A STRIPAK component Strip regulates neuronal morphogenesis by affecting microtubule stability

During neural development, regulation of microtubule stability is essential for proper morphogenesis of neurons. Recently, the striatin-interacting phosphatase and kinase (STRIPAK) complex was revealed to be involved in diverse cellular processes. However, there is little evidence that STRIPAK components regulate microtubule dynamics, especially in vivo. Here, we show that one of the core STRIPAK components, Strip, is required for microtubule organization during neuronal morphogenesis. Knockdown of Strip causes a decrease in the level of acetylated α-tubulin in Drosophila S2 cells, suggesting that Strip influences the stability of microtubules. We also found that Strip physically and genetically interacts with tubulin folding cofactor D (TBCD), an essential regulator of α- and β-tubulin heterodimers. Furthermore, we demonstrate the genetic interaction between strip and Down syndrome cell adhesion molecule (Dscam), a cell surface molecule that is known to work with TBCD. Thus, we propose that Strip regulates neuronal morphogenesis by affecting microtubule stability.

additional proteins in a mutually exclusive manner to form distinct STRIPAK complexes involved in diverse functions 9 . Although our knowledge of the composition of the striatin family complexes has increased greatly, much remains to be determined regarding the function of these complexes and the roles of the various striatin family-associated proteins 9 . There have been some indications that STRIPAK complexes might affect microtubule organization. For example, dMob4, the Drosophila Mob3 homolog is required for microtubule morphology at neuromuscular junctions, peripheral nerves and muscles 12 . Furthermore, STRIP2 knockdown in PC3 prostate cancer cells altered microtubule organization and induced cell elongation 13 . It was also reported that Dynein, a minus-end directed motor, associates with the STRIPAK complex using an affinity-purification mass spectrometry analysis 10 . However, the precise molecular mechanism of microtubule regulation by STRIPAK complexes in vivo is unclear.
We previously demonstrated that Drosophila Strip, the homolog of mammalian Strip1 and 2, regulates dendrite branching and axon elongation in Drosophila olfactory projection neurons 14 . We revealed that Strip serves as a platform for early endosome organization during axon elongation. The shorter axon phenotype caused by strip knockdown was suppressed by the expression of constitutive active form of Rab5, one of the key regulators of early endosome fusion. However, the suppression was partial, and the dendrite overbranching phenotype was not suppressed. We hypothesized that Strip might also form complexes with other molecules that affect axon elongation and dendrite branching. Thus, we speculated that Strip might affect microtubule organization, since microtubules are crucial components of axons and dendrites, and microtubule dynamics should be strictly regulated during neural development.
Here we show that Strip forms a complex with microtubules and affects their stabilization. Furthermore, we reveal that strip genetically interacts with tubulin folding cofactor D (TBCD), one of the tubulin-folding cofactors that assist the tubulin heterodimer formation and control the availability of tubulin subunits and microtubule stability 15 , to regulate neuronal morphogenesis. Moreover, we show that Strip cooperates with Down syndrome cell adhesion molecule (Dscam), a cell surface molecule whose intracellular domain binds to TBCD and it is likely that TBCD mediates Dscam functions by affecting microtubule dynamics 16 .

Results
Strip is localized along microtubules. To understand the relationship between Strip and microtubule, we first examined the subcellular localization of Strip in Drosophila S2 cells. It is difficult to evaluate whether Strip is localized on microtubules because endogenous Strip seems to be distributed throughout the cytoplasm (Fig. 1a). Therefore, we utilized the extraction method 17 to remove the cytosolic components and visualize cytoskeleton-associated proteins. As we expected, Strip was localized along microtubules (Fig. 1b). To further investigate this relationship, we performed the microtubule co-sedimentation assay. When in vitro-polymerized microtubules with no Strip were incubated with the extracts of S2 cells and pelleted by centrifugation through a glycerol cushion, S2 cell-derived Strip was associated with the microtubules (Fig. 1c). These data demonstrate that Strip forms a complex with microtubules.
Strip affects organization and stability of microtubule. We next examined the role of Strip on microtubule by treating S2 cells with strip dsRNA. Most of the microtubules showed radial projections from the cell center to the periphery in control dsRNA-treated S2 cells (Fig. 2a,b). In contrast, radial projections towards cell periphery were mostly absent and overall microtubule morphology was rounded in strip dsRNA-treated S2 cells (Fig. 2a,b). We further investigated the localization of microtubule plus end tracking protein, EB1 to monitor the microtubule organization and found that EB1 distribution was obviously altered in strip dsRNA-treated S2 cells (Fig. 2c,d). EB1 was localized at the periphery of wild-type S2 cells in a small comet-like pattern while the number of EB1 comets was significantly reduced and sometime irregular EB1 accumulations were observed in strip dsRNA-treated S2 cells.
Multiple post-translational modifications in tubulins are crucial for dynamics and organization of microtubules 18 and can be used to monitor different populations of microtubules. Therefore, we tried to investigate the level of acetylated and tyrosinated α-tubulin in strip dsRNA-treated S2 cells. Acetylated α-tubulins exist in stable, long-lived microtubules that are resistant to nocodazole 19,20 . Tyrosinated α-tubulins are found in newly formed microtubules since tubulin detyrosination occurs after the incorporation of tubulin subunits into the microtubule lattice 21 . Interestingly, the level of acetylated α-tubulin was greatly reduced when cells were treated with strip dsRNA, indicating that microtubule stability was decreased when strip was knocked down (Fig. 2e-g). The level of tyrosinated tubulin was not drastically changed in strip dsRNA-treated cells (Fig. 2f). strip and tubulin folding cofactor D cooperatively regulate neuronal morphogenesis. To investigate whether Strip also influences the stability of microtubules in vivo, we performed a mosaic analysis with a repressible cell marker (MARCM)-based analysis 22 in Drosophila olfactory projection neurons (PNs), an excellent model system for studying the molecular mechanisms of neuronal morphogenesis [23][24][25] . We analyzed MARCM single-cell clones of PNs whose dendrites target the DL1 glomerulus in the antennal lobe (DL1 PNs, Fig. 3a). Axons of the DL1 PNs target the mushroom body and the lateral horn where the axon exhibits stereotypical L-shape branching (Fig. 3a). As we reported previously, single-cell clones of PNs homozygous for strip dogi (hereafter, strip dogi PN) show defects in axon elongation and dendrite branching 14 . In strip dogi PNs, an additional dendrite branch appeared from the proximal side of the dendrite shaft (overbranching phenotype: 10.26%, n = 4/39, Fig. 3b) and two-glomerular-targeting dendrites were also observed at a low frequency 14 . The axon of strip dogi PNs did not elongate or form L-shaped branches in the lateral horn (97.2%, n = 35/36, Fig. 3b).
To determine whether Strip is involved in microtubule dynamics during neuronal morphogenesis, we examined the relationship between strip and TBCD. TBCD is one of the five tubulin-folding cofactors and assists in the formation of tubulin heterodimers 26 . Recently, we reported that an optimum level of TBCD is crucial for neuronal morphogenesis. Interestingly, PNs homozygous for TBCD 1 mutation or overexpressing TBCD show similar phenotypes with strip dogi PNs; both PNs exhibit dendrite overbranching phenotype 16 . Furthermore, TBCD 1 PNs exhibits shorter axon phenotype as strip dogi PNs. Based on these phenotypic similarities between strip and TBCD mutant, we hypothesized that Strip could function with TBCD to regulate neuronal morphogenesis. First, we examined the physical interaction between Strip and TBCD by performing a co-immunoprecipitation assay using S2 cell lysate, and found that Strip and TBCD can form a protein complex (Fig. 3c-e). Next, we examined the subcellular localization of Strip and TBCD-myc in S2 cells and found that endogenous Strip and TBCD-myc are mostly co-localized (Fig. 3f). We then examined the genetic interaction between strip and TBCD by expressing short-hairpin RNA for TBCD (shRNA-TBCD) or overexpressing TBCD in strip dogi PNs. Although PNs homozygous for TBCD 1 mutation exhibited defects in dendrite branching and axon elongation, shRNA-TBCD expression in DL1 PNs did not cause obvious phenotypes either in dendrites or axons, probably owing to its weak knockdown efficiency (Fig. 3g). In addition, TBCD overexpression also did not cause obvious phenotype (Fig. 3i), which was different from our previous report 16 that used different UAS-TBCD transgenic line (See Methods for more details). When shRNA-TBCD or TBCD was expressed in strip dogi PNs, the phenotypes of strip dogi PNs were significantly enhanced. The penetrance of two-glomeruli-targeting dendrites was increased (2.56%, n = 1/39 for strip dogi PNs, Fig. 3b; 37.5% n = 9/24 for shRNA-TBCD expressing strip dogi PNs, Fig. 3h; 30% n = 3/10 for TBCD expressing strip dogi PNs, Fig. 3j), and some axons showed 'extremely short axon' phenotype, that is, the axons did not exit the antennal lobe; this was never observed in strip dogi PNs (9.52%, n = 2/21, for shRNA-TBCD expressing strip dogi PNs, Fig. 3h; 20%, n = 2/10, for TBCD expressing strip dogi PNs, Fig. 3j). These results suggest that Strip interacts with TBCD during neuronal morphogenesis of PNs.  Strip cooperates with Dscam in mushroom body neurons. As we reported previously, TBCD forms complex with Dscam and cooperates to regulate morphology of mushroom body neurons 16 . The mushroom body is an olfactory learning and memory centre 27 and consists of three types of neurons, α/β, α′ /β′ , and γ neurons 28 . α/β neurons extend one axon into the α lobe and the other axon to the β lobe, that are labelled by anti-FasII antibody (Fig. 4a). Here we hypothesized that Strip, like TBCD, also acts in the downstream of Dscam. To test this hypothesis, we first generated MARCM clones of mushroom body α/β neurons homozygous for strip dogi (strip dogi mushroom body neurons). We found that strip dogi mushroom body neurons exhibited axon segregation phenotypes ( Fig. 4b; 11.8% (n = 8/68) of strip dogi clones exhibited less axons into the α lobes) that were similar to those in Dscam mutant mushroom body neurons 29,30 . We further provide the genetic evidence that Dscam and strip interact in mushroom body neurons. The 'α and β lobes missing' phenotypes were observed when Dscam was overexpressed in the mushroom body neurons (Fig. 4c,d,f), which was suppressed in the TBCD 1 heterozygous background 16 . Similarly, strip dogi heterozygous background could also suppress 'α and β lobes missing' phenotype ( Fig. 4e,f). Next we performed co-immunoprecipitation experiment. Anti-Strip antibody could co-immunoprecipitated endogenous Strip with Dscam-GFP (Fig. 4g). We could further confirm this interaction by immunoprecipitating Dscam-GFP by anti-GFP antibody; endogenous Strip was co-immunoprecipitated with Dscam-GFP (Fig. 4h). From these results, we concluded that Strip cooperates with Dscam in mushroom body neurons, possibly through interacting with TBCD.

Discussion
To form a functional neural circuit, the morphology of each neuron should be strictly regulated to wire with the correct synaptic partners. Microtubules are fundamental structural components of axons and dendrites and their dynamics directly affect neuronal morphology. In this report, we show that Strip, one component of the STRIPAK complex, affects microtubule stability in Drosophila S2 cells. Furthermore, we report that Strip regulates dendrite branching and axon elongation by interacting with TBCD in the olfactory projection neurons. Moreover, we observed genetic interactions between strip and Dscam in the mushroom body neurons, suggesting that Strip and TBCD regulate microtubule stability in the downstream of Dscam.
We recently found that TBCD can bind to the intracellular domain of Dscam and it is likely that TBCD mediates Dscam functions by affecting microtubule dynamics 16 . Although Dscam has been extensively studied, the downstream pathways are not well known. The Dock/Pak signaling pathway, one of the known pathways acting downstream of Dscam, does not seem to be required for dendrite targeting and axon guidance of olfactory projection neurons 31 . Strip seems to act with TBCD in the downstream of Dscam during neuronal morphogenesis, and link outer cellular environment with microtubules. strip and TBCD mutants show similar phenotypes in axon elongation and dendrite branching, and Strip and TBCD genetically and physically interact with each other. TBCD assists in the formation of tubulin heterodimers, and affects microtubule stability by controlling the availability of tubulin subunits because concentration of free tubulin dimers can affect microtubule dynamics 15 . Although mammalian TBCD does not bind to microtubules 32 , Alp1, a homolog of TBCD in Schizosaccharomyces pombe, is colocalized with microtubules 33 . Furthermore, overproduction of Cin1p, the homolog of TBCD in Saccharomyces cerevisiae, resulted in an increased sensitivity to benomyl 34 , an antifungal compound that weakly inhibits microtubule assembly 35 . Thus, TBCD and its homologs have been shown to regulate microtubule dynamics.
Strip seems to affect microtubule stability by interacting with several molecules in addition to TBCD. We previously reported that Strip also forms a complex with Glued, the homolog of mammalian p150 Glued,14 . Glued is one of the components of the dynactin complex required for dynein motor-mediated retrograde transport along microtubules. Glued has the CAP-Gly domain that is common in several + TIPs and regulates initiation of retrograde transport in neuronal cells 36,37 . Interestingly, mammalian TBCB, another member of tubulin-folding cofactor family, also has the CAP-Gly domain and interacts with p150 Glued, 38 . + TIPs are a structurally and functionally diverse group of proteins that are distinguished by their specific accumulation at microtubule plus ends, or growing ends 39,40 . + TIPs control different aspects of microtubule dynamics and form links between microtubule ends and other cellular structures 2 . Some of + TIPs also participate in microtubule-actin crosstalk, such as CLIP-170-formin interaction and EB1-RhoGEF2 interaction 2 . There is accumulating evidence that Strip and also other STRIPAK components (Mst3, Mst4, and Ccm3) regulate the actin network 9,13 . For example, Strip1 and 2, homologs of Drosophila Strip, were recently identified as regulators of the actomyosin contraction that regulates cell comets (normal), intermediate (weak), fewer comets/irregular EB1 accumulation (severe). ***P < 0.0001, chi-square test. Scale bar is 7.5 μ m. (e) S2 cells were treated with control or strip dsRNA for 8 days and immunostained with anti-acetylated-α-tubulin (green) and anti-α-tubulin (magenta) antibodies. Scale bar is 7.5 μ m. (f) Immunoblots of lysate of S2 cells treated with control or strip dsRNA for 8 days. Since Strip seems to slightly affect the total α-tubulin level, Histone H3 is used as a loading control. (g) Densitometric analysis of acetylated-α-tubulin. The level of acetylated-α-tubulin is normalized to α-tubulin. This reduction of the level of acetylated-α-tubulin in strip dsRNA-treated cells was confirmed for 10 times.  41 . Taken together, Strip and STRIPAK seem to form a giant complex with TBCD and Glued at growing ends of microtubule to stabilize them and serve as a linker between microtubules and actin networks to regulate proper neurite branching and elongation.
strip knockdown resulted in a decrease in the level of acetylated α-tubulin. Therefore, we examined the possibility that Strip directly affects the acetylated tubulin level by regulating acetylation or deacetylation enzymes. We investigated the genetic interaction between strip and HDAC6, one of the enzymes responsible for the deacetylation of α-tubulin 42 , however, the shorter axon phenotype of strip knockdown was not suppressed when combined with a HDAC6 mutation (data not shown). Thus, Strip does not seem to directly affect the acetylation of α-tubulin, but affect the stabilization of microtubule by interacting with TBCD, Glued, and other molecules. Although controversial, some reports indicate that acetylation is the result, and not the cause, of stabilization 19,43,44 .
Many diseases are linked to microtubules. For example, defects in retrograde transports along microtubules cause neurodegenerative diseases, such as motor neuropathy 7B 45 and Perry syndrome 46 . Furthermore, Dscam is implicated in the cognitive disabilities in Down syndrome. Moreover, STRIPAK complexes have been linked to a number of clinical conditions and diseases 9 . Cancer genome sequencing also identified frequent mutations in human STRIP2, and based on the mutation frequency and types, STRIP2 was classified as an oncogene 47 . Thus, further investigation of Strip, STRIPAK, and the microtubule complex would yield new insights into the mechanisms of various diseases.

Methods
Fly strains. Flies were maintained under standard laboratory condition (25 °C). UAS-Dscam17.  was a kind gift from Tzumin Lee. We previously generated strip dogi 14 . UAS-TBCD used in previous study 16 is different from the one we used in this report. They are independent transgenic lines generated from the same construct 16 . The former one is inserted in third chromosome, and the one we use in this study is on second chromosome (UAS-TBCD weak ). To generate UAS-shRNA-TBCD #1 , we used the protocol previously described 48 . The target sequence of the shRNA-TBCD #1 is 5′ -GGAGCTGAATGAACTAATAATA -3′ , which is different from the shRNA we used in the previous study 16 . The fragment was subcloned into the pUAST-attB vector. Transgenic flies were raised by BestGene.
Clonal analysis. We used the MARCM method 49 . Briefly, we crossed MARCM-ready flies that contain FLP recombinase, an FRT site, GAL4, tubulin 1α promoter-GAL80, and UAS-mCD8GFP to a line containing the corresponding FRT and mutation of interest for MARCM analysis. To analyze DL1 PNs, we used the GH146-Gal4 driver and heat-shocked flies (37 °C, 1 h) at 0-24 h after larval hatching (ALH). The anterodorsal PN single-cell clones generated during this time period innervated the DL1 glomerulus. The penetrance of overbranching phenotype of strip dogi PNs was considerably reduced in this study compared to the previous report 14 . This was because of the difference in the heat-shock timing. In the previous study 14 , we heat-shocked flies 0-30 h ALH for the phenotype analysis of strip dogi PNs. To analyze mushroom body α/β neurons, we used the OK107-Gal4 driver and heat-shocked flies (37 °C, 1h) at 0-24 h after puparium formation. Dissections were performed on both sexes of adults aged 1-10 days.
Immunohistochemistry. The fixation, immunostaining, and imaging of the fly brains were carried out as previously described 50 . Briefly, the brains were dissected in 0.3% Triton X-100 (vol/vol) in phosphate-buffer (PBT), followed by fixing with 4% paraformaldehyde (wt/vol) in PBT at room temperature (RT) for 20 min, and subsequently blocked in PBT with 5% normal goat serum (vol/vol) for 1 h at RT. Primary and secondary antibody incubations were carried out in PBT for overnight at 4 °C. endogenous TBCD. (d) Immunoprecipitation of S2 cell lysates expressing c-myc tagged TBCD (TBCD-myc). TBCD-myc was co-immunoprecipitated when endogenous Strip was precipitated by anti-Strip antibody. (e) Immunoprecipitation of S2 lysates expressing Strip. Endogenous TBCD was co-immunoprecipitated when Strip was precipitated by anti-Strip antibody. (f) Representative image of S2 cells expressing TBCDmyc. Magenta: TBCD-myc, green: endogenous Strip. (g,h) Representative images of single-cell clones of DL1 PNs expressing shRNA against TBCD (shRNA-TBCD) in wild type (g) or strip dogi (h) PNs. (i,j) Representative images of single-cell clones of DL1 PNs overexpressing TBCD in wild type (i) or strip dogi (j) PNs. The right panels in (h) and (j) show PNs exhibiting the 'extremely short axon' phenotype, i.e., the axon did not exit the antennal lobe (white circle) and did not enter the mushroom body (MB, yellow dotted circle) or the lateral horn (LH, white dotted circle). Scale bar is 25 μ m. Genotypes: (a) y w, hs-FLP122  For Drosophila S2 cells, cells were cultured on concanavalin A-coated coverslips, fixed, and immunostained as previously described 51 . Briefly, S2 cells were grown for 1 h at 26 °C on concanavalin A-coated coverslips, and were fixed with 4% paraformaldehyde (wt/vol) in 0.3% Triton X-100 (vol/vol) in phosphate-buffer saline (PBST) at RT for 10 min, and subsequently blocked in PBST with 5% normal goat serum (vol/vol) for 30 min. For investigating the colocalization of Strip and TBCD, we transfected S2 cells with Gal4-actin promoter and pUAST-TBCD-myc 16  dsRNA generation and treatment. The dsRNA design, production, and treatment were performed as follows 51 . Briefly, 700-900 bp gene-specific sequences with T7 RNA polymerase sequence at both ends were amplified by PCR, and in vitro transcription was performed using T7 RiboMAX Express RNAi System (Promega) according to the manufacturer's protocol. For strip dsRNA, templates for in vitro transcription were generated using the primers 5′ -TAATACGACTCACTATAGGGCCTGCATAAACCTGCTGCGC-3′ and 5′ -TAATACGACTCACTATAGGGCTAGAGGGCGTCCCAGTCGG-3′ . For TBCD dsRNA, templates for in vitro transcription were generated using the primers 5′ -TAATACGACTCACTATA GGGGTGGTTTACCTCTCCAACCAACGG-3′ and 5′ -TAATACGACTCACTATAGGGCTGTATGCC TGGATGTTCTCGCGG-3′ . For control dsRNA, primer sequences were used to amplify the sequence from the bacterial cloning plasmid pBluescript SK 51 . For Fig. 2, Drosophila S2 cells (1.0 × 10 6 ) were cultured in a 35-mm dish; 30 μ g control or strip dsRNA was applied every 2 days for 8 days. For Fig. 3c, Drosophila S2 cells (1.0 × 10 7 ) were cultured in a 100-mm dish; 120 μ g control or TBCD dsRNA was applied every 2 days for 6 days.
Image acquisition. For all experiments, images were obtained using TCS-SP5 confocal laser scanning confocal microscopy (Leica) with a PL APO CS 40×/1.25 Oil lens (Leica) for PN and mushroom body neuron imaging or a PL APO CS 63x/1.40 Oil lens (Leica) for imaging S2 cells. Fields of view were randomly selected and each biologically independent experiment was repeated more than two times. Fig. 2a-d, all images were blinded prior to classification to avoid experimental bias. The investigator who conducted the blind test was different from the investigator who performed the experiments. Statistical analysis. Statistical analysis was performed using Prism 5 software (GraphPad). The chi-square test was used to compare data variations between control and strip dsRNA-treated S2 cells and data variations of phenotypes of mushroom α and β lobes.