Tissue-associated self-antigens containing exosomes: Role in allograft rejection

Abstract

Exosomes are extracellular vesicles that express self-antigens (SAgs) and donor human leukocyte antigens. Tissue-specific exosomes can be detected in the circulation following lung, heart, kidney and islet cell transplantations. We collected serum samples from patients who had undergone lung (n=30), heart (n=8), or kidney (n=15) transplantations to isolate circulating exosomes. Exosome purity was analyzed by Western blot, using CD9 exosome-specific markers. Tissue-associated lung SAgs, collagen V (Col-V) and K-alpha 1 tubulin (Kα1T), heart SAgs, myosin and vimentin, and kidney SAgs, fibronectin and collagen IV (Col-IV), were identified using western blot. Lung transplant recipients diagnosed with bronchiolitis obliterans syndrome had exosomes with higher expression of Col-V (4.2 fold) and Kα1T (37.1fold) than stable. Exosomes isolated from heart transplant recipients diagnosed with coronary artery vasculopathy had a 3.9 fold increase in myosin and a 4.7fold increase in vimentin compared with stable. Further, Kidney transplant recipients diagnosed with transplant glomerulopathy had circulating exosomes with a 2 fold increased expression of fibronectin and 2.5 fold increase in Col-IV compared with stable. We conclude that circulating exosomes with tissue associated SAgs have the potential to be a noninvasive biomarker for allograft rejection.

1. Introduction

Exosomes are small vesicles (40–100nm in diameter) generated by reverse budding of early endosomes (ie, multi-vesicular bodies) and secreted by fusion with the cell membrane [1]. All mammalian cells release exosomes; they can be found in vivo in most body fluids, including plasma, urine, and saliva. The cargo within an exosome depends upon the cell from which it originated. One function of exosomes is releasing unwanted cell material; however, they can also transfer proteins, mRNAs, and micro RNAs (miRNAs) to the surrounding milieu. Exosomes from cancer cells have been shown to induce angiogenesis, invasion, and metastasis; they have even been shown to confer drug resistance [2]. After allogeneic heart and skin transplantation, donor-derived exosomes containing donor major histocompatibility complex can cross dress recipient antigen-presenting cells, leading to activation and proliferation of alloreactive T cells via the semi-direct pathway of allorecognition [3, 4].

Recent reports from our laboratory [5] and others [6] demonstrated exosomes in the sera and bronchoalveolar lavage (BAL) fluid of human lung transplant recipients (LTxRs) diagnosed with acute and chronic rejection with distinct antigenic properties and presence of mRNA and miRNA. The exosomes isolated from human LTxRs diagnosed with rejection contained not only donor-mismatched human leukocyte antigen (HLA), but also the lung-restricted self-antigens (SAgs) collagen type V (Col-V) and K-alpha 1 tubulin (Kα1T), indicating that exosomes are secreted by transplanted lungs [5]. Another study reported the presence of tissue-specific exosomes following islet cell transplantation, and suggested that exosomes may be biomarkers that allow noninvasive monitoring of immunologic rejection of islet tissues post-transplant [7].

Studies from our laboratory and others have demonstrated immune responses to tissue-restricted antigens in transplant recipients diagnosed with transplant glomerulopathy (TG) [8], chronic rejection following human kidney transplant [9], and coronary artery vasculopathy (CAV) [10, 11] after heart transplantation (HTx). Therefore, we hypothesized that tissue-restricted SAgs expressing exosomes may be present not only in the circulation during or prior to rejection of human lung and islet transplantations but also in the circulation during or prior to rejection following renal (RTx) and HTx. In this communication, we demonstrate the presence of exosomes containing Kα1T and Col-V in the sera of human LTxRs diagnosed with bronchiolitis obliterans syndrome (BOS), Col-IV and fibronectin (FN) in renal transplant recipients (RTxRs) diagnosed with TG, and cardiac myosin (Myo) and vimentin (Vim) in HTx recipients (HTxRs) diagnosed with CAV.

2. Methods

2.1. Sample collection

This study was approved by the Institutional Review Boards at Washington University School of Medicine and St. Joseph’s Hospital. All patients provided informed consent. Peripheral blood samples were collected from patients who underwent LTx, HTx, or KTx. For this preliminary study, 30 LTxRs, 8 HTxRs, and 15 RTxRs were available. Of the 20 LTxRs, 10 were clinically diagnosed with BOS, and 10 were in stable condition with functioning allografts and 5 patients developed DSA and 5 patients without DSA considered as stable. Of the 8 HTxRs, 5 were diagnosed with CAV; 3 were stable with no evidence of rejection. Nine of the RTxRs had biopsy-proven TG as described by Angaswamy et al [8], and the remaining 6 RTxRs had a biopsy that confirmed the absence of rejection pathology and a well-functioning kidney (Table 1).

Table 1.

Demographic Data of Patients

	Chronic rejection	Stable
Lung
Total No (N=30)	20	10
Age (Years)	53.3±15.1	51.4±15.8
Male	8 (80%)	5 (50%)
female	2 (20 %)	5 (50%)
End stage disease
COPD	4 (40%)	4 (4%)
IPF	3 (30%)	4 (40%)
CF	3 (30%)	2 (20%)
DSA+	5
DSA−		5
Heart
Total No (N=8)	5	3
Age (Years)	52.5±36.7	56.6±19.7
Male	4 (90%)	3 (100%)
female	1 (10%)	0
Etiology
CAV	5 (100%)	3 (100%)
Kidney
Total No (N=15)	9	6
Age (Years)	41.7±18.1	49.1±21.2
Male	8(90%)
female	1(10%)	6(100%)
Cause of ESRD
Diabetes	3	1
PKD	3	2
Reflux/obstruction	1	1
FSGS	2	2

Data Represented as mean±stddev or %. COPD:Chronic Obstructive Pulmonary Disease; IPF: idiopathic pulmonary fibrosis CF: cystic fibrosis; DSA: Donor specific antibodies; CAV: cardiac allograft vasculopathy; ESRD: end stage renal disease, FSGS, focal segmental glomerulosclerosis; PKD, polycystic kidney disease

2.2. Exosome isolation

Circulating exosomes were isolated from serum samples (200μl) using ultracentrifugation followed by sucrose cushion. Purity was confirmed by electron microscopy and CD9 staining was performed as we have described previously [5]. In brief, serum was centrifuged at 2,000g followed by 10,000g for 30 mins at 4°C to remove cell debris. Following this, serum was diluted with PBS and centrifuged to 100,000g for 70 mins at 4°C to isolate exosomes. Exosome pellets were suspended in phosphate buffered saline and exosome concentration was analyzed using the bicinchoninic acid method. Calnexin is used as a negative control [12]. All of the exosomes used in this study did not contain calnexin.

2.3. Tissue-associated Sags

Circulating exosomes were isolated from human LTxRs and analyzed for the presence of the lung-associated SAg, Kα1T, which is a gap junction protein to which antibodies (Abs) develop during rejection; and for the presence of Col-V, which is a sequestrated antigen normally intercalated by Col-I and exposed during rejection after LTx [13]. We selected two self-proteins to which immune responses have been shown to develop by us [10] and by others [14, 15] during HTx rejection: cardiac Myo and Vim. For RTxRs, we selected two renal-restricted SAgs: Col-IV, a renal tubular basement membrane protein [16], and FN, to which immune responses have been shown to develop in RTxRs diagnosed with TG [8].

2.4. Western blot analysis

To determine the presence of tissue-associated SAgs in exosomes, we performed Western blot analysis using specific Abs to Col-V, Kα1T, cardiac Myo, Vim, Col-IV, and FN. Ten μg of proteins (lung, heart, and kidney) were used for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene difluoride (PVDF) membrane. To detect protein in the isolated exosomes, we used primary Abs to lung-associated SAgs: anti-rabbit Col-V (Abcam, ab7046) and anti-mouse Kα1T (Santa Cruz, 12462-r) immunoglobulin G (IgG); Abs to cardiac-associated SAgs: anti-mouse-Myo (Abcam, ab15–100) and anti-mouse Vim (BD Pharmingen, 550513); Abs to kidney-associated SAgs: FN (Sigma, F3648) and anti-Col-IV (Abcam, ab-6586). Goat anti-rabbit/mouse IgG (Cell Signaling, 7074S; Jackson, 115-035-062) conjugated with horse peroxidase was used as secondary Abs. Blots were developed using enhanced chemiluminescent immunoblot detection kit. J Image Software (NIH) was used for densitometry analysis and semi-quantitation of the signal.

2.5. Statistical Data analysis

Graph Pad Prism 6 (GraphPad Software, Inc, CA) was used to perform data analysis. Optical density of exosomes containing lung, cardiac and kidney SAgs as well as for the differences between BOS and stable LTxRs, CAV and stable HTxRs, TG and stable RTxRs were compared using Mann-Whitney or two-tailed student’s t-test. Statistical data in each cohort was expressed as mean±standard deviation. P-values less than 0.05 were considered statistically significant in each comparative analysis. The fold changes were calculated after normalization of mean optical density of exosomes containing SAgs with CD-9.

3. Results

3.1. Circulating exosomes isolated from human LTxRs diagnosed with rejection contain the lung SAgs Col-V and Kα1T

To determine whether circulating exosomes were induced in LTxRs diagnosed with chronic rejection (ie, BOS), we isolated exosomes from the sera collected from 10 LTxRs at the time of BOS diagnosis. As shown in Figure 1A, Western blot analysis demonstrated that all 10 LTxRs who were diagnosed with BOS had significantly increased levels of exosomes containing Col-V (p=0.0024) and Kα1T (p=0.0002) compared with stable LTxRs. CD9 was used as the loading marker for the amount of exosomes. Semi-quantitation by densitometry for both Col-V and Kα1T clearly demonstrated significantly increased levels of the lung tissue-restricted SAgs, Col-V and Kα1T, in LTxRs with BOS compared to stable LTxRs (Fig. 1B). Figure 1C shows that the lung SAgs present in the exosomes are specific to LTxRs diagnosed with BOS, since the exosomes contained no kidney-associated SAg (ie, Col-IV).

Figure 1. Expression of lung SAgs in exosomes.

Exosomes were isolated from sera of patients diagnosed with BOS. 10 μg of protein was used for SDS-PAGE and transferred to a PVDF membrane. Primary Abs to specific SAgs was used to detect SAg expression in exosomes. A) Western blot analysis: P: patient and S: stable. CD9, an exosome marker used as loading control, (B) Densitometry analysis: densitometry analysis was carried out using J image software (NIH), (C) Immunoblot: Col-IV, kidney-specific SAg used as a negative control.

3.2. Exosomes from LTxRs with de novo donor-specific antibodies (DSA) have higher expression of exosomes with lung SAgs

To determine the role of DSA in induction of circulating exosomes, we selected 5 LTxRs who developed DSA after transplant and 5 patients without DSA. Serum samples were collected for exosome isolation and characterization. Exosomes were isolated from the sera from both groups and assessed for expression of lung SAgs. As shown in Figure 2A, DSA+ patients had more lung SAgs than DSA− patients. Densitometry analysis showed a 5.2-fold increase in Col-V expression (p=0.0329) and 3-fold increase in Kα1T expression (p=0.0101) in DSA+ patients compared with DSA− stable LTxRs (Fig. 2B).

Figure 2. Higher expression of lung SAgs in exosomes isolated from DSA+ LTxRs.

Exosomes were isolated from LTxRs who developed de novo DSA (n=5) and these exosomes were subjected to immunoblot. A) Western blot analysis: specific Abs to Col-V and Kα1T were used to detect expression of lung SAgs in exosomes. CD9 was used as loading control. B) Densitometry analysis: Image J is used to determine the specific band intensity and fold change of protein was calculated.

3.3. Exosomes containing cardiac SAgs Myo and Vim in the circulation of HTxRs diagnosed with CAV

We collected peripheral blood from 5 HTxRs diagnosed with CAV and 3 time-matched stable HTxRs without evidence of rejection. Circulating exosomes isolated from serum were tested for the cardiac SAgs, Myo and Vim. As shown in Figure 3A, significantly higher levels of the cardiac SAgs, Myo (p=0.0002) and Vim (p=0.0211), were present in exosomes isolated from HTxRs with CAV compared with stable HTxRs. Densitometry analysis (Fig. 3B) shows a 3.9-fold increase in Myo and a 4.7-fold increase in Vim in HTxRs diagnosed with CAV compared with stable HTxRs. These results demonstrate that exosomes containing cardiac SAgs were present in the sera of HTxRs undergoing chronic rejection. Specificity for tissue-associated SAgs was evident, as the exosomes isolated from HTxRs diagnosed with CAV did not contain the kidney SAg, Col-IV (Fig. 3C).

Figure 3. Exosomes contain cardiac SAgs.

Exosomes were isolated from sera of patients using ultracentrifugation. 10 μg of protein was used for SDS-PAGE and transferred to a PVDF membrane. Primary Abs to specific to SAgs were used to detect expression of SAgs in exosomes. A) Western blot analysis: P: patient and S: stable. CD9, an exosome marker used as loading control, (B) Densitometry analysis: densitometry analysis was carried out using J image software (NIH), (C) Immunoblot: Col-IV, kidney-specific SAg used as a negative control.

3.4. Circulating exosomes with FN and Col-IV in RTxRs diagnosed with TG

We demonstrated de novo development of Abs to renal-associated SAgs, Col-IV and FN, in serum samples collected from RTxRs who had biopsy-proven TG (independent of development of Abs to mismatched donor HLA) [8]. For this retrospective preliminary analysis, we obtained serum from nine RTxRs diagnosed with TG and six biopsy-proven stable RTxR for exosome isolation. The kidney-associated SAgs, FN (p=0.0156) and Col-IV (p=0.0349), were seen in samples isolated from patients diagnosed with TG, but not from stable RTxRs (Fig. 4A). Densitometry analysis presented in Figure 4B demonstrates a 2-fold increase in the expression of FN and a nearly 2.5-fold increase in Col-IV in RTxRs diagnosed with TG compared with stable RTxRs. These results, although obtained from a relatively small number of RTxRs, support our conclusion that circulating exosomes with renal tissue-associated SAgs are present in patients with TG, a group of RTxRs at increased risk for chronic rejection. Specificity for kidney-associated SAgs was apparent, as the exosomes isolated from RTxRs with TG did not contain the lung SAg, Col-V (Fig. 4C).

Figure 4. Presence of kidney SAgs in exosomes.

(A) Western immunoblot for kidney-associated SAgs Col-IV and FN and for exosome-specific marker CD-9 from exosomes isolated from serum samples of stable KTxRs and from KTxRs diagnosed with TG. CD9 used as loading control, B) Densitometry analysis: densitometry analysis was carried out using J image software (NIH), (C) Immunoblot: Col-V, lung-specific SAg used as a negative control.

4. Discussion

Data presented in this article demonstrate that exosomes are released into the circulation during allograft rejection, and that these exosomes contain tissue associated SAgs after lung, heart, and kidney transplant. Previous reports from our laboratory have shown that Abs to HLA and/or tissue-associated SAgs develop prior to clinical diagnosis of lung allograft rejection [17, 18]. Furthermore, a recent report from our laboratory demonstrated circulating exosomes containing donor HLA as well as Kα1T and Col-V in LTxRs diagnosed with acute rejection and chronic rejection (ie, BOS) but not in stable LTxRs. The exosomes isolated from LTxRs experiencing allograft rejection also contained immunoregulatory miRNA, which suggests that circulating exosomes may be involved in allograft rejection [19, 20]. However, the lungs pose a unique challenge in the transplant field due to their direct exposure to outside milieu, and immune responses post-LTx may, therefore, not be applicable to other solid organs (eg, the heart or kidneys), as those organs are not exposed to the outside milieu.

In this communication, we show that circulating exosomes containing SAgs are present in the circulation not only after LTx, but also in HTxRs and RTxRs diagnosed with rejection. Therefore, immune responses leading to chronic rejection can occur in response not only to mismatched donor HLA, but also in response to tissue-associated SAgs (ie, Kα1T and Col-V) after LTx, cardiac Myo and Vim after HTx, and Col-IV and FN after RTx.

Studies from our group and others have demonstrated that exosomes isolated from patients undergoing chronic rejection demonstrated various SAgs, immunoregulatory miRNA, and mRNA involved in immune cascade [6, 21–25]. Recent findings from our group demonstrated that exosomes isolated from sera and BAL fluid of LTxR diagnosed with chronic rejection contains mismatched donor HLA and the lung-associated SAgs, Col-V and Kα1T, stable LTxRs did not have these same features [5]. In addition to our publication, others have reported exosomes present in BAL fluid and identified an association between these exosomes and the development of immune responses to transplanted lungs [6]. Further, we demonstrated that circulating exosomes from LTxRs diagnosed with BOS has distinct properties in comparison to stable [26].

Kennel et al [27] showed that serum exosomal proteins are different in acute cardiac allograft rejection when compared with healthy controls, patients with heart failure, and stable HTxRs. Recent data from our laboratory showed that administration of exosomes containing SAgs caused graft failure in a murine syngeneic graft [28]. We have also demonstrated that exosomes from HTxRs diagnosed with CAV have higher expression of the cardiac-associated SAgs, Myo and Vim, compared with findings in patients not diagnosed with CAV [29].

Several groups have reported isolating circulating exosomes from blood and urine samples from patients diagnosed with kidney disease. These exosomes contained RNA and protein that may be potential biomarkers for renal disease [30–32]. A recent report demonstrated that antibody-mediated acute rejection in RTxRs is correlated with increase in C4d⁺ plasma endothelial microvesicles, and the authors suggested that this may be a marker for antibody-mediated rejection [33]. This suggests that the development of DSA, which can bind to the transplanted organ, can indeed induce exosome release, perpetuating further immune activation. Further supporting this theory, Abs binding to HLA class-1, or lung SAgs, induced exosomes releases in vitro from airway epithelial cells which contains lung SAgs [34].

Another earlier report analyzing urinary exosomes from patients with acute renal allograft rejection demonstrated altered protein contents, suggesting that exosomes in urine with an altered protein profile can indicate rejection [35, 36]. Urinary CD133+ (stem cell marker) extracellular vesicles was found to be increased in patients with slow graft function in RTx and contained glomerular and proximal tubular markers [37]. In addition to post-transplant exosome analysis, many reports demonstrate circulating exosomes in patients with various autoimmune diseases and cancers, suggesting that exosomes may play an important role in the disease process, cancer biology, and metastasis [38–40].

Therefore, based on the in vitro data using airway epithelial cells we conclude that both alloantibodies and Abs to SAgs can induce exosomes. Exosomes released by allografts may contribute to augmentation of the host’s immune responses against the transplanted organ leading to their rejection. Recent studies have shown that recipient antigen presenting cells displaying allogeneic MHC molecules acquired from donor exosomes (MHC cross-dressing) can also play a key role in the development of immune response towards allograft rejection [41]. The results presented in this manuscript demonstrate that circulating exosomes induced after lung, heart, or kidney transplant may play an important role in allograft rejection, because the exosomes were shown to contain tissue associated SAgs. The exosomes carry tissue-associated SAgs, indicating that the circulating exosomes have originated from the transplanted organs. Because exosomes are released by endothelial cells during antibody-mediated rejection [33] it is likely that exosomes can be induced by Abs capable of binding to the transplanted organs.

Limitation of this report is that it is a cross sectional analysis and does not provide results of when the circulating exosomes are demonstrable prior to rejection of the organs. Another limitation is the small number of HTxRs and RTxRs analyzed which need further confirmation. In addition, several important questions regarding the immunobiology of exosomes released into the circulation after transplantation remain unanswered. For example, the kinetics of exosome release (ie, is it before Ab development? or are Abs and cellular immune responses needed for the release of exosomes?) is currently unknown. The half-life of exosomes present in the circulation with donor HLA and tissue-associated SAgs are also unclear. However, it is evident that exosome biology will play an important role in allograft immunity, as exosomes carry donor HLA, tissue-associated SAgs, and various immunoregulatory molecules.

In conclusion, our results demonstrate that circulating exosomes isolated from lung, heart and kidney transplant recipients demonstrate unique properties compared to stable transplant recipients with respect to the presence of tissue associated SAgs (Col-V and Kα1T following LTx), (cardiac Myo and Vim following HTx) and (Col-IV and FN following KTx). We propose that the induction and release of circulating exosomes following transplantation will play an important role in inducting and perpetuating immune responses against allo and tissue restricted SAgs leading to the immune-pathogenesis of chronic rejection following solid organ transplantation.

Abbreviations

Ab

antibody

BAL

bronchoalveolar lavage

BOS

bronchiolitis obliterans syndrome

CAV

coronary artery vasculopathy

Col-V

collagen type V

DSA

donor-specific antibodies

FN

fibronectin

HLA

human leukocyte antigen

HTx

heart transplant

HTxR

heart transplant recipient

IgG

immunoglobulin G

Kα1T

K-alpha 1 tubulin

LTxR

lung transplant recipient

miRNA

micro RNA

Myo

myosin

PVDF

polyvinylidene difluoride

RTx

renal transplant

RTxR

renal transplant recipient

SAg

self-antigen

SDS-PAGE

sodium dodecyl sulfate polyacrylamide gel electrophoresis

TG

transplant glomerulopathy

Vim

vimentin