Antibodies Eluted from Acutely Rejected RenalAllografts Bind to and Activate Human Endothelial CellsNewton Lucchiari, Nicolas Panajotopoulos, Chen Xu,Helcio Rodrigues, Luiz Estevam Ianhez, Jorge Kalil,and Denis Glotz
ABSTRACT: This study was designed to investigate howantiendothelial antibodies (EAbs) are involved in acuteirreversible renal graft rejection. Eluates from 25 renalallografts, lost by irreversible rejection (n 5 22) and byrenal vein thrombosis (controls n 5 3), were tested againsta panel of cultured human umbilical vein endothelial cells(HUVEC). All patients were under immunosuppression atthe time of nephrectomy. EAbs binding and membraneexpression of adhesion molecules ELAM-1 and VCAM-1were analyzed by flow cytometry (FACS) and by semi-quantitative RT-PCR for mRNAs coding for those mol-ecules. The absence of anti-HLA antibodies against thedonor was ascertained at transplant, and before and afternephrectomy by the negativity of specific crossmatches per-formed using the most sensitive techniques. EAbs elutedfrom eight rejected kidneys bound to HUVEC. They didnot induce any cytotoxicity, but their incubation withHUVEC (4 h at 37°C; 2.5 mg/ml) led to upregulation of
INTRODUCTIONThe importance of antibodies in human renal allograftrejection has not been clearly defined [1]. They are
thought to play a pivotal role in hyperacute rejection,associated with recipient presensitization and demonstra-ble reactivity with donor cells [2, 3]. Their participationhas also been implicated in the development of thevascular lesions conspicuous in later rejections [4, 5].Data establishing the relationship of circulating antibod-ies to graft fate in man are largely circumstantial, andinclude disappearance from circulation of preexistingantibodies when a transplantation is performed, the ap-pearance or rise in titer of antibodies after removal of agraft, as well as the deposition of immunoglobulin invascular, glomerular, and tubulo-interstitial areas [5, 6].
Preformed antibodies in the recipient bind to anddestroy the donor endothelium. The first antigenic tar-gets were identified as ABO system antigens, but class I
From the Laboratory of Transplantation Immunology, Heart Institute(N.L., N.P., H.R., J.K.), Department of Medicine, Division of Allergy andClinical Immunology (J.K.), and Unity of Renal Transplantation (L.E.I.)University of Sao Paulo School of Medicine, Sao Paulo, Brasil and Unite deRecherche Immunopatologie Humaine, INSERM U 430, Hopital Broussais,Paris, France (C.X., D.G.).
Address reprint requests to: Dr. Newton Lucchiari, Laboratorio de Imu-nologia de Transplantes, Instituto do Coracao da Faculdade de Medicina daUniversidade de Sao Paulo, Rua Dr. Eneas de Carvalho Aguiar 500 3°andar, 05403-000 Sao Paulo, Brazil; Fax: (155) 11-282-9350; E-Mail:[email protected].
Received December 27, 1999; accepted February 1, 2000.
Published by Elsevier Science Inc. PII S0198-8859(00)00109-9
HLA antigens have subsequently been identified as ad-ditional targets. The role of other non-HLA antibodies intransplant patients has not been so extensively investi-gated. An endothelial cell-specific alloantigen system hasbeen widely implicated in both HLA identical [7, 8] andnonidentical [9] graft failure. We have previously re-ported that in two cases of HLA identical living relateddonor kidney transplants, who lost their grafts by hu-moral rejection, crossmatches performed T lymphocytes,with or without anti-human globulin, B lymphocytes,and monocytes remained negative [10], but antiendothe-lium antibodies could be detected.
The antigens recognized are part of a complex minorhistocompatibility antigen system called vascular endo-thelial cell system (VEC) that is partially shared byperipheral blood monocytes. There is accumulating evi-dence in renal transplantation [11, 12], as well as incardiac transplantation [13], implicating anti-VEC anti-bodies even in hyperacute rejection. Moreover, the pre-transplant presence of such anti-VEC antibodies isstrongly linked to accelerated rejection in cardiac trans-plantation [14].
Understanding about the mechanisms involved in vas-cular rejection has greatly benefited from studies in thefield of xenotransplantation, demonstrating that activa-tion of the endothelial cells is a major event leading tothrombosis and graft loss. Activation of endothelial cellcan be induced by interleukin-1 (IL-1), tumor necrosisfactor-alpha (TNF-a), or lipopolysaccharide (LPS) [15].Bach et al. [16] suggested that endothelial cell activationmay occur because of the recipient xenoreactive naturalantibodies binding to the graft’s endothelium. This ac-tivation induces a number of events such as overexpres-sion of leukocyte adhesion molecules, endothelial leuko-cyte adhesion molecule-1 (ELAM-1), vascular celladhesion molecule-1 (VCAM-1), and intercellular adhe-sion molecule-1 (ICAM-1). Normal endothelial cells ex-press few ELAM-1 and VCAM-1 in vitro or in vivo [17].Increased expression in vivo has been demonstrated, forinstance, in patients with acute rejection of renal allo-grafts [18].
In this study we investigated, in patients with irre-versible rejections of renal allografts (without any donor-specific anti-HLA antibodies pre- and post-transplant),the participation of EAbs and their potential role incytotoxicity and activation of the endothelial cell.
PATIENTS AND METHODSKidney specimens were obtained from 11 cadaveric, 9living related, and 5 living unrelated donor allografts: 22specimens demonstrated irreversible rejection and 3 hadthrombosis of the renal vein (used as a control groupbecause they showed no signs of rejection). Nineteen
were first grafts and six were second grafts. Fifteenpatients were men and ten were women. The diagnosiswas established based on clinical and biopsy characteris-tics, as described in the Banff classification [19].
All patients were undergoing immunosuppressiontreatment at the time of the nephrectomy, of which 17were receiving double therapy (azathioprine and pred-nisone) and 8 triple therapy (azathioprine, prednisone,and cyclosporin). The characteristics of the patients andcontrols are summarized in Table 1.
Histology and ImmunofluorescenceSpecimens from kidney biopsies were routinely evaluatedby light microscopy and histological evidence of rejec-tion graded according to the Banff working classification[19]. Sections (3 mm) of formalin-fixed tissues werestained with hematoxylin and eosin, and periodic acid-Schiff following standard procedures [20]. For immuno-fluorescence studies, snap-frozen tissue stored at 270°Cwas cut at 5 m in a cryostat at 220°C and dried for 2 hat 37°C. Antibodies tested were specific for IgG, IgM,IgA, C3, C4, C1q, k- and l-light chains, and fibrinogen(all from Behringwerke AG, Marburg, Germany).
Lymphocytotoxicity TestLymphocytotoxic antibodies in serum specimens weredetected using both standard NIH and anti-human glob-ulin techniques [21, 22]. For crossmatching, recipientsera were tested against T and B donor cells before andafter dithiothreitol (DTT) treatment (5 mM, 30 min, at37°C).
Elution ProcedureCortical segments were minced and washed extensivelywith phosphate-buffered saline, pH 7.4; the 1000 gsediment was eluted with 0.02 M citrate pH 3.2 at 37°Cfor 1.0 h with constant stirring [23]. After centrifugationat 4000 g for 30 min, at 4°C, the supernate was neutral-ized, dialyzed overnight against phosphate-buffered sa-line, and concentrated by lyophilization (Edwards, SaoPaulo, Brazil) before testing.
Endothelial Cell Isolation and CultureHuman umbilical vein endothelial cells (HUVECs) wereobtained from normal vaginal deliveries or Cesarean sec-tions. Briefly, umbilical cords were collected within 12 hof delivery for endothelial cell isolation using the classi-cal procedures already described [24]. The cord vein wascleared of red blood cells by perfusion with lactatedRinger’s solution, followed by incubation with collage-nase (GIBCO, Grand Island, NY) 1 mg/ml in phosphatebuffered saline for 30 min at 37°C in 5% CO2 and roomair. The vein was then perfused with several volumes ofHank’s balanced salt solution to remove the cells. A pellet
519Antiendothelial Antibodies Involved in Acute Irreversible Renal Graft Rejection
was formed by centrifuging at 200 3 g for 5 min. Afterthe cells grew to visual confluence they were passaged byincubating the cell layer with 0.05% trypsin solution inEDTA buffer for 2 min at 37°C. Trypsin was then inac-tivated with 199 medium containing 20% fetal calf serum.
Activation of endothelial cells was tested using kidneyeluates, serum of a young healthy volunteer (normalserum), medium 199, as well as two positive controls(TNF used at 0.1 ng/100 ml, and serum from an hyper-immunized patient) and patients with acute vascularrejection. Tests were performed with cell preparationsoriginating from ten different donors. Cultured cellswere used until the second passage was reached.
Complement-mediated Cytotoxicity Assayof EluatesCytotoxicity was measured by a flow cytometry assayusing propidium iodide (PI) (Molecular Probes, Eugene,OR) as a marker of cell lysis. Primary human endothelialvein cells (100 ml, 2 3 106 cells/ml) were incubated for120 min at 37°C with 50 ml, at 2.5 mg/ml of eluate andserum controls, diluted (1:100) in RPMI using Falconplastic tubes. After washing with medium (RPMI with20% FCS), rabbit complement (Biotest, Dreieich, Ger-
many) was added, and incubation was allowed to proceedfor 1 h at 37°C. Cell fluorescence was measured in aFACScan (Becton Dickinson, San Jose, CA) flow cytom-eter. This analysis enabled the determination of thepercentage of dead cells. A control tube without serum oreluate enabled the determination of the spontaneouslysis. The percent of specific cytotoxicity was calculatedusing the formula:
% cytotoxicity
5% obtained lysis 2 % spontaneous lysis
100% 2 spontaneous lysis
Analytical MethodsConcentrations of pure preparations of antibodies weremeasured using the bicinchoninic acid protein assay[25]. Purity of antibodies (eluates) was ascertained bypolyacrylamide gel electrophoresis. The eluates were con-centrated by overnight lyophilization against a pH 8.5buffer. Eluates were electrophoresed on 10% SDS-PAGE.The eluates were diluted in the presence of a reducingagent (dithiothreitol, 150 mM). Human IgG and IgM(Dako, Carpinteria, CA) were used as controls. Ten mi-
TABLE 1 Summary of characteristics of antibody allograft eluates tested and clinical data
NT 5 not tested; TF 5 technical failure; AVR 5 acute vascular rejection; CVR 5 chronic vascular rejection; ACR 5 acute cellular rejection; LRD 5 live relateddonor; LURD 5 live unrelated donor; C 5 cadaveric.
520 N. Lucchiari et al.
crograms of eluted protein were applied on 10% sepa-rating polyacrylamide gel runs according to Laemmli’stechnique [26]. Proteins were fixed and stained withCoomassie blue [27].
Quantitative Ig Concentrations, Assay of IgM, IgG,and IgA
The IgG, IgM, and IgA in eluates was assayed by im-munonephelometry (Turbox reagent, Merck, Dreieich,Germany).
Determination of Antibody Binding toEndothelial Cells
The presence of antibodies in the test eluates capable ofbinding to the HUVECs was determined by flow cytom-etry, a minimum of 2500 cells for sample was analyzed.Cells were incubated with eluates at a concentration of2.5 mg/ml, normal serum (negative control), and serumfrom hyperimmunized patient (positive control) and pa-tients with acute vascular, acute cellular, and chronicvascular rejection for 3 h at 37°C, and washed (threetimes). The cells were then incubated for 60 min, at 4°Cin the presence of fluorescein-conjugated rabbit F(ab9)2fragment to human immunoglobulins (IgG, IgA, IgM)(Organon Teknika Corporation, Durham, NC) diluted1/100 in PBS buffer. The sample was analyzed by flowcytometry on a FACScan flow cytometer (Becton Dick-inson). A minimum of 2500 cells per sample was ana-lyzed. PI (1 g/ml in medium) was added to exclude anydead cells or debris from the analysis. Cells were consid-ered positive if their fluorescence was greater than thefluorescence of cells incubated with negative control 1 2SD.
Determination of Endothelial Cells Expression ofVCAM-1 and ELAM-1 Molecules byFlow Cytometry
HUVECs at second passage were incubated for 4 h witheluates at a concentration of 2.5 mg/ml or various con-trols. HUVECs were washed (three times) with PBS andincubated in the presence of antivascular cell adhesionmolecule-1 (anti-VCAM-1 [clone 1/6] or anti-ELAM-1[clone H18/7] mouse monoclonal antibody [BectonDickinson] for 30 min, at 4°C. The cells were extensivelywashed with PBS, and stained with fluorescein-conju-gated anti-mouse immunoglobulin F(ab9)2 diluted at1/100 (Silenus, Hawthorn, Australia), absorbed on hu-man-Ig). After incubation, the cells were detached with30 M EDTA for 3 to 4 min at 37°C and analyzed by flowcytometry as described above.
Reverse Transcription and Polymerase ChainReaction AnalysisReverse transcription. The mRNAs were extracted fromendothelial cells (one million cells) after incubation for4 h with the eluates. Total RNAs were extracted by themethod of Chomczynski et al. [28]. The RNAs weredissolved in sterile water and their concentration deter-mined by absorbance at 260 nm. RNA solutions wereprepared from cells incubated with negative and positivecontrols and from three patient’s samples (1 mg/4 ml).They were reverse transcribed into cDNA with oligo(dT)and MMLV reverse transcriptase (Gibco-BRL, Gaithers-burg, MD). The reverse transcription products were am-plified with primers for human VCAM-1 (upper primer:59GAC CTT CAT CCC TAC CAT TG 39; lower primer:59ACT TCC TTT CTG CTT CTT CC 39; productlength, 372 base pairs [bp]), and for human ICAM-1(upper primer: 59CGA CTG GAC GAC AGG GATTGT 39; lower primer: 59ATT ATG ACT GCG GCTGCT ACC 39; product length, 290 bp). Primers se-quences were found in the GenBank dabase (NationalCenter for Biotechnology Information; National Insti-tutes of Health, Bethesda, MD).
Polymerase chain reaction. Two microliters of each cDNAwere amplified to a total volume of 25 ml containing 50pmol oligonucleotide primers, 10 mmol/liter of eachdNTP, 1 3 PCR buffer (10 mmol/liter Tris-HCl, pH8.3, 50 mmol/liter KCl), 40% DMSO, 1 mmol/literMgCl2, and 2.5 U ampliTaq polymerase (Perkin Elmer-Cetus, Emeryville, CA). The intensities of the cDNAbands for each protein were normalized to the GAPDHband intensities. The primers used for the amplificationof GAPDH were: 59GTG AAG GTC GGA GTC AAC G39 (sense) and 59GGT GAA GAC GCC AGT GGA CTC39 (antisense). The expected PCR product size was 299bp for GAPDH. PCR was carried out in a PerkingElmer-Cetus DNA thermal cycler with the sequence 1min at 94°C, 2 min at 55°C, and 3 min at 72°C. Theappropriate number of cycles for remaining within theexponential phase for further amplifications was deter-mined for both primers with the most concentratedsample. Fifteen microliters of each PCR product wasthen run on an ethidium bromide-stained agarose gel.All experiments were performed in duplicate.
Quantification of Detected SignalsPolaroid photographs of ethidium bromide-stained gelswere digitized into gray-scale images. The amount ofnucleic acids were determined by densitometry and nor-malized with GAPDH. The amount of nucleic acids wasproportional to the log of the optic density. The sum ofthe logarithms of the pixel values was used to estimate
521Antiendothelial Antibodies Involved in Acute Irreversible Renal Graft Rejection
the amount of nucleic acids band. Analysis was per-formed on a Macintosh Quadra 840 AV computer (AppleComputers, Cupertino, CA) using the public domainNIH Image 1.51 program (developed at the U.S. Na-tional Institute of Health and available on the internet byanonymous ftp [www.zippy.nimh.nih.gov]).
RESULTSPatients (n 5 25) presented with acute vascular rejection(n 5 9), chronic vascular rejection (n 5 8), acute cellularrejection (n 5 5), and technical failures (control group,n 5 3). The histological findings of kidney biopsies weregraded using the Banff working classification [19].
In all cases of acute vascular rejection, immunofluo-rescence studies showed vascular deposits of IgM and C3.
Donor-specific Anti-HLA AntibodiesCrossmatches performed against donor T lymphocytes,with or without anti-human globulin, and B lympho-cytes were negative in all serum specimens collectedbefore transplantation, during rejection, and 1 and 2weeks after nephrectomy.
Characterization of EluatesTable 1 shows the type of immunoglobulins obtainedfrom the eluates. Eight out of the nine eluates of kidneyslost from acute vascular rejection were composed of IgMs
(p , 0.001). Moreover, eight out of the nine eluatescomposed of IgMs were associated with acute vascularrejection (p , 0.001). Thus, these data suggest a strongcorrelation between acute vascular rejection and the pres-ence in the rejected kidney of IgMs.
Antibody Binding to Endothelial CellsEight out of the 25 eluates tested showed binding toHUVECS (Figure 1). All those eluates showed the pres-ence of immunoglobulins of the IgM isotype and wereharvested from kidneys with acute vascular rejection.
Cytotoxic TestsNone of the eluates tested had any cytotoxic activity onendothelial cells, in contrast to the serum of an hyper-immunized patient.
Expression of VCAM-1 and ELAM-1 byFACS AnalysisExpression of VCAM-1. Incubation of endothelial cellswith culture medium alone was associated with stainingfor VCAM-1 in .10% of the cells; incubation withnormal human serum led to an increase to 40% of thecells. Incubation with TNF resulted in VCAM-1 expres-sion in 78% of the cells, and seven of eight HUVEC-binding eluates caused a stimulation in endothelial cellsvarying from 55 to 89% (see Figure 2A).
Results were independent of the origin of endothelialcells because there was no difference in stimulation be-tween the various endothelial cell cultures, as shown bythe standard deviation bars.
Expression of ELAM-1. As for VCAM-1, incubation ofHUVECs with medium alone led to a minimal number
FIGURE 1 Cytometric evaluation of antibodies capable ofbinding to the HUVECs. Cells were incubated with eluates ata concentration of 2.5 mg/mL or normal serum control orpositive control (serum of an hyperimmunized). The bar chartshows the mean 6 SD of the positive fluorescence of the cellsfrom three donors and duplicate experiments.
522 N. Lucchiari et al.
of positive cells; TNF induced ELAM-1 expression inclose to 80% of the cells, and similarly seven of the eightHUVEC-binding eluates induced a strong expression ofELAM-1 (see Figure 2B).
Correlation VCAM/ELAM. It is of note than induction ofVCAM and ELAM were strictly correlated (r 5 0.77;p , 0.05), in all eluates but one inducing expression ofboth adhesion molecules. On the other hand, all HUVECnon-binding eluates did not induce any significant in-crease in membrane expression of VCAM and/or ELAM(data not shown).
Semiquantitative RT-PCR. Most of the eluates did nothave enough amounts for further examination. For thisreason only three eluates were tested: 3, 4, and 6. Theresults show that the amplification obtained forVCAM-1 was increased 59-fold for eluatum 4, 49-foldfor eluatum 3, and 35-fold for eluatum 6 as compared
FIGURE 2 Cytometric evaluation of VCAM-1 andELAM-1 expression on HUVECs. HUVECs were incubatedfor 3 h in the presence of medium alone (negative control),TNF (positive control), normal human serum, or eluates (col-umns 2 to 9). After washing, cells were incubated with anti-body against VCAM-1 (A) or ELAM-1 (B). The bar chartshows the mean (ELAM-1 5 62.78; VCAM-1 5 72.40) 6 SD(ELAM-1 5 15.70; VCAM-1 5 9.56) of the positive fluores-cence of the cells from the ten donors and duplicate experi-ments.
523Antiendothelial Antibodies Involved in Acute Irreversible Renal Graft Rejection
with control. With ICAM-1, the increase was 12.7-foldfor eluatum 3, 12.4-fold for eluatum 4, and 8.2-fold foreluatum 6, as compared with negative control (Figures3A and 3B).
DISCUSSION
Allograft rejection continues to be a critical problem intransplantation. The underlying immunological mecha-nisms are diverse and involve both antibody-mediatedand cellular events.
Sensitization against HLA antigens is an importantrisk factor in transplants. The presence of preformedantibodies against HLA class I antigens of the donor isknown to lead to hyperacute rejection [29]. These anti-bodies are produced following a previous transplant,
FIGURE 3 Expression of messenger RNA encodingVCAM-1 and ICAM-1 adhesion molecules. Eluates 3, 4, 6,and positive control (TNF and serum of hyperimmunizedpatient) induced an overexpression of mRNA coding forVCAM-1 (A) and ICAM-1 (B). The intensities of the cDNAbands for each protein were normalized to the GAPDH bandintensities.
524 N. Lucchiari et al.
transfusion, or pregnancy. The methods used to detectthese antibodies are obviously a critical factor, but aremainly directed to HLA antigens. Thus, despite all theprogress in crossmatch techniques and immunosuppres-sive therapy, rejection can still occur due to mechanismsmediated by undetected antibodies. Iwaki and Terasaki[30] have shown that 8% of the first grafts and 14% ofregrafts never function, despite a negative crossmatchagainst mononuclear cells. The authors suggest that an-tibodies not detected by the conventional crossmatchmethods caused a percentage of these losses. In thisstudy, the absence of anti-HLA antibodies against thedonor was ascertained at transplant, during rejection, andbefore and after nephrectomy by the negativity of cross-matches performed using the most sensitive techniques.This study indeed confirms that antibodies capable ofbinding to the graft can be found in cases of acutevascular rejection, even when crossmatches are negative.
Initial studies by our group showed that non-HLAantigens expressed by endothelial cells could have animportant role in the rejection process of allografts inABO and HLA compatible individuals. Moraes andStatsny [31] described specific antibodies against thedenominated endothelium/monocyte system in serumfrom patients with renal transplants. The presence ofthese antibodies was detected in some cases by directimmunofluorescence [32] and in others by cytotoxicitymediated by complement [33], in pre- and post-trans-plant sera.
The mechanisms by which these antibodies are harm-ful to the graft have not been fully elucidated. Theseantibodies could mediate the aggression against thetransplanted organ via lysis by complement [34], or evenby antibody-dependent cell-mediated cytotoxicity(ADCC) [35, 36]. Some studies have shown the presenceof cytotoxic antibodies against endothelial cells in theserum of patients with transplants, in association withrenal [33–37] and cardiac rejection [38]. Alternatively,as demonstrated in xenotransplantation, antibodies couldactivate the endothelial cells without being cytotoxic. In1996, von Willebrand et al. [18], demonstrated that, inacute rejection of the kidney graft, adhesion moleculessuch as ICAM-1, VCAM-1, ELAM-1, and platelet-endo-thelial cell-adhesion molecule-1 (PECAM-1) are stronglyinduced, especially in endothelial cells and may have amajor role in the effector phase of the rejection process.
In this study, we could detect anti-endothelial IgMantibodies in eight out of nine (88.8%) eluates fromkidneys lost from acute vascular rejection, but in none ofthe 13 (0.0%) kidneys lost from other types of rejection(p , 0.0001). Conversely, eight of the nine cases of acutevascular rejection were associated with the presence ofIgMs. This strongly suggests that these antibodies areinvolved in the loss of the allograft. Although these
antibodies are of the IgM isotype, and thus can activatecomplement, we were unable to demonstrate any cyto-toxicity towards endothelial cells. On the contrary, usingadhesion molecules as markers of activation, we demon-strated that these antibodies may induce rejectionthrough activation of endothelial cells.
Upregulation of those molecules was demonstratedboth at the protein level, by showing an increased mem-brane expression, and at the molecular level, by showinga major increase in the specific mRNAs coding for thosemolecules. Both adhesion molecules were regulated inthe same fashion, as all eluates increased the expression ofboth molecules. Moreover, upregulation of the mem-brane expression of adhesion molecules was associatedwith an increase of their specific mRNAs.
None of the HUVEC non-binding eluates led to acti-vation of the endothelial cells, suggesting that only thepresence of antibody is responsible for the upregulation ofadhesion molecules and not occasional technical failures.
In summary, we demonstrate herein, in eluates ofkidneys lost from acute vascular rejection, the presence ofIgM antibodies are capable of binding to human endo-thelial cells and provoking their activation without anydemonstrable cytotoxicity. These results strengthen thehypothesis that anti-endothelium antibodies may repre-sent a major cause of vascular rejection, and suggest thatthey could act through endothelial cell activation morethan through classical cytotoxicity.
REFERENCES
1. Karuppan SS, Ohlman S, Moller E: The occurrence ofcytotoxic and non-complement-fixing antibodies in thecrossmatch serum of patients with early acute rejectionepisodes. Transplantation 54:839, 1992.
2. Patel R, Terasaki PI: Significance of the positive cross-match test in kidney transplantation. N Engl J Med280:735, 1969.
3. Willians GM, Hume DM, Hudson RP Jr, Morris PJ,Kano K, Milgrom F: Hyperacute renal-homograft rejec-tion in man. N Engl J Med 279:611, 1968.
4. Jeannet M, Pinn VN, Flax MH, Winn HJ, Russel PS:Humoral antibodies in renal allotransplantation in man.N Engl J Med 282:111, 1970.
5. Garavoy MR, Gailiunas P, Carpenter, Busch GJ: Immu-nologic monitoring of transplant rejection: correlation ofin vitro assays with morphologic changes on transplantbiopsy. Nephron 22:208, 1978.
6. Andres GA, Accinni L, Hsu KC, Penn I, Porter KA,Rendall JM, Seegal BC, Starzl TE: Human renal transplantsIII. Immunophatologic studies. Lab Invest 22:588, 1970.
7. Cerilli J, Bay W, Brasile L: The significance of the mono-cyte crossmath in recipients of living-related HLA iden-tical kidney graft. Hum Immunol 7:45, 1983.
525Antiendothelial Antibodies Involved in Acute Irreversible Renal Graft Rejection
9. Baldwin WM, Claas FHJ, Paul LC, et al.: All monocyteantigens are not expressed on renal endothelium. TissueAntigens 21:254, 1983.
10. Kalil J, Guilherme L, Neumann J, Rosales C, Marin M,Saldanha L, Chocair PR, Ianhez LE, Sabbaga E: Humoralrejection in two HLA identical living related donor kid-ney transplants. Transplant Proc 21:711, 1989.
11. Paul LC, van Es LA, Kalffa MW, Graeff J: Intrarenaldistribution of endothelial antigens recognized by anti-bodies from renal allograft recipients. Transplant Proc11:427, 1979.
12. Paul LC, Baldwin WM III, van Es LA: Vascular endothe-lial alloantigens in renal transplantation. Transplantation40:117, 1985.
13. Brasile L, Rodmann E, Shield CF, Clarke J, Cerilli J: Theassociation of antivascular endothelial cell antibody withhyperacute rejection: a case report. Surgery 99:639, 1986.
14. Dunn MJ, Crisp SJ, Rose ML, Taylor PM, Yacoub MH:Anti-endothelial antibodies and coronary artery diseaseafter cardiac transplantation. Lancet 339:1566, 1992.
15. Pober JS, Cotran RS: The role of endothelial cells ininflammation. Transplantation 50:537, 1990.
16. Bach FH, Blakely ML, van der Werf W, Vanhove B,Stuhlmeier KM, Hancock WW, de Martin R, Winkler H:Discordant xenografting: a working model of problemsand issues. Xeno 1:8, 1993.
17. Fuggle S, Sanderson JB, Gray DWR, Richardson A, Mor-ris PJ: Variation in expression of endothelial adhesionmolecules in pretransplant and transplanted kidneys-cor-relation with intragraft events. Transplantation 55:117,1993.
18. von Villebrand E, Lautenschlager I, Krogerus L, Hayry P,Isoniemi H, Salmela K: Adhesion molecules and activa-tion markers in acute rejection of human renal allografts.Transplant Immunol 4:57, 1996.
19. Solez K, Axelsen RA, Benediktsson H, et al.: Internationalstandardization of nomenclature and criteria for histologicdiagnosis of renal allograft rejection: the Banff workingclassification of kidney transplant pathology. Kidney Int44:411, 1993.
20. Bergastrad B, Bohman SO, Farnsworth A, et al.: Renalhistopathology in kidney transplant recipients immuno-suppression with cysclosporin A: results of an intenationalworkshop. Clin Nephrol 24:107, 1985.
21. Hopkins KA: Basic microlymphocytotoxicity test. In Za-chary AA, Terisi GA (eds). ASHI Laboratory Manual, 2nded. Lenexa, KS: American Society of Histocompatibilityand Immunogenetics, 1990:195.
22. Noreen HJ: Crossmatch test. In Zachary AA, Terisi GA(eds). ASHI Laboratory Manual, 2nd ed. Lenexa, KS:American Society of Histocompatibility and Immunoge-netics, 1990:307.
23. McPhaul JJ Jr, Dixon FJ: Characterization of humananti-glomerular basement membrane antibodies in renalallograft rejection. J Clin Invest 49:308, 1970.
25. Smith PK, Krohn RI, Hermanson GT, Malia AK, GartnerFH, Provenzano MD, Fujimoto EK, Goeke NM, OlsonBJ, Klenk DC. Measurement of protein using bicinchonicacid. Anal Biochem 150:76, 1985.
26. Laemmli UK: Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227:680, 1970.
27. Fairbanks G, Steck TL, Wallach DFH: Electrophoreticanalysis of the major polypeptides of the human erythro-cyte membrane. Biochemistry 10:2606, 1971.
28. Chomczynski P, Sacchi N. Single-step method of RNAisolation by acid guanidium thiocynate-phenol-chloro-form extration. Anal Biochem 162:156, 1987.
30. Iwaki Y, Terasaki PI: Primary nonfunction cadaver kidneytransplantation: evidence for hidden hyperacute rejection.Clin Transplant 1:125, 1987.
31. Moraes JR, Statsny P: Eight groups of human endothelialcell alloantigens. Tissue Antigens 8:272, 1976.
32. Baldwin WM III, Soulillou JP, Claas FHJ, Peyrat MA,van Es LA, van Rood JJ: Antibodies to endothelial anti-gens in eluates of 88 human kidneys: correlation withgraft survival and presence of T- and B-cell antibodies.Transplant Proc 13:1547, 1981.
33. Soullilou JP, Pyrat MA, Guenel J: Studies of pre- andpostgraft antibodies against endothelial cells (unbilicalcord) in kidney allografts. Transplant Proc 13:1551,1981.
34. Baldwin WM III, Pruitt SK, Brauer RB, Daha MR,Sanfilippo F: Complement in organ transplantation.Transplantation 59:797, 1995.
35. Miltenburg AMM, Meijer-Paape ME, Weening JJ, DahaLA, van Es MR, van der Woude FJ: Induction of anti-body-dependent cellular cytotoxicity against endothelialcells by renal transplantation. Transplantation 48:681,1989.
36. Yard B, Spruyt-Gerrtse M, Claas F, Thorogood J, BruijnJA, Paape ME, Stein SY, van Es LA, van Bockel JH,Kooymans-Coutinho M, Daha MR, van der Woude FJ:The clinical significance of allospecific antibodies againstendothelial cells detected with an antibody-dependentcellular cytotoxicity assay for vascular rejection and graftloss after renal transplantation. Transplantation 55:1287,1993.
