Jonathan P. Fryer

Removal of baboon and human antiporcine IgG and IgM natural antibodies by immunoadsorption : results of in vitro and in vivo studies

The safe and effective removal of xenoreactive antibodies in the peritransplant period is likely to be critical for the clinical application of xenotransplantation involving disparate donor species, such as the pig. In an effort to develop an improved method for antibody removal in xenotransplantation, we have studied reusable antihuman antibody (Ig) columns in vitro and in vivo. Two types of columns were tested: (1) an antihuman Ig column containing polyclonal sheep antihuman IgG (heavy- and light-chain-specific) conjugated to sepharose CL-4B (Ig-Therasorb), and (2) an antihuman Ig column using polyclonal antihuman IgM (mu-chain-specific) conjugated to sepharose. Passage of human or baboon plasma through the Ig-Therasorb column resulted in 97.5% and 78.4% mean reductions in total IgG and IgM, respectively. Reductions in total IgG and IgM correlated with lowering of antipig IgG (54-486 fold) and IgM (9-54 fold) antibody titers as assessed by pig endothelial cell ELISA. The ability of the Ig-Therasorb to significantly reduce IgM may be attributed to the light chain specificity of this column. With the anti-IgM column, marked reductions in total (82.6-83.9%) and antipig (27-54 fold) IgM in human and baboon plasma occurred, while levels of total and xenoreactive IgG were slightly affected. Other than a dilutional effect, neither column resulted in significant reduction in albumin, fibrinogen, factor 5, and factor 8. Repeated in vivo use of either column in baboons achieved reductions in IgG and IgM that closely followed the results of our in vitro studies. No subject morbidity or mortality occurred. Use of the Ig-Therasorb column with immunosuppression in two baboons receiving pig renal xenografts achieved sustained reductions in antipig antibodies and prevented hyperacute rejection. Subjects were sacrificed at 11 and 13 days posttransplant with functioning xenografts and were found to have no evidence of vascular xenograft rejection. We conclude that anti-Ig columns represent a safe and effective method for antibody removal, without several of the limitations of other antibody removal techniques. Also, columns appear to be safe for repeated antibody removal in the posttransplant period.

Beyond hyperacute rejection : accelerated rejection in a discordant xenograft model by adoptive transfer of specific cell subsets

If hyperacute rejection is prevented in the guineapig (GP)-to-Lewis rat (Lew) cardiac xenograft (CXg) model, an accelerated rejection involving cellular infiltration occurs in 3 to 4 days. In previous work using an adoptive transfer model, we found that this accelerated rejection was facilitated by either sensitized splenocytes or sensitized serum. In the current study, in an attempt to determine which splenocyte subset(s) facilitated this process, sensitized splenocytes, with or without subset depletion were injected, into complement- and natural antibody-depleted Lew recipients of GP CXgs. Graft survival was 4.18 +/- 0.75 days with no injection (n = 11), 4.13 +/- 0.99 days with naive splenocytes (n = 8), 1.80 +/- 0.45 days with sensitized splenocytes (n = 5), 2.67 +/- 1.03 days with CD4(W3/25+) depletion of the sensitized splenocytes (n = 6), 3.13 +/- 0.84 days with CD8(OX8+) cell depletion (n = 8), 4.70 +/- 0.68 days with macrophage depletion (n = 10), and 4.22 +/- 0.41 days with B cell depletion (n = 9). Cellular infiltrates, hemorrhage, myocyte necrosis, and endothelial deposition of IgG, IgM, and fibrin were seen in rejected grafts. In most groups, infiltrating cells consisted of CD4 (W3/25+), CD8 (OX8+), IL2R+ cells, macrophages, and natural killer (NK) cells. However, in the macrophages-depleted group, activated (ED2+) macrophages and NK cells were significantly reduced. Total IgM, anti-GP IgM, and anti-GP IgG rebounded in all groups over several days but were not consistent at the time of rejection. Lewis rats rejecting GP CXgs early had lower final titers than those rejecting later. Total IgG titers rebounded to baseline by posttransplant day 1 and were therefore similar in all groups at the time of rejection. These findings suggest that this accelerated rejection requires interaction between macrophages and B cells, since depletion of either significantly alters the rejection tempo. A possible explanation is that xenoreactive IgG antibodies, synthesized by sensitized B cells, bind their target antigens--but also bind sensitized macrophages through their Fc region, thus causing rejection by antibody-dependent cell-mediated cytotoxicity.

Cellular rejection in discordant xenografts when hyperacute rejection is prevented: analysis using adoptive and passive transfer

Hyperacute rejection of discordant xenografts occurs rapidly, precluding cellular infiltration. Thus the role of cellular rejection in discordant xenografts is debated. Using adoptive transfer of sensitized splenocytes and passive transfer of sensitized serum, we evaluated the influence of cellular and humoral elements on cellular infiltration and rejection in the guinea-pig-to-rat discordant xenograft model. Guinea-pig hearts were transplanted into Lewis rats. Pretransplant, rats underwent splenectomy and plasma exchange and were started on daily cobra venom factor injections. Xenografts rejected faster after adoptive (1, 2, 2 and 2 days) or passive (1, 1, 2 and 2 days) transfer than controls (4, 4, 4 and 4 days; p < 0.05). Macrophages and neutrophils were predominant in early prerejection specimens. Over time, cellular infiltrates were dominated by mononuclear cells. Natural killer cells were present in all groups, as were interleukin 2 receptor positive cells. Our data suggest that either sensitized serum or sensitized cells are capable of initiating an accelerated rejection characterized by cellular infiltration. Despite subtle differences, the population of infiltrating cells was similar in each group. Thus, although rejection may be initiated by either cellular or humoral influences, the ultimate result is the same. We have, therefore, established a small animal model to study cellular rejection in discordant xenografts. This model will help evaluate the role of cell subsets and xenoantibodies in xenograft rejection and will help determine the precise relationship between the two when hyperacute rejection is prevented.

A prospective study of FK506 versus CsA and pig ATG in a porcine model of small bowel transplantation

Rejection remains a major barrier to successful bowel transplantation, and immunosuppressive protocols are far from standardized. In 88 nonrelated outbred pigs, we compared the effects of two immunosuppressive regimens--one with FK506, the other with cyclosporine (CsA) and pig antithymocyte globulin (ATG)--on incidence and severity of rejection in the early, critical posttransplant period. Group A (n = 14) was nonimmunosuppressed (controls). Group B (n = 17) received pig ATG (10 mg/kg/day x 10 days), CsA (3 mg/kg/day), prednisolone (2 mg/kg/day), and azathioprine (2.5 mg/kg/day); prednisolone and azathioprine were each reduced by 50% at 8 and 15 days posttransplant. Trough CsA whole-blood concentrations were > or = 400 ng/ml for the first 7 days, > or = 200 ng/ml thereafter. Group C (n = 13) received FK506 (0.2 mg/kg/day) and prednisolone (2 mg/kg/day); prednisolone was reduced by 50% at 8 and 15 days. FK506 whole-blood concentrations were > or = 20 ng/ml. All immunosuppression in groups B and C was given intravenously. We performed orthotopic small bowel transplants with systemic venous drainage. Recipient bowel was resected distal to the second portion of the duodenum and proximal to the rectum at transplant; bowel continuity was restored by duodenojejunostomy; ileostomy was created distally to allow access for daily biopsies. We graded interstitial and vascular rejection separately, according to a scoring system (no, mild, moderate, and severe rejection). Rejection-free graft survivals at 7, 14, and 21 days posttransplant were 38%, 19%, and 0% in group A; 93%, 93%, and 62% in group B; and 100%, 91%, and 82% in group C (P < 0.001). Comparing rejection in the immunosuppressed groups, group C (FK506) had a stronger tendency toward rejection than group B (CsA-ATG); significant differences between groups B and C were, however, noted only on individual days posttransplant, not over time. The death rate due to irreversible rejection was not significantly different in groups B and C (P = 0.8), but was significantly better in both of these immunosuppressed groups than in group A (P < 0.001). Pig survival was significantly longer in group C than in B (P = 0.001) due to a lower infection rate in group C. Posttransplant serum interleukin 2 and 7 levels did not correlate with rejection grades. Graft-versus-host reaction was noted only in the skin in 29% of group A, 73% of group B, and 77% of group C pigs; liver and native bowel were not involved.(ABSTRACT TRUNCATED AT 400 WORDS)

The emergence of xenotransplantation

The field of transplantation is faced with a growing shortage of human organs as the list of potential recipients continues to increase. Those currently listed can already expect long waits; some die waiting. Xenotransplantation is a potential solution to this widening donor-recipient disparity. Consequently, in recent years, there have been several clinical attempts using organs from nonhuman primates and pigs. The results with nonhuman primates as donors have been encouraging, but it is unlikely that these species will provide a long-term solution to the organ shortage. Most recent xenotransplantation research has therefore shifted to more phylogenetically disparate species, such as pigs, as potential donors. The major barrier to transplantation between members of disparate species combinations has been hyperacute rejection (HAR). The elements of humoral immunity involved in this rejection process include (1) naturally occurring antibodies directed against carbohydrate and other antigens expressed on pig endothelium, and (2) the complement system, which is activated by binding of natural antibodies to their targets. Several elegant strategies to prevent HAR are being developed. The creation of transgenic pigs, whose cells express human regulators of complement activation, is one such strategy. Another promising approach has been to remove antidonor antibodies from the recipient by absorption with some recently characterized carbohydrate epitopes of porcine endothelial xenoantigens. Recent experimental work indicates that HAR can successfully be prevented by inhibition or depletion of complement. A delayed type of xenograft rejection, characterized by endothelial cell antibody deposition and cellular infiltration, occurs over the next three to four days. The likely mechanisms involved in delayed xenograft rejection include antibody-dependent cell-mediated cytotoxicity and the phenomenon of endothelial cell activation.

The role of monocytes and macrophages in delayed xenograft rejection

Abstract: Despite the development of successful strategies for averting hyperacute rejection (HAR) in both small and large animal xenograft models, a delayed xenograft rejection (DXR) ultimately occurs. This process is characterized by endothelial cell activation and graft infiltration with activated monocytes and natural killer (NK) cells. We evaluated the role of monocytes and macrophages in a guinea pig‐to‐rat model of DXR. Our results suggest that specific interactions between these cells and the xenograft occur that result in their activation, since adoptive transfer of xenoactivated splenocytes significantly accelerated both DXR and allograft rejection, while adoptive transfer of alloactivated splenocytes did not. Furthermore, while normal splenocytes caused antibody‐dependent cell‐mediated cytotoxicity (ADCC) of xenogeneic endothelial cells, xenoactivated splenocytes caused significantly greater endothelial cytotoxicity by antibody‐independent mechanisms. Both normal and xenoactivated splenocytes were significantly less cytotoxic if adherent cells, consisting predominantly of monocytes and macrophages, were first removed. In vivo recipient macrophage depletion, using liposome‐encapsulated dichloromethylene diphosphonate, did not influence DXR and this may indicate that nonphagocytic circulating monocytes may be more important in DXR. However, adoptive transfer of splenocytes from a macrophage depleted, xenoactivated donor did not accelerate xenograft rejection, while splenocytes from a nondepleted xenoactivated donor did, thereby supporting the importance of monocytes and macrophages in this I phase of xenograft rejection.

Combined transplantation of small and large bowel : FK506 versus cyclosporine A in a porcine model

Clinically, FK506 is superior to CsA after solitary small bowel transplantation (SBTx). Development of diarrhea after SBTx has been the rationale for adding the colon to small bowel grafts. However, the additional lymphoid and bacterial content transferred with total small plus large bowel transplants (TBTx) might aggravate the alloimmune response-rejection and graft-versus-host disease (GVHD)-and increase the risk of infection. We studied the incidence of rejection, GVHD, and infection after TBTx and the impact of CsA versus FK506. We performed orthotopic TBTx with portal drainage after total enterectomy in outbred Yorkshire Landrace pigs, divided into 3 groups: control pigs (n=6) received no immunosuppression; CsA pigs (n= 14) received CsA (5 mg/kg), antilymphocyte globulin (10 mg/kg for 10 days), prednisone (2 mg/kg), and AZA (2.5 mgtkg); and FK506 pigs (n=9) received FK506 (0.2 mg/kg) and prednisone (2 mg/kg). Trough CsA whole blood levels were >400 ng/ml for the first 7 days and >200 ng/ml thereafter. FK506 levels were > 15 ng/ml. We excluded from further analysis 5 early deaths (<3 days) due to anesthesiologic (n=2) or technical reasons (n=3). Median survival of control pigs was 9.5 days (range, 4-13). Cyclosporine did not extend survival: median, 9 days (range, 5-31) (P=0.6). FK506 prolonged survival: median, 37 days (range, 21-49) (P<0.001 vs. control and CsA pigs). Of FK506 pigs, 60% gained weight (+75 g/day), whereas 100% of controls and 75% of CsA pigs lost weight (-550 g/day and -300 g/day, respectively). All control pigs died of rejection within 2 weeks versus none of the FK506 pigs. However, 36% of CsA pigs died of rejection. Groupwise comparison showed less rejection in FK506 versus control pigs (P<0.001) and in FK506 versus CsA pigs (P<0.03), but no difference between CsA and control pigs. None of the control pigs died of GVHD versus 18% of CsA pigs (by day 31) and 37% of FK506 pigs (by day 49). Groupwise comparison showed increased GVHD in FK506 versus control pigs (P<0.001) and a tendency toward increased GVHD in FK506 versus CsA pigs (P=0.08). None of the control pigs died of infection alone versus 22% of CsA pigs (by day 31) and 67% of FK506 pigs (by day 49). Groupwise comparison showed increased infection in FK506 versus control pigs (P<0.001). We detected significant endotoxemia early and late postoperatively. But we saw no specific correlation between endotoxemia, rejection, GVHD, or infection. Based on this study, we have drawn several conclusions: (1) In untreated pigs, TBTx provokes a severe rejection response, but no lethal GVHD. (2) Cyclosporine and particularly FK506 pigs have a high incidence of infection and lethal GVHD, a complication that we had not seen after solitary SBTx. (3) FK506 is superior to CsA in controlling rejection and in prolonging graft and recipient survival; FK506, however, does not reduce GVHD, but rather tends to augment it. (4) TBTx causes endotoxemia. As with solitary SBTx, FK506 is superior to CsA after TBTx. However, longterm survival is difficult to achieve on FK506 recipients because of the development of GVHD and infection.