Background: Contrast medium-induced acute kidney injury (CI-AKI) is one of the most common causes of hospital-acquired acute renal failure. However, the pathogenesis of CI-AKI remains unclear. Asymmetric dimethylarginine (ADMA) is an endogenous nitric oxide synthase (NOS) inhibitor that is largely metabolised by dimethylarginine dimethylaminohydroxylase (DDAH) in humans. Two isoforms of DDAH exist, namely, DDAH-1 and DDAH-2. In the present study, we examined whether the DDAH-2/ADMA/NOS pathway is involved in the pathogenesis of CI-AKI.
Methods and Results: Exposure to the contrast medium iopromide led to increase in creatinine and blood urea nitrogen (BUN) levels, accumulation of ADMA, increase in reactive oxygen species (ROS) generation, and an inflammatory response in mice kidney tissue. The injection of adenovirus-harbouring DDAH-2 lowered renal ADMA levels and had a reno-protective effect against contrast-medium injury by decreasing cell apoptosis, ROS, and fibrosis. By contrast, contrast medium-induced renal injury was exacerbated in heterozygous DDAH-2 knockout mice. In the in vitro study, overexpression of DDAH-2 increased the levels of nitrite and intracellular cGMP, while the DDAH-2 knockdown induced the opposite effect. These findings were also observed in the in vivo sample.
Conclusions: Our findings provide the first evidence that the DDAH-2/ADMA/NOS pathway is involved in the pathogenesis of CI-AKI and that the protective effect of DDAH-2 probably arises from the modulation of NOS activity, oxidative stress, and the inflammatory process.
Contrast medium-induced acute kidney injury (CI-AKI) is a complication that occurs after cardiac catheterisation and may be associated with an increased risk of in-hospital and long-term adverse events [1,2]. The pathophysiology of CI-AKI is multifaceted and not entirely clear. Renal vasoconstriction and medullary hypoxia, partly mediated by a decrease in nitric oxide (NO) bioavailability and an increase in oxidative stress, may be contributing mechanisms . Asymmetric dimethylarginine (ADMA) is a well-characterised circulating endogenous inhibitor of NO synthase (NOS) [4,5]. This inhibitor may increase oxidative stress by uncoupling electron transport between NOS and l-arginine, thereby decreasing both the production and availability of endothelium-derived NO and thus leading to endothelial dysfunction [6,7]. In an animal study, microperfusion with ADMA inhibited NOS in the macula densa and enhanced the maximal tubuloglomerular feedback response, leading to vasoconstriction of the afferent arterioles, which in turn caused a decrease in the glomerular filtration rate and flow . In elderly patients, the accumulation of ADMA was associated with increased renovascular resistance and decreased renal perfusion . Several studies have shown that plasma ADMA levels may predict the progression of renal injury in patients in the early stages of chronic kidney disease, independent of the estimated glomerular filtration rate and other traditional risk factors [10,11]. Moreover, we and other researchers have shown that elevated plasma ADMA levels might be associated with increased risk of CI-AKI following cardiac catheterisation . In sum, evidence suggests that ADMA might be involved in the development of CI-AKI, but the underlying molecular mechanism has seldom been explored.
In the human body, 80% of ADMA is metabolised to citrulline by dimethylarginine dimethylaminohydroxylase (DDAH), and the remainder is excreted by the kidneys . The two isoforms of DDAH are DDAH-1 and DDAH-2. Knockdown of DDAH-2 has been reported to blunt the endothelium-derived relaxing factor and NO responses of mesenteric resistance arteries to acetylcholine , whereas the overexpression of DDAH-2 reduces ADMA levels, affects tissue NO metabolism, and prevents ADMA- and angiotensin II-induced vascular lesions . In the kidneys, DDAH-1 is expressed mainly in the renal proximal tubules, whereas DDAH-2 is expressed strongly in the endothelium, vascular smooth muscle cells, and adventitia of blood vessels, as well as in the macula densa cells, distal tubules, and collecting ducts of the kidney . This distinct localisation pattern of DDAH and NOS isoforms may contribute to the site-specific regulation of NO generation within the nephron in response to contrast medium . In light of the abundance DDAH-2 in the vascular tissue and the considerable amount of vascular tissue in the kidneys, we investigated the axis of DDAH-2/ADMA/NOS in the development of CI-AKI in vitro and in vivo.
In the present study, we only focus on iopromide-induced acute kidney injury. Although newer iso-osmolar contrast medium is thought to be associated with a reduced incidence of CI-AKI, several meta-analysis including randomised controlled trials of iodixanol comparing with low-osmolar contrast media (including iopromide) have shown that iodixanol was associated with only a nonsignificant reduction in the incidence [18,19]. Moreover, iopromide is still one of the most commonly used contrast media and explore the strategy to alleviate iopromide-induced acute kidney injury is of great worth.
Human aortic endothelial cells (HAECs) were purchased from PromoCell (Heidelberg, Germany) and rat renal proximal tubular cell lines (NRK-52E) were from Bioresource Collection Research Center (BCRC, Hsinchu, Taiwan). HAECs were cultured in endothelial cell (EC) growth medium MV and NRK-52E cultured in Dulbecco’s modified Eagle’s medium with 5% bovine calf serum. The medium was also supplemented with 100 units/ml penicillin and 100 μg/ml streptomycin (HyClone, Logan, UT) in a humidified 95% air/5% CO2 incubator at 37°C.
Iopromide (Ultravist 370, Bayer Healthcare, Berlin, Germany), ADMA (D4268), N-acetylcysteine (NAC, A9165), and l-NAME (N5751) were purchased from Sigma–Aldrich. Antibodies against DDAH-1 (ab127579, Abcam), DDAH-2 (GTX54028, GeneTex), cleaved caspase-3 (CC3; 9664, Cell Signaling), endothelial NOS (eNOS) (32027, Cell Signaling), phospho-eNOS (9571, Cell Signaling), F4/80 (sc-377009), myeloperoxidase (MPO) (sc-390109), COL1A1 (sc-25974), CD3 (sc-20047), GAPDH (G8795, Sigma–Aldrich), β-actin (A5441, Sigma–Aldrich) for immunoblotting, and DDAH-2 (14966-1-AP, proteintech), MPO (ab208670, abcam) for immunohistochemistry were purchased. Two set of three siRNA oligonucleotides for human DDAH-2 (sc-40474) and control siRNA (sc-37007) were purchased from Santa Cruz Biotechnology. It is a mixture of below sequences, Sense: GCUGACAGAUCACCCAUAUtt, Antisense: AUAUGGGUGAUCUGUCAGCtt, Sense: GUAGGAUAGUAUAGGAAGUtt, Antisense: ACUUCCUAUACUAUCCUACtt, Sense: GGAAGGAGGGUUAGAUAGAtt, Antisense: UCUAUCUAACCCUCCUUCCtt. Transfection of siRNAs into HAECs was performed using GenMute siRNA transfection reagent (SignaGen).
Generation of DDAH-2 adenovirus
Recombinant adenovirus vectors were constructed, propagated, and titered as previously described . Briefly, pBHGloxΔE1,3Cre (Microbix), including the ΔE1 adenoviral genome, was cotransfected with the pDC shuttle vector containing the gene of interest into HEK293 cells using Lipofectamine 2000 (Invitrogen). Through homologous recombination, the test genes were integrated into the E1-deleted adenoviral genome. The viruses were propagated in HEK293 cells. We produced replication-defective human adenovirus type 5 (devoid of E1)-harbouring DDAH-2. Adenovirus-harbouring β-galactosidase (LacZ) was used as a control.
C57BL/6 mice were obtained from the National Laboratory Animal Center of Taiwan, and heterozygous DDAH-2 knockout mice (Ddah2tm1.1(KOMP)Vlcg) were purchased from the Knockout Mouse Project repository. DDAH-2 knockout mice are C57BL/6 background and carried the tm1.1 allele. The tm1.1 allele is a LacZ reporter and a nonconditional knockout of exon 2 of DDAH-2 (https://www.komp.org/geneinfo.php?geneid=53703). When the male mice were crossed with wild-type female breeders, both wild-type and heterozygous DDAH-2 mice were generated, and these mice were used for the experiment. Genomic DNA was isolated from tail biopsies obtained using a DNeasy kit when the mice reached 3 weeks of age. A polymerase chain reaction was performed for genotyping the offspring using primer pairs 5′-TTGACTGTAGCGGCTGATGTTG-3′ and 5′-GGTAAACTGGCTCGGATTAGGG-3′. All animal experiments were performed in animal center of Taipei Veterans General Hospital in accordance with the approved guidelines by the Institutional Animal Care and Use Committee (IACUC) of Taipei Veterans General Hospital (#2016-124).
Animal model of CI-AKI
The mice model of contrast-medium-induced acute kidney injury was a modified version of a previously described model . In summary, C57BL/6 mice were anaesthetised with intraperitoneal Zoletil/Xylazine (10 mg/kg + 5 mg/kg) and placed on a heating pad to maintain the core body temperature between 36.5 and 37.5°C. They were then subjected to ischaemia–reperfusion injury (IRI) by left renal pedicle clipping for 45 min using a microclamp, and Ad-DDAH-2 or Ad-LacZ (1 × 109 pfu/40 ul) was injected into the left renal artery using a 30G needle during the ischaemic phase. After removal of the clamp, reperfusion of the kidneys was confirmed visually. The incision was closed, and the mice were returned to the animal room. To maintain fluid balance, all mice were supplemented with 1 ml of prewarmed phosphate-buffered saline (PBS) administered intraperitoneally directly after surgery. Five days after transduction, the right kidney was removed through another incision, and 300 µl of contrast medium iopromide (370 mg iodine/ml and 770 mOsmol/kg H2O, approximately 600 ml of contrast for a 60-kg human adult) was injected into the tail vein using a 29G needle. Subsequently, 100 µl of blood was collected from the submandibular area using a 5-mm lancet, and serum was separated for use in blood urea nitrogen (BUN) and creatinine assessments everyday for 5 days. The mice were then killed by carbon dioxide inhalation, and their kidneys were perfused with PBS for further management.
Half of each kidney was homogenised in a RIPA buffer and centrifuged at 15800×g for 30 min at 4°C. The supernatant was collected for an immunoblot and measurement of ADMA. The other half of the kidney was fixed with formalin, embedded in paraffin, and sectioned at 6-μm thickness for staining. For the knockout experiment, mice were subjected to IRI of the left kidney for 45 min and followed by right nephrectomy and contrast-medium injection 5 days later, performed as described.
X-gal staining of frozen tissue sections was performed as previously described . Briefly, mice were perfused with PBS, and the harvested tissues were fixed in 4% paraformaldehyde for 4 h at 4°C. All samples were transferred to a 15% sucrose PBS solution overnight and subsequently to a 30% sucrose PBS solution overnight at 4°C. The samples were then mounted in an optimal cutting temperature mounting medium (Fisher Scientific) and stored at −80°C before sectioning. Subsequently, 6 μm of cyro-section was washed with the staining buffer for 10 min at room temperature, incubated with X-gal solution for 4 h at room temperature, and counterstained using Fast Red (Sigma, St. Louis, MO).
Serum BUN and creatinine were measured using the Spotchem EZ (SP-4430, ARKRAY Inc., Japan) dry-chemistry analyser. ADMA levels were measured using the previously described competitive enzyme-linked immunosorbent assay (DLD Diagnostika GmbH, Hamburg, Germany) .
Commercial ELISA assays kit for TGF-β1 (MB100B), TNF-α (MTA00B), IL-1β (MLB00C), IL-6 (M6000B), and MCP-1 (MJE00) were from R&D. All homogenate samples for each mouse were stored at −80°C and were measured simultaneously using the same ELISA to avoid variation. The concentration was expressed as pg or nmol/mg total protein of homogenate.
Cell or tissue proteins were extracted using a RIPA lysis buffer (50 mM Tris/HCl pH 7.4, 0.1% sodium dodecyl sulphate [SDS], 1% NP-40, 150 mM NaCl, 0.5% sodium deoxycholate, and 1 mM ethylenediaminetetraacetic acid supplemented with protease inhibitors). After complete homogenisation on an ice rotator, samples were centrifuged at 15800×g for 30 min at 4°C to precipitate cell debris. The supernatants were transferred into fresh tubes, and protein concentrations were determined using a bicinchoninic acid assay (BCA). Proteins were fractionated using 10% SDS/polyacrylamide gel electrophoresis and were electro-transferred on to a polyvinylidene fluoride membrane. After blocking with Tris-buffered saline containing 5% nonfat milk, the membranes were probed with a primary antibody at 4°C overnight. A horseradish peroxidase–conjugated secondary antibody was used to detect luminochemicals. Band intensities were quantified using densitometry. The results were normalised to β-actin (Sigma). Densitometric analysis was performed using ImageJ software (National Institutes of Health).
Histological analysis and immunohistochemistry
Haematoxylin and Eosin (H&E) and Masson’s trichrome stainings were performed following standard procedures. Collagen was evaluated using Masson’s trichrome staining. The collagen fraction was calculated as the ratio of the sum of the total area of interstitial fibrosis to that of the total connective tissue area plus the renal glomerular and tubular cell in the entire visual field of the section. Approximately five cross‐sections were examined in each kidney. The formalin-fixed paraffin-embedded renal sections were deparaffinised and rehydrated in PBS. Pretreatments included microwave antigen retrieval in a 10 mmol/l citrate buffer for 20 min. Immunohistochemical staining was performed using the HRP-labelled polymer system (DAKO EnVision™ + System-HRP), and sections were counterstained with Haematoxylin.
Evaluation of apoptosis in tissue sections
DNA fragmentation was detected in situ using TUNEL. Briefly, deparaffinised sections were incubated with proteinase K, and DNA fragments were labelled with fluorescein–conjugated dUTP using TdT (Roche Molecular Biochemicals). Nuclear density was determined by manual counting of DAPI-stained nuclei in six fields for each animal using the 40× objective, and the number of TUNEL-positive nuclei was counted by examining the entire section using the same power objective.
Cell proliferation reagent WST-1
Cell viability was quantified using the cell proliferation reagent WST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5 -tetrazolio]-1,3-benzene disulfonate) (Dojindo, Kumamoto, Japan). Cells were plated in 96-well microplates at 5 × 104 cells/well, and then incubated for 24 h. Contrast medium of the indicated concentrations was added to the wells and incubated for 18 h. Thereafter, 10 μl of WST-1 solution was added for 3 h, and absorbance was measured in a microplate at 450 nm.
Lipid peroxidation assay
Lipid peroxidation in the serum or renal homogenates was determined by measuring the production of malondialdehyde (MDA) using a commercial kit, (Thiobarbituric Acid Reactive Substances (TBARS) Assay Kit from Cayman Chemical Item Number 10009055). The TBARS assay was conducted based on the manufacturer’s instructions, and in brief MDA reacts with thiobarbituric acid (TBA) forming a coloured product measured at absorbance 530 nm. The level of lipid peroxides was expressed as µM or nmol/mg protein of MDA.
Renal tubular injury analysis
The presence and extent of tubular injury was evaluated semiquantitatively as described by Zhang et al. . H&E stain was examined under light microscopy (at a final magnification of 200×) and each mouse, at least five fields were checked. The tubular injury score was calculated according to the following grades: grade 0, normal; grade 1, <25%; grade 2, 25–49%; grade 3, 50–74%; grade 4, ≥75%.
Reactive oxygen species
Intracellular reactive oxygen species (ROS) production was measured as previously described . In summary, cells were loaded with 10-μM 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA) (Molecular Probes) in the dark for 1 h at 37°C. Cells were thrice washed with Hank’s balanced salt solution and were subsequently incubated in prewarmed growth medium for an additional 30 min at 37°C. ROS generation was detected using a fluorescence microscope following the oxidation of H2DCFDA (excitation, 504 nm; emission, 529 nm). To avoid photo-oxidation of H2DCFDA, the fluorescence images were collected using a single rapid scan, and identical settings were used for all samples.
Determination of NO production and intracellular cGMP
Accumulated nitrite (NO2−) is the stable breakdown product of NO. In the present study, NO2− in culture media was measured by mixing an equal volume of Griess reagent and then incubating the mixture at room temperature for 15 min. Azo dye production was analysed using an SP‐8001 UV/VIS spectrophotometer (Metertech, Taipei, Taiwan) with absorbance at 540 nm. Sodium nitrite was used as a standard. Intracellular levels of cGMP in ECs were assessed using an enzyme immunoassay kit, and amount normalised to protein content as determined by the Bradford assay.
Data were expressed as mean ± SEM. The differences in the data between the two groups were determined by a Student’s t test. Comparison among all groups was assessed by one-way ANOVA followed by Bonferroni’s post hoc test. BUN and creatinine between two groups at different times after the contrast medium injected were compared using a compound symmetry covariance linear mixed models. All statistical analyses were performed using SPSS 18.0 for Windows (SPSS Inc., Chicago, IL). A P-value of <0.05 was considered statistically significant.
Creation of adeno-DDAH-2 virus and its overexpression
in vitro and in vivo
To elucidate the role of DDAH-2 in CI-AKI, Ad-DDAH-2 was created, and it could increase DDAH-2 expression in HAECs (Figure 1A). To test its function in vivo, left kidney of C57BL/6 mice was subjected to IRI, during ischaemic phase Ad-DDAH-2 or Ad-LacZ was injected into left renal artery. Five days later, left kidney was taken to examine the expression of DDAH-2 by means of immunoblot and immunohistochemistry (Figure 1B,C). DDAH-2 is expressed stronger in vascular tissue. The expression of DDAH-1 was not influenced by DDAH-2 in vivo (Figure 1B). Renal ADMA (0.10 ± 0.01 nmol/mg protein) and MDA (0.13 ± 0.01 nmol/mg protein) level were also measured in Ad-DDAH-2 group and did not show any difference when comparing with group of Ad-LacZ, even the control group without IRI (Figure 1D). Above results also revealed that overexpression of DDAH-2 by Ad-DDAH-2 in kidney did not change renal ADMA and MDA levels 5 days after IRI without contrast medium.
Creation of Adeno-DDAH-2 virus and its overexpression
in vitro and in vivo
Overexpression of DDAH-2 alleviates contrast medium iopromide-induced acute kidney injury
In the mice model of CI-AKI, serum BUN and creatinine were measured consecutively for 5 days. After exposure to iopromide, serum creatinine and BUN levels peaked in the first day and then gradually declined. In the Ad-DDAH-2 group, the rise of serum creatinine levels was modest but significantly lower compared with that in the Ad-LacZ group (P=0.02). Lesser increases in BUN levels were observed in the Ad-DDAH-2 group, although this difference did not reach statistical significance (P = 0.28) (Figure 2A). In addition, after exposure to iopromide, renal ADMA level was significantly increased in the Ad-LacZ group (0.45 ± 0.01 nmol/mg protein), but the condition could be improved in the Ad-DDAH-2 group (0.21 ± 0.02 nmol/mg protein) (Figures 1D and 2B). This finding suggested that iopromide might increase ADMA level in the kidney and could be degraded more through overexpression of DDAH-2. Masson’s trichrome staining showed significantly less accumulation of collagen in the renal tissue in the Ad-DDAH-2 group (3.41 ± 0.27 vs. 5.14 ± 0.32% in the Ad-LacZ group, Figure 2C,D). Moreover, the expression of α-1 type I collagen (CoL1A1) and CC3 in immunoblot analysis and the number of TUNEL-positive nucleus (12.38 ± 1.26 vs. 21.55 ± 1.53% in the Ad-LacZ group) in the renal cortex area also significantly decreased in the Ad-DDAH-2 group (Figure 2E,F). Because of differential CoL1A1 expression, TGF-β1 of renal homogenate was also examined and significantly decreased from 51.37 ± 7.49 pg/mg protein in the Ad-LacZ group to 6.86 ± 1.63 pg/mg protein in the Ad-DDAH-2 group (Figure 2G). All of these suggested that overexpression of DDAH-2 may have alleviated CI-AKI and resulted in less fibrosis and apoptosis.
Overexpression DDAH-2 alleviates iopromide-induced renal fibrosis and apoptosis
We further examined the expression of inflammatory cells and cytokines in kidneys. After exposure to iopromide, the levels of inflammatory cells and cytokines increased significantly; however, in the Ad-DDAH-2 group, only modest increases in levels of inflammatory cells and cytokines were observed. Immunoblot analyses revealed that the levels of F4/80, MPO, and CD3 rose significantly less in the Ad-DDAH-2 group; the levels decreased to 35.12 ± 4.16, 44.86 ± 5.84, and 36.96 ± 11.97%, respectively, compared with the group of Ad-LacZ with iopromide (Figure 3A,B). Likewise, after iopromide exposure, the overexpression of DDAH-2 significantly alleviated the increases in TNF-α, IL-1β, IL-6, and MCP-1. The levels of TNF-α, IL-1β, IL-6, and MCP-1 significantly decreased from 39.50 ± 1.99, 156.52 ± 8.46, 119.84 ± 7.28, and 182.47 ± 9.86 pg/mg protein in the group of Ad-LacZ with iopromide to 31.36 ± 2.12, 118.48 ± 8.09, 92.46 ± 6.27, and 144.57 ± 7.45 pg/mg protein in the group of Ad-DDAH-2 with iopromide, respectively (Figure 3C). Moreover, to examine the ROS after exposure to iopromide, renal MDA was checked and showed significant increase in the Ad-LacZ group (0.23 ± 0.01 nmol/mg protein), and it could be improved in the Ad-DDAH-2 group (0.17 ± 0.01 nmol/mg protein) (Figures 1D and 3D). It pointed that iopromide could increase ADMA and ROS levels in the kidney and could be alleviated through overexpression of DDAH-2. Due to involvement of inflammatory response and ROS, renal tubular injury was also evaluated and showed lower tubular injury in Ad-DDAH-2 group (Figure 3E). These findings suggested that the overexpression of DDAH-2 may have attenuated CI-AKI through down-regulation or inhibition of the inflammatory response, ROS production, and cell injury.
Overexpression DDAH-2 attenuates iopromide-induced inflammation responses and ROS production
Knockout of DDAH-2 aggravates contrast medium iopromide-induced acute kidney injury
Characterisation of DDAH-2 heterozygous knockout (DDAH-2 KO) mice was done before in vivo experiments. It was found that DDAH-2 KO mice are subviable. Approximately 1/8 heterozygous knockout mice could be got after cross heterozygous DDAH-2 knockout with wild-type mice. Cardiac, renal, and aortic DDAH-2 expression were significantly decreased in DDAH-2 KO mice (Figure 4A,B). Furthermore, serum ADMA, BUN, creatinine, and MDA were also increased from 0.60 ± 0.07 µM, 28.75 ± 1.25 mg/dl, 0.33 ± 0.03 mg/dl, and 2.30 ± 0.25 µM in the wild-type to 0.81 ± 0.09 µM, 49.00 ± 0.91 mg/dl, 0.55 ± 0.03 mg/dl, and 4.19 ± 0.36 µM in the DDAH-2 KO mice, respectively, although serum ADMA did not reach statistical significance (Figure 4C).
Characteristics of DDAH-2 KO mice
After exposure to the contrast medium, serum creatinine and BUN levels were still significantly higher in the DDAH-2 KO group than in the wild-type group, although baseline differences existed (Figure 5A). IRI also caused elevation of serum creatinine and BUN, but more modest. ADMA levels in renal homogenates significantly increased from 0.20 ± 0.02 to 0.29 ± 0.03 nmol/mg protein in wild-type mice after exposure to iopromide. Furthermore, ADMA levels increased to 0.48 ± 0.01 nmol/mg protein in DDAH-2 KO mice exposure to iopromide (Figure 5B). A greater amount of collagen (14.57 ± 1.15 vs. 5.03 ± 0.57% in the wild-type group) accumulated in the kidney of the DDAH-2 KO group after exposure to the contrast medium (Figure 5C,D). Moreover, the expression of CoL1A1 and CC3 in immunoblot analysis and the number of TUNEL-positive nucleus (29.64 ± 1.89 vs. 20.62 ± 1.98% in the wild-type group) in the renal cortex area also significantly increased in DDAH-2 KO mice after exposure to the contrast medium (Figure 5E,F). Renal TGF-β1 increased from 16.29 ± 7.77 pg/mg protein in the group of wild-type to 59.09 ± 7.00 pg/mg protein in the group of DDAH-2 KO as well (Figure 5G). All these suggested that knockout of DDAH-2 may have aggravated CI-AKI and resulted in more fibrosis and apoptosis.
Knockout DDAH-2 aggravates iopromide-induced renal fibrosis and apoptosis
Inflammatory response, ROS production, and cell injury after contrast-medium exposure was also examined in DDAH-2 KO mice. The immunoblot analyses revealed that F4/80, MPO, and CD3 significantly increased by 3.07 ± 0.54, 2.72 ± 0.39, and 1.76 ± 0.26 times, respectively, in the DDAH-2 KO compared with wild-type with iopromide (Figure 6A,B). Likewise, the levels of TNF-α, IL-1β, IL-6, and MCP-1 significantly increased from 40.44 ± 7.05, 93.10 ± 8.18, 110.99 ± 4.98, and 165.62 ± 9.49 pg/mg protein in the wild-type with iopromide to 63.97 ± 5.45, 128.31 ± 7.74, 145.19 ± 5.35, and 232.84 ± 16.93 pg/mg protein in the DDAH-2 KO with iopromide, respectively, (Figure 6C). Renal MDA was also significantly increased in the DDAH-2 KO with iopromide (0.28 ± 0.02 vs. 0.21 ± 0.01 nmol/mg protein in the wild-type with iopromide), and tubular injury was higher in the group of DDAH-2 KO with iopromide (Figure 6D,E). All of these findings supported the possibility that DDAH-2 knockout aggravated CI-AKI through up-regulation of the inflammation response, ROS production, and cell injury.
Knockout DDAH-2 aggravates iopromide-induced inflammation responses and ROS production
DDAH-2 alleviates iopromide-induced apoptosis through NO-dependent ROS modulation
Exposure to the contrast medium iopromide led to HAECs death in a concentration (dose)-dependent manner (Figure 7A). The iopromide concentration of 60 mg Iodine/ml was used in in vitro experiments because of no obvious decrease in cell number. Firstly, it was noted that cellular ADMA level significantly increases from 0.33 ± 0.04 to 0.49 ± 0.05 nmole/mg protein after treating with iopromide (Figure 7B). As the in vivo results, DDAH-2 overexpression attenuated the CC3 expression of HAECs subjected to iopromide, and DDAH-2 knockdown exhibited the opposite effect (Figure 7C). DDAH-1 expression was not influenced by DDAH-2 overexpression or knockdown in vitro. Above data suggested that cellular DDAH-2/ADMA is associated with HAECs apoptosis when exposed to iopromide. The protective effect of DDAH-2 was also certified in the NRK-52E cells (a rat renal proximal tubular cell line) (Supplementary Figure S1). To explore the protective mechanism of DDAH-2, HAECs were transduced with either Ad-LacZ or Ad-DDAH-2 and treated with or without NAC, l-NAME (an inhibitor of NOS), or ADMA before iopromide administeration. Exposure to iopromide increased the expression of CC3 in the Ad-LacZ group, and its expression decreased when NAC was added. However, no additional protective effect was observed when NAC was added to the Ad-DDAH-2 group (Figure 7D). Intracellular ROS markedly increased after iopromide exposure and was effectively decreased by both DDAH-2 overexpression and ROS scavenger NAC. Furthermore, the effect of the ROS decrease was blocked by the addition of l-NAME (Figure 7E). These results suggested that the antioxidative effect of DDAH-2 was regulated by NOS activity in the HAECs. Both nitrite and cGMP levels in the HAECs significantly increased in the Ad-DDAH-2 group, and pretreatment of l-NAME abrogated these effects. In contrast, the addition of ADMA did not significantly reduce the nitrite or cGMP levels of HAECs in the presence of Ad-DDAH-2 (Figure 7F). In vivo, the renal nitrite level of control mice (Ad-LacZ) decreased after exposure to iopromide, while overexpression of DDAH-2 might partially reverse the effect. Contrary, renal nitrite level of DDAH-2 KO mice were more reduced after exposure of iopromide (Figure 7G). Finally, both eNOS and phosopho-eNOS (Ser1177) were examined and phosopho-eNOS showed increase after DDAH-2 overexpression, and decrease after DDAH-2 knockdown (Figure 7H). In summary, these findings suggested that the protective effect of DDAH-2 in CI-AKI might resulted from the increase in NO production and modulation of ROS generation through NOS.
DDAH-2 alleviates iopromide-induced ROS and cellular apoptosis through NO-dependent ROS modulation
In the present study, we found that exposure to the contrast medium, iopromide, may have led to ADMA accumulation, increased ROS generation, and an inflammatory response in mice kidney tissue. Overexpression of DDAH-2 may have lowered renal ADMA levels and exhibited a reno-protective effect against contrast-medium injury in the mice model, wherein cell apoptosis and fibrosis decreased. The knockout of DDAH-2 exhibited the opposite effect. The protective effect of DDAH-2 may have resulted from the increase in NO bioavailability in modulations of ROS generation and the inflammatory process. Our findings provide the first evidence that the DDAH-2/ADMA/NOS axis may be involved in the pathogenesis of CI-AKI.
Several studies have demonstrated that the administration of a contrast medium is associated with a significant reduction in NO synthesis, which may be related to CI-AKI development [26–28]. ADMA is an endogenous inhibitor for all three isoforms of NO synthase . In addition to decreasing the bioavailability of NO and resulting in vasoconstriction and endothelial dysfunction, ADMA increases oxidative stress by uncoupling electron transport between NO synthase and l-arginine [6,29], and it may inhibit the mobilisation, differentiation, and function of endothelial progenitor cells . In this study, exposure to contrast medium significantly increased renal ADMA levels in mice. A previous study demonstrated that microperfusion of individual loops of Henle in a rat with ADMA inhibited NO synthase in the macula densa and enhanced the maximal tubuloglomerular feedback response, leading to vasoconstriction of the afferent arterioles and thereby causing a decrease in the glomerular filtration rate and flow . Moreover, accumulation of ADMA in elderly human participants/patients is associated with increased renovascular resistance and decreased renal perfusion . Therefore, it is reasonable to hypothesise that ADMA is involved in the pathogenesis of CI-AKI and that thus targetting DDAH can lead to clarification of the underlying mechanism. Our results revealed protective effects of DDAH-2 on the kidney against contrast-medium injury; however, a recent study reported that reduction in renal tubular ADMA metabolism by proximal tubule-specific DDAH-1 knockout exhibited a protective effect against progressive kidney function decline , which appears to contradict our results. The distinct distributions and functions of DDAH-1 and DDAH-2 are well known . Recent studies have shown that serum ADMA levels are regulated by DDAH-1, which is expressed at sites of ADMA metabolism in the kidney cortex and liver. By contrast, NO activity/endothelium function is regulated primarily by DDAH-2, which is expressed strongly in blood vessels [14,33]. Moreover, contrast medium causes direct thick ascending limb cell damage with accompanying increased oxidative stress and decreased NO bioavailability, and it impairs the tubulovascular feedback of the macula densa . DDAH-2, rather than DDAH-1, is the predominant DDAH isoform in these tissues, which suggests that modulating DDAH-2 might enable the management of CI-AKI. Nevertheless, further studies may be necessary to elucidate the cell-specific function of DDAH isoforms.
ROS are generated during renal parenchymal hypoxia induced by contrast medium, which causes direct tubular and vascular endothelial injury followed by decreased NO bioavailability [35,36]. This phenomenon can aggravate renal parenchymal hypoxia and augment endothelial dysfunction and dysregulation of tubular transport [37,38]. Elevated ADMA levels have been associated with increased vascular superoxide production and endothelial NOS uncoupling [15,39]. This mechanism may impair DDAH activity and increase protein arginine methyltransferase (PRMT) activity through ROS in ECs . Both of these effects lead to a further increase in the production of ADMA and a decrease in NO bioavailability. Here, we found that contrast medium (iopromide) increased the ROS amount and ADMA levels in the kidney and that these increases may have further aggravated endothelial dysfunction and parenchymal hypoxia. In turn, these effects may have led to cellular apoptosis, including in ECs, and induction of an inflammatory reaction.
Research has suggested that inflammation plays a critical role in the pathogenesis of CI-AKI. In the mouse model of nephrotoxicity, TNF-α might be central to activation of the inflammatory cytokine response [41,42]. Other inflammatory mediators, including IL-6 [43,44], MPO , and IL-1β [45,46] are also involved in contrast-medium-induced nephrotoxicity. Elevated levels of MCP-1 (monocyte) and CD3 (lymphocyte) were observed in our CI-AKI model. Previous studies have demonstrated that ADMA may profoundly impair NO synthesis of neutrophils, resulting in increased superoxide generation and release of MPO . Overexpression of DDAH-2 attenuated the LPS-induced increase in ADMA levels and oxidative stress in both ECs and the mouse lung . These findings suggest a close association between DDAH-2/ADMA/NOS axis and inflammation. Similarly, DDAH-2 may protect the kidney from CI-AKI by attenuating the inflammatory process, as revealed in our study. However, further investigation delineating the molecular mechanisms of how DDAH-2/ADMA/NOS regulating the inflammatory response is warranted.
The CI-AKI mouse model in our study was modified from the model used by Linkermann et al.’s  research group. In the CI-AKI mouse model used by Linkermann et al. , a 30-min IRI was induced using bilateral renal pedicle clamping, and exposure to contrast medium for 24 h was applied after reperfusion. A previous study showed that IRI-induced oxidative stress may reduce DDAH-1 expression and cause ADMA accumulation . Thus, our study results regarding the mechanism of CI-AKI may have been confounded by the use of an animal model that combined the IRI and exposure to the contrast medium. However, single injection of contrast medium causes overt renal damage only in rabbits. Rats and mice require an additional insult to the kidney to manifest clinical CI-AKI . The most common type of additional insult in rats and mice is ischemia. We combined a left-side 45-min IRI and right nephrectomy, followed by contrast medium injection, and found that the animal model was more reliable with the rise and fall of renal function biochemistry (BUN, creatinine), similar to the change in human CI-AKI. All control animals received IRI without contrast-medium injection. Nevertheless, exploring the mechanism of CI-AKI by using a pure animal model that develops CI-AKI after contrast-medium exposure may be more informative.
It should be mentioned that 5 days after IRI, ADMA level was not different in groups of sham, Ad-LacZ, and Ad-DDAH-2 (Figure 1D), but significantly different between wild-type and knockout mice (Figure 5B). It may result from different method of gain- or loss-of-function. The gain-of-function is through intra-arterial injection of adenovirus, and overexpression only in the kidney, but loss-of-function is with the heterozygous knockout mice whose lots of tissue show lower expression of DDAH-2. Hence, the deleterious effects of DDAH-2 knockout mice may be owing to endothelial dysfunction of other tissue besides kidney, however, it is limited to differentiate it in our study. Organ-specific knockout mice may answer the question in detail.
Our results suggested that therapies targetting DDAH-2/ADMA/NOS might be beneficial in ameliorating CI-AKI. Although specific ADMA-lowering or DDAH-activating therapies are not yet available, the cAMP phosphodiesterase 3/4 inhibitor, tolafentrine, which might modulate the promoter region of the DDAH-2 gene, has been shown to promote endothelial regeneration by increasing DDAH-2 expression/activity and reducing ADMA levels . Additional studies of the potential beneficial effects of phosphodiesterase inhibitor on CI-AKI are warranted. Antioxidants for decreasing the levels of ROS and increasing the DDAH-2 expression/activity might be promising, but contradictory results have been reported regarding use of intravenous-acetylcysteine to prevent CI-AKI . By contrast, accumulating evidence suggests that statins may have distinctive anti-inflammatory properties, increase NOS bioavailability, improve oxidative injury, and reduce ROS generation [53,54]. Previous studies have also demonstrated that statins might significantly reduce CI-AKI risk in patients with diabetes and chronic kidney disease . Our earlier study showed that rosuvastatin might improve endothelial function and that appeared to be associated with its effect of reduction plasma ADMA levels . However, ADMA may trigger a nicotinamide adenine dinucleotide phosphate (NADPH) oxidase pathway and increase ROS production, thereby restricting the statin‐conferred protection of eNOS activation, NO production, and angiogenesis . This evidence, along with our findings, suggests that DDAH-2/ADMA/NOS might be involved in the mechanism underlying the protective effect of statins against CI-AKI. Further studies of these phenomena are warranted.
Strength and limitation of the study
Our findings provide the first evidence that the DDAH-2/ADMA/NOS axis may be involved in the pathogenesis of CI-AKI. However, only iopromide was studied. Furthermore, it was not explored whether the beneficial function of DDAH-2 could be independent of ADMA in CI-AKI or not. Further investigation is warranted to delineate the signalling pathways.
Our findings demonstrated that the DDAH-2/ADMA/NOS pathway is involved in the pathogenesis of CI-AKI in vitro and in vivo, and the protective effect of DDAH-2 probably results from the modulation of NOS activity, oxidative stress, and the inflammatory process.
ADMA, an endogenous NOS inhibitor, is largely metabolised by DDAH in humans, and can predict the risk of CI-AKI in patients undergoing cardiac catheterisation in our earlier clinical study. However, no cellular or animal experiment was designed to explore the mechanism.
In the present study, it showed that DDAH-2 may modulate NOS activity, oxidative stress, and the inflammatory process. To our knowledge, it is the first evidence to address that DDAH-2/ADMA/NOS axis may be involved in the pathogenesis of CI-AKI.
With our significant findings, specific ADMA-lowering or DDAH-activating therapies may be promising in the future research, even not available now.
This work was supported by the Ministry of Science and Technology of Taiwan [grant numbers MOST-103-2314-B-075-046, 106-2314-B-075-044, 107-2314-B-075-060]; and the Veterans General Hospitals and University System of Taiwan Joint Research Program [grant numbers VGHUST102-G7-5, 104-G7-5; 105-G7-3].
The authors declare that there are no competing interests associated with the manuscript.
H.-H.L.: Experimental studies and data acquisition. T.-S.L.: Statistical analysis. S.-J.L.: Manuscript review and study concepts. Y.-C.Y.: Data acquisition. T.-M.L.: Study design and manuscript editing. C.-P.H.: Study design and manuscript preparation.
blood urea nitrogen
contrast medium-induced acute kidney injury
α-1 type I collagen
- DDAH-2 KO
DDAH-2 heterozygous knockout
endothelial nitric oxide synthase
human aortic endothelial cell
Haematoxylin and Eosin
N omega-Nitro-L-arginine methyl ester
nitric oxide synthase
reactive oxygen species
sodium dodecyl sulphate
thiobarbituric acid reactive substances
Terminal deoxynucleotidyl transferase dUTP nick end labeling