Efficient gene transfer to hepatocytes without use of viral vectors would
have great advantage for liver targeted gene therapy. We have utilized isolated
liver perfusion in vivo to transfect hepatocytes with naked plasmid DNA. This
technique involves total vascular isolation of the liver, with delivery of the
plasmid via either the portal vein (PV), IVC, or bile duct, with or without
outflow occlusion. The plasmids pCILuc (luciferase gene) or pCILacZ (ß-galactosidase
gene) were injected in normal saline with 15% mannitol. Animals were sacrificed
at 24 hours, livers were harvested and prepared for analysis by homogenation and
extraction of the enzyme, or by tissue staining for X-Gal. Results in the mouse
(expressed as ug luciferase/liver) with 100 ug plasmid showed increased
expression if there was outflow occlusion from the liver. Results in a rat model
with 750 ug plasmid were similar.
PV+occl. PV-occl. IVC+occl. IVC-occl.
Mouse 3.73 0.005 17.34 2.83
Rat 53.5 1.50
Bile duct injection in the mouse and rat yielded 15.39 ug/liver and 1.3
ug/liver, respectively. Of variables measured (DNA concentration, duration of
injection, volume of infusate), the amount of intraparenchymal pressure
generated during injection of the plasmid seemed the most important, with the
optimal pressure being 31-40mm Hg. The effect of pressure plateaued above 40mm
Hg.
intraparenchymal pressure (mm Hg) mean light units (X 10^{9})
10-20 51
21-30 121
31-40 1910
>40 580
ß-galactosidase activity measured with X-Gal staining showed uptake
and activity distributed fairly evenly in the mouse liver, while rat liver
showed a preference toward peri-acinar positivity with PV injection and central
vein positivity with IVC injection. In summary, we have developed a direct DNA
delivery technique which allows hepatocytes to be transfected in vivo, does not
require removal of any liver tissue to be effective, and provides levels of gene
expression that are 10-100 times higher than previously reported with other
direct gene transfer techniques.
