Secondary Logo

Journal Logo

Functional Genomic Analysis of Peripheral Blood During Early Acute Renal Allograft Rejection

Günther, Oliver P.1,2; Balshaw, Robert F.1,3; Scherer, Andreas4; Hollander, Zsuzsanna1,2; Mui, Alice1,5,6; Triche, Timothy J.7; Freue, Gabriela Cohen1,3; Li, Guiyun8; Ng, Raymond T.1,9; Wilson-McManus, Janet1,10; McMaster, W Robert1,5,11; McManus, Bruce M.1,2,10; Keown, Paul A.1,8,10,12,13for the Biomarkers in Transplantation Team

doi: 10.1097/TP.0b013e3181b7ccc6
Clinical and Translational Research
Free

Background. Acute graft rejection is an important clinical problem in renal transplantation and an adverse predictor for long-term graft survival. Peripheral blood biomarkers that provide evidence of early graft rejection may offer an important option for posttransplant monitoring, optimize the utility of graft biopsy, and permit timely and effective therapeutic intervention to minimize the graft damage.

Methods. In this feasibility study (n=58), we have used gene expression profiling in a case-control design to compare whole blood samples between normal subjects (n=20) and patients with (n=11) or without (n=22) biopsy-confirmed acute rejection (BCAR) or borderline changes (n=5).

Results. A total of 183 probe sets representing 160 genes were differentially expressed (false discovery rate [FDR] <0.01) between subjects with or without BCAR, from which linear discriminant analysis and cross-validation identified an initial gene signature of 24 probe sets, and a more refined set of 11 probe sets found to classify subject samples correctly. Cross-validation suggested an out-of-sample sensitivity of 73% and specificity of 91% for identification of samples with or without BCAR. An increase in classifier gene expression correlated closely with acute rejection during the first 3 months posttransplant. Biological evaluation indicated that the differentially expressed genes encompassed processes related to immune response, signal transduction, and cytoskeletal reorganization.

Conclusion. Preliminary evidence indicates that gene expression in the peripheral blood may yield a relevant measure for the occurrence of BCAR and offer a potential tool for immunologic monitoring. These results now require confirmation in a larger cohort.

1 PROOF Centre of Excellence, University of British Columbia, Vancouver, BC, Canada.

2 James Hogg iCAPTURE Centre, University of British Columbia, Vancouver, BC, Canada.

3 Department of Statistics, University of British Columbia, Vancouver, BC, Canada.

4 Spheromics, Kontiolahti, Finland.

5 Infection and Immunity Research Centre, University of British Columbia, Vancouver, BC, Canada.

6 Department of Surgery, University of British Columbia, Vancouver, BC, Canada.

7 Department of Pathology and Laboratory Medicine, Los Angeles Childrens Hospital, University of California, Los Angeles, CA.

8 Department of Medicine, University of British Columbia, Vancouver, BC, Canada.

9 Department of Computer Science, University of British Columbia, Vancouver, BC, Canada.

10 Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada.

11 Department of Medical Genetics, University of British Columbia, Vancouver, BC, Canada.

12 Immunology Laboratory, University of British Columbia, Vancouver, BC, Canada.

This work was supported by Genome Canada and partnered grants from Novartis Pharma, Basel, and IBM, NY.

The authors declare no conflict of interest.

13 Address correspondence to: Paul A. Keown, M.D., Division of Nephrology, University of British Columbia, 855 W 12th Avenue, Vancouver, BC V5Z 1M9, Canada.

E-mail: keown@interchange.ubc.ca

Received 15 April 2009. Revision requested 6 May 2009.

Accepted 23 June 2009.

The success and clinical utility of renal transplantation have increased dramatically during the past decades, driven principally by advances in immunosuppression (1–3). Patient and graft survival exceed 95% and 90%, respectively, during the first year, and more than 80% of patients now remain free from acute rejection. Infectious complications have diminished in frequency and severity, and there has been a corresponding improvement in both quality of life and overall cost effectiveness (4–7). The use of potent immunosuppression has important adverse consequences, however, and attention has focused on individualizing therapy to optimize function and survival. This focus requires precise estimation of recipient risk, selection of appropriate prophylaxis, accurate and early diagnosis of rejection, and speedy intervention to prevent graft damage. The graft is at particular risk of physiologic and immunologic injury during the first weeks posttransplant, both of which may jeopardize long-term success. Understanding and distinguishing the biological processes involved through the use of specific and discriminant diagnostic procedures, and identifying selective inhibitors of key molecular steps in these injurious processes are, therefore, critical to achieving success (8, 9). Unfortunately, current approaches do not fulfill these goals. Routine laboratory tests used for graft monitoring are largely nonspecific and do not discriminate between mechanisms of injury (10, 11). Graft biopsy remains the primary diagnostic tool for confirming acute rejection, but the complexities, cost, and potential risks limit its use as a routine monitoring procedure (12, 13), whereas concerns of intrarater interpretation limit its diagnostic accuracy particularly in discriminating milder degrees of injury (14–18).

The development of high-throughput microarray technology, permitting simultaneous measurement of changes in expression of multiple genes within the human genome, provides the opportunity for novel insight into disease processes and molecular pathways of tissue injury (19–21). Recent advances have improved the diagnostic sensitivity, specificity, and accuracy of histologic diagnosis using this technology (22–24), and both biomarker panels and individual biomarkers have been identified within the allograft to improve the diagnostic, prognostic, and potentially therapeutic categorization of acute rejection (25–28). Whether noninvasive sources, such as the peripheral blood, will offer similar discriminant information remains to be determined (29–32), although the potential seems promising, and the field of convergent functional genomics is consequently a focus of intense investigation in many disease states (21, 33). This study examines the differential patterns of gene expression in normal subjects and patients with or without early biopsy-confirmed acute rejection (BCAR) after renal transplantation and proposes a potential classifier (biomarker panel) for the rejection episode.

Back to Top | Article Outline

MATERIALS AND METHODS

Study Design

This prospective longitudinal study was conducted at the University of British Columbia and was approved by the human research ethics board at that institution. All subjects who received a renal transplant from January 2005 to December 2007 were invited to participate, and those who agreed and signed consent forms were enrolled into the study. Patients were followed up routinely at the transplant center, and blood and urine samples were obtained before and serially posttransplant at 0.5, 1, 2, 3, 4, 8, 12, and 26 weeks, then every 6 months until year 3, and also at the time of suspected rejection. Graft tissue was obtained at pretransplant and at the time of all biopsies performed posttransplant. All samples were stored in a biolibrary until required for analysis. Patient samples were then selected for analysis of gene, protein, or metabolite expression or other investigation as required. Blood samples from normal healthy controls served as comparators. Samples from cases, controls, and comparators were treated identically.

Back to Top | Article Outline

Analysis Population

To ensure homogeneous phenotypes and to minimize biological variability for this analysis, patients were considered eligible if they were receiving their first kidney transplant; were less than 75 years; were not receiving immunosuppression before transplantation; had not received pretransplant immunologic desensitization; had received a kidney transplant from a deceased or human leukocyte antigen non (HLA)-identical living donor; had a negative anti-human globulin-complement-dependent cytotoxicity (AHG-CDC) antidonor T-cell crossmatch; had not received depleting antibody induction therapy with antithymocyte globulin or muromonab (OKT3); were able to receive oral medication, had immediate graft function, and had no clinical or laboratory evidence of infections, disease recurrence, and other major comorbid events.

The study used a closed cohort case-control design (34) to compare the differential gene expression in subjects with or without BCAR during the first 3 months posttransplant. Patients with BCAR (cases) diagnosed during the first 12 weeks posttransplant were matched 1:2 with controls selected sequentially using identical criteria who did not have evidence of clinical or BCAR during the same period of observation. All rejection episodes were diagnosed by conventional clinical and laboratory parameters, were confirmed by biopsy, and graded according to the Banff ’97 working classification of renal allograft pathology (35). Banff categories 2 and 4 (antibody-mediated or acute/active cellular rejection) were considered significant. Subjects with borderline changes (category 3) were analyzed separately. All baseline demographic and follow-up data were recorded in the transplant program electronic database, and there was no loss to follow-up during the period of study.

Back to Top | Article Outline

Immunosuppression

Immunosuppression consisted of basiliximab at 20 mg intravenously (IV) on days 0 and 4, with tacrolimus 0.075 mg/kg twice daily and mycophenolate 1000 mg twice daily Drug concentrations were measured by tandem mass spectrometry; the tacrolimus dose was adjusted to achieve 12-hr trough levels of 8 to 12 ng/mL for the first month posttransplant, 6 to 9 ng/mL for the second month, and then 4 to 8 ng/mL thereafter. First graft and nonsensitized subjects received methylprednisolone 125 mg IV on the day of transplantation, oral prednisone of 1 mg/kg on day 1, and declining to zero by day 3 posttransplant. Rejection episodes were treated with methylprednisolone 500 mg IV daily for 3 to 5 days. Steroid-resistant rejections were treated with OKT3 5 mg IV or ALG 15 mg/kg IV daily for 7 to 10 days.

Back to Top | Article Outline

Gene Expression

Whole blood samples drawn into PAXgene tubes (BD, Oakville, Canada) at the scheduled time points and at the time of suspected rejection, and similar blood samples from a cohort of disease-free control subjects of comparable ages and sexes to the transplant recipients were stored at −80°C in the biomarker biolibrary until selected for analysis. Total RNA was extracted from the samples using a PAXgene Blood RNA kit (Qiagen, Mississauga, Canada), and the integrity and concentration of the RNA samples determined using the Agilent 2100 BioAnalyzer (Agilent Technologies, Palo Alto, CA). Approximately 1.5 μg of total RNA per sample was transported on dry ice to the College of American Pathologists/Clinical Laboratory Improvement Amendments certified Microarray Core Laboratory, Genome Core at the Children’s Hospital, Los Angeles, CA, where the gene expression studies were performed. RNA was labeled using the Affymetrix cDNA Synthesis Kit (Affymetrix Inc., Santa Clara, CA), and the cRNA fragments were hybridized on Affymetrix Human Genome U133 Plus 2.0 arrays and scanned.

Back to Top | Article Outline

Gene Expression Analysis

The microarray analysis produced one Cel file per sample with 54,613 probe sets that analyzed more than 47,000 transcripts and variants. Quality of the samples, hybridization, chips, and scanning were reviewed using the tools in the BioConductor packages affy version 1.16.0 and affyPLM version 1.14.0. Samples where the Cel files suggested quality concerns were excluded from the analysis and reanalyzed using a different aliquot of the same sample. Background correction, normalization, and summarization of the data from the Cel files were performed using the Robust multiarray average method (10) in the Bioconductor affy package version 1.16.0. Expression values were analyzed on the log base 2 scale. To ensure stability, preprocessing of the 58 samples used in this analysis was performed in the context of a larger pool of 416 samples taken from the same biomarker project. A raw expression filter left 21,771 probe sets with a signal intensity of 26=64 in at least 3 of 416 samples.

Back to Top | Article Outline

Statistical Analysis

Three moderated t-tests: linear models for microarray data (LIMMA) (36), robust LIMMA, and significance analysis of microarrays (37) were applied separately to the filtered data to detect genes with significant differential expression between subjects with or without BCAR. The final list of differentially expressed probe sets included only those that were determined to be significant by all three methods with an FDR less than 1%. The biomarker panel probe sets were then determined by using forward selection discriminant analysis on the list of differentially expressed probe sets (PROC STEPDISC, SAS version 9.1, SAS Institute, Cary, NC). The forward selection process added one probe set at each step until no additional probe sets were found to improve the discrimination ability at the default α=0.15 level.

An 11-fold cross-validation of the entire process of classifier construction was used to evaluate the performance of the principal classifier based on the biomarker panel. Samples were randomly divided into 11 disjoint sets, each consisting of one sample from subjects with and two without BCAR, mirroring the one-to-two distribution in the overall study cohort. For each of the 11 disjoint sets, a new classifier was constructed in the same manner as the principal classifier: identification of a list of differentially expressed probe sets based on three moderated t-tests, followed by forward selection discriminant analysis. The classification accuracy (sensitivity and specificity) of each of the 11 classifiers was then determined based on the three samples left out at each fold. Sensitivity and specificity for the principal classifier were estimated by averaging the performance across the 11-fold cross-validation samples.

Biological interpretation was aided by knowledge mining using MetaCore analysis (GeneGo, Inc, CA), public databases (e.g., PubGene: www.PubGene.com), and literature mining, for example, PubMed (www.pubmed.gov). Gene Ontologies and Networks in MetaCore were prioritized based on their statistical significance with respect to the size of the intersection of the dataset and the set of genes or proteins corresponding to the Gene Ontology category or network (https://portal.genego.com/help/P-valuecalculations.pdf).

Back to Top | Article Outline

RESULTS

Subjects

Of the 305 subjects who received a renal transplant during the period of observation, 27 (8.9%) developed BCAR, Banff grade more than or equal to 1a during the first 3 months posttransplant, whereas a further 24 (7.9%) had only borderline changes. A total of 11/27 (40.74%) subjects with grade more than or equal to 1a rejection on biopsy (range: 3–10 days, mean: 7 days; histology: grade 1a=7, 1b=1, 2a=3, C4d=2) fulfilled the case selection criteria with immediate graft function, and absence of infection or other confounding comorbid events, as did 5/24 (20.83%) subjects with borderline changes on biopsy (range: 5–7 days, mean: 6 days). A further 22 subjects who had immediate graft function, with no clinical or BCAR for at least 6 months after transplantation, and no confounding clinical comorbid events, were selected as matched controls, and 20 normal control subjects served as a comparator group. Demographic details are shown in Table 1. There was no evidence of immunologic presensitization. Graft function was significantly inferior in cases with BCAR at the first week posttransplant (27±10 vs. 42±13 mL/min/1.73 m2, P=0.004), but it was comparable between cases and controls by month 3 (48±11 vs. 51±8 mL/min/1.73 m2, P=0.359) and remained clinically stable with good allograft function throughout the 12-month period of observation (54±13 vs. 53±15 mL/min/1.73 m2 at month 12, P=0.859).

TABLE 1

TABLE 1

Back to Top | Article Outline

Microarray Expression

The peripheral blood samples were selected from each of the cases with BCAR at the time of biopsy for acute rejection and from the respective controls without BCAR at a time-point identical to the respective case, and those samples were compared with samples from normal comparators. Microarray analysis of the samples from patients with or without BCAR at an FDR less than 0.01 identified a total of 239 probe sets that were differentially expressed using LIMMA, 575 probe sets with robust LIMMA, and 2677 probe sets using significance analysis of microarray. The intersection of the three methods found a more restricted set of 183 probe sets, which were differentially expressed between cases (BCAR) and controls (no BCAR) for all three analytical methods. Of the 183 significantly differentially expressed probe sets, 182 were overexpressed in subjects with BCAR, whereas one (1565484_x_at coding for the epidermal growth factor receptor) was underexpressed (Fig. 1).

FIGURE 1.

FIGURE 1.

Unsupervised two-way hierarchical clustering and principal component analysis based on these probe sets showed discrete separation between normal subjects, patients with BCAR, and those without BCAR (Fig. 2A, B). When samples from subjects with borderline changes were introduced, they were distributed heterogeneously among the cases and controls with and without BCAR (Fig. 2C). The biological processes encompassed by the 183 differentially expressed probe sets, representing approximately 160 genes, are shown in Figure 3. Combination of overlapping networks in which probe sets were shared identified three major biological categories implying involvement of processes related to immune responses, signal transduction, and cytoskeletal reorganization. Analysis of gene-gene and protein-protein networks based on exploration of public knowledge databases (www.GeneGo.com) revealed that the cytokine-activated Jak-Stat pathway, interferon (IFN) signaling, lymphocyte activation, proliferation, chemotaxis, and apoptosis were prominently represented among the 183 differentially expressed probe sets.

FIGURE 2.

FIGURE 2.

FIGURE 3.

FIGURE 3.

Back to Top | Article Outline

Classifier Selection

Although many genes were associated with BCAR, colinearity implied that not all were necessary to develop a classifier for this event. Forward selection discriminant analysis was, therefore, used to identify a linear discriminant function consisting of a more parsimonious classifier from among the 183 differentially expressed probe sets initially documented. The principal 24 probe sets identified within this classifier, and their respective genes, are shown in Table 2.

TABLE 2

TABLE 2

Cross-validation of the entire gene set using the same reductive process was used to enhance the robustness of this classifier and to estimate the out-of-sample performance. The 11 probe set lists produced by this process contained a mean of 103 probe sets, and the six most significantly differentially expressed of the original 183 probe sets (trophoblast noncoding RNA, FKSG49, AVIL, SIGLEC9, ANP32A, and SLC25A16) were present in each list. Forward selection discriminant analysis identified a group of 11 classifiers with a union of 87 probe sets. Eleven of these probe sets, depicted in Table 2, were contained within the original 24 probe set classifier. Cross-validation yielded an overall mean sensitivity of 73% and specificity of 91%, for the identification of samples with or without BCAR.

Performance of the final 11 probe set classifier is shown in Figure 4. Diagnostic accuracy improved rapidly with addition of sequential probe sets (Fig. 4A), and the linear discriminant scores for the full 11 probe set classifier showed clear separation of the samples with and without BCAR (Fig. 4B). Finally, longitudinal monitoring during the first 3 months posttransplant showed a significant increase in classifier score at the time of BCAR (P=0.001), with a subsequent return to the baseline value after treatment and resolution of the rejection episode. No comparable increase occurred in subjects who did not experience BCAR, and there was no significant difference between these curves at any other time posttransplant (Fig. 4C).

FIGURE 4.

FIGURE 4.

Back to Top | Article Outline

DISCUSSION

This feasibility study, conducted in a group of carefully phenotyped patients without confounding comorbidities, demonstrates the potential of microarray analysis using peripheral blood samples to document the biological processes invoked during graft rejection and to develop molecular markers of BCAR. The pilot data presented here suggest that it is possible to identify a panel of genes that are significantly up-regulated in BCAR, and which can correctly classify samples with high cross-validation specificity. The data are consistent with cellular and molecular signaling pathways known to participate in the immune and inflammatory processes associated with graft rejection. The biological functions of the genes differentially expressed during rejection encompass three major biological categories of processes related to immune signal transduction, cytoskeletal reorganization, and apoptosis, and emphasize the participation of the cytokine-activated Jak-Stat pathway, IFN signaling, and lymphocyte activation, proliferation, chemotaxis, and adhesion.

The Jak tyrosine kinase-Stat transcription factor pathway is essential for immune cell development, proliferation, and function (38). We observed up-regulation of all four mammalian Jak family kinases and STAT3, STAT5 and STAT6 in patients with BCAR. STAT3 and STAT5 mediate signaling from receptors for cytokines, such as interleukin (IL)-6 and IL-2, and are important in T-, B- and natural killer-cell activation. STAT6 mediates signaling from the IL-4 receptor, and the observed up-regulation of both STAT6 and IL-4 receptor in BCAR suggests activation of Th2-specific T-cell responses. Although acute rejection is classically ascribed to cytotoxic T-cell mediated events, it is clear that Th2/STAT6 processes are also important (39). Genes involved in IFN signaling that are up-regulated in BCAR include the IFN-inducible guanylate-binding protein, the IFN-response factor 1, and STAT1. IFN-γ plays a central role in rejection, enhancing T-cell cytotoxicity and natural killer cell activity and increasing major histocompatibility complex (MHC) class I antigen expression on graft and host cells (40), and posttransplant measurement of IFN-γ levels by the ELISPOT assay may provide important information about immune status (11, 10). Two MHC class I genes, HLA-E and HLA-G, were overrepresented in rejection. Their products, which exist in soluble forms which, have immunomodulatory function (41) and are increased in acute organ graft rejection, acute graft-verus-host disease after bone marrow transplantation, and certain autoimmune diseases. Further work will be needed to determine whether sHLA-E and sHLA-G are expressed during acute rejection in our patients.

T-cell activation and proliferation involve actin remodeling (42, 43). On MHC-peptide or TCR engagement, the actin cytoskeleton is bundled at the site of engagement and is essential to forming the immune synapse. This is mediated by structural proteins, such as SLP-76, and ADAP, CDC42EP and by the actin bundling protein LCP-2 (44). The actin cytoskeleton is remodeled to link to the integrin-receptor complex through proteins such as talin and paxillin. The genes encoding these proteins are up-regulated in acute rejection. Stabilization of integrin-dependent adhesion triggers gene expression changes, some of which are permissive for IFN–γ signaling, mediated through the Janus Kinase/Signal Transducers and Activators of Transcription (JAK/STAT) pathway (45), for which we also found evidence in the present dataset. AVIL (Advillin), one of the most highly differentially expressed genes, codes for a member of the gelsolin/villin family of actin regulatory proteins with structural similarity to villin (46). This Ca2+ -regulated actin-binding protein may play an important regulatory role in the signaling pathway initiated by Mac-1 binding. Actin cytoskeleton rearrangements parallel other signaling events in activated T cells, such as mobilization of intracellular calcium stores and activation of molecular pathways, which lead to the transcriptional enhancement of IL-2 expression. We found that probe sets for many genes encoding proteins in this pathway, including CAMKK2, MAP3K2, and JUN-B, were up-regulated in graft rejection.

Apoptosis or cell death, another central theme detected in this dataset, was represented by caspase 4, presenilin1, NACHT leucine rich repeat and PYD containing 1 (NLRP1), and tumor necrosis factor receptor 1. ANP32A (Acidic nuclearphosphoprotein 32 family, member a), a highly differentially expressed gene, encodes a protein with proapoptotic function (47) and has been described to participate in the granzyme A mediated apoptosis pathway and the Wnt signaling pathway (48). To our knowledge, this is the first report of the involvement of this gene or protein in apoptosis linked to acute rejection. The apoptotic signal presumably arises both from the immune-mediated death of organ parenchymal cells and the activation-induced cell death of alloreactive T cells, by which the immune system limits the inflammatory response. The apoptotic signature detected in the peripheral blood during rejection may, thus, represent a combination of T-cell activation (tumor necrosis factor receptor 1 is a T-cell coreceptor) and activation induced cell death of cells, which have transited from the organ. Interestingly, SIGLEC-9 (Sialic-acid binding Ig-like lectin 9), another of the most differentially expressed genes, encodes a cell-adhesion molecule expressed on blood leukocytes which is up-regulated during inflammation and negatively regulates T cell and other leukocyte function through induction of apoptosis and other mechanisms (49, 50). SLC25A16 (solute carrier family 25 member 16) (mitochondrial carrier; Graves’ disease autoantigen) has been identified by sequence homology as a new member of the mitochondrial transporter superfamily whose role remains to be fully identified (51).

Eight of the eleven genes identified in the classifier are known to be closely involved in relevant biological processes. Trophoblast noncoding RNA has been shown to suppress MHC class II expression in mice through inhibition of CIITApIII activity (52–54) and was recently found to be a target gene for TP53 (p53), suggesting involvement in apoptosis or cell cycle control (55). ZNF438 acts as a transcriptional repressor and binds to E-box sequences in the immunoglobulin heavy chain enhancer and the regulatory regions of many tissue-specific genes (56). Calcium or calmodulin-dependent protein kinase kinase 2, β (CAMKK2 beta) is ubiquitously expressed and regulates activation of the transcription factor NFkappaB (57), which in turn controls many biological processes including transcription of inflammatory cytokines, promotion of cell survival, and resolution of inflammation (58). LMAN2 (lectin, mannose-binding 2), an intracellular lectin, which functions as a chaperone protein and transmembrane cargo receptor in the endoplasmic reticulum and golgi apparatus, is structurally associated with the immunoglobulin-binding protein Grp78, involved with the assembly and transport of immunoglobulins and maintenance of T-cell viability after TCR activation (59–61). 237442_at encoded on chromosome 10 seems to function in signal transduction from Ras activation to actin cytoskeletal remodeling, and mediates Rap1-induced formation of an integrin activation complex and T-cell adhesion (62, 63). JunB is a proto-oncogene and a key component of AP-1, which serves as a transcriptional activator of various cytokine genes, such as IL-2, IL-4, and IL-10, and a regulator of T-cell function (64). PRO1073 (CCNL1, cyclin 1) is a transcriptional regulator for the pre-mRNA splicing process, which seems to be involved in the regulation of RNA polymerase II. It functions in association with cyclin-dependent kinases and has a role in the second step of splicing and is inhibited by the chronic kidney disease (CKD)-specific inhibitor p21 (65, 66). ITGAX (integrin, α X complement component 3 recetor 4 subunit, CD11 c) is a fibrinogen receptor and combines with ITGB2 to form a leukocyte-specific integrin referred to as inactivated-C3b receptor 4, which is expressed in the cells of hematopoietic and immune lineage (67). C3b may deposit in damaged tissue, and play an essential role in leukocyte homing during tissue injury. The function of the proteins corresponding to FKSG49, FKSG49/LOC730444, and 1558448_a_at is as yet unknown.

These results from whole blood identify immune and inflammatory mechanisms, which are consistent with previous findings from microarray analyses of renal biopsies, although the transcript sets reported differ according to tissue source and sampling criteria. Sarwal et al. (25) reported that genes associated with apoptosis were increased in renal biopsies during acute rejection and found transcript groups indicating lymphocyte infiltration and activation driven by NF-kappaB and IFN-γ. Flechner et al. (31), investigating gene expression from biopsies and peripheral blood leukocytes, showed that apoptosis and immune cell infiltration were major themes in the graft tissue, whereas immunity or inflammation predominated in the blood cells. Halloran and coworkers (28, 68) have identified transcripts in the kidney tissue associated with cytotoxic T lymphocytes, IFN-γ signaling, and epithelial cell injury in both mouse and human.

The methodology used the current discovery study was designed to ensure rigorous phenotypic comparison and diagnostic simplicity, but this design naturally entails potential limitations. The use of whole blood as the sampling matrix, while avoiding the need for cell separation and consequent potential alterations in gene expression, does not permit discrimination in differential expression because of changes in cell function, type, or number. Changes in cell populations may produce important biological variations, however, and while the differential expression reported was evident only at the time of BCAR, suggesting that this is a robust genomic biomarker, this will be the subject of a subsequent analysis. The rigorous selection to ensure phenotypic homogeneity also limited the sample numbers interrogated in the present discovery study. This study does not show the performance of these biomarkers under conditions of clinical complexity, such as for example patients with associated delayed graft function or coincident infection. The avoidance of corticosteroids is not yet routine therapy, and the wide distribution of borderline samples may reflect heterogeneous biology causing minor histologic changes or reduced sensitivity of blood compared with tissue expression in identifying borderline rejection. Evaluation of a larger patient set will now enable us to compare the transcript signals observed in these settings. If confirmed in a broader multicenter Canadian study now underway, we can conclude that the profiling of gene expression or measurement of selected transcripts by more specific methods may offer a promising approach to monitor the immunologic course in patients after kidney transplantation.

Back to Top | Article Outline

ACKNOWLEDGMENTS

The authors thank the members of the Genome Canada Biomarkers in Transplantation Team, Drs. David Landsberg, John Gill, and R. Jean Shapiro, and the members of the Biomarkers Group in Novartis Pharma for their contribution to and critical review of the manuscript. They also thank the members of our Scientific Advisory Committee for their critical oversight and scientific guidance in this project: Kathryn Wood, Oxford University, UK; Ruedi Aebersold, Institute of Molecular Systems Biology, University of Zürich, Switzerland; John Quackenbush, Dana-Farber Cancer Institute, Boston, MA; Leigh Anderson, Plasma Proteome Institute, Washington, DC; Eric Olson, University of Texas—Southwestern, Dallas, TX; Maria Rosa Costanzo, Midwest Heart, Edward Heart Institute, Naperville, IL; Gunther Engel, Novartis Pharmaceuticals, Basel, Switzerland; and George Schreiner, Raven Biotechnologies, San Francisco, CA.

Back to Top | Article Outline

REFERENCES

1.US Renal Data System U. Annual Data Report: Atlas of End-Stage Renal Disease in the United States. Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases 2006.
2.Australia and New Zealand Dialysis and Transplant Registry A. ANZDATA Registry Report 2006. Adelaide, South Australia, Australia and New Zealand Dialysis and Transplant Registry 2006.
3.Canadian Organ Replacement Register C. Treatment of end-stage organ failure in Canada 1995 to 2004. Ottawa, Ontario, Canadian Institute for Health Information 2006.
4.Parasuraman R, Yee J, Karthikeyan V, et al. Infectious complications in renal transplant recipients. Adv Chronic Kidney Dis 2006; 13: 280.
5.Yao G, Albon E, Adi Y, et al. A systematic review and economic model of the clinical and cost-effectiveness of immunosuppressive therapy for renal transplantation in children. Health Technol Assess 2006; 10: iii-iv, ix-xi, 1.
6.Winkelmayer W, Weinstein M, Mittleman M, et al. Health economic evaluations: The special case of end-stage renal disease treatment. Med Decis Making 2002; 22: 417.
7.Keown P. Improving quality of life—The new target for transplantation. Transplantation 2001; 72(12 suppl): S67.
8.Gaston R. Current and evolving immunosuppressive regimens in kidney transplantation. Am J Kidney Dis 2006; 47(4 suppl 2): S3.
9.Zand M. Immunosuppression and immune monitoring after renal transplantation. Semin Dial 2005; 18: 511.
10.Rodrigo E, Arias M. A practical approach to immune monitoring in kidney transplantation. Minerva Urol Nefrol 2007; 59: 337.
11.Girlanda R, Mannon R, Kirk A. Diagnostic tools for monitoring kidney transplant recipients. Semin Nephrol 2007; 27: 462.
12.Furness P, Philpott C, Chorbadjian M, et al. Protocol biopsy of the stable renal transplant: A multicenter study of methods and complication rates. Transplantation 2003; 76: 969.
13.Curtis J. Protocol biopsy: Today’s fine-needle aspiration biopsy? Transplantation 2007; 84: 677.
14.Furness P, Taub N, Assmann K, et al. International variation in histologic grading is large, and persistent feedback does not improve reproducibility. Am J Surg Pathol 2003; 27: 805.
15.Gough J, Rush D, Jeffery J, et al. Reproducibility of the Banff schema in reporting protocol biopsies of stable renal allografts. Nephrol Dial Transplant 2002; 17: 1081.
16.Gaber L. Role of renal allograft biopsy in multicenter clinical trials in transplantation. Am J Kidney Dis 1998; 31(6 suppl 1): S19.
17.Veronese F, Manfro R, Roman F, et al. Reproducibility of the Banff classification in subclinical kidney transplant rejection. Clin Transplant 2005; 19: 518.
18.Colvin R, Cohen A, Saiontz C, et al. Evaluation of pathologic criteria for acute renal allograft rejection: Reproducibility, sensitivity, and clinical correlation. J Am Soc Nephrol 1997; 8: 1930.
19.Hayden P, El-Meanawy A, Schelling J, et al. DNA expression analysis: Serial analysis of gene expression, microarrays and kidney disease. Curr Opin Nephrol Hypertens 2003; 12: 407.
20.Henger A, Schmid H, Kretzler M. Gene expression analysis of human renal biopsies: Recent developments towards molecular diagnosis of kidney disease. Curr Opin Nephrol Hypertens 2004; 13: 313.
21.Kurian S, Grigoryev Y, Head S, et al. Applying genomics to organ transplantation medicine in both discovery and validation of biomarkers. Int Immunopharmacol 2007; 7: 1948.
22.Irizarry R, Hobbs B, Collin F, et al. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003; 4: 249.
23.Smyth G, Michaud J, Scott H. Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinformatics 2005; 21: 2067.
24.Gentleman R, Carey V, Bates D, et al. Bioconductor: Open software development for computational biology and bioinformatics. Genome Biol 2004; 5: R80.
25.Sarwal M, Chua M, Kambham N, et al. Molecular heterogeneity in acute renal allograft rejection identified by DNA microarray profiling. N Engl J Med 2003; 349: 125.
26.Einecke G, Reeve J, Mengel M, et al. Expression of B cell and immunoglobulin transcripts is a feature of inflammation in late allografts. Am J Transplant 2008; 8: 1434.
27.Famulski K, Broderick G, Einecke G, et al. Transcriptome analysis reveals heterogeneity in the injury response of kidney transplants. Am J Transplant 2007; 7: 2483.
28.Mueller T, Einecke G, Reeve J, et al. Microarray analysis of rejection in human kidney transplants using pathogenesis-based transcript sets. Am J Transplant 2007; 7: 2712.
29.Aquino-Dias E, Joelsons G, da Silva D, et al. Non-invasive diagnosis of acute rejection in kidney transplants with delayed graft function. Kidney Int 2008; 73: 877.
30.Marrer E, Dieterle F. Promises of biomarkers in drug development—A reality check. Chem Biol Drug Des 2007; 69: 381.
31.Flechner S, Kurian S, Head S, et al. Kidney transplant rejection and tissue injury by gene profiling of biopsies and peripheral blood lymphocytes. Am J Transplant 2004; 4: 1475.
32.Brouard S, Mansfield E, Braud C, et al. Identification of a peripheral blood transcriptional biomarker panel associated with operational renal allograft tolerance. Proc Natl Acad Sci USA 2007; 104: 15448.
33.Le-Niculescu H, Kurian S, Yehyawi N, et al. Identifying blood biomarkers for mood disorders using convergent functional genomics. Mol Psychiatry 2009; 14: 156.
34.Etminan M, Samii A. Pharmacoepidemiology I: A review of pharmacoepidemiologic study designs. Pharmacotherapy 2004; 24: 964.
35.Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713.
36.Smyth GK. Limma: Linear models for microarray data. In: Gentleman R, Carey V, Dudoit S, et al eds. Bioinformatics and computational biology solutions using R and bioconductor. NY, Springer 2005, pp 397.
37.Tusher V, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 2001; 98: 5116.
38.Hennighausen L, Robinson G. Interpretation of cytokine signaling through the transcription factors STAT5A and STAT5B. Genes Dev 2008; 22: 711.
39.Zhou P, Szot G, Guo Z, et al. Role of STAT4 and STAT6 signaling in allograft rejection and CTLA4-Ig-mediated tolerance. J Immunol 2000; 165: 5580.
40.Hidalgo L, Halloran P. Role of IFN-gamma in allograft rejection. Crit Rev Immunol 2002; 22: 317.
41.Puppo F, Contini P, Ghio M, et al. Soluble HLA class I molecules/CD8 ligation trigger apoptosis of CD8+ cells by Fas/Fas-ligand interaction. Scientific World J 2002; 2: 421.
42.Cannon J, Burkhardt J. The regulation of actin remodeling during T-cell-APC conjugate formation. Immunol Rev 2002; 186: 90.
43.Burbach B, Medeiros R, Mueller K, et al. T-cell receptor signaling to integrins. Immunol Rev 2007; 218: 65.
44.Peterson E. The TCR ADAPts to integrin-mediated cell adhesion. Immunol Rev 2003; 192: 113.
45.Ivaska J, Bosca L, Parker P. PKCepsilon is a permissive link in integrin-dependent IFN-gamma signalling that facilitates JAK phosphorylation of STAT1. Nat Cell Biol 2003; 5: 363.
46.Silacci P, Mazzolai L, Gauci C, et al. Gelsolin superfamily proteins: Key regulators of cellular functions. Cell Mol Life Sci 2004; 61: 2614.
47.Kim H, Jiang X, Du F, et al. PHAPI, CAS, and Hsp70 promote apoptosome formation by preventing Apaf-1 aggregation and enhancing nucleotide exchange on Apaf-1. Mol Cell 2008; 30: 239.
48.Stelzl U, Worm U, Lalowski M, et al. A human protein-protein interaction network: A resource for annotating the proteome. Cell 2005; 122: 957.
49.von Gunten S, Yousefi S, Seitz M, et al. Siglec-9 transduces apoptotic and nonapoptotic death signals into neutrophils depending on the proinflammatory cytokine environment. Blood 2005; 106: 1423.
50.Ikehara Y, Ikehara S, Paulson J. Negative regulation of T cell receptor signaling by Siglec-7 (p70/AIRM) and Siglec-9. J Biol Chem 2004; 279: 43117.
51.Prohl C, Pelzer W, Diekert K, et al. The yeast mitochondrial carrier Leu5p and its human homologue Graves’ disease protein are required for accumulation of coenzyme A in the matrix. Mol Cell Biol 2001; 21: 1089.
52.Geirsson A, Paliwal I, Lynch R, et al. Class II transactivator promoter activity is suppressed through regulation by a trophoblast noncoding RNA. Transplantation 2003; 76: 387.
53.Geirsson A, Lynch R, Paliwal I, et al. Human trophoblast noncoding RNA suppresses CIITA promoter III activity in murine B-lymphocytes. Biochem Biophys Res Commun 2003; 301: 718.
54.Geirsson A, Bothwell A, Hammond G. Inhibition of alloresponse by a human trophoblast non-coding RNA suppressing class II transactivator promoter III and major histocompatibility class II expression in murine B-lymphocytes. J Heart Lung Transplant 2004; 23: 1077.
55.Adamsen B, Kravik K, Clausen O, et al. Apoptosis, cell cycle progression and gene expression in TP53-depleted HCT116 colon cancer cells in response to short-term 5-fluorouracil treatment. Int J Oncol 2007; 31: 1491.
56.Zhong Z, Wan B, Qiu Y, et al. Identification of a novel human zinc finger gene, ZNF438, with transcription inhibition activity. J Biochem Mol Biol 2007; 40: 517.
57.Chen B, Wu W, Ho F, et al. Inhibition of interleukin-1beta -induced NF-kappa B activation by calcium/calmodulin-dependent protein kinase kinase occurs through Akt activation associated with interleukin-1 receptor-associated kinase phosphorylation and uncoupling of MyD88. J Biol Chem 2002; 277: 24169.
58.Schlottmann S, Buback F, Stahl B, et al. Prolonged classical NF-kappaB activation prevents autophagy upon E. coli stimulation in vitro: A potential resolving mechanism of inflammation. Mediators Inflamm 2008; 2008: 725854.
59.Knarr G, Kies U, Bell S, et al. Interaction of the chaperone BiP with an antibody domain: Implications for the chaperone cycle. J Mol Biol 2002; 318: 611.
60.Nawa D, Shimada O, Kawasaki N, et al. Stable interaction of the cargo receptor VIP36 with molecular chaperone BiP. Glycobiology 2007; 17: 913.
61.Takano S, Ando T, Hiramatsu N, et al. T cell receptor-mediated signaling induces GRP78 expression in T cells: The implications in maintaining T cell viability. Biochem Biophys Res Commun 2008; 371: 762.
62.Banno A, Ginsberg M. Integrin activation. Biochem Soc Trans 2008; 36(pt 2): 229.
63.Parmo-Cabañas M, García-Bernal D, García-Verdugo R, et al. Intracellular signaling required for CCL25-stimulated T cell adhesion mediated by the integrin alpha4beta1. J Leukoc Biol 2007; 82: 380.
64.George J, Braun A, Brusko T, et al. Suppression by CD4+CD25+ regulatory T cells is dependent on expression of heme oxygenase-1 in antigen-presenting cells. Am J Pathol 2008; 173: 154.
65.Loyer P, Trembley J, Grenet J, et al. Characterization of cyclin L1 and L2 interactions with CDK11 and splicing factors: Influence of cyclin L isoforms on splice site selection. J Biol Chem 2008; 283: 7721.
66.Loyer P, Trembley J, Katona R, et al. Role of CDK/cyclin complexes in transcription and RNA splicing. Cell Signal 2005; 17: 1033.
67.Puri T, Quigg R. The many effects of complement C3- and C5-binding proteins in renal injury. Semin Nephrol 2007; 27: 321.
68.Famulski K, Sis B, Billesberger L, et al. Interferon-gamma and donor MHC class I control alternative macrophage activation and activin expression in rejecting kidney allografts: A shift in the Th1-Th2 paradigm. Am J Transplant 2008; 8: 547.
Keywords:

Gene expression; Acute rejection; Whole blood; Biomarkers; Kidney transplantation

© 2009 Lippincott Williams & Wilkins, Inc.