Disruption of Creb-Dependent Transcription Alters Brain Anatomy

Background: In the nervous system, many signaling cascades converge upon and activate the CREB family of transcription factors. Subsequent CREB-dependent gene expression has been implicated in many complex processes, including cell growth/survival, synaptic plasticity, and memory formation. Mutations of genes in the CREB signaling pathway are associated with neurodevelopmental disorders, including Rubinstein-Taybi syndrome, Coffin-Lowry syndrome, and other “Rasopathies”.

Objectives: To use whole-brain MRI to characterize whether loss of CREB affects mouse brain anatomy, and whether these effects depend on CREB gene dosage.

Methods: We used male and female mice with targeted deletions of two CREB isoforms (CREBαδ mutants). All 3 genotypes (CREB+/+, CREB+/-, and CREB-/-, n=12-16/genotype/sex, total N=82) were used. At 2-4 months of age, mice were perfusion-fixed for high-resolution (40 µm isotropic) ex-vivo MRI. Automated algorithms were used to align the images and compute the volumes of 62 segmented structures for each brain. The Jacobian determinant of the deformation field was used to measure expansion/contraction at each voxel. For each structure or voxel, ANOVAs were performed to determine whether that structure/voxel volume differed by genotype. Gene dosage effects were modeled by treating genotype as an ordered factor and testing for linear versus quadratic genotype effects. Multiple comparisons were controlled using a False Discovery Rate of 5%.

Results: Total brain volume was reduced by 2.5% in CREB-/- mice versus CREB+/+ mice (p<0.001) but did not differ significantly between CREB+/- and CREB+/+ mice. The loss of brain volume in CREB-/- mice was driven by a significant decrease in volume throughout the cerebral cortex, olfactory bulbs, basal ganglia, amygdala, hippocampus, and major white matter tracts, including the corpus callosum and anterior commissure. These areas remained significantly smaller in CREB-/- mice versus CREB+/+ mice even after normalizing for brain volume. In general, the absolute volume of these areas did not differ between CREB+/+ and CREB+/- mice. Notable exceptions include some white matter tracts including the lateral olfactory tract and internal capsule. As in CREB-/- mice, these were significantly smaller in CREB+/- versus CREB+/+ animals.

Intriguingly, the arbor vita and cerebellum were significantly larger in CREB+/- versus CREB+/+ mice, and were larger still in CREB-/- animals, indicating a gene-dose effect.

Conclusions: Loss of CREB produces a substantial volume loss throughout the brain, including cortical and subcortical areas, and white matter tracts. Even after accounting for differences in total brain volume, most of these regions were smaller in CREB-/- versus CREB+/+ mice. This suggests these areas either fail to develop fully or undergo degeneration. CREB+/- mice were overall phenotypically similar to CREB+/+ mice. The exceptions included white matter tracts, suggesting they may be more sensitive to disruption of CREB-dependent transcription. Overall, disrupting CREB has widespread effects on brain anatomy which may underlie the intellectual impairments observed in related neurodevelopmental disorders.