A Comparison of Inflammatory, Cytoprotective and Injury Gene Expression Profiles in Kidneys From Brain Death and Cardiac Death Donors

Abstract

Background 

The superior long-term survival of kidneys from living donors (LDs) compared with kidneys from donation-after-brain-death (DBD) and donation-after-cardiac-death (DCD) donors is now well established. However, comparative studies on transcriptional changes that occur at organ retrieval and during and after cold ischemia (CI) are sparse.

Methods 

Using a rat model, we used qRT-PCR to examine expression levels of inflammatory, cytoprotective, and injury genes at different time points after organ retrieval. Cleaved caspase-3 was used to evaluate early apoptosis in DCD and DBD kidneys.

Results 

Immediately after retrieval, we found massive up-regulation of proinflammatory genes interleukin-1β, interleukin-6, tumor necrosis factor-α, monocyte chemotactic protein-1, P-selectin, and E-selectin in DBD compared with LD and DCD kidneys. A significant increase in the expression of injury markers Kim-1, p21, and the cytoprotective gene heme oxygenase-1 accompanied this. Bax was increased in DCD kidneys, and Bcl-2 was decreased in DBD kidneys. After 2 hr of CI in the LD group and 18 hr in the DBD and DCD groups, gene expression levels were similar to those found after retrieval. During 18 hr of cold storage, expression levels of these genes did not change. In DCD and DBD kidneys, early apoptosis increased after CI.

Discussion/Conclusion 

The gene expression profile in DBD kidneys represents an inflammatory and injury response to brain death. In contrast, DCD kidneys show only mild up-regulation of inflammatory and injury genes. These results may imply why delayed graft function in DCD kidneys does not have the deleterious effect it has on DBD kidneys.

The superior long-term survival of kidney transplants from living donors (LDs) compared with those from donation-after-brain-death (DBD) or donation-after-cardiac-death (DCD) donors is now well established (^1–3). Living donors are healthy during organ procurement, and retrieved kidneys are exposed to minimal warm and cold ischemia (CI) times. Brain-dead donors experience hemodynamic instability, hormone dysregulation, and immunologic reactivity, which contribute to organ damage and inferior posttransplantation function and graft survival compared with LDs (^4–8). Organs from DCD donors do not develop the physiologic changes of DBD donors but are exposed to prolonged warm ischemia times (WITs), which increase the incidence of delayed graft function (DGF). In addition, organs from both postmortem groups are exposed to prolonged CI compared with the organs from LDs (^{9, 10}). Although there is no difference in the incidence or severity of acute rejection between DBD and DCD kidneys, DCD kidneys have a 73% chance of DGF compared with 27% in DBD kidneys (¹¹). However, the long-term graft survival of renal transplants from DCD donors is similar to those from DBD donors (^11–15). Improved technology and an increasing gap between supply and demand for donor organs have prompted an increased use of organs from both LDs and DCD donors (^{4, 16}).

The pathophysiology that leads to organ damage after brain death is relatively well established (^{5, 17–19}), and strategies to protect the organs of a DBD donor are emerging (^20–23). In contrast, experimental studies on the changes that occur in organs from DCD donors are sparse (²⁴). A drawback in clinical studies that compare perioperative gene expression patterns during transplantation is that they are confined to a relatively small time span and a limited number of time points. Usually “t=0 biopsies” are obtained at the end of the cold storage period, followed by a postreperfusion biopsy 30 to 60 min after reperfusion of the graft (¹⁹). These time points are not able to take into account the major pathophysiological perturbances that occur during the periods of brain death, cardiac arrest, organ retrieval, and the subsequent cold storage period. Experimental studies have tried to overcome these limitations by studying gene expression at explantation but have mainly focused on DBD donors (^{25, 26}).

The aim of this study therefore was to compare expression levels of genes indicative of inflammation (interleukin [IL]-6, IL-1α, tumor necrosis factor [TNF]-α, monocyte chemotactic protein [MCP]-1, E-selectin, P-selectin, toll-like receptor [TLR]4, high-mobility group box [HMGB]1), cytoprotection (heme oxygenase [HO]-1), and injury (hypoxia-inducible factor [HIF]-1α, vascular endothelial growth factor [VEGF], Bax, Bcl-2, p21, and Kim-1) at the time of kidney retrieval, after clinically relevant CI times and over a time course during cold storage.

RESULTS

Expression of Proinflammatory, Cytoprotective, and Injury Genes After Retrieval

Directly after retrieval, proinflammatory genes showed a significant increase in DBD kidneys compared with LD kidneys (Fig. 1). Intrarenal IL-1β messenger RNA (mRNA) levels were increased in DBD kidneys (13.7-fold, P=0.032) versus LD kidneys. Interleukin-6 expression was massively increased in the DBD kidneys compared with the LD kidneys (417-fold, P=0.003). Tumor necrosis factor-α and MCP-1 (not shown) were increased (4.7-fold, P=0.004; and 81-fold, P=0.015, respectively). In DCD kidneys, MCP-1, IL-1β, and IL-6 were not significantly different compared with LD kidneys, whereas TNF-α was slightly but significantly lower (P=0.003). Expression of the adhesion molecules P-selectin and E-selectin was highly elevated in DBD kidneys, but not in DCD kidneys (not shown).

FIGURE 1: mRNA expression levels of genes indicative of inflammation, cytoprotection, and kidney injury were determined by qRT-PCR, directly after retrieval of the kidneys in LDs and in DCD and DBD donors. To compare the individual expression levels of each animal, ΔCt values were normalized to the average ΔCt of the LD group at time point 0. Subsequently, the fold increase was calculated using 2^ΔΔCt. Results are expressed as mean±SEM of seven animals per group. *P<0.05, **P<0.01 compared with LD. mRNA, messenger RNA; LDs, living donors; DBD, donation after brain death; DCD, donation after cardiac death.

Cytoprotective gene HO-1 was highly up-regulated in DBD kidneys compared with LD kidneys (36.7-fold, P=0.004), whereas no difference was found between the DCD group and the LD group. Neither HIF-1α nor VEGF expression was significantly influenced by DBD or DCD compared with LD. The antiapoptotic gene Bax (not shown) and proapoptotic Bcl-2 (not shown) showed no significant differences between the groups. Cell cycle inhibitor p21 was significantly elevated in both the DBD (5.0-fold, P=0.032) and DCD kidneys (1.6-fold, P=0.007). These changes were accompanied by a significant up-regulation of the kidney injury marker Kim-1 in the DBD group (48-fold, P=0.021).

Gene Expression Levels After Clinically Relevant CI Times

Next, we investigated the expression levels after CI relevant to human kidney transplantation. Living-donor kidneys were analyzed after 2 hr of CI, and the DBD and DCD groups were analyzed after 18 hr of CI (^{27, 28}).

Compared with LD, IL-1β, IL-6, TNF-α, and MCP-1 were increased in the DBD kidneys (43-fold, P=0.003; 997-fold, P=0.003; 9.7-fold, P=0.003; 115-fold, P=0.004). Interleukin-1β and IL-6 were increased in the DCD kidneys but to a lesser extent (2.3-fold, P=0.004; 10-fold, P=0.003). Tumor necrosis factor-α and MCP-1 showed no significant difference in expression between the DCD and the LD kidneys. P-selectin was significantly increased in the DBD kidneys (129-fold, P=0.003) and DCD kidneys (2.9-fold, P=0.008). E-selectin was increased in the DBD kidneys (64-fold, P=0.003). Toll-like receptor 4 expression was elevated in the DBD kidneys (4.4-fold, P=0.037) but not in DCD kidneys. High-mobility group box 1 was not significantly different.

After DBD, HO-1 was up-regulated (21-fold, P=0.001), whereas in the DCD donors, HO-1 was slightly but significantly lower compared with the LDs (P=0.014). Neither HIF-1α nor VEGF expression was significantly influenced by DBD or DCD. Bax was expressed higher in the DCD kidneys compared with the LD kidneys (1.3-fold, P=0.016). Bcl-2 was lower in the DBD kidneys compared with the LD kidneys (0.7-fold, P=0.032). Expression of p21 was higher in both DBD and DCD donors (9.1-fold, P=0.003; and 2.5-fold, P=0.003) as compared with LDs. In DBD donors, Kim-1 was increased (319-fold, P=0.003), whereas in DCD donors, it was slightly increased (5.3-fold, P=0.010).

Gene Expression Profiles During an 18-Hr Cold Storage Period

To address whether changes in gene expression would occur during cold storage of the kidneys, gene expression of HO-1, HIF-1α, Bax, Bcl-2, TLR4, HMGB1, and VEGF (not shown) was examined after 0, 2, 4, 6, 12, and 18 hr of cold storage (Fig. 2). Gene expression levels did not significantly change after retrieval and during cold storage of the kidney grafts. These data show that biologic processes in the donor are the main determinants of gene expression in the graft before transplantation.

FIGURE 2: Gene expression profiles in kidneys remain stable during cold preservation. Messenger RNA expression levels of HO-1, HIF-1α, Bax, Bcl-2, TLR4, and HMGB1 were determined directly after retrieval of kidneys from the donor and at 2, 4, 6, 12, and 18 hr of cold preservation in kidneys of LDs and of DCD and DBD donors. To compare the individual expression levels of each animal, ΔCt values were normalized to the average ΔCt of the LD group at the respective time points. Subsequently the fold increase was calculated using 2^ΔΔCt. Results are expressed as mean±SEM of seven animals per group. HO-1, heme oxygenase-1; HIF-1α, hypoxia-inducible factor-1α; TLR4, toll-like receptor 4; HMGB1, high-mobility group box 1; LDs, living donors; DBD, donation after brain death; DCD, donation after cardiac death.

Apoptosis During Cold Storage

The DCD and DBD kidneys collected directly after retrieval and after 18 hr of CI were stained for cleaved caspase-3 (Fig. 3). In DCD kidneys, the average number of apoptotic cells increased after 18 hr of CI compared with kidneys directly after retrieval (P<0.01). The DBD kidneys showed a similar trend.

FIGURE 3: Early apoptosis was analyzed using a cleaved caspase-3 staining in DCD and DBD kidneys after retrieval and after 18 hr of cold ischemia. Directly after retrieval, DCD kidneys showed less apoptotic cells compared with DCD kidneys after 18 hr of cold preservation. The DBD kidneys show a similar trend. *P<0.01 compared with time point 0. DBD,donation after brain death; DCD, donation after cardiac death.

DISCUSSION

To better understand the relative contributions of brain death, prolonged WITs, and CI in DBD and DCD donors, we studied expression profiles of genes representative of inflammation, cytoprotection, and injury.

Directly after kidney retrieval, proinflammatory genes showed a massive up-regulation in the DBD kidneys compared with LD and DCD kidneys. A significant increase in the expression of Kim-1, p21, and cytoprotective gene HO-1 accompanied this. After clinically relevant CI times, the gene expression profiles remained stable.

The high levels of proinflammatory and injury genes in DBD donors are in contrast with the clinically high incidence of DGF in DCD donors compared with DBD donors. The prolonged WIT experienced by DCD donors (²⁹) was not reflected by changes in gene expression in our study. Instead, DCD kidneys showed increased apoptosis.

In the clinical setting, the average WIT in DCD donors is approximately 20 min (³⁰). Although the prolonged WIT in organs from DCD donors is considered the main reason for the increased incidence of DGF in these grafts, the deterioration in organ quality induced by 20 min of WIT in this study is not reflected by changes in gene expression.

During cold preservation, hypothermia and the lack of oxygen preclude the transcription of DNA into mRNA. The use of preservation fluids preserves cellular integrity and may prevent breakdown of RNA, and this may explain the stable mRNA levels during cold preservation (^31–34).

The most striking finding in our study is the difference between DBD and DCD donors in renal transcription of inflammatory markers, such as IL-1β, IL-6, MCP-1, and E-selectin. High levels of inflammatory cytokines in the donor are associated with decreased graft survival (^{26, 35, 36}). The high levels of IL-6 in the serum of DBD donors are believed to be caused by the production in the injured brain and subsequent leakage through the disrupted blood-brain barrier (^{17, 37}). We show that it is likely that the kidney itself significantly contributes to systemic IL-6 levels and systemic inflammation observed during brain death. This inflammation resulted in kidney injury, as shown by the up-regulation of Kim-1. In DCD kidneys, expression of IL-1β, IL-6, and P-selectin was significantly up-regulated compared with LD but to a much lesser extent than that in DBD donors.

Toll-like receptor 4 expression was significantly up-regulated in DBD donors. The important role of TLR4 in mediating renal ischemia-reperfusion injury has been shown both in animal models and after human transplantation (^{38, 39}). High-mobility group box 1 has been identified as ligand for TLR4 present on tubular epithelial cells and is involved in the cellular signaling through TLR4, which leads to the production of proinflammatory cytokines (³⁸). High-mobility group box 1 secretion is stimulated by IL-1β and by stressed cells or cells undergoing necrosis (^{40, 41}). Up-regulation of both TLR4 and HMGB1 has been observed after reperfusion, and targeting both may diminish ischemic injury (³⁹). Although we found that the up-regulation of many cytokines can be induced by TLR4 activation, such as TNF-α and MCP-1, we found low levels of HMGB1, which might suggest that it is not involved in the systemic inflammatory response observed during DBD in our model or that low levels of HMGB1 are able to activate TLR4. In addition, TLR4 can also be activated by other alarmins (⁴⁰).

Compared with the severe inflammatory response induced during 6 hr of brain death, 20 min of WIT in our DCD model induced only mild changes, which are almost indistinguishable from the transcriptional changes found after LD. Because the duration of the asystolic period in DCD donors determines the severity of the damage to the kidney (²⁴), it is likely that 20 min of asystole used in our study is not sufficient to induce a massive up-regulation of proinflammatory gene expression. In addition, human DCD donors may experience other injuries, which influence the inflammatory response.

Because of the higher incidence of DGF in DCD donors (^{9, 10, 42}), it has even been suggested that DGF in DCD kidneys predisposes for favorable graft outcome (⁴³). The absence of overt inflammatory changes and significantly lower Kim-1 expression in our study may explain why the impact of DGF is reduced in DCD kidneys.

We found a significant up-regulation of HO-1 in DBD donors at the time of retrieval, which remained stable during 18 hr of CI. Several studies describe that the overexpression of HO-1 protects kidney transplants against the deleterious consequences of ischemic injury (^44–46). In LD, elevated HO-1 expression in the transplant demonstrated a protective effect on renal ischemia-reperfusion injury (^{25, 47}). In contrast, the increased levels of HO-1 in DBD kidneys did not prevent renal damage. This suggests that DBD causes renal damage to such an extent that HO-1 expression is insufficient and may be viewed as a marker of injury (²⁵).

Both HIF-1α and VEGF itself were not significantly up-regulated in DBD donors. This implies that during 6 hr of brain death, the inflammatory response coincides with the up-regulation of HO-1, not VEGF. Interestingly, p21 was also strongly induced in both postmortem donors. p21, which binds stoichiometrically to PCNA and inhibits its action in DNA replication, is a key regulator of proliferation after renal ischemic injury (⁴⁸). We previously showed that there is a critical and temporal balance between the expression levels of p21 and proliferation in the regenerating tubular compartment after ischemia (⁴⁹). The up-regulation of p21, like HO-1, in our model may be envisioned as an indicator of damage, and the lower induction in DCD kidneys may give an advantage after transplantation, when proliferation of the damaged tubular compartment is warranted (^{48, 49}).

The proapoptotic gene Bax was slightly, but significantly, elevated in DCD versus LD. Expression of the antiapoptotic gene Bcl-2 was decreased in DBD compared with LD. These results suggest that induction of apoptosis may be caused by Bax induction during cardiac death, whereas decrease of Bcl-2 may be responsible in DBD donors.

We showed that in both postmortem donors, the number of apoptotic cells increases between retrieval and 18 hr of cold storage. Despite the stable gene expression levels, there are cellular changes during cold storage, resulting in a higher rate of apoptosis, which may account for the deteriorating graft quality during prolonged CI.

Our data strongly support the pretreatment of the DBD donor as a promising strategy to improve graft function and survival. Indeed, low-dose dopamine pretreatment of the donor before kidney transplantation results in a better postoperative kidney function and reduces the incidence of DGF (²¹). To reduce the effect of inflammatory cytokines, pretreatment with methylprednisolone has been used in deceased donors, which resulted in a better outcome after human liver transplantation (²⁰). In addition to this, induction of supraphysiologic levels of the cytoprotective gene HO-1 in DBD donors has a beneficial effect on graft function after transplantation (⁴⁴).

Our study is limited by the use of healthy, young rats without comorbidity. In this study, we focused on the influence of CI in different donor types. In a future study, the use of a transplantation model is required to explore the effect on reperfusion injury. Extensive histology could give insight at a cellular level.

Although rodent models of DBD, DCD, and LD are unable to capture the full complexity of human donors, they are a tool to dissect the relative contributions of brain death and ischemia to pretransplantation organ damage. Here, we show a massive up-regulation of inflammatory, injury, and cytoprotective genes in DBD kidneys, but not in LD or DCD kidneys at the time of graft retrieval, which remains stable during preservation. The difference between DBD and DCD donors in inflammatory gene expression may underlie the better outcome after DGF in DCD kidneys (⁵⁰) and lend further support to the antiinflammatory treatment of DBD donors (^{20, 21}).

MATERIALS AND METHODS

Experimental Design

The experimental protocol was approved by the Animal Experiments Committee under the Dutch National Experiments on Animals Act and complied with the 1986 directive 86/609/EC of the Council of Europe. Male Brown Norway rats of 12 to 14 weeks old, weighing 250 to 300 g, were purchased from Harlan-CPB (the Netherlands). Rats were randomly assigned to an LD, DBD, or DCD group (n=7 per group). The LD group served as a control for both postmortem groups.

After kidney retrieval, samples were stored in a University of Wisconsin solution; time point 0 samples were directly snap-frozen in nitrogen. At 2, 4, 6, 12, and 18 hr, the kidneys were collected and snap-frozen until further use. Gene expression levels after retrieval were measured in time point 0 samples. To measure expression levels after clinically relevant CI, 2 hr of CI was used for LD and 18 hr of CI for the postmortem donors. Messenger RNA expression levels were examined by qRT-PCR. For immunohistochemistry, DCD and DBD kidneys were used directly after retrieval and after 18 hr of CI.

Experimental Models

Animals were anesthetized with isoflurane. After intubation, anesthesia was maintained using a mixture of N₂O-O₂-2% isoflurane. Animals were pressure control ventilated on a Siemens Servo 900C ventilator with 14 cm H₂O peak inspiratory pressure, 4 cm H₂O positive end expiratory pressure with a frequency of 40 breaths per minute.

In the DBD group, a frontolateral trepanation was made, and a balloon catheter (Fogarty Arterial Embolectomy Catheter: 5F; BaxterHealthcare Co., Edwards Life Sciences, Irvine, CA) was introduced in the extradural space and slowly inflated, causing a gradually increasing intracranial pressure (⁵¹). Herniation of the brainstem and DBD was confirmed by dilated and fixed pupils, the absence of corneal reflexes, and an apnea test. Brain-dead animals received ventilation but no further anesthesia. Intra-arterial pressure was continuously monitored by means of a PE50 catheter placed in the carotid artery. Only rats with stable mean arterial pressure (>80 mm Hg) during 6 hr of brain death were included in the experiment. After 6 hr, both kidneys were removed, flushed with phosphate-buffered saline, and stored in a University of Wisconsin solution at 4°C.

In the DCD group, the animals were catheterized using a PE50 catheter placed in the carotid artery, and both kidneys were exposed by means of a midline incision that was then temporarily closed with a few sutures until retrieval of the kidneys. Cardiac arrest was induced by isoflurane overdose and lasted for 20 min, starting when the blood pressure had dropped 5 to 7 mm Hg, the time commonly needed before starting with cold storage of DCD kidneys. The DCD donors are type III, controlled; waiting for cardiac arrest, of the Maastricht criteria (⁵²). In the LD group, nephrectomy was performed after 1 hr of mechanical ventilation.

mRNA Expression Analysis

Kidney sections were homogenized in 1.0-mL Trizol (Invitrogen; Life Technologies, Bleiswijk, Netherlands). Total RNA was isolated according to the manufacturer’s guidelines, precipitated, and subsequently dissolved in diethylpyrocarbonate-treated water. RNA concentration was determined using a Nanodrop ND-1000 ultraviolet-visible spectrophotometer. For complementary DNA synthesis, M-MLV Reverse Transcriptase (Invitrogen) was used according to the manufacturer’s protocol with 250 ng of random hexamers (Promega, Madison, WI) and RNaseOUT (Invitrogen). All temperature-dependent steps were performed using a Biometra T-gradient cycler (Biometra GmbH, Goettingen, Germany). Complementary DNA samples were stored at −20°C. Quantitative real-time PCR was performed using a BioRad Icycler (Hercules, CA) with SYBR Green Incorporation. Each sample was tested in duplo, two times.

ΔCt values of genes of interest were calculated as described by Pfaffl et al. (⁵³) using β-actin as a reference gene. ΔCt values were normalized to the average ΔCt of the LD. Living-donor kidneys 2 hr after CI were used in comparison to the postmortem kidneys after 18 hr of CI and the respective time points in the time course. The fold increase was calculated using the Pfaffl equation, 2^ΔΔCt. Results are expressed as mean±SEM.

Primers

Primers indicative for inflammation are TLR4, HMGB1, IL-1β, IL-6, TNF-α, MCP-1, P-selectin, and E-selectin. Primers indicative for cytoprotection are HO-1, VEGF, and HIF-1α. Kim-1, p21, Bax, and Bcl-2 were used as markers indicative for kidney injury. Used primers are shown in Table S1 (SDC,https://links.lww.com/TP/A975).

Immunohistochemistry

Frozen sections, 5-µm thick, were stained with a polyclonal antibody against cleaved caspase-3 (Asp175; Cell Signaling, Danvers, MA), an early apoptotic marker. Cleaved caspase-3 was diluted 1:300 and visualized with a horseradish peroxidase–conjugated secondary antibody (dilution 1:500, goat-anti-rabbit IgG/HRP; DAKO, Denmark). Slides were scored at a magnification of 200× by two independent observers blinded to the treatment. Results are expressed as mean±SEM.

Statistical Analysis

Data were analyzed using a Kruskal-Wallis analysis of variance, followed by a Mann-Whitney U test. Analysis was performed using SPSS version 20.0 (IBM, Armonk, NY). A P≤0.05 was considered statistically significant.