Editing the next generation

Chinese scientists recently announced the first genome editing of human embryos — but before we condemn these experiments, we should examine all the facts.

Illustration by  John Sharp

Illustration by John Sharp

The phrase “designer babies” has a horrific futurist ring to it, science turning the creation of life into something unnatural. It might seem like a far-flung future, but the creation of genetically modified children recently came a step closer when a team from China’s Sun Yat-sen University, led by Junjiu Huang, reported the first genome editing of human embryos.

Like it or not, we now have the technology to perform this kind of experiment; in vitro fertilisation (IVF) is relatively common and accessible, and genome editing technologies allow us to delve even further into rewriting what it means to be human. Whether we should be performing these experiments is another matter entirely.

IVF involves combining sperm and egg outside the womb prior to insertion, and is typically performed for couples who have had difficulty conceiving. Families with serious heritable disorders have used this process to select against embryos with the disorder, a type of selection that is relatively uncontroversial. Much more interesting is the ethical dilemma of couples wanting a child with something traditionally considered a disorder, such as two deaf parents who wish to have a deaf child.

If you accept the use of IVF, the line that tends to be drawn in the sand regarding “what we can do to children” is related to health. This is how Australian legislation functions: negative selection against embryos carrying a debilitating disorder is acceptable, but positive selection ‒ selecting for particular traits ‒ is not. Similarly, choosing embryo sex is only allowed when there is a risk of genetic abnormalities, although in some countries this law is more relaxed. 

The first IVF baby was born in 1978.   ZEISS Macroscopy/Flickr  (CC BY-NC-ND 2.0)

The first IVF baby was born in 1978. ZEISS Macroscopy/Flickr (CC BY-NC-ND 2.0)


Editing the genome takes us one step further than merely selecting for or against disorders. Should this technology become precise enough for use in viable human embryos, parents could theoretically tweak the genomes of children to their “best”. 

To edit the DNA of human embryos, Huang’s team used a technique called Crispr (clustered regularly interspaced short palindromic repeats), which was first used for eukaryotic cells in 2013. It works by harnessing the immune system processes of bacteria.

“Bacteria don’t fight viruses the way we do,” explains Professor Merlin Crossley, Dean of Science at the University of New South Wales. “Instead, they have anti-viral software that can recognise foreign sequences, or viral code. They can find the code and cut it out of their genomes – just as your computer might delete a computer virus.”

Crispr enzymes work by using a molecular guide, made from RNA, to target and slice up DNA in an organism’s genome. This RNA guide represents the ‘code’ and the enzyme does the ‘deleting’.

“We have now harnessed this technology to cut out bad genes from other genomes,” says Prof Crossley. “We can also use an inbuilt repair system to ‘swap in’ new genes like new computer programs.”

Huang’s team targeted a gene that, when mutated, causes a severe form of inherited anaemia. Despite being rejected by the major scientific journals Science and Nature due to ethical objections, the researchers did almost everything right. Tests in human cell lines (not embryos) suggested that the guide RNA was accurate and effective, resulting in almost no cuts away from the target site. There was no reason to believe that the embryos would be different.

However, not only was the success rate lower in the embryos, but the genome was altered in places outside the target sequence. Attempts to use another guide sequence to actively change the gene had a success rate of less than 15%. Even when an alteration was successful, only some of the embryo’s cells were affected.

Realistically, the low success rate of Crispr, and the difficulty in actually correcting a gene, could have been foreseen. A previous attempt to use Crispr to fix liver cells in adult mice with liver disease only altered 1 in 250 cells. Because liver cells grow like weeds when healthy, the growth of the altered cells nevertheless resulted in healthy mice.

Crispr exploits the immune system machinery of bacteria like  E. coli .  Source: NIAID/Flickr (CC BY 2.0)

Crispr exploits the immune system machinery of bacteria like E. coli. Source: NIAID/Flickr (CC BY 2.0)


The mixed results of these studies suggest that Crispr is a long way from being used in the clinic, something with which even the researchers would agree. Nevertheless, this most recent study has sparked extensive media discussion with varying amounts of accuracy.

First of all, Crispr wasn’t the first attempt at altering the human genome. Although its simplicity and inexpensiveness has made Crispr the current darling of the scientific world, before its development there were other ways to directly edit the genome.

The old-fashioned method of using viruses to insert a healthy gene copy into the genome has a long history of research, and has been investigated in adult humans to fix various diseases, including a form of adult-onset blindness and haemophilia. However, it's difficult to control exactly where a gene inserts into the genome, and errors can cause disruption of genes and side effects like cancer.

Transcription activator-like effector nucleases (TALEN) and zinc finger nucleases can also alter the genome. As both these methods require a pair of proteins to match and then cut DNA, there are often fewer off-target effects. Both TALENs and zinc fingers are more expensive and complicated than Crispr, but zinc finger nucleases have been used to delete HIV receptors from cells to make them immune to HIV.

Other forms of gene therapy have not been without their problems. “We had restriction enzymes, another bacterial anti-viral system,” says Prof Crossley. “It is true that restriction enzymes cut DNA too, but they pulverise it. It is like taking a book of Shakespeare and pulling out every page. 

“This was useful for a long time when it was all we had – researchers are good at searching through and identifying the correct pages of Shakespeare. But now we can go in and alter ‘to be or not to be’ to ‘to be and be better’.”

One group has used Crispr in mice with a dominant cataract disease. In dominant diseases, individuals are affected even if only one copy of the disease gene is mutated. This means there is often a second healthy copy, so making the mutated gene non-functional could effectively cure the condition. The relative simplicity of removing a harmful gene, rather than repairing a broken one, means that Crispr is more likely to prove useful for these diseases. Indeed, mice injected with Crispr at the embryonic level passed on their “cured” state to their offspring. 

Prof Crossley believes Crispr will eventually allow scientists to edit single “words” in the genome. However, no one is quite sure how many unintended mutations in the genome, known as off-target effects, will occur.

With advances in this technology being made on a monthly basis, even waiting another year could have greatly increased the accuracy of the embryo editing experiments. There are a few techniques that could rectify the current high rates of off-target effects. One technique uses two modified Crispr enzymes that only cut one strand of DNA. This is more accurate but might be less effective if your goal is to correct genes. Another study published in February this year found small molecules that increase both the efficacy and accuracy of Crispr genome modification.

Ultimately, however, the controversy surrounding this experiment hasn’t been about the experiment design or principled strength, but whether genetic alteration of human embryos should be happening at all. 

An unedited human fetus.  Source:  lunar caustic /Flickr (CC BY-SA 2.0)

An unedited human fetus. Source: lunar caustic/Flickr (CC BY-SA 2.0)


Prior to the announcement from Huang’s lab, there were rumours that similar experiments on human embryos had been conducted in US labs. Nothing more has been heard of these, especially not after a self-imposed moratorium of which many researchers have spoken out in support.

However, the Chinese experiments were perhaps less dystopian than the media uproar would suggest. Huang and his team used Crispr to edit embryos rejected from an IVF clinic. Ordinarily, only one sperm fertilises an egg, but these embryos were the result of two sperm fertilising an egg, producing cells with three copies of a genome. Had this occurred naturally, it would have resulted in either miscarriage or early death.

It is because these cells could not have resulted in a living, breathing human that the ethics is in a slight grey area. These experiments were first and foremost proof of an incredibly exciting concept.

Furthermore, while much has been written about the unprecedented level of control that embryo editing would give us, this may be placing too much importance on the role of genetics. Realistically, when a parent already controls the morality, education, and nutrition of their child, altering some chemical bases to provide them with blue eyes or type O blood isn’t that much more of a jump.

The real concern emerges when you realise that any alterations affecting the germline – so not just your offspring, but theirs as well – are likely to be whatever is valued highly by society at the time. There is a current trend in some Asian countries towards products and cosmetic surgeries that produce whiter skin and wider eyes. These trends are likely to move on, making decisions based on current ideals of beauty potentially harmful to future generations.

One of the big arguments against editing the germline is that future generations do not (and cannot) give consent to have their genomes altered. Somatic gene therapy, or gene therapy only affecting your body cells and not any potential offspring, is a much more palatable idea, and already used today for therapeutic purposes.

If germline editing were to become widespread, one can easily imagine a Gattaca-like society in which class divides become entrenched by genetic engineering. But is that likely? Crispr is already a relatively cheap and straightforward process, making it potentially available to all people in the future regardless of class.

At the very least, these Crispr experiments provide a wake-up call, alerting us that right around the corner is a world in which we must make some tough moral decisions. While science might appear to be the limiting factor, we as a society have a lot of issues to work through before we can safely wield any future genome editing technology. Let’s start there.