• Alloantibody;
  • arteritis;
  • Banff lesions;
  • graft rejection


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The natural history and pathogenesis of the pathologic lesions that define rejection of kidney transplants have not been well characterized. We studied the evolution of the pathology of rejection in mouse kidney allografts, using four strain combinations across full major histocompatibility complex (MHC) plus nonMHC disparities, to permit more general conclusions. Interstitial infiltrate, MHC induction, and venulitis appeared by day 5, peaked at day 7–10, then stabilized or regressed by day 21. In contrast, tubulitis, arteritis, and glomerulitis were absent or mild at days 5 and 7, but progressed through day 21, indicating separate regulation and homeostatic control of these lesions. Edema, hemorrhage, and necrosis also increased through day 21. All lesions were T-dependent, failing to develop in T-cell-deficient hosts. Allografts into immunoglobulin-deficient hosts manifested typical infiltration, MHC induction, and tubulitis at days 7 and 21, indicating that these lesions are alloantibody-independent. However at day 21 kidneys rejecting in immunoglobulin-deficient hosts showed decreased edema, arteritis, venulitis, and necrosis.

Thus the three groups of lesions are: T-cell-mediated interstitial infiltration, MHC induction, and venulitis, which develops rapidly then stabilizes; slower but progressive T-cell-mediated tubulitis and arteritis; and late antibody-mediated endothelial injury, which contributes to late edema, arteritis, and venulitis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Despite recent successes in kidney transplantation, rejection remains a strong predictor of long-term graft survival (1). The Banff consensus process (2,3) and the CCTT process (4) are defined systems for evaluating the histologic lesions of human kidney transplant rejection. The Banff classification is now widely used, and grades interstitial infiltrate, tubulitis, arteritis and glomerulitis as the main histologic features of rejection. The key definition of rejection is the invasive lesions (tubulitis, endothelial arteritis) (5) rather than interstitial inflammation alone (6–15). The classification of rejection lesions permitted clinicians to use biopsy-proven rejection rather than graft failure as an end point. In contrast, studies of rejection in animal models remain focused on graft survival rather than pathologic lesions. This contributes to our limited understanding of the immune mechanisms mediating lesions of allograft rejection. Alloreactive cytotoxic T cells and alloantibodies are demonstrable but their relationship to specific lesions is unclear. The resolution of the molecular pathogenesis of graft rejection has also been complicated by apparent mechanistic differences among model systems: rejection mechanisms in some models are atypical and probably unrelated to those in humans (16).

Therefore in the present study we evaluated the evolution of the lesions of mouse kidney allograft rejection across full major histocompatibility complex (MHC) and minor disparities. We used several common strains of laboratory mice, in both a life-supporting model (i.e. no remaining host kidney) and a nonlife-supporting model (i.e. the contralateral host kidney remained in situ.) The nonlife-supporting models allowed us to study the development of rejection histopathology over 21 days without premature death of animals from hyperkalemia or uremia, and with low morbidity to optimize animal well being. We also used mice lacking T cells (nude mice) or immunoglobulin to explore the relative contribution of T cells and alloantibody. The results are compatible with a model of kidney rejection with distinct components, including T-cell-mediated interstitial infiltration, T-cell-mediated invasive lesions and alloantibody-mediated microvascular injury. The results suggest that these lesions may be independently regulated.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References


Male CBA/J (CBA), C57Bl/6 (B6), 129Sv/J (129), B6.129S2-Igh-6tm1/Cgn (IgKO), BALB/c mice and nude mice were obtained from the Jackson Laboratory (Bar Harbor, ME). Mice were maintained in the Health Sciences Laboratory Animal Services at the University of Alberta and were kept on acidified water. Mice maintenance and experiments were in conformity with approved animal-care protocols. The following mice strain combinations were studied across full MHC and nonMHC disparities: CBA (H-2K, I-Ak) into BALB/c (H-2D, I-Ad), 129 (H-2KbDb, I-Ab) into CBA, C57Bl/6 (B6; H-2KbDb, I-Ab) into CBA, and CBA into B6. Syngeneic transplants (129 into 129) were also performed and served as controls for life-supporting allografts. A total of 149 allografts, including 15 IgKO mice and seven syngeneic grafts, were evaluated in this study. Number of animals used in the study groups is given in Tables 1–5.

Table 1.  Weights of nonlife-supporting and life-supporting kidney transplants
Donor kidneyCBA into BALB/c129 into CBAB6 into CBACBA into B6B6 into CBA129 into 129
  1. Right kidneys were transplanted heterotopically, either into hosts with the contralateral kidney in situ (nonlife-supporting) or into nephrectomized hosts (life-supporting). Wet tissue weights (milligrams) are given as means ± SE. Number of mice used in the experiments is shown in parentheses.

  2. 1Significantly different from control, p < 0.05 (Mann–Whitney).

  3. N/A = not available.

Control161 ± 29 (22)145 ± 53 (18)127 ± 37 (21)181 ± 24 (26)127 ± 37 (21)171 ± 19 (7)
Day 5327 ± 47 (4)1279 ± 30 (5)1302 ± 57 (8)1NA216 ± 4 (2)1N/A
Day 7304 ± 66 (16)1295 ± 58 (23)1346 ± 51 (13)1324 ± 26 (5)1325 ± 48 (13)1186 ± 29 (7)
Day 21468 ± 68 (4)1380 ± 60 (5)1430 ± 137 (5)1521 ± 129 (10)1N/AN/A
Day 42N/AN/AN/A403 ± 136 (7)1N/AN/A
Table 2.  Function of nonlife-supporting, life-supporting and syngeneic mouse kidney transplant models
 CBA into BALB/c129 into CBAB6 into CBAMean ± SEB6 into CBA129 into 129
  1. Right kidneys were transplanted heterotopically, either into hosts with the contralateral kidney in situ (nonlife-supporting) or into nephrectomized hosts (life-supporting). Values (μmol/L) are given as means ± SE. Number of mice used in experiments is shown in parentheses.

  2. 1Significantly different from day 0, p < 0.05 (Mann–Whitney).

  3. N/A = not available.

Serum creatinine (μmol/L)
 Day 030 ± 2 (2)28 ± 0 (4)  29 ± 4 (9)29 ± 0.929 ± 4 (9)28 (1)
 Day 729 ± 2 (8)N/AN/A29 ± 0.778 ± 661 (10)34 ± 9 (4)
 Day 21N/A30 ± 2 (5)  31 ± 3 (3)30 ± 0.8N/AN/A
Serum Urea (μmol/L)
 Day 05.9 ± 0.88.1 ± 1.17.8 ± 2.1 7.6 ± 0.57.8 ± 2.19.3
 Day 78.4 ± 1.7N/AN/A 8.4 ± 0.643 ± 21112.2 ± 4.0
 Day 21N/A9.2 ± 2.713.6 ± 6.510.9 ± 1.6N/AN/A
Table 3.  Histopathology of rejecting kidney transplants
HistologyDay post transplantCBA into Balb/c129 into CBAB6 into CBACBA into B6Mean ± SE
  1. Right kidneys were transplanted heterotopically into hosts with the contralateral kidney in situ (nonlife-supporting). Interstitial infiltrate, tubulitis, glomerulitis necrosis, edema and peritubular capillary congestion (PTC) were scored from 0 to 3 based on the percentage of parenchymal involvement, with 0 representing no changes, 1 representing less than 25%, 2 representing 25–75%, and 3 representing more than 75% of the total parenchyma involved. Arteritis and venulitis lesions were counted in each specimen and the mean number of affected vessels per kidney section was calculated. Values are means ± SE. Number of mice used in the experiments is shown in parentheses. For clarity, the number of mice is entered for the interstitial infiltrate only.

  2. 1Significant difference compared with day 5 and 2day 7.

  3. p < 0.05 (Mann–Whitney).

  4. N/A = not available.

Interstitial infiltrate 51.5 ± 0.6 (4)1.8 ± 0.4 (5)1.3 ± 0.5 (8)N/A1.5 ± 0.1
  71.2 ± 0.7 (14)2.4 ± 0.6 (27)2.0 ± 0.8 (28)2.2 ± 0.4 (5)2.0 ± 0.11
 103.0 ± 0.0 (10)3.0 ± 0.0 (5)N/A2.5 ± 0.7 (2)2.9 ± 0.11,2
 211.3 ± 0.5 (4)1.4 ± 0.5 (5)2.6 ± 0.5 (5)2.3 ± 0.7 (10)2.0 ± 0.21
Venulitis 51.0 ± 1.21.4 ± 0.50.5 ± 0.5N/A0.9 ± 0.2
  71.0 ± 0.91.9 ± 1.01.5 ± 0.90.6 ± 0.51.5 ± 0.1
 100.7 ± 0.71.0 ± 0.7N/A1.0 ± 0.00.8 ± 0.22
 210.3 ± 0.50.8 ± 0.81.2 ± 0.80.9 ± 0.60.8 ± 0.12
Tubulitis 50.0 ± 0.00.2 ± 0.40.1 ± 0.4N/A0.1 ± 0.1
  70.7 ± 1.00.7 ± 0.70.8 ± 0.60.4 ± 0.50.7 ± 0.11
 101.6 ± 0.50.4 ± 0.5N/A1.0 ± 0.01.1 ± 0.21,2
 212.3 ± 0.52.8 ± 0.43.0 ± 0.02.8 ± 0.62.8 ± 0.11,2
Arteritis 50.0 ± 0.00.0 ± 0.00.0 ± 0.0N/A0
  70.6 ± 0.80.0 ± 0.20.1 ± 0.30.0 ± 0.00.2 ± 0.1
 101.3 ± 1.20.4 ± 0.9N/A0.5 ± 0.70.9 ± 0.31,2
 211.5 ± 1.00.8 ± 0.40.6 ± 0.51.3 ± 0.81.1 ± 0.21,2
Glomerulitis 50.5 ± 0.60.1 ± 0.30.3 ± 0.5N/A0.2 ± 0.1
  70.4 ± 0.50.7 ± 0.51.3 ± 0.90.4 ± 0.50.8 ± 0.11
 101.1 ± 0.31.0 ± 0.0N/A1.5 ± 0.71.1 ± 0.11,2
 211.0 ± 0.82.0 ± 0.72.2 ± 1.12.8 ± 0.42.2 ± 0.21,2
Necrosis 50.0 ± 0.00.6 ± 1.00.0 ± 0.0N/A0.3 ± 0.1
  70.0 ± 0.01.0 ± 1.20.2 ± 0.40.2 ± 0.40.5 ± 0.1
 100.6 ± 0.70.2 ± 0.4N/A0.0 ± 0.00.4 ± 0.1
 211.3 ± 1.30.2 ± 0.40.6 ± 0.9 (5)1.5 ± 0.81.0 ± 0.21,2
Edema 50.0 ± 0.00.4 ± 0.50.1 ± 0.4 (8)N/A0.2 ± 0.1
  70.2 ± 0.40.6 ± 0.80.1 ± 0.5 (16)0.2 ± 0.40.3 ± 0.1
 100.9 ± 0.61.0 ± 0.0N/A1.0 ± 0.00.9 ± 0.11,2
 212.0 ± 0.02.0 ± 1.01.8 ± 0.4 (5)2.7 ± 0.72.3 ± 0.21,2
PTC 50.0 ± 0.00.2 ± 0.40.4 ± 0.5N/A0.2 ± 0.1
  70.2 ± 0.40.9 ± 0.70.3 ± 0.50.2 ± 0.40.5 ± 0.1
 101.3 ± 0.50.8 ± 0.4N/A0.5 ± 0.71.1 ± 0.11,2
 212.0 ± 0.81.2 ± 0.80.8 ± 0.80.8 ± 1.11.1 ± 0.21,2
Table 4.  Histopathology of CBA kidney transplants in B6 and B6-IgKO hosts
 Day 7Day 21
 CBA into B6 (n = 5)CBA into B6-IgKO (n = 5)CBA into B6 (n = 10)CBA into B6-IgKO (n = 10)
  1. Control kidneys mean weight was 176 ± 15 mg of wet tissue (n = 9). Scoring system of lesions is explained in footnote to Table 3, except that necrosis and peritubular capillary congestion (PTC) are now expressed as a percentage of the observed parenchyma. Values are means ± SE. Number of mice used in the experiments is shown in parentheses.

  2. 1Significant difference between groups (one-way anova) and 2between B6 and IgKO at day 21 (Bonferroni), p < 0.05.

Interstitial infiltrate2.2 ± 0.22.4 ± 0.22.1 ± 0.22.1 ± 0.3
Venulitis10.6 ± 0.20.2 ± 0.21.3 ± 0.30.4 ± 0.22
Tubulitis11.2 ± 0.21.8 ± 0.22.5 ± 0.72.3 ± 0.7
Arteritis1001.3 ± 0.20.5 ± 0.22
Glomerulitis101.0 ± 0.32.2 ± 0.31.5 ± 0.3
Necrosis0  2 ± 217 ± 7  8 ± 6
Edema1001.3 ± 0.30.2 ± 0.12
PTC0  2 ± 217 ± 10  2 ± 1
Kidney weights1 (mg)324 ± 26307 ± 25521 ± 129238 ± 312
Table 5.  Assessment of major histocompatibility complex expression in mouse kidney transplants by radio-labeled antibody binding
 Fold increase in c.p.m.
Normalized expressionDay post transplantCBA into BALB/c129 into CBAB6 into CBACBA into B6
  1. Results are presented as the ratio of transplant-bound radioactivity to control-bound radioactivity (day 0) means ± SD. Number of mice used in experiments is shown in parentheses. For clarity the number of mice has been entered for major histocompatibility complex class I only.

  2. *Single experiment.

  3. 1Significant difference compared with day 5, 2day 7, and 3between days 10 and 21 (p < 0.05 (Mann–Whitney).

  4. N/A = not available.

Donor class ID513.2 ± 0.7 (4)15.3 ± 2.9 (5)18.2 ± 5.7 (10)N/A
 D713.0 ± 2.6 (16)17.0 ± 6.6 (23)12.0 ± 3.11 (26)11.7 ± 1.6 (5)
 D109.4 ± 1.31,2 (8)11.4*N/A11.0 ± 1.1 (2)
 D2113.8 ± 1.63 (4)4.7 ± 0.51,2 (5)10.1 ± 1.61 (5)16.6 ± 9.3 (10)
Donor class IID59.2 ± 0.75.6 ± 1.22.6 ± 2.3N/A
 D715.5 ± 2.816.4 ± 2.44.8 ± 2.1110.0 ± 1.8
 D106.9 ± 1.911.8*N/A 8.8 ± 2.5
 D2115.2 ± 7.432.3 ± 0.61,23.9 ± 1.34.1 ± 0.91,2

Renal transplantation

Donor mice 9–11 weeks of age were anesthetized and the abdomen was opened through a midline incision. The right kidney was excised and preserved in cold lactate Ringer's solution. The host mice were similarly anesthetized and the right native kidney was excised. The donor kidney was anastomozed heterotopically to the aorta inferior vena cava and bladder on the right side, usually without removing the host's left kidney (nonlife-supporting kidney transplantation). The mice were allowed to recover and were killed at days 5, 7, 10, 21 or 42 post-transplant following anesthesia and cervical dislocation. In a subset of experiments the hosts' contralateral kidneys were also removed at day 3 post-transplant (life-supporting). Mice with life-supporting transplants were sacrificed by day 6–8, when appearing severely unwell. No transplant hosts received immunosuppressive therapy. Mice with technical complications or pyelonephritis at the time of harvesting were removed from the study. To avoid occasional infections IgKO hosts (15 animals) and controls (five animals) received prophylactic antibiotics (Cefazolin, 0.3–0.5 mg per animal) to prevent wound and urinary tract infection. The other mice in this report received no antibiotics.


Tissue sections (2 μm) were stained with hematoxylin and eosin (H&E) or with periodic acid-Schiff (PAS), respectively. The lesions, such as interstitial infiltrate, tubulitis, glomerulitis and peritubular capillary congestion (PTC), were scored from 0 to 3 based on the percentage of parenchymal involvement, with 0 representing no changes, 1 representing less than 25%, 2 representing 25–75%; and 3 representing more than 75% of the total parenchyma involved. We also recorded the extent of graft necrosis and edema. Arteritis and venulitis lesions were counted in each specimen and the mean number of affected vessels per kidney section was calculated.


Hybridoma cell lines were obtained from ATCC (Rockville, MD). Cell lines producing mAb 25–9-17SII (anti I-Ad), 34–4-20S (anti H-2D), 11– (anti I-Ak), 11–4.1 (anti H-2K) and 20–8-4S (H-2KbDb) were maintained in tissue culture in our laboratory. Antibodies for radio-labeled antibody binding (RABA) assays (see below) were purified from the respective supernatants by protein A affinity chromatography. Antibody concentration was determined by a modified Lowry method, adjusted to 1 mg/mL, and preparations were kept frozen at –70 °C. Anti I-Ab was purchased from Serotec (Raleigh, NC). Radioiodination was performed by the Iodogen method (Pierce Chemical Co., Rockford, IL) (17). Anti-mouse IgG (goat) HRP-linked antibody was obtained from ICN (Costa Mesa, CA) and antimouse albumin (goat) HRP-linked antibody (Bethyl Laboratories Inc., Montgomery, TX) was from Cedarline Laboratories Limited (Hornby, ON, Canada).

Radioactive antibody-binding assay

This technique has been previously reported, and its quantitative characteristics have been described (18). Briefly, kidney transplant specimens were homogenized, washed in PBS and centrifuged at 3000 r.p.m. for 20 min. The pellets were suspended in PBS at 20 mg/mL. Next, 5 mg of extract was centrifuged and then resuspended in 100 μL of radiolabeled mAb solution (1000 c.p.m./μL) and incubated on ice for 60 min. After that time the incubates were diluted with 1 mL of PBS and centrifuged at 3000 r.p.m. for 20 min. The pellets were counted in a gamma counter and the values were corrected for the nonspecific binding.


Fresh-frozen sections were fixed in acetone and blocked with 10% goat serum. Incubation with peroxidase-conjugated goat antimouse antibody (ICN) was performed for 60 min at room temperature. Slides were developed with 3′3 diaminobenzidine tetrahydrochloride and hydrogen peroxide for the peroxidase reaction, and counterstained with hematoxylin.

Statistical analysis

Data was evaluated using the SPSS 11.5 statistical software package (SPSS Inc., Chicago, IL). Means were compared using the Mann–Whitney and anova-Bonferroni tests.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We investigated kidney transplants in four mouse strain combinations, all with full MHC plus nonMHC disparities. Because we aimed to investigate the development of pathologic lesions, most of these transplants were not life supporting. The study of lesion development requires that we avoid premature and unexpected death of the hosts from renal failure. Incidental death of mice causes post mortem deterioration of the graft and compromises the quality of the histology, and is not acceptable because it renders some grafts unacceptable owing to post mortem changes. Accordingly the contralateral kidney was usually left in place.

Weight and function of life-supporting vs. nonlife-supporting kidney transplants

A total of 165 mouse kidney allografts and seven syngeneic grafts were examined. Surgery was well tolerated by the host mice: mortality during the surgery and postoperative week was 16 out of 165, all in allografts (9.7%): two from ureteric leaks, two from ruptured incisions, and 12 from unknown causes on day 3–7 post surgery. Five allografts at harvest were diagnosed as early technical failures of the anastomosis on the basis of global infarction with thrombosis of the renal artery or vein. Histologic examination determined that none of these losses was the result of rejection before the infarction. This left 149 allografts and seven syngeneic grafts for inclusion in the study.

We measured the weight of the transplants and control kidneys and assayed the blood urea and creatinine. In seven syngeneic transplants examined at day 7, the kidney weight was not increased compared with the contralateral host kidney (186 ± 29 vs. 171 ± 19) or control kidneys. Thus the transplant process per se did not induce an increase in renal weight. All rejecting kidney allografts increased markedly in weight by day 5 (177% on average compared with control kidneys), and continued to gain weight through day 21 (274%) (Table 1). The increased weight of allografts persisted on day 42 transplants (217%). Life-supporting transplants did not differ in weight gain from nonlife-supporting transplants. Weight gain was not the result of ischemic damage, and histologic lesions of ischemic injury (acute tubular necrosis) were not observed in syngeneic or allogeneic kidneys.

In hosts with life-supporting transplants, serum creatinine and urea were elevated at the time of rejection as expected (Table 2), with a 2–3-fold elevation in creatinine and a 6-fold elevation in urea. Compatible with their elevated urea levels, mice with life-supporting allografts were unwell by day 6–8. Sodium, potassium, and chloride levels were unchanged. Host mice with the left kidney remaining in situ appeared healthy throughout the period of investigation (up to 42 days), with renal function (creatinine, urea and electrolytes levels) similar to that of hosts with syngeneic life-supporting transplants. The weight gain (Table 1) and the pathology (not shown) in the rejecting life-supporting transplants was similar to that in the nonlife-supporting transplants, indicating that the use of the nonlife-supporting transplant model did not alter the histology or severity of rejection.

For studies extending beyond day 7 we used only the nonlife-supporting transplant model because of its advantages for pathology studies and for animal well-being.

Histology of rejecting transplants

There were no lesions in the syngeneic transplants. Lesions in the allogeneic transplants at days 5 and 7 were similar in the life-supporting and nonlife-supporting transplants. Thus only the nonlife-supporting transplants are shown. The lesions in the rejecting nonlife-supporting transplants were evaluated at days 5, 7, 10, and 21 post transplant (Table 3; typical lesions shown in Figure 1). Interstitial infiltrate was detected at day 5, accentuated around arteries (Figure 1A), and increased by day 7 (Figure 1B). The interstitial infiltrate peaked at day 7–10, and stabilized or regressed by day 21. The cells in the infiltrate were mononuclear, with few identifiable neutrophils. Venulitis showed a similar pattern, being present at day 5 and peaking at days 7 (Figure 1B) and 10 (Figure 1C), and then regressing.


Figure 1. Histopathology of rejecting mouse allografts. (A) Kidney transplant (129 into CBA) at day 5 showing a heavy interstitial infiltrate surrounding a large artery, with no arteritis, no glomerulitis and no tubulitis (H&E, ×400). There is a mild venulitis. Tissue components are indicated as vein (v), artery (a), tubule (t), and glomerulus (g). (B) Rejecting kidney allograft (CBA into BALB/c) at day 7 showing venulitis, mild tubulitis ([RIGHTWARDS ARROW]), and interstitial infiltration but no arteritis or glomerulitis (H&E, ×400). (C) Rejecting kidney allograft at day 10 showing a longitudinally cut vein and artery with severe venulitis and mild arteritis (PAS, ×400). (D) Severe tubulitis lesions of a rejecting kidney allograft at day 21 (CBA into BALB/c). Note mononuclear infiltrate in the tubular epithelium of Bowman's capsule ([RIGHTWARDS ARROW]) (PAS, ×600), which was considered to be a tubulitis. (E) Severe endothelialitis of a rejecting CBA into BALB/c allograft at day 21, with lymphocytes beneath the arterial endothelium ([RIGHTWARDS ARROW]) and with interstitial edema (H&E, ×600). (F) Rejecting CBA kidney transplant in a B6 host at day 7 stained with goat antimouse IgG antibody, showing intense staining of IgG in interstitium and on basolateral membranes of tubules ([RIGHTWARDS ARROW]) (×400). No staining for IgG was observed in the contralateral kidney. (G) Rejecting CBA into a IgKO kidney transplant at day 7, showing no IgG staining with goat antimouse antibody (×400). (H) Rejecting CBA into a B6 kidney transplant at day 21, showing severe tubulitis ([RIGHTWARDS ARROW]), glomerulitis, edema (e) and necrosis (n) (PAS, ×400). Fibroid necrosis is present in the glomerulus shown but was uncommon. (I) Rejecting kidney allograft (CBA into IgKO) at day 21 with severe tubulitis ([RIGHTWARDS ARROW]) and mild edema (PAS, ×400).

In contrast, tubulitis (lymphocyte invasion of the tubule epithelium) was absent or minimal at day 5, mild at day 7 (Figure 1B), but severe by day 21 (Figure 1D). Arteritis showed a similar pattern, being absent or minimal at days 5 and 7, but severe at day 21 (Figure 1E). Glomerulitis, defined by lymphocytes in the glomerular tuft, also began slowly and progressed through day 21 (Figure 1D). Mononuclear cells in the epithelium of Bowman's capsule were present and considered as tubulitis.

Patchy necrosis, interstitial edema, and peritubular capillary congestion (PTC) all progressed through day 21. Of interest, the arteries and veins remained patent in all grafts through day 21 (with the exception of those infarcted from early failures of the vascular anastomosis as detailed above).

Neutrophils and eosinophils were infrequent at all times. There was no neutrophilic tubulitis or peylitis, and no neutrophilic inflammation in the medulla.

In one combination (CBA kidneys into B6 hosts), seven grafts were examined at day 42. The large vessels were patent and the majority of the parenchyma remained viable. Interstitial infiltration, venulitis, tubulitis, arteritis, glomerulitis, edema, necrosis and congestion were similar to those in the day 21 transplants in the same combination.

The contralateral kidneys were evaluated in all hosts and were within normal limits.

Altered histopathology of transplants in genetically modified (‘knockout’) hosts

To assess the rejection mechanisms we compared the histopathology of CBA kidneys transplanted into wild-type B6 mice with those transplanted into B6 nude mice or B6 hosts lacking mature B cells and immunoglobulin (IgKO) (Table 4). The grading system was the same as that in Table 3, except that necrosis and PTC were now expressed as a percentage of the observed parenchyma.

Transplants in nude mice were histologically normal and did not show infiltrate, tubulitis or arteritis (data not shown). Infiltration and tubulitis in IgKO hosts (Figure 1I) were as severe in wild-type hosts (Figure 1H). Kidneys rejecting in IgKO hosts at day 21, however, weighed less (approximately 60% of the weights of kidneys in WT hosts), and had less arteritis, venulitis, and edema. Necrosis and PTC were also decreased in IgG knockout hosts, albeit showing some variations (Figure 1I, Table 4). In grafts in the wild-type hosts, anti-IgG stained the interstitial areas and the basolateral membranes of the tubules (Figure 1F). Albumin staining, performed with the antimouse albumin (goat) HRP-linked antibody (Bethyl Laboratories Inc.), was diffused in the interstitial areas but did not affect the basolateral membranes in either WT or IgKO hosts (not shown). There was no discernable presence of albumin in the control tissue or isografts. Thus the pattern of IgG staining was compatible with both nonspecific leakage from capillaries (similar to albumin) owing to the inflammatory process, and a specific binding of anti-MHC antibody to the epithelium. As predicted, the grafts in IgKO hosts showed no staining for immunoglobulin (Figure 1G).

Intragraft MHC expression

We assessed the induction of MHC expression through day 21 semiquantitatively using radio-labeled antibody binding (Table 5). Major histocompatibility complex induction is expressed as fold increase in the radioactivity (c.p.m.) bound to cell membranes using radio-labeled antibodies specific for donor and host class I and II molecules. The increase in c.p.m. represents approximately one half of the actual MHC increase as extrapolated from a standard curve of specific antigen (18). Donor MHC expression was strongly increased by day 5 post-transplant, then stabilized at subsequent times but remained strongly elevated even at day 21. The relative increases in class I exceeded those in class II. Host MHC expression was also increased, as expected from the heavy infiltrate with host cells. Elevated levels of MHC I and II persisted even at day 42 in B6 kidneys grafted into CBA hosts. All of these results were confirmed by immunostaining with donor-specific anti-class I and class II antibodies, which showed that most of the induced staining was in the basolateral aspect of the renal tubules and on arterial endothelium, as previously reported (19).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We studied the evolution of pathologic lesions in kidney transplant rejection across MHC and nonMHC barriers in multiple-strain combinations to permit generalizations independent of strain and combination-specific peculiarities. The lesions of rejection progressed with two distinct time courses. Interstitial infiltrate develops rapidly, with massive donor MHC induction, as well as host MHC expression in the infiltrating cells. These changes then spontaneously stabilize or decline in all combinations at day 21. Despite this, the invasive lesions of tubulitis and endothelialitis continued to increase. The slower onset but continued progression of the invasive lesions indicates that they reflect different events than the rapidly developing lesions (venulitis, interstitial infiltration) or MHC induction. Tubulitis was T-cell-dependent but independent of alloantibody. Nevertheless, absence of B cells and antibody production altered the phenotype of kidney rejection at day 21, reducing necrosis, hemorrhage, edema, venulitis and arteritis, suggesting that alloantibody damages the endothelium in the microcirculation and arteries. Our results suggest that the lesions of renal transplant rejection are all T-cell-dependent but have three patterns: T-cell-mediated inflammation (interstitial inflammation, MHC induction), which peaks rapidly and then stabilizes or regresses; T-cell-mediated invasion (tubulitis), which starts slowly but progresses through day 21; and alloantibody-mediated damage to the endothelium of the microcirculation and arteries, which also develops later.

The fact that renal transplants in some mouse strain combinations tend to spontaneously survive despite early rejection (20,21), as do some human kidney transplants (22), may reflect homeostatic mechanisms in the effector cell population. The mechanism behind such survival may involve both alloantigen-driven mechanisms and newly recognized antigen-independent programs for effector T-cell homeostasis (23,24). This finding may be related to one of the fundamental characteristics of clinical transplantation: host graft adaptation (25). The mechanisms of stabilization and involution of the infiltrate probably involve lymphocyte apoptosis via a number of homeostatic mechanisms affecting effector lymphocytes, particularly CD8+ T cells (26). The mechanisms involve perforin and IFN-γ (27), both of which are abundant in rejecting grafts. In addition, CD8 T cells in mouse kidney transplants tend to down-regulate their T-cell receptors, which may also be part of effector T-cell homeostasis (28). Apoptosis, as detected by TUNEL staining, is common in the infiltrating cells of rejecting allografts (29), and probably represents the outcome of activation of most T cells.

The differential kinetics between progression in the invasive lesions and the stabilization of interstitial infiltration and MHC induction suggest that an additional event is operating for the invasive lesions such as tubulitis. In tubulitis, effector T cells enter the epithelium and potentially interact with the basolateral membrane of the epithelial cells. One possibility is that some of such cells express CD103 to interact with E-Cadherin. The CD103/E-Cadherin system is an important adhesion system of intraepithelial T lymphocytes (30,31). CD103 in vitro is induced by TGF-β and possibly by IL-15 (32), suggesting a role for these mediators in kidney as inducers of integrin CD103. Moreover, CD103 is needed for rejection of islet allografts (33), apparently because T-cell entry into islet epithelium requires CD103. The differential kinetics of lymphocytes in tubulitis vs. interstitial lesions may then reflect engagement of T-cell receptors or integrins or both in the epithelium, protecting those cells from apoptosis.

The present findings support a role for alloantibody in damaging the graft late in the course of rejection (day 21), targeting both the microcirculation and the larger vessels. At day 7 the rejection lesions in antibody-deficient hosts were indistinguishable from those in wild-type hosts, indicating that the early pathology is the result of effector T cells, which also produce the IFN-γ that induces the massive increase in MHC expression (29). However, immunoglobulin deficiency significantly altered the pathology at day 21, with less weight gain, edema, arteritis, necrosis, venulitis, and congestion. We have recently confirmed these results in another strain of IgG-deficient mice that lacks mature B cells (B6.129P2-Igh-Jtm1Cgn). Such results could be the result of the lack of B cells as antigen-presenting cells. However, it is much more likely that our data reflect the increasingly recognized ability of alloantibody against MHC to damage the graft endothelium, particularly the peritubular capillaries (34–37). We have previously shown that cytotoxic alloantibody levels are high in these combinations at day 21. The fact that alloantibody interacts with the endothelium of the microcirculation but requires some time to produce damage may reflect the large amounts of MHC in the kidney epithelium during acute rejection, coupled with the need for the antibody to reach critical concentrations on endothelium. The vast basolateral membrane surface of the renal epithelium, with its intense MHC expression during rejection, may act as a sink for alloantibody against MHC antigens, competing for alloantibody that would otherwise harm the vessels more acutely. The fact that the basolateral membranes stained for immunoglobulin but not for albumin, whereas the interstitial edema stained for both, is compatible with this concept. Nevertheless, alloantibody levels eventually become critical and cause endothelial damage. Alloantibody can also damage arteries, indicating that arteritis can reflect either cell-mediated injury or alloantibody-mediated injury (38,39).

Thus alloimmune injury across multiple MHC differences (as is the situation in most human organ transplantation) is heterogeneous, progressive, adaptive, and time-dependent, with at least three different patterns of injury. First comes the venulitis/infiltration/MHC induction phase, with the MHC serving as an indicator of massive IFN-γ production. This process may be hard-wired for homeostatic resolution despite the massive, persistent antigen expression, in keeping with recent data from nontransplant models (23,24). Second is the slower development of T-cell-mediated invasive lesions, particularly tubulitis, but also arteritis and glomerulitis. These lesions are the main correlate with rejection injury in human renal allografts. Specialized effectors such as CD103+, CD8+ lymphocytes may be important in tubulitis lesions and their emergence and homeostasis may differ from other effectors (33). The third process is alloantibody-mediated rejection, which injures the capillaries, contributes to late graft edema and necrosis, and adds to the T-cell-mediated changes in the arteries and veins. The results underscore the emerging concept that graft rejection is not a single process or mechanism, but a composite phenotype with discreet molecular events, each with its own pace and regulation. The use of pathologic lesions rather than global graft failure in experimental transplantation should facilitate the dissection of these distinct mechanisms, and must be taken into account in tolerance strategies.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Grant support was from The Canadian Institutes of Health Research, The Kidney Foundation of Canada, The Roche Organ Transplant Research Foundation, The Royal Canadian Legion, and Hoffmann-LaRoche, Canada.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Hariharan S, Johnson CP, Bresnahan BA, Taranto SE, McIntosh MJ, Stablein D. Improved graft survival after renal transplantation in the United States, 1988 to 96. N Engl J Med 2000; 342: 605612.
  • 2
    Solez K, Axelsen RA, Benediktsson H et al. International standardization of criteria for the histologic diagnosis of renal allograft rejection. The Banff working classification of kidney transplant pathology. Kidney Int 1993; 44: 411422.
  • 3
    Solez K. International standardization of criteria for histologic diagnosis of chronic rejection in renal allografts. Clin Transplant 1994; 8: 345350.
  • 4
    Colvin RB, Cohen AH, Saiontz C et al. Evaluation of pathologic criteria for acute renal allograft rejection: reproducibility, sensitivity, and clinical correlation. J Am Soc Nephrol 1997; 8: 19301941.
  • 5
    Marcussen N, Olsen TS, Benediktsson H, Racusen L, Solez K. Reproducibility of the Banff classification of renal allograft pathology. Inter- and intraobserver variation. Transplantation 1995; 60: 10831089.
  • 6
    Gaber LW, Moore LW, Alloway RR et al. Correlation between Banff classification acute renal rejection scores and reversal of rejection. Kidney Int 1996; 49: 481487.
  • 7
    Gaber LW, Schroeder TJ, Moore LW, Shakouh-Amiri MH, Gaber AO The correlation of Banff scoring with reversibility of first, recurrent rejection episodes. Transplantation 1996; 61: 17111715.
  • 8
    Nankivell BJ, Fenton-Lee CA, Kuypers DR et al. Effect of histological damage on long-term kidney transplant outcome. Transplantation. 2001; 71: 515523.
  • 9
    Macdonald FI, Ashraf S, Picton M et al. Banff criteria as predictors of outcome following acute renal allograft rejection. Nephrol Dial Transplant 1999; 14: 16921697.
  • 10
    Kuypers DR, Chapman JR, O'Connell PJ, Allen RD, Nankivell BJ. Predictors of renal transplant histology at three months. Transplantation 1999; 67: 12221230.
  • 11
    Gaber LW, Moore LW, Gaber AO et al. Correlation of histology to clinical rejection reversal: a thymoglobulin multicenter trial report. Kidney Int 1999; 55: 24152422.
  • 12
    Gaber LW, Moore LW, Gaber AO et al. Utility of standardized histological classification in the management of acute rejection: Efficacy Endpoints Conference 1995. Transplantation 1998; 65: 376380.
  • 13
    Gaber LW, Moore LW, Gaber AO et al. A multicenter report on the utilization of Banff grading in acute renal rejection. The 1995 Efficacy Endpoints Conference. Transplant Proc 1997; 29: 184185.
  • 14
    Furness PN, Kirkpatrick U, Taub N, Davies DR. Solez K. A UK-wide trial of the Banff classification of renal transplant pathology in routine diagnostic practice. Nephrol Dial Transplant 1997; 12: 995100.
  • 15
    Waiser J, Schreiber M, Budde K et al. Prognostic value of the Banff classification. Transpl Int 2000; 13 (Suppl. 1): S106S111.
  • 16
    Le Moine A, Goldman M. Non-classical pathways of cell-mediated allograft rejection. New challenges for tolerance induction Am J Transplant 2003; 3: 101106.
  • 17
    Salacinski PR, McLean C, Sykes JE, Clement-Jones VV, Lowry PJ. Iodination of proteins, glycoproteins, and peptides using a solid-phase oxidizing agent, 1,3,4,6-tetrachloro-3 alpha,6 alpha-diphenyl glycoluril (Iodogen). Anal Biochem 1981; 117: 136146.
  • 18
    Jephthah-Ochola J, Urmson J, Farkas S, Halloran PF. Regulation of MHC in vivo. Bacterial lipopolysaccharide induces class I and II MHC products in mouse tissues by a T cell independent, cyclosporine sensitive mechanism. J Immunol 1988; 141: 792800.
  • 19
    Halloran PF, Afrouzian M, Ramassar V et al. Interferon-gamma acts directly on rejecting renal allografts to prevent graft necrosis. Am J Pathol 2001; 158: 215226.
  • 20
    Russell PS, Chase CM, Colvin RB, Plate JMD. Kidney transplants in mice. An analysis of the immune status of mice bearing long-term H-2 incompatible transplants. J Exp Med 1978; 147: 14491468.
  • 21
    Russell PS, Chase CM, Colvin RB, Plate JMD. Induced immune destruction of long-surviving H, -2 incompatible kidney transplants in mice. J Exp Med 1978; 147: 14691486.
  • 22
    Hume DM, Merrill JP, Miller BF, Thorn GW. Experiences with renal homotransplantation in the human report of nine cases. J Clin Invest 1955; 34: 327382.
  • 23
    Badovinac VP, Porter BB, Harty JT. Programmed contraction of CD8 (+) T cells after infection. Nat Immunol 2002; 3: 619626.
  • 24
    Dooms H, Abbas AK. Life and death in effector T cells. Nat Immunol 2002; 3: 797798.
  • 25
    Starzl TE, Marchioro TL, Waddell WR. The reversal of rejection in human renal homografts with subsequent development of homograft tolerance. Surg Gyne Obs 1963; 117: 385395.
  • 26
    De Saint Basile G, Fischer A. The role of cytotoxicity in lymphocyte homeostasis. Curr Opin Immunol 2001; 13: 549554.
  • 27
    Badovinac VP, Tvinnereim AR, Harty JT. Regulation of antigen-specific CD8+ T cell homeostasis by perforin and interferon-gamma. Science 2000; 290: 13541358.
  • 28
    Mannon RB, Kotzin BL, Nataraj C, Ferri K, Roper E, Kurlander RJ. Downregulation of T cell receptor expression by CD8+ lymphocytes in kidney allografts. J Clin Invest 1998; 101: 25172527.
  • 29
    Halloran PF, Miller LW, Urmson J et al. IFN-gamma alters the pathology of graft rejection: protection from early necrosis. J Immunol 2001; 166: 70727081.
  • 30
    Robertson H, Wong WK, Talbot D, Burt AD, Kirby JA. Tubulitis after renal transplantation: demonstration of an association between CD103+ T cells, transforming growth factor beta1 expression and rejection grade. Transplantation 2001; 71: 306313.
  • 31
    Hadley GA, Charandee C, Weir MR, Wang D, Bartlett ST, Drachenberg CB. CD103+ CTL accumulate within the graft epithelium during clinical renal allograft rejection. Transplantation 2001; 72: 15481555.
  • 32
    Robertson H, Kirby JA. Post-transplant renal tubulitis: the recruitment, differentiation and persistence of intra-epithelial T cells. Am J Transplant 2003; 3: 310.
  • 33
    Feng Y, Wang D, Yuan R, Parker CM, Farber DL, Hadley GA. CD103 expression is required for destruction of pancreatic islet allografts by CD8 (+) T cells. J Exp Med 2002; 196: 877886.
  • 34
    Feucht HE, Schneeberger H, Hillebrand G et al. Capillary deposition of C4d complement fragment and early renal graft loss. Kidney Int 1993; 43: 13331338.
  • 35
    Trpkov K, Campbell P, Pazderka F, Cockfield S, Solez K, Halloran PF. Pathologic features of acute renal allograft rejection associated with donor-specific antibody. Transplantation 1996; 61: 15861592.
  • 36
    Matsumoto Y, McCaughan GW, Painter DM, Bishop GA. Evidence that portal tract microvascular destruction precedes bile duct loss in human liver allograft rejection. Transplantation 1993; 56: 6975.
  • 37
    Collins AB, Schneeberger EE, Pascual MA et al. Complement activation in acute humoral renal allograft rejection: diagnostic significance of C4d deposits in peritubular capillaries. J Am Soc Nephrol 1999; 10: 22082214.
  • 38
    Russell PS, Chase CM, Colvin RB. Alloantibody and T-cell-mediated immunity in the pathogenesis of transplant arteriosclerosis. Transplantation 1997; 64: 15311536.
  • 39
    Russell PS, Chase CM, Winn HJ, Colvin RB. Coronary atherosclerosis in transplanted mouse hearts. II. Importance of humoral immunity. J Immunol 1994; 152: 51355141.