Role of Cystathionine Gamma-Lyase in Immediate Renal Impairment and Inflammatory Response in Acute Ischemic Kidney Injury

Hydrogen sulfide (H2S) is known to act protectively during renal ischemia/reperfusion injury (IRI). However, the role of the endogenous H2S in acute kidney injury (AKI) is largely unclear. Here, we analyzed the role of cystathionine gamma-lyase (CTH) in acute renal IRI using CTH-deficient (Cth−/−) mice whose renal H2S levels were approximately 50% of control (wild-type) mice. Although levels of serum creatinine and renal expression of AKI marker proteins were equivalent between Cth−/− and control mice, histological analysis revealed that IRI caused less renal tubular damage in Cth−/− mice. Flow cytometric analysis revealed that renal population of infiltrated granulocytes/macrophages was equivalent in these mice. However, renal expression levels of certain inflammatory cytokines/adhesion molecules believed to play a role in IRI were found to be lower after IRI only in Cth−/− mice. Our results indicate that the systemic CTH loss does not deteriorate but rather ameliorates the immediate AKI outcome probably due to reduced inflammatory responses in the kidney. The renal expression of CTH and other H2S-producing enzymes was markedly suppressed after IRI, which could be an integrated adaptive response for renal cell protection.


Protein Expression of CTH and Endogenous H 2 S Levels in the Kidney. Levels of renal CTH protein
in heterozygous (Cth +/− ) mice were ~40% of those in Cth +/+ mice (Fig. 1D) as previously reported 11 . Similar to its mRNA level changes, renal CTH protein levels declined after IRI in both Cth +/+ and Cth +/− mice and no CTH protein was detectable in the kidneys of Cth −/− mice (Fig. 1D). Next, we measured endogenous H 2 S levels in the kidney to assess the impact of systemic CTH deletion on H 2 S production. Kidneys of Cth +/− and Cth −/− mice displayed approximately 30% and 50% reduced H 2 S levels, respectively, compared to those of Cth +/+ mice (Fig. 1E).
The impact of CTH loss in Cellular Infiltration to the Kidneys after IRI. Renal IRI is known to associate with infiltration of granulocytes, monocytes/macrophages and other immune cells immediately after reperfusion, which contributes to inflammation and subsequent repair in injured kidneys 21 . Therefore, we characterized granulocytes and macrophages in renal IRI by flow cytometry. Whole kidney cell suspensions were immunolabelled for Ly6G and F4/80 as markers for granulocytes and macrophages, respectively. Among pre-gated singlet live cells (Fig. 3A), Ly6G-positive & F4/80-negative granulocytes as well as Ly6G-negative & F4/80-positive macrophages were detected (Fig. 3B). There were no significant differences in both granulocyte and macrophage populations between Cth +/+ , Cth +/− , and Cth −/− kidneys at 24 h after IRI ( Fig. 3C and D) or after sham surgery (Supplementary Figure 2A-C). We next performed immunohistochemistry to detect IRI-induced granulocyte infiltration (Supplementary Figure 3A) 22 . In the outer medulla after IRI, average Ly6B-positive cell numbers per view field were 12 in Cth −/− mice while 13 and 19 in Cth +/− and Cth +/+ mice, respectively, although the differences were not statistically significant (P = 0.501, Supplementary Figure 3B). Furthermore, renal levels of S100a8/a9 mRNAs for calprotectin, a heterodimeric protein that was recently found to co-localize with Ly6G in granulocytes after AKI and playing a crucial part in controlling M2 macrophage-mediated renal repair following IRI 23 , were also not significantly different (Supplementary Figure 3C,D).

The impact of CTH loss in Expression of Cytokines, Chemokines, and Adhesion Molecules.
Production of inflammatory molecules is maintained low in the normal kidney but is markedly increased under pathophysiological conditions such as ischemia 21 . We measured mRNA levels of several molecules involved in long-term outcome/repair after renal IRI. Renal expression of interleukin 1-beta (Il1b) and vascular cell adhesion molecule 1 (Vcam1) after IRI was significantly lower in Cth −/− mice compared to Cth +/− mice (Fig. 4A,B). Also, renal expression of tumor necrosis factor-alpha (Tnf) and vascular cell adhesion molecule 1 (Vcam1) was similarly lower in Cth −/− mice compared to Cth +/− mice (overall ANOVA P = 0.099 and P = 0.088, respectively) ( Fig. 4C,D). Renal expression of other important cytokines/chemokines such as interleukin 6 (Il6), chemokine (C-X-C motif) ligand 2 (Cxcl2), and chemokine (C-C motif) ligand 2 (Ccl2), were not altered among Cth genotypes (Supplementary Figure 4A-C).

The impact of CTH loss in In Vitro Macrophage Polarization.
Although the proportion of infiltrating macrophages after IRI was not significantly different ( Fig. 3D), renal mRNA expression of IL1-beta and TNF-alpha, the two major inflammatory cytokines of macrophage origin, was lower or in Cth −/− mice ( Fig. 4A,C). We hypothesized that macrophage polarization is disturbed by the lack of CTH, and thus investigated Tnf induction by the lipopolysaccharide (LPS)/interferon (IFN)-gamma in vitro treatment of bone marrow (BM)-derived macrophages from Cth +/+ and Cth −/− mice. Cth expression was induced while Mpst expression was not altered by LPS/IFN-gamma treatment in BM-derived macrophages from Cth +/+ mice (Fig. 5A,B). In contrast, Mpst expression was significantly induced by the same treatment in macrophages from Cth −/− mice (Fig. 5B), and Cbs expression was not detectable in macrophages from either mice (data not shown). Under such conditions, Tnf expression was markedly induced by LPS/IFN-gamma treatment of both Cth +/+ and Cth −/− macrophages, and the levels were significantly lower in Cth −/− macrophages (Fig. 5C), although the supernatant TNF-alpha concentrations of activated macrophages were comparable between Cth +/+ and Cth −/− mice (Fig. 5D).

Discussion
A number of studies have demonstrated the cytoprotective effects of H 2 S in myocardial, liver, brain, pulmonary, and renal IRI (reviewed by Nicholson and Calvert) 2 . Most of these studies utilized Na 2 S/NaHS as exogenous H 2 S donors and PAG as a non-specific CTH inhibitor. To overcome pharmacokinetic problems in H 2 S donor applications and specificity issues of PAG, two research groups have independently generated mice in which Cth genes have been differentially deleted 10,11 . In our study, we investigated the pathophysiological roles of CTH in renal IRI using one of those Cth −/− mice and their littermate Cth +/− and Cth +/+ mice as controls; all were the offspring from the mating between Cth +/− males and Cth +/− females that had been backcrossed over 10 generations onto a C57BL/6 background 11 . We found that the lack of CTH does not cause aggravated immediate renal functional impairments after IRI as assessed by serum creatinine levels ( Fig. 2A) and renal expression of sensitive AKI markers, Lcn2 and Havcr1 (Fig. 2B,C). Our histological examinations rather identified a moderate amelioration in renal tubular damage in Cth −/− mice (Fig. 2D,E).
While our study was underway, Bos et al. published findings with their Cth −/− mice (on a mixed strain background; the sex of mice used is not indicated) investigating the role of CTH-derived H 2 S in renal IRI 17 . They found that CTH deficiency aggravated kidney damage after IRI, which was associated with increased mortality 17 however, we did not observe such severe systemic damage after renal IRI. The reasons for this discrepancy are possibly multifaceted. First, their Cth −/− mice display age-dependent hypertension (15-20 mmHg higher systolic blood pressure vs Cth +/+ mice only after 7 weeks of age) and sex-related hyperhomocysteinemia in which females have six times the plasma homocysteine levels (120 vs 20 μ M) in males 10,17 , both of which are caused by unknown mechanisms. Hypertension per se has deleterious effects on renal IRI 24,25 nevertheless, hypertension was not properly treated in their studies 17 . This affair makes it difficult to assess the impact of reduced renal H 2 S production over elevated blood pressure on the outcome of IRI; fortunately, our Cth −/− mice display systolic normotension 11 . This fact may, at least in part, underlie differences between their and our findings. It should be noted that our Cth −/− males and females display similar serum levels of homocysteine (104-151 μ M) 11 the reasons for this difference are yet unknown but may depend on differences in genetic backgrounds and/or nutritional conditions. Second, Bos et al. performed renal ischemia by clamping both (right and left) renal arteries for 30 min, whereas we performed uninephrectomy by clamping the renal artery of the remnant left kidney for 20 min 17 . Despite the differences in surgical protocols, serum creatinine levels at 24 h after IRI were equivalent. But, importantly, all mice that underwent surgery survived in our study while Bos et al. observed 35% mortality only in Cth −/− mice 17 . Third, we used a temperature controller with heating pads to maintain a stable core temperature (which was measured continuously during surgery by a rectal probe) whereas Bos et al. used only heating pads and lamps 17 . It is well known that fluctuations in core body temperature contribute to variability in IRI and the way of maintaining body temperature during ischemia has a major impact on the outcome of IRI 26 . Fourth, their Cth −/− mice showed a massive (91%) reduction in renal H 2 S production compared to Cth +/+ mice 17 while our Cth −/− mice showed only 50% reduction (Fig. 1E). Although the methods used for H 2 S measurement substantially differ between the two studies and this precludes the direct comparison, > 90% reduction is surprising per se, considering the facts that (i) Cth −/− kidney still expresses CBS and MPST, (ii) (increased/activated) CBS could compensate for H 2 S production when CTH is inhibited or abrogated (though we did not observe compensatory Cbs mRNA induction; (Fig. 1B) and (iii) renal Cbs/Mpst expression was markedly down-regulated by ischemia/reperfusion irrespective of Cth genotypes (Fig. 1B,C) 11,12,17,27,28 . A previous study mentioned that the reduction in CBS (rather than CTH) activity may serve as the major contributor for endogenous H 2 S level reduction during renal IRI 29 .
Despite such differences, we also found some agreement with previous studies by Bos et al. 17 and others 27,30 . First, renal expression (either gene or protein) of both CTH and CBS were suppressed after renal IRI (Fig. 1A,B). It might be noteworthy that the partial or complete loss of CTH did not cause compensatory induction (or reduced repression of expression) of CBS (or MPST) during IRI (Fig. 1B,C) at least on mRNA level. Second, both Bos et al. and we did not find significant differences between Cth genotypes in the numbers of granulocyte infiltrated into injured kidneys of IRI mice (Fig. 3C and Supplementary Figure 3A-C) 17 . We also counted the numbers of F4/80-positive macrophages infiltrated into injured kidneys of IRI mice and found that macrophages behave similar to granulocytes (Fig. 3D). In contrast, renal expression of Tnf, Il1b, Icam1, and Vcam1 after IRI were lower (though overall ANOVA was just P = 0.099 and 0.088 for Tnf and Icam1, respectively) in Cth −/− mice compared to Cth +/− mice (Fig. 4A-D). TNF-alpha was initially discovered as a LPS-induced macrophage product 31 . It is also released during IRI and acts as a potent pro-inflammatory cytokine 32 , and in line, the blockade of TNF-alpha signaling is a novel promising therapeutic target in renal IRI 33 . Although intrinsic renal cells also secret TNF-alpha upon injury, monocytes/classically activated macrophages are considered as the main source of TNF-alpha in early renal IRI 34 . We found that CTH deficiency alters Tnf expression in LPS/IFN-gamma-stimulated BM-derived macrophages that intrinsically differ from LPS-stimulated peritoneal macrophages 10,35 . However, the supernatant TNF-alpha concentrations did not differ between both groups, which questions the physiological relevance of this finding. Meanwhile, renal expression of other cytokines that are known to play a role in renal IRI 36 (Il6, Cxcl2, and Ccl2) were not distinguishable between Cth genotypes (Supplementary Figure 4A-C).
Our findings are in contrast to previous results by others who use PAG for CTH inhibition. Tripatara et al. found that single intraperitoneal administration of PAG (50 mg/kg, 1 h before ischemia) prevented the renal recovery from IRI (45-min ischemia/72-h reperfusion) in a rat bilateral ischemia model 37 . More recently Han et al. showed similar deteriorative effects of PAG in renal IRI (50 mg/kg daily (i.p.), beginning 2 days after ischemia) in mice 12 . However, PAG (5 mg/kg (i.p.), twice a day for 4 successive days) exhibited nephroprotective effects in the cisplatin model of AKI in rats 38 . Similar protective effects of PAG (50 mg/kg (i.p.) at 2 h after adriamycin injection) have been observed in adriamycin-induced nephrotoxicity in rats 39 . Whereas these kidney injury models differ, they point out that PAG treatment can have multiple effects depending on the renal injury models. Moreover, the specificity of this widely used CTH inhibitor and relatively late time points after reperfusion are a matter of concern. Our model is of particular interest because we used a genetic approach to abrogate CTH specifically and investigated acute renal post-ischemic injury after 24 h, a time point where serum creatinine levels are the highest and renal Cth/Cbs expression levels are the lowest 17 .
Numerous studies have revealed cytoprotective/anti-oxidative/anti-inflammatory roles of H 2 S, but some studies also have identified pro-inflammatory roles of H 2 S that accelerate inflammatory responses; for example, Ang et al. previously reported that caerulein-induced acute pancreatic damage as well as its associated lung injury was ameliorated in Cth −/− mice compared to Cth +/+ 40 . It is possible that CTH-produced H 2 S may act as a pro-inflammatory factor in renal IRI. In addition, further studies should also clarify the impact of high levels of cystathionine and homocysteine and low levels of taurine that are common in Cth −/− mice 11 on the outcome of renal IRI [41][42][43] . In conclusion, the systemic loss of CTH in mice caused approximately 50% reduction in renal H 2 S levels but did not influence immediate outcomes of ischemic AKI; however, it reduced tubular damage moderately and suppressed the renal expression of inflammatory cytokines. Future studies should clarify the role of CTH on the long-term outcome of renal impairment in AKI.

Methods
Mice. Cth +/− and Cth −/− mice were generated and characterized earlier 11 . In this study, Cth +/− males and females were bred to obtain Cth +/+ , Cth +/− , and Cth −/− littermates. Mice were allowed free access to standard chow and water. The mice were kept in a 12:12-h light-dark cycle. All works involving animals have been approved by the Berlin Animal Review Board in 2012 (No. G 0444/12) and conducted in accordance with the American Physiological Society standards.
Renal IRI Model. Male mice (age between 12-15 weeks) were used. Anesthesia was performed with isoflurane (2.3%) in air (350 ml/min) 44 . Each mouse was operated separately to ensure similar exposure to isoflurane (35.7 ± 2.3 min, mean ± SD) 45 . In order to keep body temperature stable at 37 °C and monitor it during surgery, a temperature controller with heating pad (TCAT-2, Physitemp Instruments) was used. Rectal body temperature was continuously monitored during surgery using a sensor-based thermistor (36.9 ± 0.4 °C at beginning of the surgery, 37.0 ± 0.4 °C after uninephrectomy, 37.1 ± 0.3 °C five minutes after clamping the left renal pedicle and Scientific RepoRts | 6:27517 | DOI: 10.1038/srep27517 37.1 ± 0.1 °C at the end of surgery). After right-sided uninephrectomy, ischemia was induced by clipping the pedicles of the left kidney for 20 minutes with non-traumatic aneurysm clips (FE690K, Aesculap). Reperfusion was confirmed visually. After surgery, mice had free access to water and chow. We applied body-warm sterile physiological saline solutions and preemptive analgesia with tramadol (1 mg/kg) for every mouse. Sham operation was performed in a similar manner, except for clamping the renal pedicle. Mice with bleeding during surgery, with incomplete renal reperfusion, with excessive exposure of isoflurane of any reason, with significant temperature fluctuation during surgery, or with signs for infection 24 h after IRI, were immediately euthanized and were not used for further analysis. After 24 h of reperfusion, mice were sacrificed, and kidney and blood samples were collected for further analysis. The kidneys were divided into three portions. One third of the kidney was placed in optimum cutting temperature (OCT) compound for immunohistochemistry, one third was immersed in 4% phosphate-buffered saline (PBS)-buffered formalin for histology, and the rest was snap-frozen in liquid nitrogen for RNA preparation.
Quantitative Real-Time (qRT)-PCR. Total RNA from snap-frozen kidneys were isolated using RNeasy RNA isolation kit (Qiagen) according to manufacturer's instruction after homogenization with a Precellys 24 homogenizator (Peqlab). RNA concentration and quality was determined by NanoDrop-1000 spectrophotometer (Thermo Fisher Scientific). Two micrograms of RNA were transcribed to cDNA (Applied Biosystems). Quantitative analysis of target mRNA expression was performed with qRT-PCR using the relative standard curve method. TaqMan and SYBR green analysis was conducted using an Applied Biosystems 7500 Sequence Detector (Applied Biosystems). The expression levels were normalized to 18S or to beta-actin. Primer sequences are provided in Supplementary Table 1.
Western Blot. Sham and IRI-damaged kidneys were lysed with RIPA buffer (Sigma) supplemented with Complete ® protease inhibitor (Roche), 1 mM phenylmethylsulfonyl fluoride (PMSF), phosphatase inhibitor cocktail 3 (Sigma) and were homogenized using a Precellys 24 homogenizator. Fifty micrograms of protein samples were separated by 12% SDS-PAGE. After wet transfer, non-specific binding sites of the nitrocellulose membrane were blocked with 5% non-fat skim milk in Tris-buffered saline containing 0.1% Tween (TBST). The membrane was then incubated with primary antibody (anti-CTH, 1:500 (ab80643) Abcam or anti-CTH carboxyl terminus rabbit polyclonal antibody that recognizes amino acids 194-398 of a rat 398-amino acid CTH protein, 1:1,000 46 and anti-beta-actin, 1:2,000 (4970) Cell Signaling). Secondary antibody was from LI-COR Biosciences (anti-rabbit, 1:5,000). Images were acquired by Odyssey infrared imaging system (LI-COR Biosciences). Beta-actin was used as a loading control. Membranes were first probed with anti-CTH antibody and detected for their signals, and then stripped for re-probing with anti-beta-actin antibody (as loading controls). Successive stripping was confirmed by the absence of signals in the stripped membranes.
TNF-alpha Measurement. TNF-alpha levels in the supernatants of macrophages (that were used for qRT-PCR analyses) were measured using the Mouse TNF alpha ELISA Ready-SET-Go! ® Kit (eBioscience). Thawed samples were homogenized and centrifuged, and the supernatants were analyzed for fluorescence signals using Ex. 465 nm/Em. 525 nm using a spectrofluorometer 48 . Full spectrum was also analyzed to ensure that the measured fluorescence is indeed the product of the reaction between the probe and H 2 S. Further experiments with spiking the samples with H 2 S donor NaHS (10 and 50 μ M) were performed to determine the accuracy of our measurements. Serum Creatinine. Blood samples were taken from left ventricle at the time of termination. After clotting on room temperature for at least 15 min blood was centrifuged at 2,000 × g for 10 min to obtain serum. Serum creatinine was measured by external clinical laboratory (Labor 28 GmbH, Berlin).
Histology. Formalin-fixed, paraffin-embedded sections (2 μ m) of kidneys were subjected to Masson's trichrome stain using standard protocols. The severity of tubular injury was assessed by a renal pathologist who is blinded to the genotype of the mice. Tubular necrosis was evaluated in a semi-quantitative manner by determining the percentage of tubules in the cortex where epithelial necrosis, loss of the brush border, cast formation, and tubular dilation was observed. A five-point scale was used: 1, normal kidney; 2: 1 to 25%; 3: 25 to 50%; 4: 50 to 75%; and 5, 75 to 100% tubular necrosis.
Statistics. Statistical analysis was performed using GraphPad 5.04 (GraphPad Software) and SPSS 13.0 (SPSS) softwares. Normality of the data was evaluated by Kolmogorov-Smirnov test. To test the presence of an outlier, Grubbs' test was used. Study groups were analyzed by one-way ANOVA using Tukey's post-hoc test or by Games-Howell post-hoc test if homogeneity of variances was violated, with the exception of tubular necrosis data. Those were analyzed using Kruskal-Wallis test and Mann Whitney U-test. Data are presented as mean ± SEM, or median and interquartile range in case of tubular necrosis data. P values < 0.05 were considered as statistically significant.