I have argued previously that it is better to enforce national, rather than international, regulation of gene modification tools such as CRISPR-Cas9, since strong regulatory powers such as the FDA exist (swift action by the Chinese supports this claim). In fact, the FDA has so far chosen to regulate CRISPR technology as a drug, rather than a device, which gives it more control over specific applications and uses, while Congress excludes money in its spending bill from being used by the FDA to review applications for editing germ-line or heritable code. Nevertheless, there is no law in the U.S. that outlaws gene-edited babies.
I believe that most modern nations have adequate controls to regulate gene-modified babies, but it is important to explain why there will always be an ambiguous health benefit in creating GMO humans. Consider that disabling the CCR5 gene is not a new idea since U.S. biotech companies are already pursuing a strategy of using a form of gene modification to disrupt this gene to protect T cells from HIV infection. One important difference to this form of prophylactic gene therapy is that when applying it to living human patients a doctor can trade off the risk of the disease versus the tiny risks that gene modification tools could unintentionally alter other genes or damage functional DNA in a patient. In the case of a gene-modified baby, one is only adding the risk of inserting the CRISPR molecules, compared to a hypothetical health status of a person-to-be. The risk calculation is speculative at best.
A second important ethical difference is that previous strategies for altering CCR5 have involved somatic cells in the body that do not include the heritable or germ-line form of DNA that gets passed on to future generations, while the alleged gene modification in the Chinese case was the first made to heritable code, which may account for “legacy genetics” or passing on evolutionary advantages to descendants. But I argue that alterations to T cell receptor genes, or alterations to any other genes, will always create effects in future generations that are no more than ambiguous in the near-term.
We all have two parents. We all have two copies of most genes, so any modified version of a gene will soon get paired up with “wild type” versions of that gene in future descendants. In each new birth, genes get further shuffled in a process called “crossing over,” where chromosomes tip their arms and swap versions of genes like trading baseball cards. Thus, the effects of gene edits are quickly thrown up against new backdrops of other gene variations. For most complex traits, gene variations depend on coordination of tens or hundreds of other gene variants to achieve their affects. Any single gene variants will get broken up and shuffled into the mix of other variants through time.
Evolutionary dynamics are not trivial. In the 1970s, Lewontin and Hubby introduced the idea of balancing selection, which was extrapolated into principles that rare variants that contribute risk to various diseases may stick with us because of their compensatory benefit in contributing to heterogeneity or genetic variation within a population, or whereby risk variants contribute something positive in particular niches or contexts. Risk-causing genetic variants can also stay with us by “hitchhiking” along with beneficial ones that are positively selected for. One recent paper on schizophrenia suggests that risky mutations stay with us due to a process of background selection, whereby a lot of genetic variation is eliminated over time leaving risk variants in higher frequency. The important point is that genetic effects rarely are good or bad but depend on the shifting dynamics and backgrounds of other genetic variants.
Scientists have been working on inserting their volition into in the process, first learning to splice bits of DNA together, then using viruses to carry bits of DNA into cells, and then using gene modification techniques that work as DNA surgery. In the 1960s when Paul Berg’s lab used EcoRI (pronounced “echo R one”) to splice bits of DNA together. The resulting discussion at a California conference center called Asilomar led to a wide-ranging debate. David Baltimore noted a medical justification for the “potential for manufacture of biologicals” and a “potential for understanding the complicated disease that arise from the malfunction of cells.” Berg noted that legendary microbiologist Joshua Lederberg “gave me the impression of a child who’s being threatened to have his toys removed, and the whole theme of it, very frankly, was sort of ‘you’ve got to keep the feds out.’” It was the most famous case of scientists regulating themselves.
It is so far not clear that any regulation will keep scientists from creating GMO babies. At the time of Asilomar, attorney Peter Barton Hutt asked the nerve-wracking question “could these experiments be done by a high school teacher?” I have called this the “Hutt question,” because it is perhaps the most important question of all. In the 1970s, gene splicing could only be performed by a handful of elite experts in the field of microbiology, but CRISPR-Cas9 technology now makes gene modification something most professional geneticists are able to accomplish. Indeed, CRISPR, or CRISPR scientists may be impossible to stop; good argumentation about the nebulous benefits of GMO kids, and common sense on the folly of these hubristic endeavors, is as important an arbitrator as the regulation of laws.
Science is advancing rapidly as are the incentives for “first strike.” In 1980, a UCLA researcher named Martin Cline made the first unauthorized attempt at gene modification on a living human. In 2002, Italian fertility doctor Severino Antinori and physiologist Panos Zavos, and oddly enough, the Raëlian cult, independently claimed to have created the world’s first cloned baby, an idea that was believable since Dolly the Sheep had been cloned in 1996—though no darling little clone was ever found. The recent case of the Chinese researcher who alleged to create the first gene-engineered baby was only an extension of this ambition, and not the last scientist who will seek to make a splash with provocative rogue science.
Besides being relatively easy to use, there is more genetics information available to seek to exploit. Consider Danielle Posthuma’s work in Nature Genetics in 2017 tied 52 genes to human intelligence (though no single variant contributed more than a tiny fraction of a single percentage point to intelligence). Will college applicants begin stapling their 23 and Me results to their entrance applications? Will parents seek to engineer smarter kids in the lab? I want to convince you it is a fool’s errand. In fact, distributions of risk for mental disorders are also increasingly viewed to involve hundreds or thousands of gene variants. Thus, while the volition to improve our genomes is clearly evident by the ambition of scientists—and codified in the myths of Gattaca, Jurassic Park, Andromeda Strain and Mary Shelley’s Frankenstein—the reality is that genetic risks and advantages are not as straightforward as computer circuits.
In reality, the more problematic issue of gene modification relates to the neoliberal movement of biotechnology, whereas like computer and defense technology, government is expected to fund biotech R&D (through DARPA, NIH, Defense agencies) but urged to have a more limited role in regulation and pricing, whereas labor is driven to maximal lows and prices to maximal highs. For instance, the first genetically engineered T cell to fight cancer was priced at $475,000. The greatest ethical test of biotech regulation in our times is not how it will put stop the creation of GMO babies, a probable misadventure of ambiguous improvements, but how it relates to the equity and treatment of the living.