Local caspase activation interacts with Slit-Robo signaling to restrict axonal arborization

Abstract

In addition to being critical for apoptosis, components of the apoptotic pathway, such as caspases, are involved in other physiological processes in many types of cells, including neurons. However, very little is known about their role in dynamic, nonphysically destructive processes, such as axonal arborization and synaptogenesis. We show that caspases were locally active in vivo at the branch points of young, dynamic retinal ganglion cell axonal arbors but not in the cell body or in stable mature arbors. Caspase activation, dependent on Caspase-3, Caspase-9, and p38 mitogen-activated protein kinase (MAPK), rapidly increased at branch points corresponding with branch tip addition. Time-lapse imaging revealed that knockdown of Caspase-3 and Caspase-9 led to more stable arbors and presynaptic sites. Genetic analysis showed that Caspase-3, Caspase-9, and p38 MAPK interacted with Slit1a-Robo2 signaling, suggesting that localized activation of caspases lie downstream of a ligand receptor system, acting as key promoters of axonal branch tip and synaptic dynamics to restrict arbor growth in vivo in the central nervous system.

Figure 1.

Local caspase activation at branch points in young RGC arbors. (A, D, E, and H–R) Representative IMD ratio images of 80-hpf live RGC arbors in the optic tectum (A, D, E, and H–P) and an RGC cell body (Q and R) in the retina. Caspase activation is represented by the pseudocolors that correspond to the Venus/ECFP ratio (1.5–0.5). Red represents low caspase activation, and violet represents high caspase activation. (A, D, E, and H–P) Control (A and D), SCAT3 (E and H), SCAT3 + coMO (I and J), SCAT3 + casp-3aMO1 (K and L), SCAT3 + casp-9MO1 (M and N), and SCAT3 + p38mapkMO1 (O and P) are shown. (B–D and G) ECFP channel (B and D) and Venus channel (C and G) of SCAT3- and control-expressing arbors are also shown. The region enclosed by the dashed squares in A, E, I, K, M, O, and Q are magnified in B–D, F–H, J, L, N, P, and R. (S) Quantification of Venus/ECFP ratios across the whole arbor. Branch points were determined as covering a distance of 1 µm or less away from where a branch tip joins an arbor. Surrounding regions were determined at a distance of 1 µm away from a branch tip and 2 µm in length, covering the width of the branches. (B–D and F–H) Examples illustrating regions of interest for the quantification of the Venus/ECFP ratio at branch points (purple dashed areas) and surrounding regions (white dashed areas) for control- and SCAT3-expressing arbors, respectively. (T) Quantification of Venus/ECFP ratios at branch points (BP) and surrounding regions (S). (U and V) Presented as a ratio of branch points/surrounding regions (U) and RGC cell bodies (V). White asterisks indicate the parent axon. 19–28 RGC arbors were analyzed in S. 29–36 branch points and 87–108 surrounding regions were analyzed in T and U. 15–16 RGC cell bodies were analyzed in V. Error bars represent SEMs. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bars, 10 µm. The image in E of a SCAT3-expressing arbor is shown again in Figs. 5 A and S2 C. The magnified image in H also is shown again in Fig. 5 B.

Figure 2.

Caspase activity is undetectable in older RGC arbors. (A–D) Representative IMD ratio images of 100-hpf live RGC arbors in the optic tectum; caspase activation is represented by the pseudocolors that correspond to the Venus/ECFP ratio (1.5–0.5). Red represents low caspase activation and violet represents high caspase activation. (A–D) Control (A and B) and SCAT3 (C and D) images are presented. The region enclosed by the dashed squares in A and C are magnified in B and D. (E–G) Quantification of Venus/ECFP ratios across the whole arbor for both 80-hpf (E; data as in Fig. 1 S) and 100-hpf (F) arbors at branch points (BP) and surrounding regions (S) presented as a ratio of branch points/surrounding regions (G). White asterisks indicate the parent axon. Dorsal views are shown, and anterior is up. 10–12 RGC arbors were analyzed per condition. Error bars represent SEMs. **, P < 0.01. Bars, 10 µm.

Figure 3.

Caspase activity rapidly increases upon young RGC arbor branch tip formation. (A) Imaging analysis scheme for 80-hpf arbors expressing SCAT3-forming branches during the 5-min imaging interval. Branch points were determined as covering a distance of 2 µm or less away from where a branch tip joins, and arbor nonbranch points were 5–20 µm away from a branch tip and 3 µm in length, covering the width of the branches as for surrounding regions (Fig. 1). (B–E) Representative IMD images of an 80-hpf RGC arbor expressing SCAT3. Caspase activation is represented by the pseudocolors that correspond to the Venus/ECFP ratio 1.5–0.5 or 1.0–0.5. Red represents low caspase activation, and violet represents high caspase activation. (B–D) 5 min before branch point formation (BP T − 5 min; B) and after branch tip (BP) formation (C). The region enclosed by the dashed squares in B and C are magnified in D and E. (F and G) Quantification of the Venus/ECFP ratio (F) comparing changes in arbors between regions that do or do not lead to the formation of branch points as in the scheme in A and presented as ratios (G). (H) Imaging analysis scheme for 100-hpf arbors expressing SCAT3 that are stable during the 5-min imaging interval. (I–L) Representative IMD images of a 100-hpf RGC arbor expressing SCAT3 that possesses branch points (BP T0; I), which remain stable during the 5-min imaging interval (BP T + 5 min; J). The regions enclosed by the dashed squares in I and J are magnified in K and L. (M and N) Quantification of the Venus/ECFP ratio (M) comparing changes in arbors between stable branch tips and regions that do not form branch tips as in the scheme in H and presented as ratios (N). White arrowheads depict the region in D that will become a branch point in E. White asterisks indicate the parent axon. 37 branch point additions and nonbranch points in six arbors were analyzed for 80-hpf arbors. 30 stable branch points from seven arbors were analyzed for 100-hpf arbors. Dorsal views are shown, and anterior is up. Error bars represent SEMs. ***, P < 0.001. Bars, 10 µm.

Figure 4.

Capase-3, Caspase-9, and p38 MAPK promote RGC axon arbor dynamics. (A, C, E, G, and I) Tracings of arbors at 80–85 hpf expressing YFP-Rab3. Images represent WT (A), coMO (C), casp-3aMO1 (E), casp-9MO1 (G), and p38mapkMO1 (I) morphant arbors every hour during a 5-h time-lapse imaging. All show dorsal views. Black, stable branch tips; red, eliminated tips; green, added tips; blue, tips that were added then eliminated during imaging. (B, D, F, H, and J). The proportion of branch tips in each category. These graphs are derived from analysis of a single arbor in a single experiment. (K and L) Branch tip additions (K) and retractions (L) are shown in the graphs. (N and O) Presynaptic puncta gain (N) and loss (O) are presented as a percentage of initial branch tip or presynaptic puncta, respectively. (M and P) Total branch tip (M) and presynaptic puncta lifetimes (P) include all branch tips or presynaptic puncta present at one or more time points. (Q and R) Total branch tip (Q) and presynaptic puncta lifetime distribution (R) are shown. 9–13 RGC arbors were analyzed per condition. Tracings are dorsal views; anterior is up. Error bars represent SEMs. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bar, 10 µm.

Figure 5.

Caspase activation is reduced in RGC arbors with impaired Slit1a-Robo2 signaling. (A–F) Representative IMD ratio images of 80-hpf live RGC arbors in the optic tectum expressing SCAT3. Caspase activation is represented by the pseudocolors that correspond to the Venus/ECFP ratio (1.5–0.5). Red represents low caspase activation, and violet represents high caspase activation. WT (A and B), ast (C and D), and slit1a transMO1 (E and F) are shown. The regions enclosed by the dashed squares in A, C, and E are magnified in B, D, and F. (G and H) Quantification of Venus/ECFP ratios (G) at branch points (BP) and surrounding regions (S) presented as a ratio of branch points/surrounding regions (H). White asterisks indicate the parent axon. 15–35 branch points and 45–105 surrounding areas were analyzed per condition. Dorsal views are shown, and anterior is up. Error bars represent SEMs. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bars, 10 µm. The image in A of a SCAT3-expressing arbor is shown again in Figs. 1 E and S2 C. The magnified image in B also is shown again in Fig. 1 H.

Figure 6.

Caspases-3 and -9 interact with Slit1a-Robo2 signaling to restrict arbor size. (A–L) Maximum intensity projections of 80-hpf RGC arbors in the optic tectum expressing YFP-Rab3. WT (A), coMO (B), casp-3aMO1 (C), casp-9MO1 (D), slit1a transMO1 (low dose; E), ast/+ (F), robo2SBMO1 (low dose; G), slit1a transMO1 (low dose); casp-3aMO1 (H), slit1a transMO1 (low dose; I), ast (J), ast/+; casp-3aMO1 (K), and robo2SBMO1 (low dose); casp-9MO1 (L) are shown. (M–R) Quantification of arbor parameters: branch tip number (M), arbor area (N), total arbor length (O), arbor order (P), presynaptic puncta number (Q), and puncta density (R). 7–15 arbors were analyzed per condition. Part of the ast data is obtained from Campbell et al. (2007). Dorsal views are shown, and anterior is up. White asterisks indicate the parent axon. Error bars represent SEMs. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bar, 10 µm.

Figure 7.

p38 MAPK interacts with Slit1a-Robo2 signaling to restrict arbor and presynaptic growth. (A–F) Maximum intensity projections of 80-hpf RGC arbors in the optic tectum expressing YFP-Rab3. coMO (A), p38mapkMO1 (B), slit1a transMO1 (low dose; C), robo2SBMO1 (low dose; D), p38mapkMO1; slit1a transMO1 (low dose; E), and p38mapkMO1; robo2SBMO1 (low dose; F). (G–L) Quantification of arbor parameters: branch tip number (G), arbor area (H), total arbor length (I), arbor order (J), presynaptic puncta number (K), and puncta density (L). 6–20 arbors were analyzed per condition. Dorsal views are shown, and anterior is up. White asterisks indicate the parent axon. Error bars represent SEMs. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bar, 10 µm.

Figure 8.

Caspase-3 interacts with Slit1a-Robo2 signaling to promote arbor dynamics. (A, C, E, G, and I) Tracings of arbors at 80–85 hpf expressing YFP-Rab3 for slit1a transMO1 (low dose; A), robo2SBMO (low dose; C), casp-3aMO1 (low dose; E), slit1a transMO1 (low dose); casp-3aMO1 (lose dose; G), and robo2SBMO (low dose); casp-3aMO1 (low dose; I) morphant arbors every hour during a 5-h time-lapse imaging. Dorsal views are shown. Black, stable branch tips; red, eliminated tips; green, added tips; blue, tips that were added then eliminated during imaging. (B, D, F, H, and J) The proportion of branch tips in each category. These graphs are derived from analysis of a single arbor in a single experiment. (K, L, O, and P) Branch tip additions (K), retractions (L), presynaptic puncta gain (O), and puncta loss (P) are presented as percentages of initial branch tip or puncta number. (M and N) Total branch tip lifetime (M) and the percentage of branch tips present throughout the imaging period (N). (Q and R) Total puncta lifetime (Q) and the percentage of puncta present throughout the imaging period (R). A minimum of five RGC arbors were analyzed per condition. Dorsal views are shown, and anterior is up. Error bars represent SEMs. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bar, 10 µm.