Robert Greite,1* Anja Thorenz,1* Rongjun Chen,1Mi-Sun Jang,1Song Rong,1,3Michael J. Brownstein,7 Susanne Tewes,2Li Wang,1Bita Baniassad,1Torsten Kirsch,1Jan Hinrich Bräsen,5Ralf Lichtinghagen,6 Martin Meier,4Hermann Haller,1Katja Hueper,2* and Faikah Gueler1*
1Nephrology, Hannover Medical School, Hannover, Germany;2Diagnostic and Interventional Radiology, Hannover Medical School, Hannover, Germany;3The Transplantation Center of the Affiliated Hospital, Zunyi Medical College, Zunyi, China;
4Imaging Center, Institute of Laboratory Animal Sciences, Hannover Medical School, Hannover, Germany;5Pathology, Hannover Medical School, Hannover, Germany;6Clinical Chemistry, Hannover Medical School, Hannover, Germany; and
7Pisces Therapeutics, Rockville, Maryland
Submitted 21 September 2016; accepted in final form 19 December 2017
Greite R, Thorenz A, Chen R, Jang MS, Rong S, Brownstein MJ, Tewes S, Wang L, Baniassad B, Kirsch T, Bräsen JH, Lichtinghagen R, Meier M, Haller H, Hueper K, Gueler F.Renal ischemia-reperfusion injury causes hypertension and renal perfusion impairment in the CD1 mice which promotes progressive renal fibrosis. Am J Physiol Renal Physiol 314: F881–F892, 2018. First published December 20, 2017; doi:10.1152/ajprenal.00519.2016.—
Renal ischemia-reperfusion injury (IRI) is a severe complication of major surgery and a risk factor for increased morbidity and mortality.
Here, we investigated mechanisms that might contribute to IRI-induced progression to chronic kidney disease (CKD). Acute kidney injury (AKI) was induced by unilateral IRI for 35 min in CD1 and C57BL/6 (B6) mice. Unilateral IRI was used to overcome early mortality. Renal morphology, NGAL upregulation, and neutrophil infiltration as well as peritubular capillary density were studied by immunohistochemistry. The composition of leukocyte infiltrates in the kidney after IRI was investigated by flow cytometry. Systemic blood pressure was measured with a tail cuff, and renal perfusion was quantified by functional magnetic resonance imaging (fMRI). Mesan-gial matrix expansion was assessed by silver staining. Following IRI, CD1 and B6 mice developed similar morphological signs of AKI and increases in NGAL expression, but neutrophil infiltration was greater in CD1 than B6 mice. IRI induced an increase in systemic blood pressure of 20 mmHg in CD1, but not in B6 mice; and CD1 mice also had a greater loss of renal perfusion and kidney volume than B6 mice (P⬍0.05). CD1 mice developed substantial interstitial fibrosis and decreased peritubular capillary (PTC) density by day 14 while B6 mice showed only mild renal scarring and almost normal PTC. Our results show that after IRI, CD1 mice develop more inflammation, hypertension, and later mesangial matrix expansion than B6 mice do.
Subsequently, CD1 animals suffer from CKD due to impaired renal perfusion and pronounced permanent loss of peritubular capillaries.
acute kidney injury; CKD; fibrosis; hypertension; ischemia-reperfu-sion injury
INTRODUCTION
Acute kidney injury (AKI) can either regenerate or cause ongoing renal fibrosis (8). The extent of renal damage is influenced by the onset of hypertension, local inflammation, and ongoing nephrotoxic medication. The most common mouse model to study AKI is renal ischemia-reperfusion injury (IRI) (5, 6).
The majority of renal IRI studies have been performed in C57BL/6 (B6) laboratory mice, which do not develop hyper-tension after renal injuries. However, CD1 mice have been shown to develop marked blood pressure elevation and glo-merular injury (16) following partial nephrectomy models.
Similarly, a subgroup of patients developed hypertension after renal IRI and worse clinical outcome (4). To learn more about the effect of hypertension on IRI, we compared B6 (12) and CD1 mice in the 35-min renal pedicle clamping IRI model. To overcome mortality and be able to study the pathophysiology of AKI and chronic kidney disease (CKD) longitudinally, unilateral IRI was done. Systemic blood pressure was correlated with mesangial matrix expansion and also with renal perfusion measured by arterial spin labeling (ASL), which is a functional magnetic resonance imaging (fMRI) technique. Renal pathology, inflammation, peritubular capillary density, and fibrosis were studied by immunohistochemistry.
METHODS
Animals.Adult male CD1 mice (30 –35 g; 8 –10 wk of age) and male C57BL/6N (B6) mice (25–30 g; 10 –11 wk of age) were purchased from Charles River (Sulzfeld, Germany) and used for all experiments. Mice were housed under conventional conditions in individually ventilated cages with a 14:10-h light-dark cycle and had free access to food (Altromin 1324 standard mouse diet) and domes-tic-quality drinking water. After surgery mice were monitored daily for physical condition. Animals were cared for in accordance with our institutions guidelines for experimental animals, and all experiments had been approved by the local animal protection committee of the Lower Saxony State department for animal welfare and food protec-tion. The German guidelines are in accordance with the National Institutes of Health guidelines for animal welfare.
For functional MRI studies, unilateral IRI was performed inn⫽5 CD1 mice andn⫽10 B6 mice. Animals were examined
longitudi-* R. Greite, A. Thorenz, K. Hueper, and F. Gueler contributed equally to this work.
Address for reprint requests and other correspondence: F. Gueler, Nephrol-ogy, Hannover Medical School, Carl-Neuberg-Str.1, 30625 Hanover, Germany (e-mail: gueler.faikah@mh-hannover.de).
Am J Physiol Renal Physiol314: F881–F892, 2018.
First published December 20, 2017; doi:10.1152/ajprenal.00519.2016.
1931-857X/18 Copyright©2018 the American Physiological Society
http://www.ajprenal.org F881
Downloaded from www.physiology.org/journal/ajprenal at Biblio der Medizinischen Hochschule (193.174.111.250) on March 25, 2019.
nally atday 1andday 14after IRI. Histology for renal morphology and immunohistochemistry was done in an additionaln⫽8 –10 mice in each group at days 1, 14,and 28after IRI. Blood pressure was measured by tail cuff inn⫽6 mice in each group with a coda device (Kent Scientific).
Renal ischemia-reperfusion injury to induce AKI. IRI was per-formed in general anesthesia with isoflurane (2% for induction and 1.5% for maintenance). Mice received butorphanol (1 mg/kg sc) for analgesia. Renal pedicles were clamped with micro aneurysm clips (Aesculap, Germany) for 35 min, and afterward reperfusion was allowed and visually controlled. Mice were monitored until fully awake before they were transferred to the animal facility.
MRI to measure renal perfusion by arterial spin labeling (ASL) and kidney volume loss.MRI examinations were performed in isoflurane inhalation anesthesia using a 7-T small animal scanner (Bruker, Pharmascan). For visualization of renal morphology and quantifica-tion of kidney volumes, respiratory triggered, fat-saturated T2-weighted sequences were acquired. Kidney volumes of the ischemic kidney and the contralateral kidney without IRI were determined by manual segmentation. Renal perfusion was measured without admin-istration of contrast agent using an arterial spin labeling (ASL) technique as described previously (12). Renal perfusion maps were calculated and mean local perfusion in the renal cortex was quantified in the IRI kidney and in the contralateral kidney without IRI in milliliters per minute per 100 g renal tissue. Total perfusion for each kidney was then determined by multiplying local perfusion values times the kidney volume. Furthermore, relative kidney volume and relative renal perfusion of the kidney with IRI were calculated as the percentage compared with the contralateral control kidney without IRI.
Renal morphology.Kidneys were harvested ondays 1, 14, and28 after IRI. The middle part of the kidney was immediately fixed in 4%
paraformaldehyde (PFA) and embedded in paraffin; another part was shock frozen in liquid nitrogen for Western blot analysis. Two-micrometer paraffin sections were cut and stained with PAS according to standard diagnostic protocols. AKI scoring was done using a semiquantitative grading system: 0⫽focal AKI with⬍5% of tubuli of the cortex affected; 1⫽mild AKI with 5–25% of tubuli affected;
2⫽moderate AKI with 26 –50% of tubuli affected; 3⫽severe AKI with 51–75% of the tubuli affected; 4⫽very severe AKI with⬎75%
of tubuli affected. Masson trichrome staining (standard diagnostic protocol) was done as described previously at 14 and 28 days after IRI to estimate tubulointerstitial lesions (% of total cross sectional area,
“eye-balling” by two observers, RG and JHB). To investigate inter-stitial collagen deposition paraffin sections were stained with sirius red using standard protocols (12). For quantification regions of inter-est (ROI) were defined and large vessels and glomeruli were excluded from the analysis. The proportion of the sirius red-positive area was analyzed automatically by ImageJ software as percent of the area of interest. Immunohistochemistry was done on paraffin sections using the following primary antibodies: CD31 to stain for peritubular capillaries (Dianova), NGAL for tubular damage (Bioporto), GR-1 for neutrophils (Dianova), F4/80 for macrophages (Biolegend), and Ki-67 for cell proliferation (abcam). Antigen retrieval was achieved by incubating sections with trypsin for 15 min at 37°C or by microwav-ing (15 min at 600 W). Primary antibodies were incubated for 60 min at room temperature in the dark. Alexa Fluor-conjugated secondary antibodies for fluorescent visualization of bound primary antibodies were incubated for an additional 60 min in the dark. Analysis was performed in a blinded manner using a Leica imaging microscope. To assess NGAL expression, 10 different view fields per renal section were evaluated for the proportion of NGAL stained tubuli. Ki-67-positive tubular epithelial cells were counted in 10 high power fields per renal section. CD31-positive peritubular capillaries were assessed semiquantitatively in percent of the affected area in 10 view fields per section. Gr-1- and F4/80-positive leukocytes were scored semiquan-titatively: 0⫽less than 5 cells per view field (VF); 1⫽mild
infil-trates of up to 10 cells/VF; 2⫽moderate infiltration 11–20 cells/VF;
3⫽21–50 cells/VF; 4⫽greater than 50 cells/VF. Normal peritubular capillary network in healthy kidneys was set as 100%. Mesangial thickness was measured in 10 different glomeruli of one kidney section using Axiovision software (Rel. 4.2, Zeiss Oberkochen, Ger-many). Total kidney images were generated by the Keyence micro-scope software with stack function. All images were captured with the same window size.
Western blotting. For Western blotting, the frozen kidneys were pulverized in liquid nitrogen and resuspended in 2 ml lysis buffer (20 mM Tris buffer, pH 7.5 containing 10 mM glycerolphosphate, 2 mM pyrophosphate, 1 mM sodium fluoride, 1 mM PMSF, 1 g/ml leupep-tin, 1 mM DTT, 1 mM EDTA). Homogenates were sonicated for three 20-s bursts on ice and centrifuged at 500gfor 1 min to remove cell debris. Aliquots of the supernatants were stored at ⫺80°C. The protein amount was measured by BCA assay. Fifty micrograms protein of each sample was suspended in loading buffer and run on a 10% polyacrylamide gel and electrophoretically transferred to nitro-cellulose membrane. Membranes were blocked in 5% skim milk and 1% BSA. Primary antibody against NGAL (Bioporto) was applied overnight at 4°C. After washing with TBST buffer (50 mmol/l Tris·HCl, pH 7.5, 150 mmol/l NaCl, 0.01% Tween 20), incubation with horseradish peroxidase-conjugated goat anti rabbit secondary antibody (Dianova, Hamburg, Germany) for 1 h at room temperature was performed. The protein bands in the blot were detected with the use of an enhanced chemoluminescense kit (Renaissance, NEN Life Science, Zaventem, Belgium) according to the manufacturer’s instruc-tions. Relative density measurements were done using ImageJ soft-ware (NIH).
RNA extraction and real-time quantitative PCR. Kidney sections were stored in RNA-later immediately after organ retrieval. Total RNA was extracted using the RNeasy mini kit system (Qiagen, Hilden, Germany). Total RNA was isolated using Quiagen mini-kits.
For quantitative PCR (qPCR), 1g of DNase-treated total RNA was transcribed using Superscript II Reverse transcriptase (Invitrogen), and qPCR was performed on an Lightcycler 420 II (Roche Diagnos-tics, Penzberg, Germany) using FastStart Sybr-Green chemistry.
Gene-specific primers for connective tissue growth factor (CTGF;
QT00096131) and PAI-1 were obtained from Qiagen, and results were normalized to expression of hypoxanthine guanine phosphoribosyl transferase (HPRT; QT00166768). Quantification was carried out using QGene software (20).
Tail-cuff blood pressure measurement. Systolic blood pressure (SBP) was measured with a noninvasive tail-cuff blood pressure measurement system (CODA, Kent Scientific; Torrington, CT).
Volume pressure recording technology was used to detect changes in tail blood flow volume that corresponds to SBP and DBP.
Animals were trained to the restrainers on the warming plate on at least three different days before start of the study. Each ment contained 5 training inflations and 15 experimental measure-ments.
Flow cytometry.Atday 1after IRI mice were euthanized, and 70%
of the kidneys were used for flow cytometry analysis. After digestion in Dulbecco’s modified medium (Merck Millipore, FG0415) contain-ing 500 U/ml Collagenase II (Worthcontain-ington, 42K13638) for 20 min at 37°C, kidney pieces were homogenized on a gentleMACS dissociator (Miltenyi Biotec) and incubated at 37°C again. Samples were washed 3 times by straining through a 70-m mesh with PBS and red blood cell (RBC) lysis with lysis buffer (Biolegend) was performed. Live/
dead staining was conducted using Fixable Viability Dye eFluor506 (ebioscience, 65– 0866) 1:1,000 in PBS for 30 min in the dark at 4°C.
Cells were washed with FACS buffer (PBS with 0.5% FCS) and incubated for 20 min in the dark at 4°C in the antibody mix prepared with FACS buffer. The following antibodies were used: 1:100 anti-mouse CD49b FITC (ebioscience,clone Dx5), 1:400 anti-anti-mouse CD11c PerCP-Cyanine5.5 (ebioscience, clone n418), 1:600 anti-mouse CD11b APC-eFluor780 (ebioscience, clone M1/70), 1:600
F882 AKI INDUCED HYPERTENSION AND CKD
AJP-Renal Physiol•doi:10.1152/ajprenal.00519.2016•www.ajprenal.org
Downloaded from www.physiology.org/journal/ajprenal at Biblio der Medizinischen Hochschule (193.174.111.250) on March 25, 2019.
anti-mouse CD45 Pacific Blue (Biolegend, clone 30-F11), 1:600 anti-mouse Ly-6C PE (Biolegend, clone HK1.4), 1:600 anti-mouse Ly-6G PE/Cy7 (Biolegend, clone 1A8), 1:100 anti-mouse F4/80 APC (ebioscience, BM8), and 1:100 anti-mouse CD16/CD32 (ebioscience, clone 93). Flow cytometry was then performed with a BD FACSCanto II. For analysis of the cells, Kaluza Software was used. At least an amount of 50,000 cells were analyzed.
Statistical analysis.For statistical analysis, GraphPad prism soft-ware, version 5.0c (GraphPad Softsoft-ware, San Diego, CA) was used.
For multiple comparisons, ANOVA with post hoc Tukey correction was applied. For comparison of two groups, Student’st-test was used.
SBP was correlated with fibrosis and mesangial matrix thickness using Spearman’s correlation. Differences were considered significant at aPvalue⬍0.05. Data are presented as means⫾SE.
Fig. 1. Periodic acid-Schiff (PAS) stain atday 1revealed more AKI in CD 1 than in B6 mice with pronounced damage in the outer medulla (OM; magnification 200⫻,A, D, G; CD1,n⫽8 mice; B6,n⫽10 mice). Sham control mice revealed normal renal morphology in both strains (not shown). Immunohistochemistry (B, E, H) and Western blotting (J, K) showed similar expression of the tubular damage marker NGAL in both mouse strains. Western blot was repeated in three independent experiments withn⫽3 mice in each group (CD1 IRI, black bars; B6 IRI, white bars). Significant upregulation of Ki-67 in proliferating tubular epithelial cells was seen atday 1in B6 mice but not in CD1 mice (C, F, I). TNF-alpha mRNA expression was elevated atday 1in both strains and significantly higher in CD1 mice atday 7(L; qPCR was repeated twice inn⫽5 mice). *P⬍0.05, **P⬍0.01, ***P⬍0.001.
F883
AKI INDUCED HYPERTENSION AND CKD
AJP-Renal Physiol•doi:10.1152/ajprenal.00519.2016•www.ajprenal.org
Downloaded from www.physiology.org/journal/ajprenal at Biblio der Medizinischen Hochschule (193.174.111.250) on March 25, 2019.
***
* ***
Gr-1 score d1 F4/80 score d7
Gr-1 d1
IRI-CD1
F4/80 d7
IRI-B6control
D C
F G
J I
K L
d1 d7
IRI CD1 IRI B6 d1 d7 CD1 B6
control
d1 d7
IRI CD1 IRI B6 d1 d7 CD1 B6
control
***
neutrophils/ 100mg kidney wet weight macrophages/ 100mg kidney wet weight
** **
control CD1
*** ***
CD11b
Ly6G
neutrophils
B A
E
H
*
CD11b
F4/80
macrophages
IRI control IRI B6 control
CD1
IRI control IRI B6
F884 AKI INDUCED HYPERTENSION AND CKD
AJP-Renal Physiol•doi:10.1152/ajprenal.00519.2016•www.ajprenal.org
Downloaded from www.physiology.org/journal/ajprenal at Biblio der Medizinischen Hochschule (193.174.111.250) on March 25, 2019.
RESULTS
AKI and tubular proliferation after IRI.CD1 showed more renal damage with loss of the brush border, tubular epithelial cell flattening, detachment of tubular epithelial cells, cast formation, and obstruction of the tubuli 1 day after IRI. The AKI score was significantly higher in CD1 compared with B6 mice (Fig. 1,A, D, andG). In addition, NGAL expression as a tubular damage marker was enhanced in the cortical tubuli of both groups (Fig. 1, B, E, and H). In line with the immuno-histochemistry, Western blotting for NGAL did not reveal differences between groups (Fig. 1,JandK). Tubular epithelial cell proliferation is a hallmark of regeneration after IRI. There-fore, Ki-67⫹ tubular epithelial cells were counted at the dif-ferent time points. Within 24 h of IRI, these cells were already more abundant in B6 than CD1 mice (Fig. 1,C,F, andI). The initial upregulation of TNF-␣ on day 1 was similar in both
strains. Onday 7TNF-␣expression was significantly higher in CD1 than B6 mice, suggesting that inflammation and proin-flammatory cytokine production are important in the former.
The increase in IL-6 mRNA was similar in both mouse strains (data not shown).
Inflammation after IRI.Neutrophils are the first leukocytes to invade the tissue in IRI. One day after IRI, whole renal tissue samples from CD1 had more CD11b⫹/Ly6G⫹neutrophils than samples from B6 kidneys (Fig. 2A). By day 7, the neutrophil numbers were reduced markedly in both strains and did not show significant differences between groups (Fig. 2A). Immuno-histochemistry confirmed this result and showed that GR-1⫹ -positive neutrophils were located mainly in the outer medulla.
The contralateral kidneys, which did not undergo ischemia, served as controls. The neutrophil count was very low and comparable in CD1 and in B6 controls. Shortly after IRI,
Fig. 2. Flow cytometry of isolated leukocytes showed that IRI CD1 kidneys (black bar) had significantly more infiltrating neutrophils (A; CD11b⫹/Ly6G⫹ cells/100 mg kidney) than IRI B6 kidneys (white bar) atdays 1and7. Macrophage infiltration (B, CD11b⫹/F4/80⫹cells/100 mg kidney) was increased atday 1and even more atday 7after IRI and was significantly more pronounced in CD1 compared with B6 IRI kidneys. Representative examples for neutrophil (day 1) and macrophage (day 7) gating after IRI are shown (E, H). Contralateral kidneys of both strains showed similar low baseline values (n⫽6 in each group per time point). Immunohistochemistry for Gr-1-positive neutrophils (C, F, I; left column) revealed infiltration of the outer medulla atday 1. Macrophages (D, G, J; right column) accumulated in the outer medulla but also in the cortex (magnification 200-fold; quantificationK, L;n⫽6 in each group).
*P⬍0.05, **P⬍0.01, ***P⬍0.001.
Fig. 3. Color-coded maps of local renal per-fusion of the IRI kidney (left side) and the contralateral untouched kidney (right side) at day 1andday 14after IRI are shown in CD1 (top row,A, B;n⫽5) and in B6 mice (bottom row,C, D;n⫽10). Note that perfusion of the untouched kidney onday 1(reference perfu-sion) is higher in CD1 than B6 animals. At both time points a clear difference of renal perfusion between IRI and contralateral kid-neys could be observed in both mouse strains, which is, however, more pronounced in CD1 mice. Means⫾SE of relative perfusion in the IRI kidneys at different time points and in different groups are given (E), and significant differences are indicated (*P⬍0.05).
F885
AKI INDUCED HYPERTENSION AND CKD
AJP-Renal Physiol•doi:10.1152/ajprenal.00519.2016•www.ajprenal.org
Downloaded from www.physiology.org/journal/ajprenal at Biblio der Medizinischen Hochschule (193.174.111.250) on March 25, 2019.
neutrophils were replaced by macrophages. Flow cytometry revealed more F4/80⫹ macrophages in IRI CD1 kidneys at days 1 and 7than in B6 kidneys (P ⬍ 0.01). Immunohisto-chemistry atday 1revealed small numbers of F4/80⫹ macro-phages. More of the cells were seen byday 7, but no significant differences were seen between mouse strains. At the day 14 time point, macrophages remained elevated in both mouse strains.
Functional MRI to measure renal perfusion impairment.On days 1 and 14, relative perfusion of the IRI kidney was compared with perfusion of the control (unclipped) day 1 kidney in each animal using fMRI. On day 1, relative renal perfusion of the IRI kidney was lower in CD1 than B6 mice (51⫾9% vs. 75⫾8%, not significantP⫽0.09). Onday 14, relative renal perfusion of the IRI kidney significantly de-creased in CD1 vs. B6 mice (24⫾6% vs. 61⫾9%, P ⫽ 0.016; Fig. 3). Baseline perfusion values per kidney were significantly higher in CD1 than in B6 mice (2.38⫾0.37 vs.
0.96⫾0.06 ml/min per kidney, P⬍0.001).
Peritubular capillaries.Peritubular capillaries were stained with CD31 antibodies (Fig. 4). CD31 is expressed ubiquitously within the vascular compartment and is located mainly at junctions between adjacent cells. CD31 immunostaining is used as a marker for intact capillaries. Loss of CD31 staining reflects disturbances of the integrity of microvasculature and occurs in areas with enhanced fibrosis. At baseline, CD1 (top row) and B6 (middle row) mice had similar capillary densities, but 2 wk after IRI, CD1 mice had less CD31⫹ capillary staining than B6 mice did. This difference was even more pronounced 4 wk after IRI. By this time B6 mice had normal peritubular capillary densities (Fig. 4I).
Progression to CKD. Relative kidney volumes were mea-sured using MRI data (Fig. 5). In CD1 (top row) and B6 (middlerow) mice, kidney volumes initially increased in par-allel by 14⫾8% and 18⫾7%, respectively. This might reflect the edema formation and enhanced water content of the tissue.
Fourteen days after IRI, kidney volume loss (relative to the contralateral control) in CD1 mice was significantly greater
CD1
Fig. 4. Renal capillaries were visualized by CD31 staining and peritubular capillary density was assessed at 14 days and 4 wk after IRI. The first column shows
Fig. 4. Renal capillaries were visualized by CD31 staining and peritubular capillary density was assessed at 14 days and 4 wk after IRI. The first column shows