(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



PTRN-1 (CAMSAP) and NOCA-2 (NINEIN) are required for microtubule polarity in Caenorhabditis elegans dendrites [1]

['Liu He', 'Cell Biology', 'Department Of Biology', 'Faculty Of Science', 'Utrecht University', 'Utrecht', 'The Netherlands', 'Lotte Van Beem', 'Berend Snel', 'Theoretical Biology']

Date: 2022-11 

Abstract The neuronal microtubule cytoskeleton is key to establish axon-dendrite polarity. Dendrites are characterized by the presence of minus-end out microtubules. However, the mechanisms that organize these microtubules with the correct orientation are still poorly understood. Using Caenorhabditis elegans as a model system for microtubule organization, we characterized the role of 2 microtubule minus-end related proteins in this process, the microtubule minus-end stabilizing protein calmodulin-regulated spectrin-associated protein (CAMSAP/PTRN-1), and the NINEIN homologue, NOCA-2 (noncentrosomal microtubule array). We found that CAMSAP and NINEIN function in parallel to mediate microtubule organization in dendrites. During dendrite outgrowth, RAB-11-positive vesicles localized to the dendrite tip to nucleate microtubules and function as a microtubule organizing center (MTOC). In the absence of either CAMSAP or NINEIN, we observed a low penetrance MTOC vesicles mislocalization to the cell body, and a nearly fully penetrant phenotype in double mutant animals. This suggests that both proteins are important for localizing the MTOC vesicles to the growing dendrite tip to organize microtubules minus-end out. Whereas NINEIN localizes to the MTOC vesicles where it is important for the recruitment of the microtubule nucleator γ-tubulin, CAMSAP localizes around the MTOC vesicles and is cotranslocated forward with the MTOC vesicles upon dendritic growth. Together, these results indicate that microtubule nucleation from the MTOC vesicles and microtubule stabilization are both important to localize the MTOC vesicles distally to organize dendritic microtubules minus-end out.

Citation: He L, van Beem L, Snel B, Hoogenraad CC, Harterink M (2022) PTRN-1 (CAMSAP) and NOCA-2 (NINEIN) are required for microtubule polarity in Caenorhabditis elegans dendrites. PLoS Biol 20(11): e3001855. https://doi.org/10.1371/journal.pbio.3001855 Academic Editor: Bing Ye, University of Michigan, UNITED STATES Received: February 7, 2022; Accepted: September 27, 2022; Published: November 17, 2022 Copyright: © 2022 He et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant data are within the paper and its Supporting Information files. Funding: This work was funded by the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO) (NWO-ALW-VICI 865.10.010 to C.C.H.), by the European Research Council (ERC Consolidator Grant 617050 to C.C.H.) and by the Chinese Scholarship Council (CSC) to L.H. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Abbreviations: BSA, bovine serum albumin; CAMSAP, calmodulin-regulated spectrin-associated protein; MAP, microtubule-associated protein; MTOC, microtubule-organizing center

Introduction The microtubule cytoskeleton is vital in neurons for proper axonal and dendritic development. In axons, microtubules are mainly arranged with their plus-ends distal to the cell body. In contrast, dendritic microtubules in invertebrates are predominantly arranged with their minus-ends distal to the cell body or have a mixed orientation in vertebrates [1–3]. This difference in microtubule organization allows for selective cargo transport into axons or dendrites [4,5]. Defects in this organization can lead to problems in protein trafficking, neuronal development, and function [6–8]. Although the importance of the microtubule cytoskeleton organization is apparent, the molecular mechanism controlling differential microtubule organization between axons and dendrites is still not fully clear. During cell division, the centrosome is the main microtubule-organizing center (MTOC). However, in polarized cells such as neurons, most microtubules are organized in a noncentrosomal manner [9–11]. Several mechanisms have been proposed to organize the neuronal microtubule cytoskeleton. These include the transport of microtubules into the correct organization (also referred to as microtubule sliding), microtubule growth from the cell body into the axon, the local nucleation of microtubules in the axon, and the selective stabilization of correctly organized microtubules by microtubule-associated proteins (MAPs) [5,12–14]. The formation of the axon is typically regarded as the initial step in neuronal polarization. Stabilization of axonal microtubules is a critical early event to form the axon. Indeed, the artificial stabilization of microtubules using drugs leads to the induction of multiple axons [15,16], and the formation of axons in cultured hippocampal neurons requires the TRIM46 protein that stabilizes and bundles the microtubules in a parallel plus-end out fashion in the proximal axon [17–19]. This plus-end out microtubule organization can be propagated into the growing axon by several mechanisms: by the outgrowth of existing microtubules potentially followed by severing [5,20], by forward translocation of the microtubule bundle [21], and by the local nucleation of new microtubules along the lattice of preexisting microtubules using the Augmin complex [22,23]. The transport of small microtubule fragments from the cell body has also been observed, but so far, it is not known if this contributes to the propagation of the axonal cytoskeleton [24]. While much is known about axonal microtubule organization, how dendrites acquire and maintain their typical minus-end out oriented microtubules is less clear. In mammals, Augmin-mediated microtubule nucleation and various MAPs were shown to contribute to the mixed microtubule organization [5,22,23,25]. In invertebrate neurons, the uniform minus-end out microtubule organization may offer a simpler starting point to understand the origin of the minus-end out microtubules. Early work suggested a role for microtubule sliding; in Drosophila, microtubules can be slid by the motor protein kinesin-1 during the earliest phases of neuron development [26], and in C. elegans, a mutant for kinesin-1 loses minus-end out microtubules in dendrites [27]. Recently, the role of local microtubule nucleation has gained attention, since noncentrosomal MTOCs are found localized in the dendrites and may be important to locally nucleate the minus-end out microtubules [14,28–30]. For example, Golgi outposts and early endosomes have been shown to nucleate microtubules in Drosophila dendrites, although the contribution to the overall microtubule organization is disputed [29,31–34]. In C. elegans, the relation between local microtubule nucleation and minus-end out microtubule organization in dendrites is clearer. It was shown that RAB-11-positive endosomes localize to the dendrite growth cone to nucleate microtubules with minus-end out organization [28]. However, it is unclear how the microtubule nucleating γ-tubulin is recruited to the RAB-11 vesicles and how specifically the minus-end out microtubules are maintained in dendrites. A prominent protein family regulating microtubule stabilization is the CAMSAP family: CAMSAP1–CAMSAP3 (in vertebrates), Patronin (in Drosophila), and PTRN-1 (in C. elegans). These proteins can bind microtubule minus-ends and thereby protect them against depolymerization [35–38]. Indeed, in cultured mammalian neurons, microtubule stabilization by CAMSAP proteins was critical for neuronal polarization [25,39,40], and also in Drosophila, Patronin was found to be important for dendritic microtubule polarity [41,42]. In addition to CAMSAP, microtubules can be stabilized by MAPs that can crosslink microtubules together or connect them to cortical structures via Ankyrin or Spectraplakin proteins [19,43]. In C. elegans, we found that the UNC-33(CRMP)/UNC-119/UNC-44(Ankyrin) complex connects the microtubule cytoskeleton to the cortex in both axons and dendrites to maintain the proper polarity organization [6]. In this study, we found that PTRN-1 (CAMSAP) is important in C. elegans for the proper localization of the MTOCs vesicles and, therefore, minus-end out microtubule polarity in the growing dendrite. Moreover, we found that the Ninein homologue NOCA-2 acts in parallel with PTRN-1 to localize γ-tubulin to the MTOC vesicles. Our results suggest that microtubule nucleation from the MTOC vesicles acts together with microtubule stabilization by CAMSAP proteins to organize dendritic microtubules minus-end out.

Discussion The presence of minus-end out microtubules in dendrites is one of the characteristic features that distinguish them from axons, allowing for selective cargo transport to set up neuronal polarity. It was recently shown in C. elegans that microtubule nucleation from MTOC vesicles in the distal dendrite is essential to organize dendritic microtubules minus-end out [28]. However, how these vesicles localize to the growing dendrite tip is incompletely understood, especially since these MTOC vesicles were suggested to be transported over the same microtubules as that they nucleate. This may suggest that these processes are connected and suggests a prominent role for selective microtubule stabilization. Here we report that the localization of these MTOC vesicles depends on the microtubule minus-end stabilizing protein PTRN-1 (CAMSAP), which functions in parallel to NOCA-2 (NINEIN) to set up dendritic minus-end out microtubules. Codepletion of PTRN-1 with NOCA-2 leads to MTOC vesicles mislocalization to the cell body, which is the underlying cause of the microtubule polarity defects. Since these proteins function (partially) redundantly, this suggests that they act in different processes. Indeed, mutants for these genes have a different effect on microtubule properties in the distal dendrite and have a different localization pattern. PTRN-1 (CAMSAP) shows a punctate pattern throughout the dendrite, as expected for a protein that binds to microtubule minus-ends. We did see a clear enrichment of PTRN-1 puncta surrounding the MTOC vesicles, suggesting that there is a higher microtubule density in the distal segment (Fig 4I). NOCA-2 (NINEIN), on the other hand, localizes to the MTOC vesicles where it perfectly overlaps with the microtubule nucleator γ-tubulin (Fig 4E and 4F). Moreover, in the noca-2 mutant, we observed that γ-tubulin is not efficiently recruited to the MTOC vesicles (Figs 3G, S6B, and S6C), suggesting that NOCA-2 is involved in the recruitment of γ-tubulin for proper microtubule nucleation. However how NOCA-2 localizes to the vesicles and how it may recruit γ-tubulin is still an open question. Although the MTOC vesicles are RAB-11 positive, many RAB-11-positive endosomes localize to the cell body without obvious NOCA-2 accumulation (Fig 4E). Moreover RAB-11 depletion had only mild defects on GIP-2 localization [28]; therefore, it seems unlikely that RAB-11 directly recruits NOCA-2 and suggests that other proteins are involved. The endosomal recruitment of NOCA-2 may be aided by palmitoylation, as was previously shown for NOCA-1 [48]; http://lipid.biocuckoo.org/ predicts a high threshold palmitoylation site at C441 in NOCA-2. Alternatively, as human NINEIN can act as a dynein activator [55], NOCA-2 might act with dynein to localize to the MTOC vesicles [28]. However, we did not observe defects in MTOC vesicle clustering in the noca-2 mutant as was reported for the dynein mutant [28], nor did we see obvious changes in dynein recruitment to the vesicles (S8D Fig). In mammals, CAMSAP1, CAMSAP2, and CAMSAP3 all recognize and protect microtubules minus-ends against depolymerization and are important for neuron polarization [25,35,39,40]. However, their behavior at the minus-end is different: CAMSAP1 concentrates at the outermost ends and tracks the growing microtubule minus-ends, while CAMSAP2 and CAMSAP3 are stably deposited on the microtubule lattice, forming stretches from the minus-end and stabilizing MT lattices against depolymerization [35,59]. In Drosophila neurons, Patronin (CAMSAP) tracks and controls the microtubule minus-end, similar to CAMSAP1, and is important to populate dendrites with minus-end out microtubules [41]. In C. elegans PLM neurons, microtubule minus-ends growth was also reported in the posterior process [57]. Here, we found that most of GFP::PTRN-1 (CAMSAP) puncta were highly immobile in mature PVD neurons (S9D Fig). However, we did observe a small population of anterograde moving PTRN-1 puncta in the shaft of the anterior dendrite (S9D Fig). The speed of these movements was slower than the typical plus-end growth speeds (S9E Fig) and may represent microtubule minus-end growth, although these did not overlap with EBP-2::GFP as was the case in Drosophila (S9C Fig) [41]. Therefore PTRN-1 may contribute to the microtubule organization by stabilizing the microtubules in the dendrite shaft and potentially promoting minus-end growth to allow for the MTOC vesicle transport to the tip. In addition, we observed an enrichment of PTRN-1 puncta surrounding the MTOC vesicles that displayed very different dynamics, much slower and less processive (Fig 4I). Therefore, we do not expect the distal CAMSAP dynamics to represent microtubule growth. One potential model is that PTRN-1 (CAMSAP) stabilizes a specific subset of distal microtubules that are then used for forward translocation of the MTOC vesicles, e.g., by the UNC-116 (kinesin-1) motor over the short plus-end out microtubules in the tip [28]. However, we found that PTRN-1 localizes in a punctate pattern surrounding the MTOC vesicles and comigrate with the vesicles upon dendrite growth. This may suggest that microtubule nucleation is connected to microtubule minus-end stabilization and suggests to also consider alternative models where, e.g., pushing or pulling on microtubules may translocate the MTOC vesicles forward. Such forces could, for example, be generated by microtubule sliding over other microtubules. In Drosophila, kinesin-1 was shown to slide microtubule against other microtubules during early neuronal development, using an extra microtubule binding site in the tail [26,60,61]. Although kinesin-1 is also essential in C. elegans to organize microtubules minus-end out [27], mutating the extra microtubule binding site in the motor tail using CRISPR did no show microtubule defects arguing against a microtubule-microtubule sliding model for kinesin-1 in C. elegans (11 out of 11 animals had fully minus-end out microtubule organization). Alternatively, motors could push or pull the microtubule cytoskeleton forward if anchored to static structures. For example, in axons, the dynein motor was shown to push the distal microtubule cytoskeleton forward by anchoring to the cortex and walking to the minus-end of the microtubules [21]. Similarly, kinesin-1 may push minus-end out–oriented microtubules towards the dendrite tip by walking to the plus-end. Or alternatively, dynein may function at the growing dendrite tip to pull on the short plus-end out microtubules that emanate from the MTOC vesicles, similar to its role to position the centrosomes during cell division [62–66]. Interestingly, the forward movements of the MTOC vesicles coincided with longer lived microtubules [28], which could represent cortically captured microtubules by dynein at the tip. More work is needed to determine the mechanism how the MTOC vesicles are transported anterogradely. Also, how the MTOC vesicles are connected to PTRN-1 stabilized microtubules will be interesting to investigate further. Potentially, NOCA-2 (NINEIN) can directly connect microtubules to the MTOC vesicles as Drosophila NINEIN (Bsg25D) was reported to bind to microtubules [67]. We propose that the minus-end out microtubule organization in the PVD dendrite follows a two-step model where non-centrosomal microtubules are initially nucleated from MTOC localized γ-tubulin and subsequently stabilized by PTRN-1 (CAMSAP). The minus-end out microtubule population may be further stabilized by cortical anchoring [6] and potentially further amplified by severing proteins such as Katanin and Spastin [68]. Such a mechanism may also take place in other tissues such as the C. elegans epidermis, where PTRN-1 and NOCA-1 (NINEIN) were found to also act in parallel to organize the noncentrosomal microtubules [48]. In contrast, in the Drosophila fat body, both Patronin (CAMSAP) and NINEIN act independent of γ-Tubulin to assemble noncentrosomal microtubules [69]. This indicates that the functional connection between microtubules nucleation and stabilization may vary between cell types and organisms.

Materials and methods C. elegans strains and culturing Strains were cultured at 15°C or 20°C using OP50 Escherichia coli as a food source and imaged at room temperature. To image early developing PVD neurons, the adult animals were grown at 15°C for at least 48 h, and L2-L3 stage progeny were picked for imaging. The noca-2(hrt28) allele used in this study was made using CRISPR/Cas9-mediated mutagenesis, which deletes the entire NOCA-2 fragment (S1C Fig). The kinesin-1 allele with the mutated microtubule binding site in the tail (aa761RKKYQQ->AAAYAA) [27] was ordered from SunyBiotech and is called PHX1768 unc-116(syb1768). noca-2(hrt31[GFP]) was obtained by using CRISPR/Cas9-based genome editing [70]. The strains TV21539[tba-1(ok1135) I; wyEx8784[Punc-86::gfp::tba-1; Punc-86:mCherry::PLCdeltaPH] and TV25056[wyEx9975[Punc-86::gfp::rab-11.1 cDNA; Punc-86::mCherry::PLCdeltaPH] were gifts from Dr. Kang Shen [28]. The strain gip-1(wow25[tagRFP-t::3xMyc::gip-1]) III was a gift from Dr. Jessica L. Feldman [71]. The strain ntuIs6[Pdes-2::unc-116(G237A)::mCherry; Pdes-2::unc-104(E250K)::gfp; Podr-1::GFP]III was a gift from Dr. Chan-Yen Ou [58]. The plasmid Pdes-2::unc-116(G237A)::gfp was cloned in Pdes-2::UNC-116(G237A)::mCherry (pCH9) that was provided by Dr. Chan-Yen Ou [58]. All trains used in this study are listed in S1 Table. DNA plasmids and gRNA The DNA plasmids that were used to generate transgenic C. elegans strains and the detailed cloning information are listed in S1 Table. The gRNA used to make noca-2 mutant and GFP knock-in stains are listed in S1 Table. Pmyo2::mcherry (5 ng/μl) was used as coinjection marker to generate extrachromosomal strains. Microscopy For all imaging, C. elegans were mounted on 5% agarose pads with 10 mM tetramisole or 5 mM Levamisole solution in M9 buffer. Live imaging was performed within 60 min after mounting on a Nikon Eclipse-Ti microscope and with a Plan Apo VC, 60×, 1.40 NA oil or a Plan Apo VC 100 × N.A. 1.40 oil objectives (Nikon). The microscope is equipped with a motorized stage (ASI; PZ-2000), a Perfect Focus System (Nikon), ILas system (Roper Scientific France/PICT-IBiSA, Curie Institute) and uses MetaMorph 7.8.0.0 software (Molecular Devices) to control the camera and all motorized parts. Confocal excitation and detection are achieved using 100-mW Vortran Stradus 405 nm, 100-mW Cobolt Calypso 491 nm, and 100-mW Cobolt Jive 561 nm lasers and a Yokogawa spinning disk confocal scanning unit (CSU-X1-A1N-E; Roper Scientific) equipped with a triple-band dichroic mirror (z405/488/568trans-pc; Chroma) and a filter wheel (CSUX1-FW-06P-01; Roper Scientific) containing BFP (ET-DAPI (49000), GFP (ET-GFP (49002)) and mCherry (ET-mCherry(49008)) emission filters (all Chroma). Confocal images were acquired with a QuantEM:512 SCEMCCD camera (Photometrics) at a final magnification of 110 nm (60× objective), 67 nm (100× objective) per pixel, including the additional 2.0× magnification introduced by an additional lens mounted between scanning unit and camera (Edmund Optics). All EBP-2::GFP and TBA-1::GFP imaging was performed at 1 frame per second (fps) For the analysis of PVD neuron morphology, images were acquired using an LSM700 (Zeiss) confocal with a 20× NA 0.8 dry objective using the 488 nm and 555 nm laser lines. Quantitative image analysis Image processing and analysis was done using ImageJ (FIJI) to create kymographs or merged sum intensity images for the intensity profile measurement. Statistical analysis and graphs were made in GraphPad Prism software version 8.0. To quantify EBP-2::GFP growth orientation, kymographs were made using Kymograph Builder plugin in ImageJ. The retrogradely and the anterogradely growing EBP-2::GFP were manually counted. To quantify the microtubule polarity in the mature PVD dendrite, imaging was performed in the proximal dendrite. For the developing PVD dendrite, the whole anterior PVD region was imaged. To quantify EBP-2::GFP growing speed and frequency (S4A and S4B Fig), only the distal region (20 μm) was quantified. To quantify the main site of microtubule growth (Figs 1G, 2E and S2A), kymographs were made for the whole of anterior dendrite in the developing neurons. When EBP-2 comets mainly grew from the distal anterior dendrites, the animal was classified as “distal dendrite”; when the EBP-2 comets mainly grew from cell body, theses were classified as “cell body”; and when EBP-2 comets grew in both directions, these were classified as “Mix.” To quantify the PVD branch complexity, the entire PVD dendrite was divided into 4 segments: One posterior segment (−1) and the anterior segment was divided into 3 equal length anterior segments (+1, +2, and +3) (Fig 2F). The “branch complexity” index calculation was based on a previous study definition [45]. To measure the GIP-2 cluster intensity profiles (Fig 3I), cytosolic mKate2 was used to visualize the dendrite tip. We drew a 50 px-wide line from the tip of the anterior dendrite to measure the intensity profile of the GIP-2 cluster and the background intensity was subtracted from the region next to the neuron devoid of gut granule autofluorescence. Sequence analysis and phylogenetic tree To investigate the relation between NOCA-1, NOCA-2, and human NINEIN (UniProt #Q9Y2I6), we performed BLAST searches using HHpred (https://toolkit.tuebingen.mpg.de/tools/hhpred) [72,73]. To identify domain structures, we made use of UniProt annotations supplemented with the Aphafold2 predicted 3D structure [74]. To generate the phylogenetic tree, UniProt (https://www.uniprot.org/) was used to search for the NINEIN homologous sequences from different species. The phylogenetic tree was generated using the full length proteins and the online version of MAFFT (http://mafft.cbrc.jp/alignment/server/) [75]. Pull-down and western blot HEK293T cells (authenticated and tested negative for mycoplasma) were transfected with NOCA-2::EGFP or cotransfected with BirA together with bio-mCherry::NOCA-2 and incubated at 37°C for 24 h. Cells were harvested and washed 1× with ice-cold PBS and lysed with lysis buffer (100 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, and 1× protease inhibitor cocktail). Cell lysates were centrifuged at 13,000 rpm for 10 min, and the supernatants were incubated with GFP-Trap magnetic beads (Chromotek) or Dynabeads M-280 (Invitrogen). Beads were preblocked in buffer containing 20 mM Tris (pH 7.5), 20% glycerol, 150 mM NaCl, and 10 μg chicken egg albumin for 30 min and then washed twice with buffer containing 20 mM Tris (pH 7.5), 150 mM NaCl, and 0.1% Triton X-100. After incubating for 1 h at 4°C, the supernatant was collected (after pull-down) and the beads were washed 5 times with washing buffer. Samples were eluted with SDS/DTT sample buffer and boiled for subsequent western blot. For western blot, samples were loaded onto 8% SDS-PAGE gels and transferred to nitrocellulose membrane. Membranes were blocked with 2% bovine serum albumin (BSA) in PBS/0.05% Tween-20. For NOCA-2::EGFP transfected cells samples, primary antibodies: anti-GFP (ab290, Abcam, rabbit) and anti-γ-Tubulin (T6557, Sigma, mouse) were diluted in blocking buffer and incubated with the membranes overnight at 4°C. For Bio-mCherry::NOCA-2 transfected cells, primary antibodies: anti-mCherry (632543, Clontech, mouse) and anti-γ-Tubulin (T3559, Sigma, rabbit) were used. After primary antibody incubation, membranes were washed 3 times with PBS/0.05% Tween 20 and incubated with secondary IRDye 680LT anti-mouse or IRDye 800LT anti-rabbit antibodies for 45 min at room temperature. Membranes were then washed 3 times with PBS/0.05% Tween 20 and scanned on Odyssey Infrared Imaging system (LI-COR Biosciences).

Acknowledgments We thank Mike Boxem and Sander van den Heuvel for advice, C. elegans reagents, and infrastructure. We thank Jason Kroll and Amélie Freal for feedback on the manuscript and Bart de Haan for helping with cloning. We thank Kang Shen for helpful suggestions and sharing of strains. We thank Jessica Feldman, Chan-Yen Ou, and Alexander Dammerman for kind sharing of C. elegans strains and reagents. Some strains were provided by the CGC, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440), and some by the National Biorescource Project. We thank WormBase for curating and making available data related to C. elegans.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001855

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/