Combined HS6ST1 and HS6ST2 loss disrupts the formation and migration of gonadotropin-releasing hormone neurons

ABSTRACT

6-O sulfotransferase (HS6ST) enzymes modify heparan sulphate proteoglycans (HSPGs) to promote growth factor binding to extracellular matrix or receptors. HS6ST1 variants are reported in individuals with GnRH deficiency (GD), which is caused by the defective development of gonadotropin releasing hormone (GnRH) neurons. However, it is unknown whether and how HS6ST family members cooperate in GnRH neuron development. Here, we show that Hs6st1 and Hs6st2 are co-expressed in the mouse nasal neurogenic epithelia, where GnRH neurons and their migratory substrate, the terminal nerve, originate. The combined but not individual disruption of Hs6st1 and Hs6st2 disrupted vomeronasal organ formation with incomplete penetrance, and impaired brain entry of GnRH neurons, similar to observations in mice with loss of the morphogen FGF8 or the axon guidance cue SEMA3A, respectively. In agreement with a role for HS6ST-modified HSPG in regulating SEMA3A signalling, SEMA3A binding to nasal tissues was impaired in the absence of HS6ST1 and HS6ST2. Our results establish a mechanistic role for HS6ST1 in GD and suggest HS6ST2 as an additional susceptibility locus for GD.

Summary: Combined loss of HS6ST1 and HS6ST2 disrupts GnRH neuron formation and migration, establishing a mechanistic role for HS6ST1 in GnRH deficiency and suggesting HS6ST2 as an additional susceptibility locus.

INTRODUCTION

Gonadotropin-releasing hormone (GnRH) neurons are hypothalamic neuroendocrine cells that control reproduction in most vertebrates. During embryonic development, GnRH neurons originate in the nasal placode, a specialized ectodermal structure that also gives rise to the vomeronasal organ (VNO) and the olfactory epithelium (OE). More specifically, GnRH neurons emerge primarily from the VNO region together with neurons that form the terminal nerve (TN). These GnRH neurons become part of the so-called migratory mass: a population of cells that travels collectively from the nasal compartment into the developing forebrain. Guided by the TN, GnRH neurons migrate across the nasal–forebrain junction (NFJ) before reaching their final destination in the medial preoptic area (MPOA) of the hypothalamus (Cho et al., 2019; Herbison, 2016). Defective GnRH neuron development or function cause genetic forms of infertility characterised by GnRH deficiency (GD), the clinical features of which vary from a mild form, self-limited delayed puberty, to a severe form, Kallmann Syndrome with absent puberty and sense of smell (Boehm et al., 2015; Howard and Dunkel, 2019). To date, mutations in more than 50 genes have been linked to GD (Oleari et al., 2021a), but the disease-causing mechanisms have not yet been identified for all genes implicated in GD, and additional causative genes remain to be identified for a subset of patients.

HS6ST1 encodes a 6-O-sulfotransferase, one of several enzymes that modify glycosaminoglycan chains on heparan sulphate proteoglycans (HSPGs) by sulfation (Sarrazin et al., 2011). Individuals with heterozygous HS6ST1 variants show high phenotypic variability, ranging from DP to KS (Howard et al., 2018; Tornberg et al., 2011), suggesting that variants in this gene synergise with other genetic loci. In agreement with this idea, the heterozygous loss of Hs6st1 in mice impairs pubertal onset akin to DP but does not cause KS (Howard et al., 2018). Hs6st2 is a member of the 6-O-sulfotransferase gene family that acts redundantly with Hs6st1 during facial branchiomotor neuron migration in the developing brainstem (Tillo et al., 2016). Here, we show that Hs6st1 and Hs6st2 interact genetically to promote GnRH neuron formation and guidance along their axonal substrates to ensure that a normal number of GnRH neurons reaches the brain. These findings raise the possibility that mutations in both genes can contribute to GD.

RESULTS AND DISCUSSION

Hs6st1 and Hs6st2 are co-expressed during GnRH neuron birth and migration

To investigate the expression pattern of HS6ST1 and HS6ST2 during GnRH neuron development, we combined the analysis of single-cell RNA-sequencing (scRNA-seq) data from embryonic nasal tissues with in situ hybridization (ISH) of nasal tissue sections. For our analysis, we initially focused on embryonic day (E) 12.5, when GnRH neurons migrate within the nasal parenchyma as part of the migratory mass (Miller et al., 2010) that is composed of pioneer neurons, terminal nerve (TN) axons and neural crest-derived olfactory ensheathing cells (OECs) (Amato et al., 2024; Barraud et al., 2013; Cho et al., 2019). Thus, we analysed 18,203 cells from E12.5 craniofacial tissues (Cotney et al., 2024), which form 7 clusters based on Uniform Manifold Approximation and Projection (UMAP) (Fig. 1A). Cluster identity was confirmed by the expression of known markers for mesenchymal tissues (Lhx8, Prrx1, Foxp2 and Bmpr1b), neuroectoderm (Sox2 and Tfap2a), vascular endothelial cells (Cdh5 and Pecam1), macrophages/red blood cells (Spta1 and Fcer1g), muscle cells (Myod1and Ttn) and neural crest cells (NCCs; Sox10 and Fabp7) (Fig. S1). Hs6st1 and Hs6st2 were expressed in most cell clusters, with Hs6st1 enriched in the mesenchyme and neuroectoderm, and Hs6st2 expression was enriched in NCCs and muscle cells (Fig. 1B). To define expression of 6-O sulfotransferase enzymes in the migratory mass cell types, we sub-clustered 1916 cells belonging to neuroectoderm and NCC clusters (Fig. 1C). We found that both Hs6st1 and Hs6st2 were highly expressed by neural progenitors of the olfactory and vomeronasal epithelia, and by non-neurogenic respiratory epithelium (Fig. 1C and Fig. S2). Further, Hs6st1 was expressed by the Dcx⁺ pioneer neurons of the migratory mass, whereas Hs6st2 was expressed by Fabp7⁺ OECs (Fig. 1C). ISH using Hs6st1 and Hs6st2 probes, validated on the corresponding knockout tissues (Fig. S3), confirmed that Hs6st1 and Hs6st2 were expressed in the neurogenic epithelia of the E12.5 nose (Fig. 1D), including in the VNO, which gives rise to GnRH neurons and TN cell bodies (Cho et al., 2019; Taroc et al., 2019). In addition, both transcripts were expressed in cells migrating in the nasal parenchyma (Fig. 1D), consistent with cells in the migratory mass, such as OECs, pioneer neuronal cells and GnRH neurons (Miller et al., 2010).

Fig. 1.

Hs6st1 and Hs6st2 are co-expressed during GnRH neuron formation. (A-C) Re-analysis of the dataset Craniofacial Multiome Data (doi:10.5281/zenodo.14675171). (A) UMAP showing cells sequenced from E12.5 mouse wild-type craniofacial tissues labelled based on our refined cluster annotation. The black oval indicates cell populations used for subclustering in C. (B) UMAPs and violin plots show Hs6st1 and Hs6st2 expression in different cell populations. (C) UMAP of neuroectodermal cells and NCCs, subclustered and labelled based on marker annotation, including markers for post-mitotic neurons (Dcx) and olfactory ensheathing cells (Fabp7), and expression of Hs6st1 and H6st2 in different cell populations. (D) In situ hybridisation for Hs6st1 and Hs6st2 on coronal sections of E12.5 wild-type mouse nose. Areas outlined are shown at higher magnification in the adjacent panels. Black arrows indicate Hs6st1 or Hs6st2 expression in the vomeronasal organ (VNO) and neurogenic epithelia of the olfactory epithelium (OE); black arrowheads indicate examples of Hs6st1 or Hs6st2 expression in migratory mass cells. Scale bars: 250 μm (low magnification); 125 μm (high magnification). NCCs, neural crest cells; OE, olfactory epithelium; RBCs, red blood cells; VE, vascular endothelial; VNO, vomeronasal organ.

Similar results were obtained by re-analysing 32,650 cells from E14.5 wild-type nasal tissue (GSE234871; Amato et al., 2024), which form 8 clusters by UMAP analysis (Fig. 2A). Cluster identity was confirmed with markers for cartilage (Col2a1), mesenchyme (Runx2, Twist2 and Dcn), vascular endothelial cells (Cdh5), macrophages (Fcer1g), glial cells (Sox10 and Fabp7) and neuroectodermal cells (Sox2 and Tubb3) (Fig. S4). Both Hs6st1 and Hs6st2 were expressed across all these cell populations (Fig. 2B). We next subclustered the 3567 glial and neuroectodermal cells, and found that Hs6st1 transcripts were enriched in non-neurogenic respiratory epithelium, neural precursors and post-mitotic neurons, including in the cluster that contains Prph⁺ cells and, therefore, TN neurons (Fig. 2C and Fig. S5). Conversely, Hs6st2 transcripts were present in the respiratory epithelium, and enriched in neuronal progenitors and the Fabp7⁺ OECs (Fig. 2C and Fig. S5).

Fig. 2.

Hs6st1 and Hs6st2 are co-expressed during GnRH neuron migration. (A-C) Re-analysis of dataset GSE234871. (A) UMAP showing cells sequenced from the E14.5 mouse wild-type noses labelled based on our refined cluster annotation. Black ovals in A indicate cell populations considered for subclustering in C. (B) UMAPs and violin plots show Hs6st1 and Hs6st2 expression across different cell populations. (C) UMAP of neuroectodermal and glial cell clusters, subclustered and labelled based on our marker annotation, including markers for post-mitotic neurons (Prph) and olfactory ensheathing cells (Fabp7), and expression of Hs6st1 and H6st2 in different cell populations. (D) In situ hybridisation for Hs6st1 and Hs6st2 on coronal sections of E12.5 of wild-type mouse nose. Areas outlined are shown at higher magnification in the adjacent panels. Black arrows indicate Hs6st1 or Hs6st2 expression in the vomeronasal organ (VNO) and neurogenic epithelia of the olfactory epithelium (OE); black arrowheads indicate examples of Hs6st1 or Hs6st2 expression in migratory mass cells. Scale bars: 250 μm (low magnification); 125 μm (high magnification). OE, olfactory epithelium; VE, vascular endothelial; VNO, vomeronasal organ.

ISH on E14.5 coronal sections of the nose corroborated Hs6st1 and Hs6st2 expression in the neurogenic epithelia and cells in the nasal parenchyma (Fig. 2D). Further, expression of both genes was observed in the medial preoptic area (MPOA) of the hypothalamus (Fig. S6A), which is the final target of TN axons and GnRH neurons (Cho et al., 2019; Taroc et al., 2017). Reverse transcriptase (RT)-PCR also confirmed that Hs6st1 and Hs6st2 were expressed from E12.5 to E17.5 in the nose and forebrain, whereas a third member of the 6-O-sulfotransferase family, Hs6st3, was only weakly and transiently detected in the E12.5 nose (Fig. S6B). Finally, the analysis of published data from FACS-isolated Gnrh1-GFP neurons across developmental stages (Zouaghi et al., 2025) as well as human iPSC-derived GnRH neurons (Keen et al., 2021) identified robust levels of Hs6st1 and Hs6st2, but not Hs6st3, among the detected transcripts.

As our expression analyses with several complementary methods identified co-expression of transcripts for 6-O sulfotransferases 1 and 2 in the neurogenic epithelia of the olfactory placode, where the cell bodies of olfactory and TN axons reside, and GnRH neurons originate, these enzymes have the potential to modify the signalling of proteins known to modulate neurogenesis or neuronal migration during GnRH neuron development (Cho et al., 2019; Wierman et al., 2011).

Combined Hs6st1 and Hs6st2 loss decreases GnRH neuron number

To investigate the functional requirement of Hs6st1 and Hs6st2 for GnRH neuron development, we first examined mouse embryos lacking one or both genes at E12.5, when GnRH neurons emerge from the VNO and migrate within a migratory mass through the nasal compartment (Cho et al., 2019; Wierman et al., 2011). Immunolabeling for GnRH and PRPH revealed that 2/7 (29%) E12.5 Hs6st1^−/−;Hs6st2^−/− embryos lacked a VNO, and, accordingly, also lacked migrating GnRH neurons and TN axons (Fig. 3A). The remaining 5/7double mutant embryos (71%) instead had a VNO of normal size (wild-type versus Hs6st1^−/−;Hs6st2^−/− VNO: perimeter, 521.8±73.78 μm versus 501.1±72.50 μm; area, 18,364±4397 μm² versus 15,982±3832 μm²; ns, not significant by two-tailed unpaired Student's t-test) and exhibited TN axons emerging from the VNO (Fig. 3A).

Fig. 3.

Combined Hs6st1 and Hs6st2 loss decreases GnRH neuron number. (A) Immunofluorescent staining for GnRH (red) and PRPH (green) with DAPI counterstain (blue) of coronal sections from E12.5 heads of the indicated genotypes at the level of the nose. Areas indicated are shown at higher magnification in the corresponding adjacent panel. White arrowheads and arrows indicate examples of GnRH⁺ cells and PRPH⁺ axons, respectively, in wild type and Hs6st1^−/−;Hs6st2^−/− mutants with a vomeronasal organ (VNO); Δ indicates the absence of GnRH⁺ cells and PRPH⁺ axons in Hs6st1^−/−;Hs6st2^−/− mutant lacking a VNO. Scale bars: 200 μm (low magnification); 50 μm (high magnification). (B-D) Immunohistochemical staining for GnRH of coronal sections from E12.5 heads of the indicated genotypes at the level of the nose (B), and associated quantification of GnRH⁺ cells (C,D). (B) Representative micrographs, in which regions shown at higher magnification are outlined and shown in the corresponding adjacent panel. Black arrowheads indicate examples of GnRH⁺ cells in the nose of wild types and single mutants; Δ indicates the reduced number of GnRH⁺ cells in the nose of Hs6st1^−/−;Hs6st2^−/− mutants. Scale bars: 250 μm (low magnification); 125 μm (high magnification). Areas indicated are shown at higher magnification in the corresponding adjacent panel. (C,D) Quantification of total GnRH⁺ cell number (C) and proportion in forebrain versus nose (D); n=3 for wild type and Hs6st1^−/−; n=4 for Hs6st2^−/− and Hs6st1^−/−;Hs6st2^−/−. Data are presented as mean±s.d.; ns, not significant; ***P<0.001 [one-way (C) or two-way (D) ANOVA followed by Dunnett's post-hoc test]. FB, forebrain; OE, olfactory epithelium; VNO, vomeronasal organ.

Immunostaining followed by quantification showed a similar overall number of GnRH⁺ cells in wild-type and single-mutant littermates, but not in double-mutant littermates at this stage (Fig. 3B,C and Table S1). Thus, the number of GnRH⁺ cells in double Hs6st1^−/−;Hs6st2^−/− mutants with a VNO appeared visually reduced compared to wild-type and single-mutant embryos (Fig. 3B), and quantification confirmed that the overall number of GnRH⁺ cells was decreased in double mutants compared to wild types (Fig. 3C and Table S1). Nevertheless, the proportion of GnRH⁺ cells in the different migratory compartments was similar (Fig. 3D; wild type versus Hs6st1^−/−;Hs6st2^−/−: nose, 75.72±12.73% versus 75.97±7.64%; FB, 24.28±12.73% versus 24.03±7.64%; ns, not significant after two-way ANOVA followed by Dunnett's post-hoc test). These findings suggest that either Hs6st1 or Hs6st2 is required for the formation of a normal number of GnRH neurons.

Mechanistically, the reduced number of total GnRH neurons raised the hypothesis that the combined loss of Hs6st1 and Hs6st2 reduced the neurogenesis process that generates the VNO. Although we did not provide direct evidence for this hypothesis, we note that HS6ST1 is known to regulate the FGF8 gradient during mouse brain development (Chan et al., 2017) and is essential for neuroectoderm induction in the nose (Forni et al., 2013; Kawauchi et al., 2005). Accordingly, our re-analysis of scRNA-seq from E10.5 craniofacial tissues (Cotney et al., 2024), when the ectoderm invaginates to form the nasal pits giving rise to neurogenic (VNO and OE) and respiratory epithelia (Cho et al., 2019; Forni and Wray, 2012), confirmed that Hs6st1 and Hs6st2 are already expressed in neuroectodermal cells at E10.5, when Fgf8 expression is also high (Fig. S7). Further, the frequencies of VNO absence in Hs6st1^−/−;Hs6st2^−/− mutants examined here and in embryos lacking Fgf8 (Kawauchi et al., 2005) are almost identical (29%).

Combined Hs6st1 and Hs6st2 impairs GnRH neuron migration and disrupts terminal nerve axon patterning by affecting SEMA3A binding

Since Hs6st1 and Hs6st2 are still expressed at E14.5 in cell populations that guide GnRH neuron migration, such as TN neurons and OECs (Barraud et al., 2013; Taroc et al., 2017), we analysed mouse embryos at E14.5, when GnRH neurons migrate in a bilateral continuum from the vomeronasal organ across the nasal-forebrain junction and towards the medial preoptic area by following the TN axons (Amato et al., 2024; Casoni et al., 2016; Taroc et al., 2017; Wierman et al., 2011). As observed at E12.5, the number of GnRH⁺ cells was similar in single mutants compared to wild-type littermates, whereas double Hs6st1^−/−;Hs6st2^−/− mutants with a VNO had significantly fewer GnRH⁺ cells (Fig. 4A,B and Table S2). The remaining GnRH⁺ cells were over-represented in the nasal compartment and under-represented in the MPOA of double mutants at E14.5 (Fig. 4C; wild type versus Hs6st1^−/−;Hs6st2^−/−: nose, 37.69±6.00% versus 69.54±12.56%, ***P<0.001; NFJ 15.07±11.67% versus 25.18±8.20%; ns, not significant; MPOA, 47.24 9.42% versus 5.29±4.59%, ****P<0.01; two-way ANOVA followed by Dunnett's post-hoc test). This finding suggests that the combined loss of Hs6st1 and Hs6st2 impairs GnRH neuron migration across the NFJ into the brain. In agreement, immunostaining of coronal sections from wild-type and Hs6st1^−/−;Hs6st2^−/− embryos for PRPH, an established marker to identify TN axons (Amato et al., 2024; Oleari et al., 2019; Taroc et al., 2017), showed aberrant bundles of PRPH⁺ axons at the NFJ in double mutants (Fig. 4D) and lacked PRPH⁺ TN axons in the MPOA (Fig. 4E). Together, these findings suggest that Hs6st1 and Hs6st2 act redundantly to ensure that a normal number of GnRH neurons reaches the MPOA by promoting TN axon patterning.

Fig. 4.

Combined Hs6st1 and Hs6st2 loss impairs GnRH neuron migration and terminal nerve patterning. (A) GnRH staining of coronal sections of E14.5 heads from the indicated genotypes at the level of the nose (left), NFJ (middle) and MPOA (right). Areas outlined are shown at higher magnification in the corresponding adjacent panel. Black arrowheads indicate examples of GnRH⁺ cells at the NFJ and in the MPOA in wild type and single mutants; black wavy arrows and Δ indicate ectopic GnRH⁺ cells at the NFJ and the lack of GnRH⁺ cells in the MPOA, respectively, in Hs6st1^−/−;Hs6st2^−/− mutants. Scale bars: 250 μm (left and right panels); 125 μm (middle panels). (B,C) Quantification of the total GnRH⁺ cell number (B) and the proportion in the nose, forebrain and at the NFJ (C). n=3 wild type, Hs6st2^−/− and Hs6st1^−/−;Hs6st2^−/−; n=4 for Hs6st1^−/−. Data are presented as mean±s.d.; ns, not significant; **P<0.01, ***P<0.001 [one-way (B) or two-way (C) ANOVA followed by Dunnett's post-hoc test]. (D,E) PRPH staining with DAPI counterstaining of coronal sections from E14.5 heads of the indicated genotypes (n=3 per genotype) at the level of the nose (D) or MPOA (E). White arrows indicate TN axons contacting the olfactory bulbs (OB) and TN axons in the MPOA; white wavy arrows indicate aberrant nasal axon bundles at the NFJ; Δ indicates the absence of TN axons in the MPOA. Scale bars: 200 μm in D; 100 μm in E. (F,G) Alkaline phosphatase (AP)-binding assay on coronal sections from E12.5 heads from wild type (F) and Hs6st1^−/−;Hs6st2^−/− (G) mice (n=2 per genotype) with conditioned media containing AP-conjugated SEMA3A protein. Adjacent sections were immunolabeled for TUJ1 and counterstained with DAPI. Black arrowheads indicate binding to cells on nasal/TN axons of wild-type embryos, whereas Δ indicates the absence of binding in mutants. White arrows indicate examples of TUJ1⁺ axons in the nasal parenchyma of wild type and Hs6st1^−/−;Hs6st2^−/− mutants. Scale bars: 250 μm. MPOA, medial preoptic area; NFJ, nasal-forebrain junction; NS, nasal septum; OB, olfactory bulb; OE, olfactory epithelium; VNO, vomeronasal organ.

We next asked whether disrupted HSPG sulfation in Hs6st1^−/−;Hs6st2^−/− embryos impacted SEMA3A binding to cells in the nasal tissue, because the neural guidance molecule SEMA3A binds HSPGs (De Wit et al., 2005; Pérez et al., 2021), TN axons express the SEMA3A receptors NRP1 and NRP2 (Cariboni et al., 2011; Hanchate et al., 2012; Taroc et al., 2017), and PRPH⁺ axon clumps at the NFJ resemble those in mice lacking SEMA3A or SEMA3A signalling (Nrp1^Sema−/−;Nrp2^−/−) (Cariboni et al., 2011; Hanchate et al., 2012; Marcos et al., 2017; Oleari et al., 2019). Consistent with the possibility that loss of 6-O HSPG sulfation impairs SEMA3A signalling during GnRH neuron migration, alkaline phosphatase (AP)-conjugated, recombinant SEMA3A bound areas where PRPH⁺ axons exit the VNO and navigate in the nasal parenchyma, and this binding was abrogated in Hs6st1^−/−;Hs6st2^−/− mutants with a VNO (Fig. 4F,G).

Conclusions

Based on our expression studies and genetic analysis, we conclude that HS6ST1 and HS6ST2 cooperate to ensure VNO morphogenesis, GnRH neuron formation and TN-mediated GnRH neuron migration from the nose into the brain, likely via dual roles in FGF8 and SEMA3A signalling. This agrees with previous studies in other anatomical structures of the developing mouse head, which show that single deletion of these enzymes is insufficient to generate aberrant phenotypes, e.g. in the lacrimal gland and cranial nerves (Qu et al., 2011; Tillo et al., 2016). In the future, conditional genetic approaches will help to resolve the cell type-specific roles of these enzymes in the affected processes. Our findings also have implications for genetic screening, because heterozygous HS6ST1 variants have been identified in patients affected by KS (Tornberg et al., 2011) and self-limited DP (Howard et al., 2018). By contrast, HS6ST2 variants have been found in few patients with Paganini-Miozzo syndrome, characterized by severe neurodevelopmental impairment, myopia and mild facial dysmorphisms (Paganini et al., 2019; Sarmadian et al., 2023), with no reproductive data reported for patients with Paganini-Miozzo syndrome to date. Our findings suggest that the human orthologue of Hs6st2 is also a candidate gene for mutation screening in individuals with idiopathic KS and DP alongside screening for HS6ST1 variants, and that genetic syndromes with HS6ST2 variants should, reciprocally, be screened for HS6ST1 mutations.

MATERIALS AND METHODS

scRNA-seq analysis

Publicly available scRNA-seq datasets from mouse craniofacial tissue at E10.5-E12.5 (Zenodo Craniofacial Multiome Data; Cotney et al., 2024) and mouse nose at E14.5 (GEO accession number GSE234871; Amato et al., 2024) were analysed using Python (v3.10) with the Scanpy package (v1.11.15; Wolf et al., 2018). Low-quality cells were identified and removed according to median absolute deviation-based outlier detection for read counts, number of genes per cell and mitochondrial gene content. After per-cell normalization and cell cycle regression, batch correction across samples using the Harmonypy (v0.2.0; Korsunsky et al., 2019) was performed. After batch-aware variable gene selection (2000 features per dataset), dimensionality reduction was carried out with principal component analysis (PCA) followed by UMAP for visualization. Cells were grouped using graph-based clustering and cluster identity was assigned by manual annotation based on key marker genes.

Mouse strains

We used mice lacking Hs6st1 and/or Hs6st2 (Qu et al., 2011; Tillo et al., 2016) on a C57/Bl6 background. To obtain mouse embryos of defined gestational ages, mice were mated in the evening and the morning of vaginal plug formation was counted as embryonic day (E) 0.5. All animal procedures were performed in accordance with Animal Welfare Ethical Review Body (AWERB) guidelines under UK Home Office project and personal licences. Sample size for phenotypic analyses was determined based on previous works. No data were excluded from the analyses. Randomization protocols were not applied, but all mutants generated by Mendelian inheritance were analysed.

RNA extraction and RT-PCR

Total RNA was extracted from wild-type embryo nasal and forebrain tissues with TriFast reagent (Euroclone, EMR507100). cDNA was synthesized from 1 μg of total RNA using the High-Capacity cDNA reverse transcription kit and random hexamer primers (Thermo Fisher Scientific, 4368814). End-point PCR was performed for 30 cycles using the Quick Load 2X Master Mix (New England Biolabs, M0271L) and specific oligonucleotides for mouse Hs6st1 (5′-GCCAAGAAGAACCTGCGA-3′, 5′-GGCTGACCCTGACCTTGA-3′), Hs6st2 (5′-CATCCCTGCACGTCTGTG-3′, 5′-GCGCCATGTCTCTACGCT-3′), Hs6st3 (5′-TCATGGAGAAGAAGGATTGTCC-3′, 5′-GCTCTAGCTGTTTGGTGTGATG-3′) and the reference gene Gapdh (5′-TGGCATTGTGGAAGGGCTCATGAC-3′, 5′-ATGCCAGTGAGCTTCCCGTTCAGG-3′). Amplified products were analysed on a 1% agarose gel in 1×Tris Acetate EDTA buffer by electrophoresis.

Tissue preparation

For immunohistochemistry or in situ hybridisation (ISH), embryos were fixed for 3 h in 4% formaldehyde dissolved in PBS, cryoprotected in 30% sucrose in PBS overnight at 4°C and then embedded in optimal cutting temperature (OCT) compound (VWR, 361603E) for cryosectioning. For alkaline phosphatase (AP)-fusion protein-binding assays, freshly dissected, unfixed E12.5 embryos were cryoprotected in a sucrose gradient (10%, 20% and 30% in PBS, 30 min at each concentration at 4°C) before embedding them in OCT compound for cryosectioning.

In situ hybridisation

20 μm cryosections were incubated with anti-sense riboprobes for mouse Hs6st1 and Hs6st2 transcripts, using templates generated by PCR with primers based on the oligonucleotide sequences described above, and with one oligonucleotide modified at the 5′ with a T7 promoter sequence. Probes were generated by two rounds of 35 cycles of PCR amplification using Phusion High Fidelity DNA polymerase (New England Biolabs, M0530), followed by in vitro transcription with digoxigenin (DIG) RNA labelling kit (Roche, 11175025910), as previously described (Oleari et al., 2021b). Hybridisation was performed in 50% formamide, 0.3 M sodium chloride, 20 mM Tris (pH 7.5), 5 mM EDTA, 10% dextran sulphate and 1×Denhardt's solution overnight at 65°C. Sections were washed in washing buffer (50% formamide and 0.1% Tween 20 in 1×saline sodium citrate buffer), incubated overnight with AP-conjugated sheep anti-DIG IgG (1:1500; Roche, 11093274910) and developed overnight at 37°C with 4-nitro blue tetrazolium chloride (NBT) and 5-bromo-4-chloro-3-indolyl phosphate disodium salt (BCIP) (Roche, 11383213001 and 11383221001) dissolved in 100 mM Tris (pH 9.5), 50 mM MgCl₂, 100 mM NaCl and 1% Tween 20.

Immunoperoxidase labelling

Cryosections (20 μm) were incubated with Bloxall (Vector Laboratories, SP-6060) to quench endogenous peroxidase activity and then 10% heat-inactivated normal goat serum in PBS before incubation with rabbit anti-mouse GnRH (1:1000; Immunostar, 20075) (Taroc et al., 2019) and then goat biotinylated anti-rabbit IgG (1:400; Vector Laboratories, BA-1000). Sections were developed with the ABC kit (Vector Laboratories, PK-6100) and 3,3-diaminobenzidine (Merck, D4293). To determine the total number of GnRH neurons at E14.5, coronal sections through each entire head were immunolabelled for GnRH and all GnRH-positive cells in the nose, NFJ area and MPOA were counted, as previously reported (Oleari et al., 2019). To help distinguish individual GnRH neurons found in cell clumps at the NFJ of double mutants, high magnification images were analysed.

Immunofluorescence labelling

Cryostat sections (20 μm) were incubated with serum-free protein block (DAKO) after permeabilization of sections with 0.1% TritonX-100 and then incubated with primary followed by secondary antibodies in PBS. The primary antibodies used included: mouse anti-mouse nTUBB3 (1:200; clone TUJ1; Biolegend,801201) (Lettieri et al., 2023), rabbit anti-mouse PRPH (1:100; Merck Millipore, AB1530) (Causeret et al., 2023), chicken anti-rat PRPH (1:100; Abcam, ab39374) (Hibberd et al., 2022), rabbit anti-mouse GnRH (1:400; Immunostar, 20075) (Oleari et al., 2023). The secondary antibodies used were Alexa488-, Cy3- and Alexa647-conjugated donkey anti-chicken, anti-rabbit and anti-mouse Fab fragments, respectively (all 1:200; Jackson Immunoresearch). Nuclei were counterstained with DAPI (1:10, 000; Cell Signaling Technology, 4083).

AP-fusion protein-binding assays

AP-SEMA3A condition media were prepared as previously described (Paganoni et al., 2022). Cryostat sections from unfixed, flash-frozen embryos were fixed for 5 min in methanol, washed five times with PBS, incubated in PBS containing 10% heat-inactivated foetal bovine serum for 30 min and then reacted with AP-fusion protein for 2 h at room temperature. Sections were then washed for 5 min each with PBS, fixed with 4% formaldehyde in PBS for 2 min at room temperature and then washed 5 min each with PBS again. Endogenous AP was heat inactivated by incubation at 65°C for 3 h before tissue-bound heat-stable recombinant AP activity was detected as an insoluble reaction product after incubation with NBT/BCIP.

Image acquisition

Bright-field images were acquired using a Zeiss Axioskop2 plus microscope equipped with a TCH-5.0ICE digital camera and ISCapture software (v3.6.7, TiEsseLab). Immunofluorescence images were acquired with a Zeiss LSM900 Airyscan laser scanning confocal microscope equipped with 10×Plan-Apochromat M-27 (NA 0.45) and 40×LD C-Apochromat W Korr UV VIS IR (NA 1.1) objectives and an Axiocam 305 camera (Zeiss). DAPI, Alexa488 and Cy3 were excited at 405, 488 and 561 nm, and observed at 400-605, 410-545 and 535-617 nm, respectively. For TN analysis at E14.5, a Leica TCS SPE1 laser scanning confocal microscope equipped with 10×ACS APO objective (NA 0.30); DAPI and Alexa488 were excited at 405 and 488 nm and observed at 405-532 nm and 488-635 nm, respectively. 1024×1024 pixels images were captured in a stepwise fashion over a defined z-focus range corresponding to all visible fluorescence within the sample. Maximum projections of the z-stack were performed post-acquisition by using ZEN 3.0 Suite (Zeiss) software. Images were processed with Adobe Photoshop CC 2019 (v20.0.7, Adobe) software.

Statistics

Data are presented as mean±s.d. Statistical tests employed are outlined in the results section or in figure legends and were conducted when the experiment had been performed on a minimum of three distinct embryos per group, using Prism software (v8.2.1, GraphPad). Results were considered significant when P<0.05.

Peer review history

The peer review history is available online at https://journals.biologists.com/dev/lookup/doi/10.1242/dev.204788.reviewer-comments.pdf