Axonal injury and regeneration in the adult brain of Drosophila

Abstract

Drosophila melanogaster is a leading genetic model system in nervous system development and disease research. Using the power of fly genetics in traumatic axonal injury research will significantly speed up the characterization of molecular processes that control axonal regeneration in the CNS. We developed a versatile and physiologically robust preparation for the long-term culture of the whole Drosophila brain. We use this method to develop a novel Drosophila model for CNS axonal injury and regeneration. We first show that, similar to mammalian CNS axons, injured adult wild-type fly CNS axons fail to regenerate, whereas adult-specific enhancement of protein kinase A activity increases the regenerative capacity of lesioned neurons. Combined, these observations suggest conservation of neuronal regeneration mechanisms after injury. We next exploit this model to explore pathways that induce robust regeneration and find that adult-specific activation of c-Jun N-terminal protein kinase signaling is sufficient for de novo CNS axonal regeneration injury, including the growth of new axons past the lesion site and into the normal target area.

Figure 1.

Description of the LNv neurons and Drosophila brain explant culture on organotypic culture plate inserts. A, B, D, and E are projection images of 1 μm confocal stacks through adult Drosophila brains of the indicated genotype. A and B are compiled from separate images of the brain halves of the same brain. A, A single-cell sLNv FLP-out clone shows the cell body (arrow) and dorsally projecting axon (arrowhead) [both marked with GFP (green)] and LNv [marked by lacZ (red)]. Scale bar, 80 μm. B, A single-cell lLNv FLP-out clone shows the cell body (arrow), commissural axon (arrowhead), and optic lobe innervation (double-pointed arrows) [all marked with GFP (green)] and LNv [marked by lacZ (red)]. C, Schematic of an adult Drosophila brain indicating morphology and localization of the sLNv (green) and the lLNv (blue). D, Higher magnification of the GFP pattern in the single sLNv neuron shown in A. Scale bar, 40 μm. E, Higher magnification of the GFP pattern in the single lLNv neuron shown in B. Arrows indicate the cell bodies, and arrowheads indicate the axons. Scale bar, 80 μm. F, Lateral view of a three-dimensional reconstruction of an adult Drosophila brain. The sLNv dorsal projections and the commissural axons (white arrowheads) are on the posterior side of the brain (P). The LNv cell bodies (arrow) and the optic lobe innervation (double pointed arrow) lie anteriorly (A). Genotype of the flies is PDF–Gal4; UAS–CD8–GFP. G, Intact adult Drosophila head. H, I, Dissected adult Drosophila brain seen in bright field (H, I) or GFP filter (H′, I′) of a stereomicroscope (100× magnification). H, H′, Anterior view. Arrows point to the LNv cell bodies; double pointed arrows indicate optic lobe innervation by the lLNv. I, I′, Posterior view. Red arrowheads indicate the sLNv dorsal projections, and white arrowheads indicate the lLNv commissural axons. Genotype of the flies is PDF–Gal4, UAS–CD8–GFP; UAS–CD8–GFP. J, Schematic of the culture setup. Drosophila brains are positioned on the membrane (gray) of a low height organotypic culture plate insert (blue) that is placed in a Petri dish containing culture medium (yellow). The brain is separated from the air by a thin film of medium (inset). K, Overview image of several adult Drosophila brains positioned on the membrane of a culture plate insert. L, Higher-magnification image of one brain with bright field (top) or GFP filter (bottom; pseudocolored), showing the posterior LNv axonal tracts. M, Gross morphology of a cultured brain over the course of 20 d in culture remains intact. A GFP signal can still be detected after 20 d (ol, optic lobe). All images are taken with a stereomicroscope. The blue panels in L and M are a merge between a bright-field image pseudocolored in blue and the GFP signal in green. The genotype of the flies is PDF–Gal4, UAS–CD8–GFP; UAS–CD8–GFP.

Figure 2.

The functional characteristics of the LNvs are maintained in cultured brains. Brains were maintained in a 12 h light/dark cycle. A, Immunostaining for PDF (red) shows the presence of PDH in the sLNv dorsal projection after 7 d in culture. B, Immunolabeling of small (arrow) and large (line) LNv in a 3-d-old cultured brain. Time points ZT10 and ZT22 correspond to 10 and 22 h after lights on. A significantly reduced level of TIM (red) is seen in the brains after 10 h of light (ZT10), whereas cells show higher levels and nuclear localization during the late hours of the night (ZT22) in both the large and small LNv. C, Quantification of TIM levels in brain explants (t test for lLNv ZT10 vs ZT22, t = 1.72, p = 0.024; t test for sLNv ZT10 vs ZT22, t = 1.75, p = 0.001, n = 5 brains for each comparison). The genotype of the fly in A is PDF–Gal4, UAS–CD8–GFP; UAS–CD8–GFP and in B is yw; pdf–Gal4; dORK–NC1/+. D, Spontaneous action potentials of an unidentified surface neuron in a 3-d-old cultured brain. Large-amplitude spontaneously firing action potentials and afterhyperpolarizations are observed in whole-cell patch current-clamp recordings.

Figure 3.

Injury to the sLNv axonal bundle. A, Schematic showing the setup for the injury of sLNv axonal tracts in explanted brains on a culture plate insert. The axonal GFP pattern is observed using a stereomicroscope equipped with a GFP filter (blue ray). The microdissector (gray) tip is positioned on an axonal tract using a micromanipulator. The schematic is not drawn to scale. B, Schematic of sLNv before (top) and after (bottom) injury. cb, Cell body; D, distal part of the axon; P, proximal part of the axon; arrowhead, injury. C, Fluorescent signal in sLNv of the same brain before (top) and after (bottom) injury. Arrowhead indicates the disruption in GFP signal after axotomy. D, Bright-field (left) and GFP (right) images during (top) and after (bottom) the injury procedure. The microdissector tip is indicated by a red outline, and the injury site is indicated by an arrowhead. The genotype of the fly is PDF–Gal4, UAS–CD8–GFP; UAS–CD8–GFP. Images are made with a stereomicroscope at 200× (C) or 100× (D).

Figure 4.

Response of sLNv axons to injury. A–D, At 15 min after the lesion, all axons are clearly severed (B, C, asterisk). The distal axon ends appear uniform and continuous (D). E–H, At 1 d after the lesion, the proximal stumps form bulbous ends (G, arrows) from which axonal spikes can grow (F, arrowhead). The distal axon ends appear more vesicular (H). I–L, At 2 d after the lesion, axonal sprouts are formed (I–K, filled arrowheads). The distal axonal ends appear vesicular (L). M–P, At 3 d after the lesion, axonal sprouts are still present on the proximal stump. In some cases, axons grow away from the target area (M, filled arrowheads), and degenerated fragments of axons can be observed (open arrowheads). In other cases, axons remain stalled before the lesion site (N). The distal axonal fragments appear vesicular (P, open arrowhead). Q–T, At 4 d after the lesion, axonal sprouts of different lengths (filled arrowheads) grow laterally (Q, S) or toward the injury site (Q, filled arrowheads). The distal axons are fragmented (T). * indicates lesion site. The genotype of the flies is PDF–Gal4, UAS–CD8–GFP; UAS–CD8–GFP. All panels are projections of confocal 0.5 μm stacks. Scale bars: overview pictures, 40 μm; proximal stumps, 8 μm; distal stubs, 20 μm.

Figure 5.

Axonal outgrowth and growth cone formation in explanted larval brains. A, Morphology of the third-instar larval CNS (left) (ad, antennal disc; ed, eye disc; ol, future optic lobe; vc, ventral cord; red arrows, nerves). The GFP signal in the brain lobes (right) corresponds to the sLNv neurons. B, Morphological changes in the GFP pattern over time include the appearance of new GFP-positive cell bodies (arrows). C, Detail of the GFP expression pattern after 24 h in culture showing the sLNv cell bodies (arrows) and the dorsally projecting axons. Scale bar, 20 μm. D, Detail of the GFP expression pattern after 5 d in culture showing dorsally projecting axons of the sLNv (red arrowhead). New axonal tracts project toward the contralateral brain half (white arrowheads), the future optic lobes (white arrows), and the central brain (red arrows). Scale bar, 40 μm. E, Detail of the growth cones of the commissure. Note the presence of filopodia at the axonal end (arrowheads). Scale bar, 8 μm. A and B are stereomicroscope images. The blue panel in A is a merge between a bright-field image pseudocolored in blue and the GFP signal in green. C–E are projection images of 0.5 μm confocal stacks. The genotype of the flies is PDF–Gal4, UAS–CD8–GFP; UAS–CD8–GFP.

Figure 6.

Quantification of novel axonal outgrowth after injury. A, Injured axons were imaged ∼6 h after injury to verify the injury and capture the morphology of the cut proximal stump. Scale bar, 40 μm. B, The same brain was fixed and processed for imaging 3 d after injury, and the proximal stump was imaged again at the same magnification as A. Scale bar, 40 μm. C, By comparing the images from the same axon stump ∼6 h and 3 d after the injury, the novel axons were identified, and their point of origin from the cut axonal stump was determined (yellow circles). Based on this determination of novel growth, the percentage of brains showing regrowth was calculated by counting the number of brains that sprouted at least one novel axon from the site of injury. Scale bar, 8 μm. D, To measure the extent of regrowth, the length of a novel axon is traced by free-hand tracing (pink lines), as a criterion for how long the novel axons can grow. E, In addition to the length of the new axons, the distance these novel axons can cover in a straight line is also measured (blue line), as a criterion to measure how far the novel axons reach. F, The size of the gap induced by axonal lesion was measured in micrometers by drawing a line connecting the proximal stump and closest distal stub (n = 18). G, Measurement of the direction of normal target area to a reference line running with the axonal shaft. The results of F and G were calculated as averages (n = 18). H, I, The direction of the novel axons was measured by first drawing a reference line running parallel with the axonal shaft (H) and then measuring the angle of the novel axon to the reference line (I). Scale bars: H, 40 μm; I, 8 μm. J, Graph showing the percentage of brains regrowing at least one axon with different genetic manipulations. The value is calculated as the number of brains regrowing at least one novel axon after injury as a percentage of the total number of brains. *p < 0.05; ***p < 0.001. K, Graph showing the percentage of axonal regrowth. This is calculated as the number of newly grown axons on the total number of cut axons. To calculate the total number of cut axons, the number of examined brains was multiplied by four (the average number of sLNv in one brain hemisphere). *p < 0.05; ***p < 0.001. L, A comparative graph showing the extent of regrowth under different genetic manipulations. The black bars represent the average length of the new axons, and the gray bars represent the average computed distance of the new axons. For statistical analysis, values of the different genotypes were compared with the values for WT brains. *p < 0.05.

Figure 7.

Injury experiments at 18–28°C of PKA alleles with tubPGAL80^ts. A, A′, A″, Injury on control brain. A′, Snapshot taken from a control brain ∼6 h after injury. A″, The same brain was fixed at day 3 and imaged again. A, A high magnification of the proximal stump. Scale bars: A′, A″, 40 μm; A, 8 μm. B, B′, B″, Overexpression of PKAc.A75. B′, Snapshot ∼6 h after injury. B″, The same brain imaged at day 3. B, High magnification of proximal stump. Scale bars: B′, B″, 40 μm; B, 8 μm. C, C′, C″, Overexpression of PKAc^wt has a strong positive effect on the penetrance of regrowth but not the extent of axonal growth. C′, Snapshot of a brain ∼6 h after injury. C″, Snapshot of the same brain 3 d after injury. C″, High magnification of proximal stump. Scale bars: C′, C″, 40 μm; C, 8 μm. D, D′, D″, Overexpression of PKA^mC* induces significantly more regrowth. Also, the length of the newly grown axons is significantly bigger but not the average distance they bridge. D′, Snapshot of a brain ∼6 h after injury. D″, Snapshot of the same brain 3 d after injury. D, High magnification of proximal stump. Scale bars: D′, D″, 40 μm; D, 8 μm. Genotypes are as follows: A, A′, A″ is PDF–Gal4, UAS–CD8–GFP; tubPGAL80^ts/+; B, B′, B″ is PDF–Gal4, UAS–CD8–GFP; tubPGAL80^ts/UAS–PKAc.A75; C, C′, C″ is PDF–Gal4, UAS–CD8–GFP; tubPGAL80^ts/+; UAS–PKAc5.9F/+; and D, D′, D″ is PDF–Gal4, UAS–CD8–GFP; tubPGAL80^ts/UAS–PKAmC*.

Figure 8.

Injury experiments at 18–28°C injury experiments on JNK signaling with tubPGAL80^ts. A, A′, A″, Injury on brains expressing a Bsk^DN allele in the LNv: A′, A″, a snapshot of the brain ∼6 h after injury and 3 d after injury; A, a higher magnification of the injured proximal stump 3 d after injury. Scale bars: A′, A″, 40 μm; A, 8 μm. B, B′, B″, C′, C″, Injury on brains expressing a Hep^CA allele. B′, C′ and B″, C″, Snapshot ∼6 h and 3 d after injury, respectively. B, Higher magnification showing massive growth. Scale bars: B′, B″, 40 μm; B, 8 μm. C, D, Quantifying the extent of growth observed in brains expressing Hep^CA compared with control brains. C, Percentage of brains regrowing at least one novel axon longer than 36 μm (dark blue bar), the percentage of brains growing at least one novel axon toward the target area (light blue bar), and the percentage of brains growing at least one novel axon into the target area (pink bar). D, Schematic drawing of the different quantifications described above in C. All values are significant (*p < 0.05, ***p < 0.001). E, E′, E″, Example of Hep^CA-expressing neurons showing regrowth back into the target area. Scale bars: E′, E″, 40 μm; E, 8 μm. Genotypes are as follows: A, A′, A″ is PDF–Gal4, UAS–CD8–GFP; tubPGAL80^ts/Bsk^DN; B, B′, B″, E, E′, E″ is PDF–Gal4, UAS–CD8–GFP; tubPGAL80^ts/+; UAS–hep^CA.