The age of the red pen

IN THE summer of 2005 Karen Aiach and her husband received heartbreaking news about their four-month-old daughter, Ornella: she had a rare disorder known as Sanfilippo syndrome. The prognosis was that, from about the age of three, the disorder would gradually rob her of most of her cognitive abilities. She would probably develop a severe sleep disorder and become hyperactive and aggressive. She was unlikely to live into her teens; she certainly would not survive them.

The problem was that Ornella lacked a working copy of a specific gene. It is a gene that tells the body how to make a particular protein which is involved in clearing up cellular debris. Without that protein the cells of her body were unable to break down a complex sugar molecule, heparan sulphate. It is the build-up of that molecule in brain cells that lies behind the symptoms of the syndrome. If her cells could make that protein, the situation might, in principle, be reversed. Learning this, Ms Aiach embarked on a ten-year search for a way to correct the error in her daughter’s genome.

In almost every cell in Ornella’s body, as in every human body, there are two copies of the human genome, one from her mother, one from her father. In each of those genomes there are about 20,000 genes, each of which contains the recipe for a specific protein in the form of a sequence of chemical “letters”. To date, medicine has recognised about 6,000 diseases that can be traced to a problem with one or another of those genes—a disorder in which a missing or garbled sequence of DNA leaves the body unable to make a particular protein, or causes it to be made in an abnormal form. Some of these single-gene disorders are well known: Tay Sachs; sickle-cell anaemia; haemophilia. Others, such as Sanfilippo syndrome, are the sort of thing you learn of only when a child you care about turns out to be the one in 70,000 that it afflicts.

Since genetic engineers assembled their first tool kits in the 1970s heart-broken parents and medical researchers have longed to use such technologies to fix these faulty genes. The first clinical attempts at such “gene therapy” began in the 1990s, with viruses used to add needed genes to cells that lacked them. But this was crude stuff. The new genes could not be guaranteed to slot into the right place in the genome; this often meant they did not in practice produce much protein; it also meant there was a risk that, by disrupting other genes, they could cause cancer. There were indeed cancers in some early trials; there was also a case in which a patient died of a lethal immune reaction to the virus used to carry the gene.

Powered by the desire to do something for children like Ornella, the gene therapists have soldiered on. And in the past few years they have found themselves helped by the most impressive piece of kit yet—a system called CRISPR-Cas9.

Some years ago, biologists discovered an odd feature in the genomes of some bacteria that they described as “clustered, regularly interspaced short palindromic repeats”—CRISPR for short. Bacteria use them to make little bits of RNA, a molecule that can store sequences of letters like those that make up genes in DNA. A CRISPR RNA will bind to a piece of DNA that has a complementary sequence. A protein called Cas9, which is a sort of pair of molecular scissors, recognises the structure made when a CRISPR RNA binds to a piece of DNA and responds by cutting through the DNA at precisely that point (see diagram).

Beating bugs into scalpels

Bacteria make CRISPR RNAs that recognise the DNA of viruses which prey on them, marking that DNA for destruction by Cas9 and thus protecting the bacteria from infection. Scientists can make RNAs that target any sequence they want. And because of the way that cells repair broken DNA, if they put a new gene into a cell along with the CRISPR-Cas9 system, they can get that new gene to replace an old one. The effect is to give scientists something that works like the find-and-replace function on a word processor.

Because it is so simple and easy to use, CRISPR has generated huge excitement in the worlds of molecular biology, medical research, commercial biotechnology—and gene therapy, where it may make it possible to make changes with profound consequences. To date gene therapies have been designed to fix everyday sorts of cells, such as those of the blood, or the retina, or the pancreas. CRISPR makes it possible to think about aiming at the special cells that make sperm and eggs, or the genome of a fertilised embryo awaiting implantation in the womb. In either case the changes made would pass from one generation to the next, and the one after that, in perpetuity.

This sort of “germ-line” editing is widely seen as a bourn no ethical traveller should cross. Some scientists and research organisations want a moratorium on any work aimed at engineering the germ line; others say basic research on such things should continue, but any moves to use it in the clinic should at the very least be widely debated by society as a whole. America’s National Academies of Science are convening a gathering in December to look at the options. Genetics is a peculiarly personal science, but it is also one very prone to politics. The power of CRISPR looks sure to exacerbate that propensity.

When Jennifer Doudna of the University of California, Berkeley, Emmanuelle Charpentier, who is now at the Helmholtz Centre for Infection Research in Germany, and colleagues worked out how to turn the bacterial CRISPR system into a genome editor three years ago there were already two other techniques for making specific and precise changes to genomes. But the other techniques were time-consuming and often finicky. The new technique was as good if not better, and far quicker and easier to use. Matthew Porteus, a pioneer in gene editing at Stanford University, says research that required a sophisticated molecular biology lab three years ago can now be done by a high-school student.

All of the species, all the time

By the beginning of 2015 the regular analysis of “hot” research in biology put out by Thomson Reuters, which looks at what papers are being cited most by other scientists, had three CRISPR papers in its top ten. The technique has been applied to dozens of species, including zebrafish (much favoured by developmental biologists), yeast, fruit flies, rabbits, pigs, rats, mice and macaques—the first primates to be genetically engineered with the technique. It has been used to cure mouse versions of muscular dystrophy and a rare liver disease. Ways have been found to make the technique more reliable, more versatile and less likely to make cuts where it is not supposed to; further improvements are on the way, not least at the startup companies built around the technology.

One of CRISPR’s great attractions is that it can be used to introduce, or remove, a number of different genes at a time. Most disorders are not caused by just one gene going wrong; being able to manipulate many different genes in a cell line, plant or animal opens new avenues for the study of conditions such as diabetes, heart disease and autism where a number of genes are involved, along with the environment. In the past a mouse with as few as three genes knocked out would have taken as many years to create; now it can be done in three weeks.

CRISPR is also letting researchers get more out of other technological breakthroughs—notably the ability to make stem cells which can then be turned into the cells typical of any sort of tissue. George Church at Harvard is using CRISPR to edit the genomes of stem cells before turning them into nerve cells, so as to find the mechanisms behind a range of neurological disorders. Feng Zhang, a scientist at the nearby Broad Institute, has been using CRISPR to model Angelman syndrome, a neurological disorder.

Previous genetic-engineering technologies have tended to be species-specific; there have been lots of tools for manipulating E. coli and yeast, but they have often not been broadly applicable. This is another area where CRISPR excels; it can be used in organisms that have resisted previous attempts at engineering. This could be a big help in agriculture, spreading modification techniques to new grains, roots and fruits—Monsanto has already begun working with CRISPR to design plants with useful traits. Another biotech application is the use of CRISPR to build a “kill switch” which allows any genetic modifications made to bacteria to be removed after they have been used, either for safety or to protect intellectual property.

One particularly impressive—and potentially worrying—application is in the creation of genes that can spread themselves quickly through a population with blithe disregard for the constraints of natural selection. Engineering the CRISPR-Cas9 system itself into a creature’s genome makes it possible for an organism to edit its own genes, and there are ways that this ability can be used to “drive” a gene through a population (see article). Such a technology might, proponents say, be used to make the mosquitoes that carry malaria, or dengue fever, unable to spread the organisms responsible for causing the disease.

The applications seem limited only by the imagination. Dr Zhang says CRISPR has enormous potential for treating previously intractable diseases. For example, genome editing may make it possible to eliminate viral infections within the body, creating entirely new antiviral treatments. He also speculates that it might be possible to make red meat that is less harmful, or to engineer pig organs so that they could be transplanted into humans with much less risk of rejection. Dr Church, for his part, has speculated about using gene editing to turn elephants into mammoths—or to recreate Neanderthals.

There has been a flurry of commercial activity and investment. Large pharmaceutical companies are eyeing the technology for research. AstraZeneca has plans to use it in cell cultures to explore the function of every gene in the human genome. Among the startups, Caribou, which was founded by Dr Doudna in 2011, has raised $11m in funding and will focus on cell engineering for drug screening, agricultural and industrial biotechnology. Caribou has also formed, with pharma company Novartis and a venture-capital firm, a startup called Intellia. With $15m raised in 2014 Intellia will focus its work on gene therapies in which cells are taken from patients, edited and put back.

Crispr Therapeutics, co-founded by Dr Charpentier in Switzerland, which has raised $25m, is aiming at a similar market, as is Editas Medicine, co-founded by Dr Zhang. In early August, Editas raised $120m from a group of investors that includes Bill Gates. This comes on top of $43m the company raised in 2013. Although Dr Doudna and Dr Charpentier filed the first patent for CRISPR’s use in gene editing, Dr Zhang was granted a patent on its use in plants and animals after his institution paid for an accelerated review. This would seem to give him and the Broad Institute control over the key commercial uses of CRISPR in humans and research animals. The applicants for the other patent are challenging the ruling.

The easiest sorts of gene therapy will be those that can be done outside the body—ex vivo, in lab speak. The appeal of ex vivo work is the level of control; cells can be extracted, have their genes manipulated, and have their new genes tested before being put back. To see the sort of things that this makes possible take a look at the work being done by Sangamo Biosciences, based in Richmond, California, which has been working for a decade on an earlier, more cumbersome gene-editing technology that makes use of what are known as “zinc fingers”. It is trying to apply that technology to beta-thalassaemia, sickle-cell disease, haemophilia and HIV infection.

In clinical trials of its HIV treatment, Sangamo takes the immune cells that the virus infects out of the patient’s bloodstream and edits in a mutation that makes them highly resistant to infection. It then grows up a large number of the edited cells and infuses them back into the patient, where it is hoped they will flourish. A similar sort of approach can also be used in blood disorders such as beta-thalassaemia and sickle-cell disease which are caused by mutations in the globin gene. The idea is to extract blood stem cells from bone marrow, edit them so as to switch on the production of fetal haemoglobin (which the body stops producing shortly after birth, even if it cannot make the adult stuff) and return the stem cells to the body. It would be like a bone-marrow transplant—except that since the new genetically improved cells come from the patient’s own body there is no danger of rejection.

Similar ex vivo approaches could make gene editing a powerful tool for fighting cancer. A currently promising approach to the disease is to retrofit the immune system’s T cells with what is called a chimeric antigen receptor (CAR)—a protein that recognises tumours. This CAR-T approach is likely to evolve as CRISPR makes it possible to add more, or subtler, genetic changes to the T cells. Given the ease and speed with which RNA guides can be designed and tested, it seems only a matter of time until T cells are tailored to mutations specific to a particular patient’s cancer.

With blood cells ex vivo approaches work fine, and they may have applications in other diseases, too. But when it comes, say, to a brain disease there is no way to take the cells out, fiddle about and put them back. Instead you have to deliver your molecular editing suites to the cells where they live—to do the editing in vivo. So far attempts at therapeutic in vivo gene editing have been limited in scope. Sangamo has done a little work in mouse brains, where it has been able to repress the expression of the gene that causes Huntington’s disease. Intellia has plans to look at in vivo applications that include diseases of the eyes and nerves, as well as haemophilia and some infectious diseases.

The easiest in vivo applications of gene editing will be diseases where the damaged cells are easy to get at—for example diseases of the eye. But gene-therapy companies also have strategies for getting at harder-to-reach cells, with years of work that could now be applied to the delivery of gene-editing packages. Take Lysogene, the company Karen Aiach founded after her daughter’s diagnosis with Sanfilippo syndrome. It has a viral vector which, injected directly into the central nervous system, puts copies of the gene that children like Ornella lack directly into brain cells.

And then there is the most controversial form of editing—editing the genome of a newly created embryo, or of the cells that produce sperm and eggs. If this could be done safely it would offer the possibility of acting once and for all. By changing a gene in an early-stage embryo, or in the cell that makes an egg, you could ensure that the change is found in every cell in the adult body—including its own eggs or sperm, which would pass it to the next generation and thus on down through the ages. No one is pursuing such avenues in the clinic as yet. But the announcement in April that a Chinese group had engineered changes into non-viable human embryos as part of their research into beta-thalassaemia set alarm bells ringing

Even before that a group of scientists, which included the boss of Sangamo, had published an article in Nature calling for a voluntary moratorium on all experiments involving germ-line modification. The Centre for Genetics and Society, a non-profit in Berkeley, California, that supports responsible use of genetic technologies, opposes using CRISPR to conduct even basic research on embryos. It says that the prospect of people modified in ways that would be transmitted to their children raises grave safety, social and ethical concerns, running the risk not just of producing children with unforeseen difficulties because of side-effects but of opening the door to new forms of social inequality, discrimination and conflict.

Dr Doudna and a number of other eminent molecular biologists, such as David Baltimore of Caltech, have called for scientists to avoid any attempts at human germ-line modification, even if they are in countries where regulation might allow it, before there has been a much fuller discussion of the implications. This would not preclude using the technology for research purposes on embryos created as part of an in-vitro fertilisation programme and not intended for implantation (in Britain and a number of other countries such research is allowed on embryos up to 14 days old). Dr Baltimore was part of a group that called, in 1975, for scientists to refrain from using some of the earliest tools of genetic engineering until rules had been established; that moratorium is often touted as a worthy example of scientists thinking a new technology’s implications through before running into a thicket of practical and philosophical issues.

Crossing a line, again

Francis Collins, who runs the National Institutes of Health, America’s main government funder of biomedical research, said in April that altering the human germ-line for clinical purposes is viewed “almost universally as a line that should not be crossed”. However this may not be strictly true. Mitochondrial DNA donation, an in-vitro fertilisation technique that replaces a specific form of defective DNA from the mother with equivalent DNA from another woman, recently became legal in Britain. Like changes produced by editing the genome of an egg or early embryo, the effects of this donation will be passed on to future generations.

The issues surrounding mitochondrial DNA donation were widely discussed in Britain, and the procedure voted on in parliament. The conclusion was that the risks were small and that helping people carrying certain diseases to have healthy children mattered more than rather formless worries about “playing God”. A debate on the merits of using CRISPR for germ-line engineering in cases where there was no alternative might reach a similarly permissive conclusion. But that would depend on a number of factors.

On the technical front, CRISPR, though good, is not perfect—it can make cuts that are not desired as well as the ones that are. In research it is fine just to work with the cells and animals that came out perfectly; in the clinic you need a lower error rate. In germ-line editing, when any errors will end up in every cell in the body, the problem is particularly worrying. What is more, in most cases where there is a risk of genetic disease it will be safer, when using in-vitro fertilisation, to choose an embryo that does not have the defect than edit one that does. Only when there are a number of genes to worry about would editing seem a plausible option.

Worries about germ-line editing are fascinating, and the discussion they produce may prove important, divisive or both. But they are far from the realities of today, where genetic disorders that CRISPR might make amenable to uncontentious forms of gene therapy are destroying lives and parents are fighting to save their children.

The gene therapy Ornella eventually received from Lysogene came too late to prevent her cognitive decline. But she smiles and she is gentle, and her nights are almost normal—an improvement over her earlier years that Ms Aiach puts down to the therapy. In decades to come, the prognosis for others like her will almost certainly improve. However much the well worry about the nefarious applications of gene editing, the needs of the sick will continue to drive science and medicine forward—as they should.