Huntingtin's function in axonal transport is conserved in Drosophila melanogaster

Abstract

Huntington's disease (HD) is a devastating dominantly inherited neurodegenerative disorder caused by an abnormal polyglutamine expansion in the N-terminal part of the huntingtin (HTT) protein. HTT is a large scaffold protein that interacts with more than a hundred proteins and is probably involved in several cellular functions. The mutation is dominant, and is thought to confer new and toxic functions to the protein. However, there is emerging evidence that the mutation also alters HTT's normal functions. Therefore, HD models need to recapitulate this duality if they are to be relevant. Drosophila melanogaster is a useful in vivo model, widely used to study HD through the overexpression of full-length or N-terminal fragments of mutant human HTT. However, it is unclear whether Drosophila huntingtin (DmHTT) shares functions similar to the mammalian HTT. Here, we used various complementary approaches to analyze the function of DmHTT in fast axonal transport. We show that DmHTT interacts with the molecular motor dynein, associates with vesicles and co-sediments with microtubules. DmHTT co-localizes with Brain-derived neurotrophic factor (BDNF)-containing vesicles in rat cortical neurons and partially replaces mammalian HTT in a fast axonal transport assay. DmHTT-KO flies show a reduced fast axonal transport of synaptotagmin vesicles in motoneurons in vivo. These results suggest that the function of HTT in axonal transport is conserved between flies and mammals. Our study therefore validates Drosophila melanogaster as a model to study HTT function, and its dysfunction associated with HD.

Figure 1. The dynein interacting domain of human HTT is conserved in Drosophila melanogaster.

(A) Alignment of hHTT[600–698] and DmHTT[381–549]. The red box highlights the high homology region. (B) Western blot analyses of the co-immunoprecipitation of Dm620HTT-GFP with endogenous DIC on transfected HEK 293 cells: Dm620HTT-GFP but not control GFP co-precipitates DIC using an anti-GFP antibody. Anti-HA antibody was used as negative control and showed no Dm620HTT-GFP or DIC signal.

Figure 2. Drosophila Huntingtin N-terminal fragment associates to small vesicles.

HEK 293 were transfected with GFP or Dm620HTT-GFP and cells were subjected to subcellular fractionation (T: total, P3: vesicles, S3: cytosol, P2: larges membranes, S2: cytosol plus light membranes, P1: nuclei and S1: post-nuclear supernatant) followed by Western blotting analyses to reveal the distribution of the different proteins (GFP, Dm620HTT-GFP, endogenous HTT, DIC, α-tubulin and β-actin). Immunoblot shows that GFP is mostly in the soluble fraction (S3) whereas Dm620HTT-GFP is enriched in the vesicles fraction (P3).

Figure 3. Drosophila Huntingtin N-terminal fragment colocalizes with neuronal BDNF vesicles.

Rat cortical neurons were electroporated to express BDNF-mCherry and GFP or Dm620HTT-GFP, fixed and immunostained. Fluorescence was analyzed by confocal microscopy. (A) Confocal images showing the localization of both BDNF-mCherry and Dm620HTT-GFP in double-transfected neurons. (B) Resliced z-stack along an axon with channel intensity values (top) and immunostaining (bottom) with GFP channel (green) showing the localization of GFP (left) or Dm620HTT-GFP (right) and mCherry channel (red) showing the localization of BDNF-mCherry-containing vesicles. Whereas GFP intensity is diffuse and does not correlate with BDNF, Dm620HTT-GFP shows a punctate signal that partially co-localizes with BDNF. Stars show when the two peaks of intensity co-distribute in the line-scan and co-localize in the z-resliced images.

Figure 4. Drosophila Huntingtin N-terminal fragment associates with microtubules.

HEK 293 were transfected with GFP or Dm620HTT-GFP and cell extracts were subjected to a microtubule depolymerization – repolymerization assay to pellet microtubules and the microtubule-associated proteins which were analyzed by Western blotting. (A) Whereas GFP does not pellet with the microtubules fraction (MT) but is soluble (S), Dm620HTT-GFP co-sediments with microtubules. (B) MT binding assay in presence of Taxol or Nocodazole. Dm620HTT-GFP is not pelleted in Nocodazole-treated cells when microtubules are depolymerized.

Figure 5. Drosophila Huntingtin is dynamic and rescues loss of mammalian huntingtin in axonal transport.

(A) Dm620HTT-GFP is dynamic. Image within a microchannel, of an axon from a rat cortical neuron expressing Dm620HTT-GFP. The generated kymograph below shows the vesicular trajectories and the paths that were analyzed (green for anterograde moving vesicles, red for retrograde moving vesicles and blue for stationary vesicles). (B) Two-colors kymograph obtained from neurons co-transfected with BDNF-mCherry and Dm620HTT-GFP shows co-transport of DmHTT and BDNF-mCherry in axons. (C) Western blotting shows silencing of endogenous rat Htt gene and its replacement by fly htt in cortical neurons. (D) The generated kymograph shows the vesicular trajectory in the three conditions and the paths that were analyzed. (E) The mean velocity of anterograde and retrograde moving vesicles was calculated. (Mean+S.E.M, p* <0.05, p**<0.01 and p***<0.001).

Figure 6. Fast axonal transport is impaired in Drosophila larvae knock out for HTT.

(A) Anti DmHTT-S50 antibody recognizes Dm620HTT-GFP but not GFP when expressed in HEK 293 cells. (B) DmHTT-S50 recognizes a band of about 400 kDa in Drosophila L3 larvae brain extracts (see star). This band is absent in HTT-KO fly extracts. (C) Typical kymographs of wild-type and HTT-KO larvae with the analyzed trajectories for synaptotagmin-GFP vesicles. (D) Quantification of the mean velocity of axonal synaptotagmin-GFP vesicles in wild-type and HTT-KO larvae. (E) Distribution of synaptotagmin-GFP vesicles velocities shows a reduced number of fast-moving vesicles in HTT-KO larvae compared with wild-type flies.