Gene therapy returns with a new set of tools
Image credit: Science photo library
Gene therapy has come of age and may be about to get a second chance to show its mettle in curing disease.
In the mid-1990s gene therapy looked to be a miracle cure: a way of treating people with conditions that depended on their own DNA. The therapy would provide them with DNA to augment or replace the genes in their own cells that were causing the problems. But trials quickly exposed a problem.
Teenager Jesse Gelsinger volunteered in 1999 for a gene-therapy trial but died within four days of being injected with the experimental compound. It was an altered virus carrying copies of a corrected gene. After his death, reports of hundreds of other problems caused by gene therapy reached the US National Institute of Health (NIH). It then emerged that several of the people who took part in the earliest trials intended to treat the severe combined immunodeficiency (SCID) ‘bubble boy’ syndrome had developed leukaemia within six years of treatment.
Studies found the gene therapy and part of the virus altered to carry it integrated into the genome of some of the treated cells. This, in combination with other mutations, triggered the leukaemia. The risk of inducing cancers through the therapy saw funding for the research dry up as hopes faded for gene therapy. But it is far from the first medical technology to run into problems at an early stage of development.
“Being too early is often the same as being wrong,” said Jennifer Laird, senior director of search and evaluation at Eli Lilly, at December’s Genesis 2016 conference for medical R&D in London. Some of the most important drugs for cancer today are monoclonal antibodies, she noted. Discovered in 1975, “they didn’t deliver the first time around”.
David Venables, CEO of start-up Synpromics, says: “I was involved in gene therapy back in the 1990s. It’s gone through the hype cycle with the trough of despondency and the re-emergence that we see now. But in that time, the molecular-level understanding of the basis of disease has advanced dramatically.”
A team at St Jude’s Children’s Research Hospital in Memphis, Tennessee, USA, announced in April 2016 that they had successfully treated patients with SCID-1X, the most common form of the ‘bubble boy’ immunodeficiency, with a version of gene therapy designed to be safer than predecessors but also more effective by targeting a wider range of immune cells.
To make such treatments safer, medical research has switched to a virus that behaves similarly to the gammaretrovirus used in the original trials. But the replacement is less likely to cause unwanted mutations. “Genetic insulators used to block activation of adjacent genes make it fundamentally safer,” claims Brian Sorrentino, director of the vector production facility at St Jude’s.
Nanotechnology may come up with a way of delivering genes into cell nuclei safely without involving viruses. There are many proposals for DNA carriers based on nanoparticles that could act as substitutes for viruses. But today, the virus vector is the only practical option. Teams such as Sorrentino’s use the lentivirus because it can insert the gene into the target genome, so the treatment can, in principle, be a one-off process for each patient.
There are other virus options but they do not integrate the new gene into the DNA of the target cell. As a result, the effects of the treatment fade over time. One possibility is to use the virus purely as a carrier for genes with the ability to break through cell walls and deposit the genetic payload inside. The ability to stitch replacement DNA into the cell genome comes from elsewhere. In one case, ‘fossil’ genes may provide an answer.
In the 1980s, the Minnesota Department of Natural Resources gave local university professor Perry Hackett the job of genetically engineering fish to grow more quickly. The initial project did not last long, shut down by public disquiet over modified foods. But Hackett and colleagues kept working on a section of DNA they found in salmon that has lain dormant in the natural world for 14 million years. The DNA contained what appeared to be a transposon or ‘jumping gene’. The transposon is a naturally occurring toolkit for cutting and pasting genes into chromosomes that some scientists believe evolved as a way of providing immunity to disease.
By randomly moving DNA around, transposons can activate genes under different conditions (see ‘Gene control’ panel). Some of those changes in activity can confer resistance to disease, although others may simply kill off the affected cell completely. But the transposons themselves mutated into ineffectiveness over millions of years. By 1997, Hackett and colleagues had, after identifying the broken sequences, reactivated the dormant DNA –dubbed the ‘Sleeping Beauty’ transposon.
Because it will readily copy and paste DNA into genomes, Sleeping Beauty is an option for gene therapy. Researchers at Witten University in Germany have combined Sleeping Beauty with adenovirus, overcoming that virus vector’s inability to integrate foreign DNA into the host cell’s genome.
Although Sleeping Beauty is highly effective at inserting DNA, the process is random enough for it to break an existing gene by plonking the new genetic information in the middle of it. Another relic of nature’s attempts to deal with infection may build a more precise Sleeping Beauty.
In summer 2016, scientists at Sichuan University’s West China Hospital in Chengdu were the first to obtain government approval to modify cells using a technology called CRISPR-Cas9. A team at the University of Pennsylvania is waiting for similar approval for their own trial in the US.
The CRISPR-Cas9 gene-editing technology first evolved in bacteria, seemingly as a way to co-opt DNA from viruses and bacteria in order to fend off their infections. In 2012, Jennifer Doudna and colleagues from the University of California at Berkeley found a way to harness the DNA from those bacteria to build a gene-editing system that takes the randomness out of the equation.
Lorenz Mayr, vice president of reagents and assay development at AstraZeneca, says the company started experimenting with the technology soon after it was discovered: “Why do we think this technology is as great and important? We see it as a breakthrough technology in pharmaceuticals and biotech. We want to use precise genome editing to do better in vitro and in vivo modifications.”
A single, 20-base-long strand of DNA attaches to the Cas9 protein and guides it to a complementary sequence in the cell’s genome. Even in a genome that contains several billion base pairs, the guide sequence will typically match just one section of DNA. It makes off-target cuts very rare.
Lu You’s team aims to treat lung cancer patients using modified T-cells – key cells in the immune system that are also targets for SCID-focused gene therapy. In contrast to those gene therapies, the Sichuan cancer treatment takes cells out of the patient and edits them ex vivo. They aim to snip out a gene responsible for producing a protein that limits the T-cell’s immune response. Relieved of that limitation, they hope the modified cells – once duds have been filtered out and the remainder reinjected into the patient – will home in on the metastatic cancer cells in the bloodstream.
Other researchers expect CRISPR-Cas9 or similar ‘DNA scissor’ technologies can be used to overcome the random behaviour of Sleeping Beauty and another competing gene-insertion system from the bacterial world called PiggyBAC. That would avoid the need to filter cells and open up the possibility of in vivo gene editing.
The bad news is that this has proved difficult to demonstrate in practice. Although researchers know these enzymes work, why they work remains more elusive.
The natural form of Sleeping Beauty works very slowly, which helped prevent it causing too many tumours in ancient fish. Directed evolution, in which derivatives generated using random mutations are selected based on performance, led to a version that is 100 times faster. Work by Orsolya Barabas and colleagues at the European Molecular Biology Laboratory in Heidelberg to determine the structure of Sleeping Beauty as it works pointed to further changes, increasing efficiency by a further 130 per cent. “We are still testing several further variants,” she adds.
Unfortunately, versions of Sleeping Beauty that incorporate targeting proteins work far less efficiently and are nowhere near accurate enough in picking out the right parts of the genome. But Barabas says she has no reason to believe that the techniques cannot be combined. “I believe that understanding Sleeping Beauty’s mechanism of movement at the structural level will help us better understand the reasons [for the difficulties] and enable us to overcome these challenges.”
Enhancements to the CRISPR technique may limit how much surgery the enzyme needs to do on the target genome. This will limit the potential damage of enzyme-driven DNA surgery.
Mayr points to work by a group that disabled the DNA-slicing capability of CRISPR to replace single nucleotides – such small changes are often the source of congenital diseases. “It opens up the DNA and edits bases without breaking the DNA strands,” he says.
Although the molecular techniques are advancing rapidly, the information needed to ensure that the changes made to the genome are safe still needs to be discovered. It is information that pharmaceuticals-development organisations are trying to assemble. “We have had to build our own bioinformatics,” Mayr says. “There is a whole range of technologies we have had to build to develop these techniques.”
As with previous medical discoveries, such as monoclonal antibodies and the original forms of gene therapy, cures based on advanced gene-editing technologies like CRISPR could take well over a decade to become medically accepted. But technologies are gradually providing the tools for the body to deal with disease.
In detail: Gene control
Most of the useful DNA in humans, animals and plants is there to control the actions of genes. Genes themselves provide the template for making proteins that are used in enzymes that perform the work of keeping a body going. The role of much of the rest of the DNA regulates the ‘expression’ of those genes. Such control ensures cells do not produce too much of each type of enzyme.
Biologists have found over decades that the regulatory infrastructure extends beyond the DNA itself into the so-called epigenome. Chemicals can attach to the outside of the helix and prevent some genes from activating. Even the way that the kilometres of DNA inside each cell nucleus are wrapped change the behaviour of the genome and how the cell develops.
The key regulator of each gene lies just upstream from it. This is the promoter. Each promoter, a specific sequence of DNA, can attach to proteins that latch on to that part of the chromosome and in turn either attract or deter the enzymes that transcribe the downstream gene into proteins. Promoters have become the target of synthetic biology research because their interaction with the transcription-factor proteins can be used to construct logic gates that turn cell activities on or off based on what is happening inside the cell. Such control could lead to the next generation of gene therapy.
David Venables, CEO of Synpromics, says: “Our thesis is that while people are focused on getting genes into the cell, very little attention has been paid to controlling the gene once it’s in the cell. You want to get the appropriate gene expression. It is not always the strongest expression that you need.”
Synpromics has conducted thousands of experiments on cells in culture with tiny changes to promoter sequences with the aim of generating synthetic promoters that can carefully control genes introduced into people, plants or animals. The company is performing experiments with a number of gene-therapy companies. They are designed to reduce the virus dose needed to obtain a benefit through greater promoter efficiency and to only activate in the right kinds of cells.