Dextran Sulfate Facilitates Anti-CD4 mAb-Induced Long-Term Rat Cardiac Allograft Survival After Prolonged Cold Ischemia


* Corresponding author: Robert Rieben,


Ischemia/reperfusion injury leads to activation of graft endothelial cells (EC), boosting antigraft immunity and impeding tolerance induction. We hypothesized that the complement inhibitor and EC-protectant dextran sulfate (DXS, MW 5000) facilitates long-term graft survival induced by non-depleting anti-CD4 mAb (RIB 5/2). Hearts from DA donor rats were heterotopically transplanted into Lewis recipients treated with RIB 5/2 (20 mg/kg, days—1,0,1,2,3; i.p.) with or without DXS (grafts perfused with 25 mg, recipients treated i.v. with 25 mg/kg on days 1,3 and 12.5 mg/kg on days 5,7,9,11,13,15). Cold graft ischemia time was 20 min or 12 h. Median survival time (MST) was comparable between RIB 5/2 and RIB 5/2+DXS-treated recipients in the 20-min group with >175-day graft survival. In the 12-h group RIB 5/2 only led to chronic rejection (MST = 49.5 days) with elevated alloantibody response, whereas RIB 5/2+DXS induced long-term survival (MST >100 days, p < 0.05) with upregulation of genes related to transplantation tolerance. Analysis of the 12-h group treated with RIB 5/2+DXS at 1-day posttransplantation revealed reduced EC activation, complement deposition and inflammatory cell infiltration. In summary, DXS attenuates I/R-induced acute graft injury and facilitates long-term survival in this clinically relevant transplant model.


Induction of donor-specific transplant tolerance appears to be a potential alternative in clinical transplantation to solve the major drawbacks associated with current immunosuppressive therapies. However, attempts to transfer immunological tolerance induction strategies from experimental models to clinical settings have as yet not been very successful. In clinical organ transplantation, the establishment of tolerance is hampered by problems such as ischemia/reperfusion (I/R) injury and brain death of the organ donor. There is considerable evidence supporting the fact that I/R injury influences tolerance induction (1). Consequently, it was recently proposed that minimization of I/R injury will be a key element in tolerance induction strategies (2).

I/R injury induces an inflammatory reaction, which promotes chronic rejection and jeopardizes allograft survival (3,4). The I/R-related acute inflammation leads on one hand to acute organ damage and on the other hand boosts the host immune response by enhancing graft immunogenicity through upregulation of MHC class II antigens, ICAM-1, P- and E-selectin, as well as costimulatory molecules, particularly CD80 and CD86 (5–7). As an additional pro-inflammatory mediator also, the activation of complement plays a major role in I/R injury. While the involvement of the classical and alternative complement pathways in I/R injury is generally accepted (8–10), growing evidence suggests that also neoepitopes are exposed on EC and recognized by pattern recognition molecules such as mannan-binding lectin (MBL), either directly or via binding of naturally occurring IgM, triggering complement activation and subsequent graft damage (11,12). I/R injury-related enhancement of immune responses is in part also mediated by signaling via Toll-like receptors (TLR) (13). Moreover, soluble heparan sulfate proteoglycans (HSPG), shed from the endothelial glycocalyx due to I/R injury-induced EC activation, provoke TLR4-mediated dendritic cell activation and thereby link innate and adaptive immunity (14–16). Taken together, the activation of graft EC plays a major role in I/R-mediated regulation of the innate immune response and thus the outcome of graft survival.

Protection of the endothelium using EC protectants like low molecular weight dextran sulfate (DXS, MW 5000) has been shown to provide beneficial effects in hamster-to-rat and pig-to-human xenotransplantation models (17,18). Banz et al. demonstrated that protection of the cardiac endothelium using DXS mitigates I/R-mediated organ damage in a closed chest porcine model of acute myocardial I/R injury (19). Treatment with DXS induced targeted, site-specific cytoprotection and prevented inflammation in the above model. The beneficial effect of DXS in terms of attenuating innate immunity makes this substance a possible candidate for use in a setting of transplantation tolerance. In fact, strategies to inhibit innate immunity early after transplantation have previously been successful in experimental models. Use of the NF-κB inhibitor 15-deoxyspergualin induced allograft tolerance in an anti-CD3 immunotoxin-treated nonhuman primate model (20), and lack of MyD88 signaling induced long-term allograft survival in a mouse model (21), supporting the idea that inhibition of innate immunity provides a clinically relevant strategy to facilitate transplantation tolerance. We have, therefore, used a combined approach of attenuating I/R injury-induced innate immune response by DXS and nondepleting anti-CD4 mAb treatment (22) in order to facilitate long-term graft survival in a rat cardiac allotransplantation model with prolonged ischemia time.

Materials and Methods

Animals and cardiac transplantation

Inbred male DA (RT1av1) and Lewis (RT1 l) rats, 8–12 weeks of age, were purchased from Harlan Nederland B. V. (Horst, The Netherlands). Heterotopic heart transplantation from DA (donor) to Lewis (recipient) was carried out using a modification of the original technique described by Ono and Lindsey (23). The procured heart was kept for 20 min or 12 h on ice in Celsior (IMTIX Sangstat, Lyon, France) before transplantation. Time of rejection was defined as the day of complete cessation of heart beat. All experiments in this study were performed according to current versions of Swiss Laws on Animal Protection.

Anti-CD4 mAb and DXS treatment protocol

The anti-CD4 mAb RIB 5/2 was developed at the University of Rostock, Germany (22) and produced in vitro by EXBIO Praha (Prague, Czech Republic). Recipient rats were treated i.p. with RIB 5/2 at 20 mg/kg body weight on days—1,0,1,2 and 3 to induce long-term graft acceptance. DXS (MW, 5000) was purchased from Fluka (Buchs, Switzerland). Immediately before transplantation, the grafts were perfused with 25 mg of DXS dissolved in 2 mL of sterile PBS followed by perfusion with PBS to remove unbound DXS. After transplantation, the recipients were treated with i.v. DXS at 25 mg/kg of body weight on days 1 and 3 and then 12.5 mg/kg every other day until day 15.

Quantitative RT-PCR

Total RNA from homogenized tissue was isolated using the Miniprep Kit (Stratagene, Heidelberg, Germany) and reverse transcribed into cDNA using the murine leukemia virus reverse transcriptase (Gibco BRL, Gaithersburg, MD). Samples were analyzed using qRT-PCR as previously described (24). Reactions were run using the Model 7700 Sequence Detector (TaqMan™, Applied Biosystems, Foster City, CA). β-actin was used as a housekeeping gene for the samples. The sequences of the oligonucleotides and panels used in qRT-PCR were either described previously (25) or displayed in Table 1. Sense and antisense oligonucleotides were purchased from Metabion (Munich, Germany). Probes were obtained from Eurogentec (Cologne, Germany).

Table 1.  Oligonucleotides and panels used to quantify mRNA expression
Gene qRT-PCR primer/panel (assay no.)
CD40LRn00584362_m1 (Applied Biosystems)
CTLA4Rn00581545_m1 (Applied Biosystems)

Detection of allospecific antibody response

To analyze donor-specific alloantibody responses sera were collected from recipients at days 0, 3, 7, 21, 40 and 100. Aliquots containing 1 × 106 splenocytes from DA rats (donor strain) were incubated 45 min at 4°C with diluted recipient sera (1/4). The washed cells were then incubated 30 min at 4°C with FITC-conjugated goat anti-rat IgG (Serotec, Wiesbaden, Germany) and PE-conjugated goat anti-rat IgM (Jackson ImmunoResearch, Suffolk, UK). Cells were analyzed with a BD FACScan flow cytometer and the data quantified using FlowJo (Tree Star Inc., Ashland, OR). The levels were expressed as median channel fluorescence.

Histology and immunohistology

Tissue samples from the grafts were preserved in 4% buffered formaldehyde and stained with hematoxylin & eosin (H&E), Elastica van Gieson (EVG) and Masson's Trichrome (MT). For immunohistological analysis, cryosections (5μm) were stained with mouse anti-rat mAb against granulocytes (HIS48), monocytes/macrophages (ED1), CD4+ (W3/25), CD8+ (OX-8), all from Serotec, and T cells (15–16A1, Abcam, Cambridge, UK). Positive staining was detected using the Envision kit (DAKO). Twenty high-power viewing fields were evaluated per section and graded by eye on a 0 to 4 scale (0 = not detectable, 1 = mild, 2 = moderate, 3 = strong, 4 = very strong) by a researcher blinded to the study.

Immunofluorescence staining

Acetone-fixed cryosections were used for immunofluorescence staining. The primary antibodies used were mouse monoclonal anti-human heparan sulfate (Seikagaku, Tokyo, Japan), rabbit anti-human vWF and rabbit anti-human C3b/c (DAKO), mouse monoclonal anti-rat CD31 (Serotec) and rabbit anti-rat C9 antibody (kind gift of Prof. P. Morgan, University of Cardiff, UK). All human-specific primary antibodies were cross-reactive with the respective rat antigens. Secondary antibodies were goat anti-rabbit IgG-FITC (Southern Biotechnology Associated, Birmingham, AL), goat anti-mouse IgM-FITC (Jackson) and goat anti-mouse IgG-FITC (Sigma). Fluorescein-labeled DXS (DXS-Fluo) was produced and tested as described previously (26). Images were acquired with a fluorescence microscope (Nikon Eclipse TE2000-U) and digital camera (Nikon DXM1200 F). Images were captured with identical exposure times and settings in each experiment. Image-J software ( was used to quantify fluorescence intensities.

Complement hemolytic assay (CH50), activated partial thromboplastin time (aPTT) tests and platelet counts

Hemolytic assays (CH50) were used to determine classical complement pathway activities in rat serum samples as described previously (17). Standard activated partial thromboplastin time (aPTT) tests were performed with rat citrate plasma using Dade Actin FS reagent to evaluate the effect of DXS treatment on the coagulation system. Platelet counts were analyzed with an automated counter (K-1000, Sysmex Digitana AG, Horgen, Switzerland).

Statistical analysis

Statistical analysis was performed using the GraphPad Prism program (GraphPad Software Inc., San Diego, CA). The results are expressed as mean ± SD. Survival of the allografts was examined using Kaplan–Meier analysis and groups were compared using the log-rank test. Data were analyzed using Student's t-test. Data for gene expression were analyzed using Mann–Whitney U-test. Significance was determined with p < 0.05.


Graft survival

In the experimental groups with 20 min of cold graft ischemia, treatment with RIB 5/2 mAb only or in combination with DXS significantly prolonged allograft survival as compared with the PBS-treated vehicle control group (p < 0.001). MST of the RIB 5/2 only and RIB 5/2+DXS groups were 86 and 81 days, respectively (Table 2). In both groups 28% of the grafts survived more than 175 days. To determine the effect of DXS, graft survival was monitored after treatment with DXS only and compared with the PBS-treated control group. MST of these groups were 7.0 and 5.5 days, respectively. Although DXS treatment prolonged the graft survival compared with PBS controls, analysis of the grafts in both groups revealed signs of severe acute rejection (results not shown).

Table 2.  Survival of grafts after 20 min of cold ischemia in recipients treated with PBS, DXS and RIB 5/2 alone or in combination with DXS
GroupnGraft survival (day)Median survival time
PBS (control)45, 5, 6, 6 5.5
DXS66, 6, 7, 7, 7, 8 7.0
RIB 5/2774, 74, 84, 86, 113, >175, >17586.0
RIB 5/2+DXS769, 72, 76, 81, 107, >175, >17581.0

To be closer to the clinical situation of deceased heart donors, long-term survival induced by RIB 5/2 mAb treatment, with or without additional treatment by DXS, was challenged by prolonged graft ischemia time. After 12 h of cold ischemia, grafts were transplanted and the recipients received RIB 5/2 alone or in combination with DXS. Treatment with RIB 5/2 only was no longer sufficient for maintaining the long-term graft survival and all grafts were rejected with a MST of 49.5 days (Figure 1). Moreover, with RIB 5/2 only, the MST was significantly reduced in the 12-h ischemia group as compared with 20-min ischemia (p < 0.001). As compared with RIB 5/2 only, graft survival was significantly increased in RIB 5/2+DXS-treated recipients (MST >100 days, p < 0.05). In this group, more than 57% of the grafts survived >100 days. In another, separate series of experiments, the same protocol was repeated and survival was monitored up to 40 days. In this study, the graft survival percentage was 100% and 66% at 40 days in the RIB 5/2+DXS and RIB 5/2 only-treated groups, respectively (n = 6 per group). There was no significant difference in MST between 20-min and 12-h ischemia groups treated with RIB 5/2+DXS (p = 0.858).

Figure 1.

DXS facilitates RIB 5/2 induced long-term graft survival after 12 h of cold graft ischemia time. DA-to-Lewis heterotopic rat cardiac allotransplantation was performed with 12 h of cold graft ischemia. Recipients were treated with either RIB 5/2 only or RIB 5/2+DXS. Kaplan–Meier graft survival curves are shown for both treatment groups. *p < 0.05 as compared with RIB 5/2 only.


Histologies of long-term surviving and chronically rejected grafts were compared with naïve hearts (Figure 2A). Chronically rejected grafts in both groups with 20-min ischemia showed cell infiltration, intimal thickening and fibrosis (Figure 2B). Histological analysis of long-term surviving grafts showed no evidence of cellular infiltration, parenchymal fibrosis or intimal thickening in either group (Figure 2C). Grafts harvested after chronic rejection (MST = 49.5 days) from recipients treated with RIB 5/2 only in the 12-h ischemia group showed characteristic signs of chronic rejection. Severe cellular infiltrations were observed in all these grafts by H&E staining (Figure 2D). Moreover, EVG and MT stainings revealed marked intimal thickening and parenchymal fibrosis, respectively. In contrast, long-term surviving grafts, from recipients treated with RIB 5/2+DXS, did not show detectable cellular infiltration, intimal thickening or parenchymal fibrosis (Figure 2E). Histology of chronically rejected grafts from RIB 5/2+DXS-treated recipients was similar to chronically rejected grafts after RIB 5/2 only-treatment (not shown).

Figure 2.

Histology of chronically rejected and long-term surviving allografts. Hematoxylin and eosin (H&E), Elastica van Gieson (EVG) and Masson's Trichrome (MT) stainings were performed in both chronically rejected and long-term surviving allografts in the 20-min or 12-h ischemia groups. One representative example of six to seven independent experiments is shown, magnification: ×150.

Expression of gene markers

Chronically rejected (RIB 5/2 only, at time of rejection) and long-term surviving allografts (RIB 5/2+DXS, 100 days), both with 12-h cold ischemia, were analyzed for markers described in the context of allograft tolerance as well as rejection. In addition, grafts of both groups were analyzed at 40 days. The changes in gene expression, indicated as ‘fold changes’ in the RIB 5/2+DXS versus the RIB 5/2 only-treated groups, are shown in Table 3. Most of the observed changes did not reach statistical significance. However, the recently described gene TOAG-1, associated with long-term graft acceptance (25), was upregulated 2.7-fold (p = 0.031, n = 5) and 82.9-fold (p = 0.028, n = 4) in allografts treated with RIB 5/2+DXS at 40 days and 100 days, respectively (Table 3). The other significantly upregulated genes were CD25 (p = 0.030), CD3 (p = 0.031 and CD40 L (p = 0.031), all at 40 days.

Table 3.  Allograft gene expression in RIB 5/2+DXS-treated and RIB 5/2 only-treated groups with 12-h cold ischemia at 40 and 100 days posttransplantation
GeneCategoryRIB 5/2+DXS/RIB 5/2 only 40 day post tx Fold change (mean, n = 5)RIB 5/2+DXS (100 days)/2 only (49.5 days) Fold change (mean, n = 4)
  1. *p < 0.05.

CD25Activated T-cell related +2.8*+1.3
CD3Costimulation +3.0*+1.8
CD40LCostimulation +3.9*+2.8
CTLA4 (CD152)Costimulation+2.3+1.6
Foxp3Regulatory T-cell related+3.5+2.9
PerforinCytotoxic T-cell related+1.7+1.2
RORγtTh17-cell related+3.9+3.5
TOAG-1Unknown   +2.7*+82.9* 

Humoral immune response

Serum IgM levels against donor splenocytes showed a time-dependent increase in the RIB 5/2 group with 12 h of cold ischemia, whereas they remained more or less stable in the respective RIB 5/2+DXS group (Figure 3). Similarly, antidonor IgG levels did not change in the RIB 5/2+DXS group, whereas a steep increase was seen in RIB 5/2-only treated recipients that peaked at day 7, posttransplantation and then gradually declined with time.

Figure 3.

Humoral immune response. Sera from RIB 5/2 only and RIB 5/2+DXS-treated groups were collected at days 0, 3, 7, 21, 40 and 100. Splenocytes from DA rats were incubated with the sera at 1/4 dilution and stained with FITC-conjugated anti-rat IgG and PE conjugated anti-rat IgM. Flow cytometric analysis of serum IgG and IgM alloantibody levels are shown. Data are represented as mean ± SD (n = 4).

Effect of prolonged ischemia time on early posttransplantation changes

Grafts subjected to 20-min or 12-h cold ischemia were analyzed histologically 1 day after transplantation. Only minimal histological changes were observed in the 20-min ischemia group treated with either PBS, RIB 5/2 or RIB 5/2+DXS. These changes were characterized by mild cellular infiltration and perivascular edema, and were comparable among all the treatment groups (Figures 4A, C, E, G). In the PBS-treated control group with 12 h of graft ischemia time, we observed severe cellular infiltrations, edema, hemorrhages and myocardial necrosis (Figure 4B). RIB 5/2-only treatment reduced the severity of these histological changes, but a marked ongoing inflammation, characterized by cellular infiltration, edema and hemorrhages, was clearly observed in all grafts (Figure 4D). Strikingly, treatment by RIB 5/2+DXS as well as DXS alone reduced the inflammatory changes as compared with RIB 5/2 only (Figure 4F, H). DXS treatment thus clearly reduced early graft injury, i.e. cellular infiltration, edema and other inflammatory changes in the 12-h graft ischemia group.

Figure 4.

Histology of the allografts at day 1 posttransplantation. DA hearts subjected to either 20 min or 12 h of cold ischemia were transplanted into Lewis rats. Recipients were treated with PBS, RIB 5/2, RIB 5/2+DXS and DXS alone, and sacrificed at 1 day posttransplantation. H&E stained paraffin sections are shown, one representative picture out of three independent experiments, magnification: ×150

DXS attenuates infiltration of mononuclear cells

Since histological analysis revealed significant reduction of graft inflammatory cell infiltration in RIB 5/2+DXS treatment compared with RIB 5/2 only, we analyzed immunohistochemically the type of infiltrating cells. As shown in Figure 5, RIB 5/2+DXS or DXS-only treatment led to a significant reduction of both granulocyte (HIS48) and monocytes/macrophages (ED1) infiltration into the graft tissue. Quantitative analysis of cell infiltration revealed that the numbers of ED1- and HIS48-positive cells were significantly reduced in grafts from RIB 5/2+DXS-treated recipients compared with RIB 5/2 only-treatment (Figure 5B, p < 0.01). Only minimal infiltration by T cells (15–16A1) was observed with no difference between the two groups, and also the detection of CD4+ (W3/25) and CD8+ T cells (OX-8) revealed only minimal staining with no apparent difference between the RIB 5/2 only and RIB 5/2+DXS groups (not shown).

Figure 5.

DXS Inhibits infiltration of granulocytes and monocytes/macrophages. Grafts subjected to 12-h cold ischemia were analyzed 1 day posttransplantation. (A) Immunohistochemical staining for granulocytes (HIS48), monocytes/macrophages (ED1) and T cells (15–16A1). Results are representative of three independent experiments, magnification: ×150 (B) Quantitative analysis of the positive staining (20 high-power fields scored by eye by an independent researcher blinded to the study). Data are represented as mean ± SD (**p < 0.01, n = 3).

Activation and damage of graft endothelial cells

Graft EC were assessed 1-day posttransplantation in the 12-h ischemia groups. Staining for vWF and CD31 was circumferentially detected in blood vessels of grafts of RIB 5/2+DXS-treated recipients and this was comparable with the naïve heart of these animals (Figure 6A). In contrast, remarkably diminished stainings for vWF and CD31 were observed in allografts from RIB 5/2 only-treated recipients. As compared with the naïve hearts, staining for heparan sulfate (HS) was markedly reduced in the RIB 5/2 only-treated group (Figure 6B). Interestingly, quantitative analysis of HS in RIB 5/2+DXS-treated grafts revealed higher fluorescence intensities than the naïve hearts. This may be due to binding of DXS to the endothelium and subsequent cross-reaction with anti-HS antibody. In fact, cross-reactivity of the used anti-HS antibody with DXS was confirmed in a competitive inhibition experiment (data not shown).

Figure 6.

EC integrity and shedding of heparan sulfate. DA hearts were subjected to 12 h of cold ischemia and transplanted into Lewis rats. Recipients were treated with RIB 5/2 or RIB 5/2+DXS and sacrificed at 1 day posttransplantation. Naïve hearts from graft recipients are shown as controls. (A) EC integrity was analyzed by immunofluorescence staining for vWF and CD31. (B) Immunofluorescence staining for heparan sulfate (HS). (C) Quantitative image analysis of HS staining by Image-J software. Results are representative of three independent experiments, magnification: ×400.

Binding of DXS-Fluo to the allograft vasculature

The grafts subjected to 20 min or 12-h ischemia were perfused with 10 mg of DXS-Fluo in 2 mL of sterile PBS followed by perfusion with PBS in order to remove unbound substance. Analysis of these grafts revealed strong binding of DXS-Fluo only in the 12-h ischemic grafts and there was no binding to grafts after only 20 min of cold ischemia (Figure 7). Binding of DXS-Fluo was completely blocked after the prior perfusion of the graft with unlabelled DXS (data not shown). In addition, DXS was administered in vivo at 25 mg/kg i.v. 1 day after transplantation, followed by immediate graftectomy and analysis for binding of DXS-Fluo. Also, in these in vivo experiments, binding of DXS-Fluo was only observed in grafts subjected to 12 h of cold ischemia. Moreover, no binding of DXS-Fluo was detected in the native heart, liver, kidney, lung and spleen (data not shown).

Figure 7.

Binding of fluorescence labeled DXS (DXS-Fluo) to allografts. Donor hearts were subjected to 20-min or 12-h cold ischemia in Celsior solution followed by perfusion with 10 mg/mL of DXS-Fluo (2 mL). Tissues were embedded in OCT, snap-frozen, sectioned and analyzed. (A) DXS-fluo binding to the graft subjected to 20 min and (B) 12 h of cold ischemia. One representative example of three independent experiments is shown, magnification: ×400.

DXS site specifically attenuates deposition of complement

Deposition of C3b/c and C9 was detected in grafts of RIB 5/2-treated recipients and significantly reduced by additional DXS treatment in the 12-h ischemia group (Figure 8A). In addition, the effect of DXS on systemic complement activation was analyzed. Lewis rats were treated with a single dose of 25 mg/kg of DXS i.v. without receiving a transplant, and their serum analyzed for classical pathway complement activity by a CH50 test. Six hours after injection of DXS the CH50 values were only 11% lower than the recorded baseline value (Figure 8C; p = 0.684, 6 h vs. baseline). Also, long-term monitoring up to 5 days during ongoing DXS treatment in transplant recipients revealed no reduction of CH50 values (Figure 8D), suggesting that the used dosage and application scheme of DXS had no significant influence on systemic complement activity.

Figure 8.

Effect of DXS treatment on complement deposition. (A) DA hearts subjected to 12 h of cold ischemia were transplanted into Lewis rats, which received RIB 5/2 with or without DXS. Recipients were sacrificed at day 1 after transplantation and the grafts immunostained for C3b/c and C9. (B) Quantitative analysis of deposition of complement by Image-J software. The shown sections are representative of three analyzed grafts from each group, magnification: ×400. Complement activity in serum was monitored by CH50 assay after (C) a single dose of i.v. injection of DXS (25 mg/kg) or (D) continuous treatment until day 5.

Effect of DXS on coagulation system

The effect of DXS on the coagulation system was monitored after administration of a single dose of DXS at 25 mg/kg i.v. The aPTT values rose to >600 s for up to 90 min after DXS administration and then quickly dropped again. After 360 min the values were back to nearly baseline level (Figure 9A, n = 3). Therefore, systemic administration of DXS led to a transient inhibition of the coagulation system. In contrast, we did not observe an effect of DXS on platelet counts. Platelet counts changed from 544 ± 32 × 103 cells/μL at baseline to 600 ± 64 × 103 cells/μL 1 day after DXS injection (Figure 9B; p = 0.269, n = 3).

Figure 9.

Influence of DXS on the coagulation system. Coagulation activity and platelet counts were monitored in blood samples after a single dose of i.v. injection of DXS (25 mg/kg). Citrate blood was collected at the indicated time points and analyzed for (A) aPTT and (B) platelets. Data are represented as mean ± SD.


Early inflammatory changes associated with brain death and prolonged cold ischemia, critically influence the function and survival of allografts (27–29). Indeed, strategies aimed at prevention of graft I/R injury were shown to promote long-term allograft acceptance (20,21). In this study, a new therapeutic approach targeting innate immunity by using low molecular weight dextran sulfate in combination with the nondepleting anti-CD4 mAb RIB 5/2, significantly extended the survival of allografts transplanted after prolonged cold ischemia.

As expected, RIB 5/2 treatment was able to prolong graft survival in recipients of cardiac allografts, which were subjected to only 20 min of cold ischemia. However, the same treatment led to chronic rejection, characterized by severe inflammatory cell infiltration, intimal thickening and fibrosis, which are typical features of cardiac allograft vasculopathy (CAV), if the cold ischemia time of the grafts was extended to the clinically more relevant 12 h. This picture of CAV could be prevented and long-term graft survival reestablished if RIB 5/2 treatment was complemented by i.v. injection of DXS for 2 weeks posttransplantation. Intragraft gene expression analysis at 40 days and 100 days revealed a significant upregulation of the gene TOAG-1 at both time points, consistent with the recent finding that this gene is specifically and reproducibly upregulated in long-term surviving allografts (25). The observed upregulation of genes such as CD25, CD3 and CD40 L in long-term surviving allografts may be linked to the fact that long-term graft acceptance is a result of a complex and dynamic interplay between regulatory and effector T cells. Pro-inflammatory factors were described to be upregulated in long-term surviving grafts compared with rejected grafts and to play a dominant role in maintaining tolerance (30,31). Moreover, elevation of pro-inflammatory related genes does not necessarily reflect the rejection rate (30,32).

In our experiments, chronic graft rejection was linked to a pronounced IgM and IgG alloantibody response, which was absent in RIB 5/2+DXS-treated recipients with long-term surviving allografts. This finding is in line with recent reports (31), and we conclude that inhibition of the early innate immune response by DXS treatment may play an important role in lowering the antigraft IgG and IgM response thus prevent chronic rejection.

We showed that prolonged cold graft ischemia for 12 h, but not 20 min ischemia, critically influenced damage and activation of the graft endothelium characterized by the shedding of HSPG. As a consequence, the graft endothelium seems to become a target for binding of DXS, which then reestablishes its anti-inflammatory and anticoagulatory properties. In other words, DXS reinstates the ‘nondangerous’ (33) properties of the graft endothelium. Using a fluorescence-labeled variant of DXS, DXS-Fluo, we could indeed observe binding of DXS to graft endothelium after 12 h of cold ischemia, but not after 20 min. As shown earlier by ourselves and others (34,35), binding of DXS was inversely correlated to the endothelial expression of HSPG.

Activation of the complement system and subsequent complement-mediated tissue damage is a key feature of I/R injury. While DXS had almost no systemic effect on the complement system in our experiments, it significantly reduced local complement deposition at 1-day posttransplantation. In line with reports, which highlight the importance of local complement production (36), our results indicate that local inhibition of complement activation may protect the allograft during the tolerance induction phase and therefore facilitate the latter. Moreover, early graft infiltration by granulocytes and monocytes/macrophages critically influences the long-term alloimmune response (37,38). Therefore, it is tenable that the observed inhibition of granulocyte and monocytes/macrophages infiltration by DXS is positively correlated with long-term graft survival. Matsumiya et al. showed that DXS inhibits E-selectin mediated neutrophil adhesion to activated EC (39). Furthermore, DXS may attenuate anaphylatoxin-mediated recruitment of inflammatory cells through inhibiting complement. Consistent with this observation, it was also suggested that DXS has immunosuppressive properties. DXS was shown to inhibit the IFN-γ-induced MHC class II expression on endothelial cells (40). Although soluble DXS serves as a competitive inhibitor of the binding of IFN-γ to membrane IFN-γRs, immobilized DXS can bind with IFN-γ (41), suggesting that DXS has the capacity to bind immunoregulatory elements and subsequently target them to activated graft endothelium. However, the relative importance of DXS-mediated binding of immunoregulatory elements to the EC surface remains to be determined in separate experiments.

As expected by the fact that DXS is an anticoagulatory substance (42), it has a transient effect on coagulation in the treated graft recipients. However, administration of the indicated dose of DXS in our model did not cause any major bleeding complications. It is as yet unclear whether or not this anticoagulatory effect is beneficial for long-term graft acceptance. In view of a clinical application of DXS or similarly acting substances, the anticoagulatory effect needs to be considered, and further investigations on the need of anticoagulation to achieve the DXS-mediated attenuation of innate immunity are warranted. Generation of antidextran sulfate antibodies, which might limit the clinical use of DXS, has not been described following i.v. administration so far, despite of quite extensive use of DXS in different animal models as well as pre-clinical studies (17,43,44).

It has been reported that the use of FTY720 in combination with RIB 5/2 prevents tolerance induction (45). Therefore, the selection of drugs for synergistic treatment with tolerance induction protocols appears to be a challenge. Our novel approach of graft-targeted treatment by DXS might thus offer new, clinically relevant perspectives to attenuate innate immunity in the context of tolerance induction.


This study was supported by the EU 6th framework program RISET (Reprogramming the Immune System for the Establishment of Tolerance), the Swiss National Science Foundation (grants no. 3200B0–104228 and 3200B0–116618) and the Deutsche Forschungsgemeinschaft (DFG; SFB 650). We thank the laboratory of Professor Thomas Schaffner, University of Bern, for the preparation of histology slides, Prof. Maria-Cristina Cuturi, INSERM, University of Nantes, France, for helpful discussions and Prof. Paul Morgan, University of Cardiff, UK, for kindly providing us rabbit antibody against rat C9.