Objectives: Identify the genetics of callosal agenesis and describe aberrant patterns of fiber pathways evident in these individuals.
Methods: We have used a wide array of genetic tools, including genome-wide copy number variant (CNV) analysis, linkage analysis, cloning of chromosomal breakpoints as well as high priority gene resequencing approaches. For analysis of the anatomic phenotype in patients, we have used diffusion tensor imaging with deterministic tractography and connectome analysis.
Results: We performed high-throughput genome-wide SNP arrays on a cohort of 275 individuals with AgCC and 2349 controls. Patients with AgCC had an enrichment of gene rich CNV >1 Mb (p=6.10x10-4; odds ratio [OR] = 5.22) and 9.4% of AgCC patients had a large de novo CNV. Rare AgCC CNV overlapped significantly with CNV genes in ASD and intellectual disability (ID) (p=3.82x10-3; OR=2.12 and p=1.64x10-3; OR=5.23 respectively), and were more linked when comparing de novo AgCC CNV to ASD and ID (p=1.50x10-3; OR=7.55 and p=0.001; OR=23.4 respectively). We have also shown that many of the identified CNV in this study overlap with chromosomal intervals identified in prior AgCC studies, highlighting the recurrence of AgCC genes. Overall, these findings underscore the genetic connection between AgCC and both ASD and ID and point to pathways implicated in these disorders. Additional results on single gene analyses (both published and recent work) will be presented. We have also investigated the hypothesis that additional white matter tracts that may accompany AgCC may be equally important for the ASD phenotype seen in AgCC individuals. Using deterministic tractography and targeted volumetrics, we have shown that the cingulum bundle, as measured by volume and fractional anisotropy is reduced in size and in degree of microstructural integrity in AgCC individuals compared to matched controls. We also conducted a structural connectome analysis of AgCC patients in comparison to matched controls. To further this analysis, we compared AgCC patients to controls in which we have similated removal of all callosal connections (“virtual callosotomy”). This investigation demonstrated that the network topology of AgCC is more variable than that seen in controls and that the degree of aberrant connectivity seen in AgCC is much more pervasive than can be explained simply by the absence of callosal fibers. Indeed, modularity analysis points to weak connectivity in the bilateral cingulum bundles, confirming our deterministic tractography analysis.
Conclusions: Our genetic findings demonstrate that the biology that underlies AgCC is highly linked to both ASD and ID and that de novo CNV are a significant cause of AgCC, Tractrography and connectome analysis highlight changes in other white matter tracts present in AgCC patients and suggest that biological pathways implicated in CC development are shared by many white matter tracts throughout the brain and are thus important factors in the development of autism spectrum disorders.