JOURNAL OF MAGNETIC RESONANCE IMAGING 39:835–841 (2014)
Original Research
Longitudinal Changes in MRI Markers in a Reversible Unilateral Ureteral Obstruction Mouse Model: Preliminary Experience Muhammad E. Haque, PhD,1 Tammy Franklin, MS,1 Ujala Bokhary, MBBS,1 Liby Mathew, MS,2 Bradley K. Hack, PhD,2 Anthony Chang, MD,3 Tipu S. Puri, MD, PhD,2 and Pottumarthi V. Prasad, PhD1,4* Key Words: blood oxygenation level dependent; hypoxia; diffusion coefficients; fibrosis; kidney; mice J. Magn. Reson. Imaging 2014;39:835–841. C 2013 Wiley Periodicals, Inc. V
Purpose: To evaluate longitudinal changes in renal oxygenation and diffusion measurements in a model of reversible unilateral ureteral obstruction (rUUO) which has been shown to induce chronic renal functional deficits in a strain dependent way. C57BL/6 mice show higher degree of functional deficit compared with BALB/c mice. Because hypoxia and development of fibrosis are associated with chronic kidney diseases and are responsible for progression, we hypothesized that MRI measurements would be able to monitor the longitudinal changes in this model and will show strain dependent differences in response. Here blood oxygenation level dependent (BOLD) and diffusion MRI measurements were performed at three time points over a 30 day period in mice with rUUO. Materials and Methods: The studies were performed on a 4.7T scanner with the mice anesthetized with isoflurane before UUO, 2 and 28 days postrelease of 6 days of obstruction. Results: We found at the early time point (2 days after releasing the obstruction), the relative oxygenation in C57Bl/6 mice were lower compared with BALB/c. Diffusion measurements were lower at this time point and reached statistical significance in BALB/c Conclusion: These methods may prove valuable in better understanding the natural progression of kidney diseases and in evaluating novel interventions to limit progression.
1 Department of Radiology, NorthShore University HealthSystem, Evanston, Illinois, USA. 2 Department of Nephrology, University of Chicago, Chicago, Illinois, USA. 3 Department of Pathology, University of Chicago, Chicago, Illinois, USA. 4 Department of Radiology, University of Chicago, Chicago, Illinois, USA. *Address reprint requests to: P.V.P., Department of Radiology, NorthShore University HealthSystem, Walgreen Jr. Bldg., Suite G-507, 2650 Ridge Avenue, Evanston, IL 60201. E-mail: [email protected] Received January 18, 2013; Accepted May 1, 2013. DOI 10.1002/jmri.24235 View this article online at wileyonlinelibrary.com. C 2013 Wiley Periodicals, Inc. V
OVER THE PAST two decades, there have been tremendous advances in magnetic resonance imaging (MRI) technology in terms of improved image quality, spatial coverage and opportunities to probe tissue “function” or “physiological status.” Application of functional MRI methods to understand renal physiology has been rapidly evolving (1,2). Here, we have performed a preliminary evaluation of a structural and a functional MRI marker to monitor potential changes in a reversible unilateral ureteral obstruction (rUUO) model in mice. Unilateral ureteral obstruction (UUO) is a welldescribed model to study renal fibrosis (3–5). However, this model does not allow for study of longitudinal changes because the renal parenchyma is lost completely in a few weeks. A rUUO model that allows assessment of renal function after injury in parallel with structural and molecular studies was recently described (6). Using this model, it was identified that inbred strains of mice were either susceptible (C57BL/6) or resistant (BALB/c) to development of a chronic functional deficit after rUUO mediated injury (6). Susceptible mice demonstrated loss of renal function as assessed by blood urea nitrogen (BUN) levels comparable to that seen in the 5/6 nephrectomy model (7). Histologic studies demonstrated mild tubulointerstitial fibrosis and tubular atrophy after injury in kidneys from susceptible mice. Recent studies have demonstrated that diffusion MRI measurements are sensitive to the presence of fibrosis associated with mouse UUO model (8). The fibrotic changes in tissue lead to reduced freedom to water mobility within the tissue resulting in decreased apparent diffusion coefficient (ADC) values. Because hypoxia is thought to be a precursor for development of fibrosis (9–11), we hypothesized that susceptible
835
836
Haque et al.
mice should exhibit relatively more severe hypoxia compared with the resistant strain as evaluated by blood oxygenation level dependent (BOLD) MRI. BOLD MRI exploits the changes in magnetic properties of hemoglobin associated with its oxygenation status (12). This affects the T2* relaxation time of the neighboring water molecules and in turn influences the MRI signal on T2*-weighted images (13). The rate of spin dephasing R2* (¼1/T2*) is closely related to the tissue content of deoxyhemoglobin. Because the oxygen tension (pO2) of capillary blood can be thought to be in equilibrium with the surrounding tissue, changes estimated by BOLD MRI can be interpreted as changes in tissue pO2 (14,15). A strong correspondence has been demonstrated between renal BOLD MRI measurements in humans (14,15) and rodents (16) with earlier animal data obtained using invasive microelectrodes (17). More recently, direct comparisons of BOLD MRI measurements against microprobe measurements have been reported (18,19). In the present study, we evaluated ADC and BOLD MRI measurements in C57BL/6 and BALB/c mice subject to rUUO. For this preliminary study, we included 3 time points, viz., baseline (before UUO), 2, and 28 days after release of UUO.
there were variations in the post-rUUO time (i.e., days after removal of clip) of data acquisition. C57BL/6: Post_1: 0 day (n¼2); 1 day (n¼3); 2 days (n¼4) Post_2: 26–37 days BALB/c: Post_1: 1 day (n ¼ 7); 2 day (n ¼ 1); 3 days (n ¼ 2) Post_2: 28–38 days The MRI data were acquired on a 4.7 Tesla (T) (Bruker Biospec, Billerica, MA) system using a 35-mm mouse body coil (Rapid MR International, Columbus, OH). Images were acquired using respiratory gating. BOLD MRI: pulse sequence ¼ MGE, field of view (FOV) ¼ 3 4 cm, no. of slices ¼ 1, slice thickness ¼ 0.75 mm, matrix ¼ 256 256, repetition time (TR) ¼ 50 ms, no. of echoes ¼ 8 equally spaced (3.21–34.3 ms), and # averages ¼ 20. DWI MRI: was performed using either conventional spin echo (SE) with TR ¼ 3800 ms, echo time (TE) ¼ 24 ms, b-values ¼ 50, 150, 300 s/mm2 OR a multi-shot echo planar imaging (EPI) sequence with TR ¼ 500 ms, TE ¼ 31.7 ms, no. of segments ¼ 8, b-values ¼ 50, 150, 300, 500 s/ mm2. Other parameters were: FOV ¼ 3 4 cm, no. of slices ¼ 1, slice thickness ¼ 0.75mm, matrix ¼ 128 128, big delta ¼ 20.0 ms, little delta ¼ 4.0 ms. Data were acquired for trace measurement.
MATERIALS AND METHODS
C. Image Data Analysis
All procedures were performed in compliance with our institutional animal care guidelines. A total of 31 mice (19 C57BL/6 and 12 BALB/c) underwent rUUO surgical protocol for MRI measurements. An additional 28 animals also underwent rUUO protocol for the cell density measurements (Section F) at different time points.
All quantitative image data analysis was performed using Paravision software on the scanner platform. R2* (¼ 1/T2*) was calculated by drawing three circular regions of interest (ROI) 20 pixels in cortex on one of the mGE images. The software then estimated T2* by fitting an exponential decay function to the multiple echo data. The average of the three ROIs was used as the mean value. We did not evaluate the medullary region because of the difficulty in distinguishing cortex and medulla at Post_1 time point due to hydronephrosis (see Fig. 1). Additionally, R2* color maps were generated offline using Matlab (Mathwork, Natick MA) for illustration purposes. The R2* maps were scaled from zero to 75 s1, representing well oxygenated to severe hypoxia. For ADC measurement, three to six square ROIs, approximately 20 pixels each, were placed in the cortical region on one of the diffusion weighted images. The software then estimated ADC by fitting an exponential decay function to the multiple b value data. The average of the individual ROI values was used as the mean value. The longitudinal R2* and ADC imaging data were plotted and a repeated-measures analysis of variance (ANOVA) was constructed to assess the differences between BALB/c and C57BL/6 measurements as well as the change over time.
A. Surgical Procedure The right ureter was ligated using microvascular clips (S & T Inc., Foster City, CA). The clip was moved distally every 2 days and removed on the sixth day (day of rUUO). After 1 week of recovery period, the left ureter was obstructed permanently to allow for monitoring functional changes due to rUUO in the right kidney by simple blood test such as BUN. For more details, please refer to Puri et al (6). Blood samples were collected (by means of retro-orbital bleeding) for BUN analysis on the day of the placement of the clip, the day of removal of the clip, and at the end of the study. At the end of third MRI scan, the mice were euthanized and the right kidney removed for histological analysis.
B. Image Acquisition Protocol Of the 19 C57BL/6, we were able to acquire data at all three time points in nine (five died of unknown causes during the study period and five were terminated because the UUO did not reverse at Day 14). Similarly 10 of 12 BALB/c were scanned thrice (1 died and 1 was terminated). Due to logistical issues,
D. BUN Measurement Serum for BUN measurements was obtained by retroorbital bleeding. BUN concentrations were measured using a Beckman Autoanalyzer (Beckman Coulter, Fullerton, CA).
MRI Markers in Reversible UUO Mouse Model
837
Figure 1. Axial T2 weighted images obtained in a representative mouse (C57Bl/6) at 3 different time points (left to right): baseline, Post_1 (typically 2 days post reversal of UUO) and Post_2 (typically 28 days post reversal of UUO) after release of obstruction on the right kidney. Consistent with literature (20), we see a return of the anatomical structure of the right kidney 4 weeks following rUUO. In contrast, the renal parenchyma of the left kidney is completely lost 3 weeks after permanent ligation.
E. Histology Tissue Harvest Right kidneys were surgically removed under anesthesia. Before removal from the animal, the superior aspect of the mouse kidneys was inked black to facilitate orientation and correlation of microscopic and MRI images. The kidneys were bisected through a coronal section and processed as described below for routine light microscopic evaluation.
nuclei (cells) in each of five high power fields (0.14 mm2/hpf) were averaged to calculate the cell density. Because histology was available only for one time point in the present study, additional number of animals (28 each C57BL/6 and BALB/c mice) were added for cell density measurements. Animals from each strain went through identical surgical protocols and were terminated at different time points: Day 0 (n ¼ 6 each), 3 (n ¼ 5 each), 6 (n ¼ 5 each), 9 (C57BL/6 ¼ 4, BALB/c ¼ 7), 12 (C57BL/6 ¼ 8, BALB/c ¼ 5), and 36 (n ¼ 12 each from the MRI cohort).
Light Microscopy Bisected kidneys were fixed in 4% paraformaldehyde in 1 phosphate buffered saline (pH 7.4) for 24 h at 4 C. Following fixation, tissue samples were routinely processed and embedded in paraffin wax (TissuePrep II, Fisher Scientific). Tissue sections, of 5 mm thickness, were cut onto SuperfrostPlus slides (Fisher Scientific), dewaxed, and hydrated through a descending series of alcohols. Tissue sections were routinely stained by Mayer’s hematoxylin and Putt’s eosin (H&E), periodic acid-Schiff (PAS), and Masson trichrome. Stained slides were reviewed by a nephropatholgist (AC) for semi-quantitative assessment of interstitial fibrosis and tubular atrophy (IF/TA) using the following scoring system: 0 ¼ 50%. F. Cell Density Cell density measurements were carried out using the method described by Togao et al (8). Right mouse kidneys were harvested at different time points during the rUUO protocol and processed into coronal sections followed by staining with H&E as described above. Five high power fields (hpf, 200 total magnification) from each coronal section were randomly selected for the analysis of cell density using a color DP70 camera (Olympus). Non glomerular cortical fields were preferentially selected. In each selected field, the nuclei were counted automatically using the ImageJ software (National Institutes of Health, Bethesda, MD). Specifically, eight-bit color images were split and only the red channel was used. Images were thresholded to 95 and the “analyze particle” function was used based on a minimum pixel size of 100 and a circularity of greater than 0.3. Counted
RESULTS An anatomical image of the kidney from one representative C57BL/6 animal obtained at three different time points; baseline, Post_1 and Post_2 is shown in Figure 1. Consistent with literature (20), note the recovery of the anatomical structure of the right kidney at Post_2 following rUUO. In contrast, the renal parenchyma of the left kidney is completely lost 3 weeks after permanent ligation. R2* Maps of kidneys in a representative C57BL/6 and BALB/c mouse obtained at baseline, Post_1 and Post_2 are shown in Figure 2. All the maps were scaled from 0 to 75 s1 corresponding to the degree of oxygenation from well oxygenated (red) to hypoxic (blue). There is a trend toward a decrease in cortical R2* at Post_1 compared with baseline. Figure 3a shows a significant decrease in renal function in terms of an increase in BUN level in both strains at Post_2 as compared with baseline 27.5 6 3.4 mg/dL to 35.1 6 4.8 mg/dL (BALB/c) (P < 0.01), and 25.6 6 5.0 mg/dL to 51.0 6 10.0 mg/dL (C57BL/6) (P 0.01). The difference in BUN between the strains at Post_2 was significant (P < 0.05). There was no significant change in renal function at time point Post_1 in either strain probably due to functional compensation by contraleteral kidney. These results are in agreement with the previous report (6). A summary plot illustrating cortical R2* values at various time points in each strain is shown in Figure 3b. At Post_1, C57BL/6 and BALB/c showed a decrease of approximately 6 6 14% and 22 6 20% compared with baseline in cortical R2* respectively. At Post_2, C57Bl/6 showed a decrease of 7 6 16% and BALB/c showed 12 6 22% decrease compared with baseline.
838
Haque et al.
Figure 2. R2* Maps of kidneys in one representative C57BL/6 (top row) and BALB/c (bottom row) mice obtained at baseline and Post_1 (typically 2 days post reversal of UUO) and Post_2 (typically 28 days post reversal of UUO). The maps were scaled from 0 to 75 s1 corresponding to the degree of oxygenation from well oxygenated (red) to hypoxic (blue). Note the relative lowering of the R2* values at Post_1 in both strains.
The repeated-measures ANOVA analysis showed that there was significant change over time for Balb/C (overall P ¼ 0.024); post hoc pairwise comparison showed that Post_1 significantly decreased an estimated 10.38 s1 compared with baseline (Bonferroni corrected p¼0.024). There was no significant change over time for C57BL/6 (overall P ¼ 0.311). The change over time did not differ between Balb/C and C57BL/ 6, P ¼ 0.191. The change in cortical ADC from each strain at Post_1 and Post_2 time is summarized in Figure 3c. Note that approximately a third of the scans were performed using conventional SE-DWI and the rest with multishot EPI based DWI measurement. This was because of poor reproducibility and significant image quality issues with the SE-DWI sequence especially when the diffusion gradients were applied along the phase encode direction. For this plot, only data from a single direction (slice) were used to estimate ADC. At Post_1, both C57BL/6 and BALB/c showed a decrease of 7 6 38% and 32 6 39% (P ¼ 0.024) in cortical ADC respectively. At Post_2, a 24 6 60% increase in C57BL/6 and a 8 6 29% increase in BALB/c was observed in cortical ADC compared with baseline. These trends were consistent with the cell density data (shown in Fig. 4). Higher cell density results in lower ADC values and vice versa. The cell density data provide important validation for the trends observed in the ADC measurements, especially the apparent increase at Post_2 time point compared with baseline. The reduced cell density at the 36 day (i.e., 30 days post-rUUO) compared with Day 0 is consistent with the increased ADC values at Post_2. Repeated-measures ANOVA suggested that there was significant change over time in Balb/C (overall P ¼ 0.018); post hoc pair-wise comparison showed that Post_1 significantly decreased ADC, an estimated 0.
76 103 mm2/s compared with baseline (Bonferroni corrected P ¼ 0.040), and Post_2 significantly increased 0. 90 ¼ 103 mm2/s compared with Post_1 (Bonferroni corrected P ¼ 0.018). There was no significant change over time in C57BL/6 (overall P ¼ 0.232). The R2* and ADC data in the five C57BL/6 mice that did not reverse showed no difference in the mean values compared with the ones that reversed (Baseline R2* (s1): 33.664.5 versus 36.167.5 (P > 0.05); Post_1 R2*(s1): 29.362.0 versus 33.265.3(P > 0.05); Baseline ADC (103 mm2/s): 1.860.9 versus 1.960.6 (P > 0.05); Post_1 ADC (103 mm2/s): 1.3 6 0.3 versus 1.6 6 0.5 (P > 0.05)). Semi-quantitative assessment of interstitial fibrosis based on H&E, PAS, and trichrome staining were performed by a renal pathologist blinded to the origin of slides (strain) using a 0–3 scale (as defined in the Methods). C57BL/6 mice had a higher fibrosis score as compared to BALB/c mice (1.04 6 0.13 versus 0.29 6 0.17, mean 6 SEM) consistent with the higher level of CKD induced in C57BL/6 mice after rUUO. Figure 5 illustrates the differences on histology between the two strains.
DISCUSSION Our preliminary experience with the rUUO model for longitudinal MRI studies does support feasibility. The results are consistent with previous reports indicating differences in functional alteration and recovery among the two strains (6). BUN measurements documented significantly higher functional deficit at approximately 4 weeks following reversal in the C57BL/6 mice consistent with previous report (6). Previous studies reported significant level of fibrosis following irreversible UUO (8,21). Our
MRI Markers in Reversible UUO Mouse Model
histological data at end of Post_2 showed mild fibrosis in the C57BL/6 strain and little or no fibrosis in the BALB/c strain.
Figure 3. a: Summary of BUN measurements at three time points. Note minimal changes at Post_1 (typically 2 days post reversal of UUO), partly due to functioning contralateral kidney. At Post_2 (typically 28 days post reversal of UUO), both strains exhibit statistically significant increase in BUN values (by paired two tailed t-test) but C57BL/6 mice show higher values (P < 0.05) signifying higher functional deficit. b: Summary of cortical mean R2* measurements obtained from both C57BL/6 (n¼9) and BALB/c (n¼10) mice. Based on repeatedmeasures ANOVA, there was significant change over time in BALB/C (overall p¼0.024), post hoc pairwise comparison showed that Post_1 significantly decreased (*) an estimated 10.38 compared with baseline (Bonferroni corrected P ¼ 0.024). There was no significant change over time in C57BL/6 (overall P ¼ 0.311). Error bars represent standard deviations. c: Summary of cortical mean ADC measurements obtained from both C57BL/6 (n ¼ 9) and BALB/c (n ¼ 10) mice. Based on repeated-measures ANOVA, there was significant change over time in BALB/c (overall P ¼ 0.018), post hoc pairwise comparison showed that Post_1 significantly decreased (*) an estimated 0.00076 compared with baseline (Bonferroni corrected P ¼ 0.040), and Post_2 significantly increased (y) 0.000901 compared with Post_1 (Bonferroni corrected p¼0.018). There was no significant change over time in C57BL/6 (overall P ¼ 0.232). The increased ADC values at Post_2 compared with baseline in both strains is consistent with the cell density data in Figure 4. Error bars represent standard deviations.
839
BOLD MRI did document strain based differences at Post_1. The resistant BALB/c mice showed a significantly lower cortical R2* values at Post_1 (suggesting improved oxygenation), consistent with previous reports in swine (22) and human kidneys (23) with ureteral obstruction. A similar decrease in R2* was not evident in C57BL/6 mice, probably due to increased levels of hypoxia as suggested by increased levels of HIF-1a in a mouse model of UUO (24). Alternately, this may indicate compromised adaptability in C57BL/6. At Post_2, both strains showed minimal change compared with baseline and is consistent with previous reports suggesting regeneration of tissue (4,20) and hence normalization of the oxygenation status. The renal ADC measurements also showed strain based differences at Post_1. While BALB/c showed a significant decrease at Post_1, the decrease in C57Bl/ 6 did not reach significance. At Post_2 the ADC increased in both strains. A previous report (8) correlated the ADC measurements with cell density measurements based on histological analysis. As seen from Figure 4, the trends observed in cell density measurements were consistent with our ADC measurements at Post_1 and Post_2 time points. This is also consistent with the overall finding that rUUO is less of a model of fibrosis and may be more representative of chronic reduction in renal function. Because the renal parenchyma is compressed during obstruction (Fig. 1), it is possible that it could potentially influence both cell density and water content and these in turn could influence the ADC measurements. Changes in water content would also influence T2 and hence R2*. There are a few limitations of the present study. The validity of the assumption that vascular oxygenation reflects that of tissue may not be true under all circumstances (25). The oxygenation status in a UUO model may be dependent on multiple factors such as loss of capillaries (26) and the reduction in oxygen consumption due to the obstruction (22). While interpretation of BOLD MRI data following acute UUO is dominated by the reduced oxygen consumption (22,23), effects in chronic UUO may be more complicated. There has been a previous report studying the changes in renal oxygenation during UUO, but only in the acute phase (up to 3 h) owing the use of invasive microprobes (27). Future studies may be warranted to study the longitudinal changes using invasive microprobes using different groups of animals for the different time points. However, the difference in the observed R2* values between the two species is encouraging. Future studies should include independent measures of tissue hypoxia such as HIF-1a expression (24,28). BOLD MRI parameter R2* comprises of two components, R2* ¼ R2 þ R20 where R2 is the inherent T2 relaxation rate and R20 is the component associated with the magnetic susceptibility related to the oxygenation status of hemoglobin. Therefore, the observed decrease in R2* could be influenced by a potential change in water-content (as a result of mechanical pressure associated with the hydronephrosis) which could influence the regional R2. A decrease in water-content would result in an
840
Haque et al.
Figure 4. Cell Density during rUUO protocol. Shown is the mean cell density before starting the obstruction (Day 0), during obstruction (Day 3 and 6), and after release of obstruction (Day 8 (corresponding to Post_1), 13, and 36 (corresponding to Post_2)). Data are shown for both C57BL/6 (black line) and BALB/c (grey broken line) mice. Data in the parenthesis at each time show number of animals for each strain. Y-error bars indicate standard error of the mean (SEM).
increased R2 and similarly an increase in water-content would result in decreased R2. Future studies should evaluate R2 changes independently. Early studies with renal BOLD MRI did include measurement of R2 along with R2* (15) and confirmed minimal changes in R2. However, similar data may be necessary for each paradigm/model. The choice of time point Post_1 may have been sub-optimal. Future studies should probably acquire MRI data just before
release of UUO (comparable to irreversible UUO model) and within 1 day post-reversal (to better estimate the rate of structural and functional reversibility). Based on the cell density measurements, it is possible that, at 1 week after release of UUO, the difference between the two species may be more apparent on the ADC measurements. Future studies should include this time point. Based on the study in UUO model with ADC (8), future studies should
Figure 5. Top: Representative data from C57BL/6: Clustering of glomeruli within the renal cortex indicates significant tubular loss. There is also an increased extent of interstitial fibrosis as demonstrated by blue staining on Masson trichrome, which is accompanied by tubular atrophy. Bottom: Representative data from BALB/c: There is normal renal cortex without significant interstitial fibrosis or tubular atrophy. There is normal staining of tubular basement membranes with the Masson trichrome histochemical stain. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
MRI Markers in Reversible UUO Mouse Model
include multiple b values only > 300 s/mm2 and ideally acquire data to allow trace measurements. In conclusion, our preliminary study does support the feasibility of monitoring structural and functional changes associated with the reversible UUO model using MRI. Both BOLD MRI and ADC measurements showed strain based differences at early time points. Further studies are warranted to fully understand the temporal changes in both structural and functional parameters in the reversible UUO model. Based on the cell density data, 1 week post reversal of UUO may be a better time point to show larger differences between the two species. ACKNOWLEDGMENTS The authors thank Ms. Hongyan Du for providing statistical consult for the analysis of the data and Drs. Tongyu Ji and Limin Li for their technical help during MRI data acquisition. Work supported in part by funding from The University of Chicago-Northshore University Health Systems Institute for Translational Medicine (ITM) Pilot and Collaborative Translational and Clinical Studies (PCTCS) Award (to T.S.P. and P.V.P.) and an Early Career Award from the Howard Hughes Medical Institute to TSP. REFERENCES 1. Grenier N, Basseau F, Ries M, Tyndal B, Jones R, Moonen C. Functional MRI of the kidney. Abdom Imaging 2003;28:164–175. 2. Prasad PV. Functional MRI of the kidney: tools for translational studies of pathophysiology of renal disease. 2006;290:F958–F974. 3. Chevalier RL, Forbes MS, Thornhill BA. Ureteral obstruction as a model of renal interstitial fibrosis and obstructive nephropathy. Kidney Int 2009;75:1145–1152. 4. Chevalier RL, Kim A, Thornhill BA, Wolstenholme JT. Recovery following relief of unilateral ureteral obstruction in the neonatal rat. Kidney Int 1999;55:793–807. 5. Klahr S, Morrissey J. Obstructive nephropathy and renal fibrosis. Am J Physiol Renal Physiol 2002;283:F861–875. 6. Puri TS, Shakaib MI, Chang A, et al. Chronic kidney disease induced in mice by reversible unilateral ureteral obstruction is dependent on genetic background. Am J Physiol Renal Physiol 2010;298:F1024–F1032. 7. Leelahavanichkul A, Huang Y, Hu X, et al. Chronic kidney disease worsens sepsis and sepsis-induced acute kidney injury by releasing High Mobility Group Box Protein-1. Kidney Int 2011;80:1198– 1211. 8. Togao O, Doi S, Kuro-o M, Masaki T, Yorioka N, Takahashi M. Assessment of renal fibrosis with diffusion-weighted MR imaging: study with murine model of unilateral ureteral obstruction. Radiology 2010;255:772–780. 9. Eckardt KU, Bernhardt WM, Weidemann A, et al. Role of hypoxia in the pathogenesis of renal disease. Kidney Int Suppl 2005:S46–S51. 10. Fine LG, Orphanides C, Norman JT. Progressive renal disease: the chronic hypoxia hypothesis. Kidney Int Suppl 1998;65:S74–S78.
841 11. Nangaku M. Chronic hypoxia and tubulointerstitial injury: a final common pathway to end-stage renal failure. J Am Soc Nephrol 2006;17:17–25. 12. Pauling L, Coryell CD. The magnetic properties and structure of hemoglobin, oxyhemoglobin and carbonmonoxyhemoglobin. Proc Natl Acad Sci U S A 1936;22:210–216. 13. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A 1990;87:9868–9872. 14. Prasad PV, Chen Q, Goldfarb JW, Epstein FH, Edelman RR. Breath-hold R2* mapping with a multiple gradient-recalled echo sequence: application to the evaluation of intrarenal oxygenation. J Magn Reson Imaging 1997;7:1163–1165. 15. Prasad PV, Edelman RR, Epstein FH. Noninvasive evaluation of intrarenal oxygenation with BOLD MRI. Circulation 1996;94: 3271–3275. 16. Priatna A, Epstein FH, Spokes K, Prasad PV. Evaluation of changes in intrarenal oxygenation in rats using multiple gradient-recalled echo (mGRE) sequence. J Magn Reson Imaging 1999;9:842–846. 17. Brezis M, Agmon Y, Epstein FH. Determinants of intrarenal oxygenation. I. Effects of diuretics. Am J Physiol 1994;267(Pt 2):F1059–F1062. 18. dos Santos EA, Li LP, Ji L, Prasad PV. Early changes with diabetes in renal medullary hemodynamics as evaluated by fiberoptic probes and BOLD magnetic resonance imaging. Invest Radiol 2007;42:157–162. 19. Li LP, Ji L, Santos EA, Dunkle E, Pierchala L, Prasad P. Effect of nitric oxide synthase inhibition on intrarenal oxygenation as evaluated by blood oxygenation level-dependent magnetic resonance imaging. Invest Radiol 2009;44:67–73. 20. Cochrane AL, Kett MM, Samuel CS, et al. Renal structural and functional repair in a mouse model of reversal of ureteral obstruction. J Am Soc Nephrol 2005;16:3623–3630. 21. Chevalier RL, Thornhill BA, Wolstenholme JT, Kim A. Unilateral ureteral obstruction in early development alters renal growth: dependence on the duration of obstruction. J Urol 1999;161:309– 313. 22. Pedersen M, Dissing TH, Morkenborg J, et al. Validation of quantitative BOLD MRI measurements in kidney: application to unilateral ureteral obstruction. Kidney Int 2005;67:2305–2312. 23. Thoeny HC, Kessler TM, Simon-Zoula S, et al. Renal oxygenation changes during acute unilateral ureteral obstruction: assessment with blood oxygen level-dependent mr imaging--initial experience. Radiology 2008;247:754–761. 24. Ma YY, Sun D, Li J, Yin ZC. Transplantation of endothelial progenitor cells alleviates renal interstitial fibrosis in a mouse model of unilateral ureteral obstruction. Life Sci 2010;86:798–807. 25. Evans RG, Leong CL, Anderson WP, O’Connor PM. Don’t be so BOLD: potential limitations in the use of BOLD MRI for studies of renal oxygenation. Kidney Int 2007;71:1327–1328; author reply 1328. 26. Kang DH, Joly AH, Oh SW, et al. Impaired angiogenesis in the remnant kidney model: I. Potential role of vascular endothelial growth factor and thrombospondin-1. J Am Soc Nephrol 2001; 12:1434–1447. 27. Jensen AM, Norregaard R, Topcu SO, Frokiaer J, Pedersen M. Oxygen tension correlates with regional blood flow in obstructed rat kidney. J Exp Biol 2009;212:3156–3163. 28. Tanaka T, Nangaku M. The role of hypoxia, increased oxygen consumption, and hypoxia-inducible factor-1 alpha in progression of chronic kidney disease. Curr Opin Nephrol Hypertens 2010;19:43–50.
Original Research
Longitudinal Changes in MRI Markers in a Reversible Unilateral Ureteral Obstruction Mouse Model: Preliminary Experience Muhammad E. Haque, PhD,1 Tammy Franklin, MS,1 Ujala Bokhary, MBBS,1 Liby Mathew, MS,2 Bradley K. Hack, PhD,2 Anthony Chang, MD,3 Tipu S. Puri, MD, PhD,2 and Pottumarthi V. Prasad, PhD1,4* Key Words: blood oxygenation level dependent; hypoxia; diffusion coefficients; fibrosis; kidney; mice J. Magn. Reson. Imaging 2014;39:835–841. C 2013 Wiley Periodicals, Inc. V
Purpose: To evaluate longitudinal changes in renal oxygenation and diffusion measurements in a model of reversible unilateral ureteral obstruction (rUUO) which has been shown to induce chronic renal functional deficits in a strain dependent way. C57BL/6 mice show higher degree of functional deficit compared with BALB/c mice. Because hypoxia and development of fibrosis are associated with chronic kidney diseases and are responsible for progression, we hypothesized that MRI measurements would be able to monitor the longitudinal changes in this model and will show strain dependent differences in response. Here blood oxygenation level dependent (BOLD) and diffusion MRI measurements were performed at three time points over a 30 day period in mice with rUUO. Materials and Methods: The studies were performed on a 4.7T scanner with the mice anesthetized with isoflurane before UUO, 2 and 28 days postrelease of 6 days of obstruction. Results: We found at the early time point (2 days after releasing the obstruction), the relative oxygenation in C57Bl/6 mice were lower compared with BALB/c. Diffusion measurements were lower at this time point and reached statistical significance in BALB/c Conclusion: These methods may prove valuable in better understanding the natural progression of kidney diseases and in evaluating novel interventions to limit progression.
1 Department of Radiology, NorthShore University HealthSystem, Evanston, Illinois, USA. 2 Department of Nephrology, University of Chicago, Chicago, Illinois, USA. 3 Department of Pathology, University of Chicago, Chicago, Illinois, USA. 4 Department of Radiology, University of Chicago, Chicago, Illinois, USA. *Address reprint requests to: P.V.P., Department of Radiology, NorthShore University HealthSystem, Walgreen Jr. Bldg., Suite G-507, 2650 Ridge Avenue, Evanston, IL 60201. E-mail: [email protected] Received January 18, 2013; Accepted May 1, 2013. DOI 10.1002/jmri.24235 View this article online at wileyonlinelibrary.com. C 2013 Wiley Periodicals, Inc. V
OVER THE PAST two decades, there have been tremendous advances in magnetic resonance imaging (MRI) technology in terms of improved image quality, spatial coverage and opportunities to probe tissue “function” or “physiological status.” Application of functional MRI methods to understand renal physiology has been rapidly evolving (1,2). Here, we have performed a preliminary evaluation of a structural and a functional MRI marker to monitor potential changes in a reversible unilateral ureteral obstruction (rUUO) model in mice. Unilateral ureteral obstruction (UUO) is a welldescribed model to study renal fibrosis (3–5). However, this model does not allow for study of longitudinal changes because the renal parenchyma is lost completely in a few weeks. A rUUO model that allows assessment of renal function after injury in parallel with structural and molecular studies was recently described (6). Using this model, it was identified that inbred strains of mice were either susceptible (C57BL/6) or resistant (BALB/c) to development of a chronic functional deficit after rUUO mediated injury (6). Susceptible mice demonstrated loss of renal function as assessed by blood urea nitrogen (BUN) levels comparable to that seen in the 5/6 nephrectomy model (7). Histologic studies demonstrated mild tubulointerstitial fibrosis and tubular atrophy after injury in kidneys from susceptible mice. Recent studies have demonstrated that diffusion MRI measurements are sensitive to the presence of fibrosis associated with mouse UUO model (8). The fibrotic changes in tissue lead to reduced freedom to water mobility within the tissue resulting in decreased apparent diffusion coefficient (ADC) values. Because hypoxia is thought to be a precursor for development of fibrosis (9–11), we hypothesized that susceptible
835
836
Haque et al.
mice should exhibit relatively more severe hypoxia compared with the resistant strain as evaluated by blood oxygenation level dependent (BOLD) MRI. BOLD MRI exploits the changes in magnetic properties of hemoglobin associated with its oxygenation status (12). This affects the T2* relaxation time of the neighboring water molecules and in turn influences the MRI signal on T2*-weighted images (13). The rate of spin dephasing R2* (¼1/T2*) is closely related to the tissue content of deoxyhemoglobin. Because the oxygen tension (pO2) of capillary blood can be thought to be in equilibrium with the surrounding tissue, changes estimated by BOLD MRI can be interpreted as changes in tissue pO2 (14,15). A strong correspondence has been demonstrated between renal BOLD MRI measurements in humans (14,15) and rodents (16) with earlier animal data obtained using invasive microelectrodes (17). More recently, direct comparisons of BOLD MRI measurements against microprobe measurements have been reported (18,19). In the present study, we evaluated ADC and BOLD MRI measurements in C57BL/6 and BALB/c mice subject to rUUO. For this preliminary study, we included 3 time points, viz., baseline (before UUO), 2, and 28 days after release of UUO.
there were variations in the post-rUUO time (i.e., days after removal of clip) of data acquisition. C57BL/6: Post_1: 0 day (n¼2); 1 day (n¼3); 2 days (n¼4) Post_2: 26–37 days BALB/c: Post_1: 1 day (n ¼ 7); 2 day (n ¼ 1); 3 days (n ¼ 2) Post_2: 28–38 days The MRI data were acquired on a 4.7 Tesla (T) (Bruker Biospec, Billerica, MA) system using a 35-mm mouse body coil (Rapid MR International, Columbus, OH). Images were acquired using respiratory gating. BOLD MRI: pulse sequence ¼ MGE, field of view (FOV) ¼ 3 4 cm, no. of slices ¼ 1, slice thickness ¼ 0.75 mm, matrix ¼ 256 256, repetition time (TR) ¼ 50 ms, no. of echoes ¼ 8 equally spaced (3.21–34.3 ms), and # averages ¼ 20. DWI MRI: was performed using either conventional spin echo (SE) with TR ¼ 3800 ms, echo time (TE) ¼ 24 ms, b-values ¼ 50, 150, 300 s/mm2 OR a multi-shot echo planar imaging (EPI) sequence with TR ¼ 500 ms, TE ¼ 31.7 ms, no. of segments ¼ 8, b-values ¼ 50, 150, 300, 500 s/ mm2. Other parameters were: FOV ¼ 3 4 cm, no. of slices ¼ 1, slice thickness ¼ 0.75mm, matrix ¼ 128 128, big delta ¼ 20.0 ms, little delta ¼ 4.0 ms. Data were acquired for trace measurement.
MATERIALS AND METHODS
C. Image Data Analysis
All procedures were performed in compliance with our institutional animal care guidelines. A total of 31 mice (19 C57BL/6 and 12 BALB/c) underwent rUUO surgical protocol for MRI measurements. An additional 28 animals also underwent rUUO protocol for the cell density measurements (Section F) at different time points.
All quantitative image data analysis was performed using Paravision software on the scanner platform. R2* (¼ 1/T2*) was calculated by drawing three circular regions of interest (ROI) 20 pixels in cortex on one of the mGE images. The software then estimated T2* by fitting an exponential decay function to the multiple echo data. The average of the three ROIs was used as the mean value. We did not evaluate the medullary region because of the difficulty in distinguishing cortex and medulla at Post_1 time point due to hydronephrosis (see Fig. 1). Additionally, R2* color maps were generated offline using Matlab (Mathwork, Natick MA) for illustration purposes. The R2* maps were scaled from zero to 75 s1, representing well oxygenated to severe hypoxia. For ADC measurement, three to six square ROIs, approximately 20 pixels each, were placed in the cortical region on one of the diffusion weighted images. The software then estimated ADC by fitting an exponential decay function to the multiple b value data. The average of the individual ROI values was used as the mean value. The longitudinal R2* and ADC imaging data were plotted and a repeated-measures analysis of variance (ANOVA) was constructed to assess the differences between BALB/c and C57BL/6 measurements as well as the change over time.
A. Surgical Procedure The right ureter was ligated using microvascular clips (S & T Inc., Foster City, CA). The clip was moved distally every 2 days and removed on the sixth day (day of rUUO). After 1 week of recovery period, the left ureter was obstructed permanently to allow for monitoring functional changes due to rUUO in the right kidney by simple blood test such as BUN. For more details, please refer to Puri et al (6). Blood samples were collected (by means of retro-orbital bleeding) for BUN analysis on the day of the placement of the clip, the day of removal of the clip, and at the end of the study. At the end of third MRI scan, the mice were euthanized and the right kidney removed for histological analysis.
B. Image Acquisition Protocol Of the 19 C57BL/6, we were able to acquire data at all three time points in nine (five died of unknown causes during the study period and five were terminated because the UUO did not reverse at Day 14). Similarly 10 of 12 BALB/c were scanned thrice (1 died and 1 was terminated). Due to logistical issues,
D. BUN Measurement Serum for BUN measurements was obtained by retroorbital bleeding. BUN concentrations were measured using a Beckman Autoanalyzer (Beckman Coulter, Fullerton, CA).
MRI Markers in Reversible UUO Mouse Model
837
Figure 1. Axial T2 weighted images obtained in a representative mouse (C57Bl/6) at 3 different time points (left to right): baseline, Post_1 (typically 2 days post reversal of UUO) and Post_2 (typically 28 days post reversal of UUO) after release of obstruction on the right kidney. Consistent with literature (20), we see a return of the anatomical structure of the right kidney 4 weeks following rUUO. In contrast, the renal parenchyma of the left kidney is completely lost 3 weeks after permanent ligation.
E. Histology Tissue Harvest Right kidneys were surgically removed under anesthesia. Before removal from the animal, the superior aspect of the mouse kidneys was inked black to facilitate orientation and correlation of microscopic and MRI images. The kidneys were bisected through a coronal section and processed as described below for routine light microscopic evaluation.
nuclei (cells) in each of five high power fields (0.14 mm2/hpf) were averaged to calculate the cell density. Because histology was available only for one time point in the present study, additional number of animals (28 each C57BL/6 and BALB/c mice) were added for cell density measurements. Animals from each strain went through identical surgical protocols and were terminated at different time points: Day 0 (n ¼ 6 each), 3 (n ¼ 5 each), 6 (n ¼ 5 each), 9 (C57BL/6 ¼ 4, BALB/c ¼ 7), 12 (C57BL/6 ¼ 8, BALB/c ¼ 5), and 36 (n ¼ 12 each from the MRI cohort).
Light Microscopy Bisected kidneys were fixed in 4% paraformaldehyde in 1 phosphate buffered saline (pH 7.4) for 24 h at 4 C. Following fixation, tissue samples were routinely processed and embedded in paraffin wax (TissuePrep II, Fisher Scientific). Tissue sections, of 5 mm thickness, were cut onto SuperfrostPlus slides (Fisher Scientific), dewaxed, and hydrated through a descending series of alcohols. Tissue sections were routinely stained by Mayer’s hematoxylin and Putt’s eosin (H&E), periodic acid-Schiff (PAS), and Masson trichrome. Stained slides were reviewed by a nephropatholgist (AC) for semi-quantitative assessment of interstitial fibrosis and tubular atrophy (IF/TA) using the following scoring system: 0 ¼ 50%. F. Cell Density Cell density measurements were carried out using the method described by Togao et al (8). Right mouse kidneys were harvested at different time points during the rUUO protocol and processed into coronal sections followed by staining with H&E as described above. Five high power fields (hpf, 200 total magnification) from each coronal section were randomly selected for the analysis of cell density using a color DP70 camera (Olympus). Non glomerular cortical fields were preferentially selected. In each selected field, the nuclei were counted automatically using the ImageJ software (National Institutes of Health, Bethesda, MD). Specifically, eight-bit color images were split and only the red channel was used. Images were thresholded to 95 and the “analyze particle” function was used based on a minimum pixel size of 100 and a circularity of greater than 0.3. Counted
RESULTS An anatomical image of the kidney from one representative C57BL/6 animal obtained at three different time points; baseline, Post_1 and Post_2 is shown in Figure 1. Consistent with literature (20), note the recovery of the anatomical structure of the right kidney at Post_2 following rUUO. In contrast, the renal parenchyma of the left kidney is completely lost 3 weeks after permanent ligation. R2* Maps of kidneys in a representative C57BL/6 and BALB/c mouse obtained at baseline, Post_1 and Post_2 are shown in Figure 2. All the maps were scaled from 0 to 75 s1 corresponding to the degree of oxygenation from well oxygenated (red) to hypoxic (blue). There is a trend toward a decrease in cortical R2* at Post_1 compared with baseline. Figure 3a shows a significant decrease in renal function in terms of an increase in BUN level in both strains at Post_2 as compared with baseline 27.5 6 3.4 mg/dL to 35.1 6 4.8 mg/dL (BALB/c) (P < 0.01), and 25.6 6 5.0 mg/dL to 51.0 6 10.0 mg/dL (C57BL/6) (P 0.01). The difference in BUN between the strains at Post_2 was significant (P < 0.05). There was no significant change in renal function at time point Post_1 in either strain probably due to functional compensation by contraleteral kidney. These results are in agreement with the previous report (6). A summary plot illustrating cortical R2* values at various time points in each strain is shown in Figure 3b. At Post_1, C57BL/6 and BALB/c showed a decrease of approximately 6 6 14% and 22 6 20% compared with baseline in cortical R2* respectively. At Post_2, C57Bl/6 showed a decrease of 7 6 16% and BALB/c showed 12 6 22% decrease compared with baseline.
838
Haque et al.
Figure 2. R2* Maps of kidneys in one representative C57BL/6 (top row) and BALB/c (bottom row) mice obtained at baseline and Post_1 (typically 2 days post reversal of UUO) and Post_2 (typically 28 days post reversal of UUO). The maps were scaled from 0 to 75 s1 corresponding to the degree of oxygenation from well oxygenated (red) to hypoxic (blue). Note the relative lowering of the R2* values at Post_1 in both strains.
The repeated-measures ANOVA analysis showed that there was significant change over time for Balb/C (overall P ¼ 0.024); post hoc pairwise comparison showed that Post_1 significantly decreased an estimated 10.38 s1 compared with baseline (Bonferroni corrected p¼0.024). There was no significant change over time for C57BL/6 (overall P ¼ 0.311). The change over time did not differ between Balb/C and C57BL/ 6, P ¼ 0.191. The change in cortical ADC from each strain at Post_1 and Post_2 time is summarized in Figure 3c. Note that approximately a third of the scans were performed using conventional SE-DWI and the rest with multishot EPI based DWI measurement. This was because of poor reproducibility and significant image quality issues with the SE-DWI sequence especially when the diffusion gradients were applied along the phase encode direction. For this plot, only data from a single direction (slice) were used to estimate ADC. At Post_1, both C57BL/6 and BALB/c showed a decrease of 7 6 38% and 32 6 39% (P ¼ 0.024) in cortical ADC respectively. At Post_2, a 24 6 60% increase in C57BL/6 and a 8 6 29% increase in BALB/c was observed in cortical ADC compared with baseline. These trends were consistent with the cell density data (shown in Fig. 4). Higher cell density results in lower ADC values and vice versa. The cell density data provide important validation for the trends observed in the ADC measurements, especially the apparent increase at Post_2 time point compared with baseline. The reduced cell density at the 36 day (i.e., 30 days post-rUUO) compared with Day 0 is consistent with the increased ADC values at Post_2. Repeated-measures ANOVA suggested that there was significant change over time in Balb/C (overall P ¼ 0.018); post hoc pair-wise comparison showed that Post_1 significantly decreased ADC, an estimated 0.
76 103 mm2/s compared with baseline (Bonferroni corrected P ¼ 0.040), and Post_2 significantly increased 0. 90 ¼ 103 mm2/s compared with Post_1 (Bonferroni corrected P ¼ 0.018). There was no significant change over time in C57BL/6 (overall P ¼ 0.232). The R2* and ADC data in the five C57BL/6 mice that did not reverse showed no difference in the mean values compared with the ones that reversed (Baseline R2* (s1): 33.664.5 versus 36.167.5 (P > 0.05); Post_1 R2*(s1): 29.362.0 versus 33.265.3(P > 0.05); Baseline ADC (103 mm2/s): 1.860.9 versus 1.960.6 (P > 0.05); Post_1 ADC (103 mm2/s): 1.3 6 0.3 versus 1.6 6 0.5 (P > 0.05)). Semi-quantitative assessment of interstitial fibrosis based on H&E, PAS, and trichrome staining were performed by a renal pathologist blinded to the origin of slides (strain) using a 0–3 scale (as defined in the Methods). C57BL/6 mice had a higher fibrosis score as compared to BALB/c mice (1.04 6 0.13 versus 0.29 6 0.17, mean 6 SEM) consistent with the higher level of CKD induced in C57BL/6 mice after rUUO. Figure 5 illustrates the differences on histology between the two strains.
DISCUSSION Our preliminary experience with the rUUO model for longitudinal MRI studies does support feasibility. The results are consistent with previous reports indicating differences in functional alteration and recovery among the two strains (6). BUN measurements documented significantly higher functional deficit at approximately 4 weeks following reversal in the C57BL/6 mice consistent with previous report (6). Previous studies reported significant level of fibrosis following irreversible UUO (8,21). Our
MRI Markers in Reversible UUO Mouse Model
histological data at end of Post_2 showed mild fibrosis in the C57BL/6 strain and little or no fibrosis in the BALB/c strain.
Figure 3. a: Summary of BUN measurements at three time points. Note minimal changes at Post_1 (typically 2 days post reversal of UUO), partly due to functioning contralateral kidney. At Post_2 (typically 28 days post reversal of UUO), both strains exhibit statistically significant increase in BUN values (by paired two tailed t-test) but C57BL/6 mice show higher values (P < 0.05) signifying higher functional deficit. b: Summary of cortical mean R2* measurements obtained from both C57BL/6 (n¼9) and BALB/c (n¼10) mice. Based on repeatedmeasures ANOVA, there was significant change over time in BALB/C (overall p¼0.024), post hoc pairwise comparison showed that Post_1 significantly decreased (*) an estimated 10.38 compared with baseline (Bonferroni corrected P ¼ 0.024). There was no significant change over time in C57BL/6 (overall P ¼ 0.311). Error bars represent standard deviations. c: Summary of cortical mean ADC measurements obtained from both C57BL/6 (n ¼ 9) and BALB/c (n ¼ 10) mice. Based on repeated-measures ANOVA, there was significant change over time in BALB/c (overall P ¼ 0.018), post hoc pairwise comparison showed that Post_1 significantly decreased (*) an estimated 0.00076 compared with baseline (Bonferroni corrected P ¼ 0.040), and Post_2 significantly increased (y) 0.000901 compared with Post_1 (Bonferroni corrected p¼0.018). There was no significant change over time in C57BL/6 (overall P ¼ 0.232). The increased ADC values at Post_2 compared with baseline in both strains is consistent with the cell density data in Figure 4. Error bars represent standard deviations.
839
BOLD MRI did document strain based differences at Post_1. The resistant BALB/c mice showed a significantly lower cortical R2* values at Post_1 (suggesting improved oxygenation), consistent with previous reports in swine (22) and human kidneys (23) with ureteral obstruction. A similar decrease in R2* was not evident in C57BL/6 mice, probably due to increased levels of hypoxia as suggested by increased levels of HIF-1a in a mouse model of UUO (24). Alternately, this may indicate compromised adaptability in C57BL/6. At Post_2, both strains showed minimal change compared with baseline and is consistent with previous reports suggesting regeneration of tissue (4,20) and hence normalization of the oxygenation status. The renal ADC measurements also showed strain based differences at Post_1. While BALB/c showed a significant decrease at Post_1, the decrease in C57Bl/ 6 did not reach significance. At Post_2 the ADC increased in both strains. A previous report (8) correlated the ADC measurements with cell density measurements based on histological analysis. As seen from Figure 4, the trends observed in cell density measurements were consistent with our ADC measurements at Post_1 and Post_2 time points. This is also consistent with the overall finding that rUUO is less of a model of fibrosis and may be more representative of chronic reduction in renal function. Because the renal parenchyma is compressed during obstruction (Fig. 1), it is possible that it could potentially influence both cell density and water content and these in turn could influence the ADC measurements. Changes in water content would also influence T2 and hence R2*. There are a few limitations of the present study. The validity of the assumption that vascular oxygenation reflects that of tissue may not be true under all circumstances (25). The oxygenation status in a UUO model may be dependent on multiple factors such as loss of capillaries (26) and the reduction in oxygen consumption due to the obstruction (22). While interpretation of BOLD MRI data following acute UUO is dominated by the reduced oxygen consumption (22,23), effects in chronic UUO may be more complicated. There has been a previous report studying the changes in renal oxygenation during UUO, but only in the acute phase (up to 3 h) owing the use of invasive microprobes (27). Future studies may be warranted to study the longitudinal changes using invasive microprobes using different groups of animals for the different time points. However, the difference in the observed R2* values between the two species is encouraging. Future studies should include independent measures of tissue hypoxia such as HIF-1a expression (24,28). BOLD MRI parameter R2* comprises of two components, R2* ¼ R2 þ R20 where R2 is the inherent T2 relaxation rate and R20 is the component associated with the magnetic susceptibility related to the oxygenation status of hemoglobin. Therefore, the observed decrease in R2* could be influenced by a potential change in water-content (as a result of mechanical pressure associated with the hydronephrosis) which could influence the regional R2. A decrease in water-content would result in an
840
Haque et al.
Figure 4. Cell Density during rUUO protocol. Shown is the mean cell density before starting the obstruction (Day 0), during obstruction (Day 3 and 6), and after release of obstruction (Day 8 (corresponding to Post_1), 13, and 36 (corresponding to Post_2)). Data are shown for both C57BL/6 (black line) and BALB/c (grey broken line) mice. Data in the parenthesis at each time show number of animals for each strain. Y-error bars indicate standard error of the mean (SEM).
increased R2 and similarly an increase in water-content would result in decreased R2. Future studies should evaluate R2 changes independently. Early studies with renal BOLD MRI did include measurement of R2 along with R2* (15) and confirmed minimal changes in R2. However, similar data may be necessary for each paradigm/model. The choice of time point Post_1 may have been sub-optimal. Future studies should probably acquire MRI data just before
release of UUO (comparable to irreversible UUO model) and within 1 day post-reversal (to better estimate the rate of structural and functional reversibility). Based on the cell density measurements, it is possible that, at 1 week after release of UUO, the difference between the two species may be more apparent on the ADC measurements. Future studies should include this time point. Based on the study in UUO model with ADC (8), future studies should
Figure 5. Top: Representative data from C57BL/6: Clustering of glomeruli within the renal cortex indicates significant tubular loss. There is also an increased extent of interstitial fibrosis as demonstrated by blue staining on Masson trichrome, which is accompanied by tubular atrophy. Bottom: Representative data from BALB/c: There is normal renal cortex without significant interstitial fibrosis or tubular atrophy. There is normal staining of tubular basement membranes with the Masson trichrome histochemical stain. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
MRI Markers in Reversible UUO Mouse Model
include multiple b values only > 300 s/mm2 and ideally acquire data to allow trace measurements. In conclusion, our preliminary study does support the feasibility of monitoring structural and functional changes associated with the reversible UUO model using MRI. Both BOLD MRI and ADC measurements showed strain based differences at early time points. Further studies are warranted to fully understand the temporal changes in both structural and functional parameters in the reversible UUO model. Based on the cell density data, 1 week post reversal of UUO may be a better time point to show larger differences between the two species. ACKNOWLEDGMENTS The authors thank Ms. Hongyan Du for providing statistical consult for the analysis of the data and Drs. Tongyu Ji and Limin Li for their technical help during MRI data acquisition. Work supported in part by funding from The University of Chicago-Northshore University Health Systems Institute for Translational Medicine (ITM) Pilot and Collaborative Translational and Clinical Studies (PCTCS) Award (to T.S.P. and P.V.P.) and an Early Career Award from the Howard Hughes Medical Institute to TSP. REFERENCES 1. Grenier N, Basseau F, Ries M, Tyndal B, Jones R, Moonen C. Functional MRI of the kidney. Abdom Imaging 2003;28:164–175. 2. Prasad PV. Functional MRI of the kidney: tools for translational studies of pathophysiology of renal disease. 2006;290:F958–F974. 3. Chevalier RL, Forbes MS, Thornhill BA. Ureteral obstruction as a model of renal interstitial fibrosis and obstructive nephropathy. Kidney Int 2009;75:1145–1152. 4. Chevalier RL, Kim A, Thornhill BA, Wolstenholme JT. Recovery following relief of unilateral ureteral obstruction in the neonatal rat. Kidney Int 1999;55:793–807. 5. Klahr S, Morrissey J. Obstructive nephropathy and renal fibrosis. Am J Physiol Renal Physiol 2002;283:F861–875. 6. Puri TS, Shakaib MI, Chang A, et al. Chronic kidney disease induced in mice by reversible unilateral ureteral obstruction is dependent on genetic background. Am J Physiol Renal Physiol 2010;298:F1024–F1032. 7. Leelahavanichkul A, Huang Y, Hu X, et al. Chronic kidney disease worsens sepsis and sepsis-induced acute kidney injury by releasing High Mobility Group Box Protein-1. Kidney Int 2011;80:1198– 1211. 8. Togao O, Doi S, Kuro-o M, Masaki T, Yorioka N, Takahashi M. Assessment of renal fibrosis with diffusion-weighted MR imaging: study with murine model of unilateral ureteral obstruction. Radiology 2010;255:772–780. 9. Eckardt KU, Bernhardt WM, Weidemann A, et al. Role of hypoxia in the pathogenesis of renal disease. Kidney Int Suppl 2005:S46–S51. 10. Fine LG, Orphanides C, Norman JT. Progressive renal disease: the chronic hypoxia hypothesis. Kidney Int Suppl 1998;65:S74–S78.
841 11. Nangaku M. Chronic hypoxia and tubulointerstitial injury: a final common pathway to end-stage renal failure. J Am Soc Nephrol 2006;17:17–25. 12. Pauling L, Coryell CD. The magnetic properties and structure of hemoglobin, oxyhemoglobin and carbonmonoxyhemoglobin. Proc Natl Acad Sci U S A 1936;22:210–216. 13. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A 1990;87:9868–9872. 14. Prasad PV, Chen Q, Goldfarb JW, Epstein FH, Edelman RR. Breath-hold R2* mapping with a multiple gradient-recalled echo sequence: application to the evaluation of intrarenal oxygenation. J Magn Reson Imaging 1997;7:1163–1165. 15. Prasad PV, Edelman RR, Epstein FH. Noninvasive evaluation of intrarenal oxygenation with BOLD MRI. Circulation 1996;94: 3271–3275. 16. Priatna A, Epstein FH, Spokes K, Prasad PV. Evaluation of changes in intrarenal oxygenation in rats using multiple gradient-recalled echo (mGRE) sequence. J Magn Reson Imaging 1999;9:842–846. 17. Brezis M, Agmon Y, Epstein FH. Determinants of intrarenal oxygenation. I. Effects of diuretics. Am J Physiol 1994;267(Pt 2):F1059–F1062. 18. dos Santos EA, Li LP, Ji L, Prasad PV. Early changes with diabetes in renal medullary hemodynamics as evaluated by fiberoptic probes and BOLD magnetic resonance imaging. Invest Radiol 2007;42:157–162. 19. Li LP, Ji L, Santos EA, Dunkle E, Pierchala L, Prasad P. Effect of nitric oxide synthase inhibition on intrarenal oxygenation as evaluated by blood oxygenation level-dependent magnetic resonance imaging. Invest Radiol 2009;44:67–73. 20. Cochrane AL, Kett MM, Samuel CS, et al. Renal structural and functional repair in a mouse model of reversal of ureteral obstruction. J Am Soc Nephrol 2005;16:3623–3630. 21. Chevalier RL, Thornhill BA, Wolstenholme JT, Kim A. Unilateral ureteral obstruction in early development alters renal growth: dependence on the duration of obstruction. J Urol 1999;161:309– 313. 22. Pedersen M, Dissing TH, Morkenborg J, et al. Validation of quantitative BOLD MRI measurements in kidney: application to unilateral ureteral obstruction. Kidney Int 2005;67:2305–2312. 23. Thoeny HC, Kessler TM, Simon-Zoula S, et al. Renal oxygenation changes during acute unilateral ureteral obstruction: assessment with blood oxygen level-dependent mr imaging--initial experience. Radiology 2008;247:754–761. 24. Ma YY, Sun D, Li J, Yin ZC. Transplantation of endothelial progenitor cells alleviates renal interstitial fibrosis in a mouse model of unilateral ureteral obstruction. Life Sci 2010;86:798–807. 25. Evans RG, Leong CL, Anderson WP, O’Connor PM. Don’t be so BOLD: potential limitations in the use of BOLD MRI for studies of renal oxygenation. Kidney Int 2007;71:1327–1328; author reply 1328. 26. Kang DH, Joly AH, Oh SW, et al. Impaired angiogenesis in the remnant kidney model: I. Potential role of vascular endothelial growth factor and thrombospondin-1. J Am Soc Nephrol 2001; 12:1434–1447. 27. Jensen AM, Norregaard R, Topcu SO, Frokiaer J, Pedersen M. Oxygen tension correlates with regional blood flow in obstructed rat kidney. J Exp Biol 2009;212:3156–3163. 28. Tanaka T, Nangaku M. The role of hypoxia, increased oxygen consumption, and hypoxia-inducible factor-1 alpha in progression of chronic kidney disease. Curr Opin Nephrol Hypertens 2010;19:43–50.
