Glial Draper signaling triggers cross-neuron plasticity in bystander neurons after neuronal cell death

Abstract

Neuronal cell death and subsequent brain dysfunction are hallmarks of aging and neurodegeneration, but how the nearby healthy neurons (bystanders) respond to the cell death of their neighbors is not fully understood. In the Drosophila larval neuromuscular system, bystander motor neurons can structurally and functionally compensate for the loss of their neighbors by increasing their axon terminal size and activity. We termed this compensation as cross-neuron plasticity, and in this study, we demonstrated that the Drosophila engulfment receptor, Draper, and the associated kinase, Shark, are required in glial cells. Surprisingly, overexpression of the Draper-I isoform boosts cross-neuron plasticity, implying that the strength of plasticity correlates with Draper signaling. Synaptic plasticity normally declines as animals age, but in our system, functional cross-neuron plasticity can be induced at different time points, whereas structural cross-neuron plasticity can only be induced at early stages. Our work uncovers a novel role for glial Draper signaling in cross-neuron plasticity that may enhance nervous system function during neurodegeneration and provides insights into how healthy bystander neurons respond to the loss of their neighboring neurons.

Figure 1.. Draper is required for debris clearance after Is MN ablation.

A-D and A’-D’. Axon bundles in third instar (A) no ablation control (Is>GFP), (B) Is ablated (Is>GFP,hid,rpr), (C) draper mutant with no ablation (drpr^Δ5, Is>GFP) and (D) draper mutant with Is ablated (drpr^Δ5, Is>GFP,hid,rpr) larvae, labeled with GFP (green), Repo (glial cell marker, magenta) and HRP (neuronal marker, blue). Grey dash lines indicate the position of cross sections (A’-D’). Significant GFP positive debris accumulated in glial cells when ablating Is MNs in a draper mutant background (D and D’). E. Quantification of the number of GFP+ glial cells per animal. F(3,57)=14.09, p<0.0001, one-way ANOVA. N (larvae) =15, 14, 16, 16. F-I. VNCs of first instar larvae of displayed genotypes labeled with HRP (left panel) and GFP (right panel). Note the significant amount of GFP+ signal remaining in (I) after removing draper. J. Quantification of GFP intensity in VNC. t(12)=7.703, p<0.0001, unpaired t-test. N (VNCs) =7, 7. Error bars indicate ± SEM, ****p<0.0001.

Figure 2.. Draper is required for cross-neuron plasticity.

A-D. NMJs of MN4-Ib in third instar (A) no ablation control (Is>GFP), (B) Is ablated (Is>GFP,hid,rpr), (C) draper mutant with no ablation (drpr^Δ5, Is>GFP) and (D) draper mutant with Is ablated (drpr^Δ5, Is>GFP,hid,rpr) larvae, labeled with HRP (grey). The NMJ was expanded in control Is ablated larvae due to cross-neuron plasticity (B), and this expansion is blocked in a draper mutant background (D). E. Quantification of MN4-Ib bouton numbers in no ablation and Is ablated larvae in control and drpr^Δ5 backgrounds. Control (n = 66 and 69 NMJs), t(133)=5.030, p<0.0001, unpaired t-test. drpr^Δ5 (N = 53 and 56 NMJs), t(107)=1.838, p=0.0688, unpaired t-test. F. EPSP and mEPSP traces from no ablation and Is ablated larvae in control and drpr^Δ5 backgrounds. G. Quantification of EPSP amplitude of no ablation and Is ablated larvae in control and drpr^Δ5 backgrounds. Control, t(41)=4.924, p<0.0001, unpaired t-test. drpr^Δ5, t(48.04)=7.011, p<0.0001, unpaired t-test with Welch’s correction. Is ablated control vs Is ablated in drpr^Δ5, t(43.54)=4.075, p=0.0002, unpaired t-test with Welch’s correction. H. Quantification of quantal content of no ablation and Is ablated larvae in control and drpr^Δ5 backgrounds. Control, t(41)=2.224, p=0.0317, unpaired t-test. drpr^Δ5, t(58)=6.194, p<0.0001, unpaired t-test. Is ablated control vs Is ablated in drpr^Δ5, t(46)=5.304, p<0.0001, unpaired t-test. I. Quantification of normalized EPSP of Is ablated larvae in control and drpr^Δ5 backgrounds. Is ablated control vs drpr^Δ5, t(43.20)=3.753, p=0.0005, unpaired t-test with Welch’s correction. Is ablated control vs Ib/Ib+Is, t(29)=6.506, p<0.0001, unpaired t-test. Is ablated in drpr^Δ5 vs Ib/Ib+Is, t(37.82)=1.524, p=0.1357, unpaired t-test with Welch’s correction. J. Quantification of normalized quantal content of Is ablated larvae in control and drpr^Δ5 backgrounds. t(46)=3.730, p=0.0005, unpaired t-test. For G-I, N (NMJs) = 24, 19, 31, 29. Error bars indicate ± SEM, ns = non-significant, *p<0.05, ***p<0.001, ****p<0.0001.

Figure 3.. Draper and Shark are required in glial cells for cross-neuron plasticity.

A-H. NMJs of MN4-Ib in third instar no ablation and Is ablated larvae in control, glia draper knockdown, muscle draper knockdown, and double knockdown backgrounds, labeled with GFP (green) and HRP (magenta). NMJ expansion was observed upon Is MN ablation (B), and this expansion is absent in glia draper knockdown (D) or double knockdown (H) backgrounds. I. Quantification of MN4-Ib bouton number between no ablation and Is ablated larvae in control, glia draper knockdown, muscle draper knockdown, and double knockdown backgrounds. Control (N = 20 and 21 NMJs), t(39)=2.822, p=0.0075, unpaired t-test. Glia draper knockdown (N = 20 and 19 NMJs), t(37)=0.7525, p=0.4565, unpaired t-test. Muscle draper knockdown (N = 16 and 15 NMJs), t(29)=2.204, p=0.0356, unpaired t-test. Double knockdown (N = 21 and 16 NMJs), t(35)=0.2965, p=0.7686, unpaired t-test. J. Quantification of normalized EPSP of Is ablated larvae in control, glia draper knockdown, muscle draper knockdown, and double knockdown backgrounds. F(3, 85)=9.191, p<0.0001, one-way ANOVA. Is ablated control vs glia draper knockdown, p=0.0001. Is ablated control vs muscle draper knockdown, p=0.0042. Is ablated control vs double knockdown, p=0.0006. Is ablated control vs Ib/Ib+Is, t(37)=5.462, p<0.0001, unpaired t-test. Is ablated in glia draper knockdown vs Ib/Ib+Is, t(30)=1.483, p=0.1486, unpaired t-test. Is ablated in muslce draperknock down vs Ib/Ib+Is, t(38)=3.178, p=0.0029, unpaired t-test. Is ablated in double knockdown vs Ib/Ib+Is, t(24)=1.540, p=0.1367, unpaired t-test. K. Quantification of normalized quantal content of Is ablated larvae in control, glia draper knockdown, muscle draper knockdown, and double knockdown backgrounds. F(3, 85)=8.263, p<0.0001, one-way ANOVA. Is ablated control vs glia draper knockdown, p=0.0028. Is ablated control vs muscle draperknock down, p=0.9913. Is ablated control vs double knockdown, p=0.0052. For J and K, N (NMJs) = 27, 20, 28, 14. L. Quantification of MN4-Ib bouton number in no ablation and Is ablated larvae in control, glial shark knockdown, muscle shark knockdown, and double knockdown backgrounds. Control (N = 22 and 23 NMJs), t(43)=3.598, p=0.0008, unpaired t-test. Glial shark knockdown (N = 21 and 22 NMJs), t(41)=1.566, p=0.1250, unpaired t-test. Muscle shark knockdown (N = 23 and 19 NMJs), t(40)=3.220, p=0.0025, unpaired t-test. Double shark knockdown (N = 16 and 22 NMJs), t(36)=0.3390, p=0.7366, unpaired t-test. M. Quantification of normalized EPSP of Is ablated larvae in control, glia shark knockdown, muscle shark knockdown, and double knockdown backgrounds. F(3, 71)=5.533, p=0.0018, one-way ANOVA. Is ablated control vs glia shark knockdown, p=0.0062. Is ablated control vs muscle shark knockdown, p=0.0105. Is ablated control vs double knockdown, p=0.0093. Is ablated control vs Ib/Ib+Is, t(28.04)=4.485, p=0.0001, unpaired t-test with Welch’s correction. Is ablated in glia shark knockdown vs Ib/Ib+Is, t(30)=2.798, p=0.0089, unpaired t-test. Is ablated in muscle shark knockdown vs Ib/Ib+Is, t(30)=2.329, p=0.0268, unpaired t-test. Is ablated in double knockdown vs Ib/Ib+Is, t(25)=2.254, p=0.0332, unpaired t-test. N. Quantification of normalized quantal content of Is ablated larvae in control, glia shark knockdown, muscle shark knockdown, and double knockdown backgrounds. F(3, 71)=4.437, p=0.0065, one-way ANOVA. Is ablated control vs glia shark knockdown, p=0.0466. Is ablated control vs muscle shark knockdown, p=0.9470. Is ablated control vs double knockdown, p=0.0214. For M and N, N (NMJs) = 20, 20, 20, 15. Error bars indicate ± SEM, ns = non-significant, *p<0.05, **p<0.01, ***p<0.001.

Figure 4.. Overexpression of Draper-I boosts cross-neuron plasticity.

A. Quantification of MN6-Ib bouton number in no ablation and Is ablated larvae in control, glia draper-I overexpression, and muscle draper-I overexpression backgrounds. Control (N = 19 and 23 NMJs), t(40)=3.493, p=0.0012, unpaired t-test. Glia draper-I overexpression (N = 20 and 18 NMJs), t(36)=4.103, p=0.0002, unpaired t-test. Muscle draper-I overexpression (N = 22 and 16 NMJs), t(36)=2.818, p=0.0078, unpaired t-test. B. Quantification of normalized EPSP of Is ablated larvae in control, glia draper-I overexpression, and muscle draper-I overexpression backgrounds. F(2, 34)=4.361, p=0.0206, one-way ANOVA. Is ablated control vs glia draper-I overexpression, p=0.2235. Is ablated control vs muscle draper-I overexpression, p=0.0153. Is ablated control vs Ib/Ib+Is, t(19)=2.074, p=0.0519, unpaired t-test. Is ablated in glia draper-I overexpression vs Ib/Ib+Is, t(22)=4.041, p=0.0005, unpaired t-test. Is ablated in muslce draper-I overexpression vs Ib/Ib+Is, t(20)=5.061, p<0.0001, unpaired t-test. C. Quantification of normalized quantal content of Is ablated larvae in control, glia draper-I overexpression, and muscle draper-I overexpression backgrounds. F(2, 34)=12.27, p<0.0001, one-way ANOVA. Is ablated control vs glia draper-I overexpression, p=0.2951. Is ablated control vs muscle draper-I overexpression, p<0.0001. For B and C, N (NMJs) = 11, 14, 12. D. Quantification of MN6-Ib bouton number in no ablation and Is ablated larvae in control, glia draper-II overexpression, and muscle draper-II overexpression backgrounds. Control (N = 19 and 19 NMJs), t(36)=4.319, p=0.0001, unpaired t-test. Glia draper-II overexpression (N = 23 and 17 NMJs), t(38)=1.626, p=0.1122, unpaired t-test. Muscle draper-II overexpression (N = 16 and 20 NMJs), t(34)=0.1763, p=0.8611, unpaired t-test. E. Quantification of normalized EPSP of Is ablated larvae in control, glia draper-II overexpression, and muscle draper-II overexpression backgrounds. F(2, 34)=2.731, p=0.0795, one-way ANOVA. Is ablated control vs glia draper-II overexpression, p=0.0680. Is ablated control vs muscle draper-II overexpression, p=0.7116. Is ablated control vs Ib/Ib+Is, t(21)=2.606, p=0.0165, unpaired t-test. Is ablated in glia draper-II overexpression vs Ib/Ib+Is, t(21)=1.008, p=0.3250, unpaired t-test. Is ablated in muscle draper-II overexpression vs Ib/Ib+Is, t(19)=2.162, p=0.0436, unpaired t-test. F. Quantification of normalized quantal content of Is ablated larvae in control, glia draper-II overexpression, and muscle draper-II overexpression backgrounds. F(2, 34)=8.539, p=0.0010, one-way ANOVA. Is ablated control vs glia draper-II overexpression, p=0.0130. Is ablated control vs muscle draper-II overexpression, p=0.5614. For E and F, N (NMJs) = 13, 13, 11. G. Quantification of MN6-Ib bouton number in no ablation and Is ablated larvae in control, glia draper-III overexpression, and muscle draper-III overexpression backgrounds. Control (N = 20 and 16 NMJs), t(23.64)=2.129, p=0.0439, unpaired t-test with Welch’s correction.. Glia draper-III overexpression (N = 16 and 22 NMJs), t(36)=3.261, p=0.0024, unpaired t-test. Muscle draper-III overexpression (N = 21 and 14 NMJs), t(33)=2.852, p=0.0074, unpaired t-test. H. Quantification of normalized EPSP of Is ablated larvae in control, glia draper-III overexpression, and muscle draper-III overexpression backgrounds. F(2, 38)=0.2634, p= 0.7698, one-way ANOVA. Is ablated control vs glia draper-III overexpression, p=0.7729. Is ablated control vs muscle draper-III overexpression, p=0.8508. Is ablated control vs Ib/Ib+Is, t(22)=2.291, p=0.0319, unpaired t-test. Is ablated in glia draper-III overexpression vs Ib/Ib+Is, t(22)=2.299, p=0.0314, unpaired t-test. Is ablated in muslce draper-III overexpression vs Ib/Ib+Is, t(21)=2.214, p=0.0380, unpaired t-test. I. Quantification of normalized quantal content of Is ablated larvae in control, glia draper-III overexpression, and muscle draper-III overexpression backgrounds. F(2, 38)=4.981, p=0.0120, one-way ANOVA. Is ablated control vs glia draper-III overexpression, p=0.4070. Is ablated control vs muscle draper-III overexpression, p=0.0089. For H and I, N (NMJs) = 14, 14, 13. Error bars indicate ± SEM, ns = non-significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 5.. MN11-Ib showed cross-neuron plasticity upon Is ablation.

A and B. NMJs of MN11-Ib in third instar control (Is>GFP) and Is ablated (Is>GFP,hid,rpr) larvae labeled with GFP (green) and HRP (magenta). Note the larger NMJs in Is ablated larvae. C. Quantification of MN11-Ib bouton number in no ablation (N = 31 NMJs) and Is ablated (N = 32 NMJs) larvae. t(61)=3.408, p=0.0012, unpaired t-test. D. EPSP and mEPSP recordings of muscle 11 from no ablation and Is ablated larvae. E. Quantification of mEPSP frequency. T(21)=0.5408, p=0.5943, unpaired t-test. F. Quantification of mEPSP amplitude. T(21)=0.4446, p=0.6611, unpaired t-test. G. Quantification of EPSP amplitude. T(18.02)=3.840, p=0.0012, unpaired t-test with Welch’s correction. H. Quantification of quantal content. T(21)=3.657, p=0.0015, unpaired t-test. For E-H, N (NMJs) = 10, 13. Error bars indicate ± SEM, ns = non-significant, **p<0.01.

Figure 6.. Acute Is MN ablation induces functional plasticity, but not structural plasticity.

A. Schematic of heat-shock induced Is MN ablation. B-F and B’-F’. NMJs of MN4-Ib in late third instar larvae (hs-FLP,UAS-GFP/+;UAS-FRT stop FRT-hid-2A-rpr/+;Is-GAL4/+) with (B, B’) no heat-shock, (C, C’) embryo heat-shock, (D, D’) first instar heat-shock, (E, E’) second instar heat-shock and (F, F’) third instar heat-shock, stained with GFP (green), HRP (magenta), and DLG (grey). In embryos and first and second instar heat-shocked larvae, the Is NMJs were fully cleared, while Is synaptic debris remains in third instar heat-shocked larvae (F and F’). G. Quantification of MN4-Ib bouton number in late third instar larvae with Is MNs ablated at different developmental stages. F(4, 127)=10.23, p<0.0001, one-way ANOVA. Control vs embryo heat-shock, p<0.0001. Control vs first instar heat shock, p=0.3021. Control vs second instar heat-shock, p=0.7310. Control vs third instar heat-shock, p=0.8756. N (NMJs) = 33, 36, 13, 25, 25. H. Comparison of the normalized EPSP of late third instar larvae with Is MNs ablated at different developmental stages to Ib/Ib+Is baseline. Embryo heat-shock, t(27)=7.126, p<0.0001, unpaired t-test. First instar heat-shock, t(21)=6.622, p<0.0001, unpaired t-test. Second instar heat-shock, t(22)=9.485, p<0.0001, unpaired t-test. Third instar heat-shock, t(23)=8.557, p<0.0001, unpaired t-test. I. Quantification of normalized quantal content of late third instar larvae with Is ablated at different developmental stages. F(4, 67)=9.109, p<0.0001, one-way ANOVA. Embryo heat-shock vs third instar heat-shock, p=0.0002. Non-significant for the others. This result suggested an increase of cross-neuron plasticity when acutely ablated Is MNs. For H and I, N (NMJs) = 19, 17, 11, 12, 13. Error bars indicate ± SEM, ns = non-significant, ***p<0.001, ****p<0.0001.

Figure 7.. Cross-neuron plasticity bolsters larval locomotion speed.

A. Representative crawling traces of control and Is ablated larvae. Note the control larvae perform more turns than Is ablated larvae. B. Quantification of crawling speed of control and Is ablated larvae. t(84)=3.107, p=0.0026. C. Quantification of turn frequency of control and Is ablated larvae. t(72.67)=4.518, p<0.0001. For C and D, N (larvae) = 45, 41. D. Schematic of heat induced roll behavior difference in control and Is ablated larvae. E. Quantification of the number of rolls of control and Is ablated larvae. t(16.59)=3.647, p=0.0021. N (larvae) = 12, 12. Error bars indicate ± SEM, **p<0.01, ****p<0.0001.