Frequently Asked Questions
We often see the same old arguments being bandied about in discussions of human germline genetic modification. For example, every online discussion seems to have at least one reference to GATTACA, vague fears of unknown consequences, or giving children functional wings. In addition, influential and otherwise-knowledgeable people have asked us some of these questions and it seems there are a lot of common misconceptions about the science. Below, please find our responses to such topics.
- Technical Questions
- What is a designer baby?
- Why aren’t you using this to cure hereditary diseases?
- What about using it to cure infertility?
- Why not wait for somatic gene therapy?
- What’s wrong with CRISPR microinjection?
- Isn’t every single human trait controlled by thousands of genes, so no intervention you make will have a significant effect on the child?
- Will you be performing animal testing?
- Why not reprogram male stem cells into germ stem cells, insert them into the testis, and have them produce sperm that can be collected for natural IVF? You could even make them kill and replace the already-present germ stem cells so the patient’s sperm is 100% transgenic.
- What about in vitro gametogenesis?
- Why not iterative embryo selection, a la Nick Bostrom’s proposals?
- Why not sperm-mediated gene transfer (SMGT)?
- What about testis-mediated gene transfer (TMGT)?
- How about somatic cell nuclear transfer (SCNT)?
- What are the shortcomings of embryo selection by PGD?
- What amazing enhancements will eventually be possible?
- How could you ensure that a genetically modified child will be able to have unmodified children themselves?
- Ethical & Legal Questions
- How much will it cost? Won’t the rich create a generation of superchildren who will take over the world and plunge us into a plutocratic dystopia?
- Won’t children born with enhanced genetics experience stigma from their peers and pressure from their parents?
- How is modifying an unborn child’s genome ethical, when they can’t consent?
- How can you ensure that the procedure is safe?
- Humans don’t understand biology, why do you think you can improve on nature?
- Why aren’t you an elite, well-funded team of top embryologists and stem cell biologists?
- Why not wait until a societal consensus is reached, a legal framework is in place, and the biology has been perfected?
- Is all this legal?
- Won’t this create a ‘genetic arms race’ between countries seeking to enhance their own citizens?
- But GATTACA?
- Why try to improve the species, when humans are perfectly adapted by evolution (or created by a deity)?
- Isn’t this eugenics?
- Why are you promoting this work out in the open, rather than in a secret underground or offshore lab?
- What is our vision of the ideal outcome?
- How will we prevent couples from freezing eggs and waiting for new biotechnology breakthroughs to give their child an advantage?
- What if people want to perform cosmetic or other marginal changes to their child?
- Won’t gene editing affect designer babies’ “human nature”?
Technical Questions
What is a designer baby?
A designer baby is any human that has had their DNA edited at a very early stage in development. This is in contrast to embryo selection, where parents choose one or more embryos from among the handful that have been conceived during IVF, based on biochemical and statistical analysis of each embryo’s genome. It is also in contrast to gene therapy, where a human is injected with an engineered virus that delivers a small segment of DNA to be incorporated into some cells of the body.
The process for parents desiring a designer baby is identical to that of infertile couples looking to have children by IVF. Parents donate their respective germ cells (eggs and sperm) which are combined in the clinic to create embryos. The genetic counselors and embryologists at the clinic discuss potential genetic interventions with the parents, who select genes and traits that they believe will give the best quality of life to their future child.
After the parents give their informed consent, the cells inside the previously-created embryos are extracted and grown in vitro in the clinic, where they are monitored until there are enough to perform the gene editing. Proteins, DNA and RNA are added to the cells, performing the edits selected by the parents. The cells are grown for a while longer, during which time they form “colonies” that have each descended from a single, edited cell.
Each of these colonies has a few cells removed, which undergo genome sequencing and karyotypic analysis to ensure that they possess the chosen edits, and are free from any off-target or otherwise dangerous mutations. Then a few other cells from the colony, all genetically identical, are injected into another early-stage embryo. These injected cells coax those already present in this embryo to form the baby’s placenta, while the edited cells go on to form the fetus. At the clinic, this embryo is implanted into the mother’s uterus, and from there it proceeds exactly like any other pregnancy.
Why aren’t you using this to cure hereditary diseases?
Preimplantation genetic diagnosis (PGD) already allows us to prevent nearly all diseases from being passed on to the next generation. While it may not completely eliminate a recessive allele from the next generation, it ensures that the child will not suffer from the disease. See here for additional arguments.
In addition, multiple other laboratories are already investigating the use of germline gene therapy in cases where PGD is ineffective, e.g. beta thalassemia. They have the backing of patient advocacy groups, and correcting single mutations like this can be accomplished well enough with simpler methods.
Certainly, our approach is applicable in cases where parents are carriers of multiple diseases; as parents have perhaps two dozen embryos to choose from, the probability of finding one that’s completely untouched approaches zero. While this is an intriguing avenue, each unique combination of alleles requires a unique approach to fix.
We prefer to be completely honest about our intentions - while the debate still centers around what qualifies as an ‘acceptable’ use of germline gene editing, it is inextricably linked to enhancement. As such, we must embrace the association. Correcting or preventing hereditary disorders could already vastly increase the number of quality-adjusted life years in the future, but genomic enhancement has even more potential.
What about using it to cure infertility?
True, society seems to have agreed that a parent who is naturally infertile (i.e. without the aid of IVF) is free to potentially pass down this condition to their children. It is likely that many of the first people to elect for germline genetic modification will be infertile couples, because they are already committed - physically and financially - to the rest of the IVF procedure.
However, in terms of what we would be able to do - “fixing” a mutated gene is significantly harder than inserting a functional copy. Regulators and bioethicists, who are barely in favor of germline modification to begin with, are unlikely to accept the ‘easy’ route - inserting a brand-new replacement gene. Depending on the infertile parent’s genome, we may be able to include a cure.
Also, the question of whether the effort in identifying the mutation, determining a remedy, and applying it will be cheaper than planning for IVF costs in the next generation remains to be seen. In the vast majority of IVF cases, the underlying genetic cause is not investigated, if it is indeed a loss-of-function mutation and not a compatibility issue between the parents’ gametes.
Why not wait for somatic gene therapy?
There have been some advances in gene therapy, but it still suffers from several major drawbacks.
Cost: Much of the price tag of new gene therapies (millions of dollars) results from their limited patient pool, often targeting the ‘orphan’ diseases that are considered ‘low-hanging fruit’, i.e. relatively easy to fix with a virally-delivered gene. There are very few of these disorders that are both amenable to gene-replacement therapy, and have been studied well enough to determine their cause. Expenses for R&D, and pharma executives’ new luxury cars, must therefore be spread over a smaller patient pool. The cost of the previously standard treatment (e.g. protein transfusions or hospital care) can often cost thousands of dollars per year, making the single upfront investment in gene therapy worthwhile for patients - though often a single injection isn’t sufficient.
Certainly with a larger target population, R&D costs per head will lower, and economies of scale come into play. Whether this will be enough to make it affordable remains to be seen. The closest analog we have, recombinant protein therapies, are still laughably expensive even 30 years after their introduction. One can again argue that these therapies target only a small population, e.g. those suffering from a particular type of cancer, but even drugs with a broad benefit, such as those targeting PCSK9 to reduce heart disease, are still well into the 5-figure dollar price per year.
Effectiveness: Gene therapy will have great difficulty reaching every cell in the body, especially the brain or stem cells. Delivering any therapeutic to the brain requires state-of-the-art medical facilities - even if the gene therapy virus could be grown in a 5-gallon bucket, the injection is literally brain surgery. Stem cells, too, exist in well-protected niches throughout the body, understandably shielding them from pathogenic agents or the various torments of environment. Over time, the non-stem cells that take up the gene therapy will reach the Hayflick limit and succumb to turnover.
Gene therapy viruses are also limited in their payload capacity and the reliability of their genome integration targeting. Perhaps a perfect, engineered virus that is able to bypass the host’s immune system and deliver the gene with 100% efficiency lies in our future, but we have decades before that’ll happen. That’s a long time to be twiddling our thumbs. The field certainly stumbled at first, but it’s not exactly sprinting ahead these days.
Safety: As noted above, the immune system presents a significant barrier to routine gene therapy. Most people already have a pre-existing immunity to the common viruses used (e.g. AAV), but even with a new virus the immune system is remarkably good at keeping foreign materials in check. Indeed, an engineered virus that is able to completely evade the immune system is a terrifying prospect, far more so than a flu virus tweaked to add a little more virulence, as it could wreak havoc previously unseen in human history.
Today’s gene therapy requires heavy immune suppression to even get the virus safely into the body, or at least to prevent the patient from developing a massive immune response on re-dosing of the therapy. Gene therapy is already considered a single-shot treatment, even with the best immunosuppression medicine has to offer. And even if you flood the body with gallons of pure virus in that one attempt, you’re never going to treat all the cells.
But the most critical aspect of immunogenicity is that the body will react to the delivered transgenic protein. If a person suffers from a disease because their body doesn’t produce a certain protein, when you introduce that protein to their body the immune system will see it as foreign. To their white blood cells, that livesaving protein is no different from an invading pathogen or toxin. Unlike the reactions to the gene therapy vector, which exits the body in a matter of weeks, the therapeutic protein will continue to be produced in the body for years, a tempting target for white blood cells.
Now, you can work around this by supplying a continuous dose of anti-rejection drugs, or hoping that the body will develop an immune tolerance. But those drugs come with myriad side effects, not least of which is an increased susceptibility to infectious disease. And immune tolerance is no sure thing; we still have no reliable way to get the body to ignore a foreign protein, and it can still decide one day to become intolerant and begin a campaign of destruction. Hell, the rampant incidence of autoimmune disorders is enough of a warning: sometimes our bodies will attack themselves, targeting proteins that they have no reason to.
We must still be thankful that our bodies act this way - the ability to detect, neutralize, and protect from any conceivable manner of virus or bacterium is one of the most stunning displays of biological ingenuity known. Those autoimmune disorders, as mentioned elsewhere in the book, are just an inherited overreaction from our ancestors, who had to survive daily onslaughts of tuberculosis, cholera, tapeworms and others, without the benefit of modern medicine. With no enemies to fight, the immune system begins to turn on its host. But if we can dial down this aberrant response just a little, we can begin to thrive in a modern world without eczema, IBD, multiple sclerosis, and many other diseases.
For limited applications - targeting a single organ (usually the liver), or producing a circulating enzyme or hormone, gene therapy will have its niche. Using it to improve the quality of life for otherwise healthy individuals is not happening any time soon, especially not for the masses.
What’s wrong with CRISPR microinjection?
Microinjection into a single-cell embryo (as used by Jiankui He) is a mess of a procedure that has no long-term potential.
First, there is no way to check the outcome of the mutation in 95% of the embryo. You can sample a small fraction of the cells without killing the embryo itself, and are thus hoping that statistics will indicate whether the modification performed as expected, without a dangerous off-target mutation hidden in one of the remaining cells. The cells extracted from an embryo to perform PGD are also not a truly random sampling, as they are not picked individually at random from throughout the inner cell mass, thus disallowing proper statistical analysis.
Second, as above, you have no idea or guarantee whether all the cells have been modified (mosaicism). The importance of this can vary, but in many cases, an organ or tissue that’s only partially modified isn’t enough to prevent a disease or provide an enhancement. It can also be difficult to determine which cells and fractions of tissues have been modified, if not impossible. Replacing e.g. a mutated liver enzyme only works if the liver is formed from the pool of modified cells. In limited cases, like production of a secreted enzyme, this may not be a major issue, but otherwise it is a strong downside.
Third, performing more than a single knock-out by microinjection becomes exponentially harder. Multiple changes, or knock-in of a beneficial gene, have a vanishingly small chance of working. For example, with 75% efficiency, the success rate of performing five simultaneous mutations drops to 24%, while multiplying mosaicism questions five-fold. Knock-in, which is necessary to deliver beneficial genes rather than simply ‘fixing’ one nucleotide, has far lower efficiency due to the size of the insert, which must travel to the nucleus through Brownian motion. With enough eggs available (not that humans have more than a few dozen to work with) this doesn’t matter as much, hence why it is commonly used in livestock - but the previous problems remain.
In conclusion, CRISPR microinjection has received outsized attention, with its side effects (listed above) being conflated with designer babies in general. A faux press release from 10 years in the future by Jamie Metzl still assumes “up to three single gene mutations” being made. It also laughably assumes PGD (already being performed today by e.g. Genomic Prediction) will cost $5 million, which is a worrisome mischaracterization from someone who is on the WHO’s “expert advisory committee” for human gene editing.
The reason JK He targeted CCR5 was not some core belief about HIV proliferation. It’s just one of the only genes we can knock out to a positive effect, and CRISPR microinjection is only efficient when performing a simple knockout. Even simple myostatin knockout presents obvious downsides, i.e. weak connective tissue. In order to make human germline modification worthwhile, we need a more advanced technique.
There are many teams working to improve the efficiency and capacity of microinjection, and we do not mean to diminish their achievements and ongoing work; but it is fundamentally hamstrung.
Isn’t every single human trait controlled by thousands of genes, so no intervention you make will have a significant effect on the child?
Certainly, complex traits (height, intelligence, personality, appearance, athletic performance, immune system etc) are affected by many genes. But there are cases where even a single base pair change can have a dramatic influence. Refer to George Church’s list of protective alleles for a partial synopsis. What we are targeting are those few dozen mutations that have been exhaustively shown to have an outsized impact on a given trait, either in an extant human subpopulation or in many animal studies.
Every month, new mutations are being discovered and characterized from unique families that provide them with a tangible benefit, such as requiring less sleep while being otherwise normal. These discoveries have little practical use for adults already born, but by editing an embryo they can be easily introduced into the genome.
Some argue that e.g. introducing a CCR5 knockout into an East Asian individual is dangerous enough to be an argument against Jiankui He’s experiments - purely because we can’t assess the risks involved. These people would of course have no arguments against a European and Asian parent having a child together, with the same result, but they are nonetheless reminiscent of anti-miscegenation defenders in the 20th century.
Potential enhancements that have only been studied in animal models are, naturally, more risky. Indeed, a genetic tweak that makes a mouse better at remembering where the cheese is is unlikely to translate well into humans. Cognitive (and especially “moral”) enhancement will have to be carefully considered, as suitable models simply do not exist. But, as genetics is strongly linked to IQ, there is room to work.
Outside the minefield of neuronal modification, animals (especially primates) can be an excellent test of whether an allele is protective against cancer, or whether a new gene can neutralize some toxin. Natural selection is painfully slow. Worse, it is brutally unfair - and those who argue against that point are either lucky, or misunderstand biology.
Will you be performing animal testing?
To some degree, yes: refinement of the technique in mice will provide insights without having to expend valuable human eggs. However, in other mammals - dogs, cows, pigs, etc. - it becomes more difficult. The research on in vitro culture of embryonic stem cells is focused almost entirely on mice and humans. Mice are the primary testbed for mammalian cell biology techniques, so protocols and reagents are plentiful. For humans, growth of ESCs is routine in studying disease and developmental biology, and protocols to keep human embryos alive and healthy have been developed over 40 years of IVF.
In contrast, researchers have had little reason to investigate ESC culture (or IVF) for other animals; and both of these need to be reliable for our testing purposes. Transgenic animals are far easier to generate using cruder techniques, both because the eggs are practically unlimited in number, and because people are mostly apathetic to the ‘failures’ that may occur. The plentiful supply of human eggs and embryos, left over from IVF cycles and sitting in ultralow freezers, are often ultimately discarded. Thus, as human testing is the most relevant, we prefer not to spend 5-10 years developing protocols in other mammals as an intermediate form of research.
Why not reprogram male stem cells into germ stem cells, insert them into the testis, and have them produce sperm that can be collected for natural IVF? You could even make them kill and replace the already-present germ stem cells so the patient’s sperm is 100% transgenic.
Every step in that procedure is still theoretical. Reprogrammed cells, even if they were derived from adult stem cells, still possess countless mutations - many benign, but some, especially in genes that have been deactivated during differentiation, deleterious. Aside from concerns around the reprogramming process, adult somatic cells are still a risk. Yes, sperm stem cells have resulted in sperm that conceive children even at 70 years of age. Even if you manipulated those cells, other concerns remain.
Programming the cells to seek out and destroy the other, unmodified cells in the testis is still fiction. Some limited studies have demonstrated a similar technique in a different tissue, but the amount of engineering required will, again, take decades of research, and we are not content to twiddle thumbs for that long.
Other concerns include not being able to modify the mother’s chromosomes (aside from also-theoretical gene drive approaches), the potential cost of the entire procedure, invasive extraction of testicular stem cells (as will likely be necessary), and the risk in having all germline cells in the testis killed off. A high price to pay to keep Catholics on the cutting edge of our transhuman future, or others who believe IVF is immoral.
What about in vitro gametogenesis?
This is the only other promising technique on the horizon, perhaps only a few years away from yielding sperm and/or eggs that are functionally equivalent to their natural counterparts. Assuming we can genetically modify them with similar ease to embryonic stem cells, they will then meet all necessary requirements for having a child from ‘natural’ IVF (mixing the sperm and eggs in a dish, followed by implantation), a far less costly technique than ICSI.
The potential drawbacks are: sourcing un-mutated cells from both parents (see previous question for the problems inherent there), expensive or finicky culture conditions for the cells (likely a 3D tissue-simulating matrix), and ensuring that the reprogrammed cells are faithful replicas of their naturally-occurring equivalents.
No one has succeeded in growing a full, functioning organ in vitro from programmed cells, and while germ cells are arguably the closest we have epigenetically to embryonic cells in the adult body, it will take quite some time to perfect. Scientists have not successfully generated fully mature germ cells from adult tissue in vitro, even in mice. Additionally, there are no ‘egg stem cells’, which places limits on womens’ access to the technology.
Why not iterative embryo selection, a la Nick Bostrom’s proposals?
A few reasons, but first a background: Bostrom suggests growing embryos in vitro and, using in vitro gametogenesis, generating sperm/egg cells from each embryo, and ‘breeding’ them together, repeating the cycle hundreds or thousands of times. This would be guided by SNP analysis, based on beneficial alleles identified by population-wide genome sequencing and correlated with certain markers (intelligence, or others), and reduce the generation time from ~25 years to a matter of weeks.
First, generating germ cells from embryos is still speculative, though we can envision it in the near future. Whether these cells will accumulate mutations or epigenetic aberrations after extended in vitro culture remains to be seen - already, embryos used in IVF are less robust than those from natural conception. Though still vanishingly rare, such concerns will be amplified to worrying levels after thousands of cell divisions.
Second, the genomic references used in selection of the embryos is highly speculative. Genome-wide association studies (GWAS) used to indicate whether an allele has an effect on a trait often break down upon further analysis, especially when one considers the relations between each allele and every other one. For example, two alleles that each appear to boost intelligence alone, when combined, could have a null or even opposite effect. The effect of combining all beneficial mutations in one individual is based on pure speculation.
Third, a benefit of sexual reproduction is that each generation is born and reaches adulthood before potentially passing on its genes. After a thousand simulated generations, we have no idea how the child will actually act. Evolution is indeed slow, but microevolution, even over a few generations, can still have profound effects. One need only look at selective breeding of dogs. We could predict a collection of alleles that boost intelligence and combine them in one embryo, but what if the resulting adult is incapable of speech, suicidal, epileptic, or contracts dementia at age 30?
Fourth, enrichment of specific alleles does not mean that they are the only DNA being selectively passed on to the next generation. Thousands or millions of basepairs on a chromosome near the SNP being assayed will get carried along with it, leading to a severe reduction in genetic diversity among the theoretical children. Sadly this is intentional - very few of the SNPs are directly causative of the traits they signal, but instead are only correlated with a mutation somewhere nearby that has the desired effect.
Why not sperm-mediated gene transfer (SMGT)?
If it sounds too good to be true, it probably is. SMGT was presented as an astonishingly simple way to perform genome modification: collect sperm, transfer them into saline solution, and add DNA. This would lead to plasmid uptake by the sperm, which would then integrate the DNA into their genome and continue on to fertilize an egg.
Followup studies by other labs have found that the efficiency of the technique is vastly lower than initially reported. Where this lies on the spectrum between simple differences in technique and outright fraud, is unknown. The lab responsible for the technique claims it also applies to larger animals than the initial mice, such as pigs.
Obviously, the potential advantages are great - the preparation and delivery of sperm could even be done at home with basic tools (centrifuge, pipette, turkey baster). But the mechanics of getting DNA into the nucleus of a sperm, and having it integrate, raise many doubts. For use in humans, it is unacceptable.
What about testis-mediated gene transfer (TMGT)?
Our initial idea was indeed to deliver a plasmid into the testis, which would integrate into the cells that would go on to differentiate into sperm. While this remains a viable alternative to VelociHomo, it suffers from several shortcomings that led us to shelve it as a backup.
The downside most resonant with potential recipients is that TMGT involves an injection into the testicle, which bears the risk of damage and is unpleasant. All DNA transfection agents (liposomes, etc.) must strike a balance between efficiency and safety, as the most effective agents in delivering DNA into cells will also kill a significant fraction of them. Electroporation is one alternative delivery agent that may avoid cytotoxicity, but we need not go into detail about the combination of electricity and testes.
The second downside is that of delivery capacity, at least relative to VelociHomo: the larger a plasmid is, the more difficulty it has getting inside a cell and then a nucleus, roughly following an exponential decay of size vs. efficiency. A plasmid of >20 kilobases is unlikely to make it inside tissue in vivo at a useful level, limiting how much we can deliver. Also, it is necessary to include a transgenic “tag” that allows us to identify modified sperm, which both eats into the payload capacity, and means that base editing or knockout modifications cannot be included easily.
Other minor downsides: this technique only allows us to modify one set of chromosomes in the embryo, rather than both. For some applications this is suitable, but it is by no means ideal. We cannot check genome integrity prior to conception, though this is even more true for CRISPR microinjection. Development of sorting techniques for modified sperm, and selection of modified sperm for IVF, will take some effort - though these are not insurmountable problems. Finally, the risk of testicular damage is nonzero.
How about somatic cell nuclear transfer (SCNT)?
This method is familiar to most as the public’s definition of “cloning” - the nucleus of an egg is removed, and replaced by the nucleus of a donor cell, e.g. from an adult animal, and is then implanted into a womb and forms a fetus. While this has some utility in animals, the risk of side effects is far too high to make it useful in humans. These side effects are caused by both the accumulated mutations in the donor cell’s genome, when cells are taken from an adult animal, and from irregularities that occur in manipulation of the nucleus and its gene expression profile after transplantation.
The VelociHomo method could theoretically be adapted to solve both of these problems: somatic mutations are not an issue when the donor cells are derived from a newly-formed embryo, and the differences in gene expression between an early embryonic cell and a single-cell embryo are as minimal as can be practically achieved. This leads to much higher expected efficiencies, and fewer irregularities from the epigenetic reprogramming that occurs.
Nonetheless, the above approach will only be considered if ESC transplantation into a host embryo is unworkable, as the required nuclear manipulation bears its own risks. Of note, it has a minor ethical advantage in that a ‘host’ embryo is not required in the process, as would be the case with the traditional VelociHomo technique - any healthy egg would suffice.
What are the shortcomings of embryo selection by PGD?
This option, available today, aims to choose the “best” embryo among the handful generated by an IVF cycle. By removing a small number of cells from each embryo at the ~100 cell stage, clinics are able to obtain genome sequence data for each. This is then plugged into ~the algorithm~, which is trained on correlation between existing peoples’ genomes who have been screened for qualities like height, intelligence, and risk of e.g. heart disease. Parents are able to select the embryo containing the balance of traits they consider important.
Certainly this is a promising first step, although many doubts remain about its efficacy. Studies are beginning to come out that predict an expected gain in height of 2.5 centimeters, or an increase in IQ of 2.5 points. Of course, this is one of many technologies that will benefit from widespread genome sequencing (coupled with medical assessment) among a diverse population - much of this research relies on the UK Biobank’s data, which is obviously biased toward a subpopulation. But parents are still limited both by their own genomes, and the small number of eggs available.
What amazing enhancements will eventually be possible?
It is difficult to speculate on what is possible in any science - whether one believes in artificial general intelligence, affordable fusion power, faster-than-light travel, or alchemy. However, biology contains a vast number of possibilities that we can already see are attainable, even in the absence of future breakthroughs.
Without solving e.g. the protein folding problem, we can still imagine some upgrades that would greatly benefit humanity. For example:
Preventing cancer entirely, perhaps by a robust cellular detection system, or by eliminating oncogenes from somatic cells after development is complete. Something similar to the DRACO concept would allow a universal apoptosis-like system to kill cancers before they start, while oncogenes (which are just proliferation genes from early development that are no longer needed) could be excised from each somatic cell automatically, causing no harm to the organism but making cancers much less likely.
Preventing both cancer and aging by storing the genome in triplicate - mutation is the primary driver of cancer, and plays either a partial or total role in aging, depending on the theory. Today, if a nucleotide is mutated (other than in an obvious way that the cell can correct, e.g. pyrimidine dimers), the repair machinery does not know which base pair is correct and which is the mutant. By using a triple helix (or more likely a redundant set of chromosomes), along with an engineered DNA repair complex, each base could be checked against multiple copies to determine which is correct - similar to RAID in computing.
Enhancing metabolism has several facets, but perhaps most interesting would be the elimination of “junk” DNA. While vast regions of such non-coding DNA are important for correct function, much more is simply dispensable, a byproduct of evolution and ancient viruses that only serves to take up energy during replication. While no eukaryotic organisms have yet been synthesized with such a change, the benefits to energy requirements would be worthwhile.
Virus resistance could be achieved with codon elimination, though this is not a silver bullet - current efforts in the field eliminate only the least-used codons, which are not often used by viruses as they slow replication. Additionally, viruses are well-known for their ability to mutate, often at the bleeding edge of viability. A bad actor could inject a million viral particles into a supposedly immune host (or cell line) who has had their rare codons eliminated, and quickly generate a virus strain that found a workaround.
There are countless modifications from the animal world that would benefit aspects of humanity if correctly implemented - hibernation, endogenous synthesis of vitamins, radiation resistance, and simply more efficient versions of enzymes: a bacteria with a generation time of 20 minutes has had inconceivably more time to evolve its basic enzyme machinery than humans, with our generation time of two decades. We have been effectively frozen in time for hundreds of millions of years, while bacteria have kept on evolving. Also, cuttlefish-like skin that can display colors with a high refresh rate and resolution, controlled by thought, appeals to some.
How could you ensure that a genetically modified child will be able to have unmodified children themselves?
In the case that parents elect for this option, and absent any concerns about restricting the child’s reproductive freedom, we have three options:
First, the child could be made sterile through genetic intervention. A reserve of unmodified embryonic stem cells can be kept in cold storage, and in 20+ years when the child wants to become a parent, those cells can be unfrozen and differentiated into germ cells in vitro. While this technology is not available today, we can only expect that two decades of research will offer a solution. One key benefit of the VelociHomo method is that both modified and unmodified ESCs are available before birth.
Second, if the modification is effected via a plasmid integration, the DNA could be flanked by recombinase recognition sites that can cause it to be excised by an enzyme, e.g. the Flp-out system, solely in germ cells. However, point mutations etc. can not be reverted to wild-type with this method, and there is always risk of off-target effects from the DNA-modifying enzyme acting in an uncontrolled manner in vivo.
Third, as almost a form of genetic birth control, critical genes for the production of mature germ cells could be locked behind an inducible genetic circuit. For example, mature germ cells express a protein that inhibits conception, unless an agent like doxycycline is administered, which triggers a self-reinforcing genetic circuit that counteracts the previous infertility-causing gene.
Each of these options is likely practical with enough effort, though the risk to the child from extraneous genetic interventions, purely to prevent gene pool “contamination”, seems counterproductive.
Ethical & Legal Questions
How much will it cost? Won’t the rich create a generation of superchildren who will take over the world and plunge us into a plutocratic dystopia?
As with any new technology, the costs will probably first be borne by those most able to shoulder them. Ultimately, we expect the costs will not be much more expensive than the requisite IVF procedure. Until then, the advantage given to those first children by genetic modification will pale in comparison to the advantage of their inherited wealth. The first generations will not have a lifespan and IQ of 200 each.
Faster and wider adoption of the technology for human germline genetic modification is the best way to reach the most parents. Decades of incessant discussion (or repression) of the technology, in order to ensure equitable distribution, do little to help those who are at child-rearing age and cannot simply put plans of starting a family on hold. No previous technology has ever been kept from the public simply because it is too expensive, and early adopters have necessarily borne the cost.
Alongside spontaneous zombie apocalypse, the general populace’s greatest fear around designer babies is a ruling class of hyper-wealthy genetic superhumans. The best solution (to the superhuman overlords, not the zombies) is widespread R&D into perfecting and streamlining the technology, performed as quickly as is practical, to make it as cheap and widely available as possible. The eye-watering cost of somatic gene therapy today, which some opponents point to, only endures because the market is so limited.
Won’t children born with enhanced genetics experience stigma from their peers and pressure from their parents?
While it’s an open question whether the parents (and their children, once they are cognizant of their genome) would broadcast this information to the world, this is something society has dealt with before. A child of wealthy and/or famous parents may receive undue attention in school and in public, but no restrictions on their birth are made. Similarly, if a child has exemplary intellectual or athletic ability, this is recognized in their peers regardless of whether it is due to genetic engineering or the genetic lottery.
Whether designer babies will be pressured by their parents is a more interesting question, as there is a risk that, having invested money into the procedure, parents will have increased expectations of their child. We will hopefully see studies on this in the PGD children who are already being born, thanks to companies like Genomic Prediction. However, one must consider the immense expectations that parents already place on their naturally-born children. The stereotype of first-generation Asian-American overparenting is perhaps the most salient example - though it is near-universal, to some degree, in upper-middle class families. If this were a real concern, private education and tutoring would be heavily regulated already.
How is modifying an unborn child’s genome ethical, when they can’t consent?
The child didn’t consent to being conceived, either. Nor do children today consent to the risk of having parents who didn’t get tested for hereditary diseases, or will abuse them, or deprive them of the ability to flourish. But, as with the entirety of human history, most parents strive to provide the best life for their children, and this merely a new way of doing so.
The bioethicists most fervent in supporting the idea of designer babies were the most vehement in their condemnation of Jiankui He’s work. Of course, the benefit of theoretical philosophy is in rarely having to consider its practical application, and that things are not nearly as absolute in the real world.
The enhancements we are targeting are those that, absent the debate about the general ethicality of germline modification, would be gladly accepted by the vast majority of people, were they fortunate enough to receive them before birth.
Certain of the enhancements we aim to provide could be controlled with a genetic on/off switch. These do not have a 100% guarantee of effectiveness, and likely other mutations will be more reliant on these switches (of which there are a limited number available) to function optimally, but the option is there.
How can you ensure that the procedure is safe?
We can’t ensure that anything is absolutely safe. Saying that the procedure must be perfect is simply a veiled call for it never to be tried. We must weigh the risks with the benefits, much as we do when boarding a plane, driving a car, purchasing medical insurance, or indeed having a child. There are always degrees of risk, and those are up to informed parents to decide. Saying that we can’t be sure what a particular gene will do to a child means nothing. It happens whenever people have children, as their genes randomly mix and present new phenotypes.
One major benefit of our approach is that we can perform a full genome sequencing of the child before the cells are even placed into an embryo. Karyotyping, epigenetic analysis, RNA sequencing, etc. provide us with unprecedented analysis to make sure the child is as healthy as possible. This is already vastly safer than natural conception, where no genetic sequencing is done for the child until it is (possibly) tested by amniocentesis. In contrast to first-generation CRISPR microinjection techniques (a la Jiankui He), we can be confident in the genomic integrity of the child and every cell in their body.
Humans don’t understand biology, why do you think you can improve on nature?
Nature, and natural selection, cause more human suffering than anything - except other humans. Life has its upsides, but no one can deny that there is incomprehensible human suffering because of nature. Even when nature tries to help, we are left with remnants of evolution from thousands or millions of years ago, whether it is heart disease and obesity from adaptation to starvation, or countless autoimmune diseases from the chance survival of pandemics.
Our understanding of biology is not perfect, nor will it ever be. However, it is insulting to say that science has no grasp on the mechanisms involved. Over 100 years of research, with millions of experiments, have not been for naught. Biology holds a special place in our minds because we are composed of it, so it is harder to designate as a ‘technology’. But those with nightmares of untold horrors from biological engineering should look at what nature inflicts on us naturally.
Since the dawn of sentience and civilization, we have altered our natural condition immeasurably more than could be accomplished with biological tinkering. Whether it’s agriculture to protect us from famine, shelter from weather, antibiotics for infection, transport for travel, telecommunications or the rest of technology, one cannot argue that humans do not improve on nature as it relates to our well-being. Certainly, technology results in errors and shocks, but one cannot seriously argue that we would be better off without it.
Those privileged enough to think that we can delay the advance of technology have their opinions, and we have ours.
Why aren’t you an elite, well-funded team of top embryologists and stem cell biologists?
We have found, unsurprisingly, that the people best qualified to do this work are unwilling to expose themselves to the risk, be it legal, scientific, or social. We have spoken with people in the various relevant fields, and while many offer their tacit support of what we’re trying to do, none are willing to expose themselves to the potential fallout: loss of career, freedom, or life.
Nonetheless, embryology methods are well-established and have been in use for decades around the world. Stem cell biology is trickier, but at this point can be considered routine. Protocols are freely available, and there is no ultra-specialized equipment required.
While the common expectation of R&D is that it takes hundreds of millions of dollars, a great deal can be accomplished on a small budget. As long as it tends toward engineering rather than basic science, and as long as the people are competent. Pharmaceutical companies have picked their low-hanging fruit, and so require millions of labor-hours to see new results; but a previously unexplored territory can be bountiful.
Why not wait until a societal consensus is reached, a legal framework is in place, and the biology has been perfected?
Recent calls for a broad consensus on germline gene editing, on an international scale, will not have any effect. Getting every country in the world to agree (and it would have to be all of them) has never happened. Nor has worldwide prohibition of anything ever worked: for psychoactive drugs, the result of a worldwide ban has been the massive enrichment of criminal elements, be they mafia, cartels, or drug dealers. Also, the massive incarceration of otherwise-innocent users for victimless crimes, and the risk to their health from unregulated substances.
Other examples are rare, notably weapons proliferation (nuclear, biological, chemical) which has been either ineffective, or self-limiting due to the difficulty in acquiring nuclear weapons, or the marginal utility of biological/chemical agents.
Legal restrictions are almost always reactive, but hopefully they are based on rational decision-making and not an ill-placed fear of the unknown.
Biology has always been more art than science, at best a crude reverse-engineering of a machine that has coalesced over billions of years. But we still are still subject to the vagaries of ‘natural’ biology every day, hoping that we don’t get cancer, or catch the flu, or have a child with a severe disability. The only way to get beyond that, is to try.
What we do see is people who distrust science in general, almost entirely from a base fear of the unknown. Nuclear power and vaccinations are a good example of this, though we see it in avoidance of GMO foods. The business practices of agricultural companies and the reliance on monoculture get conflated with the potential unknown of genetic editing. Corporate motivations and industrial farming are not subject to a fear of the unknown, as we accept that capitalism exists and food is required.
Is all this legal?
Restrictions on stem cell and embryonic research are permissive, up to a point - that being implantation of a genetically modified embryo into a uterus. There is also the ‘14-day’ rule, which limits in vitro growth of an embryo to two weeks. If one works in a laboratory funded by the National Institutes of Health, there are additional restrictions.
For our purposes, the current regulatory environment provides ample room to refine the necessary techniques. Extracting embryonic stem cells from an embryo, culturing and genetically modifying them, and placing them into another embryo (even with genetic modifications being made) are all perfectly legal.
This allows us to perform extensive testing of genome integrity, screening for off-target effects, and the ability of modified cells to integrate into an embryo. With sufficient funding, we expect to generate exhaustive data over several years and thousands of experiments. The current legal ‘red line’ is at implantation of an edited embryo into a uterus.
Won’t this create a ‘genetic arms race’ between countries seeking to enhance their own citizens?
Maybe. As we’ve seen throughout modern history, advances in technology are often driven by such competition. Relevant laws are enforced at the national level, and since a unanimous moratorium is unlikely to happen, much less an enforced one, some countries will take the lead and drive others to catch up.
There is a spectrum of thought on human genetic modification around the world. A country with a well-funded, nationalized healthcare system and non-Christian worldview will likely be the first to legalize and regulate the use of editing technologies. Purely from a financial perspective, the potential for cost savings in healthcare makes it an investment worth considering.
A generation of genetically-modified supersoldiers is a less-likely outcome, despite its repeated appearance in science fiction. The vast majority of modern warfare does not revolve around front-line combat between people with rifles, but the capital invested in equipment, logistics, training, and military technology. While said riflemen will be a factor for a long while yet, their role shrinks every day.
But GATTACA?
This film is often people’s first touchstone in their distrust of germline genetic editing. While the future-noir setting is entertaining, if we look at it in context, it has some flaws. The story revolves around a protagonist who wants so badly to go into space that he is willing to risk the lives of his fellow crew members - were he to die of his diagnosed heart condition during the mission.
There are very good reasons why, even today, astronauts and pilots are screened for characteristics that have been deemed important. Billions of dollars of equipment, and often the lives of others, are resting on their shoulders. The story of a scrappy protagonist who cheats on entrance exams to become an airline pilot because they really want to be a pilot, and then kills hundreds of people when their heart explodes and the plane crashes…is not quite so heartwarming.
Why try to improve the species, when humans are perfectly adapted by evolution (or created by a deity)?
We humans are not well-adapted to our current environment. Technological innovation changes our living situation on a regular basis. Starting with agriculture, fire and cooking, modern food is plentiful and in secure supply. However, like every other animal, we are programmed to constantly prepare for winter or famine, leading to obesity, heart disease, and diabetes. Notably, we have lost the enzymes responsible for producing Vitamin C and uricase as an adaptation to the last ice age, causing all sorts of metabolic disorders. We must work for years to build muscle, even with an endless surplus of calories.
An interesting conclusion of the human genome project was that the greatest change in our genes was not in those expressed in the brain, but rather in the immune system. One need only look at the myriad of autoimmune and inflammation-linked diseases that plague us far worse today than communicable diseases or parasites. This adaptation suited closely-packed communities with poor waste disposal, living side by side with zoogenic domesticated animals - but now it causes more harm than good. The claims of “humans only lived to 30” are patently false; average life expectancy was brought down to that number by rampant childhood disease.
Much as some people like to think that their ancestors procreated because of their genetic superiority, in all likelihood it was because those ancestors were lucky enough to survive horrific plagues and famines. These mutations are holdovers from pre-modern humans, and they served us well before. The past ten thousand years of civilization is not nearly enough time to evolve. People now move all over the world in hours, but we see a gradient of skin color in native inhabitants that almost perfectly matches the average sun exposure. Such adaptations, including those less visible, would seem rather important for a person to thrive in a given locale.
Other enhancements that we propose would adapt humans to more modern society, e.g. the elimination of body odor. But we can also plan for the future: if humans want to be a multi-planetary civilization, it will be far more practical to adapt genetically rather than technologically. Instead of expending massive amounts of limited energy (both electrical and edible) to centrifuge humans at 1g and run in place on a treadmill while in space, we can enhance their physiology to require neither. If there is a chance we can keep humans in a form of stasis, and potentially reach other solar systems, it will be with germline genetic engineering.
Isn’t this eugenics?
No. The word ‘eugenics’ gets thrown around a lot these days, but what it actually refers to is government restriction of reproductive freedom. We argue that making it illegal for parents to genetically modify their children is, in fact, the closest parallel we have today to classical eugenics.
The reason people have an aversion to eugenics, and rightfully so, is because countries used genocide and forced sterilization to prevent reproduction by populations that they didn’t like. We have no intention of doing anything of the sort, nor does human genetic editing enable such use cases.
For those arguing that “positive eugenics” is just as abhorrent - that is, promoting the interbreeding and procreation of chosen individuals - consider that humans already do this on a daily basis, albeit on a more individual scale. Whether a mate is selected for sexual, monetary, or interpersonal characteristics, such selection is all around us. Of course, broader societal influence is also a factor in this.
Why are you promoting this work out in the open, rather than in a secret underground or offshore lab?
Regardless of how our project turns out, we have two driving goals: first, to advance the field as much as possible, as fast as possible. The more developed human genetic editing becomes, the more useful, more safe, and more affordable it becomes.
Second, we want to raise awareness of the reality of what gene editing means for our species. The public debate today is focused around whether or not to issue a blanket moratorium, and assumes that the field will not advance at all from its primitive stage. The backlash that will result from this lack of reasoned debate and critical thinking is dangerous and short-sighted.
What is our vision of the ideal outcome?
Speaking generally, we hope that one or more countries sensibly regulate human gene editing so that it can be safely developed. They will begin to offer it to their citizens, as well as medical tourists who are at least permitted by their own governments to undergo such a procedure. The first genetic modifications, following what is available in our understanding of biology, and public opinion, are targeted at improving quality of life and healthspan.
Concurrent with this, regulations are put in place to prevent discrimination against both genetically modified humans, and those whose parents were not willing or able to participate. Common-sense legislation can accomplish both, though surely with some difficulty. Discrimination against the unmodified, while unlikely to be as extreme as GATTACA adherents expect, should reflect our current social norms: if someone wants to be a professional basketball player but is short, or wants to be an astronaut but fails the physical examination, we do not lament their suffering.
A program to expand access to genetic modification should be conducted by all governments capable of doing so: outside of the United States, universal healthcare is uncontroversial, and a country will soon see the benefits of reducing medical expense among the population as common diseases are reduced via editing. In countries where basic healthcare is unavailable, giving a child 10 extra years of life, or enhanced athletic capability, is sadly not a high priority for their governments regardless.
As more clinics begin to offer the procedure, prices will drop as automation and best practices continue to be adopted. A fact of the high price of biotechnology today is that a given drug only targets a small fraction of the population, leading to the cost of development being split among a relative few. While VelociHomo may seem like science fiction, ultimately the labor-intensive work can be accomplished by a skilled technician in a few hours per patient.
Once human gene editing becomes accepted by society, we can expect to see additional benefits. Widespread genome sequencing will allow us to find beneficial alleles that can be applied to the population far more efficiently than traditional procreation, while maintaining genetic diversity. With the bulk of biology research not solely focused on picking apart minutiae of oncology, as it is today, many labs can be dedicated to refining genetic modification. Also, medical research is not a zero-sum game, as some suggest.
How will we prevent couples from freezing eggs and waiting for new biotechnology breakthroughs to give their child an advantage?
This is an interesting question posed by Savulescu and others - assuming a constant stream of biotechnological breakthroughs, parents that wait e.g. 5 years to have a modified child may be able to bestow upon it previously-unknown beneficial genetic tweaks. There are soft limits to how long parents can wait and still safely bear children, even with egg freezing; surrogacy pushes back that limit, but it may still exert a delaying force on child-rearing. Of course, this concern is already present, as couples today prefer to delay parenthood longer and longer in favor of their careers, with some companies even offering egg freezing to female employees.
Savulescu’s “solution” to this problem is a legal limit on the amount of time that eggs can be stored, which is of course abhorrent and contrary to reproductive rights. He also points out that children may be resentful of their parents for not waiting; but in time we can hope that they will understand. Perhaps rather than a negative punishment, governments can substitute a positive reward for having children earlier, to counterbalance any potential missed breakthroughs.
This concern does, however, imply a future in which the vast majority of children are being genetically modified. We must also consider the even greater theoretical resentment from children who find out that their parents did not give them any genetic modification at all. As we cannot force parents to modify their children, there will always be a population of children who, through no fault of their own, are unmodified. Care must be taken to protect their rights and livelihoods.
Saying that a technology should not be used, simply because it may be better in the future, is not a substantive argument.
What if people want to perform cosmetic or other marginal changes to their child?
We already see societies where adults undergo perform skin-lightening procedures, rhinoplasties etc, so this is unfortunately a real possibility. There are two major limiting factors: first, any gene editing procedure carries with it some non-zero level of risk, which parents must weigh against the benefit of cosmetic changes that are often irrelevant or simply fall out of fashion.
The second factor is far more important: any genetic edits performed on an embryo must be backed up by rigorous scientific study in order to be considered safe and reliable. While there has been some research performed on e.g. genetic determinants of eye color, it pales in comparison to the body of medical literature. For example, biologists would have no idea what genetic tweaks to make to ensure that a child has a smaller nose, or was more or less likely to become homosexual.
As a side note, for those concerned that parents would want to guarantee their child was heterosexual - first, this is unlikely to be realistic without several major discoveries that biologists are unlikely to focus their time on. Second, if parents are willing to undergo such a procedure just to guarantee hetero children, is that really a family that would provide a loving and nurturing environment for a gay child?
Won’t gene editing affect designer babies’ “human nature”?
This concern always seems to carry the vague implication that the concerned is the one who considers gene-edited humans “less human”. They will be no less human than any other human. Indeed, a child given the best genetic alterations known to biological science, with the best intentions of their parents, is if anything more human.