To investigate the additional effects of cold ischemia on the expression profile found in DBD donor kidneys, T2 biopsies from the recipient cohort were compared with T1 biopsies from the donor cohort (Fig. 1). Unpaired analysis of DBD T2 (n = 110) biopsies, taken at the end of cold ischemia, compared to DBD T1 (n = 82) donor biopsies, revealed no significantly enriched pathways. Also, analysis of DCD T2 (n = 53) biopsies compared to DCD T1 (n = 38) biopsies did not reveal any significantly enriched pathways.
First, the impact of reperfusion on living donor kidneys was analyzed by paired analysis of T3 reperfusion biopsies (n = 34) with T1 donor biopsies (n = 34, Fig. 1). Paired analysis revealed enrichment of mitogen activated protein kinase (MAPK) and p53 signaling pathways (Table S5a, SDC,http://links.lww.com/TP/B84).
Subsequently, reperfusion of DBD donor kidneys was analyzed in two different cohorts. First, a paired analysis was performed on a group of paired T3-T2 biopsies. Second, an unpaired analysis was performed on T3-T2 biopsies of unpaired kidney biopsies. Using a cutoff of FC that is 1.1 or higher in expression level, significant enrichment of oxidative phosphorylation, ribosome pathways, MAPK signaling pathway, and glycosylphosphatidylinositol anchor biosynthesis was demonstrated and confirmed by GSEA. In addition, using FC that is 1.3 or higher, paired analysis revealed enrichment of nucleotide-binding oligomerization domain (NOD)-like receptor, MAPK, and cytokine-cytokine receptor pathways (Table S6a, SDC,http://links.lww.com/TP/B84). Analysis of the second set of unpaired biopsies confirmed NOD-like receptor and MAPK pathways and also found enrichment of complement-and-coagulation pathways (Table S6a, SDC,http://links.lww.com/TP/B84).
In kidneys from DCD donors, paired and unpaired analyses revealed significant overrepresentation of similar pathways compared to DBD in T3 versus T2 biopsies. Pathways of MAPK, NOD-like receptors, chemokine signaling, and antigen presentation were significantly enriched after reperfusion of a DCD graft (FC ≥ 1.3, Table S7a,SDC,http://links.lww.com/TP/B84). Gene set enrichment analysis confirmed the involvement of NOD-like receptor pathways in reperfusion of a DCD graft (Table S7b,SDC,http://links.lww.com/TP/B84). No additional pathways were revealed using a cutoff of FC that is 1.1 or higher.
Association of Transcriptional Profile With Delayed Graft Function
Delayed graft function is a common clinical outcome after transplantation of DCD donor kidneys. Global test analyses were performed to find association of KEGG gene sets of DBD or DCD grafts that experienced DGF. Already shortly after reperfusion of DCD grafts, multiple pathways related to ER and mitochondrial stress, RNA degradation and DNA repair were significantly associated with DGF (Table 1). No association of DGF with gene expression profiles of DCD grafts at earlier timepoints were found. Also, no association of DGF with DBD gene expression profile was found at all other timepoints. No association of pathways at all timepoints in DBD or DCD donor kidneys with primary nonfunction (PNF) was found.
Malondialdehyde Level in the Urine
To investigate the timing and extent of hypoxia in deceased donors, malondialdehyde (MDA) was measured in the urine (Fig. 4). Directly after the diagnosis of BD (T0) and before organ retrieval (T1), urinary MDA was significantly higher compared to living donors (P < 0.05). Time-dependent changes in MDA levels between the moment of BD (T0) and organ retrieval (T1) could not be demonstrated. In DCD donors, no significant difference could be demonstrated in urinary MDA levels at the moment of donor call (T0) or before withdrawal of treatment (T1) compared to living donors. No changes in MDA levels related to surgery in living donors were found (T0 vs. T1).
Enhancing longevity of renal allografts is a major challenge for clinicians in the field of renal transplantation. In the last decades, strategies to improve long-term renal allograft survival have been mainly directed to recipient-dependent mechanisms of renal allograft injury. In contrast, less efforts have been made to optimize donor organ quality. The present genome wide gene expression study reveals hypoxia and complement-and-coagulation cascades in the deceased organ donor as the major target for intervention to improve renal allograft outcome.
Hypoxia caused by cessation of renal blood flow is an inevitable consequence of kidney transplantation. Hypoxia is induced at the moment of organ retrieval until recovery of blood flow after reperfusion. Intriguingly, we now show that in DBD donors, renal hypoxia is already initiated before cessation of blood flow in the donor, comparable to DCD donor kidneys after cessation of blood flow. In both donor types, hypoxia was revealed by enrichment of glycolysis or gluconeogenesis pathways which showed significant induction of glycolytic gene expression. This is regulated via stabilization of HIF1A, which was induced in both donor types (Figure S2, SDC,http://links.lww.com/TP/B84).18,19 Additional proof of hypoxia was found by induction of mitochondrial antioxidants superoxide dismutase 2, glutathione peroxidase and upregulation and enrichment of genes related to the proteasome. The increased level of urinary MDA in DBD compared to living donors supports these genetic pathway analyses. Donor pretreatment to reduce hypoxia in the donor could be a promising strategy to improve renal allograft outcome. One such strategy could be the upregulation of antihypoxic cytoprotective genes hemeoxygenase 1 and heat shock protein-70.20 Pharmacological induction of hemeoxygenase 1 expression in experimental brain-dead donors has been shown to improve renal allograft survival.21 Another therapeutic option might be stabilization of HIF1A by prolyl hydroxylase domain inhibitors, recently extensively reviewed by Akhtar et al.22
Besides hypoxia, also complement-and-coagulation cascades were significantly enriched in deceased donor kidneys before organ retrieval. The transcriptional profile reveals a procoagulable state and activation of all complement pathways in both donor types. These results confirm our previous study in a small cohort of DBD donors in which we found significant gene induction and protein deposition of complement C3 and fibrinogen.23 Also, data from previous studies looking at complement single-gene expression in DBD kidneys after organ retrieval are in line with our results.12-16 Therefore, it seems most advantageous to target complement and coagulation in the donor to improve long-term allograft survival. Inhibition of complement in brain-dead rats has already been proven to be successful to improve renal function.24 The complement activation cascade can be targeted at different levels to attenuate complement mediated injury by using, for example, soluble complement regulator proteins, antibodies against complement components, their split products, or complement receptor antagonists. Only few of these agents have been used in patients for other indications than kidney transplantation. Recombinant or nanofiltered blood plasma–derived C1-esterase inhibitor is currently approved for routine prophylaxis of hereditary angioedema, and it blocks both the classical and the lectin pathway of complement activation, as well as proteases of the fibrinolytic, clotting, and kinin pathways.25 Because this study also shows enrichment of coagulation pathways in deceased donors, C1-esterase inhibitor is likely to be beneficial for graft outcome because of its pleiotropic functions. Another therapeutic option might be the use a monoclonal antibody directed against complement C5 or C5a receptor, thereby inhibiting C5a-mediated chemotaxis and formation of the membrane attack complex.26-28
The impact of cold ischemia on renal allograft outcome is a matter of debate. In DBD and DCD donor kidneys, no pathways were found to be significantly enriched between organ retrieval and cold ischemia. These results suggest that targeting pathways revealed by transcriptomic analysis should be initiated in the donor instead of during preservation.
In our analysis of reperfusion biopsies, an apparent difference in pathway enrichment was found between living and deceased donors kidneys. Although in living reperfusion biopsies only induction of p53 and MAPK was found, both DBD and DCD reperfusion biopsies showed enrichment of NOD-like receptor pathways. This suggests that tissue hypoxia leads to cell necrosis and the release of danger associated molecular patterns (Figure S3, SDC,http://links.lww.com/TP/B84). The DBD grafts showed significant enrichment of downregulated genes of ribosome and oxidative phosphorylation pathways. The findings in DBD biopsies are part of the cell survival program in which adaptation to hypoxia is initiated, but are not associated with DGF. In contrast, in DCD reperfusion biopsies, multiple hypoxia-related pathways were associated with DGF (Table 1). The difference in findings might be explained by ischemic preconditioning (IP) of the deceased donor organ. In IP, an organ or tissue is exposed to a fixed period of ischemia to induce ischemic tolerance and protect the organ or tissue against a subsequent episode of ischemia. In experimental and clinical ischemia-reperfusion injury, IP has generated conflicting results.29–31 It has been shown that the extent of IP is an important factor to reduce renal ischemia reperfusion injury.32,33 Several mediators are proposed among which superoxide dismutase and HIF1A, which were also induced in DBD and DCD donors in our study. Therefore, the genetic changes in deceased donors in our study might reflect IP. Hypothetically, one would expect a comparable degree of IP-mediated protection in DBD and DCD donors because the nature and extent of hypoxia-induced pathways was almost identical between both donor types. However, DBD kidneys are only partially protected, while DCD kidneys are almost not protected against DGF. Possibly, the T1 biopsy in DCD donors does not fully reflect the genetic changes induced by WIT1 because the WIT1 is too short for sufficient gene induction at the moment of biopsy (taken directly after WIT1). Therefore, the hypoxia-related genetic changes at T1 in DCD donor kidneys reflect the hemodynamic changes on the ICU. It is the subsequent WIT1, inherent to DCD donation, that finally leads to DGF. The hypoxia-related pathways induced by the WIT1 become apparent later in T3 reperfusion biopsies and are associated with DGF. Although DBD kidneys are partially protected against ischemia-reperfusion injury by IP in the donor, IP is overwhelmed by the massive ischemic insult associated with DCD transplantation, leading to irreversible graft injury. This theory is funded by the hallmark study of Murray et al. who found that preconditioning merely allows organs to withstand only marginally longer periods of ischemia compared to non-preconditioned organs.34
Major pitfalls of transcriptome analysis are potential biases due to analysis of a small cohort, tissue heterogeneity and lack of replication studies.11 Our study comprising a large cohort and sampling bias because of tissue heterogeneity is unlikely because paired (left and right) donor biopsies of the same donor showed identical gene pathway enrichment. Concerning replication, two separate cohorts of paired and unpaired transplants showed almost identical gene set enrichments. Therefore, we feel confident that the results of our analysis are representative measurements.
In conclusion, this large deceased donor study shows that hypoxia is already initiated in DBD donors before cessation of blood flow, before organ retrieval. The data suggest that the accepted dogma that hypoxic injury only occurs after ischemia-reperfusion has to be modified. Future intervention therapies should target hypoxia and complement-and-coagulation cascades in the donor to improve renal allograft outcome in the recipient.
MATERIALS AND METHODS
Patients, Kidney Biopsies, and Urine Samples
Between 2004 and 2009, kidney biopsy specimens were routinely obtained during organ recovery procedures from living, DBD, and DCD type III (awaiting cardiac arrest) donors. Transplant characteristics can be found in Table S1 (SDC,http://links.lww.com/TP/B84. All organ donors and recipients were white. Deceased donors who had stated their objection to participation in transplantation research in the Dutch Donor Registry, or whose kidneys were discarded after retrieval, were not included. All living donors were asked informed consent for taking biopsy specimens. Biopsies were taken using a 16-gauge needle (Acecut; TSK Laboratory, Japan), preserved in RNALater (Sigma, St. Louis, MO) and subsequently stored at −80°C until analysis. Biopsies were obtained from three different study cohorts: (1) deceased donor cohort, (2) recipient cohort, and (3) living cohort (Fig. 1).
In the deceased donor cohort, donor biopsies were collected from DBD and DCD donors whose kidneys were donated in the Northern part of the Netherlands, while their kidneys were transplanted elsewhere in the Eurotransplant region. In DBD donors, biopsies were taken before cessation of blood circulation, before organ retrieval (T1, n = 82). In DCD donors, biopsies were taken after cardiac arrest and the first warm ischemia time (WIT1), before organ retrieval (T1, n = 38). Of these transplants, no paired T2-T3 biopsies were available (Fig. 1).
In the recipient cohort, biopsies were collected from renal allografts that were transplanted only in the University Medical Center Groningen. Biopsies were taken at the end of cold ischemia before organ implantation (T2) and 45 to 60 min after reperfusion (T3). Of recipients receiving a DBD donor kidney, 67 paired T2-T3 and 43 (T2)/38 (T3) unpaired biopsies were available. Of recipients receiving a DCD donor kidney, 29 paired T2-T3 and 24 (T2)/35 (T3) unpaired biopsies were available. No paired T1 donor biopsies of these grafts were available (Fig. 1).
In the living cohort, donation and transplantation procedures were all performed in the UMCG. Donor biopsies were taken before donor nephrectomy before clamping the renal artery (T1, n = 37). In recipients, biopsies were taken 45 to 60 min after reperfusion (T3, n = 34). Paired T1-T3 biopsies were available for 34 transplants (Fig. 1).
In living and DBD donors, urine was collected at T1. In DCD donors, a T1 urine sample was collected before ventilator switch off. An additional paired baseline urine sample (T0) was collected in living donors before start of operation in DBD donors directly after diagnosis of BD and in DCD donors at the moment of organ call to Eurotransplant (Figure S1, SDC,http://links.lww.com/TP/B84).
Kidney biopsy specimens were mechanically disrupted before lysis and RNA was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. The quality and quantity of the RNA was determined by high throughput Caliper GX LabChip RNA kit (Caliper).
Amplification, Labeling, and Hybridization of RNA Samples
Ambion Illumina Total Prep RNA kit was used to transcribe 300 ng RNA to cRNA according to the manufacturer’s instructions. A total of 750 ng of cRNA was hybridized at 58°C for 16 hr to the Illumina HumanHT-12 v4 Expression BeadChips. BeadChips were scanned using Iscan Software (Illumina, San Diego, CA).
Microarray Data Analysis
The raw data files were processed using GenomeStudio Software and further analyzed using GeneSpring GX 12.0 software (Agilent, Santa Clara, CA). Data normalization was performed using the default GeneSpring GX 12.0 median shift normalization to the 75th percentile (applying a log2 transformation) and baseline transformation using the median of all samples. GeneSpring’s built-in statistical software was used to identify lists of differentially expressed genes. Data were filtered using FC of 1.1 or higher or 1.3 and a Benjamini-Hochberg False Discovery Rate less than 0.01.
For a comparison between the three different donor types at T1, one-way analysis of variance and Tukey post hoc tests were performed. For a comparison between the donor cohort (T1) and the recipient cohort (T2), unpaired t tests were performed. For transplants of which paired biopsies were available (T2 and T3), paired t test were performed. For transplants of which only one timepoint biopsy was available (T2 or T3), unpaired t tests were performed (Fig. 1).
The data discussed in this publication have been deposited in the Gene Expression Omnibus (NCBI) and are accessible online through GEO Series accession number GSE43974 at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE43974).
Pathway analyses were performed using two different approaches. First, lists of genes that were found to be differentially expressed between groups were loaded into the web-based, client or server application database for DAVID (http://apps1.niaid.nih.gov/david/). Second, the complete set of gene expression data was loaded into the GeneSpring built-in software for GSEA.35 Permutation testing was performed with 1,000 permutations per gene set. Both approaches were used to identify significantly enriched biological pathways based on the KEGG database (https://www.genome.ad.jp/kegg). Gene sets with a Benjamini Hochberg False Discovery Rate less than 0.05 were considered to be significantly enriched. GenMAPP/MAPPFinder was used to visualize pathways.36
Association of Transcriptional Profile With Clinical Outcome
Globaltest analysis was performed to find association of KEGG gene sets of DBD and DCD grafts with DGF and PNF. Delayed graft function was defined as the requirement for dialysis within the first week after transplantation. Primary nonfunction was defined as nonfunctioning of the allograft from transplantation on. The Globaltest investigates whether the genes in the geneset have a higher association with the response than expected by chance. The null hypothesis of the Globaltest assumes that all regression coefficients of all the genes in the geneset are zero and therefore have no predictive ability for DGF or PNF. It uses the full data set instead of differential expressed genes and provides a single P value for the gene set, instead of a P value for each gene.37
Level of MDA in Donor Urine
Malondialdehyde was estimated by measuring the amount of thiobarbituric acid-reactive substances in urine. Twenty microliters of urine sample was mixed with 90 μL of a 3% sodiumdodecylsulfate solution and 10 μL of 0.05 M butylated hydroxytoluene. After pH adjustment of 2 to 3, 200 μL (0.7%) thiobarbituric acid was added, and samples were heated to 95°C for 30 min. The reaction product was precipitated with butanol and, absorbance was read at 530 nm. Levels were normalized using urinary creatinine values. For statistical analysis, the Kruskall-Wallis test was performed, followed by a Mann-Whitney posttest. All statistical tests were two-tailed with P value less than 0.05 regarded as significant.
The authors thank Joseph Milano for reviewing the article.
1. Legendre C, Canaud G, Martinez F. Factors influencing long-term outcome after kidney transplantation. Transpl Int
2014; 27: 19.
2. Opelz G, Döhler B. Multicenter analysis of kidney preservation. Transplantation
2007; 83: 247.
3. Moers C, Smits JM, Maathuis MH, et al. Machine perfusion or cold storage in deceased-donor kidney transplantation. N Engl J Med
2009; 360: 7.
4. Nijboer WN, Moers C, Leuvenink HG, et al. How important is the duration of the brain death period for the outcome in kidney transplantation? Transpl Int
2011; 24: 14.
5. Terasaki PI, Cecka JM, Gjertson DW, et al. High survival rates of kidney transplants from spousal and living unrelated donors. N Engl J Med
1995; 333: 333.
6. Le Dinh H, Weekers L, Bonvoisin C, et al. Delayed graft function does not harm the future of donation-after-cardiac death in kidney transplantation. Transplant Proc
2012; 44: 2795.
7. Naesens M, Sarwal MM. Molecular diagnostics in transplantation. Nat Rev Nephrol
2010; 6: 614.
8. Schuurs TA, Morariu AM, Ottens PJ, et al. Time-dependent changes in donor brain death related processes. Am J Transplant
2006; 6: 2903.
9. Kusaka M1, Pratschke J, Wilhelm MJ, et al. Activation of inflammatory mediators in rat renal isografts by donor brain death. Transplantation
2000; 69: 405.
10. Hoeger S, Petrov K, Reisenbuechler A, et al. The additional detrimental effects of cold preservation on transplantation-associated injury in kidneys from living and brain-dead donor rats. Transplantation
2009; 87: 52.
11. Saat TC, Susa D, Roest HP, et al. A comparison of inflammatory, cytoprotective and injury gene expression profiles in kidneys from brain death and cardiac death donors. Transplantation
2014; 98: 15.
12. Naesens M, Li L, Ying L, et al. Expression of complement components differs between kidney allografts from living and deceased donors. J Am Soc Nephrol
2009; 20: 1839.
13. Kusaka M, Kuroyanagi Y, Mori T, et al. Global expression profiles in 1-hour biopsy specimens of human kidney transplantation from donors after cardiac death. Cell Transplant
2009; 18: 647.
14. Cravedi P, Maggiore U, Mannon RB. Low-density array PCR analysis of reperfusion biopsies: an adjunct to histological analysis. Nephrol Dial Transplant
2010; 25: 4077.
15. Hauser P, Schwarz C, Mitterbauer C, et al. Genome-wide gene-expression patterns of donor kidney biopsies distinguish primary allograft function. Lab Invest
2004; 84: 353.
16. Mueller TF, Reeve J, Jhangri GS, et al. The transcriptome of the implant biopsy identifies donor kidneys at increased risk of delayed graft function. Am J Transplant
2008; 8: 78.
17. Avihingsanon Y, Ma N, Pavlakis M, et al. On the intraoperative molecular status of renal allografts after vascular reperfusion and clinical outcomes. J Am Soc Nephrol
2005; 16: 1542.
18. Semenza GL. Hypoxia-inducible factors in physiology and medicine. Cell
2012; 148: 399.
19. Eltzschig HK, Carmeliet P. Hypoxia and inflammation. N Engl J Med
2011; 364: 656.
20. van Dullemen LF, Bos EM, Schuurs TA, et al. Brain death induces renal expression of heme oxygenase-1 and heat shock protein 70. J Transl Med
2013; 11: 22.
21. Kotsch K, Francuski M, Pascher A, et al. Improved long-term graft survival after HO-1 induction in brain-dead donors. Am J Transplant
2006; 6: 477.
22. Akhtar MZ, Sutherland AI, Huang H, et al. The role of hypoxia-inducible factors in organ donation and transplantation: the current perspective and future opportunities. Am J Transplant
2014; 14: 1481.
23. Damman J, Nijboer WN, Schuurs TA, et al. Local renal complement C3 induction by donor brain death is associated with reduced renal allograft function after transplantation. Nephrol Dial Transplant
2011; 26: 2345.
24. Damman J, Hoeger S, Boneschansker L, et al. Targeting complement activation in brain-dead donors improves renal function after transplantation. Transpl Immunol
2011; 24: 233.
25. Davis AE III, Mejia P, Lu F. Biological activities of C1 inhibitor. Mol Immunol
2008; 45: 4057.
26. Castellano G, Melchiorre R, Loverre A, et al. Therapeutic targeting of classical and lectin pathways of complement protects from ischemia reperfusion-induced renal damage. Am J Pathol
2010; 176: 1648.
27. de Vries B, Matthijsen RA, Wolfs TG, et al. Inhibition of complement factor C5 protects against renal ischemia-reperfusion injury: inhibition of late apoptosis and inflammation. Transplantation
2003; 75: 375.
28. Arumugam TV, Shiels IA, Strachan AJ, et al. A small molecule C5a receptor antagonist protects kidneys from ischemia/reperfusion injury in rats. Kidney Int
2003; 63: 134.
29. Huang Y, Shan J, Wang C, et al. Can ischemic preconditioning alone really protect organs from ischemia reperfusion injury in transplantation. Transpl Immunol
2009; 20: 127.
30. Torras J, Herrero-Fresneda I, Lloberas N, et al. Promising effects of ischemic preconditioning in renal transplantation. Kidney Int
2002; 61: 2218.
31. Chen Y, Zheng H, Wang X, et al. Remote ischemic preconditioning fails to improve early renal function of patients undergoing living-donor renal transplantation: a randomized controlled trial. Transplantation
2013; 95: e4.
32. Wever KE, et al. Remote ischaemic preconditioning by brief hind limb ischaemia protects against renal ischaemia-reperfusion injury: the role of adenosine. Nephrol Dial Transplant
2011; 26: 3108.
33. Lazaris AM, Maheras AN, Vasdekis SN, et al. Protective effect of remote ischemic preconditioning in renal ischemia/reperfusion injury, in a model of thoracoabdominal aorta approach. J Surg Res
2009; 154: 267.
34. Murray E, Jenning RB, Reimer KA. Preconditioning with ischemia: a delay in lethal cell injury in ischemic myocardium. Circulation
1986; 74: 1123.
35. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A
2005; 102: 15545.
36. Doniger SW, Salomonis N, Dahlquist KD, et al. MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data. Genome Biol
2003; 4: R7.
37. Goeman JJ, van de Geer SA, de Kort F, et al. A global test for groups of genes: testing association with a clinical outcome. Bioinformatics
2004; 20: 93.
Supplemental Digital Content
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.