Early ultrastructural hepatocellular alterations after hydrodynamic tail vein injection

Background:

Hydrodynamic tail vein injection (hydrodynamic transfection, HT) is one of the latest murine gene transfer models in liver cancer research. Transfected DNA constructs are composed of two plasmids, one containing the gene of interest and the other the sleeping Beauty-transposase allowing a stable integration of this gene into the hepatocyte genome.

Rapid tail vein injection of a large volume of DNA-solution induces an acute cardiac congestion, resulting in a reflux into the liver, mainly in acinus zone 3.

It has been shown that HT induces this hydrodynamic force leading to permeabilization of the fenestred sinusoidal endothelium. Nevertheless, mechanisms of plasmid incorporation into hepatocytes remain unclear.

Methods:

In this study, 2 ml volume of empty vector or saline solution (control) were hydrodynamically injected into the tail vain of anaesthesized C57BL/6J/129Sv mice. Liver tissue was resected at different time points (1, 2, 5, 10, 20, 30 or 60 minutes after HT) and quickly fixed with buffered 1% osmium tetroxide. For electron microscopic evaluation, ultrathin sections were cut from glycidether embedded blocks.

Results:

After 1 minute, hepatocytes near the central venule in acinus zone 3 reveal some small membran bound vesicles in the cytoplasm. After few more minutes, vesicles increased in size to vacuoles and in number. Vacuoles can often be found in proximity to the nucleus. They are optically empty and contain no electron dense material. Some but other hepatocytes revealed signs of cell damage, i.e. swollen mitochondria, dilated endoplasmatic reticulum and golgi apparatus and disrupted plasma membranes, whereas most hepatocytes appeared normal. All findings were similar after the injection of the empty vector and the saline solution as well.

Discussion and conclusion:

Our ultramorphological findings indicate that hydrodynamic plasmid transfer to the hepatocytes is accomplished via active or passive endocytosis with an entrapment of the plasmid in multiple membran bound vesicles and vacuoles. Hepatocytes containing vacuoles did not show disrupted plasma membranes or other signs of cell damage, therefore mechanisms like i.e. membrane poration are unlikely, but cannot be entirely excluded.

The vesicle formation is a nonspecfic process, independent from the plasmid itself or its enclosed components. It remains to be clarified which active or passive mechanisms are involved in vesicle generation and finally how the DNA enters the hepatocellular nucleus.

Corresponding author: Ribback, Silvia

E-Mail: [e-mail]