Science fiction is awash with engineered humans, from the now-classic GATTACA to the demi-gods of Banks’ Culture; the concept is linked to that of cloning and carries similar strains of hubris and double-edged consequences. As with cloning, gene engineering is no longer science fiction. Protein and Cell just published the results of a Chinese research team that used a DNA editing technique called CRISPR/Cas9 to alter early trinuclear (triploid) IVF embryos. This technique has been used in many organisms, including mice, to successfully change specific genes. It’s a variation of gene therapy; the major difference is that in this study the repair was done at the low-number cell stage instead of postnatally.
[Parenthesis for the detail-oriented: CRISPR stands for clustered regularly interspaced short palindromic repeat, a common configuration in gene editing methods derived from bacterial defense systems. Cas stands for CRISPR-associated system – a CRISPR and its associated nuclease, which recognizes and clips the palindrome. The technique puts a target sequence with a desired nucleotide change in the CRISPR construct and introduces it plus a modified Cas enzyme into a cell or organism; the introduced system replaces the endogenous target sequence with the engineered one].
Triploid embryos, ova fertilized by two sperm, are mostly miscarried during the first trimester. The extremely few fully triploid infants that survive till birth have severe defects and without exception die a few days after delivery. The experimental triploid embryos additionally carried a thalassemia mutation in the HBB (beta-hemoglobin) gene. Thalassemic heterozygotes can lead a quasi-normal life with occasional blood transfusions, provided they are monitored. Homozygotes live a life of gruesome suffering and die before age 20 unless they undergo bone marrow transplantation.
The study documented several serious stumbling blocks, though none were unexpected: primarily low efficiency and low fidelity. Dependable introduction into cells is not trivial and the difficulty increases the more specialized the cells are, which is one reason why germline or embyronic editing is easier than its adult counterpart. Also, techniques of this type, which include RNAi, are prone to off-target effects (changes of quasi-homologous non-target sequences) and mosaicism due to expression variation – particularly with gene families, of which hemoglobins are one. As the study’s authors explicitly state, the technical issues must be competely resolved before such methods can go into clinical mode. Which leaves us with the other part: the eternal battleground between “can” and “should”.
Given the embryos’ triploidy and homozygous thalassemia, the primary ethical dilemma of tinkering with potentially viable entities did not arise in this study. Even so, Science and Nature rejected the paper summarily citing ethics concerns, and the usual people were interviewed saying the same things they said about IVF and cloning (briefly: unnatural hence unethical, slippery slopes, designer babies). Beyond the original furor over IVF babies, recall that a few months ago the UK allowed the generation of triparental embryos for people who carry mitochondrial mutations that would result in disease. And although many diseases are multigenic, others, equally devastating, would yield to such therapy.
Not surprisingly, many scientists and ethicists have called for a temporary moratorium on such experiments until consensus guidelines are developed. This happened at least once before, with recombinant DNA (the famous Asilomar conference of 1975). The original fears around gene splicing proved baseless, the grandstanding of Cambridge mayor Alfred Vellucci notwithstanding. The same is true of IVF, which has resulted in millions of perfectly normal humans, though the wars around gene therapy and GMOs are still raging, partly driven by issues other than feasibility or outcomes.
In my opinion, the meaninful dividing line is not between humans and all other animals. The real dividing line is between repair and enhancement (and what the latter really means). It’s almost certain that such methods will be tried on the less privileged first and, once perfected, will be preferentially accessible to the well-off – possibly indefinitely, if the current re-stratification of humanity by wealth persists. At the same time, it’s equally clear that the CRISPR technique has passed the proof of concept test and will eventually be used. I, for one, cannot imagine many future parents who will opt for no intervention if they are told that their child will develop Tay-Sachs, sickle-cell anemia or Huntington’s disease.
The burning question, of course, is if attributes deemed socially desirable will also be on the table with CRISPR. Thankfully, almost all suchlike attributes are polygenic and/or strongly susceptible to environmental input. Closer to the bone, a condition like monogenic deafness carries the dilemmas now associated with cochlear implants (I will not discuss “IQ” or autism, since these are not defined by single genes or, in some aspects, at the gene level and therefore don’t fall into this conversation). There is also the issue of consent, which means that adults are likelier to be eventually allowed to try exotic changes – with far greater risks attached, because of the intrinsic difficulties I discussed earlier.
At one end of this lurk the specters of eugenics and coercion – and, if financial and power stratifications escalate, the fear that humans may eventually split into Eloi and Morlocks. However, speciation requires total isolation of founder populations… and masters rarely withstand the temptation to mate with their slaves and servants, whether it’s an act of love or lust. Another fear is that the editing of an “undesirable” gene variant into extinction will have unforeseen consequences, since germline or embryonic editing is heritable. Many disease alleles have persisted because they confer advantages to heterozygotes: sickle cell to malaria, cystic fibrosis to cholera. As I never tire of repeating, “optimal” status is context-dependent. But if we fine-tune the editing techniques to the point that they become safe for routine use, re-introducing known alleles will be equally easy (creating new ones is definitely terra incognita, though these could, and should, be pre-tested in non-human systems).
On this, as with recombinant DNA, I’m a cautious optimist and venture to hope that the perfected CRISPR technique will be used with awareness and care for good – to ensure that monogenic diseases don’t lead to shortened or stunted lives. We may end up with a mosaic of guidelines, but eventually familiarity will dispel our wired fear of the new. We’ll still have to struggle with diseases that are less tractable, like dementia. And if CRISPR gives rise to a few more blue-eyed babies, I think we can live with that.