Microarray gene expression profiling of chronic allograft nephropathy in the rat kidney transplant model

Transplant Immunology 27 (2012) 75–82

Contents lists available at SciVerse ScienceDirect

Transplant Immunology journal homepage: www.elsevier.com/locate/trim

Microarray gene expression profiling of chronic allograft nephropathy in the rat kidney transplant model Laurie Erickson a, 1, Carmen Wynn a, 1, Fan Pan a, Gladys Crews a, Guliang Xia a, Toshiko Yamada a, Xiaoyan Xu a, Yizheng Tu a, Di Huang b, Yi Song b, Kouichi Tamura a, b, Hongsi Jiang a, b,⁎ a b

Department of Transplant Immunology, Astellas Research Institute of America, 8045 Lamon Ave, Skokie, IL 60077, USA Chemical and Biological Engineering, Northwestern University, 2415 Sheridan Rd., Tech E-136, Evanston, IL 60208, USA

a r t i c l e

i n f o

Article history: Received 23 January 2012 Received in revised form 12 June 2012 Accepted 13 June 2012 Keywords: Chronic allograft nephropathy Kidney transplantation Immunosuppression FK506 Microarray Gene expression profiling

a b s t r a c t Whole genome gene expression profiles were correlated with renal function and histology in a well-established animal model of chronic allograft nephropathy (CAN). Kidneys of F344 rats were transplanted into LEW recipients treated with a brief dose of FK506 (BFK). Blood and urine samples were collected weekly. Kidney grafts were harvested at an early (day 6) or late (days 30–90) phase after transplantation. BFK kidney grafts showed remarkable changes in function, histology, and gene expression profiles when compared to the isograft controls. In the early phase, renal function and histology were barely affected, yet the expression levels of 225 genes were significantly changed, reflecting both immune and non-immune pathways. In the late phase, however, 826 genes were affected in the BFK kidney grafts, including genes in the pathways of extracellular matrix and cell adhesion. Of these genes, 214 appear to be key factors for development of CAN, since they were affected at both early and late phases, including genes involved in the immune response, the inflammatory response, apoptosis, and metabolism. Kinetic studies with gene expression profiling can identify genes involved in the progressive development of chronic allograft rejection, leading to more detailed therapeutic approaches or useful biomarkers in clinical transplantation. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Despite improvement in short-term renal graft survival rates, the fate of transplanted organs in the long term has not improved significantly over the last two decades. One of the leading causes of late renal graft loss is chronic allograft nephropathy (CAN)2 which is a complex cellular and molecular process driven by various forms of alloantigen-dependent and -independent injury after transplantation [1,2]. CAN is characterized by deterioration of renal function with histological findings of glomerulosclerosis, tubular atrophy, interstitial fibrosis, vasculopathy, intimal proliferation, and transplant glomerulopathy

⁎ Corresponding author at: Astellas Research Institute of America, 8045 Lamon Ave., Skokie, IL 60077, USA. Tel.: +1 847 933 7403. E-mail addresses: [email protected] (L. Erickson), [email protected] (C. Wynn), [email protected] (F. Pan), [email protected] (G. Crews), [email protected] (G. Xia), [email protected] (T. Yamada), [email protected] (X. Xu), [email protected] (Y. Tu), [email protected] (D. Huang), [email protected] (Y. Song), [email protected] (K. Tamura), [email protected], [email protected] (H. Jiang). 1 These two authors are equal contributors. 2 CAN, chronic allograft nephropathy. BFK, brief FK506 treatment. ECM, extracellular matrix. IFTA, interstitial fibrosis and tubular atrophy. PAS, Per-Iodic Acid Schiff stain. GO, Gene Ontology. CNI, calcineurin inhibitors. 0966-3274/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.trim.2012.06.007

[3]. As there is still considerable controversy regarding the mechanisms underlying development of CAN in transplantation, understanding of the key cellular and molecule profiles associated with CAN might be able to provide critical insights into the origin of this disease and eventually to therapeutic control of CAN in clinical transplantation [4,5]. Although there is a scarcity of rodent CAN models with a comprehensive manifestation of human CAN, the Fisher 344 (F344) to Lewis (LEW) rat kidney transplant model with a brief dose of calcineurin inhibitor can replicate many features of CAN in humans, including functional and histological changes [6]. Recent evidence of graft kidney injury associated with extensive nephron loss after transplantation in the F344 to LEW rat model is quite consistent with the progressive development of human CAN [7]. In addition, our previous data of elevated alloantibody titers in plasma and increased C4d positive staining in the kidney grafts of this model, which are considered important clinical features of human antibody-mediated CAN, have further supported the F344 to LEW rat kidney transplant model as a valuable model for studying the cellular and molecular profiles of CAN in the transplant setting [8]. Gene expression profiling has already provided new insights into immune and non-immune-related conditions, resulting from studies of lymphocyte activation, acute rejection with immunosuppression in rats and mice, and aging in the human kidney [9–18]. The primary goal of this study is to dissect the complex system of CAN in the well-characterized F344 to LEW rat kidney transplant model

76

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

by using microarray-based gene expression profiling. We anticipate that a detailed analysis of the gene expression profiles of CAN in this model may allow us to investigate the complex phenomena of the CAN with an exploratory approach and to provide a more complete picture of the disease at the molecular level. 2. Objectives To understand the molecular basis of chronic allograft nephropathy, we determined the gene expression profile of the progressive development of chronic rejection in a rat model, and compared them to the histology and functional features of CAN in these rats. 3. Materials and methods 3.1. Animals and kidney transplantation Naive inbred male Lewis (LEW; RT1I) rats and Fisher (F344; RT1IV1) rats, weighing 250 to 350 g, were used as recipients and kidney donors, respectively. All rats were obtained from Harlan Sprague–Dawley (Indianapolis, IN) and kept under specific pathogen-free conditions at the NorthShore University HealthSystem animal facility. Kidney transplants were performed using the modified technique described by Fisher and Lee [19]. Briefly, the kidney was transplanted with end-to-side anastomoses of the renal artery and vein with a patch of the aorta and inferior vena cava, respectively. The donor and recipient ureters were anastomosed end-to-end. No stent was used, and the graft ischemic time was approximately 45min. All recipients were bilaterally nephrectomized at the time of transplantation. All experimental procedures were approved by the Institutional Animal Care and Use Committee of NorthShore University HealthSystem. 3.2. CAN protocol and experimental groups The CAN model is defined as donor kidneys from F344 rats transplanted into Lew recipient rats treated with a brief dose of FK506 (1 mg/kg/day for 10 days) after transplantation [6]. The experimental groups were isograft (LEW to LEW), allograft (F344-LEW) and BFK (F344 to LEW with brief FK506 treatment). Grafts were harvested at day 6, day 30 or day 90. For histology and gene expression analysis, the day 30 and day 90 timepoints are combined into the late phase of chronic rejection. 3.3. Sample harvesting and functional tests Blood and 24-h urine samples were collected once a week after transplantation. Plasma creatinine and urine protein were measured by the Creatinine Kit (Sigma, St. Louis, MO) and the Bio-Rad Protein Assay Kit (Bio-Rad, Hercules, CA), respectively. The kidney grafts were harvested on day 6, 30 or 90 after transplantation. Kidney grafts were horizontally dissected into three sections for histology, immunohistology and RNA isolation. The RNA section was immediately frozen on dry ice and subsequently stored at −80°C until processing. 3.4. Histology The histology kidney graft sections were immediately fixed in 10% neutral buffered formalin or embedded in tissue tek optimum cutting temperature compound, snap-frozen in supercooled isopentane, and stored at −80°C. Formalin-fixed and paraffin-embedded grafts were sectioned at 2 to 3 μm. Two micrometer sections from each graft were stained with hematoxylin and eosin, Verhoeff's and Masson's combined elastic trichrome, and Per-Iodic Acid Schiff (PAS) stains. All histology protocols were obtained from the Theory and Practice of Histologic Technique, or the Laboratory Methods in Histotechnology manual

from the Armed Forces Institute of Pathology. Tissue sections were coded and examined in a semi-blinded manner. Renal histologic changes were semiquantitatively scored on a scale (0–3+) for inflammation, tubulitis, vasculitis, mesangiolysis, glomerulitis, tubular atrophy, glomerulosclerosis, fibrointimal hyperplasia, and transplant glomerulopathy based on criteria adapted from the Banff classification for transplant rejection [20], and the sum score was calculated for each sample. Data are expressed as average ± SD.

3.5. RNA isolation, array hybridization and qRT-PCR Total RNA was isolated from one part of the kidney sample by the Trizol method and treated with DNAse I (Invitrogen). RNA quality was evaluated on the 2100 Expert Bioanalyzer (Agilent Technologies). Microarray analysis was performed only on high-quality RNA samples with RIN (RNA Integrity Number) numbers of 9 or higher (three samples each from day 6 Iso, Allo and BFK groups, and from day 30 Iso, Allo groups, and two samples each from the day 30 BFK group and day 90 Iso, Allo and BFK groups). Microarray gene expression profiling was performed using the Agilent platform of whole genome rat chip with 41,013 probes (Agilent Technologies). First-pass assessment was done by Agilent, resulting in 11,121 probes available for data mining. The data set of probes was mapped to known genes by elimination of multiple probes for the same gene, predicted genes, unidentified genes from cloning projects, somatic rearrangement clones, and probes corresponding to protein fragments. The verified gene list was further evaluated by Gene Ontology (GO) categories (http://www.geneontology.org), focusing on biological processes connected with transplantation. The resulting data set was sorted, selected, and queried. Selected genes were analyzed by quantitative RT-PCR (qRT-PCR) using GADPH as a housekeeping gene on the SDS7900 (Applied Biosystems) using the following commercial primers: Aif1 Rn00567 906_g1; B2m Rn00560865_m1; BMP3 Rn00567346_m1; C1s Rn00 594278_m1; C4a Rn00709527_m1; Dusp2 Rn017772516_g1; Fgl2 Rn00584935_m1;Hmox1 Rn01536933_m1; Ntrk1 Rn00572130_m1 (Applied Biosystems, Foster City, CA) and the following in-house primers: CCL2 forward TTG GCT CAG CCA GAT GCA, reverse CCA GCC TAC TCA TTG GGA TCA, probe CCC CAC TCA CCT GCT GCT ACT CAT TCA; CCL5 forward GTC GTC TTT GTC ACT CGA AGG A, reverse GAT GTA TTC TTG AAC CCA CTT CTT CTC, probe CCG CCA AGT GTG TGC CAA CCC; Cxcl10 forward GCT GAG TCT GAG TGG GAC TCA AG, reverse CTC AGC GTC TGT TCA TGG AAG T, probe ATC CCT CTC GCA AGA ACG GTG CG; IRF1 forward GGA TAT CAA CAA GGA TGC CTG TCT, reverse GCT CTT TTT CCC CAG CTT TGT AT, probe AGC TGG GCC ATT CAC ACA GGC CTA; MMP2 forward GCC TGA GCT CCC GGA AAA, reverse CCT GCG AAG AAC ACA GCC TTC T, probe ATT GAT GCC GTG TAC GAG GCC CC; MMP9 forward GCC TGA GCT CCC GGA AAA, reverse CCT GCG AAG AAC ACA GCC TTC T, probe ATT GAT GCC GTG TAC GAG GCC CC; Perforin1 forward GCA GGT CAG GCC AGC ATA AG, reverse GCA GTC CTG GTT GGT GAC CTT, probe ATT CAT GCC AGT GTG AGT GCC AGG ATT; TGFb1 forward GCT GCT GAC CCC CAC TGA T, reverse GCC ACT GCC GGA CAA CTC, probe CGC CTG AGT GGC TGT CTT TTG ACG TTA; VCAM1 forward CCT AAG GAT CCA GAG ATT CAA TTC A, reverse GGG TAA ACA TCA GGA GCC AAA C, probe TTG GGA AGC CGG TCA TGG TCA AGT. The samples used for qRT-PCR include the same RNA samples used for the microarray, plus samples from additional rats from independent experiments. The number of samples was 3–5 per group. A significant gene was considered “confirmed” if the average level in the BFK group compared to the isograft group was changed by 3-fold or more in both the RT-PCR data and the microarray data in the same direction. Each time point was counted separately.

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

77

3.6. Statistical analysis Data are expressed as average±SD, or average±SEM, as indicated. Statistical significance of proteinuria and creatinine was determined by the Student T test. Gene expression data are expressed as fold-change relative to isografts. Genes were clustered using Genespring GX version 11 (Agilent Technologies), based on hierarchical clustering using Euclidean parameters and average distance. Significance of the late phase genes was further evaluated by the Student T-test comparing the BFK to Iso groups. Early phase genes in the BFK to Iso comparison were not significant by the T-test, due to an outlier sample in the Iso group. 4. Results 4.1. Functional and histological characteristics of CAN in the BFK group Functional changes of CAN were measured by levels of plasma creatinine and proteinuria in surviving recipients every week for 90 days after transplantation. The rat survival rates were affected by rejection and by FK506 treatment. 100% of the animals in the Iso group survived to 90 days (5/5 rats), compared to 31% of the animals in the Allo group, and 71% in the BFK group (5/16 and 5/7, respectively). In the isograft control, plasma creatinine levels remained within the normal range and proteinuria was undetectable during the entire observation period. However, in the Allo and BFK groups, plasma creatinine levels increased rapidly after 7 weeks (Fig. 1a). Proteinuria was detectable at 2 weeks in the BFK group, with subsequent dramatic increases (Fig. 1b), indicating progressive functional deterioration of the kidney grafts. Although the BKF group has higher values in the proteinuria levels, there were no significant differences between the BFK and Allo groups at any timepoint in either proteinuria or plasma creatinine levels. Kidney allografts from both the Allo and BFK groups demonstrated histological evidence of CAN when compared to those of the isograft control (Fig. 2a–g). The histological effects were evaluated by measuring each graft for ten pathological characteristics of CAN, adapted from the Banff classification. Some evidence of inflammation and glomerulitis was detectable in the BFK and Allo groups compared to the Iso group even in the early phase. There is substantially more tissue damage in the late phase, as measured by the average sum score of 16.7 in the BFK group and the high values observed for transplant glomerulopathy (2.1±0.9), mesangiolysis (2.1±0.9), and fibrointimal hyperplasia (2.2±1.1). The untreated Allo group shows similar evidence of CAN, at both early and late phases, with no significant differences between the BKF and the Allo groups. 4.2. Gene expression analysis and verified gene list To understand the development of CAN, total RNA from the early and late phase kidney tissues was analyzed with the Agilent Gene Chip platform. At each time point, we determined which probes were elevated or reduced more than 3-fold in the BFK group compared to the isograft controls. For the early phase at day 6, 689 probes were affected, with 470 probes elevated and 219 probes reduced. For the late phase, a total of 2082 probes were affected, with 1078 probes elevated and 1004 reduced. On the Agilent chip, genes can be represented by more than one probe. The data set of probes was mapped to known genes and the resulting list of verified genes was further sorted into Gene Ontology categories for biological processes. The resulting list of categorized genes was sorted into three sets for detailed analysis. The late phase genes were further analyzed by T tests. Set 1 consists of the 225 genes affected at the early phase of day 6, with 195 genes up-regulated and only 30 genes down-regulated in the BFK group compared to the isograft controls. Eleven of these genes are affected exclusively at the early phase and are not further affected at the late phase (Table 1). Set 2 contains 612 genes affected exclusively at the late phase, with 302 genes up-regulated and 310 genes down-regulated (Table 1). This set represents the advanced development of chronic rejection and the deterioration of the kidney. The total number of affected genes is much higher in the late phase, consistent with the higher histological scores and the greater functional damage to the kidney. The down-regulated genes in the late phase have unexpected features. There were a disproportionate number of down-regulated genes (310) in the late phase exclusive samples, compared to the 9 down-regulated genes in the early phase exclusive samples, and the 24 genes in the both phases. Set 3 consists of 214 genes which were strongly affected at both early and late phases, with 190 genes up-regulated and 24 genes down-regulated (Table 1 and Fig. 3 and SCD Table 1). Set 3 includes the majority of genes from Set 1, which indicates that these 214 genes are associated with the development of CAN in the BFK group during the entire 90 days. Set 3 represents the genes constantly affected throughout the development of chronic rejection; logically, they would be the key molecules for therapeutic targets and/or biomarkers for CAN. Hierarchical clustering was performed on the selected genes in Set 3 to verify the relationship between groups (Fig. 3a and b). The samples in the Iso group from both

Fig. 1. Plasma creatinine levels and urine protein levels in rats with kidney grafts. (a) Plasma creatinine, measured each week. (b) Urine protein, measured by twenty-four-hour collection of urine. Average±SEM. *, Pb0.05 comparing BFK to Iso, n=3–8.

early and late phases were tightly clustered. The BFK and Allo samples were not separated from each other by clustering, although the early phase was differentiated from the late phase. 4.3. Confirmation of microarray data by qRT-PCR Selected genes from Sets 1, 2 or 3 were analyzed by qRT-PCR (Fig. 4). For 18 genes measured at both early and late phases, the correspondence between microarray hits and RT-PCR results was 92%. Two genes from non-kidney tissue, Ntrk1 (neurotrophic tyrosine kinase receptor 1) and Bmp3 (bone morphogenetic protein 3), were identified as significant in the late phase by microarray analysis, but were not confirmed by RT-PCR.

5. Discussion Chronic allograft nephropathy remains the primary obstacle to long-term survival of kidney allografts in transplantation, yet the origin, pathogenesis, treatment, and prevention methods have remained largely unknown [2]. In this study, we used a well-established rat CAN model to visualize the dynamics of the key cellular and molecule profiles underlying the development of CAN [6,7,21]. Considering that CAN is an ongoing process, we examined the functional and histological changes of the CAN at the early phase (day 6) and the late phase (days 30–90) in parallel with microarray-based gene expression profiles. In this model, F344 grafts in LEW recipients with a brief dose of FK506 showed the earliest functional changes of CAN at 2 weeks and a dramatic increase in plasma creatinine and proteinuria by 90days after transplantation. Kidney allografts from the BFK group at day 6

78

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

A

B

C

D

E

F

G Pathological Characteristic Fibrosis Inflammation Tubulitis Vasculitis Mesangiolysis Glomerulitis Tubular Atrophy Glomerulosclerosis Fibrointimal Hyperplasia Transplant Glomerulopathy Actual Sum Score a b c

Early Phase

Isograftc 0.00 + 0.00 0.17 + 0.41 0.00 + 0.00 0.00 + 0.00 0.00 + 0.00 0.33 + 0.52 0.17 + 0.41 0.00 + 0.00 0.00 + 0.00 0.00 + 0.00 0.67 + 1.21

Allograftc 0.00 1.33 0.00 1.00 0.33 1.00 0.00 0.00 0.00 0.00 3.67

+ + + + + + + + + + +

0.00 1.15 0.00 1.00 0.58 1.00 0.00 0.00 0.00 0.00 3.21

Late Phase

BFK c 0.00 + 0.00 0.67 + 0.58 0.00 + 0.00 0.33 + 0.58 0.33 + 0.58 0.67 + 0.58 0.33 + 0.58 0.00 + 0.00 0.00 + 0.00 0.00 + 0.00 2.33 + 2.52

Isograftc

Allograftc

BFK c

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

1.29 + 0.49 1.14 + 0.69 0.14 + 0.38 1.43 + 0.98 2.14 + 0.90 1.00 + 0.58 1.43 + 0.98 0.86 + 0.69 2.43 + 0.98 1.71 + 1.25 13.57 + 2.94

1.44 + 1.24 1.67 + 1.12 1.00 + 1.32 1.78 + 1.09 2.11 + 0.93 1.44 + 0.73 2.00 + 1.00 0.89 + .033 2.22 + 1.09 2.11 + 0.93 16.67 + 6.67

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

+ + + + + + + + + + +

Isograft kidney samples had negligible pathologic findings for the above variables and thus did not suffer chronic rejection. Mean average was determined from 3-9 rats per group. Average pathological characteristic score of 0 (least severe) to 3 (most severe) +/-S.D, and sum score range is 0 -30 +/-S.D.

Fig. 2. Histological evidence of chronic allograft nephropathy in kidneys from BFK group. (a) Isograft (upper left), (b) allograft (upper center) and (c) BFK allograft (upper right) at 6 days post transplant. Isografts have no sign of CAN with only minimal inflammation. BFK allografts and untreated allografts not only show moderate inflammation but also evidence of glomerulitis and tubulitis. (d) Isograft (lower left), (e) allograft, and (f) BFK allograft (lower right) at 90 days post transplant. Isografts remain relatively unchanged from day 6. BFK allografts and untreated allografts show significant pathological changes including tubular atrophy, transplant glomerulopathy, mesangiolysis, and fibrointimal hyperplasia. (Per-Iodic Acid Schiff stain, magnification ×320. The bar indicates 50 μm.) (e) Histological scoring of chronic allograft damage in isografts and BFK allografts at day 6 and day 90 after transplantation. aIsograft kidney samples had negligible pathologic findings for the above variables and thus did not suffer chronic rejection. bAverage was determined from 3 to 9 rats per group. cAverage histological score±SD.

demonstrated minor inflammation and glomerulitis. However, BFK and Allo kidney allografts at the late phase showed signs of CAN in all of the Banff categories, with considerably higher scores for transplant glomerulopathy, mesangiolysis, and fibrointimal hyperplasia. These results are consistent with other studies of rat kidney allografts [1,21,22]. In order to establish a reliable model of late phase chronic rejection, a brief dose of FK506 is required [22]. LEW recipients of F344 kidney grafts with no treatment have a low survival rate of only 31% in this study, consistent with other reports [1,4]. The animals in the Allo group display a wide variation in survival and in health. Some of these animals experience severe graft rejection and early death, while those which survive to 90 days may experience milder rejection and slightly elevated proteinuria. In contrast, the BFK group has a higher survival rate of 71%, higher proteinuria levels, and more severe histological evidence of CAN at 90 days. Overall, the F344-Lew strain combination treated with a brief dose of FK506 has proven a robust animal model for chronic kidney rejection, in comparison to the untreated combination [22]. We compared the BFK group to the isograft group to study the molecular profiles of kidney graft decline at the early and late phases after transplantation and to explore the relationship between the pathological lesions of graft kidney and the changes in the gene

expression pattern. Whole-genome microarray analysis of 41,013 probes revealed that significant changes in expression levels of 225 verified genes were detected in the BFK group compared to the isograft group in the early phase (day 6), while a total of 826 genes were affected at the late phase (days 30–90) after grafting. Among the genes, 214 genes are associated with the progressive development of CAN since they are perturbed across the entire period of 90 days. Our observation of the 225 genes affected in the early phase, along with evidence of high mononuclear infiltration in the BFK grafts, suggested that acute rejection episodes and nonspecific injury-mediated inflammation are programmed during the early phase of transplantation, despite the brief course of low dose FK506 treatment. The immune response and inflammatory response are well-represented in the Set 1 genes, including such specific categories as humoral response, T cell activation, NK cells, B cells, chemokines, complement, antigen processing, cell adhesion and cell migration. The immune response genes Aif1, B2microglobulin, Perforin 1, TGF-b, as well as the complement genes C1s and C4a were induced at least 2-fold at the early phase, and 3-fold or more at the late phase (Fig. 4). Chemokines CCL2, CCl5, and CXCL10 are strongly induced at both early and late phases, as reported in other examples in rat [23]. These results corresponded with those of rodent [22–25] and clinical studies [18,25–27] using high-density microarrays to assess the

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

79

Table 1 Number and GO category of genes affected in the early phase, the late phase or both.

Notes 1 2 3 1,2 3 1 1 3 1 1,2 3 3 3 3 1,2 3 1 1 1 3 3 2 3 3 3 2 1 2 3 3 3 2 3 1,2 2 2 3 3 1 2 3

Gene Ontology category Immune response Inflammatory response Apoptosis/anti-apoptosis Immune/inflammatory Cell cycle T cell activation Humoral response Signaling Defense response Cell migration Metabolism Transport Actin cytoskeleton Miscellaneous Cell adhesion Proteolysis Immune-complement system Immune-antigen processing NK cells G-protein coupled receptor Phosphorylation Blood coagulation Cell proliferation Blood pressure Brain/nervous system Response to wounding B cell Inflammatory chemokine Transcription Cell differentiation Hormone response Extracellular matrix Development Macrophage Angiogenesis Inflammation-acute phase Muscle Catabolism

Early exclusive

Early exclusive

Both early and late

Both early and late

Late exclusive

Late exclusive

Up 0 0 1 0 1 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Down 0 0 0 0 0 0 0 0 0 1 2 0 0 0 0 0 0 0 0 0 0 0 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0

Up 20 17 15 7 9 10 10 9 8 7 7 6 5 2 4 6 7 5 6 3 4 3 3 4 2 2 2 3 2 1 1 0 0 0 0 0 0 0

Down 0 1 0 0 2 0 0 0 0 0 9 2 1 2 2 1 0 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 1 0 0 0 0 0

Up 12 9 16 3 17 4 1 9 2 1 23 17 16 12 18 8 4 3 1 11 2 5 9 4 18 11 5 6 4 2 2 12 10 7 4 5 4 5

Down 1 6 7 0 5 1 0 5 3 0 120 51 2 10 4 10 2 0 0 5 4 3 6 11 8 6 0 0 4 3 4 2 7 2 0 0 3 15

Associated with immune response. Associated with inflammatory response. Non-specific.

transcriptome of biopsy samples from kidney transplants, which found that mononuclear cell infiltration associated with initial immunoresponse and nonspecific injury transcriptomes are dominant in this phase. Clearly, the synergy between early acute rejection episodes and non-specific injury-related inflammation in the early phase allows further events to become autonomous, leading to progressive allograft deterioration and finally to graft loss after transplantation. A distinctive gene expression pattern at the late phase was observed in the grafts from the BFK group compared to the isograft control, which may explain some of the functional and histological changes of CAN. The majority of the up-regulated genes belong to the same categories as observed at day 6, indicating an increasing number of time‐dependent injuries to the grafts. However, several categories of up-regulated genes appear exclusively in the late phase, such as muscle, macrophage, acute-phase inflammation, ECM and angiogenesis. Such pathways may be very important indicators of the deterioration pathway of the kidney during chronic rejection. ECM has been described by many groups to be of high prevalence in CAN, contributing to matrix accumulation and kidney function deterioration [28–31]. Genes in the ECM category contribute to either increased production and/or reduced degradation of the extracellular matrix, ultimately resulting in interstitial fibrosis and tubular atrophy (IFTA) which is a better characterization of CAN [32]. ECM genes identified in this analysis included MMP2 and MMP9 (Fig. 4), as well as Col1A2, Col3A1, and TIMP2, which were also identified by Rodder et al. (2011) [31] in human renal allografts with IFTA.

Antigen‐independent factors can also contribute to kidney chronic rejection. In addition to up-regulated genes in the late phase, 334 genes were down-regulated, including a great number and variety of non-immune pathways, such as transport, catabolism, and metabolism of amino acids, carbohydrates, and fatty acids. The identification of these down-regulated genes during the late phase of CAN strongly suggested that gene repression and/or the loss of molecular function can be important for graft survival [33]. Gene expression changes in metabolic pathways have been reported in studies of human kidney IFTA [31,34,35] and chronic graft nephropathy [36]. Several non-immune response genes such as Dusp2, Fgl2, Irf, and Hmox1 are increased at one or both timepoints. Irf1 was previously reported as a marker of acute heart rejection [33] and Hmox1 has been reported as a protective gene in the kidney allografts [23]. Attempts to identify specific markers in the metabolic pathways have not been confirmed by independent authors, but the importance of metabolism is clearly indicated by such consistent data. This topic should be examined more closely for the potential of future therapeutic interventions. Assuming that the regulatory molecules responsible for CAN would continue to be perturbed during an ongoing graft deterioration process, these genes should be affected at both early and late phases after transplantation [34]. Based on this logic, a total of 214 genes were identified in the overlap between early and late phases, with 190 genes up-regulated and 24 genes down-regulated in the BFK group compared to the isograft control. Up-regulated genes are

80

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

A

B

IIIIIIIIAFAAFAFFFAAFFAA eellllleeeleeeellllllll

IIIIIIIIAFAAFAFFFAAFFAA eellllleeeleeeellllllll Lbp Angpt14 Cxcl1 Gstm1 Fgg Ripk3 Tifa Tap2 Dok1 Birc3 Cd44 C4b Pcdh21 Ptger2 Bcat1 Tspan1 Cd47 Tnfsf9 Krt19 Xdh Glrx1 Upp1 Cd14 Gda Kif22 Rrm2 Ccna2 Top2a Ccnb1 Trpv2 P2ry6 Tlr2 Pycard Inpp5d Casp1 Glipr1 Tyrobp Gmfg Tnfrsf1b Hck Il18 C5ar1 Gna15 Fcgr1a Tnfaip812 Tbxas1 Cd68 C1qa Lyz2 C1qb Lcp2 Sla Btk Lyn Rgs19 Fyn Unc13d S100a4 Hmgb2 Psme2 Dusp2 Serpinb6b Cd40 Lat2 Il1b Kcnn4 Ptpn7 Ctsz Hcls1 Il21r Vav1 Trpm2 Ltb Zap70 Sept1 N/A Irgm Itgal Bcl2a1 Tap1 Psmb9 Psmb8 Psmb10 Irf1 Batf3 Il1rn RT1-T18 Cd3d Klra22 Rab27a Ptpn6 Arhgap4 Serpinb9 Lcp1 Rac2 Limd2 Prkcb1 Aif1 Cd6 Lat Cd4 Fgr Samsn1 Lilrb3 Cd2 Tubb3

Cybb Stx11 Ralgds Nampt Gch1 Il18bp Chi311 Il10ra Ccr2 Nfe2 Slpr5 Cebpe LOC24906 N/A Stat4 Cd69 Slfn3 Rhoh Laptm5 Ncf1 Cd74 Camk4 Klrd1 Ccl5 Ly49i2 Ccr5 Gzmk Sh2d2a Klre1 Ly49s3 Icos Cd8b Cd38 Klri2 Ccl3 Ccl7 Itgam Fcgr2a Slc1a3 Vcam1 Pla1a N/A Faslg Cd2 Cd5 Cxcr3 Prf1 Slpi Ctsw Klrk1 Fcgr3a Mkg7 Stat1 Cxcl10 Gzmb Il2rb Cita Gfi1 Cd8a Xd1 Gzma Gzmc Sele Apln Klrc3 Hk3 C2 Cxcl9 Dnase113 Tlr4 Klkb1 Ype14 Gpr119 Nefh Gpnmb C3 Slfn8 RT1-EC2 Lilrb4 Fcgr2b C4bpa Mmp7 Ptprq Acadsb Slc10a2 Rasd2 N/A Nox4 Fat3 Pgam2 Hsd17b1 Rxrg Mmdc4 N/A Mrps10 Krt76 Gtpbp4 Comt Klk53 Spink1 P2p Cyp2d2 Es22 Rdh2 Cyp2c11 Slco1a6 Ubd

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

81

32.00

Fold Change (Log2)

16.00 8.00 4.00

3x

2.00 1.00 0.50

0.3x 0.25 0.13 Fig. 4. qRT-PCR confirmation of microarray results for 18 genes. Microarray and RT-PCR results for both early and late phases are represented as average fold-change values of the BKF vs. Iso comparison. The location of the 3-fold increase and 3-fold decrease is shown by a marker line. Black bar: early phase RT-PCR. Light gray: early phase microarray. White bar: late phase RT-PCR. Dark gray: late phase microarray.

mostly derived from infiltrating cells of the immune system and/or inflammatory reactions in the graft tissue. Down-regulated genes may reflect disruptions in the tissue or cellular metabolism of the kidney. These observations are consistent with the evidence of progressive functional deterioration and histological changes, suggesting that these genes are critical molecules which serve to initiate and sustain CAN progression as well as to regulate the processes of graft injury and repair, eventually resulting in permanent damage and graft loss. Detailed insights into the molecular pathogenesis of these genes might indicate pathways containing biomarkers for clinical purposes and be beneficial for designing appropriate therapeutic strategies for prevention and treatment of CAN [37] Microarrays are valuable tools for the gene expression profiling, but they are not perfect. Therefore, qRT-PC was used to measure the mRNA levels of 18 genes from the 3 sets to confirm the microarray results. The correspondence between microarray hits and the qRT-PCR measurement was 92%. This value is in good agreement with other microarray-based gene expression profiles of transplant rejection [24,38–40]. However, this confirmation rate does not address the issue of genes not detected by the microarray. For example, the level of IL-2 mRNA cannot be measured with this array hybridization method, although it can be quantified by RT-PCR [33]. Calcineurin inhibitors (CNI) are reported to have deleterious effects on kidney function. The effect of the brief dose of the calcineurin inhibitor FK506 on kidney function is difficult to measure, due to the wide variation in health and survival times among the control untreated Allo group in this model. Thus our data do not allow generalization of the influence of FK506 on gene expression changes. Future studies may be needed to address the effect of CNI by using shorter-term models where the survival rates are more equal. Microarrays can be useful tools to analyze complex situations. The gene expression profiles in this study give information about abundance of both up-regulated and down-regulated genes in the early phase, the late phase, and the overlap during the development of the functional and histological features of CAN in this experimental model. Such knowledge allows immunologists to investigate the complex phenomena of chronic allograft nephropathy with a nonhypothesis-based approach, to view a more complete picture of the

disease at the molecular level, and to provide potential new therapeutic approaches or biomarkers for prevention and treatment of CAN in clinical transplantation. Acknowledgments The authors thank Dr. Janardan Khandekar for critical reading of the article, Ms. Gail Jurzec and Mr. Haeman Teo for administrative support, and Drs. Shuichi Tawara and Takao Fujimura for scientific support. References [1] Pan F, Ebbs A, Wynn C, Erickson L, Jang M, Crews G, et al. FK778, a powerful new immunosuppressant effectively reduces functional and histologic changes of chronic rejection in rat renal allografts. Transplantation 2003;75(8):1110–4. [2] Pascual M, Theruvath T, Kawai T, Tolkoff-Rubin N, Cosimi AB. Strategies to improve long-term outcomes after renal transplantation. N Engl J Med 2002;346(8):580–90. [3] Tilney NL, Whitley WD, Diamond JR, Jupiec-Weglinski JW, Adams D. Chronic rejection: an undefined conundrum. Transplantation 1991;53:389. [4] Laskowski IA, Pratschke J, Wilhelm MJ, Dong VM, Beato F, Taal M, et al. Anti-CD28 monoclonal antibody therapy prevents chronic rejection of renal allografts in rats. J Am Soc Nephrol 2002;13:519–27. [5] Mannon RB. Therapeutic targets in the treatment of allograft fibrosis. Am J Transplant 2006;6:867–75. [6] Diamond JR, Tilney NL, Frye J, Ding G, McElroy J, Pesek-Diamond I, et al. Progressive albuminuria and glomerulosclerosis in a rat model of renal allograft rejection. Transplantation 1992;54:710. [7] Ziai F, Nagano H, Kusaka M, Coito AJ, Troy JL, Nadeau KC, et al. Renal allograft protection with losartan in Fisher–Lewis rats: hemodynamics, macrophages, and cytokines. Kidney Transplant 2000;57:2618–25. [8] Wynn C, Erickson L, Pan F, Hong IC, Yamada T, Wu L, et al. Chronic kidney allograft rejection in F344 to LEW rats: relationship of alloantibodies, C4d deposition and glomerulopathy/intimal fibrosis. Am J Transplant 2008;8:340–1. [9] Schena M, Shalon D, Heller RA, Chai A, Brown PO, Davis RW. Parallel human genome analysis: microarray-based expression monitoring of 1000 genes. Proc Natl Acad Sci U S A 1996;93:10614. [10] Ollila J, Vihinen M. Stimulation of B and T cells activates expression of transcription and differentiation factors. Biochem Biophys Res Commun 1998;249:475. [11] Alizadeh A, Eisen M, Botstein D, Brown P, Staudt L. Probing lymphocyte biology by genomic-scale gene expression analysis. J Clin Immunol 1998;18:373. [12] Teague TK, Hildeman D, Kedl RM, Mitchell T, Rees W, Schaefer BC, et al. Activation changes the spectrum but not the diversity of genes expressed by T cells. Proc Natl Acad Sci U S A 1999;96:12691.

Fig. 3. Heatmap of hierarchical clustering of 214 genes from Set 3. (a) Genes 1–106. (b) Genes 107–214. Each cell indicates the expression value of the gene in that sample. I=Iso, A=Allo, F=BFK groups; e=early phase, l=late phase. Red: up-regulated; blue: down-regulated. Gene symbols are given on the right. Fold-change values, GO categories, and Entrez Gene ID numbers can be found in SDC Table 1.

82

L. Erickson et al. / Transplant Immunology 27 (2012) 75–82

[13] Rogge L, Bianchi E, Biffi M, Bono E, Chang SY, Alexander H, et al. Transcript imaging of the development of human T helper cells using oligonucleotide arrays. Nat Genet 2000;25:96. [14] Heller RA, Schena M, Chai A, Shalon D, Bedilion T, Gilmore J, et al. Discovery and analysis of inflammatory disease-related genes using cDNA microarrays. Proc Natl Acad Sci U S A 1997;94:2150. [15] Chen J-S, Coustan-Smith E, Suzuki T, Neale G, Mihara K, Pui CH, et al. Identification of novel markers for monitoring minimal residual disease in acute lymphoblastic leukemia. Blood 2001;97:2115. [16] Li S, Ross DT, Kadin ME, Brown PO, Wasik MA. Comparative genome-scale analysis of gene expression profiles in T cell lymphoma cells during malignant progression using a complementary DNA microarray. Am J Pathol 2001;158:1231. [17] Saiura A, Kohro T, Yamamoto T, Izumi A, Wada Y, Aburatani H, et al. Detection of an up-regulation of a group of chemokine genes in murine cardiac allograft in the absence of interferon-g by means of DNA microarray. Transplantation 2002;73: 1480. [18] Melk A, Mansfield ES, Szu-Chuan H, Hernandez-Broussard T, Grimm P, Rayner D, et al. Transcriptional analysis of the molecular basis of human kidney aging using cDNA microarray profiling. Kidney Int 2005;68:2667–79. [19] Fisher B, Lee S. Microvascular surgical techniques in research, with special reference to renal transplantation in the rat. Surgery 1965;58:904. [20] Solez K, Axelsen RA, Benediktsson H, Burdick JF, Croker BP, Droz D, et al. International standardization of criteria for the histologic diagnosis of renal allograft rejection: the Banff working classification of kidney transplant pathology. Kidney Int 1993;44:411. [21] Azuma H, Nadeau KC, Ishibashi M, Tilney NL. Prevention of functional, structural, and molecular changes of chronic rejection of rat renal allografts by a specific macrophage inhibitor. Transplantation 1995;60(12):1577–82. [22] Bedi DS, Riella LV, Tullius SG, Chandraker A. Animal models of chronic allograft injury: contributions and limitations to understanding the mechanisms of longterm graft dysfunction. Transplantation 2010;90(9):935–44. [23] Edemer B, Kurian SM, Eisenacher M, Lang D, Muller-Tidow C, Gabriels G, et al. Activation of counter-regulatory mechanisms in a rat renal acute rejection model. BMC Genomics 2008;9:71–87. [24] Cham CM, Xu H, O'Keefe JP, Rivas FV, Zagouras P, Gajewski TF. Gene array and protein expression profiles suggest post-transcriptional regulation during CD8 T cell differentiation. J Biol Chem 2003;278(19):17044–52. [25] Einecke G, Melk A, Ramassar V, Zhu LF, Bleackley RC, Famulski K, et al. Expression of CTL associated transcripts precedes the development of tubulitis in T-cell mediated kidney graft rejection. Am J Transplant 2005;5:1827–36.

[26] Famulski KS, Einecke G, Sis B, Mengel M, Hidalgo LG, Kaplan B, et al. Defining the canonical form of T-cell-mediated rejection in human kidney transplants. Am J Transplant 2010;10(4):810–20. [27] Halloran PF, Einecke G. Microarrays and transcriptome analysis in renal transplantation. Nat Clin Pract Nephrol 2006;2(1):2–3. [28] Bedi S, Vidyasagar A, Djamali A. Epithelial-to-mesenchymal transition and chronic allograft tubulointerstitial fibrosis. Transplant Rev 2009;22:1–5. [29] Hertig A, Flier SN, Kalluri R. Contribution of epithelial plasticity to renal transplantation-associated fibrosis. Transplant Proc 2010;42:S7–S12. [30] Wu GD, Tuan TL, Bowdish ME, Jin YS, Starnes V, Cramer DV, et al. Evidence for recipient derived fibroblast recruitment and activation during the development of the chronic cardiac allograft rejection. Transplantation 2003;76(3):609–14. [31] Rodder S, Scherer A, Korner M, Marti HP. A subset of metzincins and related genes constitutes a marker of human solid organ fibrosis. Virchows Arch 2011;458: 487–96. [32] Strutz F. Pathogenesis of tubulointerstitial fibrosis in chronic allograft dysfunction. Clin Transpl 2009;23:26–32. [33] Erickson LM, Pan F, Ebbs A, Kobayashi M, Jiang H. Microarray-based gene expression profiles of allograft rejection and immunosuppression in the rat heart transplantation model. Transplantation 2003;76(3):582–8. [34] Maluf DG, Mas BR, Archer KJ, Yanek K, Gibney EM, King AL, et al. Molecular pathways involved in loss of kidney graft function with tubular atrophy and interstitial fibrosis. Mol Med 2008;14:276–85. [35] Scian MF, Maluf DG, Archer KJ, Suh JL, Massey D, Fassnacht RC, et al. Gene expression changes are associated with loss of kidney graft function and interstitial fibrosis and tubular atrophy: diagnosis versus prediction. Transplantation 2011;91(6):657–65. [36] Donauer J, Rumberger B, Klein M, Faller D, Wilpert J, Sparna T, et al. Expression profiling on chronically rejected transplant kidneys. Transplantation 2003;76(3):539–47. [37] Nankivell BJ, Borrows RJ, Fung CL, O'Connel PJ, Allen RD, Chapman JR. The natural history of chronic allograft nephropathy. N Engl J Med 2003;349:2326–33. [38] Stroo I, Stokman G, Teske GJD, Raven A, Butter LM, Florquin S, et al. Chemokine expression in renal ischemial/reperfusion injury is most profound during the reparative phase. Int Immunol 2010;22(6):433–42. [39] Eikmans M, Roos MC, Sijpkens YWJ, Ehrchen J, Roth J, Baelde HJ, et al. Expression of surfactant protein-C, S100A9, and B cell markers in renal allografts: investigation of the prognostic value. J Am Soc Nephrol 2005;16:3771–86. [40] Lande JD, Patil J, Li N, Berryman TR, King R, Hertz MI. Novel insights into lung transplant rejection by microarray analysis. Proc Am Thorac Soc 2007;4:44–51.