Recombinant adenoviruses are highly efficient at transferring foreign genes in vivo. However, duration of gene expression is limited by the host antiviral immune response which precludes expression upon viral readministration. We tested the feasibility of prolonging gene expression by induction of central tolerance to adenoviral antigens in bilirubin-UDP-glucuronosyltransferase-1 (BUGT 1 )-deficient Gunn rats. Tolerance was induced by intraperitoneal injection of antilymphocyte serum, followed by intrathymic inoculation of one of the following: a recombinant adenovirus (Ad), adenovirus human UDP-glucuronosyltransferase (Ad-hBUGT 1 ) carrying the hBUGT 1 gene; a protein extract of the same virus; or viral infected hepatocytes. Controls received intrathymic injections of normal saline. After 12 d all groups were injected intravenously with 5 ϫ 10 9 pfu of either Ad-hBUGT 1 or adenovirus ␤ -galactosidase (Ad-LacZ) (expressing the Escherichia coli ␤ -galactosidase [LacZ] gene). In all three groups of tolerized rats, hBUGT 1 was expressed in the liver after administration of Ad-hBUGT 1 , with glucuronidation of biliary bilirubin of above 95%. Serum bilirubin levels decreased from 7.2 to 1.8 mg/d1 within 1 wk and remained low for 7 wk. Similar findings were observed following repeat injections given on days 45 and 112. In control rats serum bilirubin levels were reduced for only 4 wk, and viral readministration was ineffective. In all tolerized groups, but not in controls, there was a marked inhibition of appearance of neutralizing antibodies and cytotoxic lymphocytes against the recombinant adenovirus. Injection of wild type adenovirus-5 (Ad5) into the tolerized rats elicited a wild type-specific cytotoxic lymphocyte response. This is the first demonstration of Ad-directed long-term correction of an inherited metabolic disease following central tolerization with thymic antigen. ( J. Clin. Invest . 1996. 98:2640-2647.)

An important consideration in application of hepatocyte transplantation is whether the number of engrafted hepatocytes is sufficient to achieve the desired effect. Here we have evaluated the proliferative potential of transplanted primary hepatocytes during regeneration of hepatic lobes. Two million hepatocytes isolated from congeneic normal Wistar-RHA rats were injected into the main portal vein of deficient, jaundiced Gunn rats. The right branch of the portal vein was ligated 24 hr before hepatocyte transplantation (group A) or transiently clamped during hepatocyte injection (group B) or 24 hr after hepatocyte injection (group C). In these groups, the three lobes supplied by the right branch of the portal vein rapidly atrophied and disappeared in 4 days, whereas the remaining lobes proliferated, as shown by size increase and 5-bromo-2-deoxy-uridine uptake. Two control groups received 2 million (group D) or 20 million hepatocytes (group E) without ligation. Hepatocyte engraftment occurred in all groups. The greatest hypobilirubinemic effect was observed in group A, in which serum bilirubin concentrations were reduced to 1.7+/-0.45 mg/dl from pretransplantation levels of 6.9+/-1.2 mg/dl. This effect was even greater than that observed after transplantation of 20 times more hepatocytes without ligation (group E). Specific endonuclease digestion of a polymerase chain reaction-amplified segment of the ugt1 gene from hepatic DNA showed that up to 25% of the DNA was of donor origin. This paralleled the hepatic bilirubin-UDP-glucuronosyltransferase activity, which was above 50% of normal. The results indicate that the transplanted hepatocytes proliferate preferentially within the regenerating lobes, replacing more than 20% of the liver mass with the progeny of the transplanted phenotypically normal hepatocytes.