Abstract
Pituitary gonadotrope cells are essential for the endocrine regulation of reproduction in vertebrates. These cells emerge early during embryogenesis, colonize the pituitary glands and organize in tridimensional networks, which are believed to be crucial to ensure proper regulation of fertility. However, the molecular mechanisms regulating the organization of gonadotrope cell population during embryogenesis remain poorly understood. In this work, we characterized the target genes of NEUROD1 and NEUROD4 transcription factors in the immature gonadotrope αT3-1 cell model by in silico functional genomic analyses. We demonstrated that NEUROD1/4 regulate genes belonging to the focal adhesion pathway. Using CRISPR/Cas9 knock-out approaches, we established a double NEUROD1/4 knock-out αT3-1 cell model and demonstrated that NEUROD1/4 regulate cell adhesion and cell motility. We then characterized, by immuno-fluorescence, focal adhesion number and signaling in the context of NEUROD1/4 insufficiency. We demonstrated that NEUROD1/4 knock-out leads to an increase in the number of focal adhesions associated with signaling abnormalities implicating the c-Src kinase. We further showed that the neurotrophin tyrosine kinase receptor 3 NTRK3, a target of NEUROD1/4, interacts physically with c-Src. Furthermore, using motility rescue experiments and time-lapse video microscopy, we demonstrated that NTRK3 is a major regulator of gonadotrope cell motility. Finally, using a Ntrk3 knock-out mouse model, we showed that NTRK3 regulates gonadotrope cells positioning in the developing pituitary, in vivo. Altogether our study demonstrates that the Neurod1/4-Ntrk3-cSrc pathway is a major actor of gonadotrope cell mobility, and thus provides new insights in the regulation of gonadotrope cell organization within the pituitary gland.
Similar content being viewed by others
Introduction
In multi-cellular organisms, the basic helix-loop-helix (bHLH) super-family regroups highly conserved transcription factors, often implicated in critical developmental processes. Amongst them, the Neurod/g sub-family emerged as major regulator of neurogenesis [1], playing critical roles in neural cell fate differentiation, but also in neuron proliferation, survival, and migration. This pro-neural sub-family is composed of four Neurod members: Neurod1, Neurod2, Neurod4, and Neurod6 and three Neurogenin, Neurog1, Neurog2, and Neurog3. NEUROD1 (ND1) and NEUROD4 (ND4) along with Neurogenin factors are able to reprogram progenitor cells toward a neural fate whereas ND2 and ND6 are implicated in neuronal identity maintenance (for review [2]). These factors can functionally compensate each other, displaying partially overlapping expression patterns and sharing a highly conserved DNA binding domain [3]. Neurod/g are also involved in specification of non-neural cells such as retinal photoreceptors [4], amacrine cells [5], pancreatic endocrine β cells [6], and entero-endocrine cells [7]. Furthermore, ND1 and ND4 are known to be expressed during pituitary gland development [8,9,10].
The pituitary gland is a common feature of all vertebrates. The endocrine part of this gland contains six endocrine lineages in the mouse: somatotrope, lactotrope, thyrotrope, corticotrope, melanotrope and gonadotrope. Each of them secretes specific hormones implicated in the homeostatic regulation of important physiological functions. For example, the gonadotrope cells secrete the gonadotropins, LH and FSH, regulating gonadal steroidogenesis and gametogenesis and thus, fine-tuning the reproductive function. This gland differentiates from a cranial placode, at day 8.5 of embryonic development (E8.5) in the mouse [11]. Cranial placodes are tissular hotspots of cell division and morphogenesis, originating at the boundary between neural and non-neural ectoderms [12]. They are known to produce various types of migratory cells such as neural and, as more recently demonstrated, endocrine cells [13]. Interestingly, all cranial placodes are neurogenic except for anterior pituitary and lens ones, suggesting an ontogenetic proximity between neural and pituitary development. Early during embryogenesis (E9.5), the pituitary placode gives rise to the Rathke’s pouch, the pituitary anlage. Progenitor cells proliferate and emigrate from this pouch to give birth to the nascent pituitary gland (around E12.5) [14]. From E13.5, under morphogenetic cues, the progenitors then differentiate progressively into the different endocrine lineages [15]. During this maturation, most of the pituitary lineages, and especially the gonadotrope one, are known to interconnect into cell networks [16], suggesting that cell motility might be important to establish the complex tridimensional organization of the embryonic pituitary.
Interestingly, pituitary differentiation requires several pro-neural transcription factors, highlighting similarities between pituitary and neural development. For example, as previously mentioned, the pro-neural transcription factors ND1 and ND4 are expressed in the developing pituitary [8,9,10]. In agreement with these expression patterns, Neurod1 or Neurdo4 knock-out leads to mild differentiation defects of corticotropes [17], lactotropes, and somatotropes, respectively [10]. While pituitary differentiation has not been investigated in Neurdo1/4 double mutant, differentiation defects of the somatotrope, corticotrope and gonadotrope lineages have been observed in a triple knock-out with Ascl1, coding a closely related bHLH transcription factor [18], demonstrating the importance of these factors for pituitary endocrine lineage maturation. However, the roles of ND1 and ND4, especially during gonadotrope specification, remains poorly characterized. In this work, we have studied ND1/4 molecular function during gonadotrope cell differentiation using a classical set of three immortalized mouse cell lines recapitulating three major steps of the gonadotrope differentiation process: the αT1-1 cell line, a model of early progenitors, the αT3-1 cells, exhibiting molecular signature of immature gonadotropes and finally the LβT2 cell line, representative of mature gonadotrope cells [19]. We demonstrate here that (i) ND1 and ND4 are critical factors during gonadotrope differentiation, regulating cell motility through a Ntrk3/c-Src pathway and (ii) NTRK3 tyrosine kinase receptor is crucial for the organization of the gonadotrope cell population during pituitary organogenesis.
Results
Neurod1 and Neurod4 are differentially expressed in a cellular model of gonadotrope differentiation
We first investigated the expression of Neurod/g genes in the αT1-1, αT3-1, and LβT2 cell lines. According to RNA-seq data [20], we found that Neurod1 and Neurod4 are the only Neurod/g expressed genes during gonadotrope differentiation, mostly in immature and to a lesser extent in progenitor gonadotrope cells (Fig. 1A). We confirmed this expression pattern by RT-qPCR assay (Fig. 1B). We then investigated whether Neurod1 and Neurod4 regulatory regions were differentially activated during gonadotrope differentiation using our published ATAC-seq data [21]. Only Neurod1 proximal promoter displayed chromatin accessibility, mainly in immature gonadotrope cells (Fig. 1C, left part). In contrast, we found six accessible chromatin regions in Neurod4 locus of αT1-1 and αT3-1 cells but not in LβT2 cells (Fig. 1C, right part), consistently with gene expression pattern. This indicates that Neurod4 expression might strongly depends on several potential enhancers scattered within the locus. We then tested enhancer cis-regulatory activities using luciferase reporter assays. As expected, Neurod4 promoter region (R2) was active specifically in αT1-1 and αT3-1 cells (Fig. 1D, insert). Among the potential enhancers, only R6 dramatically increased promoter basal activity in progenitor and immature gonadotrope cells (Fig. 1D). To further demonstrate that R6 is a functional Neurod4 enhancer, we generated stable R6 deletion in αT3-1 cells using CRISPR/Cas9 (these cells are hereafter named ΔND4, Fig. 1E) and measured Neurod4 expression by RT-qPCR. R6 deletion in immature gonadotropes induced a 90% decrease in Neurod4 transcript levels (Fig. 1F), demonstrating that the R6 enhancer is necessary for Neurod4 expression in αT3-1 cells. To our knowledge, this Neurod4 enhancer had not been described yet. As this evolutionary conserved region displays chromatin accessibility in tissues such as brain and retina (according to the UCSC Genome Browser databases), it might also be a critical Neurod4 regulatory region beyond the pituitary gland. The ΔND4 cells can be considered as an immature gonadotrope Neurod4 knock-out cellular model, allowing the analysis of ND4 functions in this cell type.
A Heatmap of Neurod/g gene expression in αT1-1, αT3-1, and LβT2 cells according to published RNA-seq (GSE104513). Color scale represents the normalized expression values. B Expression of Neurod1 and Neurod4 in αT1-1, αT3-1, and LβT2 cells. Gene expression levels were measured by RT-qPCR and normalized to Ppia (Cyclophilin A). Data are the normalized mean ± SEM of four independent experiments. Neurod1 and Neurod4 transcript levels were compared between the three cell lines by ANOVA followed by Tukey multiple comparison test. ***p < 0.001. C Chromatin accessibility of Neurod1 and Neurod4 loci in αT1-1, αT3-1, and LβT2 cells. Chromatin accessibility was investigated by ATAC-seq (assay for transposase-accessible chromatin with high-throughput sequencing) in the three gonadotrope cell lines. ATAC-seq tracks are shown for Neurod1 (left part of the panel) and Neurod4 (right part of the panel) loci. Gene exons are depicted as black boxes and proximal promoters are depicted as gray box in 5 prime of the ORFs. For each cell line, identified accessible chromatin regions are depicted as colored rectangles under each track. D Cis-regulatory activity of Neurod4 potential enhancers in gonadotrope cells. αT1-1, αT3-1, and LβT2 cells were transiently transfected with a GL4 luciferase reporter cassette under the control of Neurod4 promoter, here referred as R2 (pR2-GL4) or with this construct harboring potential cis-regulatory regions cloned in 3 prime of the luciferase cassette (pR2-GL4-Rx constructs, with x = R1 to R6 except R2). The activity of pR2-GL4 was normalized to the control pGL4 empty backbone activity (inset) and then pR2-GL4-Rx activities were normalized to pR2-GL4 activity. Relative luciferase activities are the normalized ratio between firefly luciferase over renilla luciferase. ANOVA followed by Dunnett’s multiple comparison test was performed independently for each cell line. Results are the mean ± SEM of six independent experiments. Significant difference with pR2-GL4 construct: **p < 0.01; ***p < 0.001. E Deletion of the genomic sequence of Neurod4 R6 enhancer and Neurod1 ORF was carried out in αT3-1 cells using CRISPR/Cas9. For Neurod4, genotyping PCR amplified a 1100 bp amplicon in WT cells while stable Neurod4 mutant cells (ΔND4) carried a deletion of about 320 bp, removing the R6 enhancer (left part of the panel, 780 bp amplicon). For Neurod1, a 780 bp amplicon was detected in WT cells while stable Neurod1 mutant cells (ΔND1) carried a deletion of about 500 bp, which encompassed Neurod1 ATG first codon (right part of the panel, 280 bp amplicon). The double mutant αT3-1 cell line (ΔΔ) was generated by deleting R6 enhancer on the Neurod1 mutant background. F For each genotype, homozygous clones were tested for Neurod4 and Neurod1 expression by RT-qPCR. Gene expression levels were normalized to Ppia. Data are the normalized mean ± SEM of four independent experiments. Gene expression levels of mutant clones were compared to WT by ANOVA followed by Dunnett’s multiple comparison test. ****p < 0.0001.
To study ND1 roles in immature gonadotropes, we created a αT3-1 Neurod1 mutant cell line by deleting the first ~500 bp of the Neurod1 ORF with CRISPR/Cas9 (Fig. 1E, ΔND1 cells), which successfully abolished gene expression (Fig. 1F). Finally, in order to investigate potential functional compensations between the two NEUROD factors, we established a double mutant αT3-1 cell line by deleting Neurod4-R6 in ΔND1 cells (Fig. 1E, ΔΔ cells) leading to a strong down-expression of both Neurod1 and Neurod4 (Fig. 1F).
Characterization of NEUROD1 and NEUROD4 potential regulome in immature gonadotrope cells
In order to characterize the molecular function of ND1 and ND4 in αT3-1 cells, we conducted an in silico functional genomics analysis (Fig. 2A). By cross-referencing several databases, we identified αT3-1 cells expressed genes (RNA-seq data [20]) associated with potential cis-regulatory regions displaying chromatin accessibility (αT3-1 cells ATAC-seq [21]) also known to be ND1 targets in other cell types (ND1 ChIP-seq data retrieved from https://chip-atlas.org/ [22]).
A Schematic representation (created with BioRender) of the in silico functional genomics analysis used to identify ND1/4 potential target genes in αT3-1 cells. Expressed genes associated with genomic regions displaying chromatin accessibly and potentially directly targeted by NEUROD factors were identified by cross-referencing ND1 ChIP-seq (from heterologous cell types) with αT3-1 cell ATAC-seq and RNA-seq databases. Retained candidate genes are depicted using blue boxes whereas discarded genes are depicted by red boxed. B Heatmap of the expression of 38 ND1/4 potential target genes identified in silico, in WT, ΔND1, ΔND4, and ΔΔ cells. Gene expression levels were normalized to Ppia. Color scale represents the normalized expression mean of four independent experiments. Gene expression in ΔND1, ΔND4, ΔΔ cells were compared to WT (values set to 1) by ANOVA followed by Dunnett’s multiple comparison test. *p < 0.05, **p < 0.01, and ***p < 0.001. C Violin plots of relative mRNA levels of 16 identified ND1/4 bona fide target genes (genes with significant expression decrease) according to WT, ΔND1, ΔND4, or ΔΔ genotype. To assess the global effect of Neurod1 and Neurod4 deletion on gene expression, mean expression levels of the target genes in ΔND1, ΔND4, and ΔΔ genotypes were compared to WT using an ANOVA followed by Dunnett’s multiple comparison test. ****p < 0.0001. D Network pathway analysis representation using Cytoscape ClueGO software. Analysis was performed on potential ND1/4 target genes identified by the genomic in silico analysis. Each colored dot is a different pathway regrouping several potential ND1/4 target genes. Related pathways (sharing more than 25% of their genes) are connected by a gray line. Pathways in the same cluster share a common color. Pathway’s name and associated genes are described in supplementary table 1.
In order to validate ND1/4 potential target genes identified in αT3-1 cells by this in silico approach, we measured the expression of a sample of them by RT-qPCR in WT, ΔND1, ΔND4, and ΔΔ cells (Fig. 2B). The expression of 40% of these genes was significantly decreased in ΔΔ cells as compared to WT, indicating that the in silico approach successfully captured a significant proportion of ND1/4 targets in αT3-1 cells. To compare the mean effect of ND1 and ND4 deletion on regulation of gene expression, we plotted the extent of variation according to the cell genotypes (Fig. 2C). This revealed that ND4 might play a more critical role as a transcriptional regulator as compared to ND1 in gonadotrope cells, even if ND1 may still be able to compensate in part for ND4 insufficiency (ΔΔ vs ΔND4 cells, p < 0.01).
To identify ND1/4 functions in gonadotrope cells, we then performed a network pathway enrichment analysis on their potential target genes, using Cytoscape ClueGO software [23]. We identified 72 significantly enriched pathways (Fig. 2D and supplementary table 1) organized in two major networks: one clustering neuro-pathogenesis-related pathways, consistently with known ND1/4 functions, and a second network of cancer and focal adhesion-related pathways.
NEUROD1/4 regulate gonadotrope cell adhesion and motility
The in silico analysis suggests that ND1/4 could be implicated in the regulation of focal adhesion (FA) organization and thus in adhesion and motility of gonadotrope cells. In order to test such hypothesis, we first performed a cell attachment assay using WT and ΔΔ cells. We found a 3-fold increase in ΔΔ cell attachment as compared to WT cells (Fig. 3A), suggesting that ND1/4 contribute to the regulation of cell interaction with the culture surface. Given that cell-matrix adhesion is involved in cell motility, we then performed a transwell migration assay using fetal bovine serum (FBS) as chemo-attractant. We observed that ∆∆ cells displayed a significant 2-fold migration decrease (Fig. 3B) whereas migration was unaltered in ΔND1 and ΔND4 cells (supplementary Fig. S1).
A Cell adhesion assay. Fluorescent Calcein AM-labeled WT and ΔΔ cells were seeded on an uncoated 96-well transparent bottom plate for 30 min. Fluorescence intensity is the ratio of fluorescence after PBS flushing over the initial fluorescence, reflecting the fraction of attached cells. Data are the normalized mean ± SEM of 3 independent experiments. The % of ΔΔ attached cells was compared to the one of WT using a Student’s t test. ****p < 0.0001. B FluoroBlok transwell migration assay. Calcein AM-labeled WT and ΔΔ cells were seeded in FluoroBlok transwell inserts and treated with foetal bovine serum (FBS, 10% in DMEM). Left: Relative Fluorescence Unit is the ratio of post-migratory fluorescence of FBS treated over untreated cells. Data are the normalized mean ± SEM of 6 independent experiments. Migration of ΔΔ cells was compared to WT using a Student’s t test. ***p < 0.001. Right: Representative images of post-migratory fluorescent WT and ΔΔ cells with or without FBS as chemoattractant. scale bar = 100 µm. C Live-cell tracking assay. WT and ΔΔ cells were transfected with H2B-mCherry to label and track the nuclei. Cells were seeded on Matrigel-coated culture surface until attachment and cell motility was tracked using live video time-lapse microscopy over 16 h. Tracks were analyzed using ImageJ TrackMate. Each graph corresponds to 30 representative tracks of transfected WT or ΔΔ cells with the initial position of the cells centered at the origin. Units: µm. D Total cell displacement analysis. The total distance (in µm) traveled by cells over 16 h was analyzed using ImageJ TrackMate on 150 WT and 220 ΔΔ cells. ΔΔ cell mean displacement was compared to WT. ****p < 0.0001 (Mann–Whitney test). E Cell velocity analysis. Cell instantaneous speed (nm/s) was calculated from the analysis of 150 WT and 220 ΔΔ cells using ImageJ TrackMate. ΔΔ cell mean velocity was compared to WT. ****p < 0.0001 (Mann–Whitney test).
In order to further characterize cell motility, we analyzed WT and ΔΔ cell velocity and overall cell displacement using video time-lapse microscopy. While αT3-1 WT cells were motile, ΔΔ cells only wobbled at their initial position (Fig. 3C and supplementary movies M1 and M2). We observed a significant ~50% decrease of both parameters in ΔΔ cells (Fig. 3D, E), indicative of cell motility impairment.
Altogether, these results demonstrate, for the first time, that ND1/4 are critical regulators of motility and adhesion of immature gonadotropes cells.
NEUROD1/4 regulate focal adhesion number
To test the potential regulation of FAs by NEUROD1/4, we first immuno-stained key components of FAs such as vinculin (VIN), FA kinase (FAK) or Zyxin (ZYX) in WT and ΔΔ cells. F-actin cytoskeleton was labeled using Phalloidin-TRITC. F-actin, FAK, and VIN can be considered as constitutive elements of all types of FAs, whereas ZYX is described as a marker of mature FAs [24]. We observed a 50% decrease of F-actin cellular content in ΔΔ cells (Fig. 4A, E), consistent with the decrease in Actb expression (Fig. 2A). Yet, actin-based cytoskeleton structures were still observable overall. This analysis also revealed a significant 40% reduction of mean apparent cell area in ΔΔ cells (Fig. 4F), suggesting a reduction of cell spreading. To analyze FAs more specifically, we first immuno-localized VIN and observed a significant 35% increase in number of Vinculin patches in ΔΔ cells without variation of global signal intensity (Fig. 4B, G, H). This suggests that FA number is increased in ΔΔ cells, in line with cells being more adherent and also less motile. To rule out the involvement of other structures implicated in cell adhesion and motility such as lamellipodia and invadopodia, we also analyzed Cortactin distribution which showed no difference between WT and ΔΔ cells (supplementary Fig. S2). We then analyzed the distribution pattern of total FAK and observed a comparable nucleo-cytoplasmic staining with signal at FAs in WT and ΔΔ cells (Fig. 4C, I). This suggests that FAK is similarly expressed in both cell lines. This was further confirmed by RT-PCR and western blotting (supplementary Fig. S3A, B). Finally, we analyzed the distribution pattern of the mature FA marker ZYX (Fig. 4D, J, K). We found an unchanged number of ZYX-labeled FA in ΔΔ cells despite a significant 35% decrease in ZYX global cell content, suggesting that ΔΔ cells might accumulate immature FAs.
A–D Representative confocal images of A F-actin, B VIN, C FAK, and D ZYX-labeled WT and ΔΔ cells. Fixed cells were labeled using phalloidin-TRITC for F-actin and DAPI for nuclei (Red: F-actin, Blue: nuclei, Green: VIN, FAK, or ZYX). B, D White arrows indicate positive B VIN and D ZYX-labeled FAs. Areas boxed in yellow are presented in the insets. scale bar = 10 µm. E Bar graphs of the mean F-actin fluorescence per cell and F mean cell apparent area, quantified on 30 different images for each genotype. ****p < 0.0001 (Mann–Whitney test). G Bar graphs of the mean VIN fluorescence intensity per cell and H violin plots of VIN-labeled FA particle number, quantified on 50 different images for each genotype. Mean VIN intensity and number of VIN-positive FAs were compared between ΔΔ and WT cells. *p < 0.05 and ns: not significant (Mann–Whitney test). I Bar graph of the mean FAK signal fluorescence intensity per cell, quantified on 15 different images for each genotype. Mean cell FAK intensity was compared between ΔΔ and WT cells. ns: not significant (Mann–Whitney test). J Bar graph of the mean ZYX signal fluorescence intensity per cell and K violin plots of ZYX-labeled FA particle numbers were quantified on 15 different images for each genotype. ZYX- mean cell intensity and number of ZYX-positive FAs were compared between ΔΔ and WT. ****p < 0.0001 and ns: not significant (Mann–Whitney test).
Altogether, these results thus indicate that ND1/4 regulate actin cytoskeleton organization and FA number in gonadotrope cells and might be implicated in FA maturation.
NEUROD1/4 regulate focal adhesion signaling
FA maturation is known to be regulated by phosphorylation of several effectors, such as FAK and c-Src. Briefly, upon stimulation, FAK dimerizes and auto-phosphorylates on Y397 residue (as reviewed in [25]). Activated c-Src is then recruited at FAs and, acting as a scaffold, independently of any catalytic activity, facilitates Y861 phosphorylation and full FAK activation by Y576/577 phosphorylation [26]. c-Src catalytic activity is then required for FAK Y925 phosphorylation [26], promoting FA turn-over [27].
We analyzed here both FAK phosphorylation level and cellular distribution for all aforementioned phosphorylation marks. We observed a significant increase in the number of FAs positive for pY397, pY576/577, and pY925 FAK in ΔΔ cells (Fig. 5A–D, F, H, J), consistently with the overall increase in FA number in these cells. Concerning FAK phospho-forms cellular content, we observed a complex perturbation pattern in ΔΔ cells: pY861 level was similar between ΔΔ and WT cells, while both pY397 and pY925 were significantly decreased and pY576/577 was significantly increased in ΔΔ cells (Fig. 5E, G, I, K). Similar level in FAK pY861 rules out a systematic impairment of FAK phosphorylation. Decreased level of pY397 might suggest a defect of FAK early activation and might be due to a lesser cell spreading (Fig. 4F) and subsequent integrin activation. However, increased level of pY576/577 might indicate a secondary hyper-activation of FAK and suggests that c-Src scaffolding activity is intact in these cells. Finally, decrease in pY925 level is consistent with a defect in FA late maturation and suggests an impairment in c-Src kinase activity. Altogether these results suggest that FA maturation defect might disturb FAK phosphorylation dynamics. To better understand this complex perturbation pattern, we then investigated c-Src activation.
A–D Representative confocal images of pFAK-labeled WT and ΔΔ cells. Phospho FAKs were immuno-localized using A pY397, B pY576/577, C pY925, and D pY861 antibodies on fixed cells, labeled with phalloidine-TRITC and DAPI (Green: pFAK, Red: F-actin, Blue: Nuclei). White arrows indicate positive FAs. Areas boxed in yellow are presented in the insets. scale bar = 10 µm. E–K Fluorescent signal quantification: E, G, I, K Bar graph of the mean pFAK fluorescence intensity per cell. F, H, J Violin plots of the number of pFAK-positive FAs. Quantification was performed on 5 (pY397), 15 (pY576/577), 35 (pY925), and 15 (pY861) different images for each genotype. For pY861 FAK, displaying cytoplasmic signal but only a weak signal at FAs, only mean cell fluorescence quantification was conducted. Mean cell fluorescence intensity and/or positive FA number were compared between ΔΔ and WT cells. *p < 0.05, **p < 0.01, ****p < 0.0001 and ns: not significant (Mann–Whitney test). L–N Representative confocal images of c-Src, pY527, and pY416c-Src-labeled WT and ΔΔ cells. c-Src and phospho c-Src were immuno-localized using L total c-Src, M pY527, N pY416 antibodies on fixed cells, labeled with phalloidine-TRITC and DAPI (Green: c-Src or phospho c-Src, Red: F-actin, Blue: Nuclei). White arrows indicate positive FAs. scale bar = 10 µm. O–R Fluorescent signals quantification: for total c-Src and pY527c-Src, displaying cytoplasmic signal but only a weak signal at FAs, only mean cell fluorescence conducted. For pY416c-Src, displaying both cytoplasmic signal and intense signal at FAs, mean cell fluorescence intensity and particle analysis were conducted. O–Q Histograms of mean c-Src or pY527 or pY416 signal fluorescence intensity per cell and R violin plots of pY416 c-Src-labeled FAs number were quantified on 20 (c-Src), 10 (pY527) and 30 (pY416) different images of both WT and ΔΔ cells. Mean cell fluorescence intensity and positive FA number were compared between ΔΔ and WT. *p < 0.05, **p < 0.01, and ****p < 0.0001 (Mann–Whitney test). S Representative image of western blot analysis of total c-Src and phosphorylated Y416c-Src in WT and ΔΔ cells. Intensity ratio: ratio of pY416c-Src over total c-Src band intensities. T Intensity ratio was compared between ΔΔ and WT using a Student’s t test. **p < 0.01.
At the basal state, c-Src is phosphorylated on Y527, a repressive post translational modification. Dephosphorylation of pY527 is necessary for c-Src auto-phosphorylation on Y416 which leads to its activation [28]. We thus analyzed total and phosphorylated c-Src localization in WT and ΔΔ cells (Fig. 5L–R). Total and pY527c-Src displayed a cytoplasmic localization whereas pY416c-Src localized at FAs. We observed a significant, albeit slight, decrease in total or pY527c-Src contents between WT and ΔΔ cells, whereas there was an unexpected 50% decrease of pY416 labeled-FAs in ΔΔ cells. Hypo-phosphorylation of c-Src on Y416 was confirmed by western blot (Fig. 5S, T). The c-Src activation defect is consistent with the FAK phosphorylation pattern described above, suggesting that ND1/4 are implicated in c-Src activation and hence in FA organization.
NEUROD1/4 regulate gonadotrope cell motility through c-Src activation
To link c-Src hypo-activation to motility defect of ΔΔ cells, we first over-expressed WT, dominant negative (DN) K297R/Y527F or constitutively activated (CA) Y527F c-Src in ΔΔ cells and monitored cell motility using video time-lapse microscopy (Figs. 6A and S4A, B). While DN c-Src had no effect, we observed that over-expression of WT c-Src fully rescued cell motility in ΔΔ cells. Furthermore, CA c-Src transfected ΔΔ cells displayed even greater motility than WT c-Src transfected ones. This suggests that c-Src activation defect is a major component of ΔΔ cells motility impairment.
A Cell motility rescue assay by c-SRC. WT and ΔΔ cells were transfected with H2B-mCherry to label and track the nuclei. WT cells were co-transfected with GFP expression plasmid (vehicle) and ΔΔ cells were co-transfected with either GFP (vehicle), c-Src DN, c-Src WT, or c-Src CA expression plasmid. Each graph corresponds to 30 representative tracks of transfected WT or ΔΔ cells with the initial position of the cells centered at the origin. B Schematic representation (created with BioRender) of c-Src inhibition and activation mechanisms. c-Src inhibition involves both Y527 phosphorylation by the CSK kinase and Y416dephosphorylation by the PTPN13 phosphatase. Upon extracellular signaling by integrins, c-Src disengages from CSK, with concomitant Y527dephosphorylation that may involve several putative phosphatases, and finally c-Src autophosphorylates on Y416. Signal transducing adaptor proteins such as DAB1 (Disabled-1) are known to increase c-Src activation. Finally, membrane receptors such as the neurotrophin 3 tyrosine kinase receptor NTRK3 have also been shown to trigger c-Src activation. C Gene expression of the actors of c-Src inhibition and activation pathways in WT and ΔΔ cells. Gene expression levels were normalized to Ppia. Data are the normalized mean ± SEM of three independent experiments. Gene expression were compared between ΔΔ and WT cells using a Student’s t test. ns: not significant, ***p < 0.001. D Cell motility rescue assay by NTRK3. (Left) Schematic representation of the Ntrk3 isoforms expressed in αT3-1 cells. Ligand domain is depicted with red boxes, the transmembrane domain with a green box and the tyrosine kinase domain with yellow boxes. The modified exon in Ntrk3-205 isoform is represented by a blue box. (Right) Cell motility assay. ΔΔ cells were co-transfected with H2B-mCherry and either with Ntrk3-201, Ntrk3-202, or Ntrk3-205 isoform expression plasmids. Each graph corresponds to 30 representative tracks of transfected ΔΔ cells with the initial position of the cells centered at the origin. E c-SRC-NTRK3 co-immunoprecipitation assay. WT cells were co-transfected with c-Src-eGFP and either with 3XFLAG control plasmid (lane 1, vehicle) or 3XFLAG-tagged isoform expression plasmids (Lane 2, Ntrk3-201, lane 3 Ntrk3-202 and lane 4 Ntrk3-205). NTRK3 was immunoprecipitated using anti-FLAG M2 antibody-coated magnetic beads. Immuno-precipitates were analyzed by Western blot for NTRK3 precipitation (using anti-FLAG antibody) and for both total and pY416 c-Src co-precipitations (Top panels). Expression of FLAG-tagged NTRK3, total c-SRC and pY416c-SRC-eGFP were verified by western blot analysis on whole cell lysates (Bottom panels). GAPDH was used as an internal reference for sample loading. The different bands observed for NTRK3 isoforms are consistent with glycosylated forms of the receptor. F Cell velocity analysis. ΔΔ cells were co-transfected with H2B-mCherry and either with GFP (vehicle), Ntrk3-201, or a mix of Ntrk3-201 and DN c-Src expression plasmids. Cell instantaneous speed (µm/minute) was calculated from the analysis of over 100 transfected ΔΔ cells over 16 h videomicroscopy, using ImageJ TrackMate. ΔΔ cell mean velocity was compared between the three conditions by ANOVA followed by Tukey multiple comparison test. ns: not significant, ***p < 0.001.
Activation of c-Src is dependent on a balance between inhibitory and activating pathways regulating the phosphorylation ratio on Y416 and Y527 residues (Fig. 6B). Both sides involve several actors, such as kinases and phosphatases [28] as well as adaptor proteins such as DAB1 (Disabled-1) [29]. If integrin signaling is the canonical activation pathway [30], several membrane receptors such as the neurotrophin 3 tyrosine kinase receptor NTRK3, have also been described to activate c-Src [31] (Fig. 6B).
We investigated the pathways regulating c-Src activation by RT-qPCR in ΔΔ cells. Interestingly, we observed a strong 7-fold decrease in Ntrk3 expression associated with a decrease in Dab1 and an increase in Ptpn13 transcript levels (Fig. 6C). All of these changes potentially contribute to the decrease in c-Src activation in ΔΔ cells.
We then evaluated the relative importance of DAB1, PTPN13, and NTRK3 in ΔΔ cell motility defects. Over-expression of DAB1, either its WT form or a phospho-mimetic activated Y185/197D mutant [32], failed to significantly rescue ΔΔ cell motility. The same was observed with Ptpn13 knock-down by CRISPRi (Figs S4C–E), ruling out a major contribution of DAB1 or PTPN13 in ΔΔ cell motility defect. We then over-expressed NTRK3 in ΔΔ cells and assessed motility. According to RNA-seq data, αT3-1 cells express three NTRK3 isoforms (Fig. 6D): Ntrk3-201 (Ensembl browser transcript name), the full-length active receptor with both neurotrophin binding and tyrosine kinase (TK) domains, Ntrk3-205, an isoform with an additional exon in the TK domain leading to functional deficiency [33] and, finally, the inactive Ntrk3-202 short variant, without the TK domain. Only expression of the full-length Ntrk3-201 rescued cell motility in ΔΔ cells (Figs. 6D and S4A, B). Interestingly, NTRK3-201 interacts with pY416c-Src in αT3-1 cells, as demonstrated by co-immunoprecipitation experiments (Fig. 6E), whereas interactions are noticeably weaker with NTRK3-202 and NTRK3-205 isoforms (2.5-fold reduction compared to NTRK3-201 after normalized quantification). These results suggest that NTRK3 full signaling activity is thus required for c-Src activation. To test whether NTRK3-driven cell motility is dependent on c-Src activity, we measured motility of ΔΔ cells expressing NTRK3 alone or together with DN c-Src. We observed that DN c-Src blocks the positive effect of NTRK3 on cell speed (Fig. 6F), indicating that c-Src mediates NTRK3 effect on cell motility.
Altogether, these results demonstrate that immature gonadotrope cell motility is regulated by a Neurod1/4-Ntrk3-Src activation pathway.
NTRK3 regulate gonadotrope cell number and position during mouse pituitary development
To investigate a potential role of Ntrk3 during gonadotrope differentiation, we first analyzed Ntrk3-201 transcript levels in the in vitro model of gonadotrope differentiation by RT-qPCR (Fig. 7A). We observed that the active isoform of Ntrk3 is specifically expressed in αT3-1 cells, representative of E14 to E16 gonadotrope cells.
A Ntrk3 active isoform expression in the in vitro model of gonadotrope differentiation. Ntrk3-201 expression level was quantified by qPCR and normalized to Ppia. Data are the normalized mean ± SEM of three independent experiments. Ntrk3-201 expression level was compared to αT3-1 cells using an ANOVA followed by Dunnett’s multiple comparison test. ****p < 0.0001. B Gonadotrope cell population distribution in TrkC mouse embryonic pituitary at E18.5. LH beta sub-unit (LHβ), specifically expressed by gonadotrope cells, was detected by immunohistofluorescence on E18.5 TrkC+/− and TrkC−/− embryo cryosections. Nuclei were stained by DAPI. Top and middle panels: Representative images of TrkC+/− (taken as control) and TrkC−/− pituitary sections. scale bar = 100 µm. Bottom panel: schematic representation of the pituitary and surrounding structures at E18.5. The pituitary gland sits between the palate sphenoid bone (ventral) and the pars nervosa and median eminence (dorsal). Gonadotrope cells, labeled in red, are mostly localized in the pars distalis, the ventral part of the endocrine pituitary. X and Y black arrows: cartesian coordinate system used to analyze gonadotrope cell distributions. C Quantification of gonadotrope cell number in TrkC+/− and TrkC−/− embryonic pituitary at E18.5. LHβ-expressing gonadotrope cells were numerated on a minimum of six sections of three female individuals for each genotype. Histological sections were matched between genotypes for similar anatomical positions. Only sections with complete pituitary of the same surface, and containing both pars nervosa and median eminence were retained to minimize counting bias. Number of gonadotrope cells per sections were compared between genotypes. ****p < 0.0001 (Mann–Whitney test). D Distribution of the gonadotrope cell population within the pituitary of TrkC+/− and TrkC−/− embryos at E18.5 (Green; TrkC+/− and red: TrkC−/−). Each identified gonadotrope cell on each section was plotted on a standardized cartesian coordinate (X;Y) system with the antero-posterior embryo axis represented by “X” and the dorso-ventral axis by “Y”. For each genotype, histograms of gonadotrope cell distributions on both X and Y axis were estimated. Distribution on each axis were compared using χ2 comparison test.
Gonadotrope cells appear around E13.5 in the mouse developing pituitary, in a restricted ventral area of the pars distalis [34]. This population then expands in the pituitary gland and, at E18.5, gonadotrope cells are organized in structured networks throughout the whole anterior lobe of the pituitary [35]. This expansion during pituitary organogenesis might implicate both proliferation and migration events. To investigate a potential role of Ntrk3 in gonadotrope cell expansion, we took profit of a Ntrk3 constitutive knock-out mouse model (TrkCtm1a, [36]), and analyzed the gonadotrope cell population in E18.5 TrkC+/− (taken as control) and TrkC−/− embryos by immunostaining of LHβ, a specific marker of gonadotrope cells. We first observed a significant 3-fold decrease in the number of LHβ positive cell in TrkC−/− embryos as compared to heterozygous controls (Fig. 7B, C), suggesting that NTRK3 might be implicated in pituitary colonization. To test for a regulatory role of NTRK3 on cell motility, we then analyzed the distribution of gonadotrope cells in the pituitary along the antero-posterior and on the dorso-ventral axes. We observed that gonadotrope cells spread significantly less on both axes in TrkC−/− fetal pituitaries as compared to TrkC+/− (Fig. 7B, D), suggesting that NTRK3 might be implicated in gonadotrope cell migration in vivo.
These results suggest that NTRK3 is an important actor of gonadotrope population organization in the developing pituitary, being implicated in population size as well as in proper colonization of the pituitary anlage.
Discussion
The concept of a tridimensional organization of endocrine cell networks in the pituitary gland has only recently emerged [16] and gonadotrope cell networks are now believed to be crucial for proper regulation of reproduction. However, the molecular mechanisms regulating gonadotrope cell motility and, in fine the structuration of the endocrine population, remain mostly elusive.
In the current work, we investigated the molecular regulation of gonadotrope cell motility. Using an in silico functional genomic analysis combined with CRISPR/Cas9 cellular knock-out, we demonstrated for the first time that, during pituitary ontogenesis, immature gonadotrope cell motility is regulated by a molecular pathway involving the ND1/4 transcription factors, the neurotrophin NTRK3 receptor and c-Src kinase (Fig. 8A).
Schematic representation of the principal results of this study (created with BioRender). A Gonadotrope cells (in red) differentiate in the ventral part of the pituitary and then disseminate in the gland. Our work shows that gonadotrope cell motility during pituitary organogenesis is dependent on NTRK3 receptor whose expression is controlled by ND1 and ND4 transcription factors. NTRK3 interacts with c-Src kinase and promotes its activation, a prerequisite for maturation of FAs and cell migration. B NEUROD1/4 insufficiency leads to a defect in NTRK3 expression, leading to a c-Src hypo-activation, which, in turn, disturbs FAK phosphorylation pattern. These signaling disruptions induce an accumulation of immature focal adhesions, preventing cell migration.
In gonadotrope cells, ND1/4 stimulate NTRK3 expression, which in turn interacts with and activates c-Src. Full c-Src kinase activity would be then required for optimal FAK phosphorylation, leading to FA terminal maturation and disassembly. ND1/4 insufficiency, leading to a hypo-activation of c-Src kinase activity, while not affecting its scaffolding role at FAs, unbalances both FAK phosphorylation pattern and FAK distribution, leading to immature FAs accumulation and cell motility defects (Fig. 8B).
Mutation of ND1 and ND4 in gonadotrope cells affects expression of other genes of the c-Src activation pathway such as DAB1 or PTPN13. However, we showed here that, amongst them, only NTRK3 efficiently rescued cell motility, suggesting that, in immature gonadotrope cells, NTRK3 activation of c-Src might be essential to trigger cell motility.
Taking advantage of a knock-out mouse model, we further demonstrated that NTRK3 receptor is crucial to ensure the colonization of the pituitary gland by gonadotrope cells during embryogenesis. NTRK3 deletion also led to a reduced number of gonadotrope cells that might be explained by a defect in migration, proliferation, or differentiation, in agreement with NTRK3 known functions [37,38,39,40].
Because ND1, ND4, and NTRK3 are expressed in other cell types, the Neurod1/4-Ntrk3-Src pathway might also be mobilized in cells requiring motility processes throughout their differentiation. For example, this pathway is expressed in human testicular spermatids according to the human protein atlas database and, although spermatid motility is an important part of their differentiation [41], the underlying molecular mechanisms are still poorly understood.
Whereas both ND1 and NTRK3 have already been reported to regulate cell motility [39, 42,43,44,45], much less was known about ND4 cellular function. We showed here that ND4 is as potent as ND1 to regulate gonadotrope cell motility. The transcriptional activity of ND4 would, however, be more important than ND1 for gonadotrope cells since ND1 only partially compensates for ND4 absence.
Pituitary gonadotrope cells display a remarkable functional plasticity during adulthood, which is essential to regulate fertility. Gonadotrope cell networks are suspected to play a important role in this plasticity as they are reconfigurated under physiological challenges such as puberty and lactation [46]. However, little is known about the regulation of gonadotrope cell motility [47, 48] or migratory ability [49] during these events. Interestingly NTRK3 remains expressed in adult gonadotrope cells, as demonstrated by single-cell transcriptomic analysis of adult rat pituitaries [50]. NTRK3, which has recently been implicated in neural network formation [51], could thus be a major actor of gonadotrope cell network plasticity during reproductive life. It would be thus interesting to analyze NTRK3 signaling in adult pituitary during critical physiological demands.
Dysregulations of NTRK3 signaling are known to be implicated in numerous cancers [52]. Gonadotrope adenomas are frequent cancers accounting for approximately 40% of all pituitary adenomas [53, 54]. However, most gonadotrope adenomas are poorly characterized and the molecular etiology of gonadotrope cell carcinogenesis is mostly unknown. A screen of NTRK3 mutations in these cancers should open new therapeutic avenues.
In conclusion, our work provides new insights into the molecular pathways controlling gonadotrope cell positioning during their differentiation in the developing pituitary and should contribute to a better understanding of the role of gonadotrope cell networks. Further studies are needed to establish whether this pathway is also mobilized at adulthood and contributes to the adaptation of gonadotrope cell population to physiological challenges at the critical stages of the reproductive life.
Materials and methods
Materials
Primary and secondary antibodies along with primers used for RT-qPCR, PCR genotyping, cis-regulatory regions, gRNAs, and ORFs cloning are described in supplementary table 2.
Cell cultures
The αT1-1, αT3-1, and LβT2 mouse gonadotrope cell lines (generously given by P. Mellon, University of California, La Jolla, CA) were grown in high-glucose DMEM supplemented with 10% fetal bovine serum (FBS from Gibco) and 0.1% penicillin/streptomycin, at 37 °C with 5% CO2. Cells are regularly tested for mycoplasma and for proper cell identity.
In silico analysis
NEURDO1 ChIP-seq Bed files (GSM3579940 [8] and GSM1306336 [55]) were retrieved from the ChIP-atlas website [22] and cross-referenced for overlapping target regions with the αT3-1 ATAC-seq [20] using bedtools intersect [56]. Only NEURDO1 ChIP-seq data were used as there is no available ChIP-seq data for NEURDO4 in the mouse. Genomic regions were then associated with the nearest gene Transcription Start Site using Homer annotatePeaks.pl [57]. Pathway analysis and network clustering were performed using Cytoscape ClueGO software [23].
RNA extraction and mRNA quantification
Total cellular RNAs were isolated using TRIzol™ Reagent according to the manufacturer’s protocol. RNAs (2 μg) were reverse transcribed with SuperScript II reverse transcriptase (ThermoFisher Scientific) using random primers according to the manufacturer’s instructions. qPCR quantifications were performed using Takyon™ 2X MasterMix from Eurogentec. Experiments were conducted independently in at least 3 replicates.
Immunocytofluorescence
Cells were seeded at a density of 500 cells/mm² on Nunc® Lab-Tek®. Coating was performed using Corning® Matrigel® Growth Factor Reduced (GFR) basement membrane matrix diluted in PBS (1/50) for 30 min at 37 °C. Cells were fixed 48 h after plating, with 4% PFA for 15 min, and 0.1% Triton X-100 permeabilization was performed. Permeabilized cells were then incubated with primary antibodies overnight at 4 °C and then with Alexa488 secondary antibodies 1 h at room temperature (RT). F-actin was labeled using Rhodamin-Phalloidin (ThermoFisher Scientific) and nuclei were labeled using DAPI.
Images were acquired using the ZEISS LSM 700 confocal microscope on the Functional and Adaptive Biology Unit microscopy facility, using laser settings and digital gain as described for each antibody in the additional file 2.
Image analyses
Images were analyzed using the ImageJ software [58]. For each image, cell number was assessed by counting the number of DAPI positive nuclei.
According to the signal patterns, we either quantified total intensity of fluorescence and/or FA number by particle analysis. Intensity of fluorescence was quantified using the “measure” option of the ImageJ software, after setting a minimum threshold value of 5 for each image. Raw Intensity values obtained were then compared to the number of nucleus per image to assess an approximated value of intensity for each cell type. Mean cell surface was assessed using the “measure” option of the ImageJ software, after setting a minimum threshold value of 5 for each image. Area values obtained were then normalized to the number of nucleus per image to evaluate the approximated surface covered by each cell type. Particle quantification was performed using the ImageJ software by reducing background for both actin and immuno-labeled protein signals with a Rolling ball radius of 2.5 pixels. Then, threshold of the immuno-labeled protein channel was set using the “Auto Local Threshold” tool with the “Phansalkar” method (radius of 25 and specific parameters value of 0.1). Actin and immuno-labeled protein colocalization was then assessed using the “Colocalization Threshold” tool. Colocalizing particles were analyzed with the “Analyze Particles” tool using the following parameters: “size: 0.3 pixel2 – infinity” “circularity: 0.0–0.8”. The number of particles counted was then normalized to the number of nuclei within each image.
Particle analysis for FAK pY397 and FAK pY925, displaying a high signal to background ratio at FA, was performed using the Analyze Particles command directly on the immuno-labeled protein channel without prior actin colocalization. Only objects with size ranging from 0.2 to 5 pixel2 and circularity parameter ranging from 0.0 to 0.7 were considered. Finally, the total count of particles for each image was normalized to the nucleus number.
Luciferase reporter assay
Cloning of the potential cis-regulatory regions of Neurod4: potential cis-regulatory element genomic regions were amplified from mouse DNA. Neurod4 proximal promoter (R2 region) was cloned upstream the luciferase gene in the pGL4.12 backbone (Promega). R1, R3, R4, R5, and R6 potential enhancers were then cloned in the pR2-GL4 vector downstream the SV40 poly(A) signal, between the BamH1 and SalI sites. Cell transfection was performed as follows: briefly, 50,000 cells were transiently transfected in 96-well plates using Lipofectamine ® 3000 (ThermoFisher Scientific) with 100 ng/well of pGL4 plasmids along with 5 ng/well of pRL-SV Renilla plasmid as an internal control for normalization. After 48 h transfection, firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega). Each experiment was performed six times independently in triplicates. Data were normalized to Renilla for each condition. ANOVA followed by Dunnett’s multiple comparison test was performed independently for each cell line and condition.
CRISPR/Cas9 deletions of Neurod1 and Neurod4 R6 region
pLV hUbC-Cas9-T2A-GFP (#53190), and pSPgRNA (#47108) plasmids [59] were purchased from Addgene. Specific guide RNAs, on both side of Neurod1 ORF or Neurod4 R6 enhancer, were designed using CCTop online tool [60] and cloned in pSPgRNA plasmid.
About 5 × 106 αT3-1 cells were electroporated with the Cas9 expression vector and the gRNA couples expression plasmids at a ratio of 1.5:8.5 µg using a Neon® Transfection System (Invitrogen) according to the manufacturer’s protocol (two pulses at 1500 mV for 15 ms). The empty pSPgRNA was used as control. After 48 h, GFP-positive cells were sorted using the FACS Aria II on the PIC2 facility of the Unit of Functional and Adaptive Biology. Sorted cells were then plated at low density until colonies were formed. ΔND1 and ΔND4 singles colonies were then genotyped for the Neurod1 locus and for the Neurod4 R6 region, respectively, using primers described in additional file 2. To generate Neurod1 and Neurod4 ΔΔ mutants, the R6 enhancer was deleted in a ΔND1 αT3-1 clone following the same protocol.
Protein extraction and western blot
Cellular proteins were extracted using a modified RIPA lysis buffer (150 mM NaCl; 50 mMTris pH 8; 1% Triton X-100; 0.1% SDS; 1/1000 DNase, supplemented with 1/100 Phosphatase and protease inhibitor cocktail from ThermoFisher Scientific) and separated on a 10% SDS-PAGE. After transfer, nitrocellulose membranes were incubated overnight at 4 °C with primary antibodies against Src and pY416Src diluted in Tris-buffered saline containing 0.01% Tween 20 (TBS-T) supplemented with 5% milk. After extensive washing, blots were incubated with a horseradish peroxidase-conjugated secondary antibody (GE healthcare #NA934V) in TBS-T/5% milk for 60 min at RT and then washed. Immunodetection was performed using an enhanced chemiluminescence detection system (GE Healthcare). Uncropped western blot images are shown in supplementary data.
ORFs cloning
For rescue experiments, the pIREShyg3 backbone (Clonetech) was modified to express a 2A-GFP cassette. c-Src, Dab1, and Ntrk3 isoforms coding sequences were amplified from αT3-1 cDNAs and inserted in frame in the p2A-GFP IRES hyg3 backbone. Constitutively activated (CA) and Dominant Negative (DN) c-Src and phopho-mimetic DAB1 were obtained by fusion PCR mutagenesis.
WT c-Src was further subcloned in pEGFP N1 backbone in frame with GFP (pSrc-eGFP). A 3X FLAG tag was inserted in Cter of each Ntrk3 isoforms (pNtrk3-3XFLAG 201, 202 and 205).
Motility assay
Cells were transfected by electroporation with pH2B-mCherry to track nucleus position (0.5 × 106 cells, 1 µg plasmid, two pulses at 1500 mV for 15 ms). For c-Src, Ntrk3, and Dab1 rescue experiments, cells were co-transfected with c-Src, Ntrk3, or Dab1 expression plasmids. For Ptpn13 knockdown, cells were co-transfected with a pLV hUbC-Zim3KRAB hdCas9-T2A-GFP, expressing a dead Cas9 (Addgene #53191) fused with the repressor KRAB domain of the human ZIM3 gene (as described in [61]), and gRNA targeting Ptpn13 proximal promoter described in [62] and supplementary table 2). Cells were then seeded in 12-well plates coated with 1/50e Corning® Matrigel® GFR at a density of 25,000 cells/cm². mCherry fluorescent nuclei were then followed using video time-lapse microscopy at 37 °C and 5% CO2 during 16 h. Cell displacement analysis and tracking were done on the red channel using the ImageJ software with the “TrackMate” plugin, using the following parameters: LoG detector, threshold of 3.0, the HyperStack Displayer, Simple LAP tracker, 75 microns Gap-closing max distance, 15 microns linking max distance and 5 microns Gap-closing max frame gap. Only cells displaying red nuclei signal, green cytoplasmic signal and directly adhering to the matrix through the whole movie were kept for analysis.
FluoroBlock cell migration assay
150,000 cells/cm² were seeded in the top part of Corning® FluoroBlok™ Cell Culture Inserts in DMEM serum-deprived medium for 24 h. Medium in the bottom chamber was then replaced by 10% FBS DMEM medium to induce cell migration. After 16 h, cells were labeled with the fluorescent Calcein AM dye (ThermoFisher Scientific) in Hank’s Balanced Salt Solution for 2 h. Bottom Fluorescence intensity, corresponding to post-migratory cells was measured using a FlexStation3 microplate reader at λ 495 nm.
Cell adhesion assay
WT and ΔΔ cells were labeled with Calcein AM dye in Hank’s Balanced Salt Solution for 30 min. Labeled cells were seeded on an uncoated cell surface of 96-well transparent bottom plate for 30 min. Initial bottom fluorescence was read using a FlexStation3 microplate reader at λ 495 nm and then cells were flushed away by two PBS washes. Residual bottom fluorescence was then read and the relative number of attached cells was estimated by calculating the percentage of remaining fluorescence intensity.
Co-immunoprecipitation
15 × 106 WT αT3-1 cells were co-transfected with c-SRC-eGFP and NTRK3 isoform expression plasmids (pNtrk3-3XFLAG 201, 202 and 205) by electroporation for 24 h. Cellular proteins were extracted using a modified RIPA lysis buffer (150 mM NaCl; 50 mMTris pH 8; 1% Triton X-100 containing 1/100 Phosphatase and protease inhibitor cocktail). Extracts were incubated overnight with anti-FLAG M2 antibody (F1804, Sigma-Aldrich) and then with Dynabeads™ Proteine G (ThermoFisher Scientific) for 4 h. Beads were then extensively washed in modified RIPA lysis buffer and eluted with Laemmli buffer. Eluted proteins were then immuno-detected for pY416c-Src by western blot. Input protein extracts were immuno-blotted for NTRK3 (FLAG), pY416c-Src, c-Src, and GAPDH). The ImageJ software was used for band intensity quantification. Uncropped western blot images are shown in supplementary data.
Immunohistofluorescence
The TrkCtm1a is a mouse gene-trap model with a trapping cassette located in between the 4th and 5th exons of Ntrk3 gene, resulting in a full Ntrk3 knock-out model. This model is hosted in the controlled environment of the animal facility (P-PAC) in the Cancer Research Center of Lyon, with ad libitum feeding (French Ministry for Research, registered number C2EA15, protocol CECCAPP CLB_2010_22). TrkCtm1a E18.5 embryos were obtained by crossing TrkC+/− mice. TrkCtm1a E18.5 embryos were obtained by crossing TrkC+/− mice. Pregnant mice were killed by cervical dislocation. Embryos were retrieved and fixated in PBS containing 4% PFA for 24 h. Fixated embryos were incubated in 30% sucrose and then embedded in OCT media (Cellpath, Newtown, UK), before being frozen at −80 °C. Embryos were PCR sexed using the mRbm31xy primers (described in [63]) and genotyped for Ntrk3 KO using the Lar primers [36] (Supplementary table 2). Female embryos were cryosectioned (15 µm thickness cuts). Cryosections were post-fixated in PBS-PFA 4% for 15 min, blocked with 10% Goat serum diluted in PBS for 1 h and then incubated overnight at 4 °C with anti-LHβ antibody (gift of Dr J.F. Roser, University of California Davis, USA). Invitrogen Alexa Fluor antibodies were used as secondary antibodies. Nuclei were stained with DAPI and slides were mounted in ProLong Gold Antifade Mountant (ThermoFisher Scientific) for imaging. Images were acquired using the ZEISS LSM 700 confocal microscope on the Functional and Adaptive Biology Unit microscopy facility. For each genotypes, at least 6 different pituitary sections of 3 embryos were studied. To reduce position bias, images were taken at standardized anatomical positions. Only images with complete pituitary glands of similar total area along with pars nervosa and median eminence were kept for analysis. More than 800 LHβ positive cells where counted per genotype. To compare cell distributions between genotypes, gonadotrope cell positions were plotted on a Cartesian coordinate system with an origin point fixed at the ventro-posterior tips of the pars distalis (schematic representation, Fig. 7B). Cell histogram distributions were plotted on both the X and Y axis and compared between genotypes using χ2 comparison test.
Statistical analysis
Analyses were performed in double-blinded experimental setups. Statistical analyses were performed using the Prism-GraphPad software. Test types and parameters are detailed in each figure legend.
Representative image adjustment
Maximum displayed value of signal (ImageJ software) were linearly adjusted using identical parameters between test and control conditions.
Data availability
The GSM3579940 and GSM1306336 NEURDO1 ChIP-seq raw data Bed files are available in the ChIP-atlas (https://chip-atlas.org/) repository. The GSE104513 αT3-1 RNA-seq dataset is available in the Gene Expression Omnibus repository and the αT3-1 ATAC-seq dataset is accessible upon reasonable request.
References
Tutukova S, Tarabykin V, Hernandez-Miranda LR. The role of neurod genes in brain development, function, and disease. Front Mol Neurosci. 2021;14:662774.
Dennis DJ, Han S, Schuurmans C. bHLH transcription factors in neural development, disease, and reprogramming. Brain Res. 2019;1705:48–65.
Bertrand N, Castro DS, Guillemot F. Proneural genes and the specification of neural cell types. Nat Rev Neurosci. 2002;3:517–30.
Cho JH, Klein WH, Tsai MJ. Compensational regulation of bHLH transcription factors in the postnatal development of BETA2/NeuroD1-null retina. Mech Dev. 2007;124:543–50.
Cherry TJ, Wang S, Bormuth I, Schwab M, Olson J, Cepko CL. NeuroD factors regulate cell fate and neurite stratification in the developing retina. J Neurosci. 2011;31:7365–79.
Gu C, Stein GH, Pan N, Goebbels S, Hörnberg H, Nave KA, et al. Pancreatic beta cells require NeuroD to achieve and maintain functional maturity. Cell Metab. 2010;11:298–310.
Li HJ, Ray SK, Pan N, Haigh J, Fritzsch B, Leiter AB. Intestinal Neurod1 expression impairs paneth cell differentiation and promotes enteroendocrine lineage specification. Sci Rep. 2019;9:19489.
Poulin G, Lebel M, Chamberland M, Paradis FW, Drouin J. Specific protein-protein interaction between basic helix-loop-helix transcription factors and homeoproteins of the Pitx family. Mol Cell Biol. 2000;20:4826–37.
Liu J, Lin C, Gleiberman A, Ohgi KA, Herman T, Huang HP, et al. Tbx19, a tissue-selective regulator of POMC gene expression. Proc Natl Acad Sci USA. 2001;98:8674–9.
Zhu X, Zhang J, Tollkuhn J, Ohsawa R, Bresnick EH, Guillemot F, et al. Sustained Notch signaling in progenitors is required for sequential emergence of distinct cell lineages during organogenesis. Genes Dev. 2006;20:2739–53.
Kouki T, Imai H, Aoto K, Eto K, Shioda S, Kawamura K, et al. Developmental origin of the rat adenohypophysis prior to the formation of Rathke’s pouch. Development 2001;128:959–63.
Sánchez-Arrones L, Ferrán JL, Hidalgo-Sanchez M, Puelles L. Origin and early development of the chicken adenohypophysis. Front Neuroanat. 2015;9:7.
Schlosser G. Induction and specification of cranial placodes. Dev Biol. 2006;294:303–51.
Ward RD, Raetzman LT, Suh H, Stone BM, Nasonkin IO, Camper SA. Role of PROP1 in pituitary gland growth. Mol Endocrinol. 2005;19:698–710.
Zhu X, Gleiberman AS, Rosenfeld MG. Molecular physiology of pituitary development: signaling and transcriptional networks. Physiol Rev. 2007;87:933–63.
Mollard P, Hodson DJ, Lafont C, Rizzoti K, Drouin J. A tridimensional view of pituitary development and function. Trends Endocrinol Metab. 2012;23:261–9.
Lamolet B, Poulin G, Chu K, Guillemot F, Tsai MJ, Drouin J. Tpit-independent function of NeuroD1(BETA2) in pituitary corticotroph differentiation. Mol Endocrinol. 2004;18:995–1003.
Ando M, Goto M, Hojo M, Kita A, Kitagawa M, Ohtsuka T, et al. The proneural bHLH genes Mash1, Math3 and NeuroD are required for pituitary development. J Mol Endocrinol. 2018;61:127–38.
Alarid ET, Windle JJ, Whyte DB, Mellon PL. Immortalization of pituitary cells at discrete stages of development by directed oncogenesis in transgenic mice. Development 1996;122:3319–29.
Xie H, Hoffmann HM, Iyer AK, Brayman MJ, Ngo C, Sunshine MJ, et al. Chromatin status and transcription factor binding to gonadotropin promoters in gonadotrope cell lines. Reprod Biol Endocrinol. 2017;15:86.
Pacini V, Petit F, Querat B, Laverriere JN, Cohen-Tannoudji J, L’hôte D. Identification of a pituitary ERα-activated enhancer triggering the expression of Nr5a1, the earliest gonadotrope lineage-specific transcription factor. Epigenetics Chromatin. 2019;12:48.
Zou Z, Ohta T, Miura F, Oki S. ChIP-Atlas 2021 update: a data-mining suite for exploring epigenomic landscapes by fully integrating ChIP-seq, ATAC-seq and Bisulfite-seq data. Nucleic Acids Res. 2022;50:W175–182.
Bindea G, Mlecnik B, Hackl H, Charoentong P, Tosolini M, Kirilovsky A, et al. ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks. Bioinformatics 2009;25:1091–3.
Zaidel-Bar R, Ballestrem C, Kam Z, Geiger B. Early molecular events in the assembly of matrix adhesions at the leading edge of migrating cells. J Cell Sci. 2003;116:4605–13.
Mitra SK, Schlaepfer DD. Integrin-regulated FAK-Src signaling in normal and cancer cells. Curr Opin Cell Biol. 2006;18:516–23.
Brunton VG, Avizienyte E, Fincham VJ, Serrels B, Metcalf CA, Sawyer TK, et al. Identification of Src-specific phosphorylation site on focal adhesion kinase: dissection of the role of Src SH2 and catalytic functions and their consequences for tumor cell behavior. Cancer Res. 2005;65:1335–42.
Deramaudt TB, Dujardin D, Hamadi A, Noulet F, Kolli K, De Mey J, et al. FAK phosphorylation at Tyr-925 regulates cross-talk between focal adhesion turnover and cell protrusion. Mol Biol Cell. 2011;22:964–75.
Roskoski R. Src kinase regulation by phosphorylation and dephosphorylation. Biochem Biophys Res Commun. 2005;331:1–14.
Bock HH, Herz J. Reelin activates SRC family tyrosine kinases in neurons. Curr Biol. 2003;13:18–26.
Huveneers S, Danen EHJ. The interaction of SRC kinase with beta3 integrin tails: a potential therapeutic target in thrombosis and cancer. ScientificWorldJournal. 2010;10:1100–5.
Jin W, Yun C, Hobbie A, Martin MJ, Sorensen PHB, Kim SJ. Cellular transformation and activation of the phosphoinositide-3-kinase-Akt cascade by the ETV6-NTRK3 chimeric tyrosine kinase requires c-Src. Cancer Res. 2007;67:3192–200.
Feng L, Cooper JA. Dual functions of Dab1 during brain development. Mol Cell Biol. 2009;29:324–32.
Valenzuela DM, Maisonpierre PC, Glass DJ, Rojas E, Nuñez L, Kong Y, et al. Alternative forms of rat TrkC with different functional capabilities. Neuron 1993;10:963–74.
Granger A, Ngô-Muller V, Bleux C, Guigon C, Pincas H, Magre S, et al. The promoter of the rat gonadotropin-releasing hormone receptor gene directs the expression of the human placental alkaline phosphatase reporter gene in gonadotrope cells in the anterior pituitary gland as well as in multiple extrapituitary tissues. Endocrinology 2004;145:983–93.
Budry L, Lafont C, El Yandouzi T, Chauvet N, Conéjero G, Drouin J, et al. Related pituitary cell lineages develop into interdigitated 3D cell networks. Proc Natl Acad Sci USA. 2011;108:12515–20.
Ménard M, Costechareyre C, Coelho-Aguiar JM, Jarrosson-Wuilleme L, Rama N, Blachier J, et al. The dependence receptor TrkC regulates the number of sensory neurons during DRG development. Dev Biol. 2018;442:249–61.
Cui Y, Yin Y, Zou Y, Zhao Y, Han J, Xu B, et al. The Rotary Cell Culture System increases NTRK3 expression and promotes neuronal differentiation and migratory ability of neural stem cells cultured on collagen sponge. Stem Cell Res Ther. 2021;12:298.
Bové M, Monto F, Guillem-Llobat P, Ivorra MD, Noguera MA, Zambrano A, et al. NT3/TrkC pathway modulates the expression of UCP-1 and adipocyte size in human and rodent adipose tissue. Front Endocrinol. 2021;12:630097.
Zanin JP, Battiato NL, Rovasio RA. Neurotrophic factor NT-3 displays a non-canonical cell guidance signaling function for cephalic neural crest cells. Eur J Cell Biol. 2013;92:264–79.
Bartkowska K, Paquin A, Gauthier AS, Kaplan DR, Miller FD. Trk signaling regulates neural precursor cell proliferation and differentiation during cortical development. Development 2007;134:4369–80.
Qian X, Mruk DD, Cheng YH, Tang EI, Han D, Lee WM, et al. Actin binding proteins, spermatid transport and spermiation. Semin Cell Dev Biol. 2014;30:75–85.
Singh SP, Wolfe A, Ng Y, DiVall SA, Buggs C, Levine JE, et al. Impaired estrogen feedback and infertility in female mice with pituitary-specific deletion of estrogen receptor alpha (ESR1). Biol Reprod. 2009;81:488–96.
Ikematsu Y, Tanaka K, Toyokawa G, Ijichi K, Ando N, Yoneshima Y, et al. NEUROD1 is highly expressed in extensive-disease small cell lung cancer and promotes tumor cell migration. Lung Cancer. 2020;146:97–104.
Ivanov SV, Panaccione A, Brown B, Guo Y, Moskaluk CA, Wick MJ, et al. TrkC signaling is activated in adenoid cystic carcinoma and requires NT-3 to stimulate invasive behavior. Oncogene 2013;32:3698–710.
Kim WY. NeuroD regulates neuronal migration. Mol Cells. 2013;35:444–9.
Alim Z, Hartshorn C, Mai O, Stitt I, Clay C, Tobet S, et al. Gonadotrope plasticity at cellular and population levels. Endocrinology 2012;153:4729–39.
Grønlien HK, Fontaine R, Hodne K, Tysseng I, Ager-Wick E, Weltzien FA, et al. Long extensions with varicosity-like structures in gonadotrope Lh cells facilitate clustering in medaka pituitary culture. PLoS ONE. 2021;16:e0245462.
Rahamim-Ben Navi L, Tsukerman A, Feldman A, Melamed P, Tomić M, Stojilkovic SS, et al. GnRH induces ERK-dependent bleb formation in gonadotrope cells, involving recruitment of members of a GnRH receptor-associated signalosome to the blebs. Front Endocrinol (Lausanne). 2017;8:113.
Godoy J, Nishimura M, Webster NJG. Gonadotropin-releasing hormone induces miR-132 and miR-212 to regulate cellular morphology and migration in immortalized LbetaT2 pituitary gonadotrope cells. Mol Endocrinol. 2011;25:810–20.
Fletcher PA, Smiljanic K, Maso Prévide R, Iben JR, Li T, Rokic MB, et al. Cell type- and sex-dependent transcriptome profiles of rat anterior pituitary cells. Front Endocrinol. 2019;10:623.
Li G, Zhang B, Sun JH, Shi LY, Huang MY, Huang LJ, et al. An NT-3-releasing bioscaffold supports the formation of TrkC-modified neural stem cell-derived neural network tissue with efficacy in repairing spinal cord injury. Bioact Mater. 2021;6:3766–81.
Manea CA, Badiu DC, Ploscaru IC, Zgura A, Bacinschi X, Smarandache CG, et al. A review of NTRK fusions in cancer. Ann Med Surg. 2022;79:103893.
Asa SL, Ezzat S. Gonadotrope tumors. Prog Mol Biol Transl Sci. 2016;143:187–210.
Ntali G, Capatina C. Updating the landscape for functioning gonadotroph tumors. Medicina. 2022;58:1071.
Jia S, Ivanov A, Blasevic D, Müller T, Purfürst B, Sun W, et al. Insm1 cooperates with Neurod1 and Foxa2 to maintain mature pancreatic β-cell function. EMBO J. 2015;34:1417–33.
Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 2010;26:841–2.
Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell. 2010;38:576–89.
Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012;9:676–82.
Kabadi AM, Ousterout DG, Hilton IB, Gersbach CA. Multiplex CRISPR/Cas9-based genome engineering from a single lentiviral vector. Nucleic Acids Res. 2014;42:e147.
Stemmer M, Thumberger T, Keyer M, del S, Wittbrodt J, Mateo JL. CCTop: an intuitive, flexible and reliable CRISPR/Cas9 target prediction tool. PLoS ONE. 2015;10:e0124633.
Alerasool N, Segal D, Lee H, Taipale M. An efficient KRAB domain for CRISPRi applications in human cells. Nat Methods. 2020;17:1093–6.
Horlbeck MA, Gilbert LA, Villalta JE, Adamson B, Pak RA, Chen Y, et al. Compact and highly active next-generation libraries for CRISPR-mediated gene repression and activation. Elife. 2016;5:e19760.
Tunster SJ. Genetic sex determination of mice by simplex PCR. Biol Sex Differ. 2017;8:31.
Acknowledgements
This work benefited from equipment and services from the PIC2 facility and the imaging facilities from both the Unit of Functional and Adaptive Biology and the ImagoSeine core facility, the latter being a member of the France BioImaging infrastructure supported by grant ANR-10-INBS-04 from the French National Research Agency. We are thankful to Stefania Tolu for her help with the FACS experiments. We also thank Dr Pamela Mellon who developed and provided us with the gonadotrope cell lines. We are very grateful to Patrick Mehlen who generously gave us access to the Trkctm1a mouse model. We also would like to thank numerous generous colleagues for their gifts of several of the antibodies used in this work, namely Dr. Sabrina Batonnet-Pichon, Dr. Frédérique Deshayes, Dr. Hervé Enslen and Dr. Frederic Bernard. Finally, this work would not have been possible without the pioneer works of Dr. Jean-Noël Laverrière on the regulation of gonadotrope cell genes.
Funding
This work was supported by grants from the Université Paris Cité, the Centre National de la Recherche Scientifique (CNRS), and the Institut National de la Santé et de la Recherche Médicale (Inserm).
Author information
Authors and Affiliations
Contributions
CLC performed most of the experiments (ICF, IHF, PCR, videomicroscopy), acquired and analyzed data, interpreted data, and participated to preparation of the figures and the manuscript; EA contributed to the western blot analyses and to the preparation of the expression plasmids. VP with FP contributed to the luciferase assays and to the CRISPR-Cas9-mediated knockout experiments. SDB greatly helped with the cell motility experimental set-ups and analysis. NR and STD produced the TrkC mouse line and provided the mouse embryos. JCT contributed to data interpretation, and to preparation of the manuscript; DL was involved in all aspects of the study design, data interpretation, and manuscript preparation. All authors reviewed and commented on the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Le Ciclé, C., Pacini, V., Rama, N. et al. The Neurod1/4-Ntrk3-Src pathway regulates gonadotrope cell adhesion and motility. Cell Death Discov. 9, 327 (2023). https://doi.org/10.1038/s41420-023-01615-7
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41420-023-01615-7
