**Background:** Gonadotropin-releasing hormone (GnRH) deficiency (GD) is a disorder characterized by absent or delayed puberty, with largely unknown genetic causes. Known genes account for only 50% of cases. The authors aimed to obtain gene expression profiles of GnRH neurons during development to identify novel genetic determinants of GD, leveraging the observation that GnRH neurons originate in the nasal placode and migrate to the hypothalamus during embryonic development.
**Methods:** The study employed Affymetrix GeneChip arrays to obtain transcriptomic profiles of GN11 (immature, migratory) and GT1-7 (mature, secretory) immortalized GnRH neuron cell lines (n=3 per line). Primary GnRH neurons were isolated by FACS from Gnrh1-GFP rat embryos at embryonic days E14, E17, and E20 (n=1 per stage, pooled from 8-10 embryos). Differentially expressed genes (|logFC|>2, adjusted P<0.05) were filtered through GO/KEGG pathways related to cytoskeleton dynamics, migration, and axon development. ToppGene software prioritized candidates based on similarity to known GD causative genes. Exome sequencing was performed on 47 GD patients from a UK cohort (Case 1) and an additional patient identified via GeneMatcher (Case 2). Functional validation included site-directed mutagenesis to create R55* and W122* NLGN3 variants, overexpression in COS7 and GN11 cells, immunoblotting, immunocytochemistry, and neurite outgrowth quantification.
**Key Results:** Transcriptomic analysis identified 1612 differentially expressed probes between GN11 and GT1-7 cells (703 upregulated in GN11, 812 in GT1-7), mapping to 1515 unique annotated genes. Filtering and prioritization yielded 29 candidate genes (7 early, 22 late). Among these, two novel nonsense NLGN3 variants were identified: c.366G>A (p.W122*) in Case 1 and c.163C>T (p.R55*) in Case 2. Both are X-linked, hemizygous in affected males, and classified as pathogenic by ACMG criteria (CADD scores >35, DANN scores >0.99). Neither variant was found in gnomAD (v2.1.1), and NLGN3 has high intolerance to protein-truncating changes (pLi=0.98). Case 1 presented with partial GD (pubertal arrest, testes volume 10-12 ml, low-normal inhibin B), obesity (BMI 36), depression, anxiety, and ASD traits. Case 2 was prepubertal with micropenis (2.5 cm), small testes (1.5 ml bilaterally), obesity (BMI Z-score +2.1), ASD features, and learning difficulties. Functional studies showed NLGN3 is developmentally upregulated in maturing GnRH neurons (logFC=8.12 by RT-qPCR, P<0.01). In COS7 cells, WT NLGN3 produced a 110 kDa full-length protein and a 90 kDa shed ectodomain in conditioned media, while R55* and W122* produced truncated proteins of 7 and 14 kDa respectively with no detectable secreted forms. In GN11 cells, mutant proteins were retained in the endoplasmic reticulum, while WT NLGN3 localized to the plasma membrane. NLGN3 WT overexpression promoted neuritogenesis with significantly increased cell perimeter (P<0.0001) and complexity index (P<0.001) compared to GFP controls, while both mutants failed to induce these changes.
**Clinical Implications:** This study provides the first genetic evidence linking NLGN3 loss-of-function variants to GnRH deficiency, establishing a novel genotype-phenotype correlation between GD and ASD. The findings suggest that NLGN3 is essential for GnRH neuron neuritogenesis during development, and its deficiency may contribute to both reproductive and neurodevelopmental phenotypes. Clinicians should be aware of potential pubertal and reproductive disorders in children with ASD/neurodevelopmental disorders, and vice versa. The integrated transcriptomic approach identified 29 candidate genes that may yield additional genetic determinants of GD in future studies.