Transplant (Tx)-associated coronary artery disease (CAD) is the most common and serious long-term complication after heart transplantation, with an incidence of 40% at 5 years (1). The condition is characterized by the diffuse and concentric narrowing of the arteries and veins in the donor heart (2). TxCAD is a multifactorial disease where traditional risk factors for atherosclerosis interact with an immune response against the donor organ. Experimental models (3) and clinical studies have suggested a role for antibodies to both human leukocyte antigen (HLA) (4) and non-HLA antigens (5,6) in the pathogenesis of TxCAD.
Western blotting and flow cytometry have revealed that sera from patients with TxCAD contain antibodies to a number of different endothelial antigens (5,6), but the nature of the antigens remains undetermined in most cases. Some of these antigenic targets have been identified by two-dimensional electrophoresis, and these include vimentin and heat shock protein 60 (7). More recently, cDNA expression library screening (8) has been used to identify further endothelial cell antigens that may be involved in the pathogenesis of TxCAD. To date, there appears to be no commonality between antigens discovered in the two different studies (7,8).
The previous immunoscreening study was limited by the source of the cDNA library (human umbilical vein endothelial cells) and by the number of patients’ sera screened (two) (8). The present study has attempted to overcome these limitations by screening a more physiologically relevant source of endothelial cells, namely, a commercially available human coronary artery endothelial cell (hCAEC) library and a wider range of sera from TxCAD patients and normal subjects. The aim of the study was to establish the diversity and nature of antigens targeted by sera from patients with TxCAD.
MATERIALS AND METHODS
Sera were collected retrospectively from seven cardiac allograft recipients (age range, 20–61 years) who underwent transplantation between 1989 and 2001 at Harefield Hospital, Middlesex, United Kingdom. The sera were from patients diagnosed as having angiographically detectable TxCAD (defined as at least 50% stenosis of one or more coronary arteries) and were known to have antivimentin antibodies, as previously described (9). Normal sera were obtained from five healthy volunteers (age range, 20–60 years). Of the cardiac allograft recipients, those that were tested for anti-HLA antibodies were negative. Clinical data relevant to the patients’ sera is shown in Table 1. Sera were collected according to guidelines provided by the local ethical committee.
A commercially available human coronary artery endothelial cell lambda cDNA expression library (Uni-ZAP XR; Stratagene, Cedar Creek, TX) was used to screen for clones showing reactivity with sera. Before screening against the library, sera were diluted 1:200 in phosphate-buffered saline (PBS), 10% goat serum, 5% Marvel, and 0.02% sodium azide, and preadsorbed against nonrecombinant plaque lifts to reduce binding to Escherichia coli and vector phage. Nonrecombinant library phage were plated out at approximately 5×106 plaque-forming units (PFU)/150-mm plate (five plates per serum sample) using XL-1 Blue MRF’ host cells according to the supplier’s instructions (Stratagene, La Jolla, CA). Nitrocellulose membranes (Stratagene) were soaked in 10 mM IPTG for 10 min and air-dried. Membranes were then overlaid onto the plates and incubated at 37°C (98.6°F) overnight. The membranes were carefully removed and washed in PBS and 0.05% Tween 20 (T) for 10 min with gentle agitation three times to remove adherent agar. Nonspecific binding was blocked by incubating in PBS, 10% goat serum, and 5% Marvel at room temperature for 1 hr with gentle agitation. The membranes were washed in PBS-T for 10 min with gentle agitation three times and then incubated with diluted sera at 4°C (39.2°F) overnight with gentle agitation. The sera was then removed and stored at 4°C between library screenings.
Primary library screening was carried out according to the supplier’s instructions. Briefly, library phage was plated out at approximately 5×104 PFU/150-mm plate (in total, ≈2.5×105 PFU/serum sample), and IPTG-soaked nitrocellulose filters were overlaid overnight (see above). Before lifting the membranes, their orientation was marked using a sterile needle. Membranes were washed and blocked as above and then incubated with diluted sera at room temperature for 1 hr with gentle agitation. Sera were retained and the membranes washed in PBS-T for 10 min with gentle agitation three times. Bound antibody was detected using horseradish peroxidase conjugated goat anti-human polyvalent immunoglobulins (Ig) (IgG fraction) (Sigma, Poole, United Kingdom) diluted 1:1,000 in PBS, 10% goat serum, and 5% Marvel at room temperature for 1 hr with gentle agitation. After three final washes in PBS-T for 10 min with gentle agitation, the membranes were incubated with 4CN Plus substrate (Perkin Elmer Life Sciences, Zaventem, Belgium).
Positive immunoreactive plaques were easily distinguishable from blue nonrecombinant plaques and background staining. Areas of the plate corresponding to the immunoreactive plaques were cored, resuspended in 500 μL SM buffer containing 20 μL chloroform, and stored at 4°C overnight before retransfection of the host cells. Host cells were transfected with phage isolated from positive plaques and plated at between 1,000 and 200 PFU/150-mm plate and lifts taken as above. Two to three further rounds of screening were required to screen plaques to clonality.
Subcloning and Identification of Positive Inserts
Cloned phages were recovered as pBluescript phagemid by in vivo excision using the ExAssist helper phage (Stratagene) according to the supplier’s instructions. The rescued phagemid was then used to transform SOLR cells. Positive recombinants were picked and grown overnight at 37°C in 5 mL LB/50 μg/mL ampicillin. An aliquot of this culture was removed to create a bacterial stock and the plasmid DNA was isolated from the remainder using the WizardSV Minipreps DNA purification system (Promega, Southampton, United Kingdom). Presence of insert DNA was determined by restriction digest using EcoRI and XhoI followed by agarose gel electrophoresis of the restricted and unrestricted DNA. DNA sequencing was conducted at the Molecular Biology Unit, King’s College London. Insert DNA sequence was submitted to a FASTA search (v3.3t09 May 18, 2001) at Embl.
Sera from 12 subjects were immunoscreened against the hCAEC cDNA library, seven heart transplant recipients diagnosed with TxCAD and five healthy volunteers. The sera from the patients displayed high titers of antivimentin antibodies (Table 1), a marker previously reported to be associated with TxCAD (9). Five of seven TxCAD sera screened identified plaques that remained positive through three rounds of screening (Table 2). Intriguingly, two of five normal sera also identified plaques that remained positive through three rounds of screening. The identity of the clones was established by DNA sequencing and DNA database searching (Table 2). The exception was cDNA sequence obtained from a cDNA clone identified in patient 5. This cDNA sequence did not correspond to any cDNA sequence in the DNA databases, and the authors have designated it as a novel cDNA sequence (Table 2).
A diverse range of cDNA clones was identified, with the spectrum for each patient being unique (Table 2). Most of the antigens identified in patients are intracellular proteins with the notable exception of neuropilin (np) 2, which was identified as a putative antigen in TxCAD patient 2 (Table 2); np2 and np1 both function as receptors for the axon guidance factors belonging to the class 3 semaphorin family but also bind to certain heparin-binding splice forms of vascular endothelial growth factor (VEGF) (10).
It was noted that sera from the two patients (patients 2 and 4) with the lowest antivimentin titers failed to identify any reactive clones (Tables 1 and 2), perhaps reflecting the fact that antivimentin titer is a good general indicator of immunoreactivity or TxCAD, or both, as has been suggested from other studies (9).
An hCAEC cDNA library was probed with sera from seven cardiac allograft recipients to identify antigens specific to TxCAD sufferers. The hCAEC cDNA library was also probed with sera from healthy volunteers. In this study, 13 previously unidentified antigens were recognized by TxCAD sera. All of the antigens recognized were unique for each patient and there was no overlap of reactivity between different patients. There was also no overlap between the antigens isolated by TxCAD sera and those isolated by the healthy sera. This study therefore did not identify a common or dominant putative antigen found in TxCAD. This is consistent with the authors’ other study on TxCAD using a molecular screening strategy (8) in which unique autoreactivities were described.
The question arises of why cardiac transplant patients make antibodies to a variety of non-HLA antigens. It is known that an immune response to donor HLA antigens predominates in the first year after transplantation, and such a response is because of direct recognition of donor HLA antigens presented by graft-derived dendritic cells (11). As time progresses, donor-derived dendritic cells disappear and the immune response is predominantly against donor antigens presented indirectly by recipient dendritic cells (11). Studies so far have demonstrated that indirect recognition of the hypervariable region of HLA is associated with chronic rejection (12–14). In contrast, this study demonstrates a response to multiple non-HLA antigens. It is not known at present whether these antigens are polymorphic non-HLA or whether they represent autoantigens. Searches so far have not revealed the antigens in Table 2 to be polymorphic, so it is assumed that they are autoantigens.
The putative autoantigens identified include both nuclear and cytoplasmic proteins and a surface protein, np2, the first surface protein to be identified as a candidate autoantigen in TxCAD and which may be unique to endothelial cells within the cardiac vasculature. Neuropilin 2 mRNA is expressed in the human heart (15) and np2 protein in the mouse embryo aorta (16). Neuropilin 2 protein expression has not been studied in the heart or vasculature, although np1 is expressed by endothelial cells in vitro (17). Autoreactivity to this receptor is potentially interesting, given that it is a receptor for VEGF (10). However, studies on the roles of np1 and np2 in the vasculature are in the infancy stages and it is unknown whether np2 has functions similar to np1. If, like np1, np2 modulates the activity of VEGF (18), autoantibodies to np2 might be able to stimulate endothelial cell growth and thus contribute to the diffuse and concentric narrowing of the coronary arteries associated with TxCAD. However, the authors only found this reactivity in one of seven patient sera, and a larger study focusing on autoreactivity to np2 would be needed to determine whether this autoreactivity is likely to be important in TxCAD.
The other candidate autoantigens identified can be broadly described as being of two types: (1) proteins associated with cell proliferation and (2) membrane-associated proteins. The proliferative proteins include M-phase phosphoprotein and mki67a. Conceivably, autoantibodies to these proteins may perturb endothelial cell proliferative processes. This second category includes proteins involved in membrane cycling, phosphatidylinositol transfer protein-β, and cellular signaling events downstream from the traditional second messengers (arfaptin 2, rabaptin 5). These proteins that associate with membranes can be presented to antigen-presenting cells during apoptosis (19). Similarly, two antigens previously demonstrated to be associated with TxCAD, vimentin (7) and ribosomal protein 7 (8), are exposed on the surface of apoptosed cells. The prevalence of intracellular antigens in TxCAD is reminiscent of other autoimmune diseases such as systemic lupus erythematosus, and the identification of proteins cleaved during apoptosis and associated with membranes suggests a potential role for apoptosis in the development of autoreactivity in TxCAD. Indeed, histologic evidence of apoptosis is found within cardiac allografts showing TxCAD (20).
Experimental studies have confirmed that allotransplantation breaks tolerance to self-antigens and is associated with antibody and T-cell responses to autoantigens (21,22). Thus, autoimmune responses to cardiac myosin after experimental cardiac transplantation have been described (21) and responses to collagen after lung transplantation have been described (22). Previously (8) two-dimensional electrophoresis and N-amino acid sequence analysis identified vimentin as an autoantigen commonly associated with TxCAD in heart transplant recipients. It is curious that vimentin was not detected in any of the seven sera tested in this study. Although prokaryotic expression systems exclude antibody-antigen reactions dependent on protein glycosylation, vimentin is not glycosylated. All seven sera had tested positive for antivimentin antibodies by enzyme-linked immunosorbent assay. The authors have to conclude that the gene for vimentin is not represented in the cDNA library or that vimentin cannot be expressed by prokaryotic cells. The two proteins identified by sera from healthy volunteers in the present study have disparate functions, and one is classified as an autoantigen, DFS70 (lens epithelium-derived growth factor-transcription co-activator p75) (23). These results confirm that autoreactive antibodies can be found in normal serum (24). Whether the autoimmune responses to the antigens the authors have described contribute to the pathogenesis of TxCAD remains to be seen, although experimental studies suggest that responses to myosin (21) and collagen (22) accelerate graft rejection.
This study highlights the high diversity of non-HLA or autoantigens involved in TxCAD. However, the inability of cDNA library screening to detect previously identified TxCAD autoantigens implies potential limitations of this powerful technique. Consequently, the authors cannot formally exclude the possibility that they may not have detected other, possibly key antigens involved in the pathogenesis of TxCAD. Further studies will establish whether the potential autoantigens identified in this study are involved in pathogenesis of TxCAD, and results from those studies will determine the usefulness of this novel library screening strategy in identifying important autoantigens in this disease. Furthermore, efforts to limit the release of autoantigens or regulate the autoimmune response after allotransplantation may be of benefit in controlling this long-term complication.
1. Costanzo MR, Naftel DC, Protzer MR, et al. Heart transplant coronary artery disease detected by coronary angiography: A multi-institutional study of preoperative donor and recipient risk factors. J Heart Lung Transplant 1998; 17: 744.
2. Weis M, von Scheidt W. Coronary artery disease in the transplanted heart. Annu Rev Med 2000; 51: 81.
3. Russell PS, Chase CM, Colvin RB. Alloantibody- and T cell-mediated immunity in the pathogenesis of transplant arteriosclerosis: Lack of progression to sclerotic lesions in B cell-deficient mice. Transplantation 1997; 64: 1531.
4. Suciu-Foca N, Reed E, Marboe C, et al. The role of anti-HLA antibodies in heart transplantation. Transplantation 1991; 51: 716.
5. Dunn MJ, Crisp SJ, Rose ML, et al. Anti-endothelial antibodies and coronary artery disease after cardiac transplantation. Lancet 1992; 339: 1566.
6. Fredrich R, Toyoda M, Czer LS, et al. The clinical significance of antibodies to human vascular endothelial cells after cardiac transplantation. Transplantation 1999; 67: 385.
7. Wheeler CH, Collins A, Dunn MJ, et al. Characterisation of endothelial antigens associated with transplant associated coronary artery disease. J Heart Lung Transplant 1995; 14: S188.
8. Linke AT, Marchant B, Marsh P, et al. Screening of a HUVEC cDNA library with transplant-associated coronary artery disease sera identifies RPL7 as a candidate autoantigen associated with this disease. Clin Exp Immunol 2001; 126: 173.
9. Jurcevic S, Ainsworth ME, Pomerance A. Antivimentin antibodies are an independent predictor of transplant-associated coronary artery disease after cardiac transplantation. Transplantation 2001; 71: 886.
10. Neufeld G, Cohen T, Shraga N, et al. The neuropilins: Multifunctional semaphorin and VEGF receptors that modulate axon guidance and angiogenesis. Trends Cardiovasc Med 2002; 12: 13.
11. Hornick P, Lechler R. Direct and indirect pathways of alloantigen recognition: Relevance to acute and chronic allograft rejection. Nephrol Dial Transplant 1997; 12: 1806.
12. Liu Z, Colovai AI, Tugulea S, et al. Indirect recognition of donor HLA-DR peptides in organ allograft rejection. J Clin Invest 1996; 98: 1150.
13. Vella JP, Spadafora-Ferreira M, Murphy B, et al. Indirect allorecognition of major histocompatibility complex allopeptides in human renal transplant recipients with chronic graft dysfunction. Transplantation 1997; 64: 795.
14. Hornick PI, Mason PD, Baker RJ, et al. Significant frequencies of T cells with indirect anti-donor specificity in heart graft recipients with chronic rejection. Circulation 2000; 101: 2405.
15. Rossignol M, Gagnon ML, Klagsbrun M. Genomic organisation of human neuropilin-1 and neuropilin-2 genes: Identification and distribution of splice variants and soluble isoforms. Genomics 2000; 70: 211.
16. Chen H, Chedotal A, He Z, et al. Neuropilin-2, a novel member of the neuropilin family, is a high affinity receptor for the semaphorins Sema E and Sema IV but not Sema III. Neuron 1997; 19: 547.
17. Kitsukawa T, Shimono A, Kawakami A, et al. Overexpression of a membrane protein, neuropilin, in chimeric mice causes anomalies in the cardiovascular system, nervous system and limbs. Development 1995; 121: 4309.
18. Oh H, Takagi H, Otani A, et al. Selective induction of neuropilin-1 by vascular endothelial growth factor (VEGF): A mechanism contributing to VEGF-induced angiogenesis. Proc Natl Acad Sci USA 2002; 99: 383.
19. Messmer UK, Pfeilschifter J. New insights into the mechanism for clearance of apoptotic cells. Bioessays 2000; 22: 878.
20. Mayr U, Mayr M, Li C, et al. Loss of p53 accelerates neointimal lesions of vein bypass grafts in mice. Circ Res 2002; 90: 197.
21. Fedoseyeva EV, Zhang F, Orr PL, et al. De novo autoimmunity to cardiac myosin after heart transplantation and its contribution to the rejection process. J Immunol 1999; 162: 6836.
22. Yasufuku K, Heidler KM, O’Donnell PW, et al. Oral tolerance induction by type V collagen downregulates lung allograft rejection. Am J Respir Cell Mol Biol 2001; 25: 26.
23. Ochs RL, Muro Y, Si Y, et al. Autoantibodies to DFS 70 kd/transcription coactivator p75 in atopic dermatitis and other conditions. J Allergy Clin Immunol 2000; 105: 1211.
24. Ronda N, Haury M, Nobrega A, et al. Analysis of natural and disease-associated autoantibody repertoires: Anti-endothelial cell IgG autoantibody activity in the serum of healthy individuals and patients with systemic lupus erythematosus. Int Immunol 1994; 6: 1651.