Abstract
Cohesin is a multi-subunit complex that plays diverse roles in cellular replication, and genome organisation. Mutations in genes involved in the cohesin complex are linked to developmental disorders called cohesinopathies, which have a broad spectrum of developmental anomalies but share neurodevelopmental phenotypes as a common feature. Despite their prominence, a mechanistic explanation for the neurological phenotypes observed in both patients and animal models with cohesin deficiency remains elusive.
The cohesin complex is made up of four core subunits: SMC1A, SMC3, RAD21 and STAG1/STAG2. STAG1 and STAG2 are interchangeable in the cohesin complex, but each subunit confers individual properties to cohesin complex function. Cohesin has important roles in cell division through its involvement in sister chromatid cohesion and DNA damage repair. Cohesin also contributes to gene expression, helping to dictate three-dimensional genome structure. Mutations that cause complete loss of cohesin function are lethal.
Zebrafish were previously created with mutations in the RAD21 paralog rad21, and the STAG2 paralog stag2b. These unique lines allowed me to model the consequences for neurodevelopment when different cohesin components are lost. For example, changes in cohesin dosage can be investigated by heterozygous loss of the essential subunit Rad21; and changes in the type of cohesin, by heterozygous or homozygous loss of Stag2b which is compensated by other Stag proteins to change the type of cohesin available. Using 6-day-old zebrafish larvae with rad21 and stag2b loss-of-function mutations, I investigated the phenotypic consequences for head and brain development, behaviour, neuronal activity, chromatin structure, and gene expression.
Heterozygous mutation of the rad21 gene modestly shortened head length and reduced hindbrain size of larvae, but these animals retained normal brain function and had minimal changes in gene expression in the head and brain. This is consistent with the interpretation that rad21 haploinsufficiency affects rapidly proliferating tissues with high cohesin demand (as shown in previous studies) but does not significantly affect slower proliferating cells in the brain. In contrast, loss of stag2b markedly reduced the size of the telencephalon, dramatically altered neuronal activity in anatomically normal sensory regions, and altered larval behaviour. Larvae with complete loss of stag2b exhibited anxiolytic behaviour resembling cohesinopathy phenotypes. Importantly, brain function changes of stag2b mutants are not associated with reduced brain size because altered neuronal activity was present in regions with unaffected volume.
Surprisingly, no changes in chromatin structure were observed in neurons of stag2b mutants in this study. Single cell RNA sequencing showed that stag2b homozygous mutants had slower progression through the G2/M phase of the cell cycle in proliferative cells in the brain, which might account for differences in brain size but cannot account for the behavioural and neuronal activity changes observed. Instead, altered brain function was linked to dysregulated expression of protocadherin genes, which help to regulate neural wiring, a system closely linked to cohesin deficiency in other models. The zebrafish stag2b model indicates that loss of normal cohesin function may alter clustered protocadherin regulatory networks to drive changes in brain function that define cohesinopathies completely independently of co-occurring changes in neuroanatomy.
By assessing the phenotypic features of cohesin deficient zebrafish and identifying potential underlying causes this study helps to explain why cohesin deficiency drives neurodevelopmental disease. My research shows that even in survivable cohesin mutants there are consequences to cell cycle progression that cannot yet truly be disentangled from cohesin's other developmental roles, suggesting that more remains to be discovered in understanding the neuropathological aspects of cohesinopathies.