Two days after Anne Morriss took her newborn son home from hospital, she received a bone-chilling phone call. The stranger on the end of the line asked her whether she was sure her baby was still alive. Rushing to the next room, she was relieved to find the baby was fine, but the call was from a Massachusetts state physician who told her that a routine scan had revealed her baby had been born with a rare and often fatal genetic condition.
The condition, MCAD deficiency, is caused by mutations in a gene involved in fat metabolism. Some babies born with a severe version of the disease do not live for more than fortnight because their bodies cannot derive energy from fat by normal methods when their sugar stores run out. An infant with MCADD (Medium-chain acyl-CoA dehydrogenase) can simply sleep beyond the amount of sugar in his or her body, without an efficient way of converting fat into energy to keep the brain alive, says Morriss.
"That first year was a blur of anxious, sleepless nights. I worried that he wasn't going to wake up in the morning. I worried when he was sick with a cold and didn't want to eat," Morriss recalls. But because the condition was picked up so early, Alec now has as good a chance as any child to live a long and healthy life.
That was six years ago. But for Morriss those early days were the start of something bigger. When she looked into it, she discovered that her son had inherited the condition as a result of a coincidence. MCADD affects 1 in 17,000 people in the US, but the sperm donor that she and her partner had chosen, using a sperm bank, had, unknowingly, been the carrier of this rare genetic mutation. And so had she.
Conditions like MCADD might be rare, but there are many of them and taken together they affect millions of people. Now Morriss has come up with a way to minimise the chances of other parents having to go through the agony that she experienced.
Along with her business partner Lee Silver, a scientist at Princeton University, she is about to launch a company called Genepeeks that uses the DNA of sperm donors and recipients to create "virtual babies". These in-silica offspring can then be screened for hundreds of genetic diseases, before ruling out donors who could pose a risk. In the future, the team hopes to make the technology available to any couple trying to conceive.
The new technology – which they call Matchright – could be a gamechanger in reproductive health technology, allowing prospective parents to make more detailed analyses of disease risk than ever before, without even needing a real pregnancy from which to extract DNA. But it will also allow them an unprecedented glimpse at what their future offspring might be like and could theoretically be used to screen for other qualities and traits besides diseases. Falling outside regulations normally used to deal with embryo testing and screening, the new technology raises important ethical questions about privacy, partner choice and the role that computing will come to play in reproduction. "We are entering a whole new era," says Ronald Green, a bioethicist at Dartmouth College in the US, "an era where biology becomes information."
Logging on to the web to look for donor sperm can be eerily similar to doing the weekly food shop. A couple of mouse clicks on the London Sperm Bank website, say, brings up donor 1015. He is mixed race, has blue eyes and dark hair, a BA in theology and is a Christian. The "more info" tab reveals him to be well travelled, worldly, with a gift for carpentry. Once you have made your selection, simply click "add to cart" and proceed to check out.
But while you can chose height, eye colour, religion, education, whether they have freckles and even which TV series they like, when it comes to a donor's medical information, there tends to be little to go on. Blood type is standard and many sites also provide a report in the form of a questionnaire about the medical history of the donor's family. How reliable these are is unclear, however – most of us would probably have scant details about the medical conditions of our own parents, let alone grandparents, cousins and so on. The amount of genetic testing of donors varies but the clinics tend to screen for a handful of conditions – cystic fibrosis, for instance. "The tests with the most breadth go up to about 100 [diseases]," Morriss says.
But part of the problem is that it is not all about the donor – it also has to do with the recipient's genome. There are hundreds of rare heritable diseases but the number of people affected is relatively small – only 4% of the population is born with genetic diseases caused by mutations in single genes (rather than being affected by a number of genes, such as breast cancer) and, because the number of people affected is small, there are far fewer treatments. One in three children with rare diseases will not make it to their fifth birthday.
Many of these conditions – including MCAD deficiency – are said to be recessive, which means that a person will be affected by it if they inherit two faulty copies of the gene, one from each parent. Those who have only one faulty copy will not have any manifestation of the condition and are unlikely to know they are a carrier – as was the case with Morriss. If two people who are carriers reproduce, they have a one in four chance of the child being born with that condition. "Recessive disease risk isn't likely to show up on a medical history," says Morriss. "No one in my family had ever had MCAD deficiency. I was a silent carrier for the disease, and we happened to choose a donor who was a silent carrier."
It was bad luck. But it made her wonder if anything could be done and she began to teach herself about genetics through books and articles. In the meantime, Lee Silver, a molecular biologist at Princeton, had been working for more than three decades in reproduction and development, first with mice, then with humans. With the explosion in the field of human genetics in the last decade, along with computer power and computational tools, and a faster, cheaper generation of gene sequencing technologies, Silver was beginning to apply the principles he had been studying in mice to humans. At the same time Morriss was digging into the genetics behind her son's condition, Silver was, in 2008, beginning to realise how his work "could be put together in a very useful way – [applied to] sperm banks and that we could predict the risk for particular diseases in the hypothetical offspring of two individuals".
When Morriss and Silver were introduced by a mutual friend in New York, they immediately discovered their common ground and came up with the idea of Genepeeks. While Silver is the science behind the company, Morriss provides the business side – and the layman's touch.
In her words, "the technology simulates the genetics of reproduction and we literally make digital sperm and eggs and put them together to make digital babies. We then look at the disease risk that's showing up in those future children."
The company then uses this to give clients their own personal catalogue of sperm donors – "the client comes in and we make a bunch of digital babies with every single donor in our network and then we filter out all the donor matches where there is an elevated risk of disease".
The client, therefore, never gets a "test result" based on their genes or those of the donors, but a list of suitable matches. On average, the system rules out between 10% and 15% of donors, which shows that it is conservative, considering that only 4% of the population is born with these rare diseases.
Morriss makes it sound simple but for the technology to work it needs to pull off a couple of amazing tricks. For a start, it is not as simple as creating a single digital sperm and an egg based on the parents and putting them together. When an egg and a sperm fuse in real life, they swap a bunch of DNA – a process called recombination – which is part of the reason why each child (bar identical twins) is different. To recreate this process, the software needs to be run 10,000 times for each individual potential donor. They can then see the percentage of these offspring that are affected by the disease.
The Matchright system screens for more than 600 recessive paediatric conditions, says Morriss, compared with the small number normally tested for in sperm donors. She says that it is the first time in human history that these additional diseases can begin to be prevented. Together, the diseases screened for account for about 20% of infant mortalities and 18% of paediatric hospitalisations.
It is normally stated that if two people are both carriers for a recessive disease they have a one in four chance of passing it on to each child. But Silver says even the simplest diseases are much more complicated; disease and its severity depends on the combination of the two mutated versions. His system can model those subtleties, even though it will rule out any donor who conveys even the slightest risk.
To do this, the software pulls in data from publicly available databases that list known mutations for these diseases – currently more than 8 million mutations – and looks for all of them, in each digital embryo. Take cystic fibrosis, says Silver. There are 196 mutations known to affect the cystic fibrosis gene, CFTR. Matchright's test would look at all of them.
But the beauty of the new technology, says Silver, is that it also incorporates software that allows him to detect mutations that have never been implicated before and work out if qthey would be likely to lead to enough protein damage – because genes are the code for building proteins – to cause the disease. Going back to cystic fibrosis: "In addition to those 196 mutations, there are another 6,800 bases [components of the gene] that could be mutated, so we would have to analyse how risky they are if we found a person with any of those mutations."
The system will provide the most comprehensive genetic analysis to date of the potential risk of disease in a newborn, without even needing to fertilise a single egg. It gives people more confidence about disease risk, says Green, who is not involved in the work: "If someone I care for was in the market for donor sperm I might encourage them to use this technology," he says.
The whole system relies on the fact that research into mutations that cause genetic diseases are made publicly available. If new, peer-reviewed research shows that a new mutation might be involved, it will be added to the list that are screened for. "If there wasn't this worldwide community of public data resources, our company couldn't exist," Silver says.
Because this information feeds his technology, it also means that the more we understand about genetics, the more detailed his analyses can become. At the moment, Genepeeks focuses on genetic conditions from single genes, even if they can be caused by lots of types of mutations to that gene. But, says Silver, "we want our system to be ready for more complex diseases".
Take diabetes or heart disease, for example. These are complex diseases where there are many genes and environmental influences involved. "It's a lot more complicated," says Silver. "That's what we're working towards. The genetic databases are not sufficient yet to do that, but we want to be in a position that when the data is available, we can jump on that."
The idea that their technology might one day be used to screen for more complex traits does not sit well with everyone. Morriss's research has found that many sperm donor recipients want children who look like them. Green says that in the future he would not be surprised to see the list of traits people select with the service "go over from obvious harmful mutations, then hair colour, eye colour, athletic ability, skin colour", whose genetic influences we are beginning to understand.
Such an idea is not entirely new. Last year, a genetics startup, 23andMe, was granted a patent for what it calls its "family inheritance trait calculator" – a tool it bills as an engaging "way for you and your partner to see what kind of traits your child might inherit from you … and is used by our customers as a fun way to look at such things as what eye colour their child might have or if their child will be able to perceive bitter taste or be lactose intolerant. The tool offers people an enjoyable way to dip their toes into genetics."
However, when the patent was granted, there was a huge backlash, both in the media and from organisations dealing with reproductive issues in the US. For instance, in a press statement offered by the Centre for Genetics and Society in the US, director Marcy Darnovsky said: "It would be highly irresponsible for 23andMe or anyone else to offer a product or service based on this patent. It amounts to shopping for designer donors in an effort to produce designer babies. We believe the patent office made a serious mistake in allowing a patent that includes drop-down menus from which to choose a future child's traits." Eventually, 23andMe stated that it would not use its inheritance calculator in fertility treatments.
The Genepeeks technology is firmly based in medical applications for now, but that's not reflected in the patent, which includes a mind-boggling list of traits that have some component of genetic heritability – however small a part it plays. The list includes complex diseases such as cancers, stroke and asthma down to memory, hip circumference, BMI, nicotine dependence and eye and skin pigmentation. "The patent covers any disease or trait that has a genetic influence," Silver says. Thomas Murray, president emeritus of the Hastings Centre for bioethics research in the US, agrees that the list of traits on the patent goes way beyond medical traits, much in the same way the 23andMe patent does.
Because there is no embryo involved, Murray says these technologies distil the question into its purest form: "How much parental discretion is wise and good in selecting the traits in the offspring?"
Morriss says they are asked a lot if they are "in the designer baby business", but dismisses the question because people don't say they want the perfect child. "We hear, I want a healthy kid, and I want them to maybe look like me."
At the same time, though, the patent also includes a description of how "same sex and infertile couples will be able to simulate the genomic profile of their 'own' purely hypothetical child and match this profile to the ones created virtually with the selected donors".
Some might argue that it's normal for people, especially those using donor services, to want their child to look like them, but for others the idea of taking the randomness out of the process, of choosing a child based on appearance, is creepy or downright wrong.
These topics have been hotly debated in the past, in particular during the 1990s surrounding the use of preimplantation genetic diagnosis (PGD), which tends to be used by couples known to be a high risk of genetic disease, letting them fertilise a number of eggs in vitro and select the most viableone. That discussion was embedded with issues over the ethics of fertilising eggs only to discard them and what constitutes a life. This technology bypasses those issues, but has some of its own.
What is more, it also bypasses the regulation that was put in place by organisations such as the Human Fertilisation and Embryology Authority (HFEA) in the UK to regulate these kinds of practices. "If tests are carried out digitally, without an actual embryo being created, then this kind of testing wouldn't be classed as embryo testing under the terms of the HFE Act 1990 (as amended), and wouldn't be regulated as embryo testing by the HFEA. But presumably it will be subject to other forms of regulation and validation," a spokesperson for the organisation said.
Green says the idea of couples using this technology to select for superficial traits is not something that keeps him awake at night. "People have been making genetic choices on superficial information since the first person met another in a bar. This just adds a technical level to it."
What he is concerned about is what Morriss and Silver plan next – to move the technology from one that assists sperm bank users to any couple thinking about having a child.
While Morriss sees this as beneficial – future couples could check out the likelihood of their having a baby with a genetic disease, make a decision based on that, and potentially do something about it, for instance treat it early, as was the case with her son, or turn to PGD for peace of mind. "It's a really powerful insight for anybody thinking about bringing a child into the world," she says.
But Green points out it would take only a strand of a lover's hair to be able to go off and get a detailed glimpse of your future children. "In addition to liking somebody, you will run a genetic test on them, with or without consent. Are we going to allow people to be tested without their consent?"
Genepeeks is now working with two sperm banks – Manhattan CryoBank and Seattle Sperm Bank (also known as European Sperm Bank USA). The service will be offered at a fee of $1,995. New donors actively consent to participate in the programme. In the case of Manhattan Cryobank, those donors already on the books were notified of the programme and given the choice to ask questions and opt out. Consenting to some genetic screening is already part of the donor qualification process, although it's usually limited to just a handful of conditions.
The official US launch is in mid-April and then the team plans to expand it internationally. "Our plan is to work with international clients on a country by country basis, but we'll need to figure out the laws as we go, so we don't trip over anything," says Morriss.
As for the ethical issues, Morriss does not deny they are there, but believes in opening up the discussion "beyond the self-appointed ethicists". "I think everybody should be involved – the public and the scientists and the regulators."
And, of course, the companies. "We don't want to be in the perfect baby business. And for the most part nor do the people out there who want to be parents."
