Efficient CNS gene delivery by intravenous injection.

We administered recombinant SV40-derived viral vectors (rSV40s) intravenously to mice with or without prior intraperitoneal injection of mannitol to deliver transgenes to the central nervous system (CNS). We detected transgene-expressing cells (mainly neurons) most prominently in the cortex and spinal cord; prior intraperitoneal mannitol injection increased CNS gene delivery tenfold. Intravenous injection of rSV40s, particularly with mannitol pretreatment, resulted in extensive expression of multiple transgenes throughout the CNS.

The human organic anion transport protein SLC21A6 is not sufficient for bilirubin transport.

A recent study (Cui, Y., Konig, J., Leier, I., Buchholz, U., and Keppler, D. (2001) J. Biol. Chem. 276, 9626-9630) suggests that human OATP2 (SLC21A6), also known as OATP-C and LST1, mediates hepatic bilirubin transport. Because of methodologic concerns, this study was designed to examine this issue using a bilirubin transport assay that was validated in overnight cultured rat hepatocytes. These studies showed that cultured rat hepatocytes transported bilirubin with kinetics virtually identical to the transport of sulfobromophthalein. This assay was then used to quantify bilirubin transport by HeLa cells that had been stably transfected with OATP2 under regulation of a metallothionein promoter. Immunoblot analysis revealed abundant expression of OATP2 after incubation of cells for 48 h in zinc, whereas uninduced cells had no expression of this protein. In OATP2-expressing (zinc-induced) HeLa cells at 37 degrees C, the uptake of [35S]sulfobromophthalein was substantial (51.6 +/- 16.5 pmol/15 min/mg protein, n = 5) with little cell-associated ligand in non-expressing (uninduced) cells (0.54 +/- 0.16 pmol/15 min/mg protein, n = 5, p < 0.002). In contrast, there was no difference (p > 0.2) in cell-associated [3H]bilirubin in induced (OATP2-expressing) as compared with uninduced cells (11.25 +/- 3.02 pmol/15 min/mg protein versus 9.15 +/- 1.68 pmol/min/mg protein, respectively, n = 5) We obtained similar results in OATP2-transfected HEK293 cells that were used in the original report. The existence of a bilirubin transporter has been an important field of investigation for many years. Although the current study indicates that a role for OATP2 in hepatocyte bilirubin transport is unlikely, it provides new and sensitive tools that can be adapted to examine the function of putative bilirubin transporters in the future.

What can SV40-derived vectors do for gene therapy?

The limited success of gene therapy as an approach to treating human disease largely reflects the limitations of the gene delivery vectors that have been used. Poor titers, low transduction efficiency, waning transgene expression and immunogenicity have remained obstacles in the field. As a consequence, much research in normal, immunocompetent animals has not demonstrated therapeutic levels of gene delivery, and results from most human clinical trials have been predictably discouraging. Recombinant gene transfer vectors derived from SV40 virus (rSV40) are potentially of great interest for those working in gene therapy, since these vectors are not subject to many of the problems that have limited gene delivery using other vector systems. rSV40 is made at a very high titer and infects - and so transduces - almost all nucleated cell types very efficiently, regardless of lineage or whether they are resting or dividing; they integrate and are not susceptible to transgene silencing; and they elicit no detectable immune response on the part of normal animals and so can be used to deliver multiple transgenes over time and in sequence. The recent development of 'gutless' rSV40 vectors has expanded the range of potential therapeutic transgenes that can be delivered with this system and added flexibility to the expression configurations that can be accommodated. All of these functional characteristics of SV40-derived vectors have their bases in the biology of SV40 and similar viruses, and have important implications for the potential utility of rSV40 vectors in gene therapeutics. Like all viral gene delivery systems, these vectors have their idiosyncrasies and limitations. They also allow gene delivery that bypasses many of the difficulties that have plagued the field from its inception.

Durability of transgene expression and vector integration: recombinant SV40-derived gene therapy vectors.

Many applications of gene delivery require long-term transgene expression. In dividing cells, this result necessitates vector genome persistence, usually by integrating into cellular DNA. Since recombinant gene delivery vectors derived from tag-deleted, replication-incompetent simian virus-40 (SV40) provide for long-term transgene expression in resting and dividing cells, we tested whether such enduring transgene expression reflected integration into cellular genomes. Several lines of evidence suggested this likelihood. After transduction in vitro, continuously dividing cell lines and continuously stimulated primary cells uniformly showed transgene expression for many months. Mice whose livers were transduced in vivo, partially resected, and allowed to regenerate showed comparable levels of transgene expression in regenerated and preoperative livers. Thus, replicationincompetent SV40 vectors (rSV40) persist in vitro and in vivo despite extensive cell division. We tested the possibility that this persistence reflected integration directly. Southern blot analyses of genomic DNA from transduced 293 cells showed that vector genome incorporation into cell DNA happened within days of transduction. Episomal vector DNA was barely detectable 96 hours post-transduction. Inverted PCR, used to characterize vector integration points, showed vector DNA integrated randomly into the cell genome. The circular rSV40 genome opened at different points in each integrand. A significant proportion of the integrands did not contain the entire vector sequence, but rather only portions thereof. Quantitative Southern blot analysis showed approximately 3.05 transgene copies per cell. Therefore, recombinant SV40 gene delivery vectors integrate into the cellular DNA of both resting and dividing cells, and do so randomly and within days of transduction. This integration may explain long-term transgene expression.

A non-immunogenic adenoviral vector, coexpressing CTLA4Ig and bilirubin-uridine-diphosphoglucuronateglucuronosyltransferase permits long-term, repeatable transgene expression in the Gunn rat model of Crigler-Najjar syndrome.

Host immune responses limit the duration of expression of transgenes introduced by recombinant adenoviruses, preclude gene transfer upon vector readministration and cause liver injury. CTLA4Ig inhibits immune response by blocking the co-stimulatory interaction between CD28 on T cells and B7 on antigen-presenting cells. We have constructed a recombinant adenovirus, Ad-hUGT1A1-CTLA4Ig that coexpresses human bilirubin-uridinediphosphoglucuronate glucuronosyltransferase (hUGT1A1) and soluble murine CTLA4Ig, both driven by CMV immediate-early promoters. After intravenous injection of this vector (6 x 10(11) p.f.u.) into UGT1A1-deficient jaundiced Gunn rats, serum CTLA4Ig levels peaked at 1.8-2.0 mg/ml on day 7 and declined thereafter to 0.2 mg/ml by day 180. Serum bilirubin declined from mean preinjection levels of 8.0 mg/dl to 0.48-0.6 mg/dl in 3 days, remained normal for 28 weeks, and then gradually increased to 8 mg/dl by day 350. A second injection of Ad-hUGT1A1-CTLA4Ig normalized serum bilirubin. In two rats in this group that were followed longer, serum bilirubin increased to 3.1 and 3.5 mg/dl in 40 weeks, but was normalized again after a third injection. The antibody and cytotoxic lymphocyte (CTL) responses were negligible, and liver biopsy showed no inflammatory cell infiltration. Rats receiving a tertiary challenge with Ad-LacZ (expressing E. coli beta-galactosidase) (5 x 10(11) p.f.u.), 2 months after the second dose of Ad-hUGT1A1-CTLA4Ig, showed beta-galactosidase expression in over 80% of hepatocytes. In contrast, after Ad-hUGT1A1 (which expresses UGT1A1 alone) injection, serum bilirubin remained normal for only 4 weeks, and returned to preinjection levels by day 120. Bilirubin levels did not decline upon reinjection, and beta-galactosidase was not expressed after Ad-LacZ. High levels of adenovirus-specific antibodies and CTL, and hepatic inflammation were found. This is the first demonstration that coexpression of CTLA4Ig permits prolonged expression and repeatable gene transfer by an adenoviral vector.