Figure 1.
Crypt most recent common stem cell (MRCSC).
(A), Left - three whole crypts. DNA extracted from each crypt represents the MCRSC of the crypt. Right - lineage tree of the three crypts' MRCSCs (blue nodes). The root of the tree represents the MRCA (white node) of the three crypts. (B), Left – six individual cells sampled from a single crypt. DNA was extracted from each cell separately. Right - cell lineage tree of the six sampled cells (green nodes). The root of the tree (blue square) represents the crypt's computed MRCSC. (C), Left - DNA was extracted from three whole crypts and from six individually sampled cells of a fourth crypt. Right - the resulting lineage tree shows three crypt MRCSCs (blue nodes), six crypt cells (green nodes), and the computed MRCSC of these cells (blue square). The root of the tree represents the MRCA of the four crypts (white node).
Figure 2.
Hypothetical scenarios for intestinal stem cell dynamics.
Schematic representation of different possible scenarios of colon crypt dynamics. The Y axis represents cell depth. (A), Hypothetical cell lineage tree of a young mouse. In this tree, crypt MRCSCs (blue nodes) are shallow, individual cells from two crypts (red and green nodes) are separately clustered on the linage tree, and the depth of the computed MRCSC of each crypt (blue squares) is in the same range as the depth of crypt MRCSCs (blue nodes). (B, C, D), In the old mouse we depict three different hypothetical scenarios for stem cell dynamics. (B), Crypt stem cells maintain an immortal strand. As a result, the depth of crypt MRCSCs (blue nodes), crypt cells (red and green nodes) and computed crypt MRCSCs (blue squares) does not change with mouse age. (C), Crypt stem cells do not maintain an immortal strand and there is no monoclonal conversion. Hence all cells of each crypt are the progeny of the stem cell that originated the crypt two weeks after birth. As a result, crypt MRCSCs (blue nodes) and computed MRCSCs (blue squares) do not get deeper with mouse age. However, crypt cells (red and greed nodes) do get deeper when the mouse gets older. (D), Crypt stem cells do not maintain an immortal strand and there is ongoing monoclonal conversion. Hence all cells of each crypt are progeny of a fairly recent single stem cell ancestor regardless of mouse age. As a result the depth of crypt MRCSCs (blue nodes) and of crypt cells (red and green nodes) and their computed MRCSCs (blue squares) increase dramatically with mouse age.
Figure 3.
Reconstructed cell lineage trees show stem cell dynamics in colon and small intestine.
(A and B), Two reconstructed cell lineage trees of colon cells of 52 and 340 day-old mice, respectively. In the tree, DNA extracted from whole crypts is represented by blue nodes, single cells isolated randomly from an individual crypt by red and green nodes, and the computed crypt MRCSCs are represented by blue squares. Bold branches represent statistically enriched branches for cells from the same crypt. The width of the branches represents clustering significance (wider = lower p value). The Y axis measures cell depth. (C), Box plots of the depths of whole crypts in 52, 199 and 340 day-old mice. Crypt depth increases significantly between 52 and 199 day-old mice (p<10−8) and between 199 and 340 day-old mice (p<0.002). Y axis is the same as in A and B. The median depth is shown in each box plot. (D and E), Two reconstructed lineage trees of small intestine cells of 52 and 199 day-old mice, respectively. In the tree, DNA extracted from whole crypts is represented by purple nodes. The Y axis measures cell depth. (F), Box plots of the depths of whole crypts of 52 and 199 day-old mice. Crypt depth increases significantly between 52 and 199 day-old mice (p<10−6). Y axis is the same as in D. The median depth is shown in each box plot.
Figure 4.
(A), In a reconstructed lineage tree, colon crypts isolated by laser capture from submilimetric regions are clustered separately than crypts that were sampled randomly. The tree shows two groups of colon crypts isolated from two different submilimetric regions (cyan and purple) and randomly sampled crypts (blue) isolated from a 278 day-old mouse. Bold branches represent statistically enriched branches for colon crypts sampled from submilmetric regions in the colon. The most significant p value is indicated. The Y axis measures cell depth. (B), Median depths of randomly and submilimetric sampled crypts. Box plots of the depths of submilimetric (cyan and purple) and randomly (blue) sampled crypts. Y axis is the same as in A. The median depth is shown in each box plot.
Figure 5.
Colon crypts are clustered differently than B-lymphocytes, beta cells, and pancreatic duct cells.
(A), A reconstructed lineage tree of colon crypts (blue), B-lymphocytes (purple), beta cells (green), CD34 positive cells from the bone marrow (gray), and pancreatic duct cells (pink) in a 278 day-old mouse. Bold branches represent statistical enrichments on the lineage tree. The most significant p value is indicated. The width of the branches represents clustering significance (wider = lower p value). The Y axis represents depth. (B), Box plots of the depths of crypts (blue), B-lymphocytes (purple), beta cells (green), pancreatic duct cells (pink), and CD34 positive cells from the bone marrow (gray) of a 278 day-old mouse. (C), Box plots of the depths of beta cells (green) and B-lymphocytes (purple) of a 30 day-old mouse. (B and C), Y axis is the same as in A. The median depth of each cell type is shown in the box plots. Columns with different superscripts differ significantly (p<0.05), e.g. pancreatic beta cells depth is statistically significantly shallower than that of B-lymphocytes, and therefore pancreatic beta cells are marked with ‘a’ and B-lymphocytes with ‘c’.