of the TGFβ receptor kinase, was assayed at 0.1, 0.3 and 0.5 μM in duplicate culture of colon crypts with cYH2CM (50% v/v). The numbers of colonospheres and colonoids were counted on Day 6 and the colonosphere and colonoid formation efficiencies were determined. (B) Representative images of cultures on day 6. Error bars and analysis: Mean ± SEM, n = 3, * p<0.05, scale bar = 100 μm. 2-way ANOVA Fisher’s LSD test.
Fig 9.
Optimum EGFR signaling is required for colonosphere viability and colonoid formation.
The tyrphostin AG1478 was titrated with 30% v/v cYH2CM without EGF into colon crypt cultures grown in the wells of a 384 well plate, and were imaged on Day 5, quantitated and tabulated. (A) Representative bright field Extended Depth of Field (EDF) images of the colon crypt cultures showing the decreasing counts with increasing AG1478. The (B) number and (C) size of colonospheres, colonoids and cysts (a spheroid with a large and transparent pseudolumen). (B) Colonosphere counts decrease with increasing AG1478 concentration. (C) An addition of 0.1μM AG1478 leads to an increase in colonoid size. Scale/Error bars and analysis: 500μm. Mean ± SEM, n = 3 experiments, 2-way ANOVA, Dunnett's multiple comparisons test.