Researchers at Oregon Health & Science University have turned acetaminophen's toxicity into an asset, using it to select genetically modified hepatocytes in vivo.

In animal experiments, the team was able to achieve therapeutic levels of both factor IX, whose absence leads to hemophilia, and phenylalanine hydroxylase, the enzyme that is absent in phenylketonuria.

The work, which was published in the June 9, 2021, issue of Science Translational Medicine, could solve a key stumbling block for gene therapy in children -- the dilution of the therapeutic gene as organs grow.

Genetic diseases "start at birth, and there are a lot of those," Markus Grompe told BioWorld Science.

There are also attempts to use the liver as a factory for therapeutic antibodies. Several groups are working on such an approach for protective broadly neutralizing antibodies in HIV, and at the recent annual meeting of the American Society for Cell and Gene Therapy, Homology reported preclinical success in producing an anti-complement antibody targeting C5 in the liver in vivo.

But "currently, liver-directed gene therapy is something that's done for adults," Grompe said. "And it's not just that no one wants to do clinical trials in kids, it's also that the vector gets diluted out."

The liver can regenerate itself to an astonishing degree. It's a virtue that was probably born out of necessity, Grompe said: It's the first line of defense against toxins that enter the body via the gut. And those toxins used to be quite plentiful.

"Our ancestors – they ate a lot of semi-rotten, semi-poisonous stuff, because they had to," Grompe said. "And if you couldn't regenerate your liver, you would die."

For non-integrating gene therapy, though, that regenerative capacity is a problem. As cells divide, they copy their genomic DNA, but not the transgene-containing DNA. Over time, as liver cells divide to replace lost cells, fewer and fewer cells have a copy of the transgene.

In adults, the turnover of liver cells is slower. But although there are long-term large animal studies showing transgene expression in the liver for up to a decade, Grompe said that "to expect that this would last for the life of a patient is, I think, a little optimistic."

The alternative is to use vectors that integrate into the cellular genome. Those transgenes can now be targeted to specific sites within the cellular genome, meaning that integration into an oncogene, which caused leukemia in an early clinical trial, is no longer a concern. But integration into the genome happens at low levels, making it much more challenging to achieve therapeutic dose levels.

Currently, the low efficiency of integration "can only be overcome with massive doses of vector, which has its own set of problems," Grompe said.

However, given the regenerative abilities of hepatocytes, Grompe and his team reasoned that it should be possible to expand therapeutically transduced cells in vivo.

Grompe compared the approach to the conditioning that is used to prepare the bone marrow for hematopoietic stem cell transplant, killing off existing hematopoietic stem cells to give transplanted cells room to expand.

"The success that people are having with bone marrow-derived diseases are possible because of conditioning," he said. "That's an idea that really hasn't been applied to many other tissues."

The team used acetaminophen, an over-the-counter painkiller sold in multiple countries under brand names including Tylenol.

Acetaminophen is metabolized by several pathways, and one of those pathways produces the metabolite N-acetyl-p-benzoquinone imine (NAPQI), which kills the liver cells that produce it.

In their work, the team made a gene therapy vector that contained both a therapeutic transgene, and an RNA that could prevent cells from utilizing the NAPQI-producing metabolic pathway.

They showed that when they first transduced mice in vivo, and then used acetaminophen to kill off cells that had not taken up the transgene. Therapeutically modified cells, which had also taken up acetaminophen protection, then expanded in response to the death of their neighbors.

Grompe said that he and his colleagues were surprised at how far they could push the expansion of genetically modified cells.

"Tylenol toxicity happens only in a subset of liver cells, and we therefore thought that the highest number of selection that we could achieve would be about 20%," he explained. But for reasons that are not yet clear, "at least in mice, that number turned out to be upward of 50%."

Grompe said the approach could be used to reach therapeutic levels with lower doses of vector, and to control the level of therapeutic proteins by fine-tuning the number of protein-producing cells.

"In a lot of these gene therapy trials, the dose response is somewhat unpredictable," Grompe explained.

"Some patients will have a therapeutic level, and the next patient, with the same dose, doesn't."

In their factor IX experiments, the team did repeated rounds of Tylenol treatment, measuring after each round whether protein levels were at the desired levels yet. If the same approach works in humans, he said, "you could overcome the vagaries of the dose response."