Postliver transplantation acute kidney injury (AKI) severely affects patient survival, whereas the mechanism is unclear and effective therapy is lacking. The authors postulated that reperfusion induced enhancement of connexin32 (Cx32) gap junction plays a critical role in mediating postliver transplantation AKI and that pretreatment/precondition with the anesthetic propofol, known to inhibit gap junction, can confer effective protection.
Male Sprague–Dawley rats underwent autologous orthotopic liver transplantation (AOLT) in the absence or presence of treatments with the selective Cx32 inhibitor, 2-aminoethoxydiphenyl borate or propofol (50 mg/kg) (n = 8 per group). Also, kidney tubular epithelial (NRK-52E) cells were subjected to hypoxia–reoxygenation and the function of Cx32 was manipulated by three distinct mechanisms: cell culture in different density; pretreatment with Cx32 inhibitors or enhancer; Cx32 gene knock-down (n = 4 to 5).
AOLT resulted in significant increases of renal Cx32 protein expression and gap junction, which were coincident with increases in oxidative stress and impairment in renal function and tissue injury as compared to sham group. Similarly, hypoxia–reoxygenation resulted in significant cellular injury manifested as reduced cell growth and increased lactate dehydrogenase release, which was significantly attenuated by Cx32 gene knock-down but exacerbated by Cx32 enhancement. Propofol inhibited Cx32 function and attenuated post-AOLT AKI. In NRK-52E cells, propofol reduced posthypoxic reactive oxygen species production and attenuated cellular injury, and the cellular protective effects of propofol were reinforced by Cx32 inhibition but cancelled by Cx32 enhancement.
Cx32 plays a critical role in AOLT-induced AKI and that inhibition of Cx32 function may represent a new and major mechanism whereby propofol reduces oxidative stress and subsequently attenuates post-AOLT AKI.
Acute kidney injury occurs after specific surgeries and in creases patient morbidity and mortality
Anesthetized rats underwent autologous orthotopic liver transplantation in the absence or presence of treatments with a selective Cx32 inhibitor, 2-aminoethoxydiphenyl borate, or propofol
Propofol inhibited Cx32 function and attenuated postautologous orthotopic liver transplantation acute kidney injury
POSTOPERATIVE acute kidney injury (AKI) is a severe complication following liver transplantation, which influences patient survival adversely.1 Reportedly, approximately 30 to 50% of the patients undergoing liver transplantation developed AKI, however, mechanisms contribute to this complication is still unclear.2 Causes of this complication are complicated and involve multi factors, among which perioperative hypotension is considered to be one of the most important independent risk factors. Perioperative hypotension always results in renal ischemia–reperfusion (I/R) injury and mechanism of which is associated with oxidative stress.3,4 However, effective therapy to combat postliver transplantation AKI is lacking. Therefore, there is urgent need to explore the underlying mechanisms of AKI and develop effective strategies for renal protection.
Connexins are a big family of transmembrane proteins that express in all human organs and tissues. Approximately 21 isoforms have been found, and each of them forms channels with distinct regulation and permeability.5,6 Six of connexins compose a hemi-channel. Two hemi-channels in the neighboring cells dock together to form an integral gap junction (GJ), which manipulates the direct cell-to-cell signaling transfer (molecule weight is less than 1 kDa, such as calcium, cyclic adenosine monophosphate, cyclic guanosine monophosphate, glutathione, etc.). This kind of signaling transfer is not only essential for many physiological events, but also relative with development of some disease.7,8 A most recent study shows connexin32 (Cx32) expression always increases when liver or brain damage become more and more serious and these damages could be blocked by Cx32 inhibitors through decreasing oxidative stress and cell apoptosis.9,10 Cx32 expression is extremely abundant in kidney, especially in renal proximal tubules, however, roles of Cx32 in the development of AKI following liver transplantation is still unknown.11 So far, effective therapeutic strategies for perioperative renal protection, especially targeting AKI after liver transplantation are lacking.
As a commonly used anesthetic in clinical anesthesia and intensive care unit sedation, propofol not only regulates the balance between oxidants and antioxidants, but also confers protective effects for different organs via different mechanisms, such as antioxidant,12,13 antiinflammatory effects,14 inhibition of inducible nitric oxide synthase enzyme and so on.15 Xia et al.12 reported propofol reduced postischemic oxidative stress and attenuated myocardial I/R injury in patients, and Zhao et al.16 showed that propofol decreased X-ray induced cellular toxicity through inhibiting Cx32. Based on the above-mentioned findings illustrating that GJ function inhibition reduced postischemic oxidative stress and that propofol can inhibit Cx32 which is abundant in the kidney, we hypothesized that GJ composed of Cx32 played a critical role in the development of AKI after liver transplantation and propofol could protect against posttransplantation induced AKI through inhibiting Cx32, mechanism of which might be relative with propofol reduced reactive oxygen species (ROS) production and attenuated oxidative stress in the setting of AKI through Cx32 GJ inhibition. This viewpoint has never been reported until now. The new mechanistic insight gained from the current study would provide a basis for developing effective therapies to combat liver transplantation-mediated AKI.
Materials and Methods
Animals and Treatment
Animal care followed National Institutes of Health criteria for the care and use of laboratory animals in research. The study was approved by the Laboratory Animal Care Committee of Sun Yat-Sen University (Guangzhou, Guangdong, China). Male Sprague–Dawley rats (200 to 220 g) were obtained from Sun Yat-Sen University. In the initial intervention model establishment study, the animals were randomly assigned into six parallel groups (n = 8 per group) using a random number table taking into consideration of the weight of the rats. These include sham operated, reperfusion 4-, 8-, 16-, 24-, and 48-h groups. Subsequent in vivo studies were performed using the 8-h reperfusion model. The animals were then randomly assigned to four groups (n = 8 per group) to explore the role of GJ function in postautologous orthotopic liver transplantation [AOLT] renal injury, which include sham, (2-aminoethoxydiphenyl borate [2-APB]) + sham, AOLT, and (2-APB) + AOLT groups. Furthermore, to explore the effectiveness and mechanisms of propofol protection, animals were randomly assigned to four groups (n = 8 per group) which include sham, propofol + sham, AOLT, and propofol + AOLT. It is of notice that, with respect to our success rate of approximately 80 to 90% for AOLT model establishment,17,18 experiments were performed on a total of 129 animals.
Before operation, rats were treated intraperitoneally with propofol (Sigma-Aldrich, St. Louis, MO) at doses 50 mg/kg for 3 days,19 and 2-APB (Sigma-Aldrich; a relative specific inhibitor of Cx32) was given at 20 mg/kg 4 h20 before AOLT or “Scrape-and-load” assay.20 The dose of propofol (i.e., 50 mg/kg, intraperitoneally) was chosen based on a preliminary experiment showing that propofol 50 mg/kg, produced a sedative response in rats, as determined by loss of reflex responses to a painful stimulus while remaining sensitive to skin incision.19 This dosage of propofol has been widely used in in vivo studies in rats21,22 and produced beneficial effects.
Establishment of Rat AOLT Models
To minimize experimenter bias, the investigator who performed the operation was blinded to the treatment conditions. Rat AOLT models were established as previously described.17,18 Initially, we used an open face guard to administer the inhalational ether anesthesia until the rats had no response to a needle stimulus. After entering the abdominal cavity, we resected and ligated the falciform ligament of the liver, and severed the blood vessel along the esophagus. The liver was revealed until the supra hepatic vena cava (SVC) was liberated and then, the liver was placed back into its original position. We prepared a bold line to guide the SVC for blockage. When the upper region of the left renal vein was completely liberated, the inferior vena cava (IVC) was dissociated. We dissected the first hepatic portal and then separated the portal vein (PV) from the convergence of the inferior mesenteric and splenic veins. Both of the hepatic artery and biliary were also liberated successively according to their anatomic relationship. Subsequently, the portal hepatics were ligated. Microvascular clamps were used at the convergence of the inferior mesenteric, splenic veins, hepatic artery, SVC, and IVC. The PV was punctured with a 24-gauge needle in preparation for reperfusion and one 1-mm incision was made on the IVC wall as an outflow tract. Precold 4°C Ringer lactate solution was injected during reperfusion at 2.5 ml/min until the liver color turned yellow. Finally, after extracting the needle, the opening of the PV and IVC was closed using 8-0 sutures. PV, SVC, IVC, and hepatic artery were all unclamped. On average, the anhepatic phase lasted for 20 ± 1 min, which was concomitant with significantly lower mean arterial pressure in this phase (vascular clamped) than in other phases as shown in table 1.
Mean Arterial Pressure Assay
The right femoral arteries were catheterized with a polyethylene catheter (outer diameter, 0.965 mm; inner diameter, 0.58 mm) for monitoring mean arterial pressure,23 which was recorded before this operation and when PV, SVC, IVC, and hepatic artery were clamped. Also, mean arterial pressure was recorded at different time points (1 min to 30 min), when PV, SVC, IVC, and hepatic artery were unclamped.
Rats were treated with propofol (50 mg/kg) or 2-APB (20 mg/kg) intraperitoneally. Four hours later, kidneys were excised and freshly sliced. We placed a 27-gauge needle dipping into a solution containing 0.5% Lucifer Yellow (Invitrogen, Carlsbad, CA). The needle was used to mechanically damage a small area of each slice to apply dyes. After 5 min incubation, kidney slices were rinsed in saline, fixed in 4% paraformaldehyde for 30 min, frozen in optimum cutting temperature compound, cryosectioned into 10 μm sections, rinsed in saline again, mounted, and imaged by fluorescence microscopy.20 Quantitative analysis of the distance of dye spread was performed between the dye transfer front and the scrape line.
Assessment of Kidney Damage
Creatinine and blood urea nitrogen were measured in blood samples with an automatic biochemistry analyzer (Hitachi 7600- 020/7170A, Tokyo, Japan). Kidney specimens were fixed in 10% buffered formalin, embedded in paraffin, and processed for hematoxylin–eosin staining.
Measurement of Intracellular ROS Production, 15-F2t-Isoprostane, and H2O2
Intracellular ROS production was estimated by using 2,7-dichlorofluorescein diacetate (Sigma-Aldrich). Formation of dichlorofluorescin was detected by fluorescence spectrometer (HITACHI, Model No. F4500, Tokyo, Japan) equipped with a fluorescein isothiocyanate filter at 488 nm and emission at 525 nm within 15 min. Dichlorofluorescin images were obtained with a fluorescence microscope (Olympus IX71, Tokyo, Japan) at a magnification of ×200 using identical acquisition settings for each section.24 15-F2t-Isoprostane (15-F2t-Isop) and H2O2 levels are determined using commercial assay kits (Cayman Chemical Company, Ann Arbor, MI) as described.25
NRK-52E cells (kidney tubular epithelial cells) were obtained from American Type Culture Collection (Manassas, VA) and cultured in Dulbecco’s Modified Eagle’s Medium/F-12 supplemented with 10% fetal bovine serum. HeLa cells expressing Cx32 under the control of a bidirectional tetracycline-inducible promoter was characterized previously26 and cultured in Dulbecco’s Modified Eagle’s Medium supplemented with 10% fetal bovine serum. Both kinds of cells were grown at 37°C in an atmosphere of 5% CO2 in air.
Cell Treatments and Survival Assay
NRK-52E cells were seeded at low density (25,000 cells/cm2, no GJ formed) or high density (125,000 cells/cm2, GJ formed) in 24-well plates to observe effects of GJ on hypoxia–reoxygenation (H/R) injury or other biochemical changes as assessed below.26 The choosing of NRK-52E cell low density at 25,000 cells/cm2 and high density at 125,000 cells/cm2 was not only based on whether or not there was GJ formation, but also based on our preliminary study findings showing that the rates of cell proliferation peaked at 25,000 cells/cm2 and maintained at a comparable level when the density was as high as 125,000 cells/cm2 (data not shown) to exclude the possibility that NRK-52E cell growth is significantly down-regulated due to reduced cell spreading when its density is lower than 20,000 cells/cm2 or higher than 200,000 cells/cm2.27
NRK-52E cells were pretreated with connexin channel inhibitors 18-α-GA (GA), 10 μM, for 1 h (Sigma-Aldrich), 2-APB, 25 μM, for 1 h (Sigma-Aldrich); a Cx32 expression enhancer, retinoic acid (RA) 10 μM, for 24 h (Sigma-Aldrich) before inducing H/R injury and “Parachute” dye-coupling assay was performed as described below.
In the preliminary dose-finding study, cells were pretreated with propofol (Sigma-Aldrich), respectively, at 1, 5, 15, 30, and 60 μM for 1 h before being processed for “Parachute” dye-coupling assay, Cell Counting Kit-8 assay (Dojindo Molecular Technologies, Tokyo, Japan), and Western blotting analysis. And the concentration of 15 μM, of propofol, was chosen for ensuring studies in NRK-52E cells and these cells were pretreated for 1 h with 15 μM propofol before being subjected to H/R injury. The choosing of propofol concentration at 15 μM was based on our dose-finding study results showing that this concentration of propofol can profoundly decrease dye coupling and also based on our previous study showing that propofol at 15 μM can reduce the cytotoxicity of radiograph irrasiation through inhibiting GJ activity.16 This concentration of propofol (15 μM) is in the range of target plasma concentration of propofol 2 to 4 μg/ml (i.e., 11 to 22 μM) as used clinically during major surgeries.28,29
H/R of NRK-52E Cells
Renal hypoperfusion is significant during liver transplantation, which is induced by hypotension. Renal hypoperfusion always results in renal I/R injury. Thus, we employed NRK-52E cell H/R model as an in vitro study to mimic renal cell I/R injury. Before the experiment, NRK-52E cells were incubated in starving medium (serum-free and glucose-free Dulbecco’s Modified Eagle’s Medium/F-12) for 12 h and then exposed to hypoxia. In all the process of H/R, cells were incubated in starving medium. NRK-52E cells were cultured in a low-oxygen condition (95% N2 + 5% CO2) for 24 h in a humidified hypoxia incubator (Galaxy 48R; Eppendorf, Hamburg, Germany). The cells were then exposed to normal oxygen condition (95% air + 5% CO2) for reoxygenation for 4 h. After the completion of the experiments, the supernatant and cells were collected separately for further analysis.30 Control groups were cultured in normoxic conditions for 28 h, coincident with duration of H/R injury, and the supernatant and cells were also harvested separately for further analysis
Cell Counting Kit-8 Assay and Lactate Dehydrogenase (LDH) Assay
Cells were seeded at low density (25,000 cells/cm2, no GJ formed) or high density (125,000 cells/cm2, GJ formed) in 24-well plates. At the end of different stimulation, Cell Counting Kit-8 and LDH assays (Roche Diagnostics, Indianapolis, IN) were carried out according to the manufacturer’s introduction as we described.25 According to the instruction of the detection kit, to calculate percent cytotoxicity, three kinds of controls (namely background control, low control and high control) were included in each experiment. The background control determines the LDH activity contained in the assay medium. The absorbance value obtained from this control was subtracted from all other absorbance values. The low control determines the LDH activity released from the untreated cells (spontaneous LDH release), whereas the high control determines the maximum releasable LDH activity in the cells (maximum LDH release). In the high control, lysis reagent was added to the samples to get the accurate estimate of maximum releasable LDH. To determine the percentage cytotoxicity, the average absorbance values of the samples and controls were respectively calculated. And, the absorbance values of the background control were subtracted from all other absorbance values. Thereafter, the resulting extent of cytotoxicity (reflected as increased LDH activity) was calculated using the following equation: cytotoxicity (%) = (sample absorbance − low control absorbance)/(high control absorbance − low control absorbance) × 100%. Thus, the value of LDH release we presented in the current study is reflected as LDH release from treatment groups relative to high control (cells treated with lysis). Then, we standardized the results of cytotoxicity (relative LDH release) of the experimental control group as 1 and the values of other groups were presented as relative values compared to experimental control, with the purpose to easily visualize the changes of cytotoxicity (relative LDH release) resulted from changes of Cx32.
“Parachute” Dye-coupling Assay
GJ function was examined with “Parachute” dye-coupling assay as described.5
Inhibition of Cx32 Expression by Small Interfering RNA Transfection
Cells were transfected with two different kinds of small interfering RNA (siRNA) targeting rat Cx32 gene (CACCAACAACACATAGAAA and GCATCTGCATTATCCTCAA, Cx32 siRNA1 and Cx32 siRNA2) or a nonspecific, control siRNA (NC group). Transfection into NRK-52E cells was carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions.
Western blotting follows the standard procedures as described.26 Anti-Cx32 (1:3,000; Sigma-Aldrich) and secondary antibody (1:2,000; Sigma-Aldrich) were used to detect Cx32 expression. Anti-β-actin (Sigma-Aldrich) and its corresponding secondary antibody were used at 1:4,000.
Quantitative data are presented as mean ± SEM. Statistical analysis was performed by using SPSS 15.0 software (SPSS Inc., Chicago, IL) and dose–response curve fitting was made by Sigmaplot 10.0 (Systat Software, Inc., Chicago, IL). And the cure was formed by the graph properties, using the function of smoothed (spline) in the shape property. Multiple comparisons among groups were analyzed using repeated measures one-way ANOVA, followed by Tukey post hoc comparisons. In our lab, the overall success rate of rat AOLT model establishment is between 80 and 90%. According to our previous studies,17 a sample size of n = 8 is needed and is sufficient to achieve a statistical power in the study of intervention effectiveness. Therefore, when designing our studies, we initially assigned 10 rats per groups while aiming for a final n = 8 in each group, bearing in mind the rate of success of the model. A two-tailed P value less than 0.05 was considered statistically significantly different.
Remote Kidney Damage following AOLT In Vivo Studies at Different Time Points
To explore effects of liver transplantation on kidneys, a well established rat AOLT model31,32 was used in this study. In this model, inferior vena cava interruption and significant systemic hypotension occurred during the anhepatic phrase which led to blood reflux disorder and caused renal ischemic injury. In in vivo studies, we found that as reperfusion time being extended, kidney pathological damage became more and more serious gradually. Eight hours after reperfusion, kidney damage became the most severe. Tissue injury recovered at 48 h after reperfusion (fig. 1A). Liver transplantation is very complex. Hypotension is one of the most important and obvious hemodynamic changes in the process, which may result in remote organs injuries, including kidneys. Thus, rat mean arterial pressure was recorded during AOLT. Results showed that when PV, SVC, IVC, and hepatic artery were clamped, mean arterial pressure decreased dramatically while recovered gradually to the normal level as vascular unclamped (table 1). Increase of creatinine and blood urea nitrogen (fig. 1, B and C) (parameters of renal functional impairment) as well as significant elevations in tissue H2O2 (fig. 1D) and 15-F2t-isoprostane (15-F2t-Isop) (fig. 1E) production, reflecting the level of oxidative stress,13 all peaked at 8 h after post-AOLT, which mirrored the patterns of pathological injury of kidney.
Changes of Cx32 Expression after AOLT In Vivo Studies at Different Time Point
Given that Cx32 expressed richly in kidney and that its overexpression has been shown to be related with organ damage, so it was determined after AOLT. As shown in figure 1F, Cx32 protein was increased after AOLT and peaked at 8 h after reperfusion, which was coincident with the most severe kidney pathological damage and functional impairment (fig. 1, A–C).
Inhibition of Renal Cx32 GJ Attenuated Remote Kidney Damage following AOLT with Concomitant Reduction of Post-AOLT Renal Oxidative Stress
Results in figure 1 provided a clue that Cx32 might play an important role in AOLT-mediated AKI. Therefore, 2-APB, a relatively specific inhibitor of Cx32 channels20,33 was used to explore roles of Cx32 in this pathology. Eight samples are contained in every group. “Scrape-and-load” assay was employed to detect GJ function composed of Cx32 in vivo. Results indicated that dye spread was decreased obviously on kidney slices obtained from rats pretreated with 2-APB (fig. 2A). Renal pathological damage, creatinine and blood urea nitrogen were also reduced obviously (fig. 2, B and C) following the treatment with 2-APB. These data indicate that inhibition of GJ function composed of Cx32, could eliminate remote kidney damage after AOLT. Of note, 2-APB treatment also significantly reduced ALOT mediated renal productions of H2O2 and 15-F2t-Isop to a level comparable to that in sham group (all P < 0.05, 2-APB + AOLT vs. AOLT; P > 0.05, 2-APB + AOLT vs. sham; fig. 2D). Dimethyl sulfoxide, the vehicle control of 2-APB had no effect on the parameters (see fig. 1, Supplemental Digital Content 1, http://links.lww.com/ALN/B94).
Cell Damage after H/R Was Cell Density Dependent
Results in figure 2 indicated that kidney damage after AOLT could be attenuated by inhibiting Cx32 GJ function. To confirm this mechanism, we tested it on NRK-52E cells, a kind of renal tubular epithelial cell line, which mainly express Cx32. Figure 3 illustrated effects of 24 h hypoxia and 4 h reoxygenation (H24R4) on NRK-52E cell survival and LDH release at low-density (25,000 cells/cm2, no GJ formed) and high-density cell culture (125,000 cells/cm2, GJ formed) respectively, which is a widely used method to observe effects of GJ.26 After being subjected to H24R4, NRK-52E cell growth was reduced at both low-density and high-density cell culture, by approximately 43 and 71% respectively (P < 0.05 high density vs. low density; fig. 3A). When cells were in contact with one another at high density, H24R4 damage was substantially greater than that at low density, manifested as greater extent of LDH release in the high-density condition than in the low-density condition (P < 0.02; fig. 3B) when the values as adjusted with their relative control. These results indicated that posthypoxic cell damage was density-dependent, being more severe in the high-density condition where GJ was formed.
H24R4 Caused Density–dependent Cell Damage through GJ
Density-dependent cell damage is always modulated by GJ.34 Thus, we investigated roles of Cx32 gap function in H24R4 damage. Different methods were used to manipulate GJ function composed of Cx32 on NRK-52E cells. Dye coupling was reduced by 18-α-GA (GA) and 2-APB, but increased by the enhancer RA (fig. 4A). At the same time, H24R4 damage was also reduced subsequent to pretreatment with GA or 2-APB but exacerbated after RA treatment in NRK-52E cells cultured in high-density (fig. 4B). However, cell growth after H24R4 did not significantly differ at low-density cell culture irrespective of treatments with either GA or 2-APB or RA as compared with untreated control group (fig. 4B). Changes in LDH release was just opposed the changes in cell growth. Following preincubation with GA or 2-APB, LDH release declined more obviously, treatment with RA resulted in dramatic increase in LDH release at high-density cell culture (fig. 4C). Although, the same treatments were applied on low-density cell culture, they did not cause significant changes in posthypoxic LDH release (fig. 4C) and dimethyl sulfoxide, the vehicle control of 2-APB, GA and RA had no effects on the parameters in figure 4 (see fig. 2, Supplemental Digital Content 1, http://links.lww.com/ALN/B94). These findings suggested that Cx32 GJ plays a key role in posthypoxic cellular damage in NRK-52E cells.
Cx32 Gene Knock-down to Inhibit Cx32 GJ Function Prevented H24R4-mediated Cell Damage in NRK-52E Cells
To confirm the impact of Cx32 function on H/R damage, we synthesized two different Cx32 siRNAs (siRNA1 and siRNA2) to specifically knock-down Cx32 expression (fig. 5A). Blocking Cx32 expression depressed dye coupling of NRK-52E cells (fig. 5, A and B). Although Cx32 knock-down per se did not affect cell growth under control condition, it significantly attenuated H24R4-induced reduction of cell growth (fig. 5C) and reduced posthypoxic LDH release (fig. 5D). The data are indicative that Cx32 GJ plays a key role in H/R-induced cell damage.
Propofol Protected against H24R4-induced Cell Injury by Inhibiting Cx32 GJ
Propofol had recently been shown to reduce kidney I/R injury induced by hyperglycemia,35 a pathological condition that is associated with increased oxidative stress. To assess whether or not propofol can reduce H24R4-mediated renal damage via inhibiting Cx32, we first examined the effects of propofol on dye coupling of NRK-52E cells. Figure 6, A–C, showed that propofol from 5 to 60 μM decreased dye coupling in a dose-dependent manner without cytotoxicity and had no significant effect on Cx32 expression in control NRK-52E cells. By contrast, propofol at a concentration as low as 15 μM dramatically decreased dye coupling in Hela cells that stably express Cx32 only, which means that the effects of propofol on Cx32 expression and Cx32 GJ was direct and specific (fig. 6D). This result suggests that propofol can profoundly inhibit Cx32 expression in situations with Cx32 overexpression such as during AKI or acute renal cell H/R injury. Also, the facts that GJ function inhibition reduced H24R4-mediated cell damage (figs. 4 and 5) and that propofol reduced dye coupling directly in Hela-Cx32 cells (fig. 6) collectively raised the possibility that propofol could manipulate H24R4 damage via regulating Cx32 GJ function. To test this hypothesis, NRK-52E cells seeded at either low or high density were pretreated with propofol 15 μM for 1 h before being exposed to H24R4. Of note, propofol did not significantly enhance cell growth in low-density cultured cells subjected to H24R4 but most profoundly increased cell growth in high-density cultured cells subjected to H24R4 (fig. 7A). Propofol only slightly (despite statistically significantly) decreased LDH release and 15-F2t-Isop production in low-density cells subjected to H24R4 but completely cancelled H24R4-induced increases in LDH and 15-F2t-Isop that at high-density cells (fig. 7A), whereas lipid, the vehicle control of propofol, had no effects on the results in figures 6 and 7 (see fig. 2, Supplemental Digital Content 1, http://links.lww.com/ALN/B94). These results suggest that inhibition of GJ function may be a major mechanism whereby propofol attenuated oxidative stress and protected against H24R4-induced cell damage.
Effects of Propofol on ROS Induced by H24R4 Damage Mediated by GJ Composed of Cx32
The facts that organ-protective effects of propofol were related to its effect in reducing ROS- mediated damage,36 together with our current findings that propofol could protect against H24R4-induced cell damage through inhibiting GJ prompted us to explore the relationship between propofol-mediated GJ inhibition and ROS production. At low-density cell culture, neither the GJ inhibitor 2-APB nor the enhancer RA or propofol had effects on intracellular ROS production in NRK-52E cells exposed to H24R4 (fig. 7B). However, at high-density cell culture, ROS level fluctuated as Cx32 function changed (fig. 7, B and C). As shown in figure 7C, propofol similar to 2-APB profoundly inhibited H24R4-induced increase of dichlorofluorescin florescence (a reflection of H2O2 production) whereas RA further increased dichlorofluorescin florescence in high-density cells, which was quantified in figure 7B. These results demonstrated that intracellular ROS was manipulated by GJ composed of Cx32. Facts in this paragraph above suggested inhibition of GJ is a major mechanism by which propofol reduced ROS, at least at the current experimental settings. We had clarified that propofol protected NRK-52E cells at low-density cell culture, notwithstanding much less than that at high-density cell culture (fig. 7A). The facts indicted propofol protected cells through different mechanisms, one of which might be relative with GJ. When NRK-52E cells (high-density cell culture) were pretreated with 2-APB or RA in combination with propofol before H24R4, 2-APB magnified propofol protection but RA diminished it obviously (fig. 7, D and E). These results strongly suggest that inhibition of Cx32 GJ may represent a major mechanism by which propofol confers cellular or organ protection where Cx32 is overexpressed. This finding might have clinical implications in guiding the optimal application of propofol for organ protection.
Propofol Decreased Remote Kidney Damage after AOLT In Vivo via Inhibiting Cx32 GJ Function
Following our above demonstration in figures 6 and 7 that propofol could protect NRK-52E cells against H/R damage through mediating Cx32 channel function, which was related to inhibition of intracellular ROS, we wanted to further confirm the protective effects and mechanism of propofol in attenuating AOLT induced remote AKI in vivo. With “scrape-and-load” assay, we demonstrated that propofol pretreatment decreased dye coupling on kidney (fig. 8A) that was associated with significant reduction in renal Cx32 expression in rats subjected to AOLT but not in sham-operated rats (fig. 8B). Furthermore, two different methods, immunohistochemisty and immunofluorescence were carried out to visualize the expression and localization of Cx32. Both of the immunohistochemisty and immunofluorescence results (see figs. 3 and 4, Supplemental Digital Content 1, http://links.lww.com/ALN/B94) showed that in the areas of the renal tubule, Cx32 expression changed obviously after AOLT, but by contrast, Cx32 expression in the renal glomerulus was much less obvious than that in the renal tubule. Cx32 expression increased in the AOLT group, which could be inhibited by propofol, whereas lipid had no effect on renal Cx32 expression and localization. These results were coincident with the changes of Cx32 assessed by Western blotting in figure 8B. Also, remote kidney damage was attenuated in the presence of propofol treatment manifested as attenuated pathological damages (fig. 8C) and reduced post-AOLT concentrations of creatinine and blood urea nitrogen (fig. 8D). In line with its prevention of AOLT induced increase in Cx32 expression and Cx32 GJ (fig. 8A), propofol completely prevented AOLT induced increase in H2O2 and 15-F2t-Isop production, but has no effect on H2O2 and 15-F2t-Isop in sham-operated rats (fig. 8E). Findings as shown in figure 8 from in vivo studied confirmed our in vitro study findings illustrated in figure 7, indicating that inhibition of Cx32 GJ function is a key mechanism by which propofol attenuated AOLT induced renal I/R injury. The facts that propofol inhibited ROS production only in high-density cells subjected to H24R4 (fig. 7B) where Cx32 GJ function was enhanced and that the enhancer RA cancelled propofol-mediated prevention of NRK-52E cells against H/R injury (fig. 7, D and E) and cancelled propofol-mediated attenuation of ROS production in NRK-52E cells subjected to H/R (data not shown) indicates that inhibition of Cx32 GJ function is a major and novel mechanism of propofol protection against AKI. This notion is confirmed in in vivo models of AOLT where propofol prevented ALOT-induced Cx32 overexpression and increased in ROS production coincidently without affecting Cx32 expression and ROS production on sham control rats (fig. 8).
Liver transplantation is always considered to be the most effective method to cure the final-stage liver disease. The incidence of AKI is always at a high level and influences patient survival severely, but the underlying mechanism remains unclear.37 In our investigation, we chose a rat AOLT model to explore Cx32 function in renal protection. Compared with pure hepatic I/R injury model, rat AOLT model not only includes most procedures of liver transplantation, but also includes blood reflux disorder. Compared with allogenic orthotopic liver transplantation, rat AOLT model has its own advantages, such as avoiding complex condition of recipients and reject reaction, keeping good repeatability and high survival of rats. It is profitable for us to investigate the role of Cx32 GJ in kidney damage following liver transplantation and the protective effects of propofol in this pathology. Our in vitro studies, incorporating multiple manipulations, provided strong evidence to show that increased Cx32 GJ function is a major mechanism of renal H/R damage, and inhibition of Cx32 GJ function could protect against renal H/R damage. It should be noted that although 2-APB had been used widely as a selective Cx32 inhibitor, which could decrease function of Cx32 in vivo and in vitro,20 it may have other nonspecific effects such as decreasing Ca2+ release or inhibiting IP3 receptors38 when used at high doses. Our incorporation of the use of two different siRNAs specifically to knock-down Cx32 expression confirmed the role of Cx32 in mediating posthypoxic cellular injury. Another important novel finding was that propofol, as a common anesthetic used in clinic, conferred protective effects against H/R damage of renal cells through inhibiting Cx32 channels and that inhibition of Cx32 is a major mechanism whereby propofol reduces ROS and oxidative stress under pathological conditions like AOLT-induced AKI.
Our joint in vitro and in vivo studies showed that GJ composed of Cx32 plays a key role in kidney damage after AOLT. GJ was always considered to be the direct cell-to-cell transfer of electrical charge or small molecules. Intercellular molecular signals transferred between neighboring cells contribute to cell growth, differentiation, normal physiology, and response to trauma in all different organs.39 Molecular signals enhancing cytotoxicity or inducing apoptosis were called “death signals,” which were mainly manipulated by GJ, especially in cancer cells.5,40
Propagation of “death signals” through GJ had been widely explored. However, the intrinsic quality of them had not been identified. These signals were usually considered to be calcium, cells metabolites or molecules triggering activities of cellular death pathways.41 We investigated the possibility of ROS as “death signals,” which was indispensable for I/R damage.42 ROS, including oxygen radicals and nonradical compounds, exist at low levels in cells but play an important role in regulating physiologic tubular functions and renal microcirculation. However, great amount of ROS produced always disrupt the redox signaling pathway, resulting in cellular molecular structures damage. Both oxygen radicals and nonradical compounds had much less molecular mass than the upper limit of GJ permeable, which mean that both of them might be transferred through GJ, acting as “death signals” function. We showed that ROS induced by H/R damage was manipulated by GJ composed of Cx32. ROS generation under pathological conditions not only damaged cells in which they were produced, but also attacked neighboring cells through the transfer mediated by GJ. In this transfer process, ROS level was raised in cells connected with GJ, resulting in damage magnification. This kind of mechanism was called “bystander effect,”43 which was the reason why GJ could regulate kidney damage. We firstly demonstrated that kidney damage could be protected by inhibiting ROS transfer mediated by Cx32 channels. This point of view has never been reported before.
Proper anesthetics choice used in perioperative or preoperative period benefitted patients who received organ transplantation, because they contribute to organ protection through different signal pathways.44 Propofol, as a commonly used anesthetic in clinic, had already been explored for many years and considered to confer organ protection through regulating different mechanisms, such as antioxidant, antiinflammatory, inhibiting apoptosis, and calcium influx, but the in depth mechanism governing the above “mechanistic effects” is still largely unclear.45
According to the reports, protective effects of propofol were always considered to be relative with decreasing ROS level. However, we found that at low-density cell culture (no contact with each other and no GJ formed), ROS level was not affected by propofol, but propofol decreased ROS significantly at high-density cell culture (contact with each other and GJ formed). The fact that propofol regulating ROS through GJ may play a more important role against H/R damage where GJ function may be enhanced, which is totally distinct from classic signal pathways. We suppose that propofol not only affected ROS distribution through GJ directly, but also inhibited some signal transmission resulting in enhanced ROS production. As a commonly used anesthetic, propofol is not only just limited in operation room, but also is used postoperatively in intensive care unit,46 especially in patients with severe conditions like those after organ transplantation. As reported, propofol preconditioning with various time intervals had protective effects against organ I/R injuries. Sufferers, who will undergo liver transplantation, are always with anxiety and insomnia. Propofol has special curative effect for this kind of patients.46 Based on these reports, we pretreated rats with propofol at its sedative doses for 3 days to observe its effects on GJ composed of Cx32.47 Our in vivo study results showed that pretreatment with propofol attenuated AOLT-induced AKI, which provided a new strategy for organ protection.
Another protective effect of propofol was relative with calcium increase. EC50 of propofol was 15 μM (measured as total concentration: protein bound and free).16,48 At this clinically relevant anesthesia concentration, propofol could increase calcium level in cells, which always affected cell survival.49 But, our findings suggest that changes of calcium were not the direct or major way for propofol to regulate cell growth. Because, no effects of propofol were observed at low-density cell culture (no GJ formed). A large number of reports documented that calcium contributed to the closure of GJ, even acting on the GJ gating directly.50 The results demonstrated that potential calcium increase induced by propofol, if any, may have affected cell survival in a GJ-dependent manner, but not directly.
Liver transplantation was a serious attack for all the systems. During the operation, a large amount of harmful factors, such as proinflammatory cytokines and ROS and so on were released into the blood induced by I/R, which initiated remote organs injury, leading to a high mortality rate after the operation.51 In our studies, we investigated remote kidney damage after AOLT and clarified that AKI could be attenuated by inhibiting GJ function composed of Cx32, which was relative with ROS reduction. Propofol reduced postischemic or posthypoxic ROS production and oxidative damage mainly via GJ-dependent mechanism and protected against AOLT induced kidney damage. These findings might not be universal to all the remote organs, but nonetheless they may have translational applications in organ protection of liver transplantation.
In conclusion, we have conducted a series of in vitro and in vivo studies and demonstrated that Cx32 plays a critical role in AOLT-induced AKI and that inhibition of Cx32 GJ function may represent a new and major mechanism whereby propofol reduces oxidative stress and subsequently attenuates post-AOLT AKI.
This study is supported by the National Natural Science Foundation of China (Beijing, China; grant nos. 81170449 and 81401628); key project of Natural Science Foundation of Guangdong Province, China (Guangzhou, Guangdong Province, China; grant no. S2011020002780); and 985 project (Beijing, China; grant no. 82000-1188190).
The authors declare no competing interests.