Accurate assessment of renal function is important for diagnosis and monitoring kidney diseases. Recent advances in magnetic resonance imaging (MRI) allow the development of non-invasive and quantitative tools to assess renal function.¹ DCE-MRI using the low dose Gd-based contrast has been established as a reliable technique for measuring glomerular filtration rate (GFR) in individual kidneys.² Other promising markers for renal function include R2* measured with blood oxygenation level-dependent (BOLD) MRI ^3-4, and the longitudinal relaxation time T1.⁵ Unilateral ureteral obstruction (UUO) is a validated model of progressive renal fibrosis.⁶ In this study, we aimed to assess the potential changes in renal tissue properties and function in the UUO mouse model using the various functional MRI techniques.

Animal Preparation: The study contained 2 groups of male C57Bl6/J mice: 8 sham and 12 UUO. Each group was imaged at 3~4 days and 7~8 days after the surgery. The mice underwent either left ureter obstruction (ipsilateral) or sham operation. After imaging, mice were sacrificed, and kidneys from both sides were collected. Each kidney was halved transversely, and half was fixed in 10% neutral buffered formalin for histology, and another half was preserved in -80°C for gene expression.

MRI acquisition: MRI experiments were performed on a Bruker Biospec 7T 20cm horizontal bore system (Bruker, Billerica, MA) equipped with a 38mm mouse whole-body RF coil. Mice were anesthetized with 0.5~2% isoflurane and room air mixture during MR scans (normoxia: 21% O₂, 79% N₂). High resolution anatomical images were acquired using fast spin echo sequence with RARE factor 4, TR/TE1/TE2 = 3000/20/60ms, fat suppression, and respiration gating. BOLD MRI was acquired using the multi-gradient-echo (MGE) sequence with TR = 90ms, flip_angle = 30°, and TE = 3.2~24.2ms. T1-mapping was implemented using the respiration gated look-locker inversion recovery sequence with TR/TE = ~3000/4ms, flip_angle = 10°, 24 inversion pulses, and 120ms interval. DCE-MRI was implemented using FLASH sequence with TR/TE = 23/2.3ms, flip_angle = 10°, and temporal resolution 3s. After a 2-minute baseline scan, 0.03 mmol/kg Gd-DTPA (Magnevist) was injected via the tail vein. Three coronal slices of kidney tissue were acquired with a field-of-view 25.6×25.6mm² and 1mm slice thickness.

Data Analysis: The parametric maps were calculated using the customized image analysis toolkits in MATLAB. Pixel-wise R2* maps were calculated using exponential fitting to the multi-echo BOLD data. T1 maps were calculated using the three-parameter curve fitting to: M=A×(1-B×exp(-t/T1*)), where M is the signal intensity, A the scaling factor for equilibrium magnetization M0, B the correction factor for imperfect inversion, and t the effective inversion time. T1 was calculated by: T1=T1*×((B/A)-1). Single kidney GFR was quantified using a three-compartment tracer kinetic model.⁷ The model characterizes tracer retention in renal tissue as convolution of arterial input function (AIF) and impulse retention function (IRF). The kidney volume was estimated using the anatomical images. During data processing, nonlinear least-square curve fitting was used to optimize the values of the IRF parameters (including single-kidney GFR), and thus the model-estimated tracer retentions matched the measured ones from MR renography. Finally, the split renal function was computed as the ratio of the UUO-side GFR and the sum of the GFR values of both kidneys.

Fig 1a shows anatomical images of the contralateral and ipsilateral (arrow) kidneys in a UUO mouse. Ureteral obstruction resulted in marked pelvis enlargement (hyperintense region), and significant compression of the cortex and medulla layers. Fig. 1b is the BOLD image with T2* contrast. Increase of R2* is clearly visible in the cortical region of the UUO kidney (Fig. 1c). Fig. 1d shows T1 maps of a contralateral and a UUO kidney. Histology (Fig. 1g) demonstrates serious tubular damage and moderate interstitial fibrosis in the UUO kidney, which may relate to the increase of T1 shown in Fig. 1f (arrow). Compared to sham kidney, significant increases of T1 were found in the cortex (Fig. 2a) and the medulla 3~4 days and 7~8 days post UUO. Increase of R2* in the cortex (Fig. 2b) indicates renal hypoxia as a result of UUO. The preliminary GFR (Fig. 2c) and split renal function (Fig. 2d) data show a significant reduction of renal function at 3~4 days after UUO, which inversely correlates with the T1 (r= 0.63). Finally, upregulation of the RNA data (e.g., αSMA, TGFβ) confirmed the presense of fibrosis, which inversely correlates with the split renal function. The study of the in vivo quantitative imaging parameters attributed to the underlying microstructural and functional changes may better elucidate the histopathologic process that is associated with kidney injury and fibrosis in animal models.