Introduction

Hypertension has been associated with alterations in vascular function, including endothelial dysfunction and arterial remodelling. These alterations may subsequently lead to structural and functional changes of the brain, ultimately contributing to the development of cognitive impairment and vascular dementia. However, exactly how vascular dysfunction contributes to these pathologies is still not fully understood [1]. Proper neuronal function is directly dependent on the composition, turnover, and amount of the interstitial fluid (ISF) that bathes the cells. Most of the ISF is likely to be derived from ion and water leakage across the brain capillary endothelial cells. A previous study showed increased interstitial fluid flow, altered ionic composition, and a tendency towards an increased water content of the brain in hypertensive rats [2]. To further elucidate differences in brain water management between normotensive and spontaneously hypertensive rats, we aimed to use non-invasive MRI measurements to determine possible changes in diffusion and water exchange properties within brain tissue.

Methods

All animal experiments were approved by our local animal welfare committee. Male spontaneously hypertensive rats (SHR) (n = 10) and normotensive Wistar Kyoto rats (WKY) (n = 11) were purchased from Envigo at 6 weeks of age. Animals were kept until 11 months old. Blood pressure and heart rate were measured in conscious rats using a non-invasive tail-cuff system, prior to the experiments. MRI was performed on a 7T small animal MRI system (MR Solutions, Guilford, UK). The MRI protocol consisted of T2w anatomical scans followed by quantitative ADC and T1 mapping for measurements of water diffusion and water capillary exchange, respectively. T1 mapping was performed before and after (with 5 min intervals) injection of an intravascular contrast agent through a cannula placed in the tail vein. Specific parameters of the different sequences were as follows: T2w-imaging – multi-slice SE, TR/TE = 4000/45 ms, α = 90°, ETL = 7, FOV = 35×35 mm², matrix size = 256×256, slice thickness = 1 mm, number of slices = 26, NSA = 4, total acquisition time = 9 min. ADC mapping – multi-slice SE-EPI, b-values = 0/800, number of diffusion directions = 3, TR/TE = 2000/30 ms, α = 90°, FOV = 35×35 mm², matrix size = 128×128, slice thickness = 1 mm, number of slices = 5, NSA = 1, total acquisition time = 13 min. T1 mapping – single-slice IR LookLocker, TR/TE = 10/3 ms, α = 8°, ETL = 8, FOV = 35×35 mm², matrix size = 128×128, slice thickness = 1 mm, NSA = 2, acquisition time = 4:32 min. After the MRI measurements, animals were sacrificed and the brains were carefully removed from the skull. Water content was determined after removing the cerebrospinal fluid and weighing the brain before and after desiccation at 90 °C for 7 days. For image analysis, ITK-SNAP was used to segment total brain volumes, as well as volumes of several anatomical structures of interest. ADC and T1 maps were co-registered to the T2w scans using FSL, after which mean values were determined for the hippocampal brain regions.

Results

Both systolic and diastolic blood pressure, as well as heart rate, were significantly elevated in SHR. Body weight did not differ between the two strains, while the brains of SHR were lighter when compared to its normotensive control (Figure 1). In line with the brain weights, total brain volumes were significantly lower in SHR. More specifically, white matter volume was smaller in SHR, while hippocampal volume was similar between groups. In contrast, both the third ventricle and lateral ventricles were remarkably enlarged in the hypertensive strain (Figure 2). Figure 3 shows representative examples of single slice ADC maps and T1 maps, acquired at the location where the hippocampus region could easily be segmented on the corresponding T2w image. Whereas whole brain water content was significantly higher in SHR rats, we could not observe significant differences in hippocampal ADC values (Figure 4). Furthermore, although we did not yet perform a detailed analysis of water exchange parameters using all dynamic T1 data [3], we did observe that ΔR1 values between pre- and 5-10 min post-contrast T1 scans tended to be different between hypertensive and normotensive rats (p = 0.057).

Discussion & Conclusion

Brains of hypertensive rats showed increased ventricular volumes, as well as an increase in brain tissue water content. Whereas these changes indicate dysregulation of brain water management in hypertensive rats, we did not yet observe significant changes in MRI parameters explaining this imbalance. Further evaluation of brain capillary water exchange may provide further insight in the impact of hypertension on brain water management.

Acknowledgements

None