Genetic and Functional Analyses Support a Role for CYFIP1 in Autism Spectrum Conditions

Background: The 15q11.2 region has been implicated in autism spectrum disorders (ASDs) associated with 15q11-q13 duplications and in Prader-Willi and Angelman syndromes. Within the recurrent, minimal copy number variation (CNV) region of 15q11.2 there are at least 5 genes, including CYFIP1 (cytoplasmic FMRP interacting protein 1). CYFIP1 is a critical part of a complex that includes fragile X mental retardation protein (FMRP) that regulates mRNA transport and translation at the synapse, controlling critical aspects of activity-dependent protein synthesis. Objectives: In the current study we examined the 15q11.2 region in ASD and studied the function of the Cyfip1 gene using mice with a disruption of the Cyfip1 gene. Methods: We made use of genome-wide single-nucleotide polymorphism (SNP) arrays to screen for CNVs in patients with ASDs. CNVs in 15q11.2 were confirmed by multiplex ligation-dependent probe amplification (MLPA). Mice with a disruption in Cyfip1 were generated from gene-trapped embryonic stem (ES) cells. Mice were characterized using quantitative polymerase chain reaction (qPCR), immunoblotting, behavior, and detailed electrophysiology in the hippocampus. Results: A survey of 184 unrelated patients with ASDs identified 3 CNVs in 15q11.2, including a de novo deletion and two inherited duplications, all affecting the genes in the interval (TUBGCP5, CYFIP1, NIPA2, NIPA1, and WHAMML1). CNV at 15q11.2 was not observed in pseudo-controls derived from untransmitted parental chromosomes. These findings suggest a role for one or more genes in 15q11.2 in ASD, with the important association of CYFIP1 with FMRP highlighting altered dosage of this gene as a very likely candidate for ASD-associated phenotypes. Mice with a disruption of one copy of Cyfip1 (heterozygotes) showed reduced expression of Cyfip1, with no change in expression of the structurally and functionally related Cyfip2. With hippocampal electrophysiology, there were no significant differences in input/output function, paired-pulse facilitation, or various forms of long-term potentiation among the genotypes. Remarkably, metabotropic glutamate receptor-dependent long-term depression (mGluR-LTD), whether induced by paired-pulse low frequency stimulation or with the mGluR agonist DHPG, was enhanced in the heterozygotes and was independent of protein synthesis. Application of both mGluR5 and mGluR1 antagonist to slices from Cyfip1 heterozygous mice reversed the increase in mGluR-LTD. Preliminary behavioral studies suggest abnormalities in extinction in a fear-conditioning paradigm. All of the phenotypes be observed in the Cyfip1 heterozygotes are similar to those observed in mouse models of Fragile X syndrome. Conclusions: Our genetic studies are consistent with a role for 15q11.2 gene dosage abnormalities in ASDs, as well as other psychiatric conditions. Mice lacking one functional copy of Cyfip1 show enhanced mGluR-LTD that is independent of protein synthesis. This observation provides a mechanism by which gene dosage abnormalities of CYFIP1 can alter synaptic plasticity and function and implicates shared mechanisms between Fragile X syndrome and loss of a functional copy of CYFIP1. Our studies with combined mGluR1 and mGlu5 antagonists to reverse these synaptic deficits provide a potential therapeutic target in CYFIP1 dosage abnormalities and also indicate that a combined therapy may be more effective in Fragile X syndrome.