Background

Chronic renal allograft rejection is characterized histologically by transplantation-associated arteriosclerosis and glomerulosclerosis (Tx-AA and Tx-AGS). Recent studies in animal models implicate the mitogenic and pressor actions of endothelin-1 (ET-1) in Tx-AA. In humans, however, a link between elevated ET-1 secretion and Tx-AA or Tx-AGS remains unclear. In this study we analyzed expression of ET-1 in the vasculature of renal transplant patients with chronic or acute rejection and in normal controls.

Methods

Renal vascular and glomerular ET-1 was assessed by immunohistochemistry in 12 patients with clinically and histologically defined chronic rejection, in 11 patients with acute rejection, and in 5 normal kidneys. ET-1 staining was also correlated with various clinical parameters and with a morphometric index of neointima formation. ET-1 secretion was measured by ELISA in cultured human vascular cell types treated with T cell- and macrophage-associated cytokines.

Results

We found that renal allografts with chronic rejection and Tx-AA expressed 6.1-fold more ET-1 in the vasculature relative to allografts with acute rejection or to normal kidneys (P < 0.01). In Tx-AA, ET-1 was detected predominantly in the neointima, which contained mostly endothelial cells and smooth muscle cells. A strong positive correlation (r = 0.82, P < 0.01) was observed between vascular ET-1 peptide expression and hypertension in patients with chronic rejection. We also showed that macrophage-associated cytokines, but not T cell-associated cytokines, stimulated ET-1 secretion in human endothelial cells, vascular smooth muscle and mesangial cells.

Conclusions

These results demonstrate that elevated ET-1 in the neointima is associated with Tx-AA and chronic rejection. In addition, these results point to an important role for endothelial dysfunction in chronic renal allograft rejection.

METHODS

Patients and controls

We studied 23 renal transplant recipients with clinical and histologic evidence of actue (N = 11) or chronic (N = 12) allograft rejection. All studies were approved by the Institutional Review Board, University Hospitals of Cleveland. Characteristics of subjects at the time of tissue recovery are summarized in Tables 1 and 2. Chronic rejection was considered clinically in patients who had been transplanted for at least one year and who exhibited proteinuria (>500 mg/day) and/or a slow, steady increase in serum creatinine concentration from a value obtained six months post-transplant. Acute rejection was considered clinically when patients exhibited a rise in serum creatinine concentration of at least 30% above baseline over a period of less than three months. In each patient the severity of systemic hypertension was scored as follows: 0 = no hypertension, no antihypertensive medication; I = normal blood pressure on antihypertensive medication; II = above normal blood pressure on antihypertensive medications; and III = above normal blood pressure on more than two antihypertensive medications.

Kidney tissue was obtained by biopsy or nephrectomy and immediately processed for immunohistochemistry as described below. All tissue samples were blindly graded for chronic or acute renal allograft rejection using the Banff criteria²⁶. Patients with clinically diagnosed chronic rejection but with histologic evidence of both acute and chronic rejection were categorized in the chronic rejection group. The vasculature of each sample was graded for Tx-AA using the morphological scale of Russell and coworkers²⁷. Normal human kidneys (N = 4), surgically excised due to localized neoplasm, were obtained from the National Cancer Institute Cooperative Human Tissue Network at Case Western Reserve University. All tissue was distant from the neoplasm and macroscopically normal. At the time of excision, patients had normal serum creatinine concentrations and normal blood pressures.

Immunohistochemical analysis

Small blocks of fixed (10% buffered formalin) kidney were embedded in paraffin and 5 m sections were cut. De-paraffinized sections were washed in Dulbecco's phosphate buffered saline (DPBS) and incubated with 1.5% H₂O₂ in methanol to block endogenous peroxidase activity followed by five DPBS washes. To block non-specific binding, sections were incubated with 10% normal goat serum (NGS). Rabbit polyclonal antisera raised against human ET-1 (1:200 to 1:400 in DPBS/1.0% NGS; #RIN6901; Peninsula Labs, Belmont, CA, USA) or a monoclonal antibody against endothelin-1 (1:100 dilution; QED Bioscience, San Diego, CA, USA) were added in a humidity chamber at room temperature for one hour. Cross reactivities of the polyclonal antisera were: ET-1, 100%; ET-2, 7%; ET-3, 7%; proET-1, 17%. The monoclonal antibody recognized only ET-1. Both antibodies gave essentially identical results. For negative controls primary antibody was replaced with non-immune serum at the same concentration of IgG. Antibodies were then localized with a peroxidase-labeled F(ab')2 fragment goat anti-rabbit (or mouse) IgG at 1:200 (Kierkegaard and Perry Laboratories, Inc., Gaithersburg, MD, USA). The sections were finally treated for five minutes with a freshly prepared diaminobenzidine substrate solution with enhancement and washed in tap water. After lightly counterstaining in methyl green, the sections were dehydrated in ethanol, cleared in xylene, and mounted in Eukitt (Calibrated Instruments Inc., Hawthorne, NY, USA). Photomicrographs were exposed and developed so that semiquantitative comparison of relative staining intensity was possible, and densitometric analysis of ET-1 immunoreactivity was performed using NIH Image v. 5.1 for the Macintosh.

In other experiments the sections were stained with the following primary antibodies to identify specific cell types within the neointima: polyclonal anti-Factor VIII (1:100, A0082; Dako, Carpinteria, CA, USA) for endothelial cells; monoclonal anti--smooth muscle actin (1:50, clone 1A4; Sigma Chemical Co., St. Louis, MO, USA) for vascular smooth muscle cells; polyclonal anti-CD3 (1:100, A0452; Dako) for T cells; and monoclonal anti-CD68 (1:75, M0876; Dako) for monocytes/macrophages. The antibodies were then detected and the tissue processed exactly as described above.

Cell culture

Human vascular cells were propagated in culture for measurements of cytokine-stimulated ET-1 secretion. Human umbilical vein endothelial cell (HUVEC) lines were obtained from the American Type Culture Collection (ATCC, ECV 304) and cultured in Medium 199 with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin²⁸. Human aortic vascular smooth muscle cell strains from the ATCC (CRL-1999) were cultured in F12K medium with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, 10 g/ml each of insulin and transferrin, 10 ng/ml selenite, 30 g/ml endothelial growth supplement, 50 g/ml ascorbic acid, 2 mM glutamine, 10 mM HEPES, and 10 mM TES [(N-tris) hydroxymethylmehtyl-2-aminoethanesulfonic acid]. Mesangial cell strains from human kidney were isolated and characterized as previously reported by Schultz et al²⁹. Cells were maintained in RPMI 1640 medium supplemented with 17% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, 5 g/ml each of insulin and transferrin, and 5 ng/ml selenite at 37°C in 5% CO₂ incubator. Characterization was performed by phase contrast microscopy and by immunostaining for intermediate filaments and surface antigens as described²⁹. Briefly, cells were positive for desmin, vimentin and myosin, but did not stain for factor VIII, keratin, or common leukocyte antigen.

Measurements of endothelin-1 secretion

Confluent cells in 12-well plates were made quiescent for 24 hours in DMEM/0.1% fetal bovine serum. Prior to stimulation with cytokines (R&D Systems, Minneapolis, MN, USA), the quiescent media was changed to Hank's balanced salt solution (without phenol red) with 25 mM HEPES and 0.1% fetal bovine serum. One milliliter of culture supernatant was extracted with 1.5 ml acetone:1 M HCl:water (40:1:5 vol/vol) and centrifuged at 4°C for 20 minutes at 1,200 g. The supernatant was dried down, reconstituted in 250 l, neutralized to pH 7.0, and assayed immediately for ET-1. Immunoreactive ET-1 was measured in 100 l aliquots using a double antibody sandwich enzyme-linked immunosorbent assay (R&D Systems) with minor modifications. Cross reactivity in the ELISA assay were bigET-1 < 1.0%, ET-2 < 45%, and ET-3 < 14%. Absorbance readings from a media blank were subtracted from the experimental values. All assays were done in duplicate.

Data analysis

Statistical significance for densitometric analysis of ET-1 staining was calculated by an unpaired t-test. For analysis of ET-1 secretion, statistical significance was calculated by a paired t-test in which repeated experiments run at different times were compared with their own control. Data presented are mean SEM normalized to the intra-experimental control value. Statistical analysis was performed using InStat for the Macintosh (Version 2.0; GraphPad Software, San Diego, CA, USA).

RESULTS

Vascular expression of ET-1 is elevated in Tx-AA

Table 1 summarizes the characteristics of the six male and six female subjects with chronic renal allograft rejection at the time tissue was obtained. Table 2 summarizes the same information for the four male and seven female subjects with acute rejection. Subjects were similar in age (40 14 years and 41 17, respectively, mean SD) and in racial distribution. The average duration of graft survival was 6.7 5.0 years for subjects with chronic rejection compared to 0.8 1.5 years for subjects with acute rejection. Average values for serum creatinine concentration were similar in both groups (6.5 2.9 mg/dl and 5.0 3.6, respectively).

Table 1 - Patients with chronic renal allograft rejection.

Full table (44K)

Table 2 - Patients with acute rejection.

Full table (44K)

To determine whether ET-1 secretion was elevated in the vasculature of subjects with Tx-AA, ET-1 was analyzed by immunohistochemistry in renal tissue from the 12 patients with chronic rejection. ET-1 staining in a representative artery with Tx-AA and neointima formation is illustrated in the photomicrograph in Figure 1a. ET-1 staining was abundant in the neointima (arrowhead, Figure 1a), particularly in the endothelium and in distinct underlying cells. ET-1 staining in the neointima was reproducibly observed in both arteries and arterioles with Tx-AA and neointima formation. ET-1 staining in Tx-AA was occasionally found in the adventitia (arrowhead, Figure 1a), but adventicial ET-1 was not present in all vessels that contained neointimal ET-1. The pattern of ET-1 staining in Tx-AA contrasts with the primarily endothelial localization of ET-1 in the arcuate artery and in other renal vessels from normal subjects (arrowheads, Figure 1b). In renal veins of subjects with chronic rejection, ET-1 localized exclusively to the endothelium and was indistinguishable from ET-1 staining in renal veins from normal subjects (data not shown). To assess continuity of the endothelium in Tx-AA, sections were also stained with a polyclonal antibody that recognizes Factor VIII, and in all cases the endothelium was apparently intact in Tx-AA Figure 1c. In addition, staining of renal sections from chronically rejecting kidney with non-immune antisera demonstrated that the immunhistochemical analysis was highly specific Figure 1d. Taken together, these results demonstrate that ET-1 secretion is prominent in the neointima (and sometimes in the adventitia) of vessels with Tx-AA.

Figure 1.

Elevated endothelin-1 (ET-1) peptide levels in the renal vascular neointima of patients with transplant-associated arteriosclerosis (Tx-AA) and chronic rejection. (A) Immunohistochemical localization of ET-1 in the neointima (arrowhead) and adventitia (arrowhead) of a small artery with Tx-AA. The majority of arterial vessels analyzed had elevated ET-1 only in the neointima. (B) ET-1 staining localizes primarily to the endothelium (arrowheads) in an arcuate artery from normal kidney. (C) Staining for Factor VIII in vessels with Tx-AA demonstrated that the endothelium was intact. (D) Immunohistochemical staining with non-immune serum in a vessel with Tx-AA. Magnification 1400.

Full figure and legend (176K)

To determine whether elevated vascular secretion of ET-1 was specific for patients with chronic rejection, we examined vascular ET-1 secretion in 11 subjects with acute rejection lacking pathological evidence of Tx-AA (Table 2). ET-1 staining localized to the intima and was essentially normal in renal vessels of patients undergoing acute rejection (data not shown). A semiquantitative densitometric analysis of vascular ET-1 staining revealed that ET-1 secretion was unchanged in acute rejection but was elevated 6.1-fold in chronic rejection compared to vessels of normal subjects Figure 2. In fact, 2 of 11 subjects with acute rejection actually had lower vascular ET-1 than observed in normal vessels. These results suggest that elevated vascular ET-1 secretion is specific for Tx-AA and chronic rejection and not the result of generalized aberrant expression of ET-1 in donor renal vessels.

Figure 2.

Immunoreactive ET-1 in the vasculature was elevated in chronic rejection but essentially normal in acute rejection. The density of vascular ET-1 staining in representative sections from 12 patients with chronic rejection and 11 patients with acute rejection was analyzed using NIH Image. Data are mean SEM in relative densitometric units.

Full figure and legend (11K)

We next analyzed the relationship between vascular ET-1 levels in Tx-AA and various clinical and pathological parameters measured in these subjects (that is, Russell index of Tx-AA and parameters listed in Table 1). A positive correlation was observed between vascular ET-1 expression and the Russell morphometric index of Tx-AA Figure 3, r = 0.78, P < 0.01), which primarily grades the extent of neointima formation. An index of hypertension (Table 1) in patients with chronic rejection also correlated with ET-1 expression Figure 3; r = 0.82, P < 0.01). Strong correlations were not observed between serum creatinine concentration, initial type of kidney disease, age, sex, or years of graft survival.

Figure 3.

Vascular expression of ET-1 in Tx-AA correlates with the Russell morphometric index of Tx-AA and with an index of hypertension. Densitometric values of ET-1 expression in the vasculature of patients with Tx-AA were correlated with a morphometric index that grades the severity of Tx-AA (A) and with an index that assess the severity of hypertension (B; Tables 1 and 2). Significant correlations (P < 0.01) were observed in both analyses. In A, r = 0.78, and in B, r = 0.82.

Full figure and legend (17K)

Multiple cell types in the neointima stain positively for ET-1

We next asked what cell types in the neointima of Tx-AA stain positively for ET-1. Previous studies in animal models and humans have reported that the neointima of Tx-AA contains variable numbers of endothelial cells, vascular smooth muscle cells, macrophages, and T cells^1,10. Sufficient tissue existed from three patients with chronic rejection to stain serial sections for ET-1 and the following cell type-specific markers: Factor VIII (endothelial cells), smooth muscle -actin (vascular smooth muscle), CD-68 (macrophage/monocyte), and CD-3 (T cells). A semiquantitative analysis was then performed in which the total number of a specific cell type was calculated as a percentage of all cells in the vessel, and these cells were further scored for ET-1 staining. As shown in Table 3, endothelial cells comprised 8% of neointimal cells and always stained positive for ET-1. As shown above Figure 1c, the endothelium is largely intact in Tx-AA. As expected, vascular smooth muscle cells were the most populous cell type (that is, 75%) but only 35% stained positive for ET-1. Infiltrating macrophages/monocytes and T cells were also present (8 and 3%, respectively; Table 3), but positive ET-1 staining was detected in only 31% of macrophage/monocytes and was never observed in T cells. Approximately 6% of cells failed to stain for any cell type-specific marker and were thus unidentified; however, these cells were not ET-1-positive. Collectively, these results suggest that nearly all endothelial cells and approximately one-third of smooth muscle cells and macrophage/monocytes elaborate ET-1 in the neointima. Considerable heterogeneity clearly exists in the propensity of smooth muscle cells and macrophage/monocytes to elaborate ET-1.

Table 3 - Composition of cells in the neointima and percentage of these cells elaborating ET-1 immunoreactivity.

Full table (44K)

Macrophage-associated cytokines stimulate ET-1 secretion

To gain insight into possible stimuli of ET-1 secretion in Tx-AA, we tested the ability of alloreactive cytokines to up-regulate ET-1 peptide secretion in cultured endothelial cells from human umbilical veins. We chose to test cytokines that have previously been shown by Nadeua and coworkers^16,18 to be elevated shortly before the burst of ET-1 secretion in a rat model of chronic renal allograft rejection: T cell (IL-2, IL-4, IFN)- and macrophage (IL-1, IL-6, and TNF-)-associated cytokines. We observed that IL-1 was a potent stimulus of ET-1 secretion with significant stimulation occurring at three hours and continuing in a time-dependent manner at five hours (Figure 4 and Table 4). IL-6 and TNF- stimulated a modest and slower increase in ET-1 secretion at five hours (Figure 4 and Table 4). Interestingly, the T cell-associated cytokines IL-2, IL-4, and IFN failed to stimulate ET-1 secretion (Table 4). Twenty-four hour incubations with IL-2, IL-4, and IFN failed to significantly elevate ET-1 above basal levels in HUVEC and vascular smooth muscle cells (N = 3 in duplicate, data not shown). We used 12–0–tetradecanoyl phorbol 13 acetate (TPA) as a positive control in each experiment because previous experiments have shown that protein kinase C is a potent stimulus of ET-1 secretion, and in all experiments TPA stimulated an approximately threefold increase in ET-1 secretion (Figure 4 and Table 4).

Figure 4.

Interleukin-1 (IL-1) stimulates secretion of ET-1 from human umbilical vein endothelial cells (HUVEC) in culture. HUVEC cells in serum-free media were stimulated for the times indicated with 10 ng/ml IL-1, 10 ng/ml IL-6, or 100 nM TPA. The supernatants were then collected and extracted ET-1 was measured by ELISA. Controls were treated with media alone. Data are mean SEM from three independent experiments in duplicate. Symbols are: () TPA; () IL-1; () IL-6; () control; **P < 0.01; *P < 0.05.

Full figure and legend (13K)

Table 4 - Stimulation of ET-1 secretion by cytokines in human endothelial and vascular smooth muscle cells in culture.

Full table (44K)

Because the immunohistochemical results presented above demonstrate that some vascular smooth muscle cells stain positive for ET-1 in Tx-AA, we also asked whether T cell- or macrophage-associated cytokines would stimulate ET-1 secretion in these cells in culture. Only IL-1 and IL-6 significantly elevated ET-1 secretion in cultured human vascular smooth muscle cells (Table 4); however, as expected, levels of ET-1 secretion in vascular smooth muscle were always significantly less than in endothelial cells. TNF and the T cell-derived cytokines did not elevate ET-1 secretion even though TPA as a positive control always stimulated ET-1 secretion. These experiments suggest that macrophage-associated cytokines such as IL-1, IL-6, and TNF might be important stimuli of ET-1 secretion in Tx-AA.

Glomerular expression of ET-1 is elevated in Tx-AGS

Because glomeruli and mesangial cells are biological targets for ET-1, we investigated whether glomerular staining for ET-1 was elevated in Tx-AGS. Focal and segmental ET-1 staining was observed in patients (Table 1) with chronic rejection and Tx-AGS (arrows, Figure 5a). Although it was difficult to establish the precise cellular pattern of ET-1 staining in Tx-AGS, it clearly contrasted with the primarily segmental endothelial pattern of staining observed in normal human kidney (arrows, Figure 5b). Semiquantitative analysis of glomerular ET-1 staining in Tx-AGS revealed a 2.7-fold increase Figure 6. In patients with acute rejection (Table 2), glomerular ET-1 staining was essentially normal or slightly depressed Figure 6. It is also important to note that the glomerular endothelium was largely intact in Tx-AGS and stained uniformly for Factor VIII (arrows, Figure 5c). Experiments with non-immune serum confirmed that glomerular immunohistochemical staining for ET-1 in Tx-AGS was highly specific Figure 5d.

Figure 5.

Glomerular ET-1 staining is modestly elevated in patients with Tx-AGS and chronic rejection. Immunohistochemical staining for ET-1 revealed a focal and segmental glomerular localization (arrowheads) in Tx-AGS (A), which contrasts with the primarily focal endothelial localization (arrowheads) in normal kidney (B). (C) Staining for Factor VIII showed an intact glomerular endothelium (arrowheads) in Tx-AGS. (D) Control staining in Tx-AGS with non-immune serum.

Full figure and legend (169K)

Figure 6.

Glomerular ET-1 is elevated in patients with Tx-AGS and chronic rejection but not in patients with acute renal allograft rejection. Semiquantitative analysis of glomerular ET-1 peptide expression was performed as described for Figure 2 with NIH Image in the same patients with chronic and acute rejection described in Tables 1 and 2. Data are mean SEM.

Full figure and legend (9K)

The same T cell- and macrophage-associated cytokines tested earlier in endothelial and vascular smooth muscle cell cultures were also tested for their ability to stimulate ET-1 secretion in cultured human mesangial cells. As before, only the macrophage-associated cytokines IL-1 and IL-6 stimulated ET-1 secretion, and in mesangial cells IL-6 was significantly more potent that IL-1 (Table 4).

DISCUSSION

The question addressed by the present study was whether ET-1 secretion is elevated in the vasculature of patients with Tx-AA and chronic renal allograft rejection. The main finding is that ET-1 secretion is elevated in the vascular neointima in Tx-AA and chronic rejection compared with the vasculature in acute rejection or normal kidney. Elevated levels of ET-1, a vasoconstrictor peptide with potent mitogenic actions, might contribute to the fibroproliferative neointima that characterizes Tx-AA in chronic renal allograft rejection.

ET-1 is elevated in the neointima of Tx-AA

Our immunohistochemical analysis of vascular ET-1 in patients with Tx-AA and chronic rejection showed that ET-1 staining was elevated 6.1-fold over vascular staining in an age-matched population of patients with acute rejection and without pathological evidence of Tx-AA. ET-1 in Tx-AA was mostly localized to the neointima and was only rarely demonstrated in the medial layer. In some vessels significant immunoreactive ET-1 was also localized in the adventitia. Focal inflammation in the adventitia has previously been described in Tx-AA and might account for the occasionally dense staining for ET-1 we observed. It is important to note, however, that the most common finding was elevated ET-1 in the neointima as contrasted with the primarily intimal expression of ET-1 in renal vessels from patients with acute rejection and in normal kidney. It is also striking that renal venous expression of ET-1 was apparently unchanged in patients with chronic rejection, suggesting that the preproET-1 gene is differentially regulated in arteries and veins in Tx-AA.

We also determined that elevated ET-1 in Tx-AA was not a non-specific consequence of transplantation as vascular ET-1 was unchanged in patients undergoing acute rejection. Our finding of normal or slightly lower vascular ET-1 in acute rejection agrees with an earlier study by Watschinger and coworkers in which ET-1 levels were not elevated in renal transplant patients with acute rejection compared with normal kidney²¹. Although vascular ET-1 is not apparently elevated in renal transplants with acute rejection, it seems quite likely that ET-1 may be elevated immediately following transplantation where it might contribute to renal damage subsequent to ischemia and/or reperfusion. We did not have the opportunity to analyze ET-1 levels in renal tissue immediately following transplantation, but rat models of cardiac and renal transplantation show a rapid, transient burst of ET-1 immediately following surgery^14,17. In addition, ET-1 is well-known to increase in response to renal ischemia^12,13. However, our present results and previous studies in animals^14,15,16,17 and humans^19,25 suggest that the increase in vascular ET-1 that occurs in chronic rejection and Tx-AA is sustained. Two independent studies of patients with cardiac allografts clearly show elevated neointimal ET-1 in chronic rejection and Tx-AA^19,25. The pattern of neointimal ET-1 in these studies is quite similar to our localization of ET-1 in renal Tx-AA. Studies in rats with chronic rejection of cardiac^14,15 and renal^16,17 allografts also document selective and sustained elevation of ET-1 in chronic rejection. ET-1 was not localized in the renal studies^16,17, but ET-1 staining in the cardiac models^14,15 was analogous to our results in human renal vessels.

We also found a strong correlation between ET-1 staining in the neointima and progression of neointima growth (that is, Russell index) and hypertension. The correlation between ET-1 staining and growth of the neointima is suggestive of a mitogenic role for ET-1 in Tx-AA. Elevated secretion of ET-1 in the neointima is not specific for Tx-AA and has previously been shown in other diseases characterized by vascular neointima including atherosclerosis, coronary artery restenosis, and hypertensive arterial remodeling^3,30,31. Although these correlations do not prove a pathological role for ET-1, the correlation between ET-1 and hypertension is striking given the potent pressor actions of ET-1 and the prevalence of hypertension in patients with chronic rejection. Results of another recent study in human renal allograft recipients supports the hypothesis that post-transplant diastolic hypertension is a result of ET-1-mediated vasoconstriction³². The mechanisms by which ET-1 might cause hypertension in these patients has not been addressed in the present study. However, ET-1 in humans has been shown to have a potent antinatriuretic action that increases effective circulating volume³³. It is also possible that spillover of ET-1 into the circulation might have a pressor effect similar to mechanisms proposed for ET-1-induced hypertension in other settings³¹.

Another pathological hallmark of chronic renal allograft rejection is Tx-AGS. In our studies ET-1 staining was also elevated in Tx-AGS in a focal and segmental pattern. It was, however, difficult to establish the precise staining pattern in Tx-AGS and the increase over normal glomerular ET-1 staining (2.7-fold) was less than that for Tx-AA. Mesangial hypercellularity and matrix expansion are common in Tx-AGS, and given the well-documented mitogenic effects of ET-1 on mesangial cells it seems possible that elevated ET-1 might contribute to the pathogenesis of Tx-AGS. Further experiments are necessary to investigate this hypothesis.

Multiple cells express ET-1 in the neointima of Tx-AA

Previous experiments have shown that the neointima of Tx-AA is a heterogeneous mixture of cell types including endothelial cells, migrating vascular smooth muscle cells, and lesser numbers of infiltrating macrophages and T cells^1,10. Our present results confirm these studies in that the neointima in renal Tx-AA was also populated with vascular smooth muscle cells > endothelial cells = macrophages > T cells. We also found a small percentage of immunohistochemically unidentifiable cells, which has also been reported in other studies¹. The identity and source of these cells is at present unknown. In our studies endothelial cells were uniformly positive for ET-1 and T cells were negative. In contrast, vascular smooth muscle cells and macrophages displayed considerable heterogeneity in ET-1 staining: 35% of vascular smooth muscle cells and 31% of macrophage/monocytes elaborated ET-1. The reasons for this heterogeneity in ET-1 staining are unclear but might reflect differences in activation state, phenotype, or cell cycle phase.

The results of our experiments to identify ET-1-positive cells in the neointima of Tx-AA are consistent with a recent study in human cardiac allografts with chronic rejection in which ET-1 staining was predominately in endothelial cells and vascular smooth muscle cells with considerable heterogeneity observed in macrophages²⁵. However, our findings differ somewhat from a previous study in rat cardiac allografts in which mononuclear inflammatory cells were the major cell type expressing ET-1¹⁴. The reasons for this apparent discrepancy are not clear but might reflect differences in species and organ (human vs. rat and renal vs. cardiac) or in the antibodies used to identify monocytes and macrophages (CD-68 vs. ED-1). Most important, however, all of the published studies document macrophage infiltration in the neointima of Tx-AA and point to the potential importance of macrophages as a source of ET-1 in Tx-AA.

Regulation of ET-1 secretion in Tx-AA

If elevated secretion of ET-1 is playing an important role in neointima formation in Tx-AA, what might be possible stimuli of ET-1 secretion by endothelial cells? In a rat model of chronic renal allograft rejection, Tilney and coworkers^16,17,18 have shown that T cell- and macrophage-derived cytokines are elevated immediately preceding the sustained burst of ET-1 secretion. We therefore asked whether T cell- or macrophage-derived cytokines might stimulate ET-1 secretion in cultured human endothelial cells. We found that macrophage-derived cytokines IL-1, IL-6, and TNF- stimulated ET-1 secretion but that IL-1 was considerably more active. The three T cell-derived cytokines (IL-2, IL-4, IFN) we tested were inactive, which support a predominant role of macrophage-derived cytokines to stimulate ET-1 secretion by endothelial cells. It is interesting to speculate that the lack of macrophage infiltration in renal veins might explain, in part, the normal expression of ET-1 in the venous endothelium in chronic rejection. Although it is sometimes difficult to extrapolate results from cultured cells to the same cells in vivo, Hancock et al⁷ have previously observed persistent and intense expression of IL-1, IL-6, and TNF- in glomerular lesions of Tx-AGS. They proposed that IL-6 in particular, which is a potent activator of mesangial cells, might mediate mesangial expansion and proliferation in Tx-AGS. It is therefore interesting that in the present study IL-6 was identified as a potent stimulus of ET-1 secretion for mesangial cells in culture. IL-6 was in fact more potent than IL-1 in mesangial cells. Collectively these results provide further evidence that IL-6 might be an important mediator of Tx-AGS.

It is important to note, however, that our results do not prove that macrophage-derived cytokines are the relevant stimuli of ET-1 secretion by endothelial cells in Tx-AA. In fact many other stimuli might contribute including thrombin, fluid mechanical shear stress, oxidized LDL, and the persistent hypercholesterolemia observed in transplant recipients receiving corticosteroids. Another confounding factor is the ability of cyclosporine to stimulate ET-1 secretion in cultured endothelial cells and in intact kidney^34,35. A role for ET-1 has also been shown in cyclosporine-induced renal vasoconstriction and mesangial cell contraction^36,37,38,39. But several lines of evidence appear to contradict the notion that cyclosporine mediates the increase of ET-1 in Tx-AA. First, in a cohort of 25 renal allograft patients receiving cyclosporine but having stable renal function, plasma ET-1 levels were normal and not elevated as would be predicted if cyclosporine were stimulating ET-1 secretion⁴⁰. Second, in a previous study²¹ and in our present study renal transplant recipients with acute rejection had normal or slightly lower levels of tissue ET-1 despite receiving high doses of cyclosporine. Third, and perhaps most impressive, in a rat model of chronic cardiac allograft rejection that did not employ cyclosporine or any other form of immunosuppresssion, ET-1 levels in the neointima were still elevated¹⁴. Thus, the presence of cyclosporine might not be necessary for enhanced ET-1 secretion, but it might certainly contribute in the presence of other primary stimuli of ET-1 secretion.

In summary, the observations presented here, taken together with earlier studies showing elevated ET-1 in animal models of Tx-AA, suggest that increased secretion of ET-1 might contribute to the abnormal growth of cells in the neointima leading to Tx-AA and chronic rejection. Since recent studies indicate that agents that antagonize ET-1 actions might be of value in attenuating vascular cell growth and neointima formation, it will be of interest to test whether Tx-AA and chronic rejection can be improved by therapies aimed at neutralization of ET-1.

References

1. GORDON, D, Transplant Arteriosclerosis, inAtherosclerosis and Coronary Artery Disease, 1996 edited by FUSTER V, ROSS R, TOPOL EJ, Philadelphia, Lippincott-Raven, p715
2. TILNEY, NL, WHITLEY, WD, DIAMOND, JR, KUPIEC-WEGLINSKI, JW, ADAMS, DH: Chronic rejection–An undefined conundrum. Transplantation 1991 52: 389–398,  | PubMed | ISI | ChemPort |
3. WATSCHINGER, B, SAYEGH, MH: Endothelin in organ transplantation. Am J Kidney Dis 1996 27: 151–161,  | PubMed |
4. PORTER, KA, Renal Transplantation, in Pathology of the Kidney, 1992 edited by HEPTINSTALL RH, Boston, Little Brown, p 1799
5. HOSTETTER, TH: Chronic transplant rejection. Kidney Int 1994 46: 266–279,  | PubMed | ISI | ChemPort |
6. BIA, MJ: Nonimmunologic causes of late graft loss. Kidney Int 1995 47: 1470–1480,  | PubMed | ISI | ChemPort |
7. HANCOCK, WH, WHITLEY, W, TULLIUS, S, HEEMANN, U, WASOWSKA, B, BALDWIN, W, TILNEY, N: Cytokines, adhesion molecules, and the pathogenesis of chronic rejection of rat renal allografts. Transplantation 1993 56: 643–650,  | PubMed | ISI | ChemPort |
8. KASISKE, BL, KALIL, RSN, LEE, HS, RAO, KV: Histopathologic findings associated with a chronic, progressive decline in renal allograft function. Kidney Int 1991 40: 514–524,  | PubMed | ISI | ChemPort |
9. SALOMON, RN, HUGHES, CCW, SCHOEN, FJ, PAYNE, DD, POBER, JS, LIBBY, P: Human coronary transplantation-associated arteriosclerosis: Evidence for a chronic immune reaction to activated graft endothelial cells. Am J Pathol 1991 138: 791–798,  | PubMed |
10. SHI, C, LEE, W-S, HE, Q, ZHANG, D, FLETCHER, DL, NEWELL, JB, HABER, E: Immunologic basis of transplant-associated arteriosclerosis. Proc Natl Acad Sci USA 1996 93: 4051–4056,  | Article | PubMed | ChemPort |
11. TULLIUS, SG, HEEMANN, U, HANCOCK, WW, AZUMA, H, TILNEY, NL: Long-term kidney isografts develop functional and morphologic changes that mimic those of chronic allograft rejection. Ann Surg 1994 4: 425–435,
12. SIMONSON, MS: Endothelins: Multifunctional renal peptides. Physiol Rev 1993 73: 375–411,  | PubMed | ISI | ChemPort |
13. KOHAN, DE: Endothelins and the kidney: Physiology and pathophysiology. Am J Kidney Dis 1993 22: 493–510,  | PubMed | ISI | ChemPort |
14. WATSCHINGER, B, SAYEGH, MH, HANCOCK, WW, RUSSELL, ME: Up-regulation of endothelin-1 mRNA and peptide expression in rat cardiac allografts with rejection and arteriosclerosis. Am J Pathol 1995 146: 1065–1072,  | PubMed | ISI | ChemPort |
15. FORBES, R, CERNACEK, P, ZHENG, S-X, GOMERSALL, M, GUTTMANN, R: Increased endothelin expression in a rat cardiac allograft rejection model of chronic vascular rejection. Transplantation 1996 61: 791–797,  | PubMed | ISI | ChemPort |
16. NADEAU, KC, AZUMA, H, WATSCHINGER, B, SAYEGH, MH, TILNEY, NL: Endothelin and cytokine gene expression in rat renal allografts undergoing chronic failure. (abstract) J Am Soc Nephrol 1995 6: 666,
17. AZUMA, H, NADEAU, K, TAKADA, M, MACKENZIE, HS, TILNEY, NL: Cellular and molecular predictors of chronic renal dysfunction after initial ischemia/reperfusion injury of a single kidney. Transplantation 1997 64: 190–197,  | Article | PubMed | ISI | ChemPort |
18. NADEAU, KC, AZUMA, H, TILNEY, NL: Sequential cytokine dynamics in chronic rejection of rat renal allografts: Roles for cytokines and RANTES and MCP-1. Proc Natl Acad Sci 1995 92: 8729–8733,  | PubMed |
19. GIAD, A, SALEH, D, YANAGISAWA, M, CLARKE FORBES, RD: Endothelin-1 immunoreactivity and mRNA in the transplanted human heart. Transplantation 1995 59: 1308–1313,  | PubMed |
20. AARNIO, P, TUKIAINEN, P, TASKINEN, E, HARJULA, A, FYHRQUIST, F: Endothelin in bronchoalveolar lavage fluid is increased in lung-transplanted patients. Scand J Thor Cardiovasc Surg 1996 30: 113–116,
21. WATSCHINGER, B, VYCHYTIL, A, ATTAR, M, WAGNER, D, SCHULLER, M, HARTTER, E, ULRICH, W: Pattern of endothelin immunostaining during rejection episodes after kidney transplantation. Clin Nephrol 1994 41: 86–93,  | PubMed |
22. SCHERSTEN, H, HEDNER, T, MCGREGOR, C, MILLER, VD, MARTENSSON, G, RIISE, G, NILSSON, FN: Increased levels of endothelin-1 in bronchoalveolar lavage fluid of patients with lung allografts. J Thorac Cardiovasc Surg 1996 111: 253–258,  | PubMed |
23. WATSCHINGER, B, VYCHYTIL, A, SCHULLER, M, HARTTER, E, TRAINDL, O: A pathophysiologic role of endothelin in acute vascular rejection after renal transplantation. Transplantation 1991 52: 743–746,  | PubMed | ISI | ChemPort |
24. SCHILLING, M, HOLZINGER, F, FRIESS, H, SEILER, C, BUCHLER, MW: Pathogenesis of delayed kidney graft function: Role of endothelin-1, thromboxane B₂, and leukotriene B₄. Transplant Proc 1996 28: 304–305,  | PubMed |
25. RAVALLI, S, SZABOLES, M, ALBALA, A, MICHLER, RE, CANNON, PJ: Increased immunoreactive endothelin-1 in human transplant coronary artery disease. Circulation 1996 94: 2096–2102,  | PubMed |
26. SOLEZ, K, AXELSEN, R, BENEDIKTSSON, H, BURDICK, J, COHEN, A, COLVIN, R, CROKER, B, DROZ, D, DUNNILL, M, HALLORAN, P, HAYRY, P, JENNETTE, J, KEOWN, P, MARCUSSEN, N, MIHATSCH, M, MOROZUMI, K, MYERS, BD, NAST, C, OLSEN, S, RACUSEN, L, RAMOS, E, ROSEN, S, SACHS, D, SALOMON, D, SANFILIPPO, F, VERANI, R, VONWILLEBRAND, E, YAMAGUCHI, Y: International standardization of criteria for the histologic diagnosis of renal allograft rejection: The Banff working classification of kidney transplant pathology. Kidney Int 1993 44: 411–422,  | PubMed | ISI | ChemPort |
27. RUSSELL, PS, CHASE, CM, WINN, HJ, COLVIN, RB: Coronary atherosclerosis in transplanted mouse hearts: Time course and immunogenetic and immunopathological considerations. Am J Pathol 1994 144: 260–274,  | PubMed | ISI | ChemPort |
28. TAKAHASI, K, SAWASAKI, Y: Rare spontaneously transformed human endothelial cell line provides useful research tool. In Vitro Cell Dev Biol 1992 28A: 380–382,
29. SCHULTZ, PJ, DICORLETO, PE, SILVER, BJ, ABBOUD, HE: Mesangial cells express PDGF mRNAs and proliferate in response to PDGF. Am J Physiol 1988 255: F674–F684,  | PubMed | ISI | ChemPort |
30. KOHAN, DE: Endothelins in the normal and diseased kidney. J Kidney Dis 1997 29: 2–26,
31. LEVIN, ER: Endothelins. N Engl J Med 1995 333: 356–363,  | Article | PubMed | ISI | ChemPort |
32. KIRK, AD, JACOBSON, LM, HEISEY, DM, FASS, NA, SOLLINGER, HW, PIRSCH, JD: Posttransplant diastolic hypertension. Transplantation 1997 64: 1716–1720,  | Article | PubMed | ISI | ChemPort |
33. RABELINK, TJ, KAASJAGER, KAH, BOER, P, STROES, EG, BRAAM, B, KOOMANS, HA: Effects of endothelin-1 on renal function in humans: Implications for physiology and pathophysiology. Kidney Int 1994 46: 376–381,  | PubMed | ISI | ChemPort |
34. BUNCHMAN, TE, BROOKSHIRE, CA: Cyclosporine-induced synthesis of endothelin by cultured human endothelial cells. J Clin Invest 1991 88: 310–314,  | PubMed | ISI | ChemPort |
35. IWASAKI, S, HOMMA, T, KON, V: Site specific regulation in the kidney of endothelin and its receptor subtypes by cyclosporine. Kidney Int 1994 45: 592–597,  | PubMed | ISI | ChemPort |
36. PERICO, N, DADAN, J, REMUZZI, G: Endothelin mediates the renal vasoconstriction induced by cyclosporine in the rat. J Am Soc Nephrol 1990 1: 76–83,  | PubMed | ISI | ChemPort |
37. CONGER, JD, LANESE, DM: Effects of endothelin receptor antagonist on cyclosporine-induced vasoconstriction in isolated rat renal arterioles. J Clin Invest 1993 91: 2144–2149,  | PubMed | ChemPort |
38. TAKEDA, M, BREYER, MD, NOLAND, TD, HOMMA, T, HOOVER, RL, INAGAMI, T, KON, V: Endothelin-1 receptor antagonist: Effects on endothelin- and cyclosporine-treated mesangial cells. Kidney Int 1992 42: 1713–1719,
39. KON, V, SUGIURA, M, INAGAMI, T, HARVIE, BR, ICHIKAWA, I, HOOVER, RL: Role of endothelin in cyclosporine-induced glomerular dysfunction. Kidney Int 1990 37: 1487–1491,  | PubMed | ISI | ChemPort |
40. STOCKENHUBER, F, GOTTSAUNER-WOLF, M, MAROSI, L, LIEBISCH, B, KURZ, RW, BALCKE, P: Plasma levels of endothelin in chronic renal failure and after renal transplantation: Impact on hypertension and cyclosporin A-associated nephrotoxicity. Clin Sci 1992 82: 255–258,  | PubMed |

Acknowledgments

This study was supported by grant DK-46939 from the National Institutes of Health, the Dunlop Eagles Kidney Research Fund, and the Centers for Dialysis Care Research Fund. We gratefully acknowledge the excellent technical assistance of Kelly Ferguson, Angela Hunt, Bill Herman, and Yuan Wang. We also thank Julie Wolfe and Arif Nawaz for preparation of cultured human mesangial cells and Mary Russell and Mohamed Sayegh for thoughtful discussions regarding the potential role of ET-1 in chronic rejection.