Old World Monkeys are less than ideal transplantation models for testing pig organs lacking three carbohydrate antigens (Triple-Knockout)

Triple-knockout (TKO) pigs (with added protective human transgenes) are likely to be optimal sources of organs for clinical organ xenotransplantation because many humans have minimal or no natural antibody to TKO pig cells. However, Old World monkeys (OWMs) have naturally-existing antibodies directed to TKO cells. We measured anti-pig IgM/IgG binding, and complement-dependent cytotoxicity to wild-type (WT), α1,3-galactosyltransferase gene-knockout (GTKO), and TKO pig peripheral blood mononuclear cells (PBMCs) using sera from humans, several OWMs, and two New World monkeys (NWMs). Furthermore, we compared survival of GTKO (n = 5) and TKO (n = 3) pig kidneys in baboons. OWMs had significantly greater IgM binding and cytotoxicity to TKO PBMCs than humans or NWMs. Mean anti-TKO IgM was significantly higher in OWMs and significantly lower in NWMs than in humans. Cytotoxicity of OWM sera to TKO PBMCs was significantly greater than of human serum, but there was no significant difference between human and NWM sera. The median survival of TKO pig kidneys (4 days) in baboons was significantly shorter than that of GTKO kidneys (136 days) (p < 0.05). Even though considered ideal for clinical xenotransplantation, the presence of naturally-existing antibodies to TKO pig cells in OWMs complicates the transplantation of TKO pig kidneys in OWMs.

Sources of pig cells. PBMCs were obtained from (i) WT, (ii) GTKO and (iii) TKO pigs (none of which expressed any human 'protective' transgenes) (Revivicor, Blacksburg, VA). Finally, we obtained GTKO/β4NT2KO pig cells (without expression of any human 'protective' transgenes). As the volumes of all primate sera were limited, we only tested baboon serum against GTKO/β4NT2KO PBMCs. All pigs were of blood type O (nonA). Isolation of PBMCs was as described previously 11,12 . Detection of expression of xenoantigens on selected pig and primate cells by flow cytometry. PBMCs from pigs or primates were stained for expression of Gal (by isolectin BSI-B4), Neu5Gc (chicken anti-Neu5Gc mAb), and Sd a (Dolichos biflorus agglutinin, DBA), as previously described 10 .
Binding of serum IgM and IgG to pig PBMCs by flow cytometry. Serum antibody binding to PBMCs was measured by flow cytometry, as previously described 10,11 . Briefly, isolated PBMCs (1 × 10 5 cells/tube) were incubated with 20 µl of serum (20% final concentration or 20 µl of phosphate-buffered saline [PBS] [control]) for 2 h at 4 °C. After incubation, the cells were washed twice in 3 ml FACS buffer and centrifuged at 700 g for 5 min. The supernatant was discarded. To prevent non-specific binding, 10 µl of 10% goat serum was added. Detection of IgM or IgG binding was performed by further incubating the serum with fluorescein isothiocyanate conjugated goat anti-human IgM (μ chain specific) (ThermoFisher Scientific, Waltham, MA) at 1:50 dilution or IgG (γ chain specific) (ThermoFisher Scientific) at 1:50 dilution for 30 min in the dark at 4 °C. The samples were washed twice and the cells resuspended with 200 µl diluted (×3) fixation buffer (Becton Dickinson, San Diego, CA).
Data acquisition was measured by LSR II flow cytometry (Becton Dickinson, San Jose, CA). The results were expressed as the relative geometric mean (rGM), calculated by dividing the geometric mean fluorescence for each sample by that of the negative control, as previously described 11,12 . The data were analyzed by FlowJo software (Treestar. Ashland, OR). Each sample was measured x3. complement-dependent cytotoxicity (cDc) assay for pBMcs. PBMCs  where A represented the percentage of dead cells, B was the maximal percentage of dead cells (PBMCs fixed with 70% ethanol), and C was the minimal percentage of dead cells (PBMCs incubated with medium only).
CDC values at varying serum concentrations (50%, 25%, 12.5%, 6.25% and 3.125%) were calculated, and a curve was generated for each serum sample. Genetically-engineered pig-to-baboon kidney xenotransplantation (n = 8). Intravascular catheter placement in baboons, donor pig nephrectomy, and life-supporting pig kidney transplantation in baboons have been described previously 1,15 . The baboons were divided into two groups based on the pig kidney they received -a TKO group (n = 3) and a GTKO group (n = 5). All baboons received identical immunosuppressive therapy -induction with anti-thymocyte globulin (thymoglobulin) and anti-CD20mAb (rituximab) 1 , and maintenance with anti-CD40mAb, rapamycin, and low-dose corticosteroids 1 . Animal care was in accordance with the Guide for the Care and Use of Laboratory Animals published by the Institute of Laboratory Animal Research (8th edition, 2010). Protocols were approved by the Institutional Animal Care and Use Committees at the University of Pittsburgh (IACUC) (#13082323) and the University of Alabama at Birmingham (#20673). The approved protocols included the approach to pain relief and animal well-being. Rejection-free survival was defined based on histopathological findings at euthanasia. Statistical analysis. Continuous variables were expressed as mean ± standard error of the mean (SEM).
Comparisons among multiple groups were performed using a Kruskal-Wallis test for continuous variables. The receiver operating characteristic (ROC) curve was used to determine cytotoxicity using PBMC cut-off points based on both the positive data (individual human serum against WT PBMCs) and the negative data (individual human serum against autologous human PBMCs). Correlations between IgM and CDC or between IgG and CDC were analyzed by calculating a Pearson correlation coefficient. Baboon survival and rejection-free survival were estimated using the Kaplan-Meier method, and the overall differences between curves were compared using the log-rank test. A p value of <0.05 was considered statistically significant. All statistical analyses were performed using social sciences software GraphPad Prism 8 (GraphPad Software, San Diego, CA). ethical approval. The protocol for withdrawal of human serum was approved by the Institutional Review Board of the University of Alabama at Birmingham (#300001924). Protocols for pig and baboon studies were approved by the Institutional Animal Care and Use Committees of the University of Pittsburgh (#13082323) and the University of Alabama at Birmingham (#20673).
In summary, the mean IgG antibody binding to WT cells and to GTKO cells was lower in OWMs (rGM for WT 3.9 and for GTKO 1.2) and NWMs (rGT for WT 1.6 and for GTKO 2.0) than in humans (rGM for WT 9.3 and for GTKO 5.7) (p < 0.05) (Fig. 1B). The mean IgG antibody to TKO cells was significantly lower in NWMs monkeys (rGT 0.9) than in humans (rGM 1.2) (p < 0.01).
These results indicate that for transplantation of kidneys from TKO pigs, IgM antibody binding will be substantially greater in OWM recipients than in humans. In contrast, for kidneys from WT or GTKO pigs, IgM antibody binding will be similar or lower in OWM recipients than in humans. cytotoxicity to Wt, GtKo, and tKo pig pBMcs. Cytotoxicity of <6.4% (representing the best compromise between sensitivity and specificity) was selected as the cut-off point for this assay, i.e., the lower level of www.nature.com/scientificreports www.nature.com/scientificreports/ detection. The outcomes of serum cytotoxicity to WT, GTKO, and TKO PBMCs, expressed in percentages, varied between the different primate species (Fig. 2).
These results indicate that, for pig organ transplantation, the use of GTKO pigs will give similar results for both OWM and NWM recipient species and for human recipients. However, when TKO pigs are used, OWM species may yield a different outcome which is not representative of the condition in human recipients. correlation between anti-pig igM or igG and cDc. To investigate the extent to which IgM or IgG antibody is associated with the results of the CDC assay, the correlation between CDC (at 25% serum concentration) and binding (rGM) of IgM or IgG antibody was evaluated.
In humans, significant correlations were found between IgM/IgG binding to TKO PBMCs and CDC (p < 0.05) (Fig. 4). In OWMs, significant correlations were found between IgM binding to both GTKO and TKO PBMCs and CDC (p < 0.05) (Fig. 4), but not for IgG binding. In NWMs, significant correlations were found between IgM binding to WT, GTKO, and TKO PBMCs and CDC (p < 0.05) (Fig. 4), but not for IgG. Even a low antibody level in OWMs might be associated with much higher CDC of TKO PBMCs than in humans (Fig. 4).
Because there was significant correlation between IgM binding and CDC to TKO PBMCs in all primates, the CDC/IgM ratio (the CDC [%] divided by IgM binding [rGM]) was calculated as percentage/rGM (Fig. 5). The CDC/IgM ratio in OWMs was significantly higher than in humans (p < 0.01). In contrast, there was no significant difference between humans and NWMs.
In summary, when one (GTKO) or more (TKO) pig xenoantigens are deleted, there is a significant reduction in the binding of human antibodies and CDC to pig PBMCs (Fig. 6). For OWMs, the mean CDC to TKO PBMCs was significantly greater than that of GTKO PBMCs, even though the levels of antibody binding were not significantly different. The binding of NWM antibodies and CDC to TKO PBMCs were lower than to GTKO PBMCs.  www.nature.com/scientificreports www.nature.com/scientificreports/

In vivo study. IgM and IgG binding/cytotoxicity of recipient baboon serum to GTKO and TKO pig PBMCs.
In the baboons that received kidneys from TKO pigs (TKO group), the level of pre-transplant mean IgM antibody to TKO cells was almost 2 (rGM) (Fig. 7). In contrast, the mean rGM values of IgM antibody binding to GTKO cells in baboons in the GTKO group were between 3 and 5, except for B17615 (mean rGM:1.7), which was the longest survivor (260 days). In the TKO group, the pre-transplant CDC to TKO PBMCs was 50-70%, except for B1417 (19%), which was the longest survivor (61 days). In the GTKO group, the pre-transplant CDC to GTKO PBMCs was also 70-90%, except for B17315 (36%) and B17615 (16%), which were the 2 longest survivors (237 and 260 days).
Pig kidney graft survival and rejection-free survival. The 3 baboons in the TKO group required euthanasia on days 1, 4, and 61, respectively (Fig. 8A). The graft in B3917 (which had the highest cytotoxicity among the 3 baboons in the TKO group) failed within 24 h (and so must be considered as hyperacute rejection, confirmed by histopathology). B2416 developed hematemeses, requiring euthanasia on day 4 for acute gastric dilatation (a complication seen previously in NHPs 16 ). B1417 developed acute humoral xenograft rejection on day 28; rescue therapy (high-dose steroids, anti-TNF-α antibody, intravenous immunoglobulin) extending over a month was only partially successful, and the baboon was euthanized on day 61.
The grafts in the baboons in the GTKO group functioned well for several months, except in B3115, which developed hematemeses, requiring euthanasia on day 4 for acute gastric dilatation ( Table 2). As previously described 1 , in B10815 acute humoral xenograft rejection developed after withholding two doses of anti-CD-40mAb (because of neutropenia), and the baboon was euthanized on day 90. The other 3 baboons required euthanasia for infectious complications on days 136, 237, and 260, but with no clinical or histopathologic signs of rejection, except possibly mild fibrosis (Table 2), as previously described 1 . www.nature.com/scientificreports www.nature.com/scientificreports/ Baboon survival in the TKO group (median survival 4 days) was significantly shorter than in the GTKO group (median survival 136 days) (p < 0.05) (Fig. 8A). Rejection-free survival in the TKO group was also significantly shorter than in the GTKO group (p < 0.05) (Fig. 8B).
Although the number of transplants in each group was small, the results suggest that in baboons (an OWM) the immune response to a TKO kidney graft resulting in antibody-mediated rejection is stronger than to a GTKO graft, even though the pre-transplant CDC values between the TKO and GTKO groups were not significantly different (Fig. 9).

Discussion
Historically, the expression of Gal on pig vascular endothelial cells presented the most important immunobiological barrier in pig-to-primate xenotransplantation until genetically-engineered pigs were produced. For WT and GTKO pig organ transplantation, OWMs have proved to be an acceptable model with respect to translation to the condition in humans. Today, there is an increasing spectrum of pigs with genetic modifications that protect the pig tissues from the immune response (and/or correct molecular incompatibilities between pigs and primates) 9,17,18 . There are increasing in vitro data indicating that TKO pig organs will prove to be a major advance over GTKO organs for transplantation into humans 9 (Fig. 6). However, when Neu5Gc is no longer expressed in the pig (after TKO), OWMs have been found to have antibodies against an epitope in the pig, which apparently emerges upon deletion of Neu5Gc epitopes. This epitope is referred to as the "4 th xenoantigen" [8][9][10] , and most likely is a carbohydrate moiety.
We therefore compared OWM and NWM serum IgM/IgG binding to, and cytotoxicity of, WT, GTKO, and TKO PBMCs, and compared the results with those of human sera.
For OWMs, serum IgM binding and CDC to TKO PBMCs were higher than for human and NWMs (Figs. 1 and 3), and only anti-TKO IgM correlated with CDC (Fig. 4). The CDC outcome associated with anti-TKO IgM was significantly different among the 3 groups, and a main finding was that OWMs have significantly stronger cytotoxicity against TKO PBMCs than humans (Fig. 5). These data strongly suggest that antibodies against the 4 th xenoantigen are mainly IgM with strong cytotoxicity.
In contrast, most humans did not have natural antibodies to the 4 th xenoantigen, and NWMs had lower levels of IgM antibodies than OWMs. In summary, even when humans have anti-TKO antibodies, their serum cytotoxicity was much lower than that of OWMs. The cytotoxicity of NWMs serum fell between that of OWMs and humans.
Based on this vitro study, we evaluated the response to kidneys from TKO genetically-engineered pigs with six additional transgenes (which might be ideal organs for humans 9 ). To our knowledge, this is the first report of kidney transplantation in baboons using TKO pigs with an additional six transgenes. Pig kidney transplantation (with an anti-CD40 mAb-based immunosuppressive regimen) 1 was carried out in two groups of baboons that received kidneys from either TKO or GTKO pigs (each with added human transgenes) ( Table 2, Figs. 7-9). The pigs in the GTKO group expressed fewer 'protective' transgenes than those in the TKO group, and so this might be detrimental to the outcome in the GTKO group. However, the baboons in the GTKO group showed a significantly longer kidney graft survival than those in the TKO group, and a significantly lower incidence of antibody-mediated rejection. It is tempting, therefore, to suggest that anti-TKO antibodies contributed to the rejection process.
We attempted to select the baboons on the basis of having low anti-pig antibody levels, as previously described 14 . The five baboons that received GTKO pig kidneys (GTKO group) met this criterion, but the three baboons that received a TKO pig kidney (TKO group) had anti-TKO IgG. Although the three baboons in the TKO group had lower anti-TKO IgM (compared to anti-GTKO IgM in the GTKO group), the cytotoxicity in the TKO group was similar to that in the GTKO group (Fig. 9). www.nature.com/scientificreports www.nature.com/scientificreports/ Limitations of our in vivo study are the small number of transplants in the TKO group, and variation in donor GTKO pig genotype. Although none of the pigs that provided cells for the in vitro studies expressed any human transgenes, the organ-source pigs in the in vivo studies expressed several human transgenes. There were some differences between the GTKO and TKO pigs with respect to the presence of these additional transgenes, but these differences were not statistically significant (possibly because of the small numbers) (Supplementary Table 1). It is therefore difficult to conclude that expression of any of these human transgenes influenced the transplant outcome. At least, our preliminary data indicate that the expression of various human transgenes does not affect antibody binding to the cells ( Supplementary Fig. 1), which is related only to the expression of the glycan antigens.
Furthermore, the sources of kidneys were pigs all of which (except one) expressed at least one human complement-regulatory protein and all of which (except one) expressed at least one human coagulation-regulatory protein. In our experience, these are the key transgenes that protect the organ from antibody-mediated injury (in addition to lack of antigen expression). The expression of the transgenes is measured before the pigs are used as sources of kidneys, and expression is uniformly good in the great majority of cases. We therefore do not believe these variations in the organ-source pigs played a significant role in outcome.
If the baboons had not been well-immunosuppressed, it is likely that they would develop antibodies to the human proteins expressed on the pig cells. As they receive effective immunosuppressive therapy, they do not develop antibodies to pig antigens. We therefore believe it is extremely unlikely they would develop antibodies to human proteins. Although we have not specifically tested for this, we do know that, even if they become sensitized to pig antigens, they do not become sensitized to third-party allo-antigens 19 .
Our data indicate that, because of the production of antibodies to the 4 th xenoantigen, OWMs provide a greater barrier to survival of organs from TKO pigs than do humans or NWMs. This will make it difficult to provide the regulatory authorities with positive data to support the concept that TKO pig organs will provide long-term graft survival in humans. NWMs mimic the human immune response to TKO pig cells more closely than OWMs, but transplantation of pig organs into NWMs would be difficult because of the small size of these monkeys available in the USA (e.g., capuchin monkeys 2-4 kg, squirrel monkeys <1 kg). If a small pig kidney is transplanted into a NWM, any growth of the kidney might be problematic 2,3,20 . Furthermore, there has been much less experience of the use of immunosuppressive agents in NWMs than in OWMs, and some biological agents available today are not effective in NWMs (Yamamoto T, et al. manuscript submitted.).
We have previously demonstrated that baboon serum antibody binding to GTKO/β4GalNT2-KO pRBCs is significantly less than to TKO pRBCs 10 . In the present study, baboon IgM binding to, and cytotoxicity of, GTKO/ β4GalNT2-KO PBMCs were significantly less than to TKO PBMCs (Fig. 10). If GTKO/β4GalNT2-KO pigs provided the organs for transplantation into OWMs, this might resolve the problem. In conclusion, when TKO pig organs are transplanted, OWMs are less than ideal as recipients as they do not mimic the human immune response.