Creating class I MHC-null pigs using guide RNA and the Cas9 endonuclease.

Pigs are emerging as important large animal models for biomedical research, and they may represent a source of organs for xenotransplantation. The MHC is pivotal to the function of the immune system in health and disease, and it is particularly important in infection and transplant rejection. Pigs deficient in class I MHC could serve as important reagents to study viral immunity as well as allograft and xenograft rejection. In this study, we report the creation and characterization of class I MHC knockout pigs using the Cas9 nuclease and guide RNAs. Pig fetal fibroblasts were genetically engineered using Cas9 and guide RNAs, and class I MHC(-) cells were then used as nuclear donors for somatic cell nuclear transfer. We produced three piglets devoid of all cell surface class I proteins. Although these animals have reduced levels of CD4(-)CD8(+) T cells in peripheral blood, the pigs appear healthy and are developing normally. These pigs are a promising reagent for immunological research.

Reduced human platelet uptake by pig livers deficient in the asialoglycoprotein receptor 1 protein.

BACKGROUND: The lethal thrombocytopenia that accompanies liver xenotransplantation is a barrier to clinical application. Human platelets are bound by the asialoglycoprotein receptor (ASGR) on pig sinusoidal endothelial cells and phagocytosed. Inactivation of the ASGR1 gene in donor pigs may prevent xenotransplantation-induced thrombocytopenia. METHODS: Transcription activator-like effector nucleases (TALENs) were targeted to the ASGR1 gene in pig liver-derived cells. ASGR1 deficient pig cells were used for somatic cell nuclear transfer (SCNT). ASGR1 knock out (ASGR1-/-) fetal fibroblasts were used to produce healthy ASGR1 knock out piglets. Human platelet uptake was measured in ASGR1+/+ and ASGR1-/- livers. RESULTS: Targeted disruption of the ASGR1 gene with TALENs eliminated expression of the receptor. ASGR1-/- livers phagocytosed fewer human platelets than domestic porcine livers during perfusion. CONCLUSIONS: The use of TALENs in liver-derived cells followed by SCNT enabled the production of healthy homozygous ASGR1 knock out pigs. Livers from ASGR1-/- pigs exhibit decreased human platelet uptake. Deletion of the ASGR1 gene is a viable strategy to diminish platelet destruction in pig-to-human xenotransplantation.

Double knockout pigs deficient in N-glycolylneuraminic acid and galactose α-1,3-galactose reduce the humoral barrier to xenotransplantation.

BACKGROUND: Clinical xenotransplantation is not possible because humans possess antibodies that recognize antigens on the surface of pig cells. Galα-1,3-Gal (Gal) and N-glycolylneuraminic acid (Neu5Gc) are two known xenoantigens. METHODS: We report the homozygous disruption of the α1, 3-galactosyltransferase (GGTA1) and the cytidine monophosphate-N-acetylneuraminic acid hydroxylase (CMAH) genes in liver-derived female pig cells using zinc-finger nucleases (ZFNs). Somatic cell nuclear transfer (SCNT) was used to produce healthy cloned piglets from the genetically modified liver cells. Antibody-binding and antibody-mediated complement-dependent cytotoxicity assays were used to examine the immunoreactivity of pig cells deficient in Neu5Gc and Gal. RESULTS: This approach enabled rapid production of a pig strain deficient in multiple genes without extensive breeding protocols. Immune recognition studies showed that pigs lacking both CMAH and GGTA1 gene activities reduce the humoral barrier to xenotransplantation, further than pigs lacking only GGTA1. CONCLUSIONS: This technology will accelerate the development of pigs for xenotransplantation research.

Efficient selection of Gal-knockout pig cells for somatic cell nuclear transfer.

BACKGROUND: The process of selecting transgenic cells has been one of the bottlenecks in the generation of transgenic animals by somatic cell nuclear transfer (SCNT). In particular, selection for the Gal double-knockout (Gal-DKO) genotype has been time consuming and inefficient. The objective of this work was to generate a highly efficient system to select Gal-DKO cells to be used in SCNT without affecting the efficiency in production of Gal-null pigs. MATERIALS AND METHODS: Fetal liver-derived cells deficient in Gal-expression were initially selected by fluorescence-activated cell sorting (FACS) using IB4 conjugated to a fluorescent dye. Cells recovered by FACS were cultured and expanded, followed by a second round of selection using streptavidin magnetic beads and IB4 lectin biotin. RESULTS: Recovery efficiency of target cells was 0.04% for the first selection using FACS and 0.3% for the second round by magnetic beads. Full reprogramming was obtained on selected Gal-DKO cells after FACS and magnetic beads selection, when used for SCNT to produce the Gal-null piglets. Cells obtained from magnetic beads developed 48 colonies; the Gal-null genotype was found in 44 of them (91.7%). Three of these colonies were used to generate piglets by SCNT. From three recipients receiving embryos, two became pregnant and produced 17 piglets, all of them DKO. CONCLUSIONS: Sequential selection of Gal-DKO cells by FACS/magnetic beads is a highly efficient system to generate null cells. Selected cells were successfully used to generate healthy double-knockout piglets by SCNT.

Fibronectin from alpha 1,3-galactosyltransferase knockout pigs is a xenoantigen.

BACKGROUND: Antibody-mediated rejection continues to be an obstacle for xenotransplantation despite development of α1,3-galactosyltransferase knockout (GTKO) pigs. Fibronectin (Fn) from GTKO pigs was identified as a xenoantigen in baboons. N-glycolylneuraminic acid (Neu5Gc), similar to galactose α1,3-galactose, is an antigenic carbohydrate found in pigs. We evaluated human antibody reactivity and performed initial antigenic epitope characterization of Fn from GTKO pigs. MATERIALS AND METHODS: GTKO pig aortic endothelial cells (AEC) were isolated and assessed for antibody-mediated complement-dependent cytotoxicity (CDC). Human and GTKO pig Fn were purified and analyzed using immunoblots. GTKO pig and human AEC absorbed human sera were assessed for CDC and anti-GTKO pig Fn antibodies. GTKO pig proteins were assessed for Neu5Gc. Immunoaffinity-purified human IgG anti-GTKO pig (hIgG-GTKOp) Fn using a GTKO pig Fn column were evaluated for cross-reactivity with other proteins. RESULTS: GTKO pig AEC had greater human antibody binding, complement deposition and CDC compared with allogeneic human AEC. Human sera absorbed with GTKO pig AEC resulted in diminished anti-GTKO pig Fn antibody. Neu5Gc was identified on GTKO pig Fn and other proteins. The hIgG-GTKOp Fn cross-reacted with multiple GTKO pig proteins and was enriched with anti-Neu5Gc antibody. CONCLUSIONS: Removal of antigenic epitopes from GTKO pig AEC would improve xenograft compatibility. GTKO pig Fn has antigenic epitopes, one identified as Neu5Gc, which may be responsible for pathology and cross-reactivity of hIgG-GTKOp Fn. Genetic knockout of Neu5Gc appears necessary to address significance and identification of non-Neu5Gc GTKO pig Fn antigenic epitopes.

Differences in human and porcine platelet oligosaccharides may influence phagocytosis by liver sinusoidal cells in vitro.

BACKGROUND: Acute thrombocytopenia was revealed as a limiting factor to porcine liver xenotransplantation from in vitro and in vivo studies using porcine liver in human and baboon transplant models. The asialoglycoprotein receptor 1 (ASGR1) on liver sinusoidal endothelial cells (LSEC) and macrophage antigen complex-1 (Mac-1) on Kupffer cells (KC) mediate platelet phagocytosis and have carbohydrate-binding sites that recognize galactose and N-acetyl glucosamine in the beta conformation. Analysis of these receptor carbohydrate-binding domains and surface carbohydrates on human and porcine platelets may shed light on the mechanism of xenotransplantation-induced thrombocytopenia. METHODS: An amino acid sequence comparison of human and porcine ASGR1 lectin-binding domains was performed. Using fluorescent labeled-lectins, human platelets, domestic and α1,3 galactosyltransferase knockout/human decay accelerating factor, porcine platelets were characterized by flow cytometry and lectin blot analyses. After desialylation, human and porcine platelets were examined by flow cytometry to determine whether sialic acid capping of galactose and N-acetyl glucosamine oligosaccharides in the beta conformation was a factor. Further, desialylated human platelets were studied on primary porcine liver sinusoidal cells with regard to binding and phagocytosis. RESULTS: Human platelets have four times more exposed galactose ß1-4 N-acetyl glucosamine (Galß) and N-acetyl glucosamine ß1-4 N-acetyl glucosamine (ßGlcNAc) than fresh porcine platelets. Galß and ßGlcNAc moieties on porcine platelets were not masked by sialic acid. Removal of sialic acid from human platelets increased binding and phagocytosis by LSEC and KC. CONCLUSIONS: Differences between human and porcine ASGR1 and Mac-1, in combination with a significantly higher number of galactose and N-acetyl glucosamine-containing oligosaccharides on human platelets contribute, in part, to platelet loss seen in porcine liver xenotransplantation.

ASGR1 expressed by porcine enriched liver sinusoidal endothelial cells mediates human platelet phagocytosis in vitro.

BACKGROUND: Porcine liver xenografts represent a potential solution to the organ shortage, but thrombocytopenia occurs within minutes to hours after xenotransplantation, preventing clinical application. Recently, it was discovered that porcine liver sinusoidal endothelial cells (LSEC) bind and phagocytose human platelets. We examined the role of ASGR1 in binding and removing human platelets by the pig liver endothelium. METHODS: Primary porcine enriched LSEC (eLSEC) were characterized by flow cytometry, immunoblot, quantitative PCR, and immunohistochemistry using confocal microscopy. Phagocytosis inhibition assays using anti-ASGR1 and an ASGR1 substrate were performed. ASGR1 was targeted for siRNA knockdown, and ASGR1-reduced cells were tested for human platelet binding and phagocytosis. RESULTS: ASGR1 is expressed by eLSEC. Human platelet binding and phagocytosis by porcine eLSEC was inhibited by asialofetuin, but not fetuin, suggesting an interaction with galactose ß1-4 N-acetyl glucosamine. Anti-ASGR1 antibodies inhibited human platelet binding in a dose-dependent manner. Knockdown experiments using siRNA reduced ASGR1 expression in asynchronous primary eLSEC by 40%-80%. There was a 20% reduction in translated protein significantly correlated with a 21% decrease in human platelet binding. CONCLUSIONS: ASGR1 on porcine eLSEC mediates phagocytosis of xenogeneic platelets.

Gene targeting and cloning in pigs using fetal liver derived cells.

BACKGROUND: Since there are no pig embryonic stem cells, pig genetic engineering is done in fetal fibroblasts that remain totipotent for only 3 to 5 wk. Nuclear donor cells that remain totipotent for longer periods of time would facilitate complicated genetic engineering in pigs. The goal of this study was to test the feasibility of using fetal liver-derived cells (FLDC) to perform gene targeting, and create a genetic knockout pig. MATERIALS AND METHODS: FLDC were isolated and processed using a human liver stem cell protocol. Single copy α-1,3-galactosyl transferase knockout (GTKO) FLDCs were created using electroporation and neomycin resistant colonies were screened using PCR. Homozygous GTKO cells were created through loss of heterozygosity mutations in single GTKO FLDCs. Double GTKO FLDCs were used in somatic cell nuclear transfer (SCNT) to create GTKO pigs. RESULTS: FLDCs grew for more than 80 population doublings, maintaining normal karyotype. Gene targeting and loss of heterozygosity mutations produced homozygous GTKO FLDCs. FLDCs used in SCNT gave rise to homozygous GTKO pigs. CONCLUSIONS: FDLCs can be used in gene targeting and SCNT to produce genetically modified pigs. The increased life span in culture compared to fetal fibroblasts may facilitate genetic engineering in the pig.

The fate of human platelets perfused through the pig liver: implications for xenotransplantation.

BACKGROUND: Pig liver xenotransplantation could offset the shortage of livers available for orthotopic liver transplantation. Studies in pig and baboon liver xenografts revealed the main obstacle to be a lethal thrombocytopenia that occurred within minutes to hours of transplantation. METHODS: We have created a model of xenotransplantation-induced thrombocytopenia using ex vivo pig liver perfusion with human platelets. Thrombocytopenia was examined using fluorescently labeled platelets during the ex vivo perfusion and coculture with primary liver sinusoidal endothelial cells (LSEC). RESULTS: Ex vivo liver perfusion revealed that 93% of human platelets were removed from circulation after 15 min. Endothelial cells and platelets were not activated based on tissue factor release into the perfusate. Biopsies from the ex vivo perfusion at 15 and 30 min and in vitro analysis indicated that human platelets are phagocytosed by pig LSEC and degraded in phagosomes. Sixty to 120 min after the addition of platelets to the ex vivo perfusion system, we observed platelet fragments and degraded platelets in hepatocytes. Platelet phagocytosis was not mediated by opsonization as Fc blocking had no effect on platelet phagocytosis. In vitro uptake of human platelets by primary LSEC cultures peaked at 15 min followed by a greater than 55% decrease in platelet fluorescence after 3 h. Primary pig LSEC phagosomes containing human platelets were colocalized with lysosomes positive for lysosome-associated membrane protein-1 (LAMP1), indicating the formation of mature phagosomes within pig LSEC. CONCLUSIONS: Our observation of pig LSEC phagocytosis of human platelets describes a novel mechanism of large-particle uptake in the liver. The creation of a model system to study xenotransplantation-induced thrombocytopenia makes possible the investigation into mechanisms that mediate platelet loss.

HLA Class I-sensitized Renal Transplant Patients Have Antibody Binding to SLA Class I Epitopes.

BACKGROUND: Highly sensitized patients are difficult to match with suitable renal allograft donors and may benefit from xenotransplant trials. We evaluate antibody binding from sensitized patients to pig cells and engineered single allele cells to identify anti-human leukocyte antigen (HLA) antibody cross-species reactivity with swine leukocyte antigen (SLA). These novel testing strategies assess HLA/SLA epitopes and antibody-binding patterns and introduce genetic engineering of SLA epitopes. METHODS: Sensitized patient sera were grouped by calculated panel reactive antibody and luminex single antigen reactivity profile and were tested with cloned GGTA1/CMAH/B4GalNT2 glycan knockout porcine cells. Pig reactivity was assessed by direct flow cytometric crossmatch and studied following elution from pig cells. To study the antigenicity of individual class I HLA and SLA alleles in cells, irrelevant sera binding to lymphoblastoid cells were minimized by CRISPR/Cas9 elimination of endogenous class I and class II HLA, B-cell receptor, and Fc receptor genes. Native HLA, SLA, and mutants of these proteins after mutating 144K to Q were assessed for antibody binding. RESULTS: Those with predominately anti-HLA-B&C antibodies, including Bw6 and Bw4 sensitization, frequently have low pig reactivity. Conversely, antibodies eluted from porcine cells are more commonly anti-HLA-A. Single HLA/SLA expressing engineered cells shows variable antigenicity and mutation of 144K to Q reduces antibody binding for some sensitized patients. CONCLUSIONS: Anti-HLA antibodies cross-react with SLA class I in predictable patterns, which can be identified with histocompatibility strategies, and SLA class I is a possible target of genetic engineering.

Eliminating Xenoantigen Expression on Swine RBC.

BACKGROUND: The rapidly improving tools of genetic engineering may make it possible to overcome the humoral immune barrier that prevents xenotransplantation. We hypothesize that levels of human antibody binding to donor tissues from swine must approximate the antibody binding occurring in allotransplantation. It is uncertain if this is an attainable goal. Here we perform an initial analysis of this issue by comparing human antibody binding to red blood cells (RBC) isolated from knockout swine and to allogeneic or autologous human RBC. METHODS: Human sera were incubated with RBC isolated from various genetically engineered swine or from humans. The level of IgG and IgM binding to these cells were compared using either flow cytometry or a novel mass spectrometric assay. RESULTS: Mass spectroscopic quantitation of human antibody binding demonstrated that as few as 3 gene inactivations can reduce the levels human antibody binding to swine RBC that is as low as autologous human RBC. Flow cytometry showed that RBC from 2-gene knockout swine exhibited less human antibody binding than human blood group O allogeneic RBC in 22% of tested sera. Deletion of a third gene from pigs resulted in 30% of human samples having less IgG and IgM RBC xenoreactivity than alloreactivity. CONCLUSIONS: Xenoantigenicity of swine RBC can be eliminated via gene disruption. These results suggest that the gene knockout approach may be able reduce antigenicity in other pig tissues to levels that enable the xenotransplantation humoral barrier to be overcome.

Humoral Reactivity of Renal Transplant-Waitlisted Patients to Cells From GGTA1/CMAH/B4GalNT2, and SLA Class I Knockout Pigs.

BACKGROUND: Antipig antibodies are a barrier to clinical xenotransplantation. We evaluated antibody binding of waitlisted renal transplant patients to 3 glycan knockout (KO) pig cells and class I swine leukocyte antigens (SLA). METHODS: Peripheral blood mononuclear cells from SLA identical wild type (WT), α1, 3-galactosyltransferase (GGTA1) KO, GGTA1/ cytidine monophosphate-N-acetylneuraminic acid hydroxylase (CMAH) KO, and GGTA1/ CMAH /b1,4 N-acetylgalactosaminyl transferase (B4GalNT2) KO pigs were screened for human antibody binding using flow cytometric crossmatch (FCXM). Sera from 820 patients were screened on GGTA1/CMAH/B4GalNT2 KO cells and a subset with elevated binding was evaluated further. FCXM was performed on SLA intact cells and GGTA1/SLA class I KO cells after depletion with WT pig RBCs to remove cell surface reactive antibodies, but leave SLA antibodies. Lastly, human and pig reactive antibodies were eluted and tested for cross-species binding and reactivity to single-antigen HLA beads. RESULTS: Sequential glycan KO modifications significantly reduce antibody binding of waitlisted patients. Sera exhibiting elevated binding without reduction after depletion with WT RBCs demonstrate reduced binding to SLA class I KO cells. Human IgG, eluted from human and pig peripheral blood mononuclear cells, interacted across species and bound single-antigen HLA beads in common epitope-restricted patterns. CONCLUSIONS: Many waitlisted patients have minimal xenoreactive antibody binding to the triple KO pig, but some HLA antibodies in sensitized patients cross-react with class I SLA. SLA class I is a target for genome editing in xenotransplantation.

Silencing Porcine CMAH and GGTA1 Genes Significantly Reduces Xenogeneic Consumption of Human Platelets by Porcine Livers.

BACKGROUND: A profound thrombocytopenia limits hepatic xenotransplantation in the pig-to-primate model. Porcine livers also have shown the ability to phagocytose human platelets in the absence of immune-mediated injury. Recently, inactivation of the porcine ASGR1 gene has been shown to decrease this phenomenon. Inactivating GGTA1 and CMAH genes has reduced the antibody-mediated barrier to xenotransplantation; herein, we describe the effect that these modifications have on xenogeneic consumption of human platelets in the absence of immune-mediated graft injury. METHODS: Wild type (WT), ASGR1, GGTA1, and GGTA1CMAH knockout pigs were compared for their xenogeneic hepatic consumption of human platelets. An in vitro assay was established to measure the association of human platelets with liver sinusoidal endothelial cells (LSECs) by immunohistochemistry. Perfusion models were used to measure human platelet uptake in livers from WT, ASGR1, GGTA1, and GGTA1 CMAH pigs. RESULTS: GGTA1, CMAH LSECs exhibited reduced levels of human platelet binding in vitro when compared with GGTA1 and WT LSECs. In a continuous perfusion model, GGTA1 CMAH livers consumed fewer human platelets than GGTA1 and WT livers. GGTA1 CMAH livers also consumed fewer human platelets than ASGR1 livers in a single-pass model. CONCLUSIONS: Silencing the porcine carbohydrate genes necessary to avoid antibody-mediated rejection in a pig-to-human model also reduces the xenogeneic consumption of human platelets by the porcine liver. The combination of these genetic modifications may be an effective strategy to limit the thrombocytopenia associated with pig-to-human hepatic xenotransplantation.

Removal of therapeutic anti-lymphocyte antibodies from human sera prior to anti-human leukocyte antibody testing.

Both monoclonal (e.g. Orthoclone (OKT3), rituximab) and polyclonal (e.g. ATGAM, Thymoglobulin (Thymo)) anti-lymphocyte Abs (ALAs) are used extensively in organ transplantation for immunosuppression induction, desensitization, and treatment of acute rejection. ALAs often interfere with post transplant immunologic monitoring. We describe a method that uses magnetic beads to selectively remove ALAs from patient serum. Rabbit anti-mouse Fc-specific (180 mug), or rabbit anti-mouse Fab-specific (180 microg), or rabbit anti-horse heavy and light chain-specific and rabbit anti-horse F(ab')2 (200 microg) (Jackson Immunoresearch) was adsorbed to 6.7 x 10(8) Dynabeads M-280 conjugated with sheep anti-rabbit IgG (Dynal Biotech). Fifty microliters of normal human serum (NHS) with 2 microg/ml of OKT3 or 100 microg/ml ATGAM, Thymo, or rituximab were incubated with conjugated beads for several incubations. NHS containing ALAs before and after treatment by the protocol were incubated with human lymphocytes and labeled with FITC-antibody to immunoglobulin of the species used to produce the particular ALA. Residual ALA was determined using flow cytometry. Average median channel for serum with or without ALA was 11.1 and 0.120, respectively for OKT3; 64.4 and 0.344 for ATGAM; 108.5 and 0.200 for Thymo; and 1022.5 and 11.4 for rituximab. Treatment lowered the median channel for serum with OKT3 to 0.103, 0.309 for ATGAM, 0.199 for Thymo, and 12.1 for rituximab. ALAs can be effectively removed from serum by the use of magnetic beads conjugated with Ab specific for ALA thereby permitting immunologic monitoring without interference.

Rituximab for reduction of anti-HLA antibodies in patients awaiting renal transplantation: 1. Safety, pharmacodynamics, and pharmacokinetics.

BACKGROUND: Preformed HLA antibodies (Ab), reported as panel-reactive antibody (PRA), prolong patient waiting time for kidney transplantation. We hypothesized that rituximab (RTX) could reduce PRA via B-cell depletion. This initial study reports the safety, pharmacokinetics, and pharmacodynamics of RTX in patients with end-stage renal failure. METHODS: The study was an investigator-initiated single-dose, dose-escalation phase I trial of RTX in chronic dialysis patients (PRA >50%). It was approved by the Institutional Review Board and the Food and Drug Administration. Nine subjects were treated with a single dose of RTX (n=3 per group) at 50, 150, or 375 mg/m. Peripheral lymphocyte cell surface markers and HLA Ab levels (%PRA and titers) were tested using flow cytometry. RESULTS: There were four significant adverse events: a suspected histoplasmosis infection; two Tenchkoff dialysis catheter infections; and fever (38.7 degrees C) during infusion. At 2 days after RTX therapy, there was depletion of CD19 cells (pre-RTX 181+/-137 vs. post-RTX 12+/-5.6, P =0.006). In 2 (22%) of 9 subjects, there was no appreciable change in PRA. Among the other seven patients, one had a decrease in PRA from 87% to 51% with a concurrent decrease in fluorescence intensity; five patients had changes in histogram architecture suggesting loss of antibody specificity; and one patient had a fourfold decrease in PRA titer from 1:64 to 1:16 at 6 months after treatment. In addition, one of the seven patients converted a donor-specific crossmatch to negative and underwent a successful living donor kidney transplantation. CONCLUSIONS: RTX can be safely administered and may be an effective agent to reduce high-titer anti-HLA Abs in subjects awaiting kidney transplantation.

Pronase treatment facilitates alloantibody flow cytometric and cytotoxic crossmatching in the presence of rituximab.

Rituximab (RIT), a murine/human chimeric monoclonal antibody directed against human CD20 is under investigation for its role in transplantation. RIT causes B-cell crossmatches to appear positive. Pronase, a proteolytic enzyme that targets F(c) receptors removes CD20 from B cells. After CD20 is removed, RIT should not bind, making it possible to detect class I or class II antibodies on treated B cells. In this study, we incubated RIT with normal human serum (NHS, negative control) or pooled sera from highly sensitized (>50% panel reactive antibody, HLA+) subjects awaiting renal transplantation (positive control) and then performed B-cell flow cytometric crossmatches using untreated or pronase treated B cells as targets. We observed that untreated B cells incubated with RIT-spiked NHS displayed a significant increase in surface fluorescence compared with NHS without RIT, similar to the fluorescence that occurs with a positive crossmatch. In contrast, when CD20 was cleaved from the B cells with pronase, B cells displayed a negative crossmatch with the RIT-spiked NHS. In addition, there was no change in the crossmatches of pooled high panel reactive antibody (PRA) sera after pronase treatment. RIT could be used without worry about losing the ability to perform transplant immunologic monitoring.

Analysis of Ki-67 antigen expression, DNA proliferative fraction, and survival in resected cancer of the pancreas.

BACKGROUND: Prognostic markers for pancreas cancer, such as CEA, CA19-9, ploidy analysis, and S-phase determination using flow cytometry, have not been consistently predictive. We chose to evaluate nuclear proliferation, as measured by the MIB-1 monoclonal antibody and digital image analysis, as a prognostic marker in pancreatic carcinoma, and compare the findings with DNA ploidy and S-phase analysis. MIB-1 identifies the Ki67 antigen present in nuclei of cells in all phases of the cell cycle except G0. METHODS: We retrospectively reviewed 33 patients with pancreatic adenocarcinoma resected for cure between 1989 and 1994 with available fixed tissue. Sectioned tissue was stained with MIB-1, and the number of positively stained nuclei determined and expressed as a MIB-1 labeling index (LI) by quantitative image analysis. Disaggregated nuclei were analyzed by flow cytometry using standard techniques. RESULTS: MIB-1 LI for pancreas cancers was heterogeneous within and between cancers. The MIB-1 LI for the cancers was 28 +/- 15 (median 29). There was no correlation between survival and MIB-1 expression (R(2) = 0.03). Likewise, there was no correlation between MIB-1 LI and percentage of cells in S-phase, G(2)/M, or total proliferating cells (S+G(2)/M; R(2) = 0.01), nor was there a difference between MIB-1 LI and ploidy (P = 0.88). CONCLUSIONS: We conclude that in our patient population, nuclear proliferation in pancreatic cancer, as determined by expression of Ki67 nuclear antigen, does not appear to correlate with survival and is not a useful prognostic marker. Despite intuitive thoughts to the contrary, there is no correlation between cell cycle analysis as determined by flow cytometry and Ki67 expression in pancreas cancer. Current methods of assessing prognosis after curative resection of cancer of the pancreas, including lymph node and margin status, tumor size, and possibly DNA ploidy as determined by flow cytometry, are not augmented by the assessment of nuclear proliferation by image analysis using the MIB-1 monoclonal antibody.

In vivo human B-cell subset recovery after in vivo depletion with rituximab, anti-human CD20 monoclonal antibody.

Rituximab, chimeric anti-human CD20 monoclonal antibody approved for B-cell lymphoma, depletes circulating B cells. Little data exist on its use for nonmalignant diseases or B-cell subset recovery after treatment. We hypothesized that rituximab may reduce panel reactive alloantibodies (PRA) in dialysis patients awaiting renal transplantation and performed a single-dose, dose-escalation phase 1 trial. Here we report changes in lymphocyte phenotypes in subjects treated with rituximab alone. Nine subjects, 44 +/- 10 years(Yr); (5 F, 4 M) received one dose (n=3) at 50, 150, or 375 mg/m(2) without other immunosuppression. Blood was collected before dosing and intervals thereafter. No significant changes in leukocytes, total or CD3(+) lymphocytes were noted. In all, there was CD19(+) depletion by day 2(D2) (12.0 +/- 5.6 cells/mm(3) vs. 181 +/- 137, D0; p<0.01) and CD20(+) (11.0 +/- 12.0 vs. 205 +/- 116, D0; p<0.01). At 6 months (mo), CD19(+) and CD20(+) remained low (51.1 +/- 42.2, p<0.05; 75.4 +/- 38.7, p<0.05, respectively) compared to D0. CD19(+)CD5(+) cells recovered more rapidly, returning to baseline by 6mo while B memory cells (CD19(+)CD27(+)) remained low (32.3 +/- 29.0) at 1Yr (7.5 +/- 4.5; p<0.001) and 2Yr (12.1 +/- 7.9; p<0.001) after treatment. We conclude that single dose rituximab ablates B cells in high PRA dialysis patients awaiting transplantation. B-cell ablation, particularly memory B cells, was long-lasting, lagging repopulation by CD5(+) B cells.

Primary porcine Kupffer cell phagocytosis of human platelets involves the CD18 receptor.

BACKGROUND: Hepatic failure has been treated successfully with clinical extracorporeal perfusions of porcine livers. However, dog-to-pig and pig-to-baboon liver xenotransplant models have resulted in severe bleeding secondary to liver xenograft-induced thrombocytopenia. Kupffer cells (KC) are abundant phagocytic cells in the liver. KC express the CD11b/CD18 receptor, which has been implicated in chilled platelet binding and phagocytosis through interaction with platelet surface proteins and carbohydrates. We sought to identify the role of KC CD18 in liver xenograft-induced thrombocytopenia. METHODS: Primary pig KC were characterized by flow cytometry, immunoblots, and quantitative polymerase chain reaction. Pig KC were used in inhibition assays with fluorescently labeled human platelets. The CD18 receptor was targeted for siRNA knockdown. RESULTS: Domestic and α1,3-galactosyltransferase double knockout porcine KC cultures were approximately 92% positive for CD18 as detected by quantitative polymerase chain reaction and flow cytometry. Use of CD18 blocking antibodies resulted in reduction of human platelet binding and phagocytosis. Additionally, asialofetuin, not fetuin, inhibited platelet phagocytosis suggesting the involvement of an oligosaccharide-binding site. Furthermore, reduced CD18 expression by siRNA resulted in decreased human platelet binding. CONCLUSIONS: Our data suggest that primary pig KC bind and phagocytose human platelets with involvement of CD18. Further understanding and modification of CD18 expression in pigs may result in a liver xenograft with reduced thrombocytopenic effects, which could be used as a bridge to allogeneic liver transplantation.