Background FOXG1 syndrome (OMIM 613454) is a rare disease that is classified as a high-confidence syndromic cause of autism spectrum disorder (ASD). The predicted incidence of FOXG1 syndrome is 2.8-3.5 per 100,000 live births, and approximately 1200 individuals have been identified with FOXG1 syndrome. The symptoms of FOXG1 syndrome encompass many organ systems and include examples such as developmental delay, intellectual disability, microcephaly, epilepsy, movement disorders, hypotonia, and feeding and gastrointestinal difficulties. Missense, nonsense, deletion, frameshift, and duplication variants in the FOXG1 gene have been classified as pathogenic, and most of those variants are de novo. Our understanding of how genetic variants in FOXG1 lead to the diverse range of clinical symptoms observed in affected individuals remains limited, and there are currently no targeted treatments available for FOXG1 syndrome. Therefore, this study aims to uncover genotype-phenotype relationships, identify plasma biomarkers, and determine whether patient-specific variant mouse model phenotypes are reflective of patient phenotypes.

Methods Using the innovative Citizen Health platform and the FOXG1 Research Foundation Patient Registry, we have collected medical records and caregiver-reported patient data. Additionally, we have conducted plasma proteomics studies for 12 patients with FOXG1 syndrome and 126 age-matched control samples to identify possible biomarkers. We have also developed and characterized four patient-specific variant mouse models, Foxg1 c.250dupC (p.Q84Pfs31), c.483del (p.G161Gfs23), c.646G>A (p.G216S), and c.900G>A (p.W300X).

Results Genetic analysis of the FOXG1 Research Foundation Patient Registry and Citizen Health data demonstrated a diverse spectrum of genetic variants and phenotypes among the affected individuals with confirmed pathogenic variants in FOXG1. Moreover, we have discovered numerous differentially expressed proteins from our proteomics analysis, with several key pathways and processes being disrupted in FOXG1 syndrome patients, including chromatin remodeling and neuroimmune activation. In our patient-specific variant mouse models, we have identified anatomical brain deficits, motor deficits, impaired communication, impaired hippocampal-dependent learning, and social cognition deficits; thus, these models exhibit overlap with FOXG1 syndrome patient phenotypes.

Conclusion We assembled a large FOXG1 syndrome patient cohort to perform molecular characterization of FOXG1 variants and assess phenotypic impacts. We also identified potential plasma biomarkers using proteomics and developed patient-specific mouse models that recapitulate patient phenotypes. Together, these efforts advance our understanding of the pathophysiology of FOXG1 syndrome and support the future development of targeted therapeutic strategies.