Journal Logo

Original Clinical Science—General

Plasma Exosomes From HLA-Sensitized Kidney Transplant Recipients Contain mRNA Transcripts Which Predict Development of Antibody-Mediated Rejection

Zhang, Hao PhD1; Huang, Edmund MD2; Kahwaji, Joseph MD3; Nast, Cynthia C. MD4; Li, Ping PhD5; Mirocha, James MS6; Thomas, David L. BS1; Ge, Shili PhD1; Vo, Ashley A. PharmD2; Jordan, Stanley C. MD2; Toyoda, Mieko PhD1

Author Information
doi: 10.1097/TP.0000000000001834

Antibody-mediated rejection (AMR) is a major obstacle for successful transplantation in HLA-sensitized patients and a leading cause of graft failure for kidney transplant patients.1 Timely and accurate diagnosis is critical for the treatment of AMR and prevention of allograft failure. In addition to renal dysfunction, usually evidenced by serum creatinine elevation, and presence of donor-specific HLA antibody (DSA), the clinical diagnosis of AMR also requires the presence of histopathological findings in renal allograft tissue. These include evidence for microvascular inflammation and microcirculation injury.2 The latest Banff 2013 classification of AMR also recognizes molecular diagnostic markers such as “increased expression of gene transcripts in the biopsy tissue indicative of endothelial injury, if thoroughly validated,” as one of the criteria for diagnosis of both acute and chronic active AMR.2 There is increasing interest in effort to develop and validate gene signatures identified from the expression levels of large number of gene transcripts, using biopsy tissue samples of renal grafts, for the purposes of both clinical diagnosis of AMR and also elucidation of molecular biological mechanisms of AMR.3-6 However, it is important to recognize that biopsies do have inherent risk and may not always be possible, especially in patients on anticoagulation. Thus, there would be benefits to develop reliable assays using blood and urine, which are predictive of AMR and possible resolution after therapy.

Exosomes are small membrane vesicles with diameter of 30 to 100 nm, and secreted by most cell types. Exosomes exist abundantly in various types of body fluid, such as saliva, urine and blood plasma. Exosomes contain proteins, miRNA and mRNA, which are protected by the exterior membrane structures and directly related to the original cells secreting exosomes.7,8 Exosome-related research using various types of body fluid has increased over the last decade in multiple diseases, including cancer and organ transplantation.9-12 Here, we investigated the utility of mRNA transcripts in plasma exosomes as a possible noninvasive diagnostic tool to identify AMR development in HLA-sensitized kidney transplant patients.


Study Subjects

This study was approved by the institutional review board at Cedars-Sinai Medical Center (IRB numbers Pro00021002, Pro00034039). The study was conducted in accordance with the ethical guideline based on federal regulations and the common rule. Cedars-Sinai Medical Center also has a Federalwide Assurance.

152 archived ethylenediaminetetraacetic acid (EDTA)-plasma samples obtained from 64 kidney transplant recipients who were transplanted between August 2008 and July 2014 were used for this study. Among these 64 patients, 12 and 6 developed biopsy-proven AMR and AMR/cell-mediated rejection (CMR), respectively (AMR group), 8 had biopsy-proven CMR (CMR group), and 38 had no rejection (control groups). AMR and CMR were diagnosed based on the Banff 2013 and Banff 1997 classification, respectively. A diagnosis of acute/active AMR required microvascular inflammation with glomerulitis score greater than 0 and/or peritubular capillaritis score greater than 0 with positive C4d staining, or glomerulitis + peritubular capillaritis score greater than 2 with negative C4d staining, and serologic evidence of DSA. Chronic AMR required transplant glomerulopathy and/or peritubular capillary multilayering per Banff criteria.2 All biopsies were stained for C4d, with positive peritubular capillary staining in the majority of AMR patients. A diagnosis of CMR was made when the interstitial and tubular inflammation thresholds were met per Banff criteria for type I CMR and/or vascular inflammation for type II or III CMR.13 AMR was developed at median 4.1 months posttransplant (9 days to 6.5 years) and CMR at 3.8 months (1.3-6.5 months).

The desensitization protocols used for ABO compatible transplant in HLA-sensitized patients and ABO incompatible transplant in non–HLA-sensitized patients have been reported.14 Briefly, a standard protocol for HLA-desensitization consisted of 2 doses of IVIG (2 g/kg) 1 month apart with 1 dose of rituximab (1 g) in between. The protocol for ABO incompatible transplant consisted of 1 dose of rituximab (1 g) 2 weeks before initiation of 5 sessions of plasma exchange followed by 1 dose of IVIG (2 g/kg). The combination of both protocols was used for HLA-sensitized patients who received an ABO incompatible transplant. If a negative or acceptable crossmatch was achieved and/or the antiblood group titer became 1:8 or less after desensitization, the patient was transplanted.15,16

All except 1 patient in this study received induction therapy. Maintenance immunosuppression consisted of calcineurin inhibitor (tacrolimus or cyclosporine A), mycofenolate mofetil and steroids. The target levels were dependent on the type of induction as reported elsewhere.17

All patients received antiviral prophylaxis consisting of valganciclovir or aciclovir for 6 months posttransplant depending on his/her risk for viral infection and viral PCR monitoring as previously reported.18 AMR was treated with pulse steroids, IVIG and rituximab with or without plasma exchange. CMR was treated with pulse steroids. Refractory or Banff 2 rejection or higher was treated with antithymocyte globulin.

Candidate Genes

We selected 21 candidate genes for this study (Table 1). Among them, CCL3, CCL4, CD160, cytotoxic and regulatory T cell molecule (CRTAM), early growth response (EGR)1, EGR2, IFNγ, XCL1 are antibody-dependent cellular cytotoxicity (ADCC)-activated genes identified by our previous study.5 ADCC is an important mechanism of allograft injury which leads to AMR.19-21 CAV1, atypical chemokine receptor 1 (duffy blood group) (DARC), FGFBP2, GNLY, SH2D1B are AMR-associated genes as identified in previous studies of biopsy samples from renal transplant recipients.4,6 IL-6, IL-6Rα, and gp130 (IL-6 β receptor) are selected due to the important role of IL-6/IL-6R signaling in mediating AMR and DSA development.22-26 IL-10, IL-12α, IL-23α, TGFβ1, TNFα are selected due to their well-known important roles in all types of inflammatory or anti-inflammatory immune events.27-31

Candidate genes and Taqman assays

Plasma Sample Selection and Exosome RNA Extraction

For patients in the AMR and CMR groups, all the archived EDTA-plasma samples from each patient, which were collected within 1 month before diagnosis of rejection by biopsy, were included in this study. The mean number of days between plasma sample collection and biopsy for AMR and CMR groups were 13.7 ± 9.9 and 11.4 ± 5.7, respectively. The first biopsy-proven rejection of each person was investigated in this study. The mean months from transplant to rejection for AMR and CMR groups were 11.9 ± 19.7 and 3.8 ± 1.7, respectively. The mean number of plasma samples per patient included in the AMR and CMR groups were 1.4 ± 0.6 (range, 1-3 samples) and 1.9 ± 0.6 (range, 1-3), respectively, and 6 AMR and 6 CMR patients had multiple plasma samples included in this study. For most patients in the desensitized (DES) and non-DES control groups, 3 plasma samples per patient were included in this study, which were generally collected over the time course of 1 to 12 months posttransplant (2.9 ± 0.4 samples per patient; range, 2-4). Two hundred microliters of each selected plasma sample were submitted for exosome RNA extraction with exoRNeasy Serum/Plasma Midi Kit (Qiagen, Hilden, Germany) following the manufacturer's protocol.

Validation and Quality Control for Plasma Exosome RNA Samples

Exosome RNA extracted from 8 samples randomly selected among the plasma samples included in this study (2 samples from each patient group) were submitted for measurement of RNA yield and RNA integrity number (RIN), on an Agilent Bioanalyzer 2100 using the Agilent RNA 6000 Pico kit following the manufacturer's protocol.

Reverse Transcription, Preamplification, and Quantitative Polymerase Chain Reaction

cDNA was synthesized from total exosome RNA using High Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (Applied Biosystems/Thermo Fisher Scientific, Foster City, CA), and then cDNA was preamplified with the pooled TaqMan Gene Expression Assays (Applied Biosystems/Thermo Fisher Scientific) of all the genes we studied in addition to the reference gene of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) with Human GAPD Endogenous Control (Applied Biosystems/Thermo Fisher Scientific), following the standard protocols with TaqMan PreAmp Master Mix (Applied Biosystems/Thermo Fisher Scientific). The preamplification product (from exosome RNA in approximately 1.25 μL of original plasma for each gene) was then submitted for quantitative polymerase chain reaction (qPCR) for each of 21 genes using TaqMan Gene Expression Assay (Table 1) as well as GAPDH, and TaqMan Gene Expression Master Mix (Applied Biosystems/Thermo Fisher Scientific) on 7500 Real-Time PCR system (Applied Biosystems/Thermo Fisher Scientific). Total RNA prepared from a normal healthy individual's peripheral blood mononuclear cells stimulated with phorbol 12-myristate 13-acetate and ionomycin was included at each qPCR run in the entire study and used as the reference RNA. Because of the high sequence similarities between XCL1 and XCL2 (98%), and between CCL4, CCL4L1 and CCL4L2 (96% between CCL4 and CCL4L1, 100% between CCL4L1 and CCL4L2), the selected assays for XCL1 and CCL4 will detect 2 and 3 transcripts, respectively. The expression level of each gene was first normalized to GAPDH using ΔCt (cycle threshold) method for each RNA sample respectively, and then presented as relative quantity (RQ) to the expression level of the same gene in the reference RNA sample (also normalized to GAPDH) by calculating ΔΔCt.

Panel-Reactive Antibody, Anti-HLA Antibody Specificity, and DSA Score

Panel-reactive antibody (PRA) and antibody specificity assays were performed at Cedars-Sinai Medical Center HLA Laboratory using the methods previously described.32 Briefly, HLA antibodies were detected by either flow quick screen or single antigen Luminex bead assay (Luminex, Austin, TX) in serum samples. The binding levels of HLA-specific antibodies were determined by multi-analyte bead assay performed on Luminex platform. The single antigen Luminex bead assay was standardized with Quantiplex beads (One Lambda/Thermo Fisher Scientific, Canoga Park, CA). Final specificity was analyzed through HLA Visual 2.2 software (One Lambda/Thermo Fisher Scientific). DSA scores were calculated based on standard fluorescence intensities (SFI) of DSAs as previously described (score = 10 for SFI ≥ 200 000; score = 5 for 100 000 ≤ SFI < 200 000; score = 2 for 0 < SFI < 100 000; score = 0 for SFI = 0).33 If DSA against multiple donor HLA antigens were detected in the same patient, then the sum of each DSA score were used as the final DSA score for that patient. For the AMR group, the serum samples used for calculating DSA scores were collected on the same dates as the EDTA-plasma samples used for calculating the exosome RNA-based gene scores for 10 of 15 patients with available DSA results, and for the remaining 5 patients, the difference in sample collection days for DSA and gene score measurements was also minimal (mostly 2 to 3 days).

Data Analysis

Because the number of plasma samples included in the study varied for each individual patient, we first calculated the average RQ of each gene for each individual patient. For patients who had only 1 plasma sample included in this study, the RQ result from that single plasma/exosome RNA sample was used as the average RQ of that gene. Using the average RQ results of each gene for each patient, nonparametric Kruskal-Wallis H Test (KW test) was first performed to identify genes whose mRNA transcript levels (average RQ) exhibited significant difference among the 4 study groups of AMR, CMR, DES and non-DES controls. If the P value was less than 0.10 by KW test, nonparametric Mann-Whitney U test (MW test) was performed to make pairwise comparison within these 4 study groups. To calculate the gene combination score, genes showing significant elevation in AMR patients were first selected based on the KW and MW test analysis. In order for each selected gene to make equal contribution to the final gene combination score, we normalized the initial average RQ of each selected gene for each patient to the overall average RQ of the same gene among all 64 patients from all 4 study groups of AMR, CMR, DES and non-DES controls. The gene combination score for each patient was then obtained by calculating the average of normalized average RQ values from 4 selected genes. Further statistical analysis of gene combination scores was then conducted by KW and MW tests to identify the group(s) significantly different among the 4 study groups. Statistical analysis was done by Prism 6.0 (GraphPad Software, La Jolla, CA). P value less than 0.05 was considered statistically significant. Receiver operating characteristic (ROC) curve of gene score, the calculation of area under curve and all related statistical analysis were done by Stata, version 14.2 (StataCorp LP, College Station, TX).


Patient demographics are shown in Table 2. There is no significant difference among the 4 patient groups in terms of age, gender and living donor status, except race.

Summary of demographics and clinical/medication/treatment history

For induction therapy, most patients in the AMR and DES control groups received lymphocyte-depleting agents (alemtuzumab or antithymocyte globulin), while the majority of patients in the CMR and non-DES control groups received anti–IL-2 receptor antibody (basiliximab or daclizumab). All patients in the AMR group except 1 were HLA-sensitized before transplant (PRA class I [PRA I], 75% ± 38%; PRA II, 57% ± 35%) and only 1 patient in the AMR group received an ABO incompatible transplant. Most of the 17 HLA-sensitized patients received desensitization therapy with IVIG + rituximab with or without plasma exchange before transplant. The 1 non–HLA-sensitized AMR patient developed de novo PRA II (>60%) including de novo DSA at rejection, 6.5 years post transplant. Most patients in the CMR group were not HLA-sensitized (PRA I: 4% ± 11%, PRA II: 3% ± 9%), and none received pretransplant desensitization. For the DES and non-DES control groups, all 18 patients in the DES control group (PRA I, 51% ± 43%; PRA II, 25% ± 32%) received pretransplant desensitization, while none of the 20 non-DES patients (PRA I, 4% ± 8%; PRA II, 0% ± 2%) received it.

RNA quality and quantity was tested on 8 randomly selected plasma samples. The average exosome total RNA yield of these samples was 1.68 ± 0.50 ng per 200 μL plasma, and the average RIN was 1.69 ± 0.35 (Figure 1).

Total RNA amount and quality of selected exosome RNA samples extracted from 200 μL plasma samples. Horizontal long and short lines indicate mean ± standard deviation of results of 8 plasma samples (2 samples from each of the 4 patient groups).

Among the 21 candidate genes, we identified 9 genes whose mRNA transcript levels in plasma exosomes varied significantly or near significance when all 4 study groups were compared (P < 0.05 for 8 genes and P = 0.07 for SH2D1B) (Figure 2) (Table 3). These 9 genes include the IL-6 signaling-related gene (gp130), AMR-associated genes identified from renal biopsies (CAV1, DARC, SH2D1B),4,6 ADCC-associated genes identified in our study (CCL4, EGR1),5 and cytokine/inflammation-related genes (TNFα, IL-10, 1L-23α). Among the 9 genes, the mRNA transcript levels of 6 genes (gp130, CCL4, TNFα, CAV1, DARC, and SH2D1B) exhibited significant elevation and those of IL-10 and IL-23α significantly decreased in the AMR compared with some of the other 3 groups (P < 0.05). The remaining 12 candidate genes showed no statistically significant variation (P > 0.10) (Figure 3). Among the 6 genes whose mRNA transcript levels exhibited significant elevation in the AMR patient group, only gp130 gene showed significant elevation of mRNA transcripts in the AMR compared with all the other 3 study groups, whereas the other 5 genes showed significant elevation in the AMR above some but not all of the other 3 groups.

Differential levels of mRNA transcripts of 9 genes in plasma exosomes among AMR, CMR, DES control, and non-DES control patient groups. Horizontal long and short lines indicate mean ± standard deviation of results of 18 AMR, 8 CMR, 18 DES Control and 20 non-DES Control patients, respectively. Statistical analysis results by MW test for each gene are indicated on each panel. **P < 0.01, * 0.01 ≤ P < 0.05; # 0.05 ≤ P < 0.10.
Summary of statistical analysis results of all 21 candidate genes
mRNA transcript levels of the other 12 genes in plasma exosomes from AMR, CMR, DES Control, and non-DES Control patient groups. Horizontal long and short lines indicate mean ± standard deviation of results of 18 AMR, 8 CMR, 18 DES Control and 20 non-DES control patients, respectively. P > 0.10 by KW Test.

We next determined if gene scores combining multiple genes would better differentiate patients with AMR from CMR and controls. We found that the gene score combining 4 genes (gp130, SH2D1B, TNFα, and CCL4) was more robust in predicting risk for AMR compared with CMR and controls. In addition, it was also significantly lower in the CMR patients compared to AMR and controls (Figure 4A). The gene score demonstrated good discrimination for AMR, with an area under the ROC curve of 0.796 (Figure 4B). Based on the ROC curve, the best cutoff point of gene score for predicting AMR risk was a score of 1.02 or greater, which yielded a sensitivity of 77.8%, specificity of 76.1%, positive predictive value of 76.5%, and negative predictive value of 77.6%. for AMR.

Gene scores among 4 patient groups and ROC curve of gene score. A, Differential gene score based on the mRNA transcript levels of 4 selected genes (gp130, SH2D1B, TNFα, and CCL4) in plasma exosomes among AMR, CMR, DES Control, and non-DES Control patient groups. Horizontal long and short lines indicate mean ± standard deviation of results of 18 AMR, 8 CMR, 18 DES Control, and 20 non-DES Control patients, respectively. Statistical analysis results by MW test: **P < 0.01, * 0.01 ≤ P < 0.05. B, ROC curve of gene score.

In addition to the analysis described above, we also analyzed the results of all candidate genes and the combined gene score using only 1 plasma sample collected closest to biopsy instead of the average RQ of all plasma samples collected within 1 month before biopsy for each patient in AMR and CMR groups. We found that the final results were very similar between these 2 different analysis for each of these candidate genes and the gene score (Fig. S1 and S2, SDC, The mean days of plasma sample collection before biopsy were 11.6 ± 10.3 for AMR and 4.6 ± 4.4 for CMR, closer to biopsy compared with 13.7 ± 9.9 and 11.4 ± 5.7 when multiple samples were used for the analysis.

Because DSA is a major factor contributing to the development of AMR, we next examined the correlation between the gene score and DSA levels. The AMR patients had significantly higher DSA score than the DES control patients although several patients in the AMR group showed low DSA score, similar to those in DES control patients (Figure 5A). Overall, there is no significant correlation between the gene and DSA scores when AMR and DES control patients are grouped together (R2 = 0.1914) (Figure 5B), and 3 AMR patients with high DSA score showed fairly low gene score. However, 12 (80.0%) of 15 AMR patients showed the gene score of 1.02 or greater, the cutoff level for AMR risk, whereas the gene score in 14 (77.8%) of 18 DES control patients was less than 1.02.

DSA levels and correlation between gene combination score and DSA score in AMR and DES Control patient groups. A, Significantly higher DSA score in AMR than DES control patient group as expected. Horizontal long and short lines indicate mean ± standard deviation of results of 15 AMR (3 patients had no PRA/DSA data available within the selected study timeframe) and 18 DES control patients, respectively. *P < 0.05 by MW test. B, Gene score versus DSA score in AMR and DES Control patient groups. R2, coefficient of determination.


AMR remains one of the major obstacles to the long-term survival of allografts in kidney transplant patients, particularly those with HLA sensitization. Renal biopsy remains the gold standard for diagnosis of allograft rejection including AMR in kidney transplant patients. Using peripheral blood or urine, noninvasive tools capable of evaluating the overall physiologic and immunologic conditions of transplant patients including subclinical allograft rejection would be excellent alternative or supplement to traditional renal biopsy.34 Such attempts have been made for kidney transplant recipients to establish and validate various types of molecular signatures based on proteins, metabolites and mRNA transcript levels of multiple genes.35-40 Most of these research have investigated intracellular total mRNA transcripts extracted from intact cells.

Exosomes are released by most cell types in human body and present abundantly in urine and blood.8,41 Recent studies have proposed that donor-derived extra-cellular vesicles including donor dendritic cell-derived exosomes play crucial roles in allograft-targeting immune response and allorecognition by host,42-44 and a very recently published study also confirmed that donor kidney-specific exosomes could indeed be characterized in the plasma of patients receiving living-donor renal transplant.12 Therefore, the plasma exosomes are likely to contain unique molecular information including not only mRNA transcript profiles associated with allograft injuries and rejection, which may be similar to what can be identified in renal graft biopsy, but also those associated with recipient’s immunological response occurring outside of the allograft, which may not be included in the information obtained from biopsy.

The serum/plasma exosome RNA kit we used in this study uses spin columns to capture plasma exosomes for subsequent RNA extraction and purification. The validity and reliability of exosome RNA extraction from human plasma samples by this kit were well studied and confirmed by a previous study.45 According to the study, both the spin columns of this kit and the conventional ultracentrifugation method for exosome purification capture almost the same vesicle/exosome population with the same vesicle size (peak size of the captured particles/vesicles at 160 and 173 nm, respectively, with the vesicles of expected exosome size at 50 to 200 nm clearly visible under electron microscopy, as captured by both methods) and also the same range of RNA yield (1-10 ng/ml plasma).45,46 Due to the extremely limited availability of the archived plasma samples, we conducted some validation and quality control work for this exosome RNA kit using a small number of plasma samples in this study. Our RNA quality control tests confirmed that the RNA yields of our plasma samples from this spin column kit (average, 8.4 ng/mL plasma) fit perfectly into the same range mentioned above. Meanwhile, the previous study mentioned above also showed that using this spin column exosome RNA kit, high-quality RNA preparations containing intact, full-length mRNA transcripts could be achieved from plasma samples of human blood including patient plasma sample biobanked over a decade.45 Therefore, even though our exosome RNA samples did not return perfect RIN on Bioanalyzer measurement, we believe this did not necessarily indicate poor quality of our exosome RNA samples, considering the unique nature of exosome RNA samples that predominantly consist of miRNA, instead of ribosomal RNA like samples extracted from lymphocytes or tissue samples.

Studies assessing mRNA transcript profiles of multiple genes in exosomes have been conducted for various diseases of the urinary system, including renal allograft rejection.10,11,47 These studies all investigated exosomes from relatively large amount of urine samples per patient (usually 20 mL or more). So far, we have not found studies reported on assessing mRNA transcripts profile related to allograft rejection in kidney transplant patients using blood plasma exosomes. A sizable archive of plasma samples from kidney transplant patients in our group allows us to pursue a study in this direction. For this study we selected 21 candidate genes, which included genes whose mRNA transcript levels were elevated in the renal biopsies of kidney transplant patients with AMR as previously described,4-6 and other genes related to general inflammatory and/or IL-6 signaling events,27-30,48-50 the significant changes of whose expression very likely also contribute to the pathogenesis of AMR.

Among the 21 candidate genes, we identified multiple genes whose mRNA transcript levels exhibited significant changes (mostly elevation) in the plasma exosomes of AMR patients compared with the other groups of kidney transplant patients. These genes include IL-6 β receptor of gp130, chemokine/cytokine genes like CCL4, TNFα, IL-10, and IL-23α, and also other AMR-associated genes, SH2D1B, CAV1, DARC, as identified in previous published study of renal allograft biopsies.4,6 Analysis of our plasma exosome samples detected almost no IL-6 mRNA, and the levels of IL-6Rα were similar in all groups. However, there were consistently higher levels of gp130 mRNA transcripts in the plasma exosomes of AMR patients compared to other groups. A previous study also reported the association of gp130 with posttransplant acute tubular necrosis, a predisposing factor for acute rejection and reduced graft survival in kidney transplant patients.22 Therefore, we cannot rule out a role for IL-6/IL-6R signaling because this is known to be of importance in mediating AMR and DSA development in animal models and likely in human patients.23-26

Because AMR is a highly complex immunologic and pathologic event, we generated a gene score combining the transcript levels of multiple genes (gp130, SH2D1B, TNFα, and CCL4) which was significantly higher in AMR group than CMR and control groups. At the cutoff vaule of 1.02 or greater, this gene score showed good positive and negative predictive values for AMR. Meanwhile, DSA is a major risk factor associated with the development of AMR in kidney transplant patients. High DSA levels (usually SFI > 150 000 in PRA/HLA test) are associated with a risk for AMR, whereas lower levels of DSA (SFI < 150 000) do not necessarily correlate with the development of AMR.51 We did not identify a significant correlation between our gene score and the DSA levels in AMR and DES control patients as some AMR patients with high DSA showed low gene score at biopsy, and there were also 3 AMR patients in our study who did not have detectable DSA when the biopsy diagnosis of AMR was made, while their gene scores were ≥1.02. Since the development of AMR involves multiple immunologic events such as IL-6 signaling, natural killer cell function, inflammation and immune homeostasis, and chemokine/chemotaxis more than just generation of DSA, measuring a comprehensive biomarker with components from various immune functions such as our gene score in addition to monitoring DSA might be more beneficial to predict AMR than measuring DSA alone.

Overall, our current study confirmed the feasibility of quantifying mRNA transcript levels of various genes in plasma exosome of kidney transplant patients. Using our platform, differential mRNA transcriptions of multiple genes were detected in plasma exosomes among AMR, CMR, and control patients, suggesting that exosome mRNA contains information related to events occurred at allograft and/or overall medical and immunologic conditions of kidney transplant recipients. Elevated mRNA level of gp130 in plasma exosomes of AMR patients suggests the involvement of IL-6/gp130 signaling in the development of AMR in some HLA-sensitized patients. More studies are warranted using larger number of samples and covering more comprehensive set of genes to develop a plasma exosome mRNA-based diagnostic tool that can be used to predict allograft rejection for kidney transplant patients.


1. Amore A. Antibody-mediated rejection. Curr Opin Organ Transplant. 2015;20:536–542.
2. Haas M. The revised (2013) Banff Classification for antibody-mediated rejection of renal allografts: update, difficulties, and future considerations. Am J Transplant. 2016;16:1352–1357.
3. Sellares J, Reeve J, Loupy A, et al. Molecular diagnosis of antibody-mediated rejection in human kidney transplants. Am J Transplant. 2013;13:971–983.
4. Venner JM, Hidalgo LG, Famulski KS, et al. The molecular landscape of antibody-mediated kidney transplant rejection: evidence for NK involvement through CD16a Fc receptors. Am J Transplant. 2015;15:1336–1348.
5. Suviolahti E, Ge S, Nast CC, et al. Genes associated with antibody-dependent cell activation are overexpressed in renal biopsies from patients with antibody-mediated rejection. Transpl Immunol. 2015;32:9–17.
6. Dominy KM, Roufosse C, de Kort H, et al. Use of quantitative real time polymerase chain reaction to assess gene transcripts associated with antibody-mediated rejection of kidney transplants. Transplantation. 2015;99:1981–1988.
7. Tkach M, Théry C. Communication by extracellular vesicles: where we are and where we need to go. Cell. 2016;164:1226–1232.
8. Ha D, Yang N, Nadithe V. Exosomes as therapeutic drug carriers and delivery vehicles across biological membranes: current perspectives and future challenges. Acta Pharm Sin B. 2016;6:287–296.
9. Franzen CA, Blackwell RH, Foreman KE, et al. Urinary exosomes: the potential for biomarker utility, intercellular signaling and therapeutics in urological malignancy. J Urol. 2016;195:1331–1339.
10. Motamedinia P, Scott AN, Bate KL, et al. Urine exosomes for non-invasive assessment of gene expression and mutations of prostate cancer. PLoS One. 2016;11:e0154507.
11. Peake PW, Pianta TJ, Succar L, et al. A comparison of the ability of levels of urinary biomarker proteins and exosomal mRNA to predict outcomes after renal transplantation. PLoS One. 2014;9:e98644.
12. Vallabhajosyula P, Korutla L, Habertheuer A, et al. Tissue-specific exosome biomarkers for noninvasively monitoring immunologic rejection of transplanted tissue. J Clin Invest. 2017;127:1375–1391.
13. Sis B, Mengel M, Haas M, et al. Banff '09 meeting report: antibody mediated graft deterioration and implementation of Banff working groups. Am J Transplant. 2010;10:464–471.
14. Kahwaji J, Tong C, Jordan S, et al. Rituximab: an emerging therapeutic agent for kidney transplantation. Transplant Res Risk Management. 2009;1:15–29.
15. Vo AA, Lukovsky M, Toyoda M, et al. Rituximab and intravenous immune globulin for desensitization during renal transplantation. N Engl J Med. 2008;359:242–251.
16. Vo AA, Choi J, Cisneros K, et al. Benefits of rituximab combined with intravenous immunoglobulin for desensitization in kidney transplant recipients. Transplantation. 2014;98:312–319.
17. Barbosa D, Kahwaji J, Puliyanda D, et al. Polyomavirus BK viremia in kidney transplant recipients after desensitization with IVIG and rituximab. Transplantation. 2014;97:755–761.
18. Kahwaji J, Sinha A, Toyoda M, et al. Infectious complications in kidney-transplant recipients desensitized with rituximab and intravenous immunoglobulin. Clin J Am Soc Nephrol. 2011;6:2894–2900.
19. Hirohashi T, Uehara S, Chase CM, et al. Complement independent antibody-mediated endarteritis and transplant arteriopathy in mice. Am J Transplant. 2010;10:510–517.
20. Hirohashi T, Chase CM, Della Pelle P, et al. A novel pathway of chronic allograft rejection mediated by NK cells and alloantibody. Am J Transplant. 2012;12:313–321.
21. Lee CY, Lotfi-Emran S, Erdinc M, et al. The involvement of FcR mechanisms in antibody-mediated rejection. Transplantation. 2007;84:1324–1334.
22. Sadeghi M, Daniel V, Lahdou I, et al. Association of pretransplant soluble glycoprotein 130 (sgp130) plasma levels and posttransplant acute tubular necrosis in renal transplant recipients. Transplantation. 2009;88:266–271.
23. Liang Y, Christopher K, Finn PW, et al. Graft produced interleukin-6 functions as a danger signal and promotes rejection after transplantation. Transplantation. 2007;84:771–777.
24. Kim I, Wu G, Chai NN, et al. Anti-interleukin 6 receptor antibodies attenuate antibody recall responses in a mouse model of allosensitization. Transplantation. 2014;98:1262–1270.
25. Lion J, Taflin C, Cross AR, et al. HLA class II antibody activation of endothelial cells promotes Th17 and disrupts regulatory T lymphocyte expansion. Am J Transplant. 2016;16:1408–1420.
26. Pabois A, Pagie S, Gérard N, et al. Notch signaling mediates crosstalk between endothelial cells and macrophages via Dll4 and IL6 in cardiac microvascular inflammation. Biochem Pharmacol. 2016;104:95–107.
27. Ma L, Zhang H, Hu K, et al. The imbalance between Tregs, Th17 cells and inflammatory cytokines among renal transplant recipients. BMC Immunol. 2015;16:56.
28. Sinuani I, Beberashvili I, Averbukh Z, et al. Role of IL-10 in the progression of kidney disease. World J Transplant. 2013;3:91–98.
29. Kalliolias GD, Ivashkiv LB. TNF biology, pathogenic mechanisms and emerging therapeutic strategies. Nat Rev Rheumatol. 2016;12:49–62.
30. Gaffen SL, Jain R, Garg AV, et al. The IL-23-IL-17 immune axis: from mechanisms to therapeutic testing. Nat Rev Immunol. 2014;14:585–600.
31. Goriely S, Goldman M. Interleukin-12 family members and the balance between rejection and tolerance. Curr Opin Organ Transplant. 2008;13:4–9.
32. Reinsmoen NL, Lai CH, Vo A, et al. Acceptable donor-specific antibody levels allowing for successful deceased and living donor kidney transplantation after desensitization therapy. Transplantation. 2008;86:820–825.
33. Vo AA, Choi J, Kim I, et al. A phase I/II trial of the interleukin-6 receptor-specific humanized monoclonal (tocilizumab) + intravenous immunoglobulin in difficult to desensitize patients. Transplantation. 2015;99:2356–2363.
34. Gwinner W, Metzger J, Husi H, et al. Proteomics for rejection diagnosis in renal transplant patients: where are we now? World J Transplant. 2016;6:28–41.
35. Sigdel TK, Gao Y, He J, et al. Mining the human urine proteome for monitoring renal transplant injury. Kidney Int. 2016;89:1244–1252.
36. Suhre K, Schwartz JE, Sharma VK, et al. Urine metabolite profiles predictive of human kidney allograft status. J Am Soc Nephrol. 2016;27:626–636.
37. Li L, Khatri P, Sigdel TK, et al. A peripheral blood diagnostic test for acute rejection in renal transplantation. Am J Transplant. 2012;12:2710–2718.
38. Li L, Khush K, Hsieh SC, et al. Identification of common blood gene signatures for the diagnosis of renal and cardiac acute allograft rejection. PLoS One. 2013;8:e82153.
39. Matignon M, Ding R, Dadhania DM, et al. Urinary cell mRNA profiles and differential diagnosis of acute kidney graft dysfunction. J Am Soc Nephrol. 2014;25:1586–1597.
40. Sigdel TK, Ng YW, Lee S, et al. Perturbations in the urinary exosome in transplant rejection. Front Med (Lausanne). 2015;1:57.
41. Robbins PD, Morelli AE. Regulation of immune responses by extracellular vesicles. Nat Rev Immunol. 2014;14:195–208.
42. Robbins PD, Dorronsoro A, Booker CN. Regulation of chronic inflammatory and immune processes by extracellular vesicles. J Clin Invest. 2016;126:1173–1180.
43. Liu Q, Rojas-Canales DM, Divito SJ, et al. Donor dendritic cell-derived exosomes promote allograft-targeting immune response. J Clin Invest. 2016;126:2805–2820.
44. Morelli AE, Bracamonte-Baran W, Burlingham WJ. Donor-derived exosomes: the trick behind the semidirect pathway of allorecognition. Curr Opin Organ Transplant. 2017;22:46–54.
45. Enderle D, Spiel A, Coticchia CM, et al. Characterization of RNA from exosomes and other extracellular vesicles isolated by a novel spin column-based method. PLoS One. 2015;10:e0136133.
46. Noerholm M, Balaj L, Limperg T, et al. RNA expression patterns in serum microvesicles from patients with glioblastoma multiforme and controls. BMC Cancer. 2012;12:22.
47. Hu L, Skog J, Hurley J, et al. Discovery of a urinary exosome mRNA signature for the diagnosis of human kidney transplant rejection. J Extracell Vesicles. 2016;5:31552.
48. Hunter CA, Jones SA. IL-6 as a keystone cytokine in health and disease. Nat Immunol. 2015;16:448–457.
49. Scheller J, Garbers C, Rose-John S. Interleukin-6: from basic biology to selective blockade of pro-inflammatory activities. Semin Immunol. 2014;26:2–12.
50. Taniguchi K, Karin M. IL-6 and related cytokines as the critical lynchpins between inflammation and cancer. Semin Immunol. 2014;26:54–74.
51. Jordan SC, Reinsmoen N, Peng A, et al. Advances in diagnosing and managing antibody-mediated rejection. Pediatr Nephrol. 2010;25:2035–2045; quiz 2045–8.

Supplemental Digital Content

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.