‘Ghost’ viruses offer potentially better approach to gene editing | Science

Not every human disease can be tackled with a pill or shot. Some disorders would ideally be treated by delivering molecular payloads—like modified viruses carrying gene-editing tools—into defective cells, where they could rewrite target genes. Despite some early successes, researchers are still struggling to get this approach to work.

Now, researchers say they have found a better way to deliver their gene-altering cargo to cells, dramatically improving their efficacy. The advance could bolster a wide range of gene-editing projects, making it possible to treat everything from muscle disorders to hearing loss. “This is a big step forward,” says Mitchell Weiss, a gene therapy expert at St. Jude Children’s Research Hospital who was not involved with the work.

Using viruses to deliver gene-based treatments has a troubled past. In 1999, Jesse Gelsinger, an 18-year-old, died in a gene therapy trial that attempted to cure his liver disease by adding a new gene to his cells. The new gene was packaged inside an adenovirus, a type of cold virus, that triggered a massive immune reaction that killed him.

Safer adeno-associated viruses continue to be used in gene therapies, in part because they are easy to grow in the lab and effectively reach their target cells. These viruses are also being used to deliver gene-editing tools such as CRISPR that repair defective genes. But concerns persist. Experts worry these modified viruses may trigger immune reactions, as they did with Gelsinger, or integrate their genetic material into human cells, potentially causing cancer.

As an alternative, researchers have developed viruslike particles (VLPs), hollowed out viruses that have their own genomes removed. Instead of carrying DNA encoding proteins that modify genes, their cargo is the proteins and RNA themselves that make the modifications. These VLPs shouldn’t pose any threat on their own, and they’re easy to grow in the lab. But when injected into animals, they make the proper genetic changes to only about 10% of the targeted cells in most tissues, well below the level needed for a desired therapy.

Researchers led by David Liu, a chemical biologist at Harvard University, tried to see whether they could step up the VLP game. To do so, they needed to tackle two main problems: ensuring gene-editing proteins were properly packaged inside the VLPs, and properly steered to the cell’s nucleus to carry out their editing.

Other groups had tried to solve the steering problem before by attaching protein snippets, called nuclear localization sequences (NLSs), to their gene-editing cargo, which direct it to a cell’s nucleus. But this interfered with the cargo’s packaging in the first place. That’s because VLPs are constructed inside the cytoplasm of specialized cells, called producer cells, in the lab. And if NLSs are attached to the cargo proteins as the particles are being built, the proteins migrate to the producer cell’s nucleus, instead of staying packaged inside the VLPs.

illustration of a engineered viruslike particle
Newly engineered viruslike particles could revolutionize therapies designed to edit genes involved in diseases.Aditya Raguram

So Liu’s team added a short linker protein between the cargo proteins and the nuclear location sequence, which they designed so it could later be cut. Inside the producer cell, the linker prevented the sequence from steering the cargo to the nucleus, ensuring the combination would be properly packaged inside the VLP. But once delivered to the target cell, a protein-cutting enzyme, also present in the VLP, cut the linker, restoring the nucleus-steering ability of the NLS snippet. For each set of VLPs, the researchers also engineered the outer coat of the particles to express particular sugar-decorated glycoproteins that steer the particles to infect specific target cells.

After confirming that their revamped VLPs successfully edited target genes of cells in a dish, the scientists turned to mice. First, they showed that when injected into the cerebrospinal fluid of mice, the VLPs infected target neurons in the brain, making the proper genetic edits to 53% of the target cells. They next sought to edit a gene called PCSK9 involved in cholesterol metabolism and heart disease. One week after injection into the blood, the revamped VLPs successfully edited 63% of defective PCSK9 genes in targeted liver cells, a 26-fold higher efficiency than previous VLPs. The researchers also tested the treatment in mice with a genetic mutation that causes blindness. Six mice given a single VLP injection below the retina all gained partial vision, the team reported yesterday in Cell.

The new VLPs also appear safer than past approaches. There’s no evidence that they cause “off-target” edits that could result in cancer, for example, says Paula Rio Galdo, a biologist with the Center for Energy, Environmental, and Technological Research who was not involved with the work. That could be critical, she says, for convincing regulators to accept future therapies using VLPs.