Evidence for a transport barrier between ISS divisions. A) LSCM conducted after fluorescent tracer was injected into Cn and Tha. After injection, tracer distributions were limited within the two divisions. B) The transportation barrier between Cn and Tha with MRI in oblique-sagittal slices. Right panel: a high-resolution T2-weighted image. Middle and left panels: tracer-based dynamic T1-weighted images in which the tracer distribution reached maximum volumes. After paramagnetic tracer injection, the local tissue appeared as a hyper intense spot on MR. In Cn, tracer distribution was more extensive and the traced ISF flowed to the ipsilateral frontal cortex. No distribution was observed in Tha. The enhancement in Tha was localized to its anatomical division and didn’t flow to Cn. In the AQP4-knockout rats the barrier effect was still stable (C). The structural and functional parameters in different ISS divisions were disparate (D). The volume fraction was higher in Cn than Tha (P <0.01). The tortuosity was lower in Cn than Tha (P <0.05). The clearance rate was lower in Cn than Tha (P <0.05). The diffusion rate was higher than that in Tha (P <0.05). The structural and functional parameters of ISS in AQP4-knockout rats were disparate (E). In Tha, the volume fraction, tortuosity, and diffusion rate were not statistically different between AQP4-knockout and control groups. The clearance rate was lower in AQP4-knockout group than control group (P <0.05). In Cn, the volume fraction in AQP4-knockout group was higher than that in control group (P <0.05) and the tortuosity was lower than control group (P <0.05). The clearance rate in AQP4-knockout group was lower than that in control group (P <0.05), while the diffusion rate was higher than control group (P <0.05).