Huntingtin (htt), the protein mutated in Huntington’s disease, is a positive regulatory factor for vesicular transport whose function is lost in disease. Here, we demonstrate that phosphorylation of htt at serine 421 (S421) restores its function in axonal transport. Using a strategy involving RNA (ribonucleic acid) interference and re-expression of various constructs, we show that polyQ (polyglutamine)-htt is unable to promote transport of brain-derived neurotrophic factor (BDNF)-containing vesicles, but polyQ-htt constitutively phosphorylated at S421 is as effective as the wild-type (wt) as concerns transport of these vesicles. The S421 phosphorylated polyQ-htt displays the wt function of inducing BDNF release. Phosphorylation restores the interaction between htt and the p150Glued subunit of dynactin and their association with microtubules in vitro and in cells. We also show that the IGF-1 (insulin growth factor type I)/Akt pathway by promoting htt phosphorylation compensates for the transport defect. This is the first description of a mechanism that restores the htt function altered in disease.
Huntington’s disease (HD) is a devastating neurodegenerative disorder characterized by the dysfunction and death of striatal and cortical neurons in the brain. The mutation that causes HD is an abnormal polyglutamine (polyQ) expansion in the huntingtin (htt) protein (1). One key early pathogenic event in the disease is the alteration of axonal transport (2–6). Htt is found on microtubules (MTs) and is associated with proteins of the molecular motor machinery including dynein and htt-associated protein-1 (HAP1) that interacts with p150Glued and kinesin-1 (3,7–11). Wild-type (wt) htt promotes the transport of brain-derived neurotrophic factor (BDNF)-containing vesicles (3,12,13). BDNF transport is central to HD because BDNF is actively transported in cortico-striatal projecting neurons and provided to the striatum (14), where it acts as a prosurvival factor for these neurons that are particularly vulnerable in HD (15,16). The abnormal expansion in polyQ-htt alters the HAP1–p150Glued complex, leading to the molecular motors being depleted from the MTs (3,7). This reduces the intracellular transport of BDNF, resulting in a reduction of neurotrophic support and increased neuronal death (3).
The alteration of axonal transport observed in HD is linked to a defect in htt function in transport. Reducing htt levels by ribonucleic acid (RNA) interference, genetic manipulations in Drosophila or antibody blocking leads to a reduction in MT-based transport similar to that observed with the polyQ expansion (2,3,9). The loss of the htt transport function can be rescued by re-expressing wt-htt but not polyQ-htt (3). This subsequently restores the ability of neuronal cells to release neurotrophic factors and increases their survival. Therefore, compounds or signaling pathways that rescue the defective transport may be of therapeutic interest. Along this line, the transport deficit in HD can be ameliorated by treating neurons with histone deacetylase 6 (HDAC6) inhibitors (12). This effect involves tubulin acetylation and the recruitment of molecular motors to MTs but does not involve htt itself.
There are various known phosphorylation sites in htt. We recently demonstrated that phosphorylation of wt-htt at position S421 (serine 421) by the kinase Akt controls the direction of transport in neurons (13). Indeed, when htt is phosphorylated, kinesin-1 is recruited to vesicles and MTs, favoring anterograde transport. Conversely, in the absence of phosphorylation, retrograde transport is facilitated. In HD, phosphorylation of mutant htt at S421 abolishes the toxicity of mutant htt (17,18). Furthermore, the Akt pathway is altered and a decreased S421 phosphorylation is observed (17–20) suggesting that phosphorylation of htt at S421 is a key process that could participate in pathogenesis. However, how this phosphorylation leads to neuroprotection is unknown. Indeed, we could not find obvious effect of phosphorylation at S421 of mutant htt on its subcellular localization, its ability to be ubiquitinated or degraded, or its susceptibility to cleavage by caspases, as has been proposed for S434 of htt (21).
Here, we tested whether such phosphorylation could rescue the axonal transport function of htt. We show that phosphorylating mutant htt completely restores its ability to transport BDNF-containing vesicles. Mutant htt phosphorylation leads to a restoration of the interaction properties between htt and the molecular motors and a recovery of htt and p150Glued dynactin subunit capacities to interact with MTs. We also show that insulin growth factor type I (IGF-1) and its downstream kinase Akt by inducing htt phosphorylation, compensate for the transport deficit of BDNF-containing vesicles in HD. Finally, phosphorylation of polyQ-htt at S421 by restoring polyQ-htt capacity to promote transport, results in an increased BDNF release.
Htt phosphorylation at S421 restores defective transport function
We analyzed the dynamics of BDNF-eGFP-containing vesicles in mouse neuronal cells (wt, +/+) using fast three dimensional (3D) videomicroscopy followed by deconvolution (3,12). Endogenous htt may affect the findings, so we developed a strategy based on the silencing of endogenous htt and re-expression of various htt variants. Videoexperiments were performed 1 day after lipofection in conditions in which no toxicity was observed irrespective of the conditions or the constructs used (Fig. 1). We first reduced endogenous htt levels by RNA interference using siRNA targeting the mouse htt sequence (siRNA1) (Fig. 1A and B). As previously reported and in agreement with the positive regulatory role of htt in transport (2,3,9,12), reducing htt levels decreased the velocity of BDNF vesicles in cells (Fig. 1C). Next we introduced an N-terminal 480 amino acid fragment of htt with a wt glutamine stretch (17Q, wt-htt): this N-terminal fragment restored the velocity of BDNF vesicles to control values. The siRNA1 targets the endogenous mouse htt, and had no effect on the expression of the various transgenes used (Fig. 1B). The same results were obtained with a siRNA whose target sequence is outside the N-terminal 480 amino acid fragment (siRNA2) (Fig. 1A, D and E). This confirms the absence of any off target effects and that the wt-htt fragment recapitulates the transport function of the full-length protein.
We next tested (i) whether the mutant htt containing a pathological polyQ expansion (68Q, polyQ-htt) stimulated transport of BDNF vesicles and (ii) the effect of constitutive phosphorylation at S421 in this context. As expected, the velocity of BDNF-containing vesicles in presence of polyQ-htt was comparable with the siRNA1 situation (Fig. 2A; Supplementary Material, Movie 1); thus polyQ-htt was unable to stimulate transport. We then silenced endogenous htt and expressed a polyQ-htt construct containing a serine to aspartic acid mutation, polyQ-htt–S421D, mimicking constitutive phosphorylation. Expression of this polyQ-htt–S421D construct led to a complete recovery of the transport of BDNF-containing vesicles: their velocity was not significantly different from that in the presence of scramble RNA (Control, scRNA + pcDNA3) (Fig. 2A; Supplementary Material, Movie 1). Visualization of the paths followed by randomly selected individual vesicles also revealed that S421D in the polyQ context increased the displacement of BDNF vesicles such that they were similar to those in the wt (Fig. 2B).
Phosphorylation of polyQ-htt at S421 restores BDNF anterograde and retrograde transport in neurons and increases BDNF release
We investigated the physiological consequences of polyQ-htt phosphorylation at S421. We electroporated primary cultures of embryonic E17 rat cortical neurons with constructs encoding BDNF-eGFP and the various 480 amino acid N-terminal fragments of htt and tracked the vesicles in neurites. Anterograde and retrograde velocities were scored according to the direction the vesicles moved relative to the cell body (Fig. 3A). Work with primary cultures of neurons shows that most outward and inward movements correspond to anterograde and retrograde movements, respectively, according to the organization of the MTs (3,22). Anterograde and retrograde transport of BDNF in cortical neurons was altered in disease situation, and transport in both directions increased when polyQ-htt was constitutively phosphorylated at S421 (polyQ-htt–S421D) (Fig. 3A). Thus, phosphorylation at S421 is able to restore completely the capacity of htt to transport vesicles in neurons in both directions.
We next determined the capacity of neurons to release BDNF. We first depolarized cortical neurons to release the pool of vesicles present at the membrane and performed a second depolarization (K2) to release the vesicles that had been transported to the plasma membrane. The quantity of BDNF released during the second depolarization reflects transport-dependent release of BDNF (3,12). As previously shown, less BDNF was released when htt contained the polyQ expansion compared with wt-htt (Fig. 3B) (3,12). However, more BDNF was released by neurons expressing the polyQ-htt construct that is constitutively phosphorylated at S421 than by those containing polyQ-htt. BDNF production, as measured in the cell lysates, was similar in all conditions indicating that the observed differences in BDNF release were not due to differences in production (data not shown). Therefore, phosphorylating polyQ-htt at S421 leads to an increase in BDNF release.
Htt phosphorylation restores htt–p150Glued interaction
How does phosphorylation of htt restores BDNF trafficking? In the disease situation, polyQ-htt binds with a higher than normal affinity to the HAP1 protein resulting in an abnormally tight interaction between htt and p150Glued. Alteration of this complex is accompanied by the detachment of the htt–p150Glued complex from the MTs (3,7,12). We first analyzed the effect of htt phosphorylation on its interaction with p150Glued in cells transfected with various 1301 amino acid N-terminal fragments (Fig. 1A) by immunoprecipitation (IP) experiments using an anti-htt antibody. We used these 1301 amino acid N-terminal fragments as they immunoprecipitated efficiently from HEK 293 cell extracts. However, similar results were obtained using the corresponding 480 amino acid N-terminal fragments (data not shown). The efficiency of IP of the 1301 amino acid N-terminal fragments was not influenced by the presence of an abnormal polyQ stretch and the S421D mutation (Supplementary Material, Fig. S1) or by the quantity of immunoprecipitated protein (data not shown). As previously reported (3), the polyQ-containing htt showed a significantly stronger interaction than wt-htt with p150Glued. However, the interaction of polyQ-containing htt carrying the S421D mutation was similar to that of the wt (Fig. 4). Quantification showed that S421 phosphorylation significantly restored the htt–p150Glued interaction to the normal situation.
Htt phosphorylation modulates the htt–p150Glued association with MTs in vitro and in cells
BDNF transport is MT dependent and changes in BDNF dynamics associated with disease or phosphorylation do not result from a switch to actin filaments (3,12) (data not shown). Htt localizes on MTs and in disease, the localization of htt and dynactin on MTs is disrupted (3,11,12). We therefore investigated the consequences of htt phosphorylation and the subsequent modification of htt–p150Glued interaction on their association with MTs. Mouse neuronal cells transfected with mCherry-tubulin and N-terminal 480 amino acid fragments fused with GFP (green fluorescent protein) were fixed shortly after transfection to avoid overexpression artifacts and were analyzed for htt distribution on fluorescent MTs (Fig. 5A). Less GFP-polyQ-htt than GFP-wt-htt localized on MTs and more was found in the Golgi region consistent with reduced transport. This abnormal localization pattern was completely absent with the polyQ-htt–S421D. We quantified the recruitment of htt to MTs and found these differences to be statistically significant (Fig. 5C). This indicates that htt phosphorylation restores the association of htt with MTs. We also analyzed the effect of polyQ expansion and S421 phosphorylation on htt distribution with respect to BDNF-containing vesicles by preparing synaptic vesicles; there were no significant differences between GFP-polyQ-htt–S421D and GFP-wt-htt (data not shown). This agrees with the observation that polyQ expansion in htt detaches the molecular motor and the associated vesicles from MTs but has no effect on htt association with vesicles (3).
By preparing taxol-stabilized MT fractions, we previously demonstrated that the binding of the p150Glued subunit of dynactin with MTs is disrupted in the presence of polyQ-htt (3). Here, we used an alternative approach allowing the analysis of the influence of htt phosphorylation on the association of p150Glued to individual MTs polymerized in vitro (Fig. 5B). Purified tubulin was polymerized into MTs, incubated with lysates from cells transfected with the various N-terminal fragments of htt and pelleted onto coverslips (12,23,24). Consistent with previous studies (25,26), p150Glued bound with MTs in vitro. We quantified the recruitment to MTs polymerized in vitro by 3D linescan analysis of individual MTs (Fig. 5D). The binding of p150Glued with MTs was significantly lower than in the wt situation when htt contained the pathological polyQ expansion (Fig. 5B and D). Strikingly, we observed a marked increase in p150Glued binding with MTs when polyQ-htt was constitutively phosphorylated. Thus, experiments in cells and in vitro demonstrate a decreased association of the htt–p150Glued complex with MTs that is reverted when htt is constitutively phosphorylated at S421.
IGF-1 and Akt compensate for the transport defect in HD
Phosphorylation of htt at S421 restores htt–p150Glued interaction and its association with MTs and thereby rescues transport and release of BDNF. We investigated whether manipulating the IGF-1/Akt signaling pathway could compensate for the axonal transport defect of mutant htt. Both IGF-1 treatment and Akt overexpression in cells and neurons induced strong and sustained htt phosphorylation at S421 as shown by the use of a specific antibody recognizing htt phosphorylated at S421 (17) (data not shown and Supplementary Material, Fig. S2). We first assessed the consequences of IGF-1 treatment by analyzing BDNF-eGFP vesicular dynamics in wt (+/+) and mutant (109Q/109Q) mouse neuronal cells derived from knock-in mice in which a CAG expansion was inserted into the endogenous mouse htt gene (27). We visualized the paths followed by individual vesicles randomly selected and consistent with previous studies (3,12), found significantly less displacement of BDNF vesicles in 109Q/109Q than wt cells (Fig. 6D). Similarly, the velocity of BDNF vesicles was lower in 109Q/109Q cells than +/+ cells (Fig. 6A). IGF-1 treatment rescued vesicular transport in 109Q/109Q cells by significantly increasing the velocity of BDNF-containing vesicles. To establish whether our findings are relevant to HD pathogenesis, we analyzed the effect of IGF-1 on primary cultures of cortical neurons after transfection with BDNF-eGFP and either full-length htt with a wt expansion of 17Q (wt-FL-htt) or a polyQ expansion of 75Q (polyQ-FL-htt). As expected, the mean velocity of BDNF-containing vesicles was lower in the presence of an abnormal polyQ expansion than that of wt-FL-htt (Fig. 6B). When polyQ-htt-containing neurons were treated with IGF-1, the velocity was similar to that of the wt situation. We tested whether the serine/threonine kinase Akt, the downstream target of IGF-1 is also able to compensate for the transport defect observed in HD cells. We co-transfected a constitutively active form of Akt (Akt c.a.) and BDNF-eGFP in 109Q/109Q cells. Akt c.a. significantly increased the velocity (Fig. 6C) and displacement (Fig. 6D) of BDNF vesicles to wt levels (Supplementary Material, Movie 2).
Finally, we compared the ability of IGF-1 to increase transport in cortical neurons expressing polyQ-htt with or without the S421A mutation (Fig. 7). IGF-1 increased anterograde and retrograde BDNF transport in the presence of polyQ-htt to wt levels (Fig. 3 and 7), therefore recapitulating the effect of a constitutive phosphorylation at S421 (polyQ-htt–S421D) (Fig. 3). However, IGF-1 treatment had no effect in cells expressing an unphosphorylatable form (S421A mutation) of polyQ-htt (Fig. 7). This unequivocally demonstrates that the stimulatory effect of IGF-1 requires the presence of an intact S421 in the polyQ-htt protein. Similar results were obtained with neurons expressing Akt c.a. (data not shown). Although both IGF-1 and Akt increase velocity close to wt values, we found Akt to be more potent at increasing transport. The difference might be due to the type of treatment (drug addition for IGF-1 and transfection of a constitutive active form in the case of Akt). Indeed, phosphorylation at S421 was greater in the Akt c.a. than in the IGF-1 experiments (17) (data not shown and Supplementary Material, Fig. S2).
These various findings indicate that the IGF-1/Akt pathway by phosphorylating polyQ-htt at S421, corrects the defect in intracellular transport of BDNF-containing vesicles associated with HD.
Structure prediction software suggests that S421 maps in a highly disordered and exposed region of the protein (data not shown), and this is consistent with its phosphorylation having an important regulatory role. We demonstrate that htt S421 phosphorylation leads to the restoration of htt function in MT-based transport. We previously reported that in the disease situation, the abnormal polyQ expansion in htt alters the htt–HAP1–p150Glued complex, leading to the molecular motors being depleted from the MTs (3). Here, we show that phosphorylation of polyQ-htt restores its correct interaction with molecular motors and their association with MTs. Our findings further indicate that one key function of the htt protein is to act as a facilitator of MT-based transport in neurons. Indeed, htt associates with the molecular motor machinery and promotes intracellular transport of organelles (3,7–10). Also, our findings that a single phosphorylation of htt restores its function highlight the importance of the protein context as a specific component for each of the diseases due to polyQ expansion. This emphasizes the importance of studying both htt function and htt regulation by signaling pathways to elucidate the pathogenic mechanisms in HD.
Phosphorylation of the wt protein affects the direction of transport through the stimulation of BDNF anterograde transport (13). In HD, the presence of polyQ in htt leads to a reduced anterograde and retrograde transport of BDNF-containing vesicles. The loss of htt function in transport is rescued by phosphorylation. Interestingly, the presence of the pathological polyQ expansion not only decreases overall BDNF dynamics but also impairs the capacity of htt to switch the transport direction. Nevertheless, when S421 is phosphorylated both anterograde and retrograde transport are rescued to similar extent leading to the recovery of htt stimulatory function in intracellular transport and, to increased neuronal survival both in in vitro and in vivo HD models (3,17,18).
Akt, also known as protein kinase B, is a serine/threonine kinase that controls many biological processes (28). This pathway is one of the most important pathways regulating cell survival. Akt exerts its effect by directly phosphorylating components of the apoptotic pathway. For example, phosphorylation by Akt counteracts the effects of the proapoptotic Bcl2-related protein, BAD. Akt also regulates cell survival through the phosphorylation of various transcription factors including members of the forkhead family of transcription factors, the transcription nuclear factor κB (NF-κB) and the cAMP-response element binding protein transcription factors. Here, we describe a new anti-apoptotic role of Akt in the context of HD linked to the restoration of htt function in MT-based transport of trophic factors.
Correcting the abnormal intracellular dynamics in HD is a promising therapeutic approach. We have shown that enhancing BDNF secretion from the Golgi apparatus compensates for the deficit observed in HD (29); it leads to a significant increase in the striatal release and systemic blood levels of BDNF. Similarly, increasing the recruitment of molecular motors through MT acetylation upon HDAC6 inhibition increases BDNF transport and release (12). We demonstrate that htt protein can be manipulated directly to restore its function in intracellular transport more specifically. Thus, drugs that increase htt phosphorylation in the pathological situation are of interest. Calcineurin dephosphorylates S421 (18), and inhibition of calcineurin by treatment with molecules such as FK506 increases phosphorylation of S421 and prevents polyQ-induced toxicity in striatal neurons. Therefore, molecules that act as calcineurin inhibitors may have neuroprotective effects due to an increased transport. Finally, as shown with cysteamine (29), blood BDNF levels could be used as a biomarker to validate drugs aimed at restoring the defective intracellular vesicular dynamics in HD.
MATERIALS AND METHODS
Statview 4.5 software (SAS Institute Inc., Cary, NC, USA) was used for statistical analysis. Data are expressed as mean ± SEM. Complete statistical analyses may be found in Supplementary Material.
Constructs, siRNA and antibodies
Constructs encoding Akt c.a., BDNF, BDNF-eGFP and mCherry-tubulin were described elsewhere (17,30,31). BDNF-mCherry, a generous gift from G. Banker (Oregon Health and Science University, Portland, OR, USA) behaves as BDNF-eGFP and both BDNF-tagged constructs show cellular localization, processing, and secretion properties indistinguishable from those of endogenous BDNF. BDNF-mCherry was used in Fig. 1E and Fig. 7. All htt constructs are human–mouse hybrids derived from mouse htt cDNA: first exon of mouse htt has been substituted by the homologous human one in mouse full-length cDNA (32). The wt and polyQ-htt constructs 480-17Q, 480-68Q, 480-68Q-S421D, 480-68Q-S421A, FL-17Q/75Q, YFP-htt-1301-17Q/73Q have been previously described (16,17,33). GFP-480-17Q/68Q were obtained by insertion of EcoRI–BamHI fragment of 480-17Q/68Q in the peGFP-C2 vector (Clontech, Mountainview, CA, USA). eYFP-C1 was from Clontech, pcDNA3 from Invitrogen (Breda, Netherlands). GFP-480-68Q-S421D was obtained by insertion of the EcoRI–XbaI fragment of 480-68Q-S421D-HA (18) in the peGFP-C2 vector. YFP-htt-1301-73Q-S421D was obtained using QuickChange site-directed mutagenesis (Stratagene, La Jolla, CA, USA) on YFP-htt-1301-73Q as a template and complementary oligonucleotides (5′-GCCGAAGCCGTAGTGGGGATATTGTGGAACTTATAGC-3′). The siRNAs targeting mouse htt correspond to the coding region 361–380 (siRNA1) and 2803–2822 (siRNA2) of htt mouse mRNA (Acc. no. XM_132009), the scRNA Control (Eurogentec, Seraing, Belgium) used had a unique sequence which did not match with any sequence in the genome of interest. Monoclonal antibody used were against htt (clone 1HU-4C8; 32), p150Glued (clone 1, BD Biosciences, San Jose, CA, USA), α-tubulin and β-tubulin-CY3 conjugated (Sigma, St Louis, MO, USA) and mouse IgG (Upstate, Charlottesville, VA, USA). Anti-mouse secondary antibodies conjugated to HRP or to CY5 were purchased from Jackson Immunoresearch Laboratories (West Grove, PA, USA).
Cell culture and transfection
wt (+/+) and mutant (109Q/109Q) mouse neuronal cells, HEK 293 cells and primary cortical neurons from E17 rat embryos were prepared, cultured and transfected as described (3,16,17,27). Cortical neurons were electroporated with the rat neuron Nucleofector® kit according to the supplier’s manual (Amaxa, Biosystem, Köln, Germany). Mouse neuronal cells were lipofected (Lipofectamine™ 2000, Invitrogen). For htt gene replacement, cells were first electroporated with siRNA1 or siRNA2 24 h prior to lipofection. For IGF-1 treatment, cells were treated with IGF-1 (50 ng/ml; Sigma) or vehicle (culture media) 60 min before videomicroscopy. All the experiments and in particular the videoexperiments were conducted in conditions in which no overt toxicity of the various constructs nor aggregation could be detected. The co-expression of BDNF and the various constructs as well as RNAi efficiency was verified by immunostaining after the videomicroscopy experiments. Ninety five percent or more of co-expression was observed. Single cell analysis of immunostaining levels revealed no difference in htt expression between the different constructs or signs of apoptosis in the conditions used.
For IP experiments, HEK 293 cells were transfected with pcDNA3, YFP, YFP-htt-1301-17Q/73Q and YFP-htt-1301-73Q-S421D fragments and lysed in Triton X-100 buffer (pH 7.4; 160 mm NaCl; 50 mm Hepes; 2.5 mm MgCl2; 1.5 mm CaCl2; 2.5 mm KCl; 1% Triton; 1 mm PMSF; 2 µg/ml Pepstatin; 2 µg/ml leupeptin; 2 µg/ml Aprotinin; 1 mm Iodoacetamide; 10 mm β-glycerophosphate; 2 mm DTT; and 1 mm orthovanadate). Lysates (500 µg at 1 µg/μl) were precleared 1 h at 4°C with 100 µl of a 50% solution of protein G beads. Extracts were incubated for 1 h at 4°C with 2 µl of htt-4C8 antibody or 8 µl of mouse IgG prebound with 100 µl of a 25% solution of protein G sepharose beads (Sigma). Beads were washed three times with Triton X-100 buffer. Bound proteins were eluted with sodium dodecyl sulfate (SDS) loading buffer and resolved by SDS–PAGE(polyacrylamide gel electrophoresis) and transferred onto a nitrocellulose membrane. Membranes were blocked for 1 h in 5% dried milk and Tris buffer saline containing 0.1% Tween-20 and incubated overnight at 4°C in Tris buffer saline containing 0.1% Tween-20 with the primary antibody, htt (1:5000), p150Glued (1:1000) and tubulin (1:3000). After 1 h incubation with the corresponding HRP-anti-mouse secondary antibody (1:10 000), proteins were revealed with a chemiluminescent substrate (Pierce, Rockford, IL, USA).
In vitro binding assay
The method was adapted from (12). Briefly 3 µg of purified tubulin (Cytoskeleton) were polymerized in PEM buffer [100 mm piperazine-N,N′-bis(ethanesulfonic acid) pH 6.7, 1 mm ethylene glycol tetra-acetic acid (EGTA), 1 mm MgCl2] including 1 mm guanosine 5′-triphosphate (GTP), 200 µM taxol for 30 min at 37°C. MTs were pelleted at 20 000g for 30 min at room temperature (RT), resuspended in PEM buffer containing 20 µM taxol (Sigma), 1 mm GTP and 47 µM AMP-PNP and then sheared with a 25G needle. Cells were scrapped at RT in PBS (phosphate buffer solution) containing 1 mm PMSF; 2 µg/ml Pepstatin; 2 µg/ml leupeptin; 2 µg/ml Aprotinin; 1 mm Iodoacetamide; 10 mm β-glycerophosphate; 2 mm DTT; and 1 mm orthovanadate. Pellet were resuspended in PM buffer (PEM buffer without EGTA) with 1 mm CaCl2 and 0.4% NP40, incubated 30 min on ice to depolymerize MTs and centrifugated 15 min at 10 600g. Following incubation with depolymerized cellular extracts, MTs were pelleted onto a 12 mm coverslip precoated with polyD-lysine (BD Bioscience) at 27 000g for 15 min, fixed and visualized by immunofluorescence.
Videomicroscopy experiments and imaging treatment
Videomicroscopy experiments were done 1–4 days after transfection. Cells were co-transfected with BDNF-eGFP or BDNF-mCherry and various constructs of htt or the corresponding empty vectors with a DNA ratio of 1:4. Live videomicroscopy was carried out using a Leica DM IRBE microscope and a PL APO oil ×100 objective with a numerical aperture of 1.40–0.70, coupled with a piezo device (PI) and recorded with Photometrics CoolSNAP HQ2 camera (Roper Scientific, Trenton, NJ, USA) controlled by Metamorph software (Molecular Devices, Sunnyvale, CA, USA). Stacks were acquired in cultured medium at 33°C for neuronal cell and 37°C for cortical neurons. Images were collected in stream set at 2 × 2 binning with an exposure time of 50–100 ms (frequency of 2 s) with a Z-step of 300 nm. All stacks were treated by automatic batch deconvolution using the PSF of the optical system, Meinel algorithm with parameters set at seven iterations, 0.7 sigma and four frequencies. Maximal Z and time projection, animations and analyses of vesicles tracking were done with ImageJ software as previously described (12).
For image analyses of fixed samples, images were acquired at RT with a Leica DM RXA microscope with a PL APO oil ×100 NA of 1.4 objective coupled with a PI and a Micromax RTE/CCD-1300-Y/HS camera controlled by Metamorph software. The mounting medium was 0.1 g/ml Moviol 4–88 (Calbiochem, Merck Biosciences, Darmstadt, Germany) in 20% glycerol. Z-stack was of 200 nm. Deconvolution was performed as that of videomicroscopy.
For Linescan analyses, MT intensity profiles were obtained using MT profiler, a homemade plug-in to the open-source software ImageJ. A defined rectangular region of interest was placed around the MT. The connected maximum intensity path was retrieved in 3D, starting from the tip. Fluorescence signal was then quantified along this path. For quantification of p150Glued recruitment on MT, results were expressed as the density of fluorescence (sum of motor intensity along the path divided by the length of MT) and were normalized to control (wt-htt). About 50 MTs were analyzed as per condition.
Quantification of htt localization on MTs was achieved by creating a binary mask from the MT image. First, a Gaussian filter with a 5 pixel radius was used to smooth the original image. To get rid of the uneven background, a 3D top-hat filter was then applied, using a 5 × 5 × 2 pixels kernel. The MT image was further binarized using the overall minimum gray level value of the full stack as a threshold. MT gray level value was set to 1 whereas the surrounding background pixels were set to 0 to obtain the mask image. Finally, the mask was dilated by 5 × 5 pixels in order to ensure the retrieval of all the structures (objects) associated with MT in the remaining process. The number of htt objects in the cells was determined (total objects). The image of htt signal on MTs was obtained by multiplying the latter and the htt image and scoring (MT-associated objects). Quantifications were expressed as the percentage of the MT-associated objects on the total objects. All processes were automated using macros written for ImageJ extended by a homemade plug-in for 3D top-hat filtering.
BDNF immunoenzyme assays
Cortical neurons were electroporated with BDNF and the plasmids of interest. To measure transport-dependent release, cells were depolarized [treatment for 20 min with neuronal culture media containing high K+ (30 mm CaCl2, 30 mm NaCl, 28 mm KCl)], washed and incubated 30 min with neuronal culture media, and depolarized again (K2). The amount of BDNF was measured in supernatants and cell lysates using BDNF Emax Immunoassay System (Promega, Charbonnieres, France) (12) and BDNF release was expressed as K2/lysate ratio.
This work was supported by grants from Agence Nationale pour la Recherche - Maladies Rares (ANR-MRAR-018-01 to F.S.), Association pour la Recherche sur le Cancer (ARC, 3665 to S.H.), Fondation pour la Recherche Médicale (FRM) and Fondation BNP Paribas (F.S.), HighQ Foundation (F.S., S.H.), Provital - P. Chevalier (F.S., S.H.). D.Z. was supported by a FRM and Swiss National Science Foundation postdoctoral fellowships; E.C. by a MRT and FRM doctoral fellowships; H.R. by BDI-CNRS and FRM doctoral fellowships; G.L. by Ile de France postdoctoral fellowship. F.S. and S.H. are Institut National de la Santé et de la Recherche Médicale/Assistance Publique-Hôpitaux de Paris investigators.
We acknowledge L.R. Gauthier and G. Grange for help with experiments; G. Banker for reagent; F.P. Cordelières and the Institut Curie Imaging Facility for image acquisition and treatment and members of the Saudou/Humbert’s laboratory for helpful comments.
Conflict of Interest statement. The authors declare that they have no conflicts of interest.