Prolonged xenokidney graft survival in sensitized NHP recipients by expression of multiple human transgenes in a triple knockout pig

Abstract

Genetic modification of porcine donors, combined with optimized immunosuppression, has been shown to improve outcomes of experimental xenotransplant. However, little is known about outcomes in sensitized recipients, a population that could potentially benefit the most from the clinical implementation of xenotransplantation. Here, five highly allosensitized rhesus macaques received a porcine kidney from GGTA1 (α1,3-galactosyltransferase) knockout pigs expressing the human CD55 transgene (1KO.1TG) and were maintained on an anti-CD154 monoclonal antibody (mAb)-based immunosuppressive regimen. These recipients developed de novo xenoreactive antibodies and experienced xenograft rejection with evidence of thrombotic microangiopathy and antibody-mediated rejection (AMR). In comparison, three highly allosensitized rhesus macaques receiving a kidney from GGTA1, CMAH (cytidine monophospho-N-acetylneuraminic acid hydroxylase), b4GNT2/b4GALNT2 (beta-1,4-N-acetyl-galactosaminyltransferase 2) knockout pigs expressing seven human transgenes including human CD46, CD55, CD47, THBD (thrombomodulin), PROCR (protein C receptor), TNFAIP3 (TNF-α induced protein 3), and HMOX1 (heme oxygenase 1) (3KO.7TG) experienced significantly prolonged graft survival and reduced AMR, associated with dampened post-transplant humoral responses, early monocyte and neutrophil activation, and T cell repopulation. After withdrawal of all immunosuppression, recipients that received kidneys from 3KO.7TG pigs rejected the xenografts via AMR. These data suggest that allosensitized recipients may be suitable candidates for xenografts from genetically modified porcine donors and could benefit from an optimized immunosuppression regimen designed to target the post-transplant humoral response, thereby avoiding AMR.

OVERLINE: KIDNEY TRANSPLANT

One sentence summary: Xenograft survival can be substantially extended in sensitized recipients by expressing seven human transgenes in triple knockout donor pigs.

Introduction

Xenotransplantation has long been proposed as a therapeutic strategy to address the global organ shortage (1, 2). Recently, the field has undergone a renaissance, attracting both public and scientific interest. Improvements in graft and recipient survival have stemmed from two major advances: (i) utilization of immunosuppressive therapies targeting the CD40/CD154 co-stimulation pathway (3–5), and (ii) genetic engineering of porcine donors to reduce immunogenicity and inherent incompatibilities between the xenograft and the recipient (6, 7).

In the United States, Mohiuddin et al. reported a median survival of 300 days (with survival up to 900 days) in a heterotopic heart transplant baboon model using porcine donors genetically engineered to knock out the glycoprotein alpha-galactosyltransferase 1 (GGTA1) gene and to express the CD46 and thrombomodulin (THBD) human transgenes (GTKO.hCD46.hTHBD), with administration of anti-CD40 monoclonal antibody (mAb)-based immunosuppression (3). Meanwhile, in Europe, Langin et al. reported survival of up to 200 days in a life-supporting orthotopic heart transplant baboon model using a porcine donor with the same genetic edits and anti-CD40 mAb-based immunosuppression (4). In kidney xenotransplantation, a similar approach using GGTA1 knockout and human CD55 transgene (GGTA1KO.CD55Tg) donor pigs with anti-CD154 mAb–based immunosuppression prolonged graft survival up to 400 days (5, 8). Compared to earlier attempts (9), these studies demonstrated that xenografts can survive many months when the CD40/CD154 pathway is blocked, and suggesting that the translation of xenotransplantation to the clinic may be within reach (10). Having demonstrated convincing and repeatable outcomes of xenotransplantation in nonhuman primate (NHP) models, one key question regarding cross-species organ transplantation is, for whom does the potential benefit of xenotransplantation outweigh the risk?

More than 100,000 patients await kidney transplantation in the US and only one in four are likely to receive transplants (11, 12) in a given year. Despite changes to the deceased donor allocation system, highly sensitized patients, those who have developed anti-donor antibodies as a response to foreign human leukocyte antigen (HLA) exposure, remain challenging to transplant. Such patients experience increased wait times compared to non-sensitized counterparts (13, 14), particularly the most highly sensitized who have antibodies against almost all HLA antigens [calculated Panel Reactive Antibody (cPRA) 100%], making them incompatible with virtrually all organ donors. Furthermore, should these patients receive a transplant, they have an increased risk of antibody mediated rejection (AMR) after transplantation (15, 16). Although HLA sensitization is a barrier to successful allotransplantation, such patients have not been exposed to xenoantigens. Thus, theoretically, xenotransplantation might pose the same immunological risk to both sensitized and non-sensitized patients, but for sensitized patients who might otherwise remain untransplanted, xenotransplantation would potentially confer a greater relative benefit. On the other hand, if there is cross reactivity between allo- and xeno-antigens, allosensitization may prime the immune response to a xenotransplant.

Here we used an immunosuppressive regimen that previously resulted in long-term graft survival in a nonsensitized NHP xenokidney transplantation model (5, 8). We investigated whether comparable outcomes could be achieved in a well-established allosensitized NHP model. This study aimed to test the hypothesis that allosensitization does not incur additional immunological risks for xenotransplant recipients (17, 18), and to compare the effects of two cohorts of genetically modified pigs.

Results

Allosensitized NHP recipients of GGTA1KO.CD55Tg porcine donors show early antibody-mediated rejection after xenokidney transplantation

To evaluate whether allosensitization produces xenoreactive antibodies (XAbs), we sensitized recipient animals (weight: 6.45±2.12 kg; n=6) with serial skin grafts from maximally major histocompatibility complex (MHC) mismatched allo-donors (Fig. 1A and Table S1) as previously reported (18–21). Sensitization against the allo-donor was confirmed by elevation of alloantibodies measured by flow crossmatch using donor peripheral mononuclear cells (PBMCs) (Fig. 1B). Serum samples were analyzed at pre-sensitization (naïve), 28 days after first skin transplantation (Peak 1), 14 days after second skin transplantation (Peak 2), and 7 days prior to xenokidney transplantation (Pre-tx). XAbs against donor pig (GGTA1KO.CD55Tg) cells transiently increased after allo-skin transplantation but returned to baseline before xenokidney transplantation (Fig. 1B). In contrast, alloantibodies were persistently elevated up to the time of the xenokidney transplant. A larger cohort (n=41), including our previously reported allosensitized animals (19–25), demonstrated a similar pattern of an initial small rise in xenoantibody post-skin transplant, but there was no persistent anti-porcine IgG against either wild-type or GGTA1KO porcine splenocytes detectable by flow crossmatch (fig. S1, A and B).

Figure 1. Xenokidney transplantation in allosensitized rhesus recipients.

(A) A schematic representation of allosensitization. Six rhesus recipients were sensitized by skin transplantation from maximally MHC-mismatched donors (2 skin grafts from the same allo-donor placed at 8-week intervals). (B) Allo- and xeno-reactive antibodies after skin sensitization. Allosensitization was confirmed by elevation of DSA. Anti-pig antibodies in naïve, first sensitization (peak 1), second sensitization (peak 2), and before xenotransplant were measured. (C) A schematic representation of experimental design and immunosuppression regimen. Six sensitized rhesus monkeys received kidney transplantation from GGTA1KO.hCD55 transgenic pigs. Induction immunosuppression consisted of anti-CD4 mAb and anti-CD8 mAb (50 mg/kg). Maintenance immunosuppression consisted of costimulation blockade with anti-CD154 mAb (20 mg/kg), MMF, and steroids. (D) Kaplan-Meier curve of survival of xenograft in sensitized recipients (blue solid line, n=6, one censored) and allograft in sensitized recipients (red dotted line, n=4).. (E) Post-transplant serum creatinine (sCr) changes (left) and representative images of xenokidney (right) at time of rejection. A rise in serum creatinine corresponds with xenograft rejection. (F) Flow crossmatch of circulating anti-pig IgG after xenokidney transplantation. (G) Representative images of histology (H&E staining, left) and C4d immunostaining (right) at the time of rejection (scale bar: 200μm). (H) Quantification of allo- and xeno-DSA after xenokidney transplantation (n=5). Error bars represent SD. All data points indicate biologically independent animals; *p<0.05; **P<0.01; ***p<0.001 using two-tailed parametric paired t test; NS indicates no statistical significance.

At the time of xenokidney transplantation, allosensitized rhesus recipients received anti-CD4/CD8 mAbs as induction and anti-CD154 mAb, mycophenolate mofetil (MMF), and steroids as maintenance immunosuppression (Fig. 1C). Six animals received xenokidney grafts from GGTA1KO.CD55Tg porcine donors (weight: 32.06±4.04 kg). The average anti-porcine donor IgG XAbs by T and B cell flow crossmatch (TFXM and BXFM) were 2435 and 4117 mean fluorescent intensity (MFI), respectively, at the time of transplantation (Table S2. Although all porcine donors had the same genotype (GGTA1KO.CD55Tg), one of the donors expressed a different swine leukocyte antigen (SLA) type than the other two (Lr-21.22 vs. Lr 3.3) (Table S3). We did not observe hyperacute rejection in any of the six recipients. One animal (HAKJ) was censored from the post-transplant analysis because he was sacrificed due to technical complication (urine leak) without any evidence of rejection (Table S4). As shown in Fig. 1D graft survival of allosensitized macaques with xenokidney transplantation (Mean survival time (MST) = 26.8d, blue line, p=0.0027) was significantly prolonged compared to our previously reported allosensitized animals who received an allokidney transplant from their allosensitizing donor, with conventional immunosuppression (tacrolimus/MMF/steroid) and anti-CD4/CD8 mAb induction (MST = 5d, red dotted line) (19). All animals exhibited graft dysfunction as assessed by serum creatinine (sCr) (Fig. 1E). We did not observe any rapid post-transplant coagulopathy as evidenced by stable prothrombin time (PT), partial thromboplastin time (PTT,) and fibrinogen concentrations in the first 5 days post-transplantation (fig. S2A). All animals had normal hemoglobin (Hb), hematocrit (HCT), and platelet counts during the study period, except at the time of rejection (fig. S2B).

De novo or rebound anti-pig donor-specific antibody (DSA) titers rapidly increased during rejection (Fig. 1F and fig. S3). Consistent with rising XAbs, histological analysis revealed thrombotic microangiopathy (TMA) features along with evidence of AMR (g+ptc>2) together with a paucity of interstitial inflammation, suggesting a lack of T cell–mediated injury (Fig. 1G and Table S5). Although endarteritis is a key feature of T cell-mediated acute rejection (TCMR), the high vasculitis score (v3) was secondary to TMA and AMR rather than TCMR in two recipients (K508 and 620). All xenografts demonstrated evidence of transplant glomerulitis and glomerular fibrin thrombi, manifestations of TMA, peritubular capillaritis (PTC), and glomerular basement membrane duplication, suggestive of transplant glomerulopathy (Table S5). Lastly, we determined that alloantibodies against previous rhesus monkey skin donors were not increased following xenotransplant (Fig. 1H). Taken together, these data suggested that allosensitization prior to xenotransplantation, despite the lack of elevated anti-pig antibody titers at the time of xenokidney transplantation, increased the risk of post-transplant AMR and early graft failure, although xenotransplantation did not further allosensitize recipients.

Xenokidney transplantation induces germinal center responses and follicular helper T cells in sensitized recipients

To confirm involvement of the germinal center (GC) response during humoral xenograft rejection, GCs were evaluated in secondary lymphoid organs. Hyperplastic GCs in both lymph node (fig. S4, A and B) and spleen (fig. S4, C and D) were observed at the time of rejection, suggesting a rapid reconstitution of the GCs despite continuous treatment with anti-CD154 mAb, which has been observed to disrupt germinal centers (26). Consistent with this observation, cells resembling follicular helper T cells (Tfhs; PD-1⁺ICOS⁺CD4 and PD-1^hiICOS⁺CD4) were increased in the peripheral lymph nodes at rejection, particularly significant in animals beyond 3 weeks after transplantation (p=0.0048 and p=0.0003 respectively, Fig. 2A). Because surface ICOS⁺PD-1⁺ expression reflects activated CD4⁺ T cells, BCL-6⁺PD-1⁺ cells were also measured as they mark GC-Tfh cells in the lymph node. BCL-6⁺PD-1⁺CD4⁺ T cells were significantly increased in the secondary lymphoid organs at rejection sampled from animals beyond 3 weeks post-xenokidney transplantation (p=0.029, Fig. 2B). Lastly, we evaluated antibody-secreting cells (ASCs) in the lymph nodes. Congruent with GC induction and GC-Tfh cells in the lymph nodes, the frequency of ASCs was significantly increased during rejection in the lymph nodes (p=0.0219, Fig. 2C). Taken together, these results suggest that, despite immunosuppression with anti-CD154 mAb, intact GC responses in the secondary lymphoid organs may contribute to the development of humoral responses to the xenograft in allosensitized recipients.

Figure 2. Induction of germinal center response after xenotransplantation.

(A) Representative flow cytometry plots for ICOS⁺PD-1⁺ and ICOS⁺PD-1^hi CD4⁺ T cells (left panels) and quantification (right panels) for lymph node ICOS⁺PD-1⁺ and ICOS⁺PD-1^hi CD4⁺ T cell populations at pre-transplant and rejection (n=5). (B) Representative flow cytometry plots (left panels) and quantification (right panels) for PD-1^hiBCL-6⁺ Tfh cell (n=5). (C) Representative ELIspot images (left panels) and compiled data (right panel) for antibody-secreting cells (ASCs) in lymph node (LN) at pre-transplantation and rejection (n=5). Error bars represent SD. Each dot indicates biologically independent animals; Data were analyzed with all samples (black, n=5) and with samples collected at least 3 weeks after transplantation (red, n=3); p-values were calculated using two-tailed parametric paired t test.

Xenokidney transplantation promotes aberrant expansion of exhausted CD4⁺ T cells

After xenokidney transplantation, we observed a marked phenotypic change of circulating CD4⁺ T cells, with central memory T cells (T_CM; CD95⁺CD28⁺CCR7⁺CD45RA⁻) becoming the dominant population (fig. S5A). Furthermore, there was a significant increase in PD-1 expression (p=0.042, Fig. 3A, left), and we noted a heightened ICOS expression on circulating CD4⁺ T cells (Fig. 3A, right). These two surface markers are of particular interest since both are known to be expressed on Tfh cells in circulation (cTfh) (27). ICOS⁺PD-1^hi CD4⁺ T cells, which are usually only found in secondary lymphoid organs (19), gradually increased in the peripheral circulation during T cell repopulation after induction with CD4/CD8 depletional therapy (Fig. 3B). The frequency of both ICOS⁺PD-1⁺ and ICOS⁺PD-1^hi CD4⁺ T cells were significantly increased after xenokidney transplantation (p=0.031 and p=0.028 respectively, Fig. 3C). Transcriptome analysis of peripheral CD4⁺ T cells showed upregulation of exhaustion markers (including PD-1, LRBA, TACI, and TIGIT) and IL-21 (fig. S5B). Although ICOS and PD-1 expression as well as IL-21 are associated with Tfh cells, the expression of BCL-6 and CXCR5, two definitive markers of Tfhs, remained unchanged within the circulating CD4⁺ T cell pool (fig. S5C). Consistent with this, CXCR5⁺ CD4⁺ and CXCR5⁺PD-1⁺CD4⁺ T cells in the peripheral blood, as analyzed by flow cytometry, were not elevated after xenokidney transplantation despite marked increases in PD-1 and ICOS expression in this circulating CD4⁺ T cell population (fig. S5D). In summary, the phenotype of circulating CD4⁺ T cells shifted after xenotransplantation with an increase in markers associated with Tfhs and exhausted T cells.

Figure 3. Aberrant expansion of circulating ICOS⁺PD-1⁺CD4⁺ T cells after xenokidney transplantation.

(A) Representative histogram (upper) and quantitation (lower) of PD-1 (left) and ICOS (right) expression on circulating CD4⁺ T cells at pre-transplant (red) and rejection (blue). (B) Representative flow cytometry plots for ICOS⁺PD-1⁺ and ICOS⁺PD-1^hi cells in gated CD4⁺ cells during T cell repopulation. (C) Kinetics (left) and frequency (right) of post-transplant circulating ICOS⁺PD-1⁺ (top) and ICOS⁺PD-1^hi (bottom) CD4⁺ T cells. Error bars represent SD. Each dot indicates a biologically independent animal; Data were analyzed with all samples (black, n=5) and with samples collected at least 3 weeks after transplantation (red, n=3); p-values were provided using two-tailed parametric paired t test.

Transplantation with porcine kidneys carrying 3 gene KOs and seven human transgenes (3KO.7TG) prevents early onset AMR and prolongs graft survival

To evaluate whether further genetic modifications to the kidney xenograft could mitigate early onset AMR, kidneys from porcine donors lacking the genes for three major known carbohydrate xenoantigens [GGTA1, CMAH (cytidine monophospho-N-acetylneuraminic acid hydroxylase), and SDa (b4GNT2 and b4GALNT2); called 3KO] and seven human transgenes [human CD46, CD55, CD47, THBD (thrombomodulin), PROCR (protein C receptor), TNFAIP3 (TNF-α induced protein 3), and HMOX1 (heme oxygenase 1); called 7TG] were evaluated in the rhesus macaque NHP model. 3KO.7TG (EGEN-2734) porcine donors were generated employing CRISPR/Cas9-mediated gene editing (28) and were provided by eGenesis. All four (4) recipient NHPs were sensitized as described in Fig. 1, and received identical immunosuppression. Donor/recipient weight, pre-transplant DSA (TFXM and BFXM) as well as total ischemic time were comparable between the two groups (fig. S6 and Table S2). One animal (L106) was censored from post-transplant analysis due to a technical complication (mesh infection, as a consequence of donor-recipient size mismatch necessitating complex abdominal wall closure) without evidence of rejection (Table S4). Sensitized recipients receiving the 3KO.7TG kidney had significantly prolonged graft survival compared to those receiving the GGTA1KO.hCD55 (1KO.1TG) kidneys (MST 61 vs. 24 days; P = 0.0482, Fig. 4A). By one-month post-transplantation, sCr concentrations were significantly lower (p=0.0347) in recipients receiving 3KO.7TG kidneys although peak DSA was not significantly different between groups (Fig. 4, B and C). Congruent with these findings, AMR score (combined glomerulitis and peritubular capillary inflammation scores; g+ptc) and glomerulopathy/microvascular injury score (combined glomerulopathy and mesangial matrix scores; cg+mm) were significantly reduced at one-month post-transplantation (p=0.0015 and p<0.0001, respectively; Fig. 4D). Acute rejection (ACR) scores (combined tubulitis, intimal arteritis, and interstitial inflammation scores; t+v+i) were not different between groups (Fig. 4D and Table S6). C4d and C3d depositions in peritubular capillaries were not significantly different between the two groups. However, C3d deposition in glomeruli was significantly reduced in 3KO.7TG renal grafts (p=0.024; Fig. 4E). C3d deposition in glomeruli also showed a strong correlation to AMR (fig. S7).

Figure 4. Transplantation of 3KO.7TG xenokidneys results in prolonged graft survival and reduced antibody-mediated rejection.

(A). Kaplan-Meier curve of survival of xenograft in sensitized recipients with 3KO.7TG (n=3) and 1KO.1TG (N=5) pig donor. (B) Individual (left panel) and quantitated (right panel) post-xenotransplant serum creatinine (sCr). (C) Individual (left panel) and quantitated (right panel) post-xenotransplant DSA (anti-pig donor antibodies). (D) Representative PAS staining of 1KO.1TG (top left) and 3KO.7TG (top right) grafts at one-month post-transplantation and collated clustered Banff pathology gradings (bottom) for antibody-mediated rejection (g+ptc), microvascular injury (mm+cg), T-cell mediated rejection (t+v+i) at one-month post-transplantation in 1KO.1TG (blue) or 3KO.7TG (green) xenokidneys. (E) Representative C3d immunostaining of 1KO.1TG (top left) and 3KO.7TG (top right) grafts at one-month post-transplantation and collated grading (bottom) for C4d at peritubular capillaries (PTC), C3d at peritubular capillaries (PTC), and C3d at glomeruli (C3d_G) at one-month post-transplantation with 1KO.1TG (blue) or 3KO.7TG (green) xenokidneys. (F) Volcano plots (left) and hierarchical clustering (right) of differentially expressed genes at one month after xenotransplantation of 1KO.1TG (n=3) or 3KO.7TG (n=3) using human NanoString platform. (G) Differentially expressed genes between 1KO.1TG and 3KO.7TG xenografts at one-month post-transplantation. The Purple background indicates 31 downregulated genes, while the green background indicates 18 upregulated genes (including human transgenes) in 3KO.7TG (n=3) vs. 1KO.1TG (n=3) xenokidney transplantation. g, glomerulitis; ptc, peritubular capilaritis; cg, glomerular basement membrane double contours; mm, mesangial matrix expansion; t, tubulitis; v, intimal arteritis; i, interstitial inflammation. Error bars represent SD. N number indicates biologically independent animals; *p<0.05 using two-tailed parametric unpaired t test; NS indicates no statistical significance.

To investigate whether changes at the gene level might explain the differential survival outcomes seen in recipients of the two porcine donor kidney genotypes, the NanoString platform was employed. RNA samples were prepared from formalin fixed paraffin embedded (FFPE) kidney xenografts at necropsy. As expected, large sets of genes were upregulated after xenokidney transplantation regardless of the donor kidney. Significantly higher gene expression of CD46, CD47, THB), and TNFAIP3 was observed in 3KO.7TG compared to 1KO.1TG xenokidneys, most likely reflecting expression of the human transgenes. The two donor types clearly segregated in the hierarchical clustering analysis (Fig. 4F). Forty-nine differentially expressed genes (Fig. 4G) were identified, with 3KO.7TG xenografts having 14 upregulated genes (excluding human transgenes CD46, CD47, THBD, TNFAIP3; under green background) and 31 downregulated genes (under purple background) compared to the 1KO.1TG xenografts. The 3KO.7TG xenokidneys showed down-regulation of genes associated with the neutrophil degranulation pathway (S100A8, S100A9, MMP9, LCN2, LTF, ARG1, PF4, and ADAM8), the coagulation pathway (PF4, PLAT, and PPBD), and reduced transcripts associated with B cells and plasma cells (SDC1, FCGR2B, BCL3, and LILRB2), suggesting a less inflammatory phenotype and possibly fewer infiltrating immune cells. Taken together, these data suggest that the rejection phenotype is milder in 3KO.7TG xenografts and may be associated with improved regulation of anti-graft immune response.

3KO.7TG Xenokidney transplantation prevents early innate and adaptive immune cell activation

Because 3KO.7TG donors express human transgenes involved in coagulation (THBD, PROCR) and complement regulation (CD55, CD46), blood chemistry changes after xenokidney transplantation were evaluated. HgB and HCT values one-month post-transplantation declined significantly in both groups, whereas platelet concentrations fell significantly in 1KO.1TG recipients but not 3KO.7TG recipients (fig. S8, top row), suggesting a systemic microangiopathic hemolytic anemia (MAHA) in 1KO.1TG recipients and consistent with local evidence of TMA in these xenografts. Overall, there were no significant differences between the two groups in terms of phosphorus, potassium, or calcium concentrations.

To evaluate the potential mechanisms leading to prolonged graft survival in 3KO.7TG recipients, circulating immune cell populations were compared between 3KO.7TG and 1KO.1TG recipients. After induction with anti-CD4/CD8 mAbs, circulating lymphocytes declined in both groups but greater repopulation was observed with 1KO.1TG recipients (Fig. 5A). 1KO.1TG xenokidney recipients showed significantly increased numbers of circulating monocytes and neutrophils at one-month post-transplant (p=0.0168 and p=0.0175, respectively; Fig. 5A). After depletion, we observed a greater T cell repopulation in 1KO.1TG xenokidney recipients, which was potentially driven by an increase in memory T cells (T_CM and T_EM) (Fig. 5B, fig. S9, A and B). The circulating B cell frequency was significantly lower at one-month post-transplant in 1KO.1TG xenokidney recipients (Fig. 5B), likely secondary to the increased frequency of T cells. However, the numbers of circulating naïve and memory B cells were not different between the two groups (Fig. 5B, fig. S10B). The numbers of circulating plasmablasts (CD3⁻CD20⁻CD27⁺CD38⁺) were broadly similar between 1KO.1TG and 3KO.7TG recipients, with the exception of one-month post-transplant where a marginal increase was observed in 1KO.1TG recipients. (Fig. 5B). 3KO.7TG xenokidney recipients also showed a significantly lower number of ASCs in lymph nodes compared to 1KO.1TG xenokidney recipients (p=0.0192; Fig. 5C). Because the T cell repertoire showed distinct repopulation kinetics between the groups, we further analyzed T cell subpopulations. There were no differences in any CD4⁺ T cell subset between the groups (fig. S9). However, terminally differentiated memory (T_EMRA; CD45RA⁺CCR7⁻, p<0.05) and effector memory (T_EM; CD45RA⁻CCR7⁻, p=0.07) CD8 T cells were reduced in circulation in animals receiving the 3KO.7TG xenokidney graft (Fig. 5D). Unlike 1KO.1TG xenokidney recipients (fig. S6), 3KO.7TG xenokidney recipients showed a relative lack of GC structures 1-month post-transplant on histologic analysis, along with marginally decreased GC-Tfh and GC-B cells in the lymph node (Fig. 5, E and F). Taken together, these data suggest that receiving a 3KO.7TG genotype donor kidney may alleviate the post-xenotransplant response by down-regulation of the adaptive and innate immune responses.

Figure 5. Transplantation of 3KO.7TG xenokidneys results in reduced innate and adaptive immune cell responses.

(A) Absolute number of circulating lymphocytes, monocytes, and neutrophils following anti-CD4/CD8 induction in both the 1KO.1TG (n=6) and 3KO.7TG (n=4) groups. (B) Absolute number of circulating T cells, B cells, and plasmablasts following anti-CD4/CD8 induction in both the 1KO.1TG (n=6) and 3KO.7TG (n=4) groups. (C) Representative ELIspot images (top) and quantification (bottom) for ASCs from LN from 1KO.1TG (n=5) vs. 3KO.7TG (n=3). (D) Representative CD4⁺ (top) and CD8⁺ (bottom) T cell subset analysis based on the expression of CCR7 and CD45RA in 1KO.1TG (left, n=3) and 3KO.7TG (right, n=3) in the lymph node at one-month post-transplantation. Quantitated CD4⁺ (top) and CD8⁺ (bottom) T cell subsets are depicted on the right. (E) Representative histology (H&E) images showing lymph node and germinal center architectures (white dotted line) before and after xenokidney transplantation (one-month) in the 3KO.7TG group. (F) Lymph node GC (CXCR5⁺PD-1⁺ and BCL6⁺) CD4⁺ T and (BCL6⁺) B cells were analyzed in the 1KO.1TG (n=3) and 3KO.7TG (n=3) group at 1-month post-transplant. Error bars represent SD. Each dot indicates biologically independent animals; *p<0.05 using two-tailed parametric un-paired t test; NS indicates no statistical significance.

Xenokidney transplantation rejection is mediated by the post-transplant humoral response

To study the mechanisms underlying xenograft rejection in allosensitized individuals, anti-CD154 mAb treatment was discontinued following the 4^th dose on post-operative day (POD) 35. All animals exhibited low serum hemoglobin concentrations (fig. S8) and were subsequently found to be positive for simian parvovirus (SPV). After discontinuing anti-CD154 mAb, the number of monocytes and neutrophils were unchanged (Fig. 6A, top row) as measured by complete blood count, whereas the lymphocyte population significantly increased, as did CD4⁺ and CD8⁺ T cell populations (p=0.0087 and p=0.0295, respectively; Fig. 6A, bottom row). This suggests that lymphocytes are key players in graft rejection of 3KO.7TG xenokidneys, and that the introduced human transgenes control innate cell activation independent of the immunosuppression. Although total T cell numbers were significantly increased (p=0.0024, Fig. 6A bottom row), B cells numbers were unchanged in the blood (fig. S11). As T cells rebounded, the phenotype of CD8⁺ T cells changed, similar to the phenotypic homeostatic proliferation of cells in a lymphopenic environment (29). Naïve CD8⁺ T cells (T_N; CD45RA⁺CCR7⁺) were significantly reduced (p=0.0045) whereas T_EM and T_EMRA significantly increased (p=0.0322 and p=0.033, respectively) after anti-CD154 mAb discontinuation (Fig. 6B, right panel). In contrast, the CD4⁺ T cell phenotype was not affected by the withdrawal of anti-CD154 mAb therapy (Fig. 6B, left panel). These data are similar to the phenotype switching of CD4⁺ and CD8⁺ T cells during rejection in 1KO.1TG xenokidney recipients. Biopsies prior to discontinuation of immunosuppression showed no worsening of AMR (g+ptc), glomerulopathy and microvascular injury (cg+mm), or C3d scores compared to one-month biopsy scores. However, all animals eventually developed AMR following discontinuation of immunosuppression (Fig. 6C) although measures of T cell-mediated rejection (evaluated by ACR score: t+v+i) remained low after weaning immunosuppression. Consistent with the histopathological diagnosis of AMR, NanoString analysis in the xenograft showed a trend towards up-regulation of B cell related genes (including CD19, IGHM, IGHG3, IGLC, IGHG4, IGHG2, CD27, CD79A, LILRB4, LILRB2, and BCL2A1) after discontinuation of immunosuppression (Fig. 6D) although there was no significant increase in plasmablasts in the lymph nodes at euthanasia (Fig. 6E). Taken together, these data suggest that discontinuation of CD154-based costimulation blockade resulted in antibody-mediated rejection of the 3KO.7TG xenografts.

Figure 6. Rejection by AMR after discontinuation of immunosuppression in 3KO.7TG xenokidney grafts.

A. The number of circulating lymphocytes, monocytes, neutrophils, and T (total, CD4⁺, and CD8⁺) cells before and after withdrawal of anti-CD154-based immunosuppression. B. CD4⁺ (left) and CD8⁺ (right) T cell subsets analysis based on the expression of CCR7 and CD45RA following withdrawal of immunosuppression in 3KO.7TG (right, n=3) C. Representative PAS images of 3KO.7TG graft after withdrawal of immunosuppression (left) and at rejection (right). Collated clustered Banff pathology gradings for AMR (g+ptc), microvascular injury (mm+cg), ACR (t+v+i), and glomerular C3d deposition (C3d_G) at one-month post-transplantation (n=3), before withdrawal of immunosuppression (n=3), and at rejection (n=3) are also shown. D. Differential gene expression in xenografts pre- vs. post-withdrawal of immunosuppression. E. Representative flow plot (left) and quantification (right) of circulating plasmablasts at one month after xenotransplantation (n=3), after withdrawal of anti-CD154mAb-based immunosuppression, and at rejection (n=3). Error bars represent SD. N number indicates biologically independent animals; *p<0.05; **P<0.01; ***p<0.001 using two-tailed parametric paired t test; NS indicates no statistical significance.

Discussion

There is a renewed interest in the clinical application of xenotransplantation, particularly in light of the heavily publicized first case of a pig to human cardiac xenograft using a genetically modified porcine donor with ten genetic edits under emergency IND approval (30). Despite recent cases of pig-to-human kidney transplantation in brain-dead patients using the GalSafe Pig^™ (31) and a further genetically modified (10 genetic edits) pig (32), xenokidney transplantation from pigs had not been reported in a living human patient until recently but has now been reported with patient survival of less than two months (https://www.nytimes.com/2024/05/12/health/richard-slayman-death-pig-kidney-transplant.html), and details of that transplant are not yet publicly available. Although dialysis offers acceptable renal replacement for end stage renal disease (ESRD) patients, decreasing the urgency of experimental therapies, long-term dialysis is not without risk, and highly sensitized patients with ESRD wait longer than unsensitized counterparts, leading to increased mortality and morbidity while on the waitlist. In the US, a survival benefit from incompatible transplantation has been the justification for immunologically higher risk allotransplantation (33). Such highly sensitized patients may be potential candidates for xenokidney transplantation (34–36). Given that xeno-antigens are not recognized by anti-HLA antibodies, allosensitization should not carry an increased immunologic risk for xenotransplantation, although cross-reactivity of alloantibodies with xenoantigens is possible (37). In considering the homology between human and porcine MHC, and thus potential cross-reactivity, particularly with respect to class II molecules (38), it is difficult to predict in vivo outcomes. Although pig-to-NHP organ transplant models have been instrumental in advancing xenotransplantation, the impact of allosensitization has not previously been fully explored in vivo (39).

Here we showed that a genetically engineered xenograft combined with anti-CD154mAb-based immunosuppression promoted less post-transplant humoral response compared to allograft with conventional immunosuppression in highly allosensitized recipients. In addition, allosensitization with serial skin transplantations only temporarily elevated XAbs. This implies that allosensitization might not pose a significant risk for xenotransplantation. Despite highly elevated alloantibodies at the time of transplantation, sensitized recipients did not exhibit hyperacute or early accelerated xenograft rejection, as would otherwise be expected in allosensitized animals receiving an allograft without desensitization. Furthermore, elevation of alloantibody after xenotransplantation or xenograft rejection was not observed. This supports the potential application of xenotransplantation as a ‘bridging graft’ for sensitized patients with limited transplant options.

We also showed that allosensitization events promoted homogeneous early antibody-mediated rejection via de novo anti-pig antibodies. When compared to reports of non-sensitized animals that received a 1KO.1TG xenokidney transplant under the same immunosuppression regimen (8), graft rejection appeared to be somewhat accelerated in the sensitized recipients. Thus, these data suggested that allosensitization may prime the immune system to mount a humoral immune response after xenotransplantation despite a lack of persistently elevated anti-pig antibodies. Whether the mechanism for priming of B cells is the result of crossreactivity stemming from homologies between NHP and porcine MHCs or bystander activation is currently unclear. Due to the utilization of slightly different donor and recipient pre-selection criteria (5, 8), it is hard to generalize that allosensitization results in an acceleration of AMR, but it is clear that in both sensitized and non-sensitized NHP recipients, porcine xenokidneys are rejected by AMR.

Our data also showed the limitation of anti-CD154mAb monotherapy in xenotransplantation. The advent of anti-CD40 and CD154 mAb-based immunosuppression has permitted great improvements in xenograft survival (40–42) compared to tacrolimus-based therapy (43, 44) by suppressing humoral immunity through costimulation blockade. Studies targeting the CD40-CD154 pathway have demonstrated attenuation of anti-pig antibody production and absence of TMA in the graft together with a reduced incidence of systemic consumptive coagulopathy (45). Nevertheless, although blockade of CD40-CD154 does suppress GC responses (46), the increase of germinal center Tfh cells (BCL-6⁺PD-1⁺CD4 T cells) in lymph nodes, coupled with the observation of hyperplastic germinal centers in lymph nodes and spleen, indicates the persistence of an on-going GC reaction despite anti-CD154 mAb. The rapid post-transplant humoral response seen with 1KO.1TG xenografts in allosensitized NHPs suggests that anti-CD154mAb-based immunosuppression is insufficient to control the humoral response against the xenograft. In keeping with this, long-term xenograft survival has been achieved by other groups through anti-CD20mAb treatment together with CD40-CD154 signaling pathway blockade (3, 4, 44, 47). It is therefore crucial to evaluate whether a supplementary B cell targeting strategy, combined with anti-CD154mAb, can extend xenograft survival in sensitized recipients, analogous to nonsensitized recipients

Circulating Tfh cells are frequently associated with persistent AMR (19, 20, 48). In our model, elevated PD-1⁺ICOS⁺CD4 T cells in the peripheral blood were observed, suggesting that Tfh or Tfh-like cells may play a role in the rejection phenotype observed; however, because these cells lacked CXCR5, it is not clear that they are typical circulating Tfh cells. On further characterization, these cells exhibited upregulated gene expression suggesting an exhausted T cell phenotype (49). This could be the reason for the lack of T cell-mediated rejection observed in our model; however, the specific role and characteristics of this cell population require further elucidation.

Lastly, our data confirmed the benefit of additional gene editing in pig donors for xenotransplantation. The generation of the GGTA1KO (or GTKO) pigs was a breakthrough in xenotransplantation and has largely overcome the hyperacute rejection barrier in pig-to-primate (40, 41) as well as pig-to-human transplantation (31, 32). Here, kidneys from porcine donors carrying additional genetic edits (EGEN-2734, 3KO.7TG) were evaluated to determine whether the hurdles posed by allosensitization could be overcome. 3KO.7TG xenograft recipients showed prolonged graft survival over 1KO.1TG grafts, with less glomerulitis and peritubular capillaritis, two pathological findings reflective of active or chronic active AMR in allorejection (50), suggesting that the additional edits were indeed beneficial. The AMR score was also substantially reduced in 3KO.7TG xenografts at one-month post-transplantation. Although tissue deposition of complement split products, such as C4d, is often considered a diagnostic criterion for antibody-mediated injury (50, 51), C4d deposition may exist with no other evidence of AMR (21), and 3KO.7TG xenografts express human CD46 and CD55 that regulate complement downstream of C4 activation (52). We therefore evaluated C3d deposition in our xenokidney graft tissues and demonstrated that C3d deposition was decreased in 3KO.7TG xenograft recipients, highlighting the possible protective effects of the additional human transgenes.

NanoString data suggested that in 3KO.7TG xenograft recipients, multiple immune related genes, including those associated with B cell and innate cell activation, were downregulated compared to 1KO.1TG xenograft recipients, consistent with the milder rejection response seen clinically. T cell repopulation was reduced in 3KO.7TG recipients, which may reflect less effective initial depletion of these cells. However, the additional edits in 3KO.7TG xenografts could also have contributed to a reduced pro-inflammatory environment and less homeostatic T cell repopulation. Interestingly, the memory CD8⁺ T_EMRA subset (CD45RA⁺CCR7⁻) was not elevated at 1-month post-transplantation. Although a role for CD8⁺ T_EMRA T cells in xenorejection has not been described, these cells recently have been shown to identify kidney allograft recipients at high risk of rejection (53). In addition, the expansion of CD8⁺ T_EMRA cells may have contributed to AMR via FcgRIIIA (CD16) rather than TCR recognition (53). Following withdrawal of costimulation blockade, we observed a rapid increase in T lymphocytes, including CD8⁺ T_EMRA cells, and yet saw histological evidence of AMR and elevated B cell related gene expression within the xenografts, but not TCMR. Similarly, natural killer (NK) cells can mediate AMR via FcgRIIIA (CD16) engagement with bound antibody (54, 55). Loupy et al., recently demonstrated early innate cell infiltrations including NK cells in the xenograft in a decedent human model (56). Given the overlapping nature of CD8⁺ Temra and NK cells (57), further investigations will be required to better understand the precise mechanisms and components of xenograft rejection.

Despite the reports of improved xenograft survival in non-human primate models (5, 44, 58), outcomes remain heterogeneous. The inconsistent graft survival of individual recipients may reflect either the suboptimal conditions of therapeutic approaches or, alternatively, donor/recipient selection. However, the homogenous timing and frequency of AMR suggests that the main mode of rejection in xenotransplantation in this allosensitized model is antibody mediated. Anti-CD154 mAb, an agent that suppresses antibody responses in xenotransplantation and promotes long-term graft survival (40, 41), at the dose and frequency given, appeared to be relatively ineffective in controlling the post-xenotransplant humoral response in this highly allosensitized model when 1KO.1TG xenografts were transplanted. It cannot be ruled out that a higher dose or more frequent dosing of anti-CD154 mAb may have been more effective in this respect. Lastly, the lack of T cell-mediated rejection in this study, even without immunosuppression, highlights the importance of targeting the humoral response in order to achieve prolonged graft survival.

Limitations of the current study include that xenografts to unsensitized recipients were not included and hence a direct comparison to allosensitized recipients was not done. The different background SLA types of 1KO.1TG and 3KO.7TG xenografts require a caveat to ascribing the incremental benefit of 3KO.7TG being related solely to a change in transgene expression. Because host T cell and T cell-dependent B cell responses depend on SLA expression, it may, in fact, be the donor SLA type that impacts xenotransplant outcomes (38). All 3KO.7TG pigs were haplotype Lr-6.7, whereas 1KO.1TG pigs were either NIH haplotype c (Lr-3.3) or haplotype H03 (Lr-21.22). Given the high diversity of SLA and Macaca mulatta (MAMU) combinations, resulting in potential cross reactivity, and considering the relatively limited number of transplants performed, it is challenging to dissect whether the differences observed relate to a particular donor MHC type or the additional transgenic edits.

As xenotransplantation advances towards clinical trials (59, 60), highly allosensitized patients may derive great benefit from this donor source. Our data suggest that future investigations should aim at two areas. Firstly, genetic modifications of the donor need to be optimized and fully characterized. The true benefits of these added manipulations need to be evaluated, ideally in head-to-head comparisons such as those presented here (61, 62). Secondly, an immunosuppressive regimen to control post-xenotransplantation humoral responses without damaging protective immunity needs to be optimized. Although targeting the CD40/CD154 pathway has previously successfully prolonged xenograft survival in non-sensitized animals (63, 64), the humoral response in sensitized recipients treated with anti-CD154 mAb needs to be further managed.

Materials and Methods

Study design

This study aimed to test the hypothesis that allosensitization does not incur additional immunological risks for xenotransplant recipients by transplanting porcine kidneys into highly allosensitized NHPs. Further, we evaluated the effects of different genetic modifications of donor pigs on outcomes in these NHPs. All animal care, procedures, and surgeries were conducted in accordance with National Institutes of Health (NIH) guidelines and were approved by the Duke University Institutional Animal Care and Use Committee (Duke IACUC A033–20-02). Male rhesus macaques (n=10) were obtained from Alphagenesis, Inc. Six animals were allocated to receive xenokidney transplantation from reference pigs (1KO.1TG) from the National Swine Resource and Research Center (NSRRC; University of Missouri-Columbia) after allosensitization to assess overall impact of allosensitization to xenotransplantation outcome. Four allosensitized animals were later allocated to receive xenokidney transplantation from clinical grade pigs (3KO.7TG) from eGenesis Inc. to evaluate whether additional human transgenes with TKO could promote long-term xenograft survival.

Allosensitization of rhesus macaque prior to xenokidney transplantation

All animals received two sequential skin transplants from maximally MHC classes I and II mismatched pairs (Table S1) to sensitize each other as previously described (18, 65). Full-thickness skin grafts about 1-inch in diameter were exchanged between NHP pairs, and secured on dorsal skin with 3–0 or 4–0 nylon, adhesive bandage, and jackets for 7 days. Skin grafts were assessed on day 7. A second skin transplant was performed 8 weeks after first skin transplant. Allosensitization was confirmed with the presence of DSA measured by flow crossmatch.

Xenokidney transplantation in allosensitized NHPs

GGTA1KO.CD55 transgenic pigs (1KO.1TG, n=3) were obtained from the NSRRC. EGEN-2734 transgenic pigs (3KO.7TG, n=2) were generously provided by eGenesis. Six to eight weeks after second skin transplant, 10 NHPs underwent bilateral native nephrectomy followed by life-sustaining kidney transplantation. Bilateral nephrectomy and donor kidney transplantation into NHP recipients were performed using standard microvascular techniques. Briefly, the inferior vena cava (IVC) and distal aorta of the recipient were dissected to allow implantation of the porcine kidney in the lower abdomen. Prior to clamping the aorta and IVC, heparin was administered (100–200 units/kg) to prevent thrombotic complications. The xenograft was implanted using standard microvascular techniques to create an end-to-side anastomosis between the donor renal vein and the recipient vena cava with 7–0 polypropylene, followed by the anastomosis between the donor renal artery and the recipient distal aorta with 8–0 polypropylene. The xenokidney was then reperfused. The donor ureter was anastomosed to the bladder using either a tunneled ureteroneocystostomy technique or an extravesical ureteral implantation technique. An 8 cm 3.7 French ureteral stent was placed. Bilateral native nephrectomy was then completed. The abdominal cavity was inspected prior to closure to ensure hemostasis, absence of intestinal intussusception and a correct sponge and instrument count. The abdominal wall was then closed in two layers.

Induction, maintenance immunosuppression, and other treatments

All transplanted NHPs received induction therapy with depleting rhesus anti-CD4 mAb [CD4R1, PRID: AB_2716322, Nonhuman Primate Reagent Resources (NHPRR)] 50 mg/kg IV on post-operative days (POD) −3 and 0, with rhesus anti-CD8 mAb (MT807R1, PRID: AB_2716320, NHPRR,) 50 mg/kg IV on POD 0. Maintenance immunosuppression after xenotransplantation consisted of anti-CD154 mAb given at 20 mg/kg IV on days 0, 7, and 14, then biweekly, in addition to daily mycophenolate mofetil (Genentech), 15 mg/kg subcutaneously (SC) or 30 mg/kg orally twice daily and 125 mg methylprednisolone (Pfizer) was infused IV, then reduced to 0.5 mg/kg daily with intramuscular (IM) doses. Ganciclovir for rhesus CMV (rhCMV) reactivation was given at 6 mg/kg SQ daily for prophylaxis, 7.5 mg/kg twice daily for therapy (Fresenius Kabi). Suspected acute rejection episodes indicated by rise in sCr were treated with 125 mg methylprednisolone for three days.

Kidney xenograft monitoring and histology

Xenograft function was assessed by monitoring urine output daily and by serum chemistry tests at least twice per week. Protocol renal biopsies were performed one month after xenokidney transplantation. Percutaneous biopsies were performed with a 20G CareFusion Coaxial AchieveTM Automatic Biopsy System (BD Biosciences). Terminal kidney failure was defined as a prolonged rise in sCr>4 mg/dl and/or BUN>80 mg/dl at consecutive measures with anuria, at which point the animal was humanely euthanized. Lymph node and spleen biopsies were performed on all primates before transplant and at time of euthanasia. All tissue specimens were fixed in 10% neutral buffered formalin, paraffin embedded, sectioned and stained with hematoxylin and eosin (H&E) and periodic acid-schiff (PAS). In addition, C3d (clone: E28-P, Abcam), C4d (polyclonal, American Research Products), and CD3 (polyclonal, Agilent Technologies) immunohistochemistry was performed for selected biopsy samples and all necropsy specimens. An experienced transplant pathologist (ABF) evaluated and scored the histology specimens according to the Banff criteria (66) in a blinded fashion.

Viral monitoring and treatments

All animals received 6 mg/kg ganciclovir daily SC for CMV prophylaxis starting the day of transplantation. RhCMV viral titers were monitored weekly from whole blood by real time PCR as described previously (22). For rhCMV reactivation (>10,000 copies/ml), increased dose and frequency of ganciclovir (up to 7.5 mg/kg, twice daily) were given. Simian parvo virus (SPV) was tested from serum samples with PCR by VRL Diagnostic. For post-transplant anemia, we treated with epoetin alfa (Amgen) and injectable irons (10 mg/kg). When animals showed hemoglobin lower than 6 mg/ml, we provided blood transfusion from blood donor animal with compatible blood type. One unit (50cc) of blood was given per transfusion.

Blood coagulation studies and thromboelastography

Citrated blood was collected at 0h, 24h, and 4d after xenokidney transplantation as well as at euthanasia. Platelet count, fibrinogen, prothrombin time (PT), and partial prothrombin time (PTT) were measured as part of a blood coagulation panel (Antech Diagnostics).

Immune monitoring with flow cytometry

Lymph nodes were obtained from biopsy of axillary and/or inguinal lymph nodes (LNs) before kidney transplantation (and T cell depletion), one month after transplantation, and during necropsy. Spleen was also obtained during the kidney transplantation using 20G CareFusion Coaxial AchieveTM Automatic Biopsy System (BD Biosciences) and at necropsy. Blood was serially collected after transplantation. Single-cell suspensions of LN, spleen, and PBMCs were prepared and stained with Fixable Blue viability dye (Thermo Fisher Scientific), according to the manufacturer’s recommendations. Cells were washed with 2% FBS in PBS and stained with the following fluorochrome-conjugated mAbs against human: CCR7, CD3, CD4, CD8, CD19, CD20, CD27, CD28, CD38, CD45RA, CD95, CD185 (CXCR5), CD278 [inducible T-cell co-stimulator (ICOS)], CD279 (PD-1), IgD, IgG, and IgM (Table S7). Cells were fixed and permeabilized using eBioscience Foxp3/Transcription Factor Staining Buffer Set (Thermo Fisher Scientific) and stained with fluorochrome-conjugated mAbs against human BCL-6. Samples were washed and fixed with 1x stabilizing fixative solution (BD Biosciences). Flow cytometry was performed on a BD LSRFortessa or BD Symphony and analyzed using FlowJo software version 9 or 10 (Tree Star).

Monitoring antibodies against allo-donor, xeno-donor and HLA

Recipients’ serum samples from various timepoints throughout the study period were incubated with 3 ×10⁵ PBMCs from skin donor (for alloantibody) or from kidney donor (for xenoantibody). 1/50 diluted serum was used for alloantibody while neat serum was used for xenoantibody. After incubation, PBMCs were washed and serum antibody (IgG or IgM) bound to the allogenic NHP PBMCs was detected by subsequent staining with FITC-labeled anti-monkey IgG or IgM (both KPL), PE-labeled anti-CD20 (2H7), and PerCP-Cy5.5-labeled anti-CD3 (SP34–2) (both BD Biosciences). For xenoanitbody, serum antibody (IgG or IgM) bound to the xenogenic pig PBMCs was detected by subsequent staining with FITC-labelled anti-monkey IgG or IgM (KPL), anti-swine CD3e mAb (BB23–8E6–8C8, BD), anti-human CD21 mAb (B-ly4, BD) and Live/Dead Fixable Blue (Thermo Fisher Scientific) staining. MFI of anti-monkey and anti-pig IgG/IgM against T or B cells was measured on BD LSRFortessa and analyzed using FlowJo software version 9 or 10.0.

ELISpot assay for measuring antibody secreting cells

The frequency of total IgG secreting cells (or antibody-secreting cells; ASC) was measured by ELISpot assay as previously described (65). Briefly, 96-well ELISpot plates (MultiScreen^® 8 Well Strip Assay Plate, Merck Millipore) were prepared (prewetted with 70% ethanol for 2 minutes, coated with anti-IgG capture antibody for 24 hours at 4 °C, and blocked with R10 medium for 2 hours at room temperature with extensive washing between each step – 3 time with PBS-Tween20 and 3 times with PBS). RPMI suspended cells from bone marrow, lymph node, and blood were plated (2×10⁵, 1×10⁵, 5×10⁴, 2.5×10⁴, 1.25×10⁵, and 6.25×10³ cells per well in duplicates) in ELISpot plates for 24 h in a 5% CO₂ incubator at 37°C. After washing, ASC was visualized with anti-huIgG-biotin (Thermo Fisher Scientific) in 4 °C overnight and avidin-D-HRP (Vector Laboratories) for 1 hour in room temperature and developed with AEC substrate (Sigma-Aldrich). The number of spots were counted using the CTL immunospot reader (Cellular Technologies Ltd) and reported as spots per million cells.

Gene expression analysis using the NanoString Platform

We isolated total RNA from kidney xenograft FFPE blocks and measured gene expression by using the NanoString nCounter MAX platform (NanoString Technologies) as previously reported (21). We used the nCounter Human Organ Transplantation Panel (No. LBL-10743–01) to measure 758 genes covering the core pathways and process surrounding host response and rejection of transplanted tissues (including 12 internal reference genes for data normalization). Gene expression data were log2-transformed, background subtracted, and normalized to the geometric mean expression of 12 housekeeping genes by using the Rosalind (version 3.35.x) platform.

Statistical analysis

All statistical analyses were performed using Prism 9.0 (Graphpad Software). Values of P<0.05 were considered to be significant. Post-transplant survival times were plotted and compared using Kaplan-Meier survival curves and log-rank tests. Normally distributed data were evaluated using a two-tailed paired or unpaired t test. Data with more than two groups were evaluated using one-way analysis of variance (ANOVA) with the Geisser-Greenhouse correction.

Data and materials availability

All data associated with this study are present in the paper or the supplementary materials. Anti-CD4, CD8, and CD154 mAbs used in this study were provided by the NIH Nonhuman Primate Reagent Resource (R24 OD010976, U24 AI126683). Donor GGTA1KO.hCD55 transgenic pigs were obtained from the National Swine Resource and Research Center (U42 OD011140) and EGEN-2734 pigs were provided by eGenesis. NanoString data has been deposited in the Gene Expression Omnibus (GEO) under accession number 24655292.