It was previously reported that treatment with leflunomide (LF; 10 mg/kg/day) together with cyclosporine (CsA; 10 mg/kg/day) resulted in long term survival of hamster heart xenografts (Xg) in rats and that LF could be withdrawn 2 to 4 wk after transplantation. To study the mechanisms allowing withdrawal of LF, second hamster heart Xgs were transplanted 6 wk after the first xenograft. Only the rats that received LF for 4 wk accepted second Xgs (>30 days; n = 5). Hence, after 4 wk of LF, the rats developed partial B cell tolerance, as they were unable to produce T-independent (CsA-resistant) XAbs. Rejection of second Xgs (2–4 days; n = 5) in the 2-wk LF group resulted in the formation of IgM xenoantibodies (XAbs) localizing together with complement within rejected grafts. However, these XAbs did not affect first Xgs, suggesting that the latter Xgs became resistant to this IgM XAb-mediated rejection, a phenomenon referred to as accommodation. Accommodation was further confirmed as adoptive transfer of IgM XAbs, which resulted in hyperacute Xg rejection in naive rats (<1 h; n = 5), did not cause rejection in long term survivors (>30 days; n = 4). This was associated with a down-regulation of the expression on the graft endothelial cells of adhesion molecules (believed to be important expressers of xenogeneic epitopes), such as P- and E-selectins. Interestingly, these adhesion molecules reappeared after retransplanting the accommodated Xgs to naive recipients. In conclusion, depending on the duration of the LF treatment, long term survival of hamster hearts in CsA-treated rats is based in part on accommodation and in part on T-independent B cell tolerance.

Advances in immunosuppression have led to successful prevention of hyperacute Xg3 rejection (1, 2, 3) and significant prolongation of Xg survival in several experimental models (4, 5, 6). Therapeutic strategies to achieve this were mainly based on depletion or suppression of xenoreactive Abs and/or complement in combination with continuous nonspecific immunosuppressive treatment (1, 2, 3, 4, 5, 6). Because many of these immunosuppressive treatments may cause severe side effects in humans, it is believed that the eventual success in clinical xenotransplantation will depend on the possibility to taper immunosuppression after transplantation. This goal may be achieved by induction of xenotransplant accommodation or xenotransplant tolerance.

Xenotransplant accommodation refers to a phenomenon in which an Xg acquires resistance to Ab and complement-mediated humoral immunity (7). The existence of accommodation appeared from observations that in some cases, when anti-donor XAbs and/or complement had been depleted from a Xg recipient for a period of time, the graft continued to function even after the Ab titer and complement function were restored (7, 8, 9, 10, 11). In a recent study involving CsA-treated rats in which the complement system was temporarily blocked by cobra venom factor, accommodation of hamster heart Xgs was demonstrated. This accommodation involved Th2 lymphocytes and cytokines, and seemed to be maintained by the induction of anti-apoptotic genes (12). The development of T cell xenotransplantation tolerance was also recently demonstrated by grafting thymectomized, T cell-depleted mice with xenogeneic fetal pig thymus and liver tissue. Subsequently, these mice developed specific T cell tolerance for pig Ags and permanently accepted pig skin grafts (13). However, for vascularized Xgs, many non-T cell-mediated immune responses, involving T-independent B cells, NK cells, and macrophages, may lead to rejection of Xgs before T lymphocytes are even activated by xenoantigens (14, 15).

Previous studies from our and other laboratories have shown that a combination therapy of CsA with leflunomide (LF), a novel immunosuppressant, allowed for long term survival of hamster heart Xgs in rats (16, 17). This synergistic effect appeared to result from the different actions of LF and CsA to block T-independent and T-dependent XAb formation, respectively (16). In addition, it was shown that after a short course (2–4 wk), LF could be withdrawn, and xenografts survived indefinitely with CsA treatment alone (16). The mechanisms allowing for this progressive withdrawal of LF from the therapy were investigated here by experiments involving second xenotransplantation, adoptive transfer of hyperimmune serum, and retransplantation of the long term surviving Xgs.

Inbred 2- to 4-mo-old male PVG rats, weighing ±200 g, were used as recipients. Golden Syrian hamsters, weighing ±100 g, were used as donors. Animals were kept in conventional housekeeping facilities.

First, hamster hearts were transplanted in the neck of recipient rats. The recipients were anesthetized by Enflurane (Abbott S.A., Ottignies, Belgium) supplemented with oxygen through a semiclosed circuit inhalation system. The aorta and pulmonary artery of the donor were anastomosed to the carotid artery and jugular vein of the recipient by either microsuture or cuff techniques. A second hamster heart was transplanted to the contralateral side of the recipient neck or in the abdominal cavity. Retransplantation was performed by implanting long term surviving hamster heart Xgs to naive recipient rats. The function of the grafts was monitored by daily inspection and palpation. Rejection was determined by the cessation of beating of the graft and was confirmed by histology.

LF (HWA 486, Hoechst, Germany) was provided by Dr. Bartlett (Hoechst, Wiesbaden, Germany) and was suspended in 1% carboxymethylcellulose just before gavage. CsA (Sandimmun, Sandoz, Belgium) was diluted in olive oil and given by gavage.

Whole blood (0.5–2 ml) was taken by heart puncture at various times after transplantation. After 40 min at room temperature, the serum was separated from the blood by centrifugation, and aliquots were stored at −70°C until use.

The IgM and IgG isotypes of anti-hamster XAbs were determined by flow cytometry using hamster PBMC as targets. Aliquots of 0.5 × 106 PBMC were cultured for 30 min at 4°C with 100 μl of 1/10 diluted serum taken from recipient rats at various times after transplantation. After a secondary staining with FITC-conjugated anti-rat IgM (The Binding Site Ltd., Birmingham, U.K.) or anti-rat whole IgG (ST-AR 17, Serotec, Oxford, U.K.) Abs, the cells were examined by flow cytometry. Results were expressed as relative mean channel fluorescence, which was calculated as the mean channel fluorescence of stained cells divided by the mean channel fluorescence of cells incubated with control serum and counterstained with FITC-conjugated anti-rat IgM or IgG Abs.

Anti-hamster IgM XAb-containing serum (0.5 ml), taken from CsA-treated rats that rejected hamster heart Xgs with the production of IgM, but not IgG, XAbs, were injected i.v. into rats carrying long term (6 wk) surviving hamster Xgs or into naive rats carrying a newly transplanted Xg (1 h after transplantation).

Graft samples for histology were fixed in 10% formalin solution, paraffin embedded, sectioned, and stained with hematoxylin and eosin. Graft samples for immunopathology were snap-frozen in prechilled isopentane and stored at −70°C. Specimens were cut into 4-μm sections in a cryostat at −25°C, air-dried, fixed in ice-cold acetone for 10 min, and washed with PBS. For immunofluorescence staining, each section was incubated with FITC-conjugated anti-rat IgM (The Binding Site), IgG (ST-AR 17, Serotec), C3 (Cappel, Organon Teknika, Turnhout, Belgium), and anti-human C6, shown to be cross-reactive with rat C6 (18). Before use, the Abs were absorbed with hamster tissues. Sections were washed with PBS and mounted in 50% glycerol in PBS. Sections of naive hamster hearts were used as negative controls. For immunoperoxidase staining, sections were incubated overnight with a rabbit Ab for P-selectin (PharMingen, San Diego, CA) and rat mAb for E-selectin (PharMingen). The sections were then stained with labeled streptavidin-biotin reagents (Dako LSAB kit, peroxidase, Dako, Carpenteria, CA).

Results were statistically analyzed by two-sample Student’s t test or by the log rank or Wilcoxon’s test when appropriate.

Table I shows Xg survival after withdrawal of LF from the CsA/LF combination immunosuppression. As demonstrated previously (16), when CsA was combined with LF and administrated continuously, all hamster heart Xgs survived (group 1). When LF was withdrawn after 1 wk (group 2), Xgs were rejected after a mean survival time (MST) of 10 ± 1 days. In contrast, when LF was withdrawn after 2 (group 3) or 4 (group 4) wk, respectively, 7 of 9 and 12 of 12 grafts survived under CsA monotherapy. However, if both drugs were withdrawn on day 30, Xgs were invariably rejected (5 of 5) after an MST of 39 ± 2 days (group 5).

Table I.

Graft survival after withdrawal of LF from the LF+CsA combination therapy

GroupTreatmentDose (mg/kg/day)Duration (day)nGraft Survival (days)Mean ± SD (days)
LF 10 −1→a >60, >113, >60-113a 
 CsA 10 −1→a  >120, >120,  
     >130, >132  
LF 10 −1→7 9, 10, 10, 10 ± 0.7 
 CsA 10 −1→a  10, 11  
LF 10 −1→14 18, 23, >60-120a 
 CsA 10 −1→a  >60, >60  
     >120 (x5)  
LF 10 −1→30 12 >60 (x4) >60-120a 
 CsA 10 −1→a  >90 (x4)  
     >120 (x4)  
LF 10 −1→30 36, 38, 39 39 ± 2 
 CsA 10 −1→30  39, 42,  
GroupTreatmentDose (mg/kg/day)Duration (day)nGraft Survival (days)Mean ± SD (days)
LF 10 −1→a >60, >113, >60-113a 
 CsA 10 −1→a  >120, >120,  
     >130, >132  
LF 10 −1→7 9, 10, 10, 10 ± 0.7 
 CsA 10 −1→a  10, 11  
LF 10 −1→14 18, 23, >60-120a 
 CsA 10 −1→a  >60, >60  
     >120 (x5)  
LF 10 −1→30 12 >60 (x4) >60-120a 
 CsA 10 −1→a  >90 (x4)  
     >120 (x4)  
LF 10 −1→30 36, 38, 39 39 ± 2 
 CsA 10 −1→30  39, 42,  
a

Until sacrifice for histology.

While a combination of LF/CsA resulted in a continuous absence of XAb formation (Fig. 1,A), when LF was withdrawn after 1 wk, a rapid increase of mainly IgM anti-donor XAbs was noticed (Fig. 1,B). In contrast, when LF was given for 2 (Fig. 1,C) or 4 (data not shown) wk, no XAbs occurred after withdrawal of LF. However, when both LF and CsA were withdrawn after 4 wk, a rapid rise of predominantly IgG anti-donor XAbs was detected (Fig. 1 D).

FIGURE 1.

XAb formation after withdrawal of LF and/or CsA. A, LF and CsA were given continuously. B, LF was withdrawn after 1 wk; CsA was continued. C, LF was withdrawn after 2 wk; CsA was continued. D, Both drugs were withdrawn after 4 wk.

FIGURE 1.

XAb formation after withdrawal of LF and/or CsA. A, LF and CsA were given continuously. B, LF was withdrawn after 1 wk; CsA was continued. C, LF was withdrawn after 2 wk; CsA was continued. D, Both drugs were withdrawn after 4 wk.

Close modal

As a 2- or 4-wk course of LF was sufficient to block IgM XAb formation and to induce long term Xg survival in CsA-treated rats, it was hypothesized that LF had induced tolerance in T-independent B lymphocytes. To verify this hypothesis, second xenotransplantations were performed 6 wk after the first xenografting. As shown in Figure 2,A, when the second Xgs were transplanted in the rats that received only 2-wk LF treatment, the second Xgs invariably underwent an acute rejection within 4 days after transplantation (MST = 3 ± 1 days; n = 5). Interestingly, this process of second Xg rejection did not influence the survival of the first Xgs. In contrast, in the 4-wk LF-treated group, four of five second Xgs were accepted indefinitely. Again, second xenografting did not influence the survival of the first Xgs. In some cases (Fig. 2 C), the surviving first Xg was removed 3 days before second xenotransplantation in the 4-wk LF group. This did not significantly affect the acceptance of the second Xg.

FIGURE 2.

Second hamster heart Xg survival. Six weeks after first xenografting, second hamster hearts were transplanted in CsA-treated rats carrying surviving first Xgs after a 2-wk LF (group A) or a 4-wk LF induction therapy (group B). In some cases of the 4-wk LF group, the first Xg was removed 3 days before second xenotransplantation (group C). ▪, First Xgs; ○, second Xgs.

FIGURE 2.

Second hamster heart Xg survival. Six weeks after first xenografting, second hamster hearts were transplanted in CsA-treated rats carrying surviving first Xgs after a 2-wk LF (group A) or a 4-wk LF induction therapy (group B). In some cases of the 4-wk LF group, the first Xg was removed 3 days before second xenotransplantation (group C). ▪, First Xgs; ○, second Xgs.

Close modal

XAb formation after the second xenotransplantation is shown in Figure 3. After a second Xg, rats that received only a 2-wk course of LF developed significant levels of anti-donor IgM XAbs (Fig. 3,A), which correlated with the rejection of these second Xgs. In contrast, in the rats that initially received a 4-wk course of LF, no IgM XAbs were detectable (Fig. 3,B). In both cases, IgG XAbs were not detected (Fig. 3, C and D), indicating that the CsA treatment efficiently blocked T-dependent IgG XAb formation.

FIGURE 3.

XAb formation after second xenotransplantation. A andB, Anti-hamster IgM in the 2-wk LF group (A) or in the 4-wk LF group (B); C and D, anti-hamster IgG in the 2-wk LF group (C) or in the 4-wk LF group (D). (The profiles marked represent the sera taken 3 days after second Xgs; the other profiles represent the sera taken before second grafting.) One representative of three similar results from different individual rats in each group is shown.

FIGURE 3.

XAb formation after second xenotransplantation. A andB, Anti-hamster IgM in the 2-wk LF group (A) or in the 4-wk LF group (B); C and D, anti-hamster IgG in the 2-wk LF group (C) or in the 4-wk LF group (D). (The profiles marked represent the sera taken 3 days after second Xgs; the other profiles represent the sera taken before second grafting.) One representative of three similar results from different individual rats in each group is shown.

Close modal

The observation that in 2-wk LF-treated rats, IgM XAbs provoked after the second xenotransplantation destroyed only the second, but not the first, Xgs suggested that the first Xgs had become resistant to this IgM XAb-mediated rejection, and, hence, that some form of accommodation was achieved. This was further investigated by an experiment involving adoptive transfer of anti-donor IgM XAbs taken from CsA-treated rats rejecting a hamster heart Xg. We previously showed that these rats produced only IgM XAbs (16). Transfer of 0.5 ml of this serum provoked hyperacute Xg rejection in naive rats treated with CsA only (MST = 30 ± 20 min; n = 5; Table II). In contrast, the same procedure did not cause rejection of primary Xgs in the 2-wk LF-treated group when injected 6 wk after transplantation (Table II). Hence, the latter Xgs had become resistant to IgM XAbs. Together with the presence of the normal level of lytic complement in these rats (data not shown), this observation fits with the definition of accommodation.

Table II.

Xenograft survival after transfer of anti-hamster IgM XAbsa

RecipientTreatment (mg/kg/day)Graft survivalMST ± SD
Newly grafted CsA 10 30, 10, 15, 30, 60 min 30 ± 20 min 
naive rats    
Long term CsA 10 >30 days (x4) >30 daysc 
grafted ratsb    
RecipientTreatment (mg/kg/day)Graft survivalMST ± SD
Newly grafted CsA 10 30, 10, 15, 30, 60 min 30 ± 20 min 
naive rats    
Long term CsA 10 >30 days (x4) >30 daysc 
grafted ratsb    
a

Taken from CsA-treated rats rejecting hamster heart xenografts.

b

Six weeks after hamster heart transplantation in the CsA+2-wk LF group.

c

Until the end of experiments.

Interestingly, after retransplantation in recipients treated with CsA only, the accommodated Xgs were rejected within 2 to 5 days (MST = 3.5 ± 1 days; n = 5) and provoked an induction of significant levels of anti-hamster IgM XAbs (Fig. 4). Hence, it appeared that the xenoantigens that were resistant to IgM XAbs in the accommodated hearts were able to provoke an IgM XAb-mediated rejection. This suggested that the process underlying accommodation of Xgs was reversible after retransplantation.

FIGURE 4.

XAb formation after retransplantation. Accommodated hamster heart Xgs taken from the 2-wk LF group 6 wk after transplantation were retransplanted to naive rats treated with CsA only. The data are expressed as the mean ± SD of four rats per group.

FIGURE 4.

XAb formation after retransplantation. Accommodated hamster heart Xgs taken from the 2-wk LF group 6 wk after transplantation were retransplanted to naive rats treated with CsA only. The data are expressed as the mean ± SD of four rats per group.

Close modal

The histology of all rejected hamster hearts, including first, second, and retransplanted Xgs, showed similar signs of acute vascular rejection. The histology of a hamster heart rejected by an untreated rat is shown in Figure 5,A; intravascular thrombosis, interstitial hemorrhage, edema, myocardial necrosis, and polymorphonuclear cell infiltration were noticed. In contrast, long term surviving Xgs (Fig. 5 B) showed normal vessels without intima proliferation, a healthy myocardium, and the absence of immune cell infiltration.

FIGURE 5.

Histology of control rejecting hamster hearts or accommodated hamster hearts removed 2 mo after transplantation. A, Hamster hearts rejected by untreated control rats, showing edema, thrombosis, myocardial necrosis, and polymorphonuclear cell infiltration (hematoxylin-eosin stain). B, Accommodated hamster heart Xgs taken from the 2-wk LF group, showing normal vessels without intima proliferation, healthy cardiac fibers, and an absence of cellular infiltration (Orcein stain).

FIGURE 5.

Histology of control rejecting hamster hearts or accommodated hamster hearts removed 2 mo after transplantation. A, Hamster hearts rejected by untreated control rats, showing edema, thrombosis, myocardial necrosis, and polymorphonuclear cell infiltration (hematoxylin-eosin stain). B, Accommodated hamster heart Xgs taken from the 2-wk LF group, showing normal vessels without intima proliferation, healthy cardiac fibers, and an absence of cellular infiltration (Orcein stain).

Close modal

As expected, immunofluorescence staining of Xgs rejected by untreated rats revealed vessel wall deposition of IgM XAbs (Fig. 6,A) and complement C6 (Fig. 6,B). Long term surviving first Xgs showed an absence of this deposition, even after rejection of second Xgs in the 2-wk LF group (Fig. 6, C and D). In the 2-wk LF group, deposition of IgM (Fig. 6,E) and C6 (Fig. 6,F) was clearly found in rejected second hamster hearts. In contrast, in the 4-wk LF group, both the long term surviving first (Fig. 6, G and H) and the accepted second Xgs (Fig. 5, I and J) showed an absence of deposition of IgM and complement. After retransplantation in CsA-treated recipients, accommodated hamster hearts were rejected and exhibited vessel wall deposition of both IgM (Fig. 5,K) and C6 (Fig. 5,L). Immunoperoxidase staining showed that after retransplantation, there was a progressive induction of the expression by the adhesion molecules of P-selectin (Fig. 7, A–D) and E-selectin (Fig. 7, E–H) on vessel endothelial cells of accommodated hamster hearts.

FIGURE 6.

Immunofluorescence staining for IgM and C6. A andB, IgM (A) and C6 (B) in hamster hearts rejected by untreated rats. In the 2-wk LF group, surviving first Xg had no deposition of IgM (C) and C6 (D), but rejected second Xgs showed both IgM (E) and C6 (F). In the 4-wk LF group, surviving first and second Xgs both showed the absence of IgM (G and I) and C6 (H andJ). Accommodated Xgs rejected after retransplantation showed severe vessel wall deposition of IgM (K) and C6 (L).

FIGURE 6.

Immunofluorescence staining for IgM and C6. A andB, IgM (A) and C6 (B) in hamster hearts rejected by untreated rats. In the 2-wk LF group, surviving first Xg had no deposition of IgM (C) and C6 (D), but rejected second Xgs showed both IgM (E) and C6 (F). In the 4-wk LF group, surviving first and second Xgs both showed the absence of IgM (G and I) and C6 (H andJ). Accommodated Xgs rejected after retransplantation showed severe vessel wall deposition of IgM (K) and C6 (L).

Close modal
FIGURE 7.

Immunoperoxidase staining of accommodated Xgs showing P-selectin (A–D) and E-selectin (E–H) at various times after retransplantation.

FIGURE 7.

Immunoperoxidase staining of accommodated Xgs showing P-selectin (A–D) and E-selectin (E–H) at various times after retransplantation.

Close modal

Recently, we and others demonstrated that treatment with LF, a recently developed immunosuppressant, synergized with CsA to achieve permanent hamster heart survival in rats (16, 17). It was also shown that after a period of 2 to 4 wk, LF could be withdrawn without provoking rejection. Our goal here was to address the mechanisms allowing for the progressive withdrawal of LF from an LF plus CsA combination therapy in rats with surviving hamster heart Xgs.

First, the capacity of rats carrying long surviving Xgs to reject a second hamster heart graft was investigated. A 2-wk course of LF failed to prevent IgM XAb formation after second xenografting, and these XAbs bound to and provoked rejection of these second Xgs. In contrast, the primary grafts survived in the presence of these XAbs and functional complement. This fits with the definition of Xg accommodation. Accommodation was further confirmed by the experiment involving transfer of IgM XAbs. IgM XAb (0.5 ml) containing serum provoked hyperacute Xg rejection in naive rats, but did not affect the survival of the accommodated grafts. The phenomenon of accommodation was first described in recipients of Red Blood cell group ABO-incompatible kidney allografts by Alexandre and co-workers (8, 10, 11). It was also observed in some cases following transplantation of organs into patients with circulating anti-donor HLA Abs (19). Graft accommodation was demonstrated after xenotransplantation (10, 20, 21). In experiments using plasmapheresis and immunosuppression to prolong the survival of swine-to-baboon renal Xgs, one recipient had a graft function for >20 days despite the presence of circulating anti-graft Abs (10). Accommodation may have an important implication for the eventual clinical application of xenotransplantation, since it would allow progressive withdrawal of immunosuppressive regimens aimed at depleting and/or suppressing XAbs and/or complement from Xg recipients.

Until recently, the mechanisms underlying accommodation were unclear, but several possible explanations have been forwarded. One mechanism may involve a decrease in Ab-Ag interactions. This may be due to a change in the affinity or specificity of the anti-graft Abs or to a change of the expression of the target Ags in the donor organ (22). It was recently reported that accommodation of porcine endothelial cells was induced in an in vitro cell culture system by incubation with low concentrations of polyclonal human IgG Abs (23). The induced resistance of the accommodated porcine endothelial cells to complement-mediated lysis was correlated with a down-regulation of expression of an vascular cell adhesion molecule (23). A second mechanism underlying accommodation may be that with continued stimulation of endothelial cells by low levels of Abs and/or complement, the sensitivity of these cells to injury is decreased (22). Such a mechanism has been demonstrated in vitro, as continued stimulation of endothelial cells with endotoxin or IL-1 caused the cells to become resistant to restimulation (24). A third mechanism of accommodation may be one involving certain protective cellular and molecular mechanisms within the grafts, as suggested by the recent work of Bach et al. (12). It was shown that in a model also involving CsA-treated rats but in which the complement system was temporarily blocked by cobra venom factor, surviving hamster heart grafts became resistant to humoral rejection despite deposition of IgM xenoantibodies and complement activation within the graft (12). Interestingly, the endothelial cells of these accommodated Xgs showed an up-regulation of anti-apoptotic genes, and the grafts were infiltrated with Th2-type lymphocytes (12).

The data presented in the present study provide in vivo evidence supporting the regulation of xenoantigen expression as a potential mechanism for accommodation. We clearly showed that the accommodated hamster hearts exhibited a strongly decreased expression or disappearance of several xenoantigens on the endothelial cells. This was supported by two points. First, anti-hamster IgM, whether provoked (by second hamster Xgs) or adoptively transferred, failed to bind to or to cause rejection of the accommodated hamster grafts. Second, by detecting adhesion molecules that are believed to be important expressers of xenogeneic epitopes (22, 25), we found that the endothelial expression of both P-selectin and E-selectin was dramatically down-regulated in the accommodated Xgs. Interestingly, this decrease in xenoantigen expression was reversible when the accommodated grafts were retransplanted to a naive recipient. Indeed, after retransplantation, the graft endothelial cells showed a progressive re-expression of both P- and E-selectins. This process correlated well with IgM XAb formation, IgM deposition, and graft rejection. These findings are consistent with a recent report showing that antisense-mediated down-regulation of GpIIIa, another adhesion molecule of potential relevance in the human-to-pig Xg recognition, results in a significant decrease in the binding of IgM natural Abs (26). As for the mechanisms leading to the reappearance of the xenoantigens, as seen in our study, they may be related to the process of harvesting, cold storage, and reperfussion of the Xgs during retransplantation (27, 28). This increased expression of xenoantigens together with an increased vulnerability to Ab-mediated injury probably explain the rejection after retransplantation of tolerated Xgs.

When comparing the characteristics of accommodation achieved in the present study with those reported by Bach et al., the mechanisms leading to accommodation may be relevant. When accommodation was obtained by blocking complement, while XAb formation was not suppressed, accommodation apparently depended on active mechanisms, such as Th2 lymphocytes and an up-regulation of protective genes in endothelial cells (12). In contrast, when accommodation was achieved by blocking XAb formation, while leaving the complement system intact as in our study, accommodation was based on a passive mechanism involving down-regulation of xenoantigens.

In the 4-wk LF group, CsA-treated rats had lost their capacity to produce IgM anti-hamster Abs even after withdrawal of LF and transplantation of a second Xg. This was not due to the fact that in this group LF was withdrawn only 2 wk before performing the second transplantation. Indeed, we have successfully performed second transplantations without rejection in 4-wk LF-treated rats 2 and 3 mo after the first transplantation (data not shown). The absence of circulating anti-donor XAbs after second xenografting was also not due to an adsorption by the tolerated first Xg, since 1) deposition of XAbs was detected in neither the first nor the second Xgs; and 2) removal of the first Xgs 3 days before the second xenotransplantation did not result in an appearance of XAbs in the circulation (data not shown), and the second Xgs were accepted as well as in rats with the first Xgs in place. Hence, B cell unresponsiveness was induced. This unresponsiveness was limited to the T-independent (CsA-insensitive) XAb formation. Indeed, the capacity to make T-dependent XAbs remained, as withdrawal of also CsA resulted in a prompt formation of mainly IgG XAbs. In rodents as well as in other mammalian species, B cells are continuously produced in the bone marrow during postnatal life. Due to their limited lifespan, peripheral immunocompetent B cells are continuously replaced by newly formed B cells emerging from the bone marrow (29, 30). In rats, the majority of the peripheral immunocompetent B cell pool turns over in approximately 4 wk (31). Several studies have demonstrated that Ag recognition by surface IgM in immature B cells induces clonal anergy rather than activation (32, 33). A first element to explain the T-independent B cell tolerance occurring after 4 wk of LF could thus be that this period of time is needed to block the activation of peripheral immunocompetent B cells. As this pool can be expected to be renewed after 4 wk, thereafter the newly formed B cells may be tolerized by the xenoantigens. However, maintenance of the latter tolerance by these xenoantigens remains unclear in the present study. Indeed, we showed that the xenoantigens that were the mediators and also the targets of the IgM XAbs were not detectable in the long surviving Xgs. This paradox may be explained by a difference in the relative concentration needed to tolerize compared with that needed to stimulate immature B cells to form XAbs.

Another explanation for the kinetics of the B cell tolerance may be that in the hamster-to-rat model, IgM XAbs are produced by a specific subpopulation of B lymphocytes with a limited self-renewing capacity. Once tolerized by LF, this subpopulation may be unable to repopulate even after withdrawal of treatment. CD5 B lymphocytes may constitute such a subpopulation. It is generally known that CD5 B cells are a primitive B cell subpopulation with very limited self-renewing capacity (33). It has also been suggested that CD5+ B cells are important for the production of T-independent IgM Abs (34, 35). In addition, we recently found that CD5+ B cells were the major source of T-independent IgM XAbs in rats. This was based on the experiments in which SCID mice were reconstituted with a highly purified CD5+ or CD5 rat B cell population and subsequently challenged with hamster cells; only the CD5+ B cell-reconstituted SCID produced anti-hamster IgM XAbs (Y. Lin and M. Waer, manuscript in preparation).

In conclusion, we suggest that depending on the duration of immunosuppression with LF, the mechanism of long term survival of hamster hearts in rats treated with CsA is based in part on graft accommodation and in part on T-independent B cell tolerance. Transient inhibition of XAb formation in the early period after xenografting allowed for the development of accommodation in Xgs. This accommodation was based on a decreased expression of xenoantigens. In a later phase, a degree of tolerance developed for T-independent B cells. Both, accommodation and T-independent B cell tolerance may allow for the progressive withdrawal of immunosuppressants after xenotransplantation.

1

This work was supported by a grant of the ONDERZOEKSRAAD from the University of Leuven and by the National Fund for Scientific Research.

3

Abbreviations used in this paper: Xg, xenograft; XAb, xenoantibody; CsA, cyclosporin; LF, leflunomide; MST, mean survival time.

1
Leventhal, J. R., A. P. Dalmasso, J. M. Cromwell, J. L. Platt, C. J. Manivel, R. M. Bolman III, A. J. Matas.
1993
. Prolongation of cardiac xenograft survival by depletion of complement.
Transplantation
55
:
857
2
Hancock, W. W., M. L. Blakely, W. Van der Werf, F. H. Bach.
1993
. Rejection of guinea pig cardiac xenografts post-cobra venom factor therapy is associated with infiltration by mononuclear cells secreting interferon-γ and diffuse endothelial activation.
Transplant. Proc.
25
:
293
3
Hayashi, S., M. Ito, I. Yokoyama, H. Takagi.
1994
. Evidence that combination therapy using cobra venom factor, splenectomy, and deoxyspergualin is effective in guinea pig-to-rat cardiac xenotransplantation.
Transplantation
57
:
777
4
Bartholomew, A. M., A. B. Cosimi, D. H. Sachs, M. Bailin, S. Boskovic, R. Colvin, H. Hong, M. Johnson, M. Kimikawa, A. Leguern, S. Meehan, T. Sablinski, S. L. Wee, J. Powelson.
1997
. A study of tolerance in a concordant xenograft model.
Transplant. Proc.
29
:
923
5
Roslin, M. S., R. B. Tranbaugh, A. Panza, S. C. Matthew, Y. D. Kim, T. Chang, J. N. Cunningham, A. J. Norin.
1992
. One-year monkey heart xenograft survival in cyclosporine-treated baboons.
Transplantation
54
:
949
6
Van den Bogaerde, J., R. Aspinall, M. W. Wang, N. Cary, S. Lim, L. Wright, D. White.
1991
. Induction of long-term survival of hamster heart xenografts in rats.
Transplantation
52
:
15
7
Platt, J. L., G. M. Vercellotti, A. P. Dalmasso, A. J. Matas, R. M. Bolman, J. S. Najarian, F. H. Bach.
1990
. Transplantation of discordant xenografts: a review of progress.
Immunol. Today
11
:
450
8
Alexandre, G. P. J., J. P. Squifflet, M. De Bruyere, D. Latinne, R. Reding, P. Gianello, M. Carlier, V. Pirson.
1987
. Present experience in a series of 26 ABO-incompatible living donor renal allografts.
Transplant. Proc.
19
:
4538
9
Slapak, M., N. Digard, M. Ahmed, T. Shell, F. Thompson.
1990
. Renal transplantation across the ABO barrier -A 9-year experience.
Transplant. Proc.
22
:
1425
10
Alexandre, G. P., P. Gianello, D. Latinne, J. P. Squifflet, M. Lievens, L. Lambotte.
1991
. Plasmapheresis and spleenectomy in experimental renal xenotransplantation. M. A. Hardy, ed.
Xenograft 25
259
Elsevier Science Publishers, New York.
11
Alexandre, G. P., D. Latinne, P. Gianello, J. P. Squifflet, M. Lievens, J. P. Tomasi, L. Lambotte.
1991
. Preformed cytotoxic antibodies and ABO-incompatible grafts.
Clin. Transplant.
5
:
583
12
Bach, F. H., C. Ferran, P. Hechenleitner, W. Mark, N. Koyamada, T. Miyatake, H. Winkler, A. Badrichani, D. Candinas, W. W. Hancock.
1997
. Accommodation of vascularized xenografts: expression of “protective genes” by donor endothelial cells in a host Th2 cytokine environment.
Nat. Med.
3
:
1
13
Zhao, Y., K. Swenson, J. J. Sergio, J. S. Arn, D. H. Sachs, M. Sykes.
1996
. Skin graft tolerance across a discordant xenogeneic barrier.
Nat. Med.
2
:
1211
14
Fryer, J. P., J. R. Leventhal, A. P. Dalmasso, S. Chen, P. A. Simone, J. J. Goswitz, N. L. Reinsmoen, A. J. Matas.
1995
. Beyond hyperacute rejection.
Transplantation
59
:
171
15
Lin, Y., M. Vandeputte, M. Waer.
1997
. NK cell- and macrophage-mediated rejection of concordant xenograft rejection in the absence of T- and B-cell responses.
J. Immunol.
158
:
5658
16
Lin, Y., M. Vandeputte, M. Waer.
1995
. Mechanism of leflunomide-induced prevention of xenoantibody formation and xenograft rejection in the hamster to rat heart transplantation model.
Transplant. Proc.
27
:
305
17
Xiao, F., A. S. Chong, P. Foster, H. Sankary, L. McChesney, G. Koukoulis, J. Yang, D. Frieders, J. W. Williams.
1994
. Leflunomide controls rejection in hamster to rat cardiac xenografts.
Transplantation
58
:
828
18
Baker, P. J., R. F. Ochi, M. Schulze, R. J. Johnson, C. Campbell, W. G. Couser.
1989
. Depletion of C6 prevents development of proteinuria in experimental membranous nephropathy in rats.
Am. J. Pathol.
135
:
185
19
Palmer, A., K. Welsh, P. Gjorstrup, D. Taube, W. Bewick, M. Think.
1989
. Removal of anti-HLA antibodies by extracorporeal immunoadsorption to enable renal transplantation.
Lancet
1
:
10
20
Fischel, R. J., A. J. Matas, J. L. Platt, E. Perry, H. Noreen, S. J. Shumway, R. M. Bolman.
1992
. Cardiac xenografting in the pig-to-rhesus monkey model; manipulation of anti-endothelial antibodies prolongs survival.
J. Heart Lung Transplant.
11
:
965
21
Tomas, F., D. Araneda, P. Quarantillo.
1994
. Long-term survival of disparate cardiac xenografts: induction of tolerance and humoral accommodation.
Transplant. Proc.
26
:
1220
22
Platt, J. L..
1994
. A perspective on xenograft rejection and accommodation.
Immunol. Rev.
141
:
127
23
Dorling, A., C. Stocker, T. Tsao, D. O. Haskard, R. I. Lechler.
1996
. In vitro accommodation of immortalized porcine endothelial cells: resistance to complement mediated lysis and down-regulation of VCAM expression induced by low concentrations of polyclonal human IgG antipig antibodies.
Transplantation
62
:
1127
24
Busso, N., S. Huet, E. Nicodeme, J. Hiernaux, F. Hyafil.
1991
. Refractory period phenomenon in the induction of tissue factor expression on endothelial cells.
Blood
78
:
2027
25
Rosa, J. P., P. F. Bray, O. Gayet.
1988
. Cloning of glycoprotein IIIa cDNA from human erythroleukemia cells and localization of the gene to chromosome 17.
Blood
72
:
593
26
Kearns-Jonker, M. K., D. V. Cramer, L. A. Dane, J. M. Swensson, L. Makowka.
1997
. Human serum reactivity to porcine endothelial cells after antisense-mediated down-regulation of GpIIa expression.
Transplantation
63
:
588
27
Nishimura, Y., Y. Takei, S. Kawano, M. Goto, K. Nagano, S. Tsuji, H. Nagai, A. Ohmae, H. Fusamoto, T. Kamada.
1996
. The F(ab′)2 fragment of an anti-ICAM-1 monoclonal antibody attenuates liver injury after orthotopic liver transplantation.
Transplantation
61
:
99
28
Porcheron, J., J. Margonari, F. Gorry, O. Boillot.
1995
. Activation of the endothelial cell in rat liver during preservation by cold storage.
Transplant. Proc.
27
:
1658
29
Osmond, D. G., G. J. V. Nossal.
1974
. Differentiation of lymphocytes in mouse bound marrow. II. Kinetics of maturation and renewal of antiglobulin binding cells studied by double labeling.
Cell. Immunol.
13
:
132
30
Osmond, D. G..
1993
. The turnover of B cell population.
Immunol. Today
14
:
34
31
Chan, E. Y., I. C. MacLennan.
1993
. Only a small proportion of splenic B cells in adults are short-lived virgin cells.
Eur. J. Immunol.
23
:
357
32
Goodnow, C. C..
1992
. Transgebic mice and analysis of B cell tolerance.
Annu. Rev. Immunol.
10
:
489
33
Latinne, D., M. Soares, X. Havaux, F. Cormont, B. Lesnikoski, F. H. Bach, H. Bazin.
1994
. Depletion of IgM xenoreactive natural antibodies by injection of anti-μ monoclonal antibodies.
Immunol. Rev.
141
:
95
34
Hayakawa, K., R. R. Hardy.
1992
. Normal, autoimmune and malignant CD5+ B cells: the Ly-1 lineage?.
Annu. Rev. Immunol.
6
:
197
35
Nawata, Y., A. M. Stall, L. A. Herzenberg, E. M. Euqui, A. C. Allison.
1990
. Surface immunoglobulin ligands and cytokines differentially affect proliferation and antibody formation by human CD5+ and CD5 B lymphocytes.
Int. Immunol.
2
:
603