Detrimental effects of prolonged warm renal ischaemia-reperfusion injury are abrogated by supplemental hydrogen sulphide: an analysis using real-time intravital microscopy and polymerase chain reaction


Alp Sener, Departments of Surgery, Microbiology and Immunology, LHSC-UH C4-208, 339 Windermere Road, London, ON N6A 5A5, Canada. e-mail:


What's known on the subject? and What does the study add?

Hydrogen sulphide (H2S) has recently been classified as a member of the gasotransmitter family. Its physiological and pathophysiological effects are rapidly expanding with numerous studies highlighting the protective effects of H2S on ischaemia-reperfusion injury (IRI) in various organ systems, e.g. heart, liver, CNS and lungs. The mechanisms behind its protective effects reside in its vasodilatory, anti-inflammatory and anti-oxidant characteristics. These specific mechanistic profiles appear to be different across different tissues and models of IRI.

We recently showed that supplementation of preservation solutions with H2S during periods of prolonged cold renal storage and subsequent renal transplantation leads to a massive and significant survival, functional and tissue protective advantage compared with storage in standard preservation solution alone. However, there have only been a few studies that have evaluated the effects of H2S against warm renal IRI; although these studies have focused primarily upon shorter periods of warm renal pedicle clamping, they have shown a clear survival benefit to H2S supplementation. The present study adds to the existing literature by evaluating the effects of H2S in a model of warm IRI with clinically relevant, prolonged warm ischaemia-reperfusion times (1 h ischaemia, 2 h reperfusion). We show an unprecedented view into real-time renal and hepatic perfusion with intravital microscopy throughout the reperfusion period. We show, for the first time, that supplemental H2S has multiple protective functions against the warm IRI-induced tissue damage, which may be clinically applicable to both donation after cardiac death models of renal transplantation, as well as to uro-oncological practices requiring surgical clamping of the renal pedicle, e.g. during a partial nephrectomy.


  • • To determine the protective role of supplemental hydrogen sulphide (H2S) in prolonged warm renal ischaemia-reperfusion injury (IRI) using real-time intravital microscopy (IVM).


  • • Uninephrectomised Lewis rats underwent 1 h of warm ischaemia and 2 h of reperfusion during intraperitoneal treatment with phosphate buffer saline (IRI, n= 10) or 150 µmol/L NaHS (IRI+H2S, n= 12) and were compared with sham-operated rats (n= 9).
  • • Blood was collected for measurement of serum creatinine (Cr), alanine aminotransferase (ALT) and aspartate aminotransferase (AST).
  • • IVM was performed to assess renal and hepatic microcirculation.
  • • Kidneys were sectioned for histology and real-time quantitative polymerase chain reaction for markers of inflammation.


  • • The mean (sd) Cr concentration raised to 72.8 (2.5) µmol/L after IRI from 11.0 (0.7) µmol/L (sham) but was partially inhibited with H2S to 62.8 (0.9) µmol/L (P < 0.05).
  • • H2S supplementation during IRI increased renal capillary perfusion on IVM, and improved acute tubular necrosis and apoptotic scores on histology (P < 0.05).
  • • Supplemental H2S decreased expression of the pro-inflammatory markers toll-like receptor 4, tumour necrosis factor α, interleukin 8, C-C chemokine receptor type 5, interferon γ and interleukin 2 (P < 0.05).
  • • Distant organ (liver) dysfunction after renal IRI was limited with H2S supplementation: blunting of the ALT and AST surge, decreased hepatic sinusoidal vasodilation, and decreased leukocyte infiltration in post-sinusoidal venules (P < 0.05).
  • • H2S supplementation directly inhibited interleukin 8-induced neutrophil chemotaxis in vitro (P < 0.05).


  • • These findings are the first to show the real-time protective role of supplemental H2S in prolonged periods of warm renal IRI, perhaps acting by decreasing leukocyte migration and limiting inflammatory responses.
  • • The protective effects of H2S suggest potential clinical applications in both donors after cardiac death models of renal transplantation and oncological practices requiring vascular clamping.

alanine aminotransferase


aspartate aminotransferase


C-C chemokine receptor type 5


complementary DNA




donors after cardiac death


glyceraldehyde-3-phosphate dehydrogenase


hypoxanthine ribosyltransferase


Primary human umbilical venous endothelial cells


ischaemia-reperfusion injury


intravital microscopy


leukocytes per viewing field


National Center for Biotechnology Information


polymorphonuclear cells


quantitative real-time PCR


toll-like receptor 4


terminal deoxynucleotidyl-transferase-mediated dUTP nick end labelling.


Ischaemia and subsequent reperfusion induced tissue injury is an inevitable event associated with organ transplantation. The expanding deficit in the availability of healthy kidney donors has led to a surge in the use of kidneys obtained from donors after cardiac death (DCD) [1]. Unfortunately, DCD kidneys suffer from prolonged courses of warm hypoxia, which may compromise organ injury and increase the risk of delayed graft function and acute rejection [2].

Ischaemia-reperfusion injury (IRI) is initiated by renal ischaemia, leading to neutrophil activation and release of reactive oxygen species [3,4]. Subsequent reperfusion then potentiates the damage by initiating a complex inflammatory response involving vascular and humoral mediators, contributing to necrosis and apoptosis of renal tubular cells, cellular oedema and tubular obstruction [5–7]. Traditionally, strong induction immunosuppressive regimens have been implemented to overcome ischaemia induced immune cell activation in DCD kidney recipients [8,9]. However, the increasing use of DCD kidneys for organ transplantation places further emphasis on other novel strategies that improve graft survival [2]. In the past, gasotransmitters, e.g. nitric oxide (NO) and carbon monoxide (CO), have been studied and have shown potential benefits in protecting tissue from warm IRI [10–13]. More recently, the emergence of hydrogen sulphide (H2S), as the third member of the gasotransmitter family, has stimulated further interest to investigate its protective effects on IRI [13–15].

H2S is produced mainly by cystathionine γ-lyase in vascular and non-vascular smooth muscle and has been shown to play a role in vasodilatation, angiogenesis, and neuromodulation [16–18]. In recent years, the protective actions of H2S in abrogating IRI in other organ systems, through mechanism including mitochondrial membrane stabilisation via ATP-sensitive potassium (K+ATP) channels, free radical scavenging characteristics, and anti-apoptotic and anti-inflammatory cytokine release, have been reported [19,20]. Our own observations support that supplemental H2S to standard cold renal preservation solution has the ability to significantly improve both transplant recipient and graft survival, even at extremes of cold storage [14,21,22]. Although there is evidence to support a role for H2S in mitigating tissue injury associated with warm renal IRI, these studies were carried out using short, clinically insignificant periods of warm IRI, and thus require further validation [14,21,22].

In the present study, we established a model of warm renal IRI congruent with clinically relevant, prolonged warm ischaemic times, where we also show an unprecedented view into real-time renal and hepatic perfusion using intravital microscopy (IVM). Our aim was to determine whether supplementation with H2S would impart any tissue protective effects against prolonged warm renal IRI in a uninephrectomised model, as measured by serum markers of renal injury, real-time IVM, and correlated with histopathology and inflammatory cytokine profiles.


Adult male Lewis rats (200–250 g; Charles River Laboratories International Ltd, USA) were maintained in accordance with the Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council. The experimental protocol followed the guidelines of the Council on Animal Care of our institution.

Rats were randomly assigned to one of three groups: renal IRI (i.p. PBS, n= 10), IRI+H2S (i.p. PBS + 150 µmol/L NaHS, n= 12) or sham-operated (n= 9). Within these groups, rats were further subdivided to IVM examination of either the kidney (sham: n= 4, IRI: n= 6, IRI+H2S: n= 6) or liver (sham: n= 5, IRI: n= 4, IRI+H2S: n= 6). IVM analysis of both organs in one rat was avoided to limit additional inflammation due to tissue handling.


Rats were anaesthetised by inhalation of 50% O2–50% N2 mixed with 5% isoflurane and maintained under anaesthesia with 2% isoflurane during surgery and 1% isoflurane during reperfusion. Body temperature was maintained at 37 ± 1 °C using a 50-W heat lamp. A right nephrectomy was initially performed via a midline abdominal incision to remove confounding protective effects of a functioning contralateral kidney and renorenal reflex. The left renal pedicle was subsequently occluded via atraumatic clamping for 1 h followed by 2 h of reperfusion. During occlusion and reperfusion, the abdomen was filled with 10 mL PBS (or with 150 µmol/L NaHS) and the incision was sutured to minimise secondary exposures and heat loss. The dose of NaHS was determined by previous models of renal IRI [14]. Immediately before IVM recording, rhodamine 6G (50 µg/mL; given 0.1 mL/100 g weight) was injected into the penile vein for contrast enhancement of circulating leukocytes. In rats designated for IVM of liver, the abdomen was reopened to exteriorise the liver. The left lobe was exposed on a Plexiglas stage and covered with Saran Wrap to prevent dehydration. In vivo IVM was performed using a Nikon Inverted Microscope (Nikon Eclipse TE200) and recorded by a charge-coupled device camera for video. The microscope had a visualisation depth of 0.06 mm. The same procedure was followed in different experimental rats for renal IVM recordings. At the end of the experiments, rats were killed with high dose inhalation anaesthetic. Blood samples (1 mL) were taken by cardiac puncture for analysis of serum creatinine (Cr), alanine aminotransferase (ALT), and aspartate aminotransferase (AST), centrifuged (6000g for 3 min) to separate plasma/serum and stored at –80 °C for later batch analysis by London Health Science Centre Laboratories.


To determine the exogenous effects of H2S on post-ischaemic hepatic and renal microvascular morphology, previously established methods were used to measure hepatic sinusoid diameter, post-sinusoidal venule leukocyte trafficking, renal peritubular capillary perfusion and renal capillary diameters [23]. After reperfusion, IVM recorded microcirculation of the liver for 30 min and of the kidney for 2 h. In the IRI+H2S group, the H2S solution was aspirated off before recording. Twelve videos were captured at 30 frames/s for each liver assessment, including six for post-sinusoidal venules and six for the sinusoids. Six videos of renal microvasculature were recorded for kidney function. The videos, captured via x40 objective, were analysed by a single observer in a ‘blinded’ manner.

Microscopy of liver sinusoids and post-sinusoidal venules

Random views of sinusoids were recorded in the peri-central and peri-portal zones of liver acini, as well as post-sinusoidal venules for analysis as previously described [24]. Upon video playback, a representative frame with optimal resolution was chosen for assessment. Sinusoid diameter was measured from six randomly allocated frames per rat liver (J-image software, The National Institutes of Health). The liver surface was briefly epi-illuminated with a 100-W mercury lamp via 366-nm excitation and 450-nm emission bandpass filters. The video-taped images were analysed frame-by-frame to quantify post-sinusoidal venule leukocyte rolling and adherence [25,26]. Rolling and adhered leukocytes were counted using the imaging software. In total, six 30-s videos for each rat were assessed for quantification of the hepatic inflammatory response after renal IRI.

Microscopy of the renal peritubular capillaries

As previously described, renal peritubular capillary perfusion was analysed using a stereological point-counting grid, such that the density of points counted was proportional to the density of peritubular capillaries within the viewing field [27]. Capillaries were selected by drawing three parallel lines on the video monitor perpendicular to the capillary axis and counting all capillaries that crossed. Continuously perfused capillaries maintained constant erythrocyte perfusion throughout the 1-min video, whereas non-perfused capillaries were devoid of observable blood flow. Kidney peritubular capillary perfusion was expressed as a proportion of the total number of capillaries counted. For each video, one randomly selected frame was analysed for peritubular capillary diameter. Diameters were measured in three equally distanced locations along a minimum of three vessels within each picture by J-Image (The National Institutes of Health).


At the time of death, the kidney was removed and divided sagittally and half was fixed in 10% formalin. The other half was frozen at –80 °C for later mRNA analysis. The fixed tissues were sectioned, embedded in paraffin and stained with haematoxylin and eosin or with terminal deoxynucleotidyl-transferase-mediated dUTP nick end labelling (TUNEL) for apoptosis (FragEL DNA Fragmentation Detection Kit). All sections were graded for acute tubular necrosis and apoptosis by an experienced clinical renal pathologist in a ‘blinded’ fashion.


RNA isolation and complementary DNA (cDNA) synthesis

For each kidney sample, 0.5–1.0 mg of tissue was homogenised using TRIZOL® reagent and total RNA was isolated. The RNA quantity was determined by spectrophotometry at 260 nm. The cDNA was synthesized using 1 µg total RNA in a reaction that contained 1 µL Oligo dT12–18, 1 µL dNTP (10 mm) and distilled water to a volume of 12 µL. The mixture was heated to 65 °C for 5 min and chilled on ice followed by addition of 4 µL 5X first-strand buffer, 2 µL dithiothreitol (DTT) (0.1 m) and 1 µL RNase Inhibitor (40 U/µL). The contents were incubated at 42 °C for 2 min followed by the addition of 1 µL SuperScript II Reverse Transcriptase for a total of 20 µL. The reverse transcription was performed at 42 °C for 50 min and inactivated at 70 °C for 15 min. The resulting cDNA concentration was measured by spectrophotometry and the solution was diluted to 100 ng/µL with RNase-free deionized water.


Primer sequences were designed using Primer-BLAST software (National Center for Biotechnology Information [NCBI], Bethesda, MD, USA). The Stratagene Real-Time PCR System was used to detect the accumulated fluorescence from qScript One-Step SYBR Green qRT-PCR Kit (Quanta BioSciences Inc.) according to the manufacturer's instructions. Each PCR plate contained one sample of each treatment group (Sham, IRI, IRI+H2S) and a no-template control organised in triplicates. RT-PCR was initiated at 50 °C for 10 min and 95 °C for 5 min and run for 40 cycles of 95 °C for 10 s and 60 °C for 30 s. The assays were run against two housekeeping genes: hypoxanthine ribosyltransferase (HPRT1) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as endogenous controls. Given that HPRT1 was recently shown to be a more stable housekeeper gene to GAPDH in kidney cells in hypoxic conditions, we used it as our endogenous control, although we saw very minimal variability in GAPDH expression with hypoxia (results not shown) [28]. The analysis was performed using the ΔΔCt methods. The primers were constructed via Primer-BLAST (NCBI): GAPDH, F:5′-TCTTGTGCAGTGCCAGCCTC-3′ and R:5′-GTCACAAGAGAAGGCAGCCCTGG-3′; HPRT1, F:5′-CCCTCAGTCCCAGCGTCGTGATTA-3′ and R:5′-CCCCTTCAGCACACAGAGGGC-3′; toll-like receptor 4 (TLR4), F:5′-GGGGCAACCGCTGGGAGAGA-3′ and R:5′-AACCAGCGGAGGCCGTGAGA-3′; TNFα, F:5′-TTCGGGGTGATCGGTCCCAAC-3′ and R:5′-TGGTGGTTTGCTACGACGTGG-3′; interleukin 8 (also known as cytokine-induced neutrophil chemo attractant 1 [CINC1] or keratinocyte-derived cytokine [KC]), F:5′-TGCACCCAAACCGAAGTCATAGCC3′ and R:5′-GCGTTCACCAGACAGACGCCA-3′; C-C chemokine receptor type 5 (CCR5), F:5′-GGACTGAATAATTGCAGTAGTTC-3′ and R:5′-TGTTTTCGGAAGAACACAGAG-3′; interferon γ, F:5′-AGTTCGAGGTGAACAACCCACAG-3′ and R:5′-ATCAGCACCGACTCCTTTTCCG-3′; interleukin 2, F:5′-CCCAAGCAGGCCACAGAATTG-3′ and R:5′-AGTCAAATCCAACACACGCTGC-3′.


Human neutrophils (polymorphonuclear cells, PMN) were radiolabeled by incubating the cells at 2 × 107 cells/mL in a 1.11 MBq Na51CrO4/mL PBS suspension with 1 µM glucose at 37 °C for 45 min. Cells were washed with cold PBS and divided into two tubes, one additionally treated with 150 µM NaHS. Primary human umbilical venous endothelial cells (HUVEC) grown on polyethyleneterephthalate migration inserts were divided into four groups and treated accordingly in duplicates: I, Interleukin 8-negative controls in the absence of H2S (n= 6); II, Interleukin 8-negative controls with addition of 150 µm NaHS (n= 4); III, interleukin 8-positive (20 ng/mL of human recombinant interleukin 8) without H2S (n= 4), and IV, interleukin 8-positive treatment with H2S (n= 6). Interleukin 8 was used to induce strong transendothelial migration [29]. All cytokines were re-suspended in PBS with 0.1% BSA. The final added volume of M199(+) (cat#: M5017, Sigma) was 800 µL in the insert and 850 µL in the basal compartment below the insert.

After treatment with cytokines and NaHS, the migration plate was added to a hypoxic environment in a sealed Plexiglas chamber that was continuously purged (1 L/min) with an anoxic gas mixture (92% N2, 5% CO2, and 3% H2) for 4 h. Temperature was maintained at 37 °C in the chamber. HUVEC were then re-oxygenated by exposure to room air in a 37 °C CO2-cell culture incubator. To each insert was added 5 × 105 radiolabeled PMN; PMN that had been treated with 150 µm NaHS were added to inserts in which the HUVEC had also been treated with NaHS, and untreated PMN were added to inserts with untreated HUVEC. PMN were given 1.5 h to migrate, inserts were discarded and cells in the basal compartment were treated with 850 µL 1 m NaOH for lysis. Radioactivity (counts/min) in the basal compartment was assessed using a 1480 Wizard®γ-counter (Wallac, Turku, Finland), and the percentage of PMN migration was calculated by comparison to a 100% migration control of 500 000 radiolabeled PMN. The percentage migration is the mean of each duplicate.


Data in figures are presented as the mean (sd). anova and the t-test were used to statistically analyse independent groups. A P < 0.05 was considered to indicate statistical significance.



The mean (sd) serum Cr concentration was significantly higher at 2 h after warm renal IRI at 72.8 (5.0) µmol/L (P < 0.05) when compared with the sham rats (11.0 [0.7] µmol/L; Fig. 1), which was curbed in the IRI+H2S group (62.8 [3.1] µmol/L; P < 0.05 vs IRI).

Figure 1.

Exogenous H2S reduced the amount of renal dysfunction after IRI in rats. The mean creatinine concentration (µmol/L) from blood samples collected after sham operations (Sham), 1-h ischaemia/2-h reperfusion (IRI), and IRI with H2S (150 µmol/L of NaHS i.p. at time of renal pedicle clamping and during reperfusion; IRI+H2S). Data are expressed as the mean ±sd, *P < 0.05 vs Sham, †P < 0.05 vs IRI.


Renal peritubular capillary perfusion was assessed using IVM (Fig. 2). In sham-operated rats, a mean (sd) of 96.1 (2.3)% of visualised capillaries were continuously perfused capillaries, whereas kidneys that endured warm IRI had a significantly lower percentage of perfused capillaries (62.4 [16.3]%, P < 0.05), indicating a disruption in steady blood flow. Treatment with H2S at the time of warm IRI significantly improved the proportion of perfused capillaries to 83.8 (5.4)% (P < 0.05). Conversely, the population of non-perfused capillaries was significantly greater after warm IRI at 37.5 (16.3)% (P < 0.05) compared with sham levels of 3.9 (2.3)%, and was this was significantly less after NaHS treatment, at 16.2 (5.4)% (P < 0.05).

Figure 2.

Exogenous H2S improves microvascular flow and attenuates the visual changes in renal vascular morphology caused by renal IRI. The graph shows the analysis of renal peritubular capillary perfusion from IVM recordings of kidneys from sham-operated rats, rats with IRI, and rats with IRI+H2S. Peritubular capillaries within each frame were identified for inclusion based on a stereological point-counting grid. Data are expressed as the mean percentage ±sd of the total number of capillaries included, *P < 0.05 vs Sham, †P < 0.05 vs. IRI.


Warm renal IRI led to a rise in cytoplasmic vacuolization in the proximal convoluted tubules and Bowman's capsule, expansion of the Bowman's capsule and stasis within the interstitial capillary lumina, which was associated with a higher acute tubular necrosis score (1.28 [0.19] vs 0 in sham). The acute tubular necrosis score was a little less with H2S supplementation and was borderline significant, at 1.07 (0.26) (P= 0.07; Fig. 3A). TUNEL staining showed heavy localisation of apoptotic collecting ducts in the renal medulla after warm IRI, which was significantly lower when the rats were treated with H2S (apoptosis scores: 1.92 [0.25] vs 1.05 [0.25], respectively, P < 0.05; Fig. 3B).

Figure 3.

Supplemental H2S reduces histological signs of tissue injury after warm renal IRI. A, haematoxylin and eosin-staining revealed increased cytoplasmic vacuolization of the proximal convoluted tubules and Bowman's capsule, expansion of the Bowman's capsule, and stasis within the interstitial capillary lumina were observed in IRI kidneys and were associated with higher acute tubular necrosis scores, which were minimally diminished with supplemental H2S (#P= 0.07 vs IRI). B, The appearance of apoptotic cells in the renal medulla were reduced with supplemental H2S. Histological apoptosis scores are compared for kidneys of sham, IRI, and IRI+H2S groups. All representative images were obtained from the three groups were captured from the medullary region at x40 objective. Data are expressed as the mean ±sd, *P < 0.05 vs Sham, †P < 0.05 vs IRI.


RNA was isolated from kidney homogenates for RT-PCR analysis. We found lessened expression of early inflammatory markers, e.g. TLR4 and TNFα (Fig. 4). TLR4 expression was lower with H2S supplementation, from a mean (sd) of 3.39 (3.70) folds during IRI (compared with sham) to 0.26 (0.25) folds with H2S (P < 0.05; Fig. 4). TNFα expression was lower with H2S supplementation, from a mean (sd) of 1.55 (1.13) folds during IRI to 0.36 (0.29) folds with H2S (P < 0.01).

Figure 4.

Expression of innate immune markers, TLR4 and TNFα, are down-regulated with H2S supplementation after 1 h of warm renal IRI as quantified with real-time qPCR. Data are expressed as the mean ±sd, †P < 0.05 vs IRI.


Additionally, several markers involved with T cells and acquired immunity, e.g. CCR5, interferon γ, and interleukin 2, were down-regulated with H2S supplementation. CCR5 was down-regulated from 1.67 (1.35) folds (compared with sham,) during IRI to 0.61 (0.49) folds with H2S supplementation (P < 0.05) (Fig. 5). IFNγ expression was lower with H2S supplementation, from 1.43 (0.59) folds during IRI to 0.19 (0.16) folds with H2S (P < 0.01). Interleukin 2 expression was also down-regulated from 1.08 (0.71) folds during IRI to 0.14 (0.12) folds with H2S supplementation.

Figure 5.

Expressions of T-cell-related inflammatory and immune markers are down-regulated by real-time qPCR with H2S supplementation: CCR5, interferon γ (IFNγ), and interleukin 2 (IL2). Data expressed as the mean ±sd, †P < 0.05 vs IRI.


Serum ALT increased from baseline of 46.0 (8.4) U/L to 100.8 (34.2) U/L after 1 h of warm renal ischaemia in the IRI group (P < 0.05), but remained unchanged in the H2S+IRI rats, at 54.9 (12.5) U/L (P < 0.05 vs IRI; Fig. 6A). Similarly, serum AST significantly increased from a baseline of 80.6 (10.1) U/L to 672.3 (110.6) U/L after IRI (Fig. 6B, P < 0.05), but was significantly suppressed with H2S treatment, at 169 (53.4) U/L (P < 0.05 vs IRI).

Figure 6.

Liver injury and the inflammatory response after renal IRI are attenuated with exogenous H2S supplementation. A and B, the mean serum ALT and AST from blood collected after sham operation (Sham), 1-h ischaemia and 2-h reperfusion (IRI), and IRI with H2S supplementation (IRI+H2S). C, Vasodilation of liver sinusoids observed after warm renal IRI is suppressed with exogenous H2S. Sinusoid diameters were measured from IVM recordings representing the sham, IRI and IRI+H2S. D, IVM videos were analysed frame-by-frame to quantify hepatic post-sinusoidal venule leukocyte adherence (AL), rolling (RL) and total number of leukocytes (TL) per observation area. E, Representative images were captured at x40 objective from the liver of sham, IRI, and IRI+H2S treatment groups. The liver surface was epi-illuminated to visualise leukocytes labelled with rhodamin-6G. All data are expressed as the mean ±sd, *P < 0.05 vs Sham, †P < 0.05 vs IRI.

The rise in liver transaminases after warm renal IRI was complemented by a significant increase in the local inflammatory response on IVM with vasodilation of peri-portal sinusoids and a dramatic sequestration of leukocytes in post-sinusoidal venules (Fig. 6C–E). Compared with sham-operated rats, the mean peri-portal sinusoid diameter was significantly larger after warm renal IRI [sham: 7.9 (0.4) µm, IRI: 10.36 (1.1) µm, P < 0.05, Fig. 6C). Although the change in sinusoid diameter was modestly reduced upon administration of H2S solution during warm IRI, at 8.8 (0.7) µm (P < 0.05), it remained statistically higher compared with the sham value (P < 0.05). Quantification of the IRI-induced inflammatory response in the hepatic venules showed a significant rise from sham, in adhered leukocytes as well as rolling leukocytes to 39.7 (4.5) and 15.1 (0.7) leukocytes per viewing field (LPF) (both P < 0.05). The application of H2S solution reduced the effects of warm IRI, resulting in decreased adhered leukocytes and rolling leukocytes (25.2 [7.6] LPF [P < 0.05] and 6.0 [4.1] LPF [P= 0.05], respectively; Fig. 6D). The total number of leukocytes in the observation area was significantly higher after warm IRI, at 1.26 (1.4) LPF in sham vs 54.8 (4.4) LPF after warm IRI (P < 0.05), whereas the IRI+H2S group had less total post-sinusoidal leukocyte migration at 31.2 (8.7) LPF (P < 0.05).


RT-PCR of kidney homogenates showed an upregulation of interleukin 8 expression during IRI [2.16 (1.18) folds], which was reduced with H2S supplementation to 0.71 (0.75) folds (P < 0.01). In an in vitro hypoxic transendothelial migration assay, neutrophils in the presence of interleukin 8 (n= 6) had 28.16 (7.12)% migration through the endothelial layer (Fig. 7). The presence of H2S reduced the migration to 17.23 (4.48)%, a relative reduction of 38.8% (P < 0.001). Neutrophils in the absence of interleukins 8 showed minimal migration through hypoxia treated HUVECs, as was expected [1.41 (0.13)% without H2S, 1.59 (0.91)% with H2S supplementation).

Figure 7.

A, RT-PCR of kidney homogenates shows that warm renal IRI leads to an upregulation of tissue interleukin 8 (IL-8), which is clearly diminished with H2S supplementation. B, H2S supplementation inhibited interleukin 8-mediated neutrophil transendothelial migration across primary HUVECs, under hypoxic conditions. The percentage migration of neutrophil chemotaxis in the presence of interleukin 8 was reduced from 28.16 (7.12)% to 17.23 (4.48)% with H2S supplementation. Data are expressed as the mean ±sd, †P < 0.01 vs IRI, P < 0.001 vs Hypoxia.


Despite a surge of recent progress, the protective role of exogenous H2S against prolonged warm IRI is largely unknown. In the present study, we report for the first time, a visually demonstrable, real-time protective effect of H2S after warm renal IRI. Specifically, we found that H2S supplementation during warm renal ischaemia and subsequent reperfusion limited the initial solitary kidney renal dysfunction caused by IRI, and markedly improved real-time microvascular flow in the kidney peritubular capillaries and decreased changes on kidney histology in a uninephrectomised model. These protective effects were manifested as improved post-IRI serum parameters of renal and hepatic function. In concert, we also found histological and mRNA evidence for the anti-inflammatory effect of supplemental H2S in mitigating warm ischaemic tissue injury.

The beneficial and protective effect of the endogenously produced small molecules including: NO, CO, and H2S, as well as of other vasoactive agents, e.g. erythropoietin, adenosine and verapamil, against short courses of warm renal IRI has been reported [10–12,22,30–33]. As with the other two members of the small molecule family, H2S has also been shown to possess bimodal effects: at lower concentrations it is cytoprotective by inducing a hypometabolic state via a myriad of mechanisms including alterations of the electron transport chain, whereas at higher levels it is toxic and leads to cell injury and death [21,34,35]. Similar to other reports, we showed that H2S was protective in renal IRI and preserved renal function (Fig. 1) [31]. IRI is known to disrupt microvascular and endothelial function, and through IVM, we observed improvements in vasodilation and microvascular flow with H2S supplementation (Fig. 2) [23,36]. These protective effects may be attributable to a combination of interactions with NO and CO as well as through the opening of K+ATP channels, which are readily present on the surface of vasa recta pericytes in the kidney, probably to promote perfusion during states of hypoxaemia and hence preserve GFR [37,38].

Although landmark studies by Tripatara et al. [22] provide evidence to support a protective effect of H2S on renal function through limiting acute tubular necrosis due to IRI, to the best of our knowledge, we are the first to show that the greatest degree of protection conferred by H2S supplementation against warm IRI-induced tubular apoptosis was concentrated to the renal medulla (Fig. 3). This is not surprising, as this area is especially susceptible to tissue injury due its inherent low oxygen tension and high energy requirements for maintenance of the osmotic gradient required for urine concentration [39]. Studies have shown that apoptosis commonly precedes inflammation in kidney IRI [40]. Perhaps this is too early a time point after IRI-induced damage for histological signs of acute tubular necrosis to be readily observed.

Inflammation and apoptosis are two major mechanisms that play a significant role in renal dysfunction after IRI. We found that H2S significantly down-regulated markers of both innate (e.g. TLR4, TNFα, interleukin 8) and acquired immunity (e.g. CCR5, interferon γ, and interleukin 2) (Figs 4,5). IRI is known to increase TLR4 expression in response to endogenous damage-associated molecular pattern molecules (DAMPs) to initiate inflammation and apoptosis via the MyD88 pathway in tubular epithelial cells [41,42]. TNFα and interleukin 8, on the other hand, are pro-inflammatory cytokines that mediate neutrophil migration and promote renal injury after IRI [43–45]. H2S supplementation was able to dampen these responses to warm renal IRI, and down-regulate expression of TLR4 as well as other pro-inflammatory cytokines (Fig. 4). In fact, the present novel trans-well migration assay clearly showed the ability of H2S in directly inhibiting interleukin 8-induced neutrophil trans-epithelial chemotaxis (Fig. 7), which is probably due to the binding of H2S to K+ATP channels located on leukocytes, leading to decreased expression of adhesion molecules including CD11/CD18 and P-selectin on endothelial cells, and thereby minimising leukocyte rolling and adherence [46–48]. Given that leukocyte migration is one of the initial inflammatory events involved in renal IRI, the protective effect of H2S appears to reside in its ability to down-regulate multiple inflammatory cytokines involved in this important step.

The adaptive immune system is also a major contributor to IRI [49]. A recent transcriptome analysis of T-cell populations during renal IRI showed upregulation of CCR5, a chemokine receptor involved with T-cell activation and migration [50]. CCR5 blockade was shown to decrease T-cell activation and improved renal function after IRI [50]. Other well-known T-cell-mediated cytokines, e.g. interleukin 2 and interferon γ are also upregulated in IRI [44,51]. We are the first to show a clear protective role for H2S supplementation in down-regulating the expression of multiple markers of adaptive immunity and suggest that inhibition of both immune and adaptive immune responses is a major role of H2S in renal-protection during IRI.

Distant organ injury, e.g. acute liver injury, is commonly observed after renal IRI [52,53]. It has been shown that hepatic sinusoidal dilatation and leukocyte aggregation are common events in systemic inflammation and sepsis [54]. Consistent with these previous findings, we showed using IVM that 1 h of warm renal ischaemia led to elevated liver transaminases, increased leukocyte infiltration, and liver sinusoidal dilatation, all of which was curbed significantly by supplemental H2S (Fig. 6). These anti-inflammatory properties probably reside in the abilities of H2S to modulate systemic and distant inflammation via decreasing inflammatory cytokines, reducing leukocyte migration, as well as scavenging free radicals [48,55].

Interestingly, the role of H2S during inflammation is still subject to much debate, as it has also been shown to augment inflammatory conditions, e.g. acute pancreatitis, sepsis and has even been shown to enhance neutrophil migration into the peritoneal cavity [56–58]. The controversy surrounding this issue may be attributed to a variety of potential variables including H2S dosing/concentration, exposures times or even tissue specificity. Regardless, it is clear that H2S has an important role in inflammation and should be investigated as a potential target for novel therapies [59].

In summary, the present findings show that supplemental H2S has multiple protective functions in warm renal IRI including down-regulation of inflammatory markers, and diminishing renal and distant-organ dysfunction. We have also shown, for the first time, that supplemental H2S preserves microvascular flow within the kidney and limits distant-organ inflammatory responses caused by warm IRI. The mechanisms behind these protective effects appear to be rooted in reductied leukocyte migration and down-regulation of inflammatory markers including TLR4, TNFα, interleukin 8, CCR5, interferon γ and interleukin 2. The protection conferred by the addition of intraperitoneal H2S suggests that this novel mediator may have clinical applicability in both DCD models of renal transplantation, as well as in oncological practice requiring vascular clamping (i.e. partial nephrectomy for RCC), whereby its addition to standard preservation/storage solutions may have therapeutic value in improving allograft and renal functional outcomes, respectively.


This research was made possible by the Northeastern Section of the AUA and the Canadian Urological Association Foundation.


All authors declare no conflicts of interest.