The Clinical Relevance of Intragenic NRXN1 Deletions

Background: Neurexins (NRXNs) are involved in synapse formation, differentiation and function, which makes them important in brain development and function. Rare intragenic deletions affecting NRXN1 have been associated with autism spectrum disorder, but also with other neuropsychiatric disorders, such as developmental delay, language disorders, epilepsy, schizophrenia and aspecific dysmorphic features. In fact, NRXN1 deletions lack phenotypic specificity and are often inherited from a clinically unaffected parent. Therefore, it is challenging to determine to what extend a NRXN1 deletion can explain the phenotype in a particular patient.

Objectives: The current study aims to gain insight in the genotype-phenotype correlation of intragenic NRXN1 deletions.

Methods: First, an extensive literature study was performed that contains 629 heterozygous NRXN1 deletions. The study cohort was assembled through data extraction from 73 literature reports and consists of 341 probands and 140 relatives from clinical cohorts or case reports and 148 individuals from control cohorts. All individuals are characterized by clinical data, position and size of the variant and family relationships. These data were used to study the prevalence, penetrance, inheritance, segregation and phenotypic specificity of NRXN1 deletions. Second, we studied 44 additional cases with heterozygous NRXN1 deletions identified at our center in Leuven to illustrate our findings by means of case based reasoning. Moreover, the exact breakpoints of these deletions were identified through targeted sequencing.

Results: Based on published unscreened control cohorts, the population prevalence of intragenic NRXN1 deletions was estimated to be ~1/300 (0.312%; 99%CI [0.253%-0.370%]). However, 90% of these deletions were intronic, pointing out that the exonic deletions are very rare with a prevalence of ~1/3,000 (0.031%; 99%CI [0.013%-0.050%]). The observations from the extensive literature study allowed to distinguish three distinct parts in the NRXN1 gene. First, exon 6-24 deletions, that were only present in two out of over 100,000 controls, often segregated with intellectual disability (ID) in families or were de novo in sporadic cases with ID. Genotype-phenotype data in the Leuven patient cohort illustrated this observation. Taken together, this lends support for an association of exon 6-24 deletions with mainly ID with high penetrance. Second, several intragenic deletions within intron 5 were founder deletions that did not show clear evidence for an association with neurodevelopmental or psychiatric disorders in the current study. These results were obtained by exact breakpoint determination in the Leuven cohort. Third, deletions affecting exon 1-5 are susceptibility variants that may have a very variable expression (from mild to severe phenotypes) and a variable penetrance. Penetrance is estimated to be rather low, however, it is difficult to estimate since these copy number variants are also present in screened and unscreened controls and may be overrepresented in the patient population because of reporting bias.

Conclusions: The current study revealed some important correlations between the position of NRXN1 deletions on the one hand and the population prevalence, the segregation, the penetrance and the associated phenotype on the other hand. From these observations, clinical guidelines can be derived to improve the interpretation of a NRXN1 deletion for a particular patient and family.