Article Figures & Data
- Figure 1.
Identifying morphologic changes that precede neuronal death using computational quantitation. A, Representative BioStation CT images of human Islet1::GFP-positive MNs in basal (TF+) and withdrawal (TF–) conditions at key experimental time points: day 1 (1 day after TF withdrawal), day 6, and day 10 of imaging (scale bar = 40 µm). B–E, Time series plots comparing the populations of Islet1::GFP MNs in TF+ (green) and TF– (magenta) conditions. Quantitation was done at each time point for the number (#) of MNs (B), neurite length per cell (C), number of nodes per cell (D), and cell body size per cell (E). The arrow indicates time of TF withdrawal (day 1). The average fold change relative to day 0 for all of the measured parameters is shown. A significant different between TF+ and TF– over time was found by two-way repeated-measures ANOVA after Bonferroni correction in the number of MNs (p < 0.001, F = 13.63, DFn = 1; B), neurite length per cell (p < 0.01, F = 5.27, DFn = 1; C), number of nodes per cell (p < 0.001, F = 30.57, DFn = 1; D) and cell body size per cell (p < 0.001, F = 11.54, DFn = 1; E), and significant differences between TF+ and TF– were observed at the end point by Bonferroni post-tests: **, p < 0.01; ***, p < 0.001. Data presented as mean + SEM. (n = 4 biological replicate experiments, each with three technical replicates.) F, Comparison of the effect magnitudes between TF+ and TF– for number of MNs (B) and number of nodes per cell (D) at day 12. Data presented as mean + SEM. *, p < 0.05 by t test (n = 4 biological replicate experiments, each with three technical replicates). G, Extracted time series data showing fold change measurements in features between day 1 and day 12. Data presented as mean + SEM. *, p < 0.05; **, p < 0.01; ***, p < 0.001 by t test (n = 4 biological replicate experiments, each with three technical replicates).
- Figure 2.
Measuring MN rescue responses following TF addback or kenpaullone treatment. A, Experimental design of TF withdrawal and addback imaging assays. For all experiments, TF withdrawal began 1 day after the start of imaging defined as day 0. For TF addback, the TFs- BDNF, GDNF, and CNTF– were reinstated to the culture media after 6, 7, or 8 days of TF withdrawal. B, Representative images of Islet1::GFP MNs at day 1, 6, and 11 in TF+, TF addback at day 6, and TF–. C, Fold change in number (#) of MNs between day 6 and day 14 following TF addback at day 6, 7, or 8. All data presented as mean + SEM. *, p < 0.05; ***, p < 0.001 by t test; all compared to TF– conditions (n = 5 biological replicate experiments, each with three technical replicates). D, E, Time series plots depicting neurite length (D) and number of nodes per cell (E) show the MN population behavior in TF– and TF addback conditions. Key experimental time points are denoted with arrows: the start of TF withdrawal (red), day 6 addback (dark blue), day 7 addback (light blue), and day 8 addback (orange). Two-way repeated-measures ANOVA showed significant differences in neurite length per cell (p < 0.001, F = 6.555, DFn = 4; D) and number of nodes (p < 0.05, F = 3.356, DFn = 4; E) over time under the different conditions. Subsequent Bonferroni post tests indicated TF addback at day 6, 7, and 8 increased neurite length per cell (D) and the number of nodes (E) per cell compared to TF– at the end point. All data presented as mean + SEM. **, p < 0.01; ***, p < 0.001 by two-way repeated-measures ANOVA with Bonferroni correction, all compared to TF– conditions (n = 5 biological replicate experiments, each with three technical replicates). F, Experimental design of imaging assays kenpaullone addition to cells undergoing TF withdrawal. Analysis window between day 6 and day 14 was used for both evaluation of survival effect (G) and temporal quantitation of morphometric parameters (H, I). G, Fold change in # of MNs between day 0 and day 14 of kenpaullone treatment. All data presented as mean + SEM. *, p < 0.05; ***, p < 0.001 by t test; all compared to TF– conditions (n = 5 biological replicate experiments, each with three technical replicates). H, I, Time series plots depicting neurite length (H) and number of nodes per cell (I) after kenpaullone treatment. Kenpaullone (light and dark purple) was administered from the start of TF withdrawal (magenta) as denoted with arrows. All data presented as mean + SEM. **, p < 0.01; ***, p < 0.001 by two-way repeated-measures ANOVA with Bonferroni correction, all compared to TF– conditions (n = 5 biological replicate experiments, each with three technical replicates).
- Figure 3.
Classifying MNs according to their number of nodes. A, Classification of the heterogeneous MN population according to their number of nodes (0, 1, 2, 3 or 4+ nodes) determines the class assignment (class A, B, C) for each MN. B, C, A stacked temporal area plot displaying the number of MNs with 0, 1, 2, 3 or 4+ nodes, quantified at each time point throughout the imaging experiment. In this stacked area graph, the height of each colored region on the y-axis indicates the number of MNs in each number of nodes category. (B), After the first few days in TF+, the total number of MNs remains relatively constant. MNs mature over time and eventually MNs with 4+ nodes become the majority of the population. (C), In TF–, the total number of MNs gradually decreases over time and the MN population comprises a large percentage of cells with 2 or 3 nodes and relatively fewer neurons with 4 or more nodes. The magenta arrow indicates the time of TF withdrawal. D, A histogram of fold change (relative to day 0) in the number of MNs with 4 or more nodes in the TF+ and TF– conditions respectively. During the time period, the number of MNs with 4 or more nodes in TF+ and TF– conditions becomes increasingly different (p < 0.05, F = 3.949, DFn = 2 by two-way repeated-measures ANOVA with Bonferroni correction). All data presented as mean + SEM. *, p < 0.05. (n = 5 biological replicate experiments, each with three technical replicates.)
- Figure 4.
A Single-cell tracking algorithm to measure the lifespan of MNs. A, Representative images of cell tracking. Three MNs distinguished with different colors (green, purple, orange) were tracked by time lapse imaging using CL-Quant software The software was able to assign individual identities to two cells that were overlapping (middle image) using cell body size and position information collected over time. B, Schematic diagram of the class A hours measurement. Class A MNs were defined at the beginning of the imaging period and then tracked. The same analysis window (between day 6 and day 14) was used for all experiments. Using single-cell tracking with a mask for node number, the lifespan hours of class A MNs were measured and averaged during the selected analysis window. For each MN, class A hours ended when class A MNs changed to class B or C. C, D, Normalized class A hours (relative to TF+) for TF addback (C) and kenpaullone (D). Table 1 lists the number of tracked MNs for the TF addback experiments, and Table 2 provides the numbers of tracked MNs for the kenpaullone experiments. Data presented as mean + SEM. **, p < 0.01; ***, p < 0.001 by t test all compared to TF– (n = 5 biological replicate experiments, each with three technical replicates).
- Figure 5.
Tracking cell class transitions of individual MNs in TF withdrawal, TF addback, and kenpaullone conditions. Cells were categorized as either class A or class B MNs as shown in Fig. 3A and then individually tracked to determine if they remained in the same class at the end of the analysis window. Table 1 details the class transitions for all tracked MNs in the TF addback experiments, while Table 2 provides this information for the kenpaullone experiments. A, MNs identified as class A on day 6 were tracked until day 14 in each treatment condition and were categorized according to their class transition in a stacked histogram (n = 5 biological replicate experiments, each with three technical replicates). B, Stacked histogram of class transitions from day 6 to day 14 of class B MNs in each treatment condition (n = 5 biological replicate experiments, each with three technical replicates). C, The ratio of class A to A MNs (from day 6 to day 14) in each different treatment condition, relative to TF+ control, quantifies the maintenance of class A MNs. TF addback at day 6 shows a significant rescue effect, while TF addback at day 7 or 8 is less effective. Data presented as mean + SEM. *, p < 0.05; ***, p < 0.001 by t test; all compared to TF– (n = 5 biological replicate experiments, each with three technical replicates). D, The ratio of class B to A MNs (from day 6 to day 14) in the different treatment conditions relative to TF+ quantifies the rescue of class B MNs to class A MNs. Data presented as mean + SEM. *, p < 0.05; ***, p < 0.001 by t test; all compared to TF– (n = 5 biological replicate experiments, each with three technical replicates). E, Maintenance of class A to A MNs from day 0 to day 14 in different treatment conditions relative to TF+. Kenpaullone significantly maintained class A MNs as class A compared to TF withdrawal. Data presented as mean + SEM. **, p < 0.01; ***, p < 0.001 by t test; all compared with TF– conditions (n = 5 biological replicate experiments, each with three technical replicates). F, Changes in the ratio of class B to A MNs from day 0 to day 14. Kenpaullone did not rescue many class B MNs to class A. Data presented as mean + SEM. **, p < 0.01; ***, p < 0.001 by t test; all compared with TF– conditions (n = 5 biological replicate experiments, each with three technical replicates).
- Figure 6.
Characterization of key morphologic features of rescuable class B MNs using reverse tracking. A, B, Representative single cell analysis extracted using CL-Quant. Graphs show tracked number of nodes (A) and cell body size (B) of two individual MNs over time. In this plot, both MN #1 and MN #2 became class B MNs after TF withdrawal at day 6. MN #1 recovered to class A following day 6 of TF addback. C, D, Common features of rescuable class B MNs at day 6, 7 and 8. At the time point immediately before TF addback, class B MNs that could develop into class A MNs had more nodes (C) and a larger cell body size (D) compared to class B MNs that could not regrow neurites. Data presented as mean + SEM. ***, p < 0.001 by t test; all compared with TF–conditions (n = 5 biological replicate experiments, each with three technical replicates).
- Table 1.
Numbers of tracked MNs to evaluate cell class transitions following TF addback (day 6–14)
Outcome of fate transition Outcome of fate transition Conditions Total tracked # of MNs Tracked # of Class A MNs Class A to A Class A to B Class A to C Tracked # of Class B MNs Class B to A Class B to B Class B to C TF+ 110.9 ± 32.05 83.33 ± 30.16 62.27 ± 22.07 2.267 ± 1.024 18.80 ± 7.240 27.57 ± 4.540 9.533 ± 2.421 3.300 ± 0.9638 14.73 ± 2.379 TF addback at day 6 93.60 ± 29.36 58.33 ± 24.68 40.17 ± 16.36 2.033 ± 1.311 16.13 ± 7.285 35.27 ± 6.954 10.43 ± 2.834 3.167 ± 0.7853 21.67 ± 4.001 TF addback at day 7 97.97 ± 28.09 46.92 ± 16.25 26.56 ± 11.81 1.333 ± 0.5164 19.03 ± 4.254 44.67 ± 11.90 33.90 ± 17.07 2.100 ± 0.9422 21.27 ± 6.517 TF addback at day 8 110.3 ± 31.51 57.27 ± 24.46 33.90 ± 17.07 2.100 ± 0.9422 21.27 ± 6.517 55.54 ± 11.60 25.83 ± 13.03 5.600 ± 1.431 24.43 ± 6.128 TF– 101.6 ± 30.11 55.87 ± 19.52 25.83 ± 13.03 5.600 ± 1.431 24.43 ± 6.128 45.73 ± 12.47 5.267 ± 2.125 10.70 ± 4.803 29.77 ± 6.136 Data are presented as mean ± SEM (n = 5).
- Table 2.
Numbers of tracked MNs to evaluate cell class transitions during kenpaullone treatment (day 0–14)
Outcome of fate transition Outcome of fate transition Conditions Total tracked # of MNs Tracked # of Class A MNs Class A to A Class A to B Class A to C Tracked # of Class B MNs Class B to A Class B to B Class B to C TF+ 86.20 ± 17.08 42.47 ± 13.86 27.47 ± 10.07 0.6000 ± 0.2449 14.40 ± 4.128 43.73 ± 6.322 14.87 ± 4.121 1.733 ± 0.5907 27.13 ± 3.023 TF+/Ken 5uM 80.27 ± 19.01 41.07 ± 15.52 27.60 ± 11.26 0.7333 ± 0.3399 12.73 ± 4.153 39.20 ± 6.538 17.00 ± 4.791 2.333 ± 0.5055 19.87 ± 2.277 TF–/Ken 2.5uM 98.27 ± 21.23 42.80 ± 16.05 21.40 ± 8.362 2.400 ± 0.8781 19.00 ± 7.551 55.47 ± 10.46 15.07 ± 4.993 7.000 ± 1.623 33.40 ± 5.499 TF–/Ken 5uM 96.87 ± 16.16 38.27 ± 11.93 18.67 ± 5.544 1.867 ± 0.7196 17.73 ± 5.743 58.60 ± 12.21 11.80 ± 4.196 7.200 ± 2.560 39.60 ± 7.318 TF– 92.53 ± 19.25 42.47 ± 14.90 11.13 ± 3.868 4.467 ± 1.698 26.87 ± 9.443 50.07 ± 6.998 6.867 ± 1.695 4.333 ± 1.038 38.87 ± 5.904 Data are presented as mean ± SEM (n = 5).
Location Description Data structure Type of test Statistical value a Fig. 1B–E Normal (Gaussian distribution) at the end point; number (n) of biological experiments = 4 with three technical replicates Two-way repeated-measures ANOVA after Bonferroni correction Number (#) of MNs : p < 0.001, F = 13.63, DFn = 1; Neurite length per cell: p < 0.01, F = 5.27, DFn = 1; # of Nodes per cell : p < 0.001, F= 30.57, DFn = 1; Cell body size per cell : p < 0.001, F = 11.54, DFn = 1 Unpaired two-tailed t-test at the end point # of MNs : p < 0.001; Neurite length per cell: p < 0.01; # of Nodes per cell : p < 0.001; Cell body size per cell : p < 0.001 b Fig. 1F Normal, n = 4 Unpaired two-tailed t-test p < 0.05 c Fig. 1G Normal, n = 4 Unpaired two-tailed t-test # of MNs: p < 0.01, Neurite length per cell: p < 0.05; # of Nodes per cell: p < 0.0001, Cell body size: p < 0.01 d Fig. 2C Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.001, TF addback at day 6: p < 0.05 e Fig. 2D, E Normal at the end point; n = 5 Two-way repeated-measures ANOVA after Bonferroni correction Neurite length per cell: p < 0.01, F = 6.555, DFn = 4; # of Nodes per cell: p < 0.05, F = 3.356, DFn = 4 Unpaired two-tailed t-test at the end point Neurite length per cell: TF addback at day 6, day 7, and day 8: p < 0.001, respectively; # of Nodes per cell: TF addback at day 6, day 7: p < 0.001; TF addback at day 8: p < 0.01 f Fig. 2G Normal, n = 5 Unpaired two-tailed t-test TF+ and TF+/Ken 5µM: p < 0.001; TF–/2.5µM and TF–/5µM: p < 0.05 g Fig. 2H, I Normal at the end point; n = 5 Two-way repeated-measures ANOVA after Bonferroni correction Neurite length per cell: p < 0.05, F = 4.01, DFn = 4; # of Nodes per cell: not significant Unpaired two-tailed t-test at the end point Neurite length per cell: TF+ and TF+/Ken 5µM: p < 0.001; # of Nodes per cell: TF+: p < 0.001,; TF+/Ken5 µM: p < 0.01 h Fig. 3D Normal at the end point; n = 5 Two-way repeated-measures ANOVA correction after Bonferroni correction Fold change of MNs with 4 or more nodes :p < 0.05, F = 3.949, DFn = 2 Unpaired two-tailed t-test at the end point p < 0.05 i Fig. 4C Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.001, TF addback at day 6: p < 0.001 j Fig. 4D Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.01, TF+/Ken 5 µM: p < 0.01 k Fig. 5C Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.001, TF addback at day 6: p < 0.001, TF addback at day 7: p < 0.05, addback at day 8: p < 0.05, l Fig. 5D Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.001, TF addback at day6: p < 0.001,; TF addback at day7: p < 0.05, addback at day8: p < 0.05 m Fig. 5E Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.001, TF+/Ken 5 µM: p < 0.001,; TF–/Ken 2.5 µM: p < 0.01, TF–/Ken 5µM: p < 0.01 n Fig. 5F Normal, n = 5 Unpaired two-tailed t-test TF+: p < 0.01, TF+/Ken 5 µM: p < 0.01 o Fig. 6C Normal, n = 5 Unpaired two-tailed t-test TF addback at day 6, day 7, and day 8: p < 0.001, respectively p Fig. 6D Normal, n = 5 Unpaired two-tailed t-test TF addback at day 6, day 7, and day 8: p < 0.001, respectively
- Video 1.
Representative video of H9-Islet1::GFP MN, captured every 6 hours for 14 days. Single-cell tracking software successfully identifies and tracks individual MNs. Green dot indicates center of the cell body. The red line maps the MN’s trajectory over time.
- Video 2.
Morphometric single-cell tracking analysis tracks neurite length (yellow), number of nodes (magenta), and cell body size (green) of MN over time. Because of the increasing complexity of the neuritic network, the number of nodes metric is better than neurite length for tracking at the single-cell level. Images were captured every 6 hours for 14 days.
- Video 3.
Single-cell tracking analysis of neurite length (yellow), number of nodes (magenta), and cell body size (green) of MN in four experimental conditions: TF (A), TF addback at day 6 (B), 5 µm kenpaullone and TF withdrawal (C), and TF withdrawal (D). TF was withdrawn at day 1 (1 day after live imaging initiation), and TF was added at day 6 (C), and 5 μm kenpaullone was added during the entire period in which MNs were maintained in the absence of TF. Images were captured every 6 hours for 14 days. In TF (A), individual MNs increased neurite length and number of nodes, and TF addback at day 6 (B) rescues both neurite length and number of nodes. Kenpaullone (C) is less protective compared to TF addback (B). In TF withdrawal (D), the neurites retract and the cell body size and number of nodes decrease.
In this issue
