Modeling glial contributions to seizures and epileptogenesis: cation-chloride cotransporters in Drosophila melanogaster

Abstract

Flies carrying a kcc loss-of-function mutation are more seizure-susceptible than wild-type flies. The kcc gene is the highly conserved Drosophila melanogaster ortholog of K+/Cl- cotransporter genes thought to be expressed in all animal cell types. Here, we examined the spatial and temporal requirements for kcc loss-of-function to modify seizure-susceptibility in flies. Targeted RNA interference (RNAi) of kcc in various sets of neurons was sufficient to induce severe seizure-sensitivity. Interestingly, kcc RNAi in glia was particularly effective in causing seizure-sensitivity. Knockdown of kcc in glia or neurons during development caused a reduction in seizure induction threshold, cell swelling, and brain volume increase in 24-48 hour old adult flies. Third instar larval peripheral nerves were enlarged when kcc RNAi was expressed in neurons or glia. Results suggest that a threshold of K+/Cl- cotransport dysfunction in the nervous system during development is an important determinant of seizure-susceptibility in Drosophila. The findings presented are the first attributing a causative role for glial cation-chloride cotransporters in seizures and epileptogenesis. The importance of elucidating glial cell contributions to seizure disorders and the utility of Drosophila models is discussed.

Figure 1. Immunohistochemistry with confocal fluorescence microscopy reveals neuronal and glial Kcc in larvae and adults.

(A) Larval nervous system Kcc in a 0.75 µm confocal slice is seen surrounding the ventral nerve cord (VNC) and peripheral nerves (PNs), and colocalizing with neuronal somata on the periphery of the VNC. CNS: central nervous system. High-magnification PN in a 2 µm slice reveals membrane-bound GFP of neuronal processes (B1) and Kcc (B2) surrounding and colocalizing with the processes (B3). Adult posterior brain Kcc in a 0.5 µm slice (C3) depicts the general localization pattern observed in several regions; Kcc is widespread and salient around cell somata (C4) and on brain surface/edges (C5). Neuronal GFP in mushroom body somata and calyx membranes (C2) colocalizes with Kcc (C1), confirming that Kcc is in central brain neuropile and ubiquitous. Cyx: mushroom body calyx. Optic lobe Kcc in 0.5 µm slice (D3) conformed to the central brain pattern of neuronal membrane (D2) colocalization (D1). Optic medulla NC82 (D4), a neuropile marker, colocalized with Kcc (D5). Posterior brain Kcc in a 0.5 µm slice (E3) in animals expressing glial membrane-bound GFP (E2) colocalizes with GFP significantly (E1), probably in cortex (arrow heads) and surface (arrows) glia. Neuronal membrane (E4) colocalization with Kcc (E5) in the same slice is shown for comparison. High magnification micrographs of the optic lamina showed that Kcc in a 0.5 µm slice (F3) is present in glial (F2) and neuronal (F4) membranes of this brain structure (F1), (F5). Scale bars are in microns. For orientation, A: anterior, D: dorsal, L: lateral, M: medial, P: posterior.

Figure 2. Reducing kcc expression by RNAi causes behavioral and electrophysiologically-recorded seizure-like activity.

(A) Quantification of behavioral seizure-sensitivity for select GAL4/UAS genotypes, expressed as %bang-sensitive (%BS) paralysis on the y-axis, using two different UAS-kcc-RNAi transgenes. The GAL4 driver and expression domain for each genotype is shown on the x-axis. kcc-RNAi-B (black bars/text) is more effective than kcc-RNAi-V (grey bars/text) with respect to causing %BS paralysis and lethality phenotypes. (B) in vivo stimulation and recording from the giant fiber circuit of a fly mounted in dental wax for quantifying thresholds to evoked seizure-like activity. (C) High-frequency stimulus (200 Hz for 300 ms) seizure-like activity voltage thresholds, in volts high-frequency stimulus (V HFS), for select test and control genotypes. N-values are as noted for each genotype. Error bars are S.E.M. and significance for Student's t-tests is: ***  = p<0.001; n.s.  =  not significant. RNAi expression caused reduced V HFS thresholds relative to controls, thus indicating increased seizure-sensitivity. (D) Representative seizure-like discharges recorded in dorsal longitudinal muscles from flies of select genotypes, as indicated. Green insert is an enlargement of the region enclosed by green lines illustrating a muscle response following a giant fiber threshold stimulus pulse (∼2 V, 0.3 ms pulse-width). Red insert is an enlargement of the region enclosed by the first pair of red lines illustrating a failure following a single giant fiber threshold stimulus pulse. Remaining pairs of red lines indicated failures following seizures in other genotypes.

Figure 3. kcc knockdown leads to brain volume increases in 24–48 h-old adults.

Shown are representative brain enlargements via kcc RNAi in 0.5 µm confocal slices, (A2), (B2), (C2), with respect to controls, (A1), (B1), (C1), for the repo, A307, and OK107 GAL4 drivers. *: optic lobes of test or control brains were often severed to distinguish genotypes stained in same solutions. (D) Quantification of mean %volume increases for test genotype brains compared to their respective controls. Expressing kcc RNAi in glia or neurons causes whole-brain swelling. High magnification of glia with membrane-bound GFP in the wild-type mushroom body region (E2) wraps and interlaces neuronal somata and the calyx neuropile (E4). Dorsal Kcc in this same region (E3) colocalizes with cortex and surface glia (E1) and with glomeruli of the calyx (E5). Glia expressing membrane-bound GFP and kcc-RNAi-B in similar mushroom body regions (F2) are largely absent; defined cortex glia, glia wrapping the calyx, and stereotyped surface glia are few, and thus, Kcc colocalization is reduced (F1). Kcc in this region (F3) mostly colocalizes with the neuronal calyx (F4) neuropile (F5). Neuronal expression of membrane-bound GFP in the mushroom body (G1) confirms Kcc localization (G2) in the somata and calyx neuropile (G3). (H1) kcc RNAi and membrane-bound GFP expression in the mushroom body caused a significant reduction of Kcc (H2) in the calyx (H3). Moreover, brain surface Kcc is further from the calyx in this loss-of-function genotype, as seen in other genotypes with neuronal kcc RNAi expression (data not shown). (I) Quantification of Kcc knockdown in the mushroom body calyx due to kcc RNAi expression. Significance for Student's t-tests is: **  = p<0.01; ***  = p<0.001. Scale bars are in microns.

Figure 4. Giant fiber and lateral pace-making neurons deficient in Kcc are enlarged.

(A1) A 34 µm Z-stack projection in animals expressing membrane-bound GFP in the giant fiber system depicts the stereotyped morphology of the wild-type giant fiber soma, axon and dendrites (insets). (A2) Animals expressing GFP and kcc-RNAi-V in the giant fiber system have giant fiber neurons with enlarged somata and diffuse dendrites lacking defined spines (insets), as seen in this representative 34 µm Z-stack projection. (B1) Wild-type lateral pace-making neurons expressing membrane-bound GFP are nearly spherical, as shown in this representative 10 µm Z-stack projection. (B2) Lateral pace-making neurons expressing GFP and kcc-RNAi-B are enlarged, as seen in this larger (15 µm) representative encompassing Z-stack projection. (C) Quantification of soma volume for control and kcc RNAi expressing lateral pace-making neurons. Error bars are S.E.M. and significance for Student's t-test is: ***  = p<0.001. Scale bars are in microns. For orientation, A: anterior, D: dorsal, L: lateral, M: medial.

Figure 5. kcc knockdown causes swelling of third instar larval peripheral nerves.

(A1) Membrane-bound glial GFP illuminating peripheral nerves at a low magnification 40 µm slice. (A2) Peripheral nerve glia GFP in a high magnification 2 µm slice colocalizes with the pervasive Kcc (A3) that also marks the bundled neuronal processes (A4). (A5) Orthogonal view of A4 at indicated position (A4, arrowhead). (B1) Glial GFP of swollen peripheral nerves in a low magnification 40 µm slice of animals also expressing glial kcc-RNAi-B. High magnification 2 µm slice of nerve glial GFP (B2) and Kcc (B3) showing whole nerve enlargement (B4). (B5) Orthogonal view of B4 at indicated position (B4, arowhead). (C1) Glial GFP of peripheral nerve in low magnification 40 µm slice of animals also expressing glial ncc69-RNAi-V. High magnification 2 µm slice of nerve glial GFP (C2) and Kcc (C3) showing swelling and fraying in a peripheral nerve bulge (C4), as shown previously . (C5) Orthogonal view of C4 at indicated position (C4, arrowhead). (D1) Neuronal GFP of peripheral nerves in low magnification 40 µm slice. (D2) Wild-type neuronal processes in animals expressing neuronal membrane-bound GFP are tightly bundled within peripheral nerve with pervasive Kcc (D3). (D4) Kcc is within and surrounding nerve GFP+ neuronal processes of wild-type larvae. (D5) Orthogonal view of D4 at indicated position (D4, arrowhead). (E1) Neuronal GFP of peripheral nerve in low magnification 40 µm slice of animals also expressing neuronal kcc-RNAi-V. Neuronal processes within peripheral nerve (E2) and Kcc mostly on the surface (E3) do not extensively overlap (E4). (E5) Orthogonal view of E4 at indicated position (E4, arrowhead). (F) and (G) Quantification of average cross-sectional areas of peripheral nerves in control and RNAi genotypes. Error bars are S.E.M. and significance for Student's t-tests is: **  = p<0.01. White numbers indicate abdominal nerve number. Scale bars are in microns.