Quantitative Micro-Computed Tomography Imaging of Vascular Dysfunction in Progressive Kidney Diseases

Abstract

Progressive kidney diseases and renal fibrosis are associated with endothelial injury and capillary rarefaction. However, our understanding of these processes has been hampered by the lack of tools enabling the quantitative and noninvasive monitoring of vessel functionality. Here, we used micro-computed tomography (µCT) for anatomical and functional imaging of vascular alterations in three murine models with distinct mechanisms of progressive kidney injury: ischemia-reperfusion (I/R, days 1-56), unilateral ureteral obstruction (UUO, days 1-10), and Alport mice (6-8 weeks old). Contrast-enhanced in vivo µCT enabled robust, noninvasive, and longitudinal monitoring of vessel functionality and revealed a progressive decline of the renal relative blood volume in all models. This reduction ranged from -20% in early disease stages to -61% in late disease stages and preceded fibrosis. Upon Microfil perfusion, high-resolution ex vivo µCT allowed quantitative analyses of three-dimensional vascular networks in all three models. These analyses revealed significant and previously unrecognized alterations of preglomerular arteries: a reduction in vessel diameter, a prominent reduction in vessel branching, and increased vessel tortuosity. In summary, using µCT methodology, we revealed insights into macro-to-microvascular alterations in progressive renal disease and provide a platform that may serve as the basis to evaluate vascular therapeutics in renal disease.

Keywords: capillary rarefaction; chronic kidney disease; computed tomography; fibrosis; imaging; noninvasive.

Figure 1.

Experimental outline for anatomical and functional imaging of the renal vasculature during progressive kidney diseases. Three different murine models were investigated: I/R, Col4A3-deficient (Alport) mice and UUO. Using contrast-enhanced in vivo µCT, the renal relative blood volume (rBV) was determined noninvasively by quantifying the radiodensity of five spherical volumes of interests (VOIs) per kidney cortex. Upon Microfil perfusion, the 3D microarchitecture of the renal vasculature was visualized by high-resolution ex vivo µCT at a spatial resolution of 3 µm voxel side length, enabling the quantitative assessment of vessel branching, diameter and tortuosity in strict hierarchical order from the hilus (Aa. segmentales) to the periphery (afferent arterioles). Arrows demonstrate the order of rising branching points along the course of blood vessels. Vessel tortuosity was calculated as a ratio of the vessel length between two nearby branching points and the linear distance between those. CTRL, contralateral kidney.

Figure 2.

Capillary rarefaction is closely associated with renal fibrosis in I/R-induced progressive kidney fibrosis. (A) PAS staining as well as collagen I, fibronectin, αSMA and Meca-32 immunohistochemistry of I/R kidneys or sham controls. (B) Time course of quantified positively stained area fractions for collagen I, fibronectin, αSMA, and of the filled Meca-32 area fraction representing the reconstructed vascular volume on 2D histologic slices. (C) Significant correlations were found between the computationally filled Meca-32 area fraction and quantified area fractions for collagen I (left panel), fibronectin (middle panel), and αSMA (right panel) demonstrating a close association between renal fibrosis and capillary rarefaction. *P<0.05; **P<0.01; ***P<0.001. Correlation analyses were performed by calculating R² (square of Pearson correlation coefficient). PAS, periodic acid-Schiff.

Figure 3.

Capillary rarefaction is closely associated with renal fibrosis in UUO and in Alport mice. (A) Time course of quantified positively stained area fractions for collagen I, fibronectin, αSMA and of the filled Meca-32 area fraction in UUO kidneys using immunohistochemistry. (B) Time course of the same parameters in kidneys of Alport versus wild type (WT) mice. *P<0.05; **P<0.01; ***P<0.001.

Figure 4.

Longitudinal in vivo µCT imaging reveals a continuous reduction in functional renal blood vessels during progression of I/R-induced kidney injury. (A) Visualization of renal blood vessels by contrast-enhanced in vivo µCT (2D cross-sectional images in transversal (I), coronal (II) and sagittal (III) planes, as well as 3D volume renderings). (B) Longitudinal monitoring of the continuous reduction in number and functionality of renal blood vessels by in vivo µCT. (C) Noninvasive quantification of the renal rBV over time by in vivo µCT revealed a continuous reduction of vessel functionality in I/R-induced progressive fibrosis compared with kidneys of sham control mice. (D) A highly significant correlation was observed between the renal rBV determined by functional in vivo µCT and computationally filled vascular structures using ex vivo Meca-32 immunohistochemistry. **P<0.01; ***P<0.001.

Figure 5.

Functional vessels assessed by in vivo µCT are reduced in UUO and in Alport mice. (A) Visualization of the altered renal vasculature in UUO day 10 and sham control kidneys by in vivo µCT. (B) Noninvasive quantification of the renal rBV over time by in vivo µCT confirmed the continuous reduction in number and functionality of renal blood vessels in the UUO model. (C) A highly significant correlation was observed between the renal rBV and the filled Meca-32 area fraction in the UUO model. (D) Visualization of the altered renal vasculature in 8-week-old Col4a3-deficient (Alport) mice and age-matched wild type (WT) littermates by in vivo µCT. (E) Noninvasive quantification of the renal rBV in 6- and 8-week-old Alport mice and age-matched controls. (F) Also for the Alport model, a highly significant correlation was observed between the renal rBV and the filled Meca-32 area fraction. **P<0.01; ***P<0.001.

Figure 6.

Ex vivo µCT imaging and 3D quantification of the renal vasculature reveals significant alterations of renal arteries in progressive I/R induced renal injury. (A) Representative high-resolution ex vivo µCT images of sham controls and I/R day 14, 21 and 56 kidneys after Microfil perfusion (2D cross-sectional images in transversal (I), coronal (II) and sagittal (III) planes, as well as 3D volume renderings). Note the progressive rarefaction of functional vessels besides the continuous shrinkage of fibrotic kidneys over time. Scale bar, 200 µm. (B) µCT-based quantification of vascular branching points per increasing vessel order (from hilus to periphery). (C, D) µCT-based quantification of the mean vessel diameter (C) and mean vessel tortuosity (D) in sham control and I/R day 14, 21 and 56 kidneys. A continuous reduction in vessel branching and vessel size was linked to an increased vessel tortuosity during fibrosis progression. *P<0.05; **P<0.01; ***P<0.001.

Figure 7.

Alterations of renal arteries in progressive kidney damage using ex vivo µCT are confirmed in Alport mice. (A) Representative high-resolution ex vivo µCT images of kidneys from 8-week-old Alport and age-matched wild type (WT) mice upon Microfil perfusion. Note the loss of functional vessels and the shrinkage of the Alport kidney compared with the WT kidney. Scale bar, 200 µm. µCT-based quantification of vascular branching points (B), of vessel size (C) and of vessel tortuosity (D) confirmed the link between reduced branching points, reduced vessel diameters and increased vessel tortuosity during fibrosis progression. *P<0.05; **P<0.01; ***P<0.001.

Figure 8.

Alterations of renal arteries in progressive kidney damage using ex vivo µCT are confirmed in UUO induced renal injury. (A) Representative high-resolution ex vivo µCT images of Microfil perfused UUO day 10 and sham control kidneys. Scale bar, 200 µm. µCT-based quantification of vascular branching points (B), of vessel diameters (C) and of vessel tortuosity (D). **P<0.01; ***P<0.001.