Prevention of acute liver failure in rats with reversibly immortalized human hepatocytes.

Because of a critical shortage in suitable organs, many patients with terminal liver disease die each year before liver transplantation can be performed. Transplantation of isolated hepatocytes has been proposed for the temporary metabolic support of patients awaiting liver transplantation or spontaneous reversion of their liver disease. A major limitation of this form of therapy is the present inability to isolate an adequate number of transplantable hepatocytes. A highly differentiated cell line, NKNT-3, was generated by retroviral transfer in normal primary adult human hepatocytes of an immortalizing gene that can be subsequently and completely excised by Cre/Lox site-specific recombination. When transplanted into the spleen of rats under transient immunosuppression, reversibly immortalized NKNT-3 cells provided life-saving metabolic support during acute liver failure induced by 90% hepatectomy.

Treatment of nonalbumin rats by transplantation of immortalized hepatocytes using artificial human chromosome.

The shortage of organ donors has impeded the development of human hepatocyte transplantation. Immortalized hepatocytes, however, could provide an unlimited supply of transplantable cells. To determine whether immortalized hepatocytes could provide global metabolic support in end-stage liver disease, rat hepatocyte clones were developed by transduction with the gene encoding the simian virus 40 T antigen (SVLT) using the new technique of human artificial mini chromosome (HAC). Immortalized rat hepatocyte clones were developed by transduction with the gene encoding the SV40 using HAC. Many clones were obtained using this technique. From comparison of the properties of all these clones using the normal hepatocytes and reverse transcription-polymerase chain reaction (RT-PCR), the characteristics of the cell clones (at least partially characterized, and assayed for albumin, glucose-6-phosphate and dipeptidyl peptidase-4, gamma-glutamyltranspeptidase, SVLT and beta-actin expression by RT-PCR) showed no differences other than the immortalization. We compared the albumin bands of the first-day (0-day) and 30-day cells by RT-PCR, showing conditions to be stable for at least 1 month. Three experimental animal model groups were used for albumin analysis: nonalbumin rats with 2/3 hepatectomy only (R-NARs; n = 4); R-NARs with intrasplenic transplantation of 3 x 10(7) primary hepatocytes (pHTx; n = 4); and R-NARs with intrasplenic transplantation of 3 x 10(7) immortalized hepatocytes (iHTx; n = 4). All HTx groups produced albumin, but the immortalized hepatocyte group did not generate significantly elevated albumin levels compared with primary hepatocytes. The results presented herein have demonstrated an initial step toward the development of immortalized hepatocytes for transplantable cells or artificial organs using HAC technology.

Myoglobin facilitated oxygen diffusion maintains mechanical function of mammalian cardiac muscle.

Myoglobin, an intracellular iron containing protein that binds oxygen reversibly, has been shown in model systems to facilitate the diffusion of oxygen and thereby maintain the mechanical function of exercising canine skeletal muscle and of hypoxic benthic fish hearts. Since no such role has yet been established for mammalian cardiac muscle small diameter (less than or equal to 0.70 mm) isolated kitten papillary muscles were stimulated at 24 X min-1 under isometric conditions in a physiological bath maintained at 30 degrees C with an oxygen tension of approximately equal to 450 mm Hg (59.8 kPa) to obtain a level of oxygenation just adequate to meet the metabolic needs of the muscles, as confirmed experimentally. Myoglobin was inactivated by adding 2 X 10(-3) mol X litre-1 sodium nitrite to the bath to abolish the facilitated diffusion of oxygen in the presence or absence of glycolytic blockade by 10(-4) mol X litre-1 sodium iodoacetate. This resulted in a 22(8)% (with blockade) or 10(3)% (without blockade) decrease (p less than 0.05) in the maximal rate of relaxation (-dT/dtmax) of the papillary muscles. Since the depression in mechanical function was reversible by increasing the bath oxygen tension to approximately equal to 600 mm Hg (79.8 kPa) it is concluded that the myoglobin facilitated diffusion of oxygen plays a role in maintaining the mechanical function of mammalian cardiac muscle under normal conditions. Furthermore, the maximal rate of relaxation of cardiac muscle is a sensitive indicator of the presence of hypoxia.

Use of intracoronary KCl in the beating and asystolic heart to determine the mechanism of initiation of the left ventricular mechanoreceptor reflex.

To provide further evidence that the veratrum alkaloids' mechanical, positive inotropic effect and not their chemical depolarising action predominates in initiating the left ventricular mechanoreceptor (including the Bezold) reflex the effect of intracoronary KCl, a chemical depolarising agent like the veratrum alkaloids, but with a negative inotropic effect, was studied in beating and verapamil-asystolic hearts. Five dogs were placed on a total cardiac bypass, pneumonectomised and their coronary and systemic circulations isolated and perfused separately, at a constant rate, so that changes in systemic pressure reflected changes in systemic resistance. Injection of 5 mmol X litre-1 KCl into the isolated coronary circulation caused cardiac asystole and a resultant reflex rise in systemic pressure (resistance) of 26 +/- 9% (p less than 0.05) above the control of 10.5 +/- 0.7 kPa (79 +/- 5 mmHg). This pressure rise, which indicates predominance of KCl's mechanical, negative inotropic over its chemical depolarising effect, was abolished by vagotomy, indicating its reflex nature. Contrariwise, in five other pneumonectomised dogs, similarly perfused on total cardiac bypass but with cardiac asystole from intracoronary verapamil, a subsequent, similar intracoronary dose of KCl now produced a fall in systemic pressure (resistance) of 8 +/- 2% (p less than 0.005) below the control of 12.8 +/- 0.5 kPa (96 +/- 4 mmHg). This pressure fall, presumably due to chemical depolarisation of the left ventricular mechanoreceptors, was also abolished by vagotomy.(ABSTRACT TRUNCATED AT 250 WORDS)

Reflex hypotension due to regional activation of left ventricular mechanoreceptors to explain the hypotension noted in clinical myocardial ischaemia or reperfusion.

To determine if regional increases in myocardial contractility, as may occur clinically in angina pectoris, myocardial infarction, or coronary thrombolysis, can initiate the reflex hypotension that sometimes accompanies these conditions, regional injections of positive inotropic agents were made into 32(3)% of the left ventricular myocardium in seven pneumonectomised dogs on total cardiac bypass. The coronary and systemic circulations were isolated and perfused separately. The systemic circulation was perfused at a constant rate so that changes in systemic pressure reflected changes in resistance. Regional injections of doses from 0.001 to 1.0 micrograms noradrenaline in a 0.1 ml volume appreciably increased regional contractility, detected visually and by strain gauge arches, whereas global contractility (left ventricular peak dP/dt) was increased much less. This caused a fall in the systemic pressure (resistance) of 14(2)% below the control value of 78(5)mm Hg, at the largest dose. The decreases in resistance were abolished by bilateral vagotomy, proving their reflex nature. The smaller (0.0001-0.01 micrograms) doses of noradrenaline and the smallest (0.25 micrograms) dose of veratridine increased regional contractility almost without increasing global contractility, indicating that the increase in regional contractility was the major cause of the reflex decrease in systemic resistance. In one animal a decrease in contractility in a control myocardial region occurred simultaneously with the experimentally produced increase in regional left ventricular contractility. This decrease may be analogous to the increase in contractility in the non-ischaemic left ventricular myocardium that occurs simultaneously with the decrease in contractility in the ischaemic region in clinical or experimental myocardial infarction. Left ventricular mechanoreceptors in the region with increased contractility probably initiate the reflex hypotension that sometimes occurs in both circumstances. Thus in angina pectoris or acute myocardial infarction the reflex hypotension probably originates in the hyperactive non-ischaemic myocardial region, whereas in coronary arterial thrombolysis it probably originates in the newly reperfused, formerly ischaemic, region.

Effect of nitroglycerin with and without systemic hypotension on canine regional myocardial tritiated water deposition.

A marked alteration in the transmural distribution of left ventricular blood flow, with a relative increase in subendocardial and mid-wall flows, but with no change in the distribution of the relative blood flow to the two ventricles occurred when nitroglycerin was administered and the systemic arterial blood pressure in the upper body maintained near control levels in anaesthetized, open-chested dogs. The relative increase in subendocardial and mid-wall flows may have resulted from a direct action of nitroglycerin on the coronary vasculature. On the other hand, the intravenous administration of nitroglycerin, when followed by the hypotension which it produces, did not alter the transmural distribution of blood flow in the left ventricle of the dog. Blood flow to the right ventricle relative to flow to the left ventricle increased in this situation.

Changes in regional myocardial blood flow and variable development of hypertrophy after aortic banding in puppies.

Supravalvar aortic banding was performed in 6 to 12 week puppies. Sixteen animals were studied 7.3 (3.5 to 10) months later, closed-chested under morphine-chloralose, catheters being positioned in the great vessels and heart, including the left atrium for microsphere injection. Compared with 11 controls, eight dogs developed biventricular hypertrophy, four isolated left ventricular hypertrophy and four had no hypertrophy. The left ventricular systolic pressure was similar (P greater than 0.05) in these 3 banded groups (mean, 30 +/- 2 [SEM] kPa, [222 +/- 16 mmHg], n = 16). The left ventricle was divided into three coronal slices with approximately 59 samples being taken from subendocardial, midwall, and subepicardial layers and additional samples from the atria and right ventricle for regional myocardial flow measurement. As left ventricular hypertrophy increased, the subendocardial/subepicardial flow ratio decreased (r = -0.8). Heterogeneity of left ventricular regional myocardial flow, including a base-to-apex decrease in flow, present in controls, was markedly reduced in the banded dogs. Analysis of variance was found to be the most sensitive test for detecting left ventricular perfusion abnormalities since in banded dogs without hypertrophy, total and regional subendocardial/subepicardial flow ratios were not significantly different from control values, whereas the subendocardial circumferential flow pattern determined by analysis of variance was significantly different from control in these dogs (P less than 0.05).

Impairment of antigen-presenting cell function by ultraviolet radiation. II. Effect of in vitro ultraviolet irradiation on antigen-presenting cells.

The s.c. injection of 10 mM 2,4,6-trinitrobenzene sulfonic acid (TNBS) derivatized splenic adherent cells (SACs) into syngeneic mice primes for contact sensitivity or delayed-type hypersensitivity (DTH) when these animals are challenged with picryl chloride on the ear or trinitrophenol (TNP)-coupled cells in the footpad, respectively. If recipient mice are exposed to ultraviolet light (UV) irradiation and are immunized with normal TNP-treated SACs, they develop marked DTH reaction upon challenge but develop limited DTH reactions if immunized with hapten-derivatized SACs that had been obtained from UV-treated recipients. Moreover, if the SACs are obtained from normal mice but are treated in vitro with UV light (1.2 to 1.4 mJ/cm2/sec over the wavelength range 280 to 340 nm at a tube to target distance of 20 cm) these cells can neither prime nor elicit hapten-specific T cell immunity in UV-treated recipients. If UV-treated TNP SACs are used to prime UV-irradiated recipients, TNP-specific suppressor T cells are generated rather than T effector cells.

Procurement and preparation of human isolated small intestinal grafts for transplantation.

We have developed a donor operation that incorporates en bloc removal of the liver and intestine with a limited surgical resection in vivo. Over the past 18 months, we have used the following technique for the retrieval and preparation of seven isolated small intestinal allografts. The donor operation and bench preparation can be divided into three phases. During the first phase, the small intestine is removed with the liver, pancreas, and an aortic segment. In the second phase performed ex vivo, the donor liver can be separated from the specimen. The third phase involves additional bench dissection to yield an isolated intestinal allograft. The principle advantage of this technique is that it reduces potential liver injury by minimizing the surgical dissection required in vivo. Also, dividing the liver from the intestine ex vivo allows the organs to be separated in a bloodless field under controlled conditions that may be especially important when two different surgical teams are involved.

Transfusion-induced graft-versus-host disease after liver transplantation. Documentation using polymerase chain reaction with HLA-DR sequence-specific primers.

Graft-versus-host disease (GVHD) occurring after liver transplantation can pose a difficult diagnostic dilemma. Similar clinical and pathologic skin and gastrointestinal manifestations can result from other causes (i.e., drugs, infections). Treatment for each of these entities differs, and the high mortality associated with GVHD makes this distinction critical. GVHD has been assumed to result from the cotransplantation of donor lymphoid tissue along with the allograft. In most instances, the patient also receives blood products during the operation, and occasionally during the postoperative period, and the lymphoid cells in these products are also a potential source of concern. In this report, we describe a patient who developed GVHD after liver transplantation. Using molecular diagnostic techniques, we determined that the source for this GVHD was not the organ donor, but was most likely nonirradiated blood products received during the hospital course. Our results suggest that transplant recipients with concomitant hematopoietic dysfunction would benefit from irradiated blood products to reduce the likelihood of this complication.

Modulation of the major histocompatibility complex antigen and the immunogenicity of islet allografts.

Modulation of major histocompatibility complex (MHC) antigen by parenchymal cells and "passenger leukocytes" is a common feature of allograft rejection. To assess its significance we have examined the fate of antigen-presenting cell (APC)-depleted pancreatic islet allografts subsequent to increasing their expression of MHC antigens by in vitro exposure to the lymphokine interferon-gamma (gIFN). While most untreated grafts survived indefinitely, gIFN-exposed grafts were acutely rejected. Using in vitro islet cell-lymphocyte coculture assays, we attempted to dissect the underlying mechanism of enhanced islet cell immunogenicity resulting from gIFN treatment. We determined that gIFN exposure did not affect the capacity of islet cells to serve as APC for T lymphocytes, however islet cell exposure to gIFN was associated with enhanced vulnerability to allogeneic cytotoxic T lymphocyte (CTL) lysis in vitro by an CD5+ (OX-19+), CD8+ (OX-8+), CD4- (W3/25-), class I-restricted CTL. On the basis of these findings, we conclude that antigenic modulation can be a decisive factor in the survival of engrafted tissues by augmenting the interaction of the graft antigens with cytolytic effector T lymphocytes.

Treatment of surgically induced acute liver failure by transplantation of conditionally immortalized hepatocytes.

The shortage of human livers available for hepatocyte isolation limits its clinical application. The availability of cloned, conditionally immortalized hepatocytes that could be grown in culture but would lose their transformed phenotype and provide metabolic support upon transplantation would greatly facilitate the treatment of acute liver failure. Toward this goal, we transduced isolated Lewis rat hepatocytes using a replication-defective recombinant retrovirus capable of transferring a gene encoding a thermolabile mutant simian virus 40 T antigen (SV40ts). The cloned, immortalized hepatocytes proliferate at 33 degrees C. At the nonpermissive temperatures (37-39 degrees C), they stop growing and exhibit characteristics of differentiated hepatocytes. These cells did not produce tumors when transplanted in mice with severe combined immunodeficiency disease or in syngeneic rats. To induce acute liver failure, Lewis rats were subjected to 90% hepatectomy (Hpx) and given 5% oral dextrose. All rats that did not undergo hepatocyte transplantation died within 96 hr. Fifty percent of rats that received intrasplenic injection of 10 x 10(6) primary Lewis rat hepatocytes (G2, n=6) or 10 x 10(6) SV40ts-conditionally immortalized (SV40ts-ci) hepatocytes (G3, n=8) 1 day before 90% hepatectomy survived, whereas 80% of rats that received an intraperitoneal injection of 200 x 10(6) primary Lewis rat hepatocytes (G4, n=10) or 200 x 10(6) SV40ts-ci hepatocytes (G5, n=10) on the day of hepatectomy survived. Survival after intraperitoneal injection of a cellular homogenate of 200 x 10(6) primary Lewis rat (G7, n=9) or SV40ts-ci hepatocytes (G8, n=10) on the day of Hpx was 33% and 40%, respectively, whereas survival after intraperitoneal injection of 200 x 10(6) Lewis rat bone marrow cells (G6, n=7) was 29%. Thus, transplanted, conditionally immortalized hepatocytes can be as effective as primary hepatocytes in supporting life during acute liver insufficiency. This work represents the first step in developing an hepatocyte cell line that would partially alleviate the organ-donor shortage and could be of potential clinical value.

Use of gene therapy to suppress the antigen-specific immune responses in mice to an HLA antigen.

Hematopoietic chimerism has been used in the laboratory to induce life-long immunologic tolerance to donor antigens. The present study demonstrates that mice transplanted with autologous bone marrow cells retrovirally transduced to express HLA-A2.1 develop a significantly depressed immune response to this antigen while retaining normal reactivity to HLA-B7. Retrovirus-mediated transduction was performed using whole bone marrow-producer cell coculture. This approach did not result in significant gene transfer into hematopoietic progenitor cells. Despite this, the antibody response to HLA-A2.1 in mice reconstituted with genetically modified BMC was completely suppressed three months following bone marrow transplantation. Cell-mediated immunity to HLA-A2.1 was partially suppressed in three-fourths of animals tested three months later, although one animal had a CTL profile similar to that an of HLA-A2.1 transgenic mouse. Complete suppression of the antibody-mediated immune response occurred when only one-third of mice had evidence of the introduced genes in their spleen and one-tenth had the introduced sequences in their circulating WBCs by PCR. In conclusion, engineering of BMC to express donor MHC genes may be an alternative to xenogeneic BMT to induce chimerism and tolerance. More efficient transduction of bone marrow progenitor cells may result in more persistent gene expression and long-lasting transplantation tolerance in recipients of genetically modified bone marrow. Successful application of this technology may also be useful in altering immune responses to other external and self antigens.

Combined liver/small bowel transplantation using a blood group compatible but nonidentical donor.

A successful liver/small intestinal transplantation with a blood group O donor to a blood type A recipient is described. Mild graft versus host disease developed, manifested by hemolysis, but did not result in graft loss or patient mortality. This suggests that minor ABO incompatibility may be tolerated with intestinal transplantation, despite the transplantation of large amounts of lymphoid tissue.

A new technique for combined liver/small intestinal transplantation.

The most common application of small bowel transplantation is for the patient with parenteral nutrition-induced liver failure. In this setting, the small intestine is transplanted simultaneously with the liver. We identified three technical problems that we believe contributed to complications in our first eight patients. First, pancreaticoduodenectomy was challenging in the infant donor. Second, the bowel graft was prone to volvulus around the skeletonized donor portal vein. Third, in the pediatric recipient, use of the donor bowel for Roux-en-Y biliary reconstruction was associated with biliary leaks in the early postoperative period. Our surgical technique of liver/small bowel (L/SB) transplantation has evolved since our early experience in 1990. Modifications in the L/SB operation, reported briefly in 1996 and 1997, have led to easier graft preparation and have reduced the incidence of technical complications.

Extracorporeal liver perfusion using human and pig livers for acute liver failure.

BACKGROUND: Patients with fulminant hepatic failure (FHF) often die awaiting liver transplantation. Extracorporeal liver perfusion (ECLP) has been proposed as a method of "bridging" such patients to transplantation. We report the largest experience to date of ECLP using human and porcine livers in patients with acute liver failure. METHODS: Patients with FHF unlikely to survive without liver transplantation were identified. ECLP was performed with human or porcine livers. Patients underwent continuous perfusion until liver transplantation or withdrawal of support. Two perfusion circuits were used: direct perfusion of patient blood through the extracorporeal liver and indirect perfusion with a plasma filter between the patient and the liver. FINDINGS: Fourteen patients were treated with 16 livers in 18 perfusion circuits. Nine patients were successfully "bridged" to transplantation. ECLP stabilized intracranial pressure (ICP) and cerebral perfusion pressure (CPP). Arterial ammonia levels fell from a median of 146 to 83 micromol/liter within 12 hr and this reduction was maintained at least 48 hr. Pig and human ECLP lowered ammonia levels equally. Serum bilirubin levels also fell from a median of 385 to 198 micromol/liter over the first 12 hr but the response was not sustained as well with porcine livers. There was no immunological benefit to using the the filtered perfusion circuit. INTERPRETATION: These data demonstrate that ECLP is safe and can provide metabolic support for comatose patients with fulminant hepatic failure for up to 5 days. While labor and resource intensive, this technology is available to centers caring for patients with acute liver failure and deserves wider evaluation and application.

Effect of surrogate tolerogenesis on the vascular rejection of pig heart xenografts.

BACKGROUND: Organ xenografts are fulminantly rejected by antibody-mediated vascular rejection. Surrogate tolerogenesis (ST), the induction of tolerance within the donor, is effective with aorta xenografts. This preliminary study assesses the effect of ST on preformed antibodies and rejection of porcine heart xenografts. METHODS: Tolerance to the donor pig was induced by infusing recipient marrow into fetal pigs. Later, pig splenocytes were transfused and heterotopic pig hearts transplanted using chimeric or nonchimeric pigs. Anti-pig antibodies were assessed. RESULTS: With ST alone, xenografts developed cellular rejection at 4-6 days, whereas control grafts developed vascular rejection at 3-4 days (cellular vs. vascular, P<0.03). There was a reduction in preformed antibodies (P<0.03). ST combined with moderate cyclosporine prevented rejection at 9+ and 25 days in sensitized recipients compared with vascular rejection at 0.5-2 days for controls (P<0.07). CONCLUSIONS: ST seems to provide protection against vascular rejection. The cellular rejection seems sensitive to cyclosporine.

Disaccharidase activities and fat assimilation in pediatric patients after intestinal transplantation.

BACKGROUND: Intestinal transplantation has become an accepted therapy for short bowel syndrome and other types of intestinal failure. In order to assess digestive capabilities and feeding practices in a group of 22 pediatric patients after intestinal transplantation, we assessed mucosal disaccharidase activities and assimilation of total dietary lipid and vitamin E. Twelve of the patients had undergone contemporaneous liver transplantation. METHODS: Mucosal biopsies were assayed for disaccharidase activities between 15 and 412 days after transplantation in 7 of the 22 when all were receiving some enteral nutrition and were free of rejection. Coefficients of lipid absorption were determined in those patients receiving total enteral feeding (two-thirds polymeric/one-third elemental) between 43 and 1032 days after transplantation; oral vitamin E tolerance tests were done at about the same time. RESULTS: Activities of lactase, sucrase, maltase, and palatinase consistently exceeded reference ranges (P<0.05). Mean coefficient of lipid absorption equaled 86+/-12% and was not influenced by duration of time after transplantation. No patient required dietary lipid restriction. No significant absorption of vitamin E was demonstrated until 160 days after transplantation. Vitamin E absorption did correlate with length of time elapsed after surgery (r=0.64, P<0.0011). CONCLUSIONS: The results of this investigation show that, in the absence of histologic or clinical indications of allograft rejection, pediatric intestinal transplant recipients do not have primary disaccharidase deficiencies. Similarly, absorption of usual dietary lipid content is adequate once weaning from parenteral nutrition is complete. In contrast, early assimilation of vitamin E is poor. Vitamin E absorption subsequently improves, but the mechanism is obscure.

Preliminary experience with intestinal transplantation in infants and children.

OBJECTIVE: This report discusses the preliminary experience with intestinal transplantation in children at the University of Nebraska Medical Center. PATIENTS: During the past 4 years, 16 intestinal transplants have been performed in infants and children. Thirteen have been combined liver and bowel transplants, and the reminder were isolated intestinal transplants. Nearly half of the patients were younger than 1 year of age at the time of surgery, and the vast majority were younger than 5 years of age. All but one had short bowel syndrome. RESULTS: The 1-year actuarial patient and graft survival rates for recipients of liver and small bowel transplants were 76% and 61%, respectively. Eight of 13 patients who received liver and small bowel transplants remain alive at the time of this writing, with a mean length of follow-up of 263 (range, 7 to 1223) days. Six patients are currently free of total parenteral nutrition. All three patients receiving isolated intestinal transplants are alive and free of parenteral nutrition. The mean length of follow-up is 384 (range, 330 to 450) days. Major complications have included severe infections and rejection. Lymphoproliferative disease, graft-versus-host disease, and chylous ascites have not been major problems. CONCLUSIONS: Although intestinal transplantation is in its infancy, these preliminary results suggest combined liver and bowel transplants and isolated intestinal transplantation may be viable options for some patients with intestinal failure caused by short bowel syndrome or other gastrointestinal disease in whom long-term total parenteral nutrition is not an attractive option.

Use of gene therapy to induce human-mouse xenogeneic chimerism.

BACKGROUND: Bone marrow transplantation (BMT) has been used in the laboratory to overcome the immunologic barriers to xenotransplantation and results in chimerism and specific tolerance to donor antigens in lethally irradiated mice. Clinically, BMT carries the considerable risks of graft-versus-host disease and graft failure. Retrovirus-mediated gene transfer could provide a means of introducing foreign major histocompatibility (MHC) genes into host bone marrow cells (BMC) and thus accomplish the immunologic goals of BMT, without the associated risks. METHODS: Using a Moloney virus-based vector, a replication defective retrovirus was constructed that contained a complementary DNA encoding the human MHC antigen HLA-A2. Three million C57BL/6 mouse BMC were cocultured for 48 hours with 1 x 10(6) HLA-A2 virus "producer" cells in the presence of 15% WEHI supernatant (interleukin-3) and 200 units/ml interleukin-6. Putatively infected BMC were then used at 2 to 3 x 10(6) BMC/animal to reconstitute lethally irradiated syngeneic mice. RESULTS: Twelve days after reconstitution, spleen colonies were found to have integrated the full-length retroviral sequences. Thirty days after BMT, the introduced DNA could be found in the bone marrow, thymus, and spleen, and approximately 5% of T cells in the spleen expressed the HLA-A2 surface antigen. CONCLUSIONS: These data show that xenogeneic MHC genes can be introduced and expressed in mouse hematopoietic cells in vivo and indicate that gene therapy potentially may be used in the future to manipulate the immune system to induce transplantation tolerance.