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



Cbln1 regulates axon growth and guidance in multiple neural regions [1]

['Peng Han', 'School Of Life Sciences', 'Department Of Neuroscience', 'Department Of Biology', 'Brain Research Center', 'Shenzhen Key Laboratory Of Gene Regulation', 'Systems Biology', 'Southern University Of Science', 'Technology', 'Shenzhen']

Date: 2022-11 

The accurate construction of neural circuits requires the precise control of axon growth and guidance, which is regulated by multiple growth and guidance cues during early nervous system development. It is generally thought that the growth and guidance cues that control the major steps of axon development have been defined. Here, we describe cerebellin-1 (Cbln1) as a novel cue that controls diverse aspects of axon growth and guidance throughout the central nervous system (CNS) by experiments using mouse and chick embryos. Cbln1 has previously been shown to function in late neural development to influence synapse organization. Here, we find that Cbln1 has an essential role in early neural development. Cbln1 is expressed on the axons and growth cones of developing commissural neurons and functions in an autocrine manner to promote axon growth. Cbln1 is also expressed in intermediate target tissues and functions as an attractive guidance cue. We find that these functions of Cbln1 are mediated by neurexin-2 (Nrxn2), which functions as the Cbln1 receptor for axon growth and guidance. In addition to the developing spinal cord, we further show that Cbln1 functions in diverse parts of the CNS with major roles in cerebellar parallel fiber growth and retinal ganglion cell axon guidance. Despite the prevailing role of Cbln1 as a synaptic organizer, our study discovers a new and unexpected function for Cbln1 as a general axon growth and guidance cue throughout the nervous system.

Funding: This work was supported by National Natural Science Foundation of China ( https://www.nsfc.gov.cn/ ) (31871038 and 32170955 to S.-J.J.), Shenzhen-Hong Kong Institute of Brain Science-Shenzhen Fundamental Research Institutions ( http://bcbdi.siat.ac.cn/ ) (2022SHIBS0002), High-Level University Construction Fund for Department of Biology ( https://bio.sustech.edu.cn/ ) (internal grant no. G02226301), Science and Technology Innovation Commission of Shenzhen Municipal Government ( http://stic.sz.gov.cn/ ) (ZDSYS20200811144002008), and NIH ( https://www.nih.gov/ ) (R35NS111631 to S.R.J.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Here, we found that Cbln1 is expressed both in the dorsal commissural neurons (DCN) and in the floor plate (FP) of the embryonic mouse spinal cord. We generated DCN- and FP-specific Cbln1 conditional knockout (cKO) mice that demonstrated that the cell-autonomous and non-cell-autonomous Cbln1 from DCNs and FP regulate commissural axon (CA) growth and guidance, respectively. The dual roles of Cbln1 are mediated by its receptor, neurexin-2. Interestingly, the functions and mechanisms of Cbln1 in regulating axon growth and guidance were replicated in the developing cerebellar granule axon growth and the embryonic retinal ganglion cell axon guidance, respectively. Together, our findings reveal a general role for Cbln1 in regulating axon growth and guidance during early nervous system development prior to synapse formation.

The commissures in the rodent spinal cord are one of the most prominent model systems to study axon growth and guidance. In a search for the differentially expressed genes in the dorsal spinal cord of mouse embryos, we identified a gene encoding the secreted protein cerebellin-1 (Cbln1). Cbln1 is released from cerebellar parallel fibers and has previously been characterized as a synaptic organizer by forming the synapse-spanning tripartite complex Nrxn-Cbln1-GluD2 (Nrxn, neurexin; GluD2, the ionotropic glutamate receptor family member delta-2) [ 19 , 20 ]. However, whether Cbln1 is expressed and plays roles in earlier nervous system development is unknown.

The precise control of axon pathfinding is critical for the correct neural wiring during nervous system development. The stimulation of axon growth and regulation of axon guidance have been shown to require adhesion molecules, diffusible signals, and morphogens such as Netrins [ 1 – 7 ], Slits [ 8 – 10 ], Ephrins [ 11 ], Semaphorins [ 10 , 12 ], Draxin [ 13 ], Shh [ 14 ], Wnts [ 15 ], and BMPs [ 16 , 17 ]. These axon guidance molecules bind to their receptors in the axon growth cones to activate various signaling pathways that eventually change the cytoskeleton [ 18 ]. The lack of newly identified cues in the past decade has suggested that the major classes of growth and guidance cues have now been identified.

Results

Cell-autonomous Cbln1 in the dorsal commissural neurons is both required and sufficient to stimulate commissural axon growth in vivo Next, we wanted to explore the roles of Cbln1 expressed in DCNs and FP, separately. In order to specifically ablate Cbln1 from these tissues, we generated cKO of Cbln1 using tissue-specific Cre lines (Fig 2A). We used Wnt1-Cre line to specifically ablate Cbln1 from spinal DCNs, without affecting Cbln1 expression in other parts of spinal cord (Fig 2B). Cbln1 cKO in spinal DCNs does not disturb neurogenesis of these neurons, as indicated by normal numbers, distribution and patterning of dl1-4 neurons in the developing spinal cord (S2A–S2H Fig). We continued to check CA growth in DCN-specific Cbln1 cKO. We prepared open-books of developing spinal cords and used Robo3 immunostaining to label CAs. Robo3 selectively marks CAs as they navigate to and across the FP [27]. As shown in Fig 2C–2E, both lengths and numbers of CAs were decreased in Cbln1 cKO embryos compared with their littermate controls. These data suggest that Cbln1 in the DCNs is required for their own commissural axon growth. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Cell-autonomous Cbln1 is both required and sufficient to stimulate commissural axon growth in vivo. (A) Schematic drawings showing the generation of Cbln1 cKO. The coding sequence of Cbln1 is deleted after Cre-mediated recombination. (B) Specific depletion of Cbln1 in the dorsal spinal cord of Wnt1-Cre+/-,Cbln1fl/fl cKO mouse embryos. In situ hybridization of E11.5 spinal cord sections using RNA probes against Cbln1 confirmed specific ablation of Cbln1 from the DCNs. Black arrowheads indicate Cbln1 expression in control DCNs and white arrowheads highlight the missing Cbln1 expression in cKO DCNs. Expression of Cbln1 in FP (in red dotted circles) and other parts are not affected in DCN-specific Cbln1 cKO spinal cords. Scale bar, 100 μm. (C) DCN-specific Cbln1 cKO caused dramatic CA growth defects in vivo. Commissural axons were marked by Robo3 immunostaining in spinal cord open-books at E10.5. The lengths of CAs are much shorter and the numbers of CAs are much fewer in Cbln1 cKO spinal cords compared with their littermate controls. Notice that these differences are more obvious in posterior ends of spinal cords. The dotted boxed areas were shown in the higher magnification insets (C’ and C”). A, anterior; P, posterior. Scale bar, 200 μm. (D and E) Quantification of commissural axon lengths and numbers in (C). The spinal cords were divided to bins (500 μm) along the anterior-posterior (A→P) direction, and the lengths (D) and numbers (E) of CAs in each bin were quantified. All data are mean ± SEM: Cbln1fl/fl (n = 9 embryos) vs. Wnt1-Cre+/-,Cbln1fl/fl (n = 12 embryos); *p = 0.014 for Bin 6 in D; *p = 0.015 for Bin 3 in E; *p = 0.049 for Bin 4 in E; *p = 0.041 for Bin 5 in E; by multiple t tests. (F) In situ hybridization of cCbln1 in spinal cord cross-sections of St.23-24 chick embryos. Arrowheads indicate the expression of cCbln1 in DCNs. Scale bar, 50 μm. (G) Unilateral DCN-specific overexpression of cCbln1 by in ovo electroporation of pMath1-eGFP-IRES-cCbln1 enhanced commissural axon growth in chick neural tubes. Lhx9 marks dI1 DCNs and eGFP marks electroporated DCNs and their axons. The arrows point commissural axon terminals. Shown are the representative images from 10 chick embryos with pMath1-eGFP-IRES-cCbln1 and 8 embryos with control plasmid. Scale bar, 50 μm. (H and I) Quantification of commissural axon length and Lhx9+ neuron numbers in (G). All data are represented as box and whisker plots: for H, pMath1-eGFP-IRES-cCbln1 (n = 35 sections) vs. pMath1-eGFP (n = 31 sections), ****p = 5.56 × 10−12; for I, pMath1-eGFP-IRES-cCbln1 (n = 37 sections) vs. pMath1-eGFP (n = 29 sections), p = 0.32, ns, not significant; by unpaired Student t test. The data underlying all the graphs shown in the figure are included in S1 Data. CA, commissural axon; Cbln1, cerebellin 1; Cre, Cre recombinase; DCNs, dorsal commissural neurons; cKO, conditional knockout; A→P, anterior to posterior; SEM, standard error of the mean; St., stage (Hamburger-Hamilton staging for chick development). https://doi.org/10.1371/journal.pbio.3001853.g002 To further test whether Cbln1 is sufficient to stimulate CA growth in vivo, we used a model of chick neural tube. Chick Cbln1 (cCbln1) is expressed in the DCNs of developing chick neural tube, as is the case with mouse Cbln1, but is not detected in the FP of chick embryonic spinal cord (Fig 2F). We made a DCN-specific overexpression plasmid, pMath1-eGFP-IRES-MCS, by modifying a DCN-specific knockdown plasmid pMath1-eGFP-miRNA [28]. Unilateral DCN-specific overexpression of cCbln1 by in ovo electroporation of pMath1-eGFP-IRES-cCbln1 enhanced chick CA growth compared with control plasmid without changing commissural neuron numbers (Fig 2G–2I). These data suggest that Cbln1 is sufficient to stimulate CA growth. In the DCN-specific Cbln1 cKO embryonic spinal cord, other cell-autonomous growth promoting molecules or non-cell-autonomous guidance attractants/repellents are not affected. So we continued to check whether the CA growth could eventually reach the FP. As shown in S2I Fig, almost all axons reached the floor plate at E12, and there was no difference between Cbln1 cKO and control for commissural axon length or number at this stage (S2J and S2K Fig). In an open book DiI assay performed at E12, the crossing or post-crossing behaviors of CAs in DCN-specific Cbln1 cKO embryos were also similar to those of controls (S2L Fig).

Cbln1 is secreted from commissural axon growth cones and stimulates commissural axon growth in an autocrine manner Next, we continued to elucidate the mechanisms for the cell-autonomous functions of Cbln1. We hypothesized that Cbln1 was secreted from the DCNs and then acted to stimulate CA growth in an autocrine manner. To test this, we cultured DCN explants from E10.5 (a stage when most CAs have not projected to the midline yet and are called pre-crossing axons) mouse spinal cords and used Tag1 immunostaining to visualize pre-crossing commissural axons. Tag1 has been widely used as a marker for pre-crossing CAs [29,30]. Compared with control embryonic DCN explants, the CA growth of Wnt1-Cre-mediated Cbln1 cKO DCNs was significantly inhibited, indicated by decreased axon numbers and reduced axon lengths (Fig 3A–3C), which is consistent with in vivo results for DCN-specific Cbln1 cKO (Fig 2B–2E). These axon growth defects were efficiently rescued by adding a recombinant human Cbln1 protein (rhCbln1) to the cultures (Fig 3A–3C). These data suggest that the cell-autonomous Cbln1 regulates CA growth in an autocrine manner. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Cbln1 is secreted from commissural axon growth cones and stimulates commissural axon growth in an autocrine manner. (A) Extrinsic Cbln1 could rescue CA growth defects caused by cell-autonomous ablation of Cbln1 in the DCNs. DCN explants dissected from E10.5 mouse embryos were cultured in vitro and CA length was monitored by immunostaining of Tag1, a CA marker. Compared with Cbln1fl/fl, DCN explants of Wnt1-Cre+/-,Cbln1fl/fl embryos showed significant CA growth defects. These defects were rescued by adding the recombinant human Cbln1 protein (rhCbln1) to the cultures. Scale bar, 200 μm. (B and C) Quantification of Tag1+ commissural axon numbers and lengths in (A). Data are represented as box and whisker plots. For B, ****p = 1.69 × 10−6, Cbln1fl/fl (n = 14 explants) vs. Wnt1-Cre+/-,Cbln1fl/fl (n = 14 explants); ****p = 6.23 × 10−7, Wnt1-Cre+/-,Cbln1fl/fl vs. Wnt1-Cre+/-,Cbln1fl/fl + rhCbln1 (n = 16 explants); ns, not significant (p = 0.99), Cbln1fl/fl vs. Wnt1-Cre+/-,Cbln1fl/fl + rhCbln1. For C, ****p = 6.61 × 10−11, Cbln1fl/fl (n = 876 axons) vs. Wnt1-Cre+/-,Cbln1fl/fl (n = 274 axons); ****p = 6.61 × 10−11, Wnt1-Cre+/-,Cbln1fl/fl vs. Wnt1-Cre+/-,Cbln1fl/fl + rhCbln1 (n = 1,013 axons); ns, not significant (p = 0.91), Cbln1fl/fl vs. Wnt1-Cre+/-,Cbln1fl/fl + rhCbln1. By 1-way ANOVA followed by Tukey’s multiple comparison test. (D) Robust Cbln1 IF signals were detected in CAs and growth cones. Dissociated DCN neurons from E11 mouse embryos were cultured in vitro and Cbln1 IF signals were imaged after lentiviral shRNA infection. Loss of Cbln1 IF signals after shCbln1 infection indicated the specificity of Cbln1 IF signals in CAs and growth cones. Scale bar, 10 μm. (E) Quantification of axonal Cbln1 IF signals in (D). Data are represented as box and whisker plots: shCtrl (n = 55 axons) vs. shCbln1 (n = 68 axons), ****p = 5.65 × 10−31, by unpaired Student t test. (F) Cbln1 is exocytosed from CAs via lysosomes. Robust Cbln1 IF signals were detected on the CA surface of cultured DCN explants and were eliminated after blocking exocytosis with GPN treatment for 10 min. Scale bar, 50 μm. (G) Quantification of axon surface Cbln1 IF signals in (F). Data are represented as box and whisker plots: Vehicle (n = 140 axons) vs. GPN (n = 126 axons), ****p = 1.94 × 10−26, by unpaired Student t test. (H) Blocking Cbln1 exocytosis in CA with GPN for 7 h inhibited CA growth. Data are represented as box and whisker plots: Vehicle (n = 75 axons) vs. GPN (n = 51 axons), ****p = 2.15 × 10−20, by unpaired Student t test. The data underlying all the graphs shown in the figure are included in S1 Data. A.U., arbitrary unit; CA, commissural axon; Cbln1, cerebellin 1; DCN, dorsal commissural neuron; GPN, Glycyl-L-phenylalanine 2-naphthylamide; IF, immunofluorescence; rhCbln1, recombinant human Cbln1 protein; shCtrl, control shRNA; shCbln1, shRNA against Cbln1. https://doi.org/10.1371/journal.pbio.3001853.g003 We next asked whether Cbln1-induced axonal growth works locally in CAs and growth cones. Immunofluorescence of DCN neuron culture using a Cbln1 antibody detected robust Cbln1 IF signals in CAs and growth cones (Fig 3D). To confirm the specificity of these axonal Cbln1 IF signals, we generated lentiviral shCbln1 that led to dramatic knockdown of Cbln1 in cultured neurons (S3A Fig). The Cbln1 IF signals in CAs and growth cones were largely lost after knockdown of Cbln1 (Fig 3D and 3E), indicating that Cbln1 is present in CAs and growth cones. We continued to explore how Cbln1 is released from CAs. It is reported that Cbln1 co-localizes with the lysosomal enzymes cathepsin B and D in the adult mouse brain [31,32], indicating the lysosome may regulate Cbln1 secretion in CAs. To test this, we applied different lysosome inhibitors to the DCN cultures and checked their effects on Cbln1 secretion from CAs. Glycyl-L-phenylalanine 2-naphthylamide (GPN) can be specifically cleaved by cathepsin C, which leads to targeted disruption of the lysosomal membrane [32,33]. Bafilomycin A1 (BafA 1 ) blocks lysosomal functions through working as a specific inhibitor of vacuolar H+-ATPase [34]. Treatment of DCN cultures with GPN or BafA 1 , followed by an IF protocol to detect surface Cbln1 by leaving out the permeabilization steps, showed loss of Cbln1 IF signals on the CA surface (Fig 3F and 3G, S3B and S3C Fig), suggesting that Cbln1 is released from lysosomes in CAs and growth cones. Blocking Cbln1 secretion by GPN inhibited CA growth (Fig 3H), further supporting a model that Cbln1 is released from and works back on CA and growth cones to stimulate axon growth.

Non-cell-autonomous Cbln1 from the floor plate regulates commissural axon guidance The facts that the secreted Cbln1 works extrinsically and that it is expressed in the FP during CA growth to the midline suggest that Cbln1 from FP might regulate CA guidance. To test this idea, we first prepared COS7 cell lines stably expressing mouse Cbln1. High levels of Cbln1 were detected in the culture media, indicating the overexpressed Cbln1 was secreted from COS7 cells (S4A Fig). We then co-cultured the dorsal spinal cord explants from E11 mouse embryos with COS7 cell aggregates expressing Cbln1 tagged with FLAG and ZsGreen or ZsGreen alone in collagen gels (Fig 4A). The dorsal spinal cord explants growing with Cbln1-expressing COS7 cell aggregates had significantly longer axons than the control (Fig 4A and 4B). More importantly, the growth of commissural axons was attracted toward Cbln1-expressing COS7 cell aggregates, indicated by the higher axon number ratios (proximal/distal) compared with the control cell aggregates expressing ZsGreen alone (Fig 4A and 4C). These results suggest that the non-cell-autonomous Cbln1 functions as an attractive axon guidance molecule. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Non-cell-autonomous Cbln1 from the floor plate regulates commissural axon guidance. (A) Co-culture of DCN explants from E11 mouse spinal cords with COS7 cell aggregates expressing Cbln1-FLAG with ZsGreen or ZsGreen alone. Commissural axons were visualized with NFM immunostaining. Cbln1 expression attracted CA turning toward cell aggregates and also enhanced axon growth. Scale bar, 200 μm. (B and C) Quantification of CA growth and turning in (A) by measuring the axon length (B) and the axon number ratio (proximal/distal) (C). All data are represented as box and whisker plots: for B, Ctrl (n = 1,336 axons) vs. OE (n = 1,051 axons), ****p = 1.93 × 10−70; for C, Ctrl (n = 16 explants) vs. OE (n = 14 explants), ****p = 1.97 × 10−8; by unpaired Student t test. (D) Specific ablation of Cbln1 in the floor plate of Foxa2-Cre+/-,Cbln1fl/fl cKO mouse embryos was confirmed by in situ hybridization of E11.5 spinal cord sections. Expression of Cbln1 in the floor plate was completely lost in the cKO spinal cord (white asterisk) compared with the control embryos (black asterisk). Black arrowheads indicate the unchanged Cbln1 expression in DCNs of both genotypes. Scale bar, 100 μm. (E) The axon guidance defects of pre-crossing commissural axons were observed by Tag1 immunostaining in floor plate-specific Cbln1 cKO and control embryos at E11.5. Higher magnification views of the floor plate region in the white dotted boxes are also shown (bottom). The pair of yellow arrowheads denotes the thickness of the VC. The double-arrowed line measures the distance between the point of intersection (of the main pre-crossing CA bundle with the ventral edge of spinal cord) and the midline (indicated by the dotted line). Scale bars, 50 μm. (F and G) Quantification of the VC thickness and the distance from the main bundle intersection point to the midline. The VC thickness was normalized to the height (dorsal to ventral) of spinal cord. All data are represented as box and whisker plots: Cbln1fl/fl (n = 62 sections) vs. Foxa2-Cre+/-,Cbln1fl/fl (n = 60 sections), ****p = 1.69 × 10−6 for F, ****p = 1.07 × 10−7 for G, by unpaired Student t test. (H) DiI labeling of E11.5 spinal cord open-books traced commissural axon guidance behaviors during midline crossing. The region between 2 white dotted lines indicates the floor plate. The white, yellow, and purple arrowheads indicate the CAs with aberrant behaviors such as ipsilateral turning, slower growing or winding crossing, respectively. Scale bar, 50 μm. (I) Quantification of the percentages of CAs with different guidance behaviors. All data are mean ± SEM and represented as histogram: Cbln1fl/fl (n = 45 DiI injections) vs. Foxa2-Cre+/-,Cbln1fl/fl (n = 32 DiI injections), ****p = 3.36 × 10−17 for normal crossing, ****p = 8.86 × 10−10 for ipsilateral turning, ****p = 6.14 × 10−7 for slower growing, *p = 0.033 for winding crossing, by unpaired Student t test. The data underlying all the graphs shown in the figure are included in S1 Data. CA, commissural axon; Cbln1, cerebellin 1; cKO, conditional knockout; Ctrl, control; DCNs, dorsal commissural neurons; NFM, neurofilament; OE, overexpression; SEM, standard error of the mean; VC, ventral commissure. https://doi.org/10.1371/journal.pbio.3001853.g004 To assess the in vivo functions of the non-cell-autonomous Cbln1, we generated FP-specific Cbln1 cKO mice. We utilized Foxa2-CreERT line that has been used to induce Cre recombinase expression specifically in floor plate cells in response to tamoxifen (TM) treatment [35,36]. Cbln1 expression was specifically ablated from the FP in these cKO embryos, without affecting its expression in other parts of spinal cord including the DCNs (Fig 4D). The neural patterning or neurogenesis was not disturbed by ablation of Cbln1 from the FP (S4B–S4E Fig). However, examination of commissural axon trajectories using Tag1 immunostaining in E11.5 spinal cords revealed significant axon guidance defects in the midline and ventral spinal cord. First, the thickness of the ventral commissure (VC) was significantly reduced in the FP-specific Cbln1 cKO embryos compared with their littermate controls (Fig 4E and 4F). Second, the intersection of the main CA bundle with the ventral commissural funiculus was shifted laterally in the FP-specific Cbln1 cKO embryos compared with their littermate controls (Fig 4E). The distances between the point of intersection and the midline were quantified, showing a significant increase in the FP-specific Cbln1 cKO embryos (Fig 4G). These phenotypes were also evident by NFM immunostaining (S4F–S4H Fig). These axon guidance defects suggest that Cbln1 from the FP indeed works as an axon guidance cue in the developing spinal cord. To observe more clearly the CA guidance behaviors in the FP-specific Cbln1 cKO embryos, we performed DiI labeling of DCNs in the open-book spinal cords at E11.5. As shown in Fig 4H and 4I, there was a significant decrease of the number of normal crossing CAs in the Cbln1 cKO. Meanwhile, the numbers of commissural axons showing guidance defects such as ipsilateral turning, slower growing or winding crossing were significantly increased in Cbln1 cKO embryos compared with their littermate controls (Fig 4H and 4I). All these data support the idea that the non-cell-autonomous Cbln1 derived from the FP works as an axon guidance cue for CAs in the developing spinal cord.

Cell-autonomous Cbln1 from cerebellar granular cells is required for parallel fiber growth We wondered whether the function of Cbln1 to regulate axon development is a general mechanism which also works in other brain regions during development. The studies on Cbln1 so far have been focused on its functions as a synaptic organizer in cerebellum. Whether Cbln1 is expressed and exerts its functions at earlier stages of cerebellar development remains unexplored. We first checked expression of Cbln1 in earlier cerebellar development. As shown in Fig 6A, high and specific Cbln1 expression was detected in the P4-P8 cerebellar granule cells in the inner granule layer (IGL) by in situ hybridization. Immunofluorescence using a Cbln1 antibody showed that Cbln1 protein is enriched in the molecular layer (ML) of cerebellum (Fig 6B), suggesting that Cbln1 protein is expressed and secreted by cerebellar granule cell (GC) axons. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. Cell-autonomous Cbln1 is required for cerebellar granule cell axon growth. (A) In situ hybridization of Cbln1 in cerebella during P4 and P8. Cbln1 mRNA is specifically and highly expressed in granule cells, esp. in the IGL. Scale bars, 500 μm. (B) High level of Cbln1 protein is detected in the ML of P8 cerebellum, which is expressed and secreted by GC axons. Higher magnification of the boxed area is shown in (B’). Scale bars, 500 μm (B) and 100 μm (B’). (C) Ablation of Cbln1 expression in Cbln1 cKO mouse cerebella. In situ hybridization of Cbln1 in P8 cerebellum confirmed the ablation of Cbln1 from IGL. Scale bar, 500 μm. (D and E) Extrinsic Cbln1 could rescue GC axon growth defects caused by cell-autonomous ablation of Cbln1 in cerebellar GC neurons. P8 GC neurons were dissected and cultured in vitro. GC axons were imaged at 2 time points (0 and 4 h). The growth rate of GC axons from Cbln1 cKO cerebella was significantly slower than that of control. This defect was rescued by adding the recombinant human Cbln1 protein (rhCbln1) to the cultures. Quantification data are represented as box and whisker plots (E): Cbln1fl/fl (n = 114 axons) vs. Wnt1-Cre+/-,Cbln1fl/fl (n = 191 axons), ****p = 6.18 × 10−37; Wnt1-Cre+/-,Cbln1fl/fl + rhCbln1 (n = 68 axons) vs. Wnt1-Cre+/-,Cbln1fl/fl (n = 191 axons), ****p = 1.63 × 10−21; Cbln1fl/fl vs. Wnt1-Cre+/-,Cbln1fl/fl +rhCbln1, **p = 0.0098; by 1-way ANOVA followed by Tukey’s multiple comparison test. Scale bar, 20 μm (D). (F) Neurogenesis is not disturbed in the Cbln1 cKO cerebellum at P8. Immunostainings of the PC marker Calbindin and the granule cell marker NeuN showed no difference between Cbln1 cKO and control cerebella, suggesting that the neurogenesis of PCs and GCs in cerebellum is not affected. Scale bars, 500 μm. (G and H) Lengths of PFs labeled by DiI were significantly decreased in Cbln1 cKO mice at P6. The white arrowheads indicate the terminals of DiI-labeled PFs. Quantification of PF lengths is shown as box and whisker plots (H): n = 190 axons for Cbln1fl/fl mice, n = 132 axons for Wnt1-Cre+/-,Cbln1fl/fl mice; ****p = 1.09 × 10−13; by unpaired Student t test. Scale bar, 100 μm (G). The data underlying all the graphs shown in the figure are included in S1 Data. Cbln1, cerebellin 1; cKO, conditional knockout; GC, granule cell; h, hours; IGL, inner granule layer; ML, molecular layer; PC, Purkinje cell; PF, parallel fiber; rhCbln1, recombinant human Cbln1 protein. https://doi.org/10.1371/journal.pbio.3001853.g006 Next, we tested the possible roles of Cbln1 in earlier cerebellar development. We generated Cbln1 cKO in cerebellum using the Wnt1-cre line [21,42], which resulted in the efficient knockout of Cbln1 from GCs (Fig 6C). Axon growth rates of Cbln1-deficient GCs in vitro were significantly decreased compared with control neurons (Fig 6D and 6E), suggesting that the cell-autonomous Cbln1 is required for GC axon growth. Similar to Cbln1 on commissural axons, extrinsic application of the recombinant hCbln1 (rhCbln1) protein to the GC axons could efficiently rescue this axon growth defect (Fig 6D and 6E). These results support a similar model as in the developing spinal cord that Cbln1 secreted from cerebellar GC axons works back to stimulate GC axon growth in the developing cerebellum. Detection of Nrxn1, 2, and 3 expression in GCs at IGL (S6A and S6B Fig) implied that neurexins would mediate the autocrine function of Cbln1 to stimulate GC axon growth in the developing cerebellum as in the spinal cord. We continued to carefully examine the Cbln1 cKO cerebella. Immunostaining of the Purkinje cell (PC) marker Calbindin and the granule cell (GC) marker NeuN showed no difference between Cbln1 cKO and control cerebella at P8 (Fig 6F), suggesting that the neurogenesis of PC and GC in the cerebellum is not impaired. To investigate whether the in vitro regulation of GC axon growth by Cbln1 was recapitulated in vivo, we examined parallel fiber (PF) development in Cbln1 cKO mice by DiI labeling. Compared with control mice, the DiI-labeled parallel fiber lengths in Cbln1 cKO mouse pups at P6 were significantly decreased (Fig 6G and 6H), indicating that the parallel fiber growth was impaired in Cbln1 cKO cerebella. All these data suggested that cell-autonomous Cbln1 from granule cells is required for parallel fiber growth in the developing cerebellum, just as cell-autonomous Cbln1 from commissural axons stimulates CA growth in the developing spinal cord.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001853

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/