One of the greatest challenges facing the renal transplant community is chronic allograft injury (CAI); characterized by interstitial fibrosis and tubular atrophy (IFTA), it leads to a progressive decline in renal transplant function, and the loss of 5% of renal transplants annually (1). The development of new and effective immunosuppressive agents targeted primarily against T lymphocytes has led to a significant reduction in acute rejection rates; however, this has had minimal impact on long-term outcome (2). Current therapeutic strategies focus on minimizing the risk of CAI, but specific treatments to combat IFTA remain elusive and factors that drive it are unclear. Understanding the mechanisms which regulate IFTA may identify novel therapeutic targets and strategies.
Galectin-3 (gal-3) is a β-galactoside-binding lectin of approximately 30 kDa that is expressed in fibrotic tissue of diverse causes: it is a potent mitogen for fibroblasts in vitro (3), and mice deficient in gal-3 develop reduced fibrosis in vivo in lung, liver, and kidney models, suggesting that gal-3 plays a key role in this process and represents a novel target for intervention to block the initiation and progression of fibrosis (4–6). Our work in renal fibrosis has suggested that gal-3 may regulate the development of chronic inflammation by modulating alternative macrophage activation (4, 7). Alternative macrophage activation is implicated in diverse disease pathologies such as asthma, organ fibrosis, and granulomatous diseases (8, 9). Disruption of the gal-3 gene specifically restrains interleukin (IL)-4/IL-13-induced alternative macrophage activation in vitro and in vivo without affecting interferon-γ and lipopolysaccharide-induced classical activation or IL-10-induced deactivation. By contrast, classical macrophage activation with lipopolysaccharide inhibits gal-3 expression and release (7). Although gal-3 has been shown to have an effect in other models of renal fibrosis, its role in CAI is unknown, and forms the focus of the study. We examined the consequences of transplanting a BM12 kidney (gal-3 wild type) into a C57BL/6 gal-3 null mouse.
Model of Chronic Renal Allograft Injury Is Characterized by IFTA
Characterized by a single class II mismatch, transplantation of a BM12 kidney into a C57BL/6 recipient leads to the development of IFTA. We have demonstrated a significant loss of tubules at 4 and 8 weeks (Fig. 1a): 4 weeks: Syngeneic: 159.3±3.3, allografts: 63.6±8.8 tubules per low power field (100×, lpf), P<0.0001; 8 weeks: Syngeneic: 143.3±1.19, allografts: 65.5±6.6 tubules per lpf, P<0.0001) and the development of interstitial fibrosis in allografts compared with syngeneic controls (Fig. 1b): (Syngeneic: 0.01±0.01, allografts: 11.8%±0.8% picrosirius red (PSR) staining per high power field (200×, hpf); P<0.0001; 8 weeks: Syngeneic: 0.03±0.02, allografts:14.18%±1.4% per hpf, P<0.0001, n=6.
Gal-3 Expression Is Upregulated in CAI
Gal-3 was minimally present in a few collecting ducts in syngeneic controls (Fig. 1c). It was markedly upregulated in the renal interstitium in allografts (Fig. 1d, main picture), and particularly at sites of inflammation (Fig. 1d, inset) compared with syngeneic controls at 4 and 8 weeks (Syngeneic: 1.14±0.22, allograft: 9.87%±1.30% per hpf, P=0.002; 8 weeks: Syngeneic: 0.99%±0.07%, allograft: 8.24±1.03, P=0.002, n=6). Transplantation of kidneys from BM12 donors into gal-3 null mice was associated with a significant reduction in gal-3 expression to levels similar to those detected in syngeneic controls in the renal interstitium (Fig. 1f), and was notably absent at sites of inflammation (Fig. 1e), (Gal-3 null at 4 weeks: 2.13%±0.29% per hpf, P=0.001, compared with allograft controls, 8 weeks: 1.20±0.18, P=0.0002, n=6).
Abrogation of Gal-3 Expression in Renal Allografts Leads to Preservation of Renal Tubules
Control allografts demonstrated histological features of CAI, including interstitial fibrosis, tubular atrophy, and chronic inflammatory cell infiltrate, and these changes were reduced in gal-3 null mice (Fig. 2a–c). (Tubular count, 4 weeks: Gal-3 null: 108.8±8.77, allografts: 63.6±8.8 tubules per lpf, P=0.008, n=6, 8 weeks: Gal-3 null: 100.3±9.35, allografts: 65.5±6.6 tubules per lpf, P=0.01, n=6). Development of CAI was associated with reduced tubular cell proliferation, and induction of tubular apoptosis over time, and this was unaffected by gal-3 expression (Fig. 2e,f).
Abrogation of Gal-3 Expression Protects Against Interstitial Fibrosis
Transplantation of BM12 donor kidneys into gal-3 null mice resulted in significant reduction in interstitial fibrosis, assessed after staining with the pan-collagen marker, PSR (4 weeks: Gal-3 null: 7.32±1.01, allografts: 11.8%±0.8% per hpf; 8 weeks: Gal-3 null: 9.04±0.54, allografts: 14.2%±1.4% per hpf P=0.01, n=6) as shown in Figure 3(a–d). Other markers of fibrosis were examined: collagen I (Fig. 3e) and α-smooth muscle actin (Fig. 3f) were both upregulated in allografts and reduced in the gal-3 null group.
Abrogation of Gal-3 in Recipient Mice Does Not Alter the Number of Macrophages Infiltrating Into the Graft, but Reduces the Number Expressing YM1, a Marker of Alternative Macrophage Activation
There was no difference in the number of F4/80-positive infiltrating macrophages in the gal-3 null group, compared with allografts at 4 and 8 weeks (Fig. 4a). Although the number of macrophages infiltrating the transplanted kidney was similar irrespective of gal-3 expression, there was a reduction in YM1 expression in gal-3 null recipients compared with allografts, (4 weeks: Gal-3 null: 9.88±1.2, allografts: 39.5%±1.7% per hpf; P=0.0001; 8 weeks: Gal-3 null: 9.90±2.08, allografts: 27.4%±3.34% per hpf, P=0.006, n=6, Fig. 4b–d). Dual staining demonstrated colocalization of YM1 and F4/80 in the allografts (Fig. 4e), and this was not detected in allografts.
Abrogation of Gal-3 in Recipient Mice Does Not Alter the Number of T Cells Infiltrating Into the Graft, But Does Alter the T-Cell Population
There was no difference in the total number of CD3+ve T cells (Fig. 5) in the gal-3 null group compared with the allograft controls. When examining the T-cell population, we demonstrated a significant upregulation in the number of CD4+ve T lymphocytes (Th2) in the allograft group, which was reduced with abrogation of gal-3 (4 weeks: Gal-3 null: 3.68±0.53, allografts: 6.92%±1.3%; 8 weeks: Gal-3 null: 3.83±0.34, allografts: 7.09%±1.3% per hpf (×800) P=0.02, n=5). We went on to examine the expression of Th2 cytokines that are important in the induction of alternative macrophage activation, namely IL-4 and transforming growth factor (TGF)-β. There was significantly increased expression of IL-4 at 8 weeks in the allograft group, and this was reduced with abrogation of gal-3 (Fig. 5c–d). Although there seemed to be a reduction in TGF-β expression in the gal-3 null group, this difference was not statistically significant.
Significant improvements in acute rejection rates mean that CAI, characterized by IFTA, has surmounted this as the major challenge facing renal transplantation. Specific strategies aimed at halting or reversing the fibrotic process would be beneficial in the management of this condition.
In this study, we have demonstrated that transplantation of BM12 donor kidneys into C57BL/6 mice with a single class II mismatch results in the development of IFTA. This is associated with significantly increased expression of gal-3, increased numbers of circulating CD4+ve T lymphocytes, with a modest increase in the expression of profibrotic cytokines such as IL-4 and TGF-β, and the switch to an YM1-expressing M2 macrophage population. Conversely, transplantation of BM12 donor kidneys into gal-3 null mice results in significant renal tubular preservation, although this was not reflected in a reduction in tubular apoptosis or increased proliferation, suggesting that this may simply reflect reduced interstitial fibrosis. There was reduced renal myofibroblast accumulation and fibrosis compared with allograft controls, suggesting that host gal-3 and its regulation of bone marrow-derived cells infiltrating the transplanted kidney plays a role in the development of interstitial fibrosis.
Several studies have demonstrated a role for macrophages in CAI (10, 11). Macrophages are involved in all stages of the inflammatory process, including fibrosis, tissue repair, and healing. In the context of renal transplantation, we have previously demonstrated that macrophage depletion in an experimental model of acute renal allograft rejection was associated with preservation of the renal microvasculature, and a reduction in acute rejection (12), and there is no doubt that macrophage infiltration persists in CAI, as observed in this and other studies (13). Recent work indicates that macrophages can be functionally distinguished into two broad phenotypes based on cell surface markers and cytokine profile (14, 15). Classically activated (M1) macrophages are induced by the Th-1 lymphokines interferon-γ inducing tissue injury. In contrast, alternatively activated (M2) macrophages are induced by Th-2 lymphokines such as IL-4, IL-10, IL-13, and TGF-β, promoting tissue fibrosis (16).
These results suggest a role for gal-3 in the development of renal fibrosis in the context of CAI. Gal-3 expression is upregulated in the unilateral ureteric obstruction mouse model of progressive renal fibrosis, and its absence is associated with a reduction in myofibroblast accumulation and renal fibrosis (4). Gal-3 is associated with the development of fibrosis in other organs, such as the liver, in which its expression was spatially and temporally related to the induction and resolution of experimental hepatic fibrosis. Disruption of the gal-3 gene blocked myofibroblast activation and procollagen (I) expression both in vivo and in vitro (5).
Gal-3 has been linked to IL-4-mediated alternative macrophage activation, which drives the resolution phase of inflammation and is associated with upregulation of the mannose receptor (17), YM-1 (chitinase-like lectin), and FIZZ1 (resistin-like secreted protein) (18). A number of studies provide evidence for an association between alternative macrophage activation and enhanced fibrosis (8), and IL-4/IL-13 activated macrophages upregulate profibrotic genes, stimulating production of fibronectin, and other matrix proteins. Recent work by Duffield and coworkers (19) demonstrated macrophage differentiation into three subpopulations depending on their Ly6C expression, and, in their model of unilateral ureteric obstruction, YM1 was not found to be expressed in the profibrotic Ly6C-lo population. There is evidence to support that targeting of gal-3 pathway through specific inhibitors blocks IL-4-induced macrophage activation (7) and reduces fibrosis. In addition, gal-3 has been shown to modulate T-cell function (20) and this has been borne out in this study, in which the presence of gal-3 was associated with an increased number of CD4 positive T cells, and associated production of IL-4, which were significantly reduced with abrogation of gal-3.
It must be acknowledged that not all authors have found gal-3 to be profibrotic, and mediating injury. Okamura et al. (21) have demonstrated a protective effect of gal-3 in a model of renal fibrosis, but the majority of studies have found the opposite effect (4, 22).
Modified citrus pectin is a naturally occurring inhibitor of gal-3 carbohydrate binding that has been shown to inhibit gal-3 function in vivo (23, 24). Most recently, it has been shown that administration of modified citrus pectin was protective in a model of experimental acute kidney injury, specifically associated with reduced fibrosis, macrophage infiltration and proinflammatory cytokine expression (25). At present, the effect of gal-3 inhibitors on organ rejection is not known. Given our previous work showing that gal-3 regulates renal fibrosis after ureteric obstruction (4) coupled with the present findings, further research into the potential of gal-3 inhibitors in preventing fibrosis in allograft injury after transplant is warranted.
The retention of a single native kidney in this study means that it remains a transplant-independent model, and no functional readouts could be obtained. Although others have published models in which the native kidney is removed some days after transplantation (26), this remains a technically challenging step, and its lack is a limitation of the study. In addition, we acknowledge that the model is dependent on a single Class II mismatch between donor and recipient, and there may be minor genetic differences between individual C57Bl/6 mice. To optimize this, the breeding of gal-3 null and C57Bl/6 wild types to generate C57Bl/6 homozygotes would be required.
This study does not attempt to address the contribution of donor-derived gal-3, and thus the role of resident donor-derived macrophages in promoting fibrosis. This would be an interesting adjunct to this study. However, the results do suggest a significant role for recipient-derived infiltrating cells, such as alternatively activated macrophages that express gal-3, in mediating IFTA.
Our results lead us to conclude that strategies to inhibit gal-3 in the kidney may lead to the development of exciting antifibrotic therapies, with potential benefit in combating CAI, which is currently responsible for the loss of 5% of renal transplants annually, with no effective treatment.
MATERIALS AND METHODS
All mice were bred in the University of Edinburgh and were maintained in 12-hr light/12-hr dark cycles with free access to food and water. All procedures were performed in accordance with Home Office Guidelines (Animals Scientific Procedures Act, 1986). Generation of gal-3 null mice by gene-targeting technology has been described previously (27). Strain-matched controls (C57BL/6) and donor mice (BM12) were bred in house.
Murine Model of CAI
Renal transplants were performed as previously described (28). We have developed a model of CAI, which adopts the use of congenic murine strains, characterized by a single class II mismatch: donor BM12 mice kidneys, (H-2bm12), were transplanted into C57BL/6 recipients (H-2b) (n=6) (control allograft group). This strain combination has been previously used in a cardiac allograft model (29). Transplants performed between C57BL/6 mice formed the syngeneic control group (n=6). Gal-3 null mice are bred on the C57BL/6 background, and transplants were performed between BM12 donors and gal-3 null mice (Gal-3 null group, n=6). All mice were male, aged 8 to 10 weeks, and weighing approximately 25 g.
To describe the surgical technique briefly, after intraperitoneal administration of anesthetic with ketamine and metomodine in normal saline, the donor kidney was isolated and perfused with cooled University of Wisconsin solution. After unilateral nephrectomy in the recipient, the donor renal artery and renal vein were anastomosed to recipient aorta and inferior vena cava, respectively, with a mean warm ischemic time of 30 min. The donor ureter was anastomosed to recipient bladder. Mean total ischemic time was 40 min. Intravenous heparin was administered to both donor and recipient (10 units in 400 μL normal saline). Mice received subcutaneous normal saline and were maintained in a warm box at 27°C for 48 hr postoperatively. All mice had a single intact native kidney and so the model is not transplant-dependent. All experiments were performed in accordance with the UK Government Home office regulations. The technical success rate of the model was 75%, defined as survival to the end of the experiment and the presence of a viable kidney on completion. The animals were kept under close observation postoperatively and culled if recovery was slow. Mice were culled at 4 (n=6 per groups) and 8 weeks (n=6 per group). These time points were selected based on time course experiments undertaken in our laboratory, demonstrating progressive tubular loss and fibrosis between 4 and 8 weeks, with the highest number of infiltrating cells detected between 2 and 4 weeks (data not shown).
Whole kidneys were cut longitudinally and fixed in methyl Carnoy's solution (60% methanol, 30% chloroform, and 10% acetic acid) or paraformaldehyde for YM1 antibody before being embedded in paraffin. Tissue sections (4 μm) were stained with hematoxylin-eosin for assessment of tubular atrophy and stained with PSR for pan collagen. Before staining with primary antibody, antigen retrieval was performed with antigen unmasking solution (Vector, Peterborough, UK). Tissue endogenous peroxidase was blocked in 3% hydrogen peroxide solution (Sigma-Aldrich, Poole, UK) and tissue sections were blocked with avidin/biotin and protein blocking kit (Vector, Peterborough, UK). Collagen I expression was studied using goat anti-type I collagen-UNLB polyclonal antibody (1:100 dilution; Southern Biotech, USA) and myofibroblast activation using mouse monoclonal anti-α-smooth muscle actin clone 1A4 (1:1500 dilution; Sigma-Aldrich, Poole, UK). Gal-3 expression was quantified after staining with rat anti-mouse gal-3 clone 8942F (1:1000 dilution, Cedarlane, Canada). Tubulointerstitial macrophage infiltration was quantified after immunostaining for the murine macrophage marker F4/80 (1:200 dilution, Caltag Laboratories, Northampton, UK). Alternative activation (M2) of macrophages was assessed after staining with rabbit anti-mouse YM1 (1:200, Stemcell Technologies, Grenoble, France). T- and B-lymphocyte infiltrates were quantified with rabbit anti-mouse CD3 (1:400 dilution, Dakocytomation, Denmark) and rat anti-mouse B220 (1:200 dilution, BD Biosciences, Oxford, UK), respectively.
All primary antibodies were incubated at 4°C overnight with subsequent incubation with biotinylated species-specific secondary antibody (1:300 dilution; Vector, Peterborough, UK) at room temperature for 30 min. After washing, sections were incubated with Vectastain ABC Elite reagent (Vector, Peterborough, UK) for 30 min at room temperature, before washing and staining with diaminobenzidine (Dako UK, Ely, UK). Hematoxylin counterstaining was performed before mounting. Normal murine spleen tissue sections were used as positive controls.
Kidney frozen sections (10 μm) were fixed with ice-cold acetone for 10 min and air dried at room temperature. The sections were washed with phosphate-buffered saline for 5 min and permeabilized with 0.1% Triton in phosphate-buffered saline at room temperature for 10 min. Blocking was performed with 3% hydrogen peroxide, avidin/biotin/protein blocking kit (Vector, Peterborough, UK), and with Fc-block diluted in antibody diluent (Vector, Peterborough, UK). CD4 T cells were stained with rat anti-mouse CD4 (1:50 dilution, BD Biosciences, Oxford, UK), incubated at 4°C overnight and followed by goat anti-rat Alexa Fluor 568 (Caltag, Burlingame, CA) at room temperature for 30 min. F4/80 and YM1 double staining with rat anti-mouse F4/80 (1:100 dilution, Caltag Laboratories, Northampton, UK) and rabbit anti-mouse YM1 (1:100, Stemcell Technologies, Grenoble, France) primary antibody at 4°C overnight and followed by goat anti-rat Alexa Fluor 568 and donkey anti-rat Alexa Fluor 488 (Caltag, Burlingame, CA) at room temperature for 30 min. All slides were mounted with DAPI Vectashield mounting medium (Vector Laboratories) and finally examined by fluorescent microscopy (Carl Zeiss Meditec, Göttingen, Germany).
Tubules were photographed and counted using Image J software (Cell_counter plugin; ImageJ 1.36b; National Institutes of Health, Bethesda, MD) on five nonoverlapping lpfs (×100 magnification) from each section. Using well-validated methods that have been previously published (4) PSR, F4/80, CD3, α-smooth muscle actin, and type I collagen staining were quantified in 10 hpfs (×200) fields, using Color Range tool on Photoshop CS4. YM1 was analyzed by counting the number of YM1-positive cells per hpf (×200). All sections were analyzed in a blinded manner. All histological figures shown were taken at 4 weeks.
Real-Time Reverse-Transcriptase Polymerase Chain Reaction
Total RNA from frozen kidneys was prepared using RNeasy kits (Qiagen, Crawley, UK) and reverse transcribed into cDNA using random hexamers (Applied Biosystems, UK). For analysis of transcripts, cDNA was analyzed using a SYBR green-based quantitative fluorescence method (Applied Biosystems), The following primer pairs were used:
- Mouse β actin: forward 5′-AGAGGGAAATCGTGCGTGAC reverse 5′-CAATAGTGATGACCTGGCCGT-3′;
- Mouse IL-4: forward 5′-CACGGATGCGACAAAAATCA-3′
- Reverse 5′- AGGACGTTTGGCACATCCAT-3′
- TGF-β: forward 5′- GACTCTCCACCTGCAAGACCA-3′
- Reverse 5′- GGGACTGGCGAGCCTTAGTT-3′
Graph Pad Prism was used for statistical analyses. Unpaired t tests were performed for two groups comparisons, data are expressed as mean±SEM, and a P value less than 0.05 considered statistically significant. The P values are presented graphically as *P less than 0.05, **P less than 0.005, and ***P less than 0.0005.
The authors acknowledge Prof. Sarah Howie for her scientific support and help with manuscript preparation, and Mr. M. Clay and Mr. G. Borthwick for their assistance and technical support with management of the mice.
1. Ponticelli C, Villa M, Cesana B, et al.. Risk factors for late kidney allograft failure. Kidney Int 2002; 62: 1848.
2. Meier-Kriesche HU, Schold JD, Kaplan B. Long-term renal allograft survival: Have we made significant progress or is it time to rethink our analytic and therapeutic strategies? Am J Transplant 2004; 4: 1289.
3. Sasaki S, Bao Q, Hughes RC. Galectin-3 modulates rat mesangial cell proliferation and matrix synthesis during experimental glomerulonephritis induced by anti-Thy1.1 antibodies. J Pathol 1999; 187: 481.
4. Henderson NC, Mackinnon AC, Farnworth SL, et al.. Galectin-3 expression and secretion links macrophages
to the promotion of renal fibrosis. Am J Pathol 2008; 172: 288.
5. Henderson NC, Mackinnon AC, Farnworth SL, et al.. Galectin-3 regulates myofibroblast activation and hepatic fibrosis. Proc Natl Acad Sci USA 2006; 103: 5060.
6. Farnworth SL, Henderson NC, Mackinnon AC, et al.. Galectin-3 reduces the severity of pneumococcal pneumonia by augmenting neutrophil function. Am J Pathol 2008; 172: 395.
7. MacKinnon AC, Farnworth SL, Hodkinson PS, et al.. Regulation of alternative macrophage
activation by galectin-3. J Immunol 2008; 180: 2650.
8. Nair MG, Cochrane DW, Allen JE. Macrophages
in chronic type 2 inflammation have a novel phenotype characterized by the abundant expression of Ym1 and Fizz1 that can be partly replicated in vitro. Immunol Lett 2003; 85: 173.
9. Raes G, Brys L, Dahal BK, et al.. Macrophage
galactose-type C-type lectins as novel markers for alternatively activated macrophages
elicited by parasitic infections and allergic airway inflammation. J Leukocyte Biol 2005; 77: 321.
10. Ozdemir BH, Demirhan B, Gungen Y. The presence and prognostic importance of glomerular macrophage
infiltration in renal allografts. Nephron 2002; 90: 442.
11. Beckmann N, Cannet C, Zurbruegg S, et al.. Macrophage
infiltration detected at MR imaging in rat kidney allografts: Early marker of chronic rejection? Radiology 2006; 240: 717.
12. Qi F, Adair A, Ferenbach D, et al.. Depletion of cells of monocyte lineage prevents loss of renal microvasculature in murine kidney transplantation. Transplantation 2008; 86: 1267.
13. Adair A, Mitchell DR, Kipari T, et al.. Peritubular capillary rarefaction and lymphangiogenesis in chronic allograft failure. Transplantation 2007; 83: 1542.
14. Duffield JS. The inflammatory macrophage: A story of Jekyll and Hyde. Clin Sci (Lond) 2003; 104: 27.
15. Ferenbach D, Kluth DC, Hughes J. Inflammatory cells in renal injury and repair. Semin Nephrol 2007; 27: 250.
16. Goerdt S, Orfanos CE. Other functions, other genes: Alternative activation of antigen-presenting cells. Immunity 1999; 10: 137.
17. Stein M, Keshav S, Harris N, et al.. Interleukin 4 potently enhances murine macrophage
mannose receptor activity: A marker of alternative immunologic macrophage
activation. J Exp Med 1992; 176: 287.
18. Raes G, Noel W, Beschin A, et al.. FIZZ1 and Ym as tools to discriminate between differentially activated macrophages
, Develop Immunol 2002; 9: 151.
19. Lin SL, Castano AP, Nowlin BT, et al.. Bone marrow Ly6Chigh monocytes are selectively recruited to injured kidney and differentiate into functionally distinct populations. J Immunol 2009; 183: 6733.
20. Hsu DK, Chen HY, Liu FT. Galectin-3 regulates T-cell functions. Immunol Rev 2009; 230: 114.
21. Okamura DM, Pasichnyk K, Lopez-Guisa JM, et al.. Galectin-3 preserves renal tubules and modulates extracellular matrix remodeling in progressive fibrosis. Am J Physiol Renal Physiol 2011; 300: F245.
22. Nishi Y, Sano H, Kawashima T, et al.. Role of galectin-3 in human pulmonary fibrosis. Allergol Int 2007; 56: 57.
23. Liu HY, Huang ZL, Yang GH, et al.. Inhibitory effect of modified citrus pectin on liver metastases in a mouse colon cancer model. World J Gastroenterol 2008; 14: 7386.
24. Inohara H, Raz A. Effects of natural complex carbohydrate (citrus pectin) on murine melanoma cell properties related to galectin-3 functions. Glycoconj J 1994; 11: 527.
25. Kolatsi-Joannou M, Price KL, Winyard PJ, et al.. Modified citrus pectin reduces galectin-3 expression and disease severity in experimental acute kidney injury. PLoS One 2011; 6: e18683.
26. Brown K, Sacks SH, Wong W. Tertiary lymphoid organs in renal allografts can be associated with donor-specific tolerance rather than rejection. Eur J Immunol 2010; 41: 89.
27. Hsu DK, Yang RY, Pan Z, et al.. Targeted disruption of the galectin-3 gene results in attenuated peritoneal inflammatory responses. Am J Pathol 2000; 156: 1073.
28. Mannon RB, Kopp JB, Ruiz P, et al.. Chronic rejection of mouse kidney allografts. Kidney Int 1999; 55: 1935.
29. Win TS, Rehakova S, Negus MC, et al.. Donor CD4 T cells contribute to cardiac allograft vasculopathy by providing help for autoantibody production. Circ Heart Fail 2009; 2: 361.