The KIF1A homolog Unc-104 is important for spontaneous release, postsynaptic density maturation and perisynaptic scaffold organization

The kinesin-3 family member KIF1A has been shown to be important for experience dependent neuroplasticity. In Drosophila, amorphic mutations in the KIF1A homolog unc-104 disrupt the formation of mature boutons. Disease associated KIF1A mutations have been associated with motor and sensory dysfunctions as well as non-syndromic intellectual disability in humans. A hypomorphic mutation in the forkhead-associated domain of Unc-104, unc-104bris, impairs active zone maturation resulting in an increased fraction of post-synaptic glutamate receptor fields that lack the active zone scaffolding protein Bruchpilot. Here, we show that the unc-104brismutation causes defects in synaptic transmission as manifested by reduced amplitude of both evoked and miniature excitatory junctional potentials. Structural defects observed in the postsynaptic compartment of mutant NMJs include reduced glutamate receptor field size, and altered glutamate receptor composition. In addition, we observed marked loss of postsynaptic scaffolding proteins and reduced complexity of the sub-synaptic reticulum, which could be rescued by pre- but not postsynaptic expression of unc-104. Our results highlight the importance of kinesin-3 based axonal transport in synaptic transmission and provide novel insights into the role of Unc-104 in synapse maturation.

Kinesins are microtubule based molecular motors that transport various cargos including membranous organelles, protein complexes and messenger RNAs 1 . Thus, they are of fundamental importance for the establishment, plasticity, injury response and survival of neuronal networks 2,3 . C. elegans Unc-104 4,5 , Drosophila Unc-104/Imac 6 and the mammalian kinesin-3 family members KIF1A and KIF1B 7,8 are the predominant motor proteins for the fast anterograde transport of membranous organelles. KIF1A knockout mice display severe neurological abnormalities including motor and sensory disturbances and die shortly after birth 8 . In Drosophila complete loss of unc-104 function in motor neurons leads to an arrest of synaptogenesis and embryonic lethality 6 .
In humans recessive, autosomal dominant and spontaneous mutations in KIF1A have been associated with hereditary spastic paraplegia (HSP) [9][10][11][12][13] , and hereditary sensory and autonomic neuropathy type IIC (HSAN2C) 14 . Moreover, mutations in KIF1B has been implicated in Charcot-Marie-Tooth disease 15 . These diseases primarily affect nerve cells that have long axons and are thus most dependent on efficient cargo transport, consistent with kinesin-3's important role for long-range intracellular trafficking. Apart from these symptoms in the peripheral nervous system, HSAN2C related mutations in KIF1A were also described to affect the central nervous system causing mental retardation and brain atrophy 9,16,17 . While unc-104's mammalian homolog KIF1A is important in facilitating synaptogenesis in an experience dependent manner 18 , Unc-104 has been proposed to be part of the molecular machinery that regulates activity-dependent feedback in Drosophila photoreceptors 19 . Thus, defects in plasticity might be of major pathological relevance in the context of disease-related loss of KIF1A function.
We have previously reported a hypomorphic allele, unc-104 bris which causes partial loss of kinesin-3 function and thus permits detailed analysis of synapse maturation at the larval stage unlike the unc-104 null mutant that dies at late-embryonic stage 6 . The recessive point mutation unc-104 bris at the forkhead associated (FHA) domain causes morphological changes at the neuromuscular junction (NMJ) such as increased NMJ length and synaptic bouton number as well as reduced bouton size 20 . In addition, unc-104 bris have increased proportion of postsynaptic glutamate receptor fields that are unapposed by presynaptic active zone (AZ) scaffolding protein Bruchpilot (Brp) [20][21][22][23] . Furthermore, we have recently shown that these Brp-negative synapses lack all major AZ components investigated, such as Liprin-α , Cacophony and SRPK79D, suggesting a crucial role of unc-104 in presynaptic maturation 23 .
Despite the mounting evidence on the consequences of unc-104 loss-of-function on the maturation of the presynaptic terminal, little is known about its impact on postsynaptic development. In the present study, we examine functional and structural defects at unc-104 bris mutant NMJs with emphasis on structural deficits in the postsynaptic compartment. We found defects in synaptic transmission and altered glutamate receptor composition as a consequence of partial loss of kinesin-3 function. In addition, the postsynaptic compartments are characterized by the reduced abundance of discs large (Dlg), Dorsal B (DorB) and α -Spectrin (α -Spec) as well as reduced folding complexity of the subsynaptic reticulum (SSR). These defects can be rescued by the presynaptic but not postsynaptic expression of unc-104. Collectively, our data reveal new functions of kinesin-3 based transport in synapse development and may have important implications for our understanding of the molecular pathway underlying neurological diseases caused by impaired kinesin-3 function.

Results
Synaptic transmission is impaired in unc-104 bris mutants. Unc-104 bris mutant larvae demonstrate a distinct synapse maturation defect. At wild-type NMJs only the youngest synapses, in sum approximately 5%, lack the AZ organizing protein Brp 24 . In contrast, at the NMJs of unc-104 bris mutant larvae about 25% of the postsynaptic densities (PSDs) are unapposed by Brp 20,23 . In order to examine the functional consequences of impaired kinesin-3 function, we performed intracellular current clamp recordings at 3 rd instar larval NMJs. The amplitude of evoked and miniature excitatory junctional potentials (EJPs and mEJPs) was reduced by 72% and 50% respectively in unc-104 bris mutants compared to controls (Fig. 1a-d). The quantal content, which is an estimation of the number of synaptic vesicles (SVs) released per evoked release event, was decreased by 47% at unc-104 bris mutant NMJs (Fig. 1e). Furthermore, we observed a 20-fold reduction in mEJP frequency in unc-104 bris larvae. All above described functional defects were rescued by ectopic pan-neuronal expression of UAS-unc-104-mcherry (Fig. 1c,f). Expression of the rescue construct in the control background did not affect junctional potentials but resulted in a 42% increase in mEJP frequency (Fig. 1c,f).
Unc-104 bris mutants display altered glutamate receptor composition. Multiple presynaptic structural defects have so far been identified in the context of loss of unc-104 function 6,20,23 . However, the importance of Unc-104 for postsynaptic development is largely unexplored. We next investigated whether impaired synaptic transmission at unc-104 bris mutant NMJs might be caused by altered postsynaptic glutamate receptor composition and/or clustering. To this aim we first performed immunohistochemistry at the NMJ using antibodies against Brp and obligate glutamate receptor subunit IIC (GluRIIC). We observed a 17% reduction of the average GluR field size at the NMJ (Fig. 2a,b). The reduction in size is reflected by plotting all GluR fields based on their size, where unc-104 bris mutant NMJs have an increased percentage of small GluR fields (Fig. 2c). To address the role of presynaptic Brp presence on the size of GluR fields, we limited our size analysis only to those GluR fields that were positive for Brp. Intriguingly, we did not observe a difference in GluR field size in the subset of synapses that were positive for Brp (Fig. 2a,d).
Two types of ionotropic glutamate receptors have been identified at the Drosophila NMJ synapses. Apart from the three obligatory subunits, they either contain GluRIIA or GluRIIB as a fourth subunit to form a functional heterotetrameric receptor 25 . The Drosophila NMJ contains a mixture of IIA-and IIB-type glutamate receptors. We sought to assess whether differential incorporation and stabilization of glutamate receptor complexes could be observed at unc-104 bris mutant NMJs. To visualize GluRIIB and GluRIIA, we utilized a double transgenic line expressing GluRIIB-GFP and GluRIIA-mRFP under the control of their native promotors 24,26 . The intensity of GluRIIA-mRFP was reduced in unc-104 bris mutant synapses (Fig. 2e, arrowheads and f). Localization and abundance of GluRIIB-GFP remained unchanged in unc-104 bris mutants (Fig. 2e, arrowheads and g). It has been reported that receptor composition changes during synapse maturation at Drosophila NMJ: IIA-type receptors are predominant in immature synapses, whereas the IIA-and IIB-type receptor ratios are balanced as synapses mature 26 . We quantified the IIA-and IIB-type receptors for each PSD and found that consistent with previous reports, small synapses in control group were "IIA-rich" compared with mature ones which had a more balanced glutamate receptor composition (Fig. 2h). In contrast, IIA-receptors are not enriched at small synapses of unc-104 bris mutant larvae (Fig. 2h). Irrespective of their size, synapses of unc-104 bris mutant larvae contain less IIA-type receptors than control (Fig. 2i), while no changes were observed in the amount of GluRIIB per synapse between control and unc-104 bris mutants (Fig. 2j). IIA-and IIB-type glutamate receptors exhibit distinct electrophysiological properties 25,[27][28][29] , with the IIA-type receptors showing higher single channel conductance than IIB-type receptors 28 . Therefore, the altered receptor composition may contribute to the decreased EJP and mEJPs sizes in unc-104 bris mutants.
Scientific RepoRts | 7:38172 | DOI: 10.1038/srep38172 Impaired abundance of post synaptic markers in unc-104 bris mutants. Membrane associated guanylate kinase-like proteins (MAGUKs) including PSD protein 93 and 95 (PSD-93, PSD-95), SAP97 and SAP102 are thought to be key organizers of molecular scaffolds at excitatory synapses 30 . We examined the localization of Dlg 31 , the Drosophila homolog of PSD-95, at unc-104 bris mutant synapses to further investigate the impact of reduced Unc-104 function on postsynaptic development. We found that Dlg was severely reduced at unc-104 bris mutant NMJs (Fig. 3a). Next, we examined two additional postsynaptic markers, the Drosophila NF-κ B protein DorB 31,32 and the SSR enriched, cytoskeletal protein α -Spec 33 . Similar to Dlg, the levels of both DorB and α -Spec were also reduced at the SSR (Fig. 3b,c).
Scientific RepoRts | 7:38172 | DOI: 10.1038/srep38172 Unc-104 bris mutants display gross morphological changes in SSR membranes. Type I boutons are surrounded by the SSR, which consists of densely stacked muscle membranes that are thought to be important for the stability of the neuromuscular terminal 31,34 . Loosening and vacuolization of the SSR has been described in NMJ dismantling during metamorphosis 35 . Given that partial loss of Dlg has been associated with reduced convolution of SSR membranes 31,36 , we investigated the ultrastructure of the extrasynaptic compartment of larval NMJs using electron microscopy. While wild-type boutons were surrounded by a tight array of muscle membranes, the SSR structure was less dense in unc-104 bris mutant NMJs suggesting that partial loss of Unc-104 function results in gross morphological changes in postsynaptic compartments (Fig. 4a). The ultrastructure of the SSR could be rescued by pre-but not by postsynaptic expression of unc-104 cDNA, suggesting that SSR defects are secondary to loss of presynaptic Unc-104 function (Fig. 4a). Taken together, our results show that partial loss of Unc-104 function resulted in reduced levels of postsynaptic proteins and impaired SSR morphology.

Discussion
Unc-104 bris mutant NMJs are characterized by a strong reduction in EJP amplitude (Fig. 1a,b). This is consistent with previous reports that unc-104 bris mutant larvae are severely impaired in locomotion; larval lethality is preceded by almost complete paralysis at L3 stage 20 . Notably, the strong reduction in EJP amplitude (Fig. 1a,b) and the number of vesicles released (Fig. 1d) per action potential are not being partially compensated in the postsynaptic compartment by concomitant increase in the mEJP amplitude (Fig. 1c), i.e. the response per vesicle. This is surprising as it has been previously reported that upon decrease in quantal content due to lack of Brp, mEJP amplitudes are scaled up 37 . Loss of Brp from a subset of synapses is a primary defect in unc-104 bris larvae 20 . We observed a strong reduction of mEJP frequency at unc-104 bris mutant NMJs (Fig. 1c,f). Kinesin-3 is the major transporter of SVs, and previous reports show several SV markers are severely reduced at unc-104 mutant NMJs 6,23 . Loss of SVs from the NMJ as well as alterations in the size of the readily releasable pool are possible causes for the decreased mEJP frequency at unc-104 bris mutant NMJs. While there is direct ultrastructural evidence for the loss of SVs from boutons in unc-104 null mutant embryos 6 , we did not observe any reduction in the amount of SVs at central synapses in unc-104 bris mutant larvae 23 . Interestingly, while presynaptic over-expression of the UAS-unc-104-mcherry transgene in wild-type background did not cause any changes in EJP or mEJP amplitudes, it resulted in an increase in mEJP frequency (Fig. 1c,f). These data suggest that the frequency of spontaneous release is likely particularly sensitive to the abundance of readily releasable pool of SVs at the NMJ which might be regulated by kinesin-3 based axonal transport.
In the nervous system, homeostatic signaling is an essential feedback mechanism to ensure stable synaptic activity in a highly variable environment. Homeostatic signaling has been shown to compensate for perturbation of synaptic excitability through multiple mechanisms including changing the efficacy of SV release, ion channel density and neurotransmitter receptor composition [38][39][40][41] . The conductivity of IIA-type glutamate receptors is much larger than IIB-type receptors, and change in the ratio between these two receptors has been shown to modulate quantal size 27 . Thus, a change in receptor composition is a potential mediator of postsynaptic homeostasis. Synaptic transmission blockage via presynaptic expression of TNT leads for example to higher abundance of the GluRIIA at the PSD 26 . Likewise, the ratio of GluRIIA/GluRIIB containing glutamate receptors changes in dsyd-1 mutants due to a simultaneous increase in GluRIIA and decrease in GluRIIB abundance 42 . However, no such change was observed in liprin-α mutants, that display a similar reduction in glutamate release 42 . Alternatively, this phenotype might thus be caused by a loss of a maturation signal that is present in liprin-α mutants, but lacking in dsyd-1 mutants. Despite the severely impaired presynaptic AZ assembly 23 , no sign of postsynaptic compensation is observed in unc-104 bris mutants. Hence, the observed reduction in GluRIIA clustering is likely responsible for the decreased mEJP amplitude, and in turn contributes to the impaired EJP (Fig. 1a-c).
We have previously shown that the unc-104 bris mutation causes alteration of the gross NMJ structure, resulting in impaired synapse maturation along with severe reduced anterograde transport of SVs and dense core vesicles 20,23 . Interestingly, expression of Rab3, an SV-associated protein that is depleted from the nerve terminals of unc-104 bris mutants partially ameliorates presynaptic and postsynaptic defects 23 . This suggests that some of the phenotypes observed in unc-104 bris mutants are likely due to the depletion of Unc-104 cargo at presynaptic terminal. The present study shows a severe loss of mEJP frequency in unc-104 bris NMJs (Fig. 1c,f). Given that mEJPs are an instructive signal that guides synaptic development 43 , some of the observed morphological defects in unc-104 bris could be a result of reduced mEJP frequency (Fig. 1c,f). As presynaptic expression of Unc-104 rescues postsynaptic Dlg and SSR phenotypes (Figs 3b and 4a), it would be interestingly to address in a future study, if this rescue is dependent on specific presynaptic Unc-104 cargo.
Dlg, one of the major proteins localized to the SSR, is a known interaction partner of Gtaxin 44 , the Drosophila homolog of mammalian syntaxin-18, which regulates endoplasmic reticulum (ER) membrane trafficking 45 . It is intriguing to postulate that that Dlg and Gtaxin may cooperatively regulate addition of ER membrane to the SSR 46 , and via this membrane trafficking pathway ER characteristics might be transferred to the SSR 46 . Consistently, ample evidence suggests that local translation might occur at the SSR [47][48][49] . This raises the exciting possibility that the reduced SSR complexity observed at unc-104 bris mutant NMJs (Fig. 4) might be a result of impaired membrane addition to the SSR. The insufficiently developed SSR at unc-104 bris mutant NMJs might be less efficient in local translation of key proteins required for synaptic development. Consistent with this notion, the glutamate receptor subunit IIA, which is strongly reduced at unc-104 bris mutant NMJs (Fig. 2f), has been shown to be translated in local, subsynaptic translation aggregates 47 .
The membrane of insect muscles is invaginated at most synapses to frame a complex array of tubes and folds, the SSR. Although it is morphologically similar to subsynaptic folds that have been proposed to be important to amplify synaptic signals at the vertebrate NMJ, its primary cellular function might be different 50 . At the Drosophila NMJ, the SSR gradually develops at Type I boutons. It is more pronounced at Type Ib boutons, displaying more fully developed and deeper stacked SSR than Type Is terminals 34,51 . Type II and III fibers that often run parallel with Type I innervations might be partially embedded in the SSR emerging from the Type I innervation but do not form a multi-layered complex SSR on their own 51 . The SSR is also absent from synapses at indirect flight muscles 50 , suggesting that its role is modulatory rather than essential for neurotransmission. We found that the SSR in unc-104 bris mutant larvae is less dense than in control larvae (Fig. 4). While loosening of the SSR has been associated with NMJs dismantling during metamorphosis 35 , it is not very likely that the reduced SSR complexity at unc-104 bris mutant NMJs is due to ongoing neurodegeneration 20 . Similar with a previously described HSP Type 10 (SPG10) Drosophila model, larval locomotion was also severely impaired in unc-104 bris mutant larvae 20,52 . However, while robust structural defects were observed upon disturbance of Kinesin-1 based transport 52 , no signs of synapse dismantling were observed in unc-104 bris mutant  20 . Therefore it is more likely that the observed reduction of SSR associated proteins (Fig. 3a-c) and the reduction of SSR density (Fig. 4) at unc-104 bris mutant NMJs reflects a defect in synapse maturation and/or plasticity rather than being a sign of neurodegeneration.
The degree of behavioral impairment and the lethality of unc-104 bris larvae is similar to those described for a Drosophila model for SPG10 caused by dominant negative mutations in kinesin heavy chain (khc), the Drosophila homolog of human KIF5A 52 . Notably, although both disease models are characterized by a partial loss of kinesin function related to two forms of HSPs, different pathological progressions were observed. SPG10 model larvae display a pronounced dystonic posterior paralysis that is reminiscent of the ascending paralysis observed in SPG10 patients 52 . While SPG30 is also an ascending motoneuron disease that firstly and primarily affects nerve cells that innervate the feet of the patients, unc-104 bris larvae suffer from a more generalized paralysis, i.e. NMJs in segments innervated by longer axons seem not to be more severely affected 20 . This complex phenotype highlights that more detailed studies will be necessary to fully dissect the specific impact of kinesin-3 dysfunction on neural function and human pathology.
While truncation of KIF1A has been associated with another neurological disease HSAN2C 14 , various mutations of KIF1A have been associated with SPG30 as well as non-syndromic intellectual disability accompanied by variable additional symptoms including progressive encephalopathy and brain atrophy 9,16,17,53,54 . Recently, an amino acid exchange (S69L), proposed to be important for the ATP binding of KIF1A, has been shown to underlie SPG30 10 . A detailed comparison of different disease related mutations may give valuable insight on the specific impact of disrupting the function of different domains of KIF1A.
The amino acid mutated in unc-104 bris larvae has been suggested to further stabilize kinesin-3 dimers by electrostatic interaction with the E499 residue 55 . Defects observed in unc-104 bris larvae are not restricted to the axonal compartment, but include defects in dendrite maturation in sensory neurons 20 , a phenotype most relevant in the context of HSAN2C. We thus suggest that the animal disease model presented herein, rather than being a precise model of any of the human diseases associated with KIF1A dysfunction, might have broad implication for neurological diseases that are associated with impaired stability of kinesin-3 dimer including of HSP, HSAN2C and intellectual disability.

Methods
Fly Stocks. Flies were cultured on standard soft media seeded with live yeast at 25 °C unless otherwise indicated. w 1118 , elav-Gal4, 24B-Gal4 and unc-104 d11024 were obtained from the Bloomington Drosophila Stock Center. GluRIIA-mRFP 24 and GluRIIB-GFP 26 with endogenous promoters were obtained from Stephan Sigrist (FU Berlin). UAS-unc-104-mcherry was a generous gift from Thomas Schwarz (Harvard University).
Immunohistochemistry and microscopy. Middle 3 rd instar stage larvae were dissected in Ca 2+ -free HL3 solution and fixed in 4% formaldehyde in PBS for 3 minutes (for staining with native fluorescent proteins) or for 10 minutes (for staining with only immunofluorescent labeling). Correct NMJs were identified as previously described 56 . Primary antibody incubation was done overnight at 4 °C in PBS containing 0.05% Triton-X and 5% normal goat serum. Fillets were then washed and incubated with fluorescent-conjugated secondary antibodies at room temperature for 2 hours. Larval fillets were mounted on a glass slide in mounting medium (Vectashield, Vector). Primary antibodies used were: mouse monoclonal anti-Dlg at 1:100, mouse monoclonal anti-Brp (NC82) at 1:100, mouse anti-α -Spec (3A9) at 1:100 (Developmental Studies Hybidoma Bank), rabbit anti-GluRIIC at 1:2000 (Stephan Sigrist), rabbit anti-DorB at 1:4000 (Steven Wasserman). Fluorescent-conjugated secondary antibodies used were: goat anti-mouse Alexa 488 or Alexa 568 and goat anti-rabbit Alexa 488 or Alexa 568 (Molecular Probes) or goat anti-mouse Atto 647 (Sigma). Goat anti-HRP conjugated with Cy3 (Dianova) was added together with secondary antibodies. All fluorescent-conjugated antibodies were diluted at 1:500. Images were captured and processed essentially as previously described 20,23,57,58 . In detail, a Zeiss LSM 710 confocal microscope with a 40x plan apochromat 1.3 N.A. oil objective and the ZEN software was used to acquire images at voxel size: 100 nm * 100 nm * 500 nm; pinhole: 1 AU, average: 2-4. ImageJ 1.43 (NIH) was used for image processing and analysis.
Electron Microscopy. Sample preparation and image acquisition were performed essentially as previously described 52 . In brief, fillets were fixed with 4% PFA (in PBS) for 10 min at room temperature followed by fixation in 2.5% glutaraldehyde (in PBS) overnight at 4 °C. Next, samples were treated with 1% osmium tetroxide in 100 mM phosphate buffer, pH 7.2, for 1 h on ice. Larval fillets were rinsed with water, treated with 1% aqueous uranyl acetate (UA) for 1 h at 4 °C, dehydrated through a graded series of ethanol concentrations, and stored in liquid Epon overnight. Next, muscles 4 of segment 4 were dissected with sharp insect pins, embedded in Epon, and polymerized for 48 h at 60 °C. Ultrathin sections were stained with UA and lead citrate and viewed in a Philips CM10 electron microscope.
Electrophysiology. Current clamp intracellular recordings were performed on muscle 6 segment A2 of mid 3 rd instar stage larvae as previously described 52 . The larvae were pinned and stretched in a Sylgard-coated perfusion chamber and visualized on an Olympus BX51WI microscope. "Bee-stinger" sharp electrodes (10-15 MΩ), made of borosilicate glass (outer diameter 1.5 μ m) were filled with 3 M KCl. Only cells with resting potentials between − 55 and − 80 mV and input resistance higher than 4 MΩ were included in the analysis. Recordings were performed in HL3 Stewart saline 59 containing (in mM): 70 NaCl, 5 KCl, 20 MgCl 2 , 10 NaHCO 3 , 5 trehalose, 115 sucrose, and 5 HEPES; the concentration of Ca 2+ was 1, pH adjusted to 7.2. All experiments were performed at 18 °C. Stimulation of the segmental nerve was executed by pulling the cut end of the nerve into a self-made suction electrode (5-6 μ m in diameter) filled with HL3 and passing a brief (0.3 ms) bi-polarizing pulse across the nerve. Stimulation was accomplished with an ISO-STIM 01D stimulus isolation unit (NPI electronics GmbH, Scientific RepoRts | 7:38172 | DOI: 10.1038/srep38172 Tamm, Germany). The signal was acquired with an Axoclamp 900A amplifier (Axon Instruments), digitized with a Digidata 1440A analog to digital board, and recorded with a PC using pClamp 10. 3 (Axon Instruments), and analyzed with AxoGraph X software. The amplitudes of the EJPs were corrected for nonlinear summation. The quantal content was estimated by dividing the averaged corrected EJP by the averaged mEJP amplitude 60 .
Statistical Analysis. Statistical analysis was performed using the software PRISM 6. Sample errors are given as standard error of the mean (SEM). The following alpha levels were used for all tests: *p < 0.05; **p < 0.01; ***p < 0.001. Data were first tested for normality and then analyzed by either the student's t-test for two groups or by a one-way analysis of variance followed by a Tukey's multiple comparison test. Non-normally distributed data were analyzed by using either a Mann-Whitney test for two groups or a Kruskal-Wallis test for multiple groups.
Use of experimental animals and human subjects. This study did not involve the use of human subjects or samples from human donors. All experiments were performed in accordance with University guidelines and regulations.