INDUCTION OF ANTIGRAFT AND ANTIRECIPIENT ANTIBODY RESPONSES AFTER FULLY ALLOGENEIC AND SEMIALLOGENEIC RAT SMALL BOWEL TRANSPLANTATION

Abstract

Background. 

Given the potential influence of alloantibodies on organ graft outcome, this study investigated the induction of antigraft and antirecipient antibodies after allogeneic and semiallogeneic rat small bowel transplantation.

Methods. 

Fully allogeneic, unidirectional rejection and unidirectional graft-versus-host disease (GvHD) heterotopic small bowel transplantation was performed using DA, PVG, and (PVGxDA)F₁ donor-recipient combinations. Serum was obtained before and at time points after transplantation and incubated with blood from untransplanted DA and PVG rats. Antibody binding to T cells was detected by whole blood flow cytometry using FITC-conjugated anti-rat IgM murine monoclonal antibody. Antibody levels were determined by reference to a standard curve of fluorescent intensity generated using a serum sample with known anti-target cell IgM activity. Data are presented as arbitrary units/ml (AU/ml).

Results. 

In the PVG→DA combination, five of six DA recipients had detectable anti-graft (PVG) antibodies by day 4 after transplantation (mean 72 AU/ml) and all animals were positive by day 6 (976 AU/ml). Antirecipient (DA) antibodies were also induced, however, they were only apparent after 6 days in five of eight animals (90 AU/ml). Antigraft (DA) antibody responses were also induced in the DA→PVG combination (day 6–218 AU/ml), however no antirecipient (PVG) response was apparent. Transplantation induced antirecipient (DA) antibodies in the unidirectional GvHD model (day 6–90 AU/ml) and an anti-graft (PVG) response in the unidirectional rejection model (day 6–60 AU/ml). However, the latter was quantitatively lower than that generated in the PVG→DA combination (day 6–976 AU/ml).

Conclusions. 

Antigraft and antirecipient antibody responses are simultaneously induced after fully allogeneic small bowel transplantation, despite rejection being the predominant clinical feature. Further studies are required to elucidate their influence on graft outcome.

It has been known since the 1960s that the presence of preexisting donor-specific cytotoxic antibodies can lead to hyperacute rejection and graft loss (¹). Subsequent work demonstrated that donor-specific antibodies can also be generated after transplantation and that these can induce rejection and graft loss, by mechanisms that appear to involve endothelial cell binding, complement activation and the sequestration of neutrophils and platelets (²).

The development of de novo donor-specific antibodies has been associated with a poor graft prognosis (^{3, 4}). Five to 10% of renal allograft recipients experience alloantibody-mediated acute rejection (⁵) and the 1-year graft survival rate after an episode of alloantibody-mediated acute rejection despite aggressive immunosuppressive therapy is 15 to 50% (^{5, 6}).

In their pioneering work, Monchik and Russell (⁷) demonstrated that serum from acutely rejecting rat small bowel transplant recipients demonstrated cytotoxic activity against donor strain splenocytes. However, little work has since been done on the possible influence of antibody on the outcome of small bowel transplantation. The generation of antibodies de novo in the posttransplant period and their potential influence on graft outcome is particularly relevant to small bowel transplantation as, unlike the majority of solid-organ grafts, the lymphoid tissue of the transplanted small bowel has the capacity to generate recipient-specific immune responses. Although the generation of recipient-reactive antibodies after small bowel transplantation in a unidirectional GvHD model has been demonstrated (⁸), antibody generation in the more clinically relevant fully allogeneic situation, in which graft rejection predominates, has not yet been evaluated.

This study used flow cytometry to monitor the generation of donor- and recipient-specific antibody responses in fully allogeneic, unidirectional graft-versus-host disease (GvHD) and uni-directional rejection small bowel transplant models.

MATERIALS AND METHODS

Small bowel transplantation.

Heterotopic small bowel transplantation was performed using adult male PVG (RT1^c), DA (RT1^a), and (PVGxDA)F₁ rats (B & K Universal Ltd, Hull, UK), essentially as previously described (^{7, 9, 10}). Animals were housed under standard conditions and were given normal rat food and water ad libitum. Treatment was in compliance with the Animals (Scientific Procedures) Act of 1986.

The small bowel from immediately distal to the duodeno-jejunal junction to immediately proximal to the ileo-cecal junction was isolated on a vascular pedicle comprising the portal vein and superior mesenteric artery on an aortic cuff. The vasculature was flushed with Marshall’s perfusion solution and the bowel lumen cleared with chlorhexidine solution (Baxter Healthcare Ltd., Thetford, UK). The graft was stored in Marshall’s solution at 4°C until transplanted. End-to-side aorto-aortic and porto-caval anastomoses were performed in the recipient and the small bowel graft was transplanted in a heterotopic position with cutaneous stomata at both luminal ends. Animals were monitored regularly and sacrificed if necessary. The survival time was recorded as the number of days from the transplant to the day of sacrifice, or to the day before death if the animal died overnight. The following strain combinations were used: 1) PVG donors into DA recipients (fully allogeneic; n=8); 2) DA donors into PVG recipients (fully allogeneic; n=4); 3) PVG donors into (PVG x DA)F₁ recipients (unidirectional GvHD; n=5); 4) (PVG x DA)F₁ donors into DA recipients (unidirectional rejection; n=5); 5) DA donors into DA recipients (isograft controls; n=5)

Detection of circulating allo-antibodies by flow cytometry.

Peripheral blood was obtained at the times indicated and the serum was stored at −20°C until analysis for circulating strain-specific IgM antibody levels by whole blood flow cytometry.

For the flow cytometric assay, 10 μl heparinized blood from untransplanted DA and PVG rats were incubated with 20 μl standardized pooled mouse serum (Serotec Limited, Oxford, UK) for 15 min at room temperature. Cells were washed by centrifugation and the cell pellet was resuspended in the residual volume. Cells were incubated for 15 min with 3 μl of test serum or a standard serum having a known high target cell activity (see below), after which they were washed. The binding of strain-specific IgM to the cell surface was detected by adding 3 μl FITC-conjugated anti-rat-IgM-specific murine monoclonal antibody (mAb; Serotec). After washing, cells were incubated for 15 min with 20 μl pooled mouse serum, washed, and B cells were stained using a PE-conjugated murine mAb reactive with rat B cells (OX33; Serotec). This enabled the exclusion of B cells expressing surface immunoglobulin from the subsequent fluorescent intensity analysis. Erythrocytes were lysed with Erythrolyse (Serotec) and the cells washed before flow cytometric analysis.

Samples were analyzed on a FACScan flow cytometer (BD Immunocytometry Systems, Oxford, UK). The forward and side scatter and fluorescence data for 5000 cells within a live gate placed around the lymphocyte and monocyte regions were acquired using “Consort 30” software (BD Immunocytometry Systems). An analysis gate was placed around the OX33 negative cell population (T cells) within the lymphocyte region and the mode fluorescent intensity of these cells was recorded. The fluorescent intensity of T cells reflected the levels of antitarget cell IgM in the serum sample. Negative controls comprised cells stained with FITC-conjugated anti-rat IgM mAb alone. Sera resulting in a mode fluorescent intensity of 5 or more were considered to be positive.

Serum samples were prescreened and those having high levels of the appropriate antitarget cell IgM antibody were identified. Standard fluorescent intensity curves were generated by incubating peripheral blood with serial dilutions of the standard sera. Antibody levels in a particular sample were determined using ASSAYZAP data analysis software (BIOSOFT, Cambridge, UK). Standard curves were generated on each occasion that serum samples were analyzed.

Statistical analysis.

Data are presented as means±SEM. Statistical differences in antibody levels between pre- and post-transplant time-points were determined using the one-sample t test, as antibody levels in pretransplant sera were determined using serum pooled from the respective animal groups. Antibody levels in the experimental groups were compared using the independent-samples t test. P <0.05 were considered to indicate statistical significance. Statistical analyses were performed using SPSS 10.0.5 for Windows (SPSS UK Limited, Woking, UK).

RESULTS

Recipient survival.

In this study, all animals in the PVG→DA group were killed on day 6. Previous work from our laboratory has demonstrated that the median survival of small bowel allograft recipients in this strain combination is 7 days (^{9, 10}). The median survival in the DA→PVG group was 8 days (^{8, 11}). All animals in the fully allogeneic transplantation combinations displayed macroscopic evidence of rejection on postmortem examination. DA→DA isograft controls survived indefinitely.

Recipient survival after transplantation in the PVG→(PVG x DA)F₁ unidirectional GvHD group was 10, 12, 12, 13, and 13 days, with death resulting from overt signs of GvHD (hunched posture, redness of paws and ears). Survival after small bowel transplantation in the (PVG x DA)F₁→DA unidirectional rejection group was 13, 16, >27, >96, and >97 days; the graft became an encapsulated mass in the long-term survivors.

Recipient- and graft-specific antibody after fully allogeneic transplantation.

No anti-PVG or anti-DA IgM antibody was detected in the serum of any rats before transplantation (data not shown). In the PVG→DA group, an anti-graft (PVG) antibody response was detectable in five of six animals assayed 4 days after transplantation (72±29 AU/ml), whereas no anti-recipient (DA) antibody was detected at that time (Fig. 1). By day 6, all DA recipients of PVG grafts had detectable levels of anti-graft (PVG) IgM antibodies (976±273 AU/ml) and five of eight animals also developed antirecipient (DA) IgM antibodies (90±33 AU/ml;Fig. 1). It is not possible to directly compare levels of anti-DA antibody with levels of anti-PVG antibody as the units were defined by necessarily different arbitrary standards. There was no alloantibody response at any time point after transplantation in the DA→DA isograft group.

Figure 1: Antigraft (PVG) and antirecipient (DA) IgM antibody responses after fully allogeneic small bowel transplantation in the PVG→DA strain combination. Data are expressed as AU/ml and are presented as individual (diamonds) and mean (circles) levels. No antigraft or antirecipient antibodies were detected in pretransplant sera. Significant differences from pretransplant levels are indicated (one-sample t test).

Transplantation also induced a progressive, albeit variable appearance of antigraft antibodies in PVG recipients of DA grafts (Fig. 2). In contrast to the PVG→DA strain combination, the antirecipient response was limited and anti-PVG antibody was only detected in a single animal 8 days after transplantation (152 AU/ml). These findings clearly demonstrate a detectable, albeit slightly delayed induction of an antirecipient (GvHD) antibody response in the fully allogeneic transplantation combination, despite the fact that rejection is the predominant clinical feature.

Figure 2: Antigraft (DA) IgM antibody responses after fully allogeneic small bowel transplantation in the DA→PVG strain combination. Except for one animal on day 8 after transplantation (152 AU/ml), no antirecipient (PVG) IgM antibodies were detected. Data are expressed as arbitrary units/ml and are presented as individual (diamonds) and mean (circles) levels. No antigraft or antirecipient antibodies were detected in pretransplant sera. Significant differences from pretransplant levels are indicated (one-sample t test; n.s, nonsignificant).

Recipient- and graft-specific antibody in unidirectional GvHD.

Antirecipient (DA) antibody was first detected in the serum 6 days after transplantation in the unidirectional GvHD model, and levels continued to rise until all the animals had succumbed to acute GvHD at around day 12 (Fig. 3). The levels of antirecipient (DA) antibody generated in the first 6 days after transplantation in the unidirectional GvHD model were the same as those generated in the fully allogeneic combination (90±41 and 90±33 AU/ml respectively;Figs. 1 and 3). No anti-PVG (antigraft) IgM was found at any time point after transplantation in this strain combination (data not shown).

Recipient- and graft-specific antibody in unidirectional rejection.

Antigraft (PVG) antibody was first detectable on day 6 after transplantation in the (DAxPVG)F₁→DA (unidirectional rejection) model, although this was not of statistical significance (60±37 AU/ml;Fig. 4). Levels were significantly higher than pretransplantation after 10 days (78±16 AU/ml), however, the variability in data and reduced number of animals at later time points rendered any subsequent differences from pretransplant levels to be of no statistical significance (Fig. 4). At no time point after transplantation were anti-DA (antirecipient) antibodies detected (data not shown). The antigraft (PVG) antibody response in the unidirectional rejection strain combination on day 6 after transplantation was markedly lower than the antigraft (PVG) response induced in the fully allogeneic (PVG→DA) strain combination at the same time point (60±37 vs. 976±273 AU/ml, P= 0.012;Figs. 1 and 4).

Figure 3: Antirecipient (DA) IgM antibody responses after semiallogeneic transplantation in the PVG→(PVG x DA)F₁ (unidirectional GvHD) strain combination. No antigraft (PVG) IgM antibodies were detected in any animal at any time point. Data are expressed as AU/ml and are presented as individual (diamonds) and mean (circles) levels. No antigraft or antirecipient antibodies were detected in pretransplant sera. Significant differences from pretransplant levels are indicated (one-sample t test; n.s, nonsignificant).

Figure 4: Anti-graft (PVG) IgM antibody responses after semiallogeneic transplantation in the (PVG x DA)F₁→DA (unidirectional rejection) strain combination. No antirecipient (DA) IgM antibodies were detected in any animal at any time point. Data are expressed as AU/ml and are presented as individual (diamonds) and mean (circles) levels. No antigraft or antirecipient antibodies were detected in pretransplant sera. Significant differences from pretransplant levels are indicated (one-sample t test; n.s, nonsignificant).

DISCUSSION

This study has demonstrated that antirecipient and antigraft IgM antibodies are simultaneously generated after fully allogeneic rat small bowel transplantation. We have also demonstrated that antirecipient and antigraft antibodies are induced in unidirectional GvHD and rejection models, respectively. These findings extend the report of antirecipient antibody induction after small bowel transplantation in LEW→(LEWxBN)F₁ (unidirectional GvHD) rat strain combination (⁸).

The pathogenic role of donor-specific antibodies in the clinical situation is somewhat uncertain. Some investigators have suggested that they are deleterious (^2–5), whereas others have suggested that they are an epiphenomenon resulting from the inflammatory process of the acute rejection response (^{11, 12}). However, despite the debate surrounding the contribution of alloantibodies to graft rejection, humoral responses can directly influence graft outcome, as the transfer of donor-specific immune serum to B cell-deficient mice accelerates cardiac graft rejection (¹³) and induces the rejection of rat renal, cardiac, and skin grafts (^{14, 15}).

Although the large lymphoid component of the small bowel graft led to the suggestion that GvHD may present a significant problem after small bowel transplantation, save for a few isolated instances, GvHD has not been a common clinical occurrence and rejection remains the predominant feature (¹⁶). However, the immunological competence of small bowel allografts remains intact and capable of generating antirecipient responses. GvHD has been reported after a clinical case of combined liver/small bowel transplantation (¹⁷) and an antirecipient blood group (anti-A) antibody response has been reported after clinical small bowel transplantation from a blood group O donor into a group A recipient (¹⁸). This study has demonstrated that PVG grafts are capable of generating antirecipient (DA) antibody responses in the fully allogeneic PVG→DA strain combination model, despite the overwhelming clinical manifestations of rejection.

To date, attention has predominantly focused on the potential effects of IgG alloantibodies on late allograft loss. Given that IgM antibody is present in recipient liver and deposited in native intestine in a unidirectional GvHD small bowel transplant model (⁸) and that intestinal ischemia-reperfusion injury is dependent on the classic complement pathway and IgM (¹⁹), it may be that IgM alloantibody influences the outcome of small bowel transplantation. Our previous studies demonstrating a localized antirecipient response in transplanted rat small bowel allografts support the potential involvement of GvHD responses after fully allogeneic small bowel transplantation (²⁰).

Small bowel transplantation is followed by extensive cellular exchanges between graft and host tissues (^21–24). However, the rate and extent of cellular migration after transplantation using PVG and DA rats is dependent on the donor-recipient combination used (^{25, 26}). Transplantation of PVG grafts into DA recipients is associated with a greater degree of cellular migration, suggesting a stronger potential for GvH reactivity in this combination (²⁶). This strain-dependent potential for GvH reactivity after fully allogeneic small bowel transplantation appears to be confirmed by the observation that PVG grafts can mount an antirecipient (DA) antibody response, whereas DA grafts do not appear capable of generating antirecipient (PVG) responses. This may result from the much lower and more transient emigration of DA leukocytes into PVG recipients in the DA→PVG strain combination (²⁵).

The implications of these findings are that the appearance of circulating recipient-specific antibody is to some extent dependent on the emigration of graft lymphoid cells into the recipient. It is likely that these cells have been activated within the graft, as the MHC haplotype mismatch in the fully allogeneic combination does not allow the co-operation of graft B cells with recipient T cells.

It is interesting to note that the capacity of PVG and (PVGxDA)F₁ grafts to elicit anti-DA (recipient) antibodies was equivalent, whereas the response of DA recipients to PVG grafts was far greater than the response to (PVGxDA)F₁ grafts. The reasons for the differing responses are currently uncertain. The shared MHC haplotype in the semi-allogeneic combination allows co-operation between recipient B cells and graft T cells (²⁷). It may be that the reduced intensity of PVG antigens in the (PVGxDA)F₁ grafts is involved or that the exchange migration of cells is more limited than that present in the fully allogeneic combination. An alternative possibility is that the systemic inflammatory response and the simultaneous generation of anti-graft and anti-recipient antibodies in the fully allogeneic PVG→DA combination further fuels the generation of anti-graft antibody responses.

The majority of studies into the immunological consequences of small bowel transplantation have focused on the cell-mediated components of acute rejection and GvHD. To our knowledge this is the first study to demonstrate the presence of circulating antigraft and antirecipient antibodies after fully allogeneic small bowel transplantation in which rejection is the predominant clinical feature. Detection of circulating antibody in all forms of organ transplantation has always been problematic, primarily due to the capacity of the graft to bind recipient antibody. This is not a benign process and the endothelial reaction undoubtedly contributes to impairment of vascularity and the development of chronic rejection changes in the long-term.

Clearly, there is a need for more research into the role of antibody responses in intestinal transplantation particularly with regard to their influence on acute and chronic rejection. This study used a flow cytometric approach for identifying these antibodies and their cytotoxic capacity has yet to be determined. However, irrespective of any active role in these pathological events, the presence of such antibodies in peripheral blood may provide useful surrogate markers of cellular immune responses in clinical practice.