Rompani and Cepko. 10.1073/pnas.0709979104. |
Fig. 8. Number of clones containing one, two, or three subtypes of HCs, arranged by number of HCs per clone and age of infection. (a) clones derived from E1.5 infections. (b) Clones derived from E3 infections. (c) Clones derived from E4 infections. (d) Clones derived from E5 infections. Blue = three subtypes. Red = two subtypes. Yellow = one subtype. Green = total. x axis = number of HCs in a given clone.
Fig. 9. Percentage of all HCs that are each subtype, per age of injection, and per clone. (a) E1.5 injection; (b) E3 injection. (c) E4 injection. Blue = percentage of H1, red = percentage of H2 cells, yellow = percentage of H3 cells. The percent of each subtype, relative to the total number of HCs in a clone, in each clone from a given age of infection were averaged, and a standard deviation was calculated.
Fig. 10. The percentage of clones with each HC subtype from each age of infection is shown. x axis = age of injection. y axis = Percent of clones with the subtype. Blue tetrahedron = % of clones with H1, red square = % of clones with H2, green triangle = % of clones with H3.
Fig. 11. Calretinin is a marker for H1 HCs, TrkA for H3 HCs, and Prox1 for all HCs. (a-c) HCs in clones stained for Calretinin, showing H1 (a), H2 (b), and H3 cells (c). (d-f) Clones stained with anti-TrkA, showing H1 (d), H2 (e), and H3 cells (f). (g-i) Clones stained with anti-Prox1, showing H1 (g), H2 (h), and H3 cells (i).
Fig. 12. A portion of a retinal flat mount with a clone composed of only two HCs is shown. The clone was derived from an E4 infection and is composed of two H1 cells. Scale bar = 40 mm. Inset = higher magnification to demonstrate the H1 cell morphology, with the arrows indicating the cell bodies and the arrowheads the processes.
Movie 1. Computer-generated model of lineage tracing in the avian retina to demonstrate the greater tangential migration of early-born cells over late-born cells.
Movie 2. 3D rendering of clone shown in Fig. 2 d-f. Red = H1 cells, Yellow = H2 cells, Purple = H3 cells.
Table 2. The frequencies and combinations of HC subtypes in clones that contained only one or two HCs
Subtypes found in one HC-containing clones are heavily skewed toward the H2 subtype. (a) The frequency of each HC subtype observed in the clones with only one or two HCs are listed. The frequencies of HC subtypes in all clones generated by infection at E3-E5 are also listed, as clones with only one or two HCs were generated by infection at these ages. In addition, the frequencies of HC subtypes in all clones that had only one or two HCs are considered, because these clones likely were generated by progenitor cells infected near the end of the HC-producing window, given that they made very few HC and thus would give the best correction for any order of birthdates for HC subtypes. The observed number of each subtype of HC in clones with only one HC was compared to the number of each subtype predicted by the frequency of each subtype among all clones from E3-E5 infection, or among clones from this age of infection with only one or two HC. A c2 test was used to determine if the observed number of each subtype was as predicted, with the associated P values listed. Subtypes found in two HC-containing clones are heavily skewed away from the H2 subtype. (b) The observed number of each subtype of HC in clones with 2 HC was compared to the number of each subtype predicted by the frequency of each subtype among all clones from E3-E5 infection, or among clones from this age of infection with only one or two HC (from a). A c2 test was used to determine if the observed number of each subtype was as predicted, with the associated P values listed. (c) Pairs of cells found in two HC-containing clones are heavily skewed toward H1 pairs and H3 pairs. The observed numbers of pairs of HC subtypes in clones with two HC were compared to the pairs predicted based upon either the frequency of each subtype in observed clones with only two HC or upon the frequency each subtype in clones with one or two HCs, or upon the frequency of each subtype in all clones from infection at E3-E5. A c2 test was used to determine if the observed number of each pair of subtypes was as predicted, with the associated p values listed for the H1+H1, H1+H3, and the H3+H3.
Table 3. Analysis of even and odd numbers of each subtype in clones from infections at all ages
(a) The number of clones with the indicated number of each subtype, i.e., the number of clones with a particular number of subtypes, and how often a clone with that number of subtypes included an H1, H2, or H3. (b) An analysis of the frequency of clones with even and odd numbers of each subtype, including data from all clones. The expected values for an even and odd number of each subtype in the total collection of clones is shown. A c2 analysis and associated p value are shown. H2 cells were more often found as an odd number of H2 cells, and H3 cells were more often found as an even number of H3 cells in clones than predicted by chance. (c) An analysis of the even and odd number of each subtype in clones with more than one of each indicated subtype is shown. The expected values for an even and odd number of each subtype in the total collection of clones is shown. If one cell clone is excluded, H2 cells were found in even or odd numbers as predicted by chance, while H1 and H3 cells were found more often in even numbers, likely due to their production as pairs in a terminal division. A c2 analysis and associated P value are shown. (d) An analysis of the even and odd number of each subtype in clones with more than three of each indicated subtype are shown. The expected values for an even and odd number of each subtype in the total collection of clones is shown. A c2 analysis and associated P values are shown. Clones with four or more H1 or H3 cells were more likely than clones with H2 cells to have H1 and H3 cells as even numbers, perhaps due to their genesis as pairs, but cell death likely pruned these clones to reduce the significance of even numbers of H1 and H3 cells.
SI Text
Immunocytochemistry
Retinas were fixed in 4% PFA for 30 min at room temperature, washed three times with PBS, and then incubated overnight in 30% sucrose at 4°C. Retinas were subsequently removed from the mesh, placed onto slides, and with some residual sucrose on the retinas, freeze-thawed in dry ice three times to increase signal intensity. Retinas were then immobilized between mesh again and washed three times in PBS for 30 min; blocked with 10% NGS, 1%BSA, and 0.1% TritonX; then incubated with rabbit anti-GFP antibody (1:1,000, Molecular Probes) for 3 days at room temperature in the dark. Retinas were washed three times in PBS for 30 min again and incubated with goat anti-rabbit A488 (1:500, Molecular Probes) and DAPI (2 mg/ml) overnight at room temperature in the dark. They were then washed three times in PBS for 30 min and mounted with ProLong Gold (Molecular Probes) between coverslips and left to cure overnight at room temperature in the dark before imaging. All solutions except the ProLong Gold were 0.45 mm filtered to prevent particulate staining artifacts.
Statistics
:c2 statistics for the significance of clones containing one or two HCs was conducted with an expectation that all HC subtypes should be randomly distributed according to their frequency in clones with a particular number of HCs (SI Table 2). For the probability of odd- vs. even-number of HC subtypes, the expected was a 50/50 split between number of clones with an odd and even number of a particular subtype. Further statistics were conducted as described in SI Tables 2 and 3.