Counterpoints to Lovell-Badge & Daley’s CRISPR baby rationales

CRISPR babyTwo prominent scientists, Robin Lovell-Badge and George Daley, have been amongst the most outspoken proponents of leaving the door open to heritable human genetic modification via CRISPR. While they each have articulated their reasons in somewhat different ways at times, their core reasons arguing in favor of future heritable CRISPR appear largely the same.

In this post I tackle each of these arguments in favor of leaving the door open to “CRISPR babies” with science-based counterarguments. I also raise larger risks to going down this path including social justice issues and unintentionally paving the road to trait enhancement. There may also be one insurmountable technological hurdle here too in terms of safety.

Antonio Regalado annotation of Lovell-Badge's reasons heritable human CRISPR
Antonio Regalado annotation and his views of Lovell-Badge’s reasons for heritable human CRISPR. You need to scroll through Antonio’s various tweets to get to this one.

This past week Lovell-Badge outlined the main reasons he thinks producing CRISPR babies (and hence people bearing heritable genetic modifications) may be beneficial. His position was included in a very interesting piece where he systematically goes through the unfolding of the He Jiankui CRISPR baby situation.

Science journalist Antonio Regalado helpfully annotated the piece (including his own views) with red text in the margin, highlighting some of the key Lovell-Badge arguments to leave the door open to human CRISPR (see screenshot from Twitter.) Below I’ve concisely summarized these points and my counter arguments to Lovell-Badge’s. A HT to Antonio for his thoughtful take on this, with which I generally agree.

  1. Argument #1. If one or especially both parents are homozygous for serious disease-causing mutations then PGD won’t work. Counterargument. It is true that under those extremely rare and sometimes hypothetical circumstances that embryo screening methods such as PGD by itself would be impossible or close to impossible to use to yield a healthy new baby. However, in my view, CRISPR gene editing is likely to be impossible or close to impossible in that context too. Why? If both parents-to-be are homozygous for disease-causing mutations that means all embryos will be homozygous mutant too so you’d then have the extremely tough hurdle of CRISPR’ing both alleles in each embryo somehow back to WT. Also, according to Shoukhrat Mitalipov lab’s work, early human embryos often may not utilize exogenously introduced homology-directed repair templates meaning repair of homozygous mutations in human embryos may be impossible in some cases.  Also keep in mind that you must be able to do PGD on your CRISPR’d human embryos or what you are doing won’t be safe because you’d be “flying blind.” Finally, the potential parents-to-be who have disease-causing mutations may produce embryos that may already much less viable to start with, interfering with gene editing and PGD. If only one parent has a homozygous or heterozygous mutation then PGD is almost always going to be a more effective and definitely safer option (more below).
  2. Argument #2. We can give prospective parents who are mutation carriers more WT IVF embryos to try to use to make a healthy baby (or some say even to deal with infertility in a non-life threatening genetic disease context). Counterargument. I’d dispute the notion of needing to do CRISPR when only one parent carries a mutation because in that case PGD probably will work better. The notion that CRISPR’ing a mutation will safely and effectively give you significantly more WT embryos to work with for reproduction is unrealistic in my view at this point and for the foreseeable future given the state of the technology. Also, keeping it real, there may be an insurmountable hurdle here: is it really possible to rule out risky mosaicism in any specific CRISPR’d embryo without destroying the embryo? Not that I know of. Can someone resolve this roadblock? Finally, as to infertility, should we be promoting human genetic modification specifically to try to deal with just infertility itself outside the context of a life-threatening genetic disease? I’m skeptical that is a wise thing to start doing.
  3. Argument #3. We’ll need heritable CRISPR of babies to make ‘savior siblings’. Counterargument. Savior siblings are children produced by parents mainly or only because they have a small chance of being useful in saving an already existing family member with a fatal illness such as a brother or sister who desperately needs a stem cell transplant, but there’s currently no identified living donor who’s a match. It’s true that there are cases where producing the needed savior sibling will only be potentially achievable via genetic modification of embryos with CRISPR, but this is a very ethically complex path to go down. Even conceiving non-gene-edited children for the sole reason to possibly serve as savior siblings is itself ethically complicated, although I personally can see circumstances where for I would definitely be supportive of that. However, throwing in genetic modification into this future hypothetical scenario makes things far thornier in my view. There would definitely be risks to the potential savior sibling and making a healthy savior sibling with CRISPR may be technologically nearly impossible in many cases.

There are some larger issues that make heritable human CRISPR very problematic.

Cost. At just a practical level, cost is going to be a major problem. Being able to safely and effectively use CRISPR to make children with beneficial genetic changes, assuming it’s possible and there is societal consensus in coming decades) will cost a fortune for each attempt. The price of PGD (embryo screening) is roughly $10K-$20K USD, but often way more, and this is already a barrier to families wanting to employ it to try to prevent transmission of genetic diseases. However, making CRISPR babies in the needed meticulous, regulatory-compliant manner could easily be about 100-times more expensive at millions of US dollars each. There are many problems associated with this kind of astronomical cost including social justice issues.

Enabling attempts at trait modification. Finally another bigger picture thing to contemplate is whether it is wise to heritably genetically modify a few children in the future with only hypothetical benefit to them, while at the same time you are likely enabling potentially broad use of CRISPR gene editing for attempts at other things that society is clearly against such as trait enhancement. He Jiankui’s misguided efforts are a perfect illustration that we cannot count on people being wise or even rational about the heritable use of CRISPR in humans.

Looking ahead. While I disagree with Lovell-Badge and Daley about heritable human genetic modification, it’s good to have healthy discussions and debates on this important issue. Unfortunately, recently the “big wig” meetings run by National Academies mostly seem to not include gene editing scientists or policy/ethics researchers who are more skeptical.

It’s unclear if the meeting organizers are truly open to other opinions any more (they sometimes seem to unfairly dismiss those who disagree with them as being motivated by ignorance, ideology or fear…or having watched too many sci-fi movies) or to pursuing societal consensus on the path forward.

Finally, I have yet to hear a strong argument against a temporary moratorium on implantation of CRISPR’d human embryos. It’s an imperfect step, but such a moratorium would do more good than harm.

5 thoughts on “Counterpoints to Lovell-Badge & Daley’s CRISPR baby rationales”

  1. Paul,
    I was going to write about the analogy to the three parent embryos in the U.K. , of course, I learned about the issue on this very blog!

    Why does the Institution of Science seem obsessed with enabling some people to reproduce their DNA when their is plenty of reproductive capacity around, especially when it seems their DNA isn’t a good fit for that task?

    On argument #3.
    Our plastic brains endow us with the potential to not need ‘savior siblings’ and aspire to engineering societies that take good care of people when they are sick.

  2. I would like to see someone repeat Shoukhrat Mitalipov’s experiment.

    He claimed that when a CRISPR-Cas9 construct and a template for the wild type sequence is injected into a human egg along with a mutant sperm, the sperm’s mutation is corrected.

    The quirky bit is that he says that the mutation is repaired by copying the egg’s wild type DNA sequence, not the sequence of the template injected with the sperm.

    This is a mystery, since the pronuclei of the egg and the sperm are light years apart (in cellular space) when the correction supposedly occurred. I can’t get my embryologist brain around this one. I want the mystery to be solved.

    1. It seems the idea now for this mechanism is that the gene editing happens later (relatively long time after sperm/CRISPR-Cas9 injection), perhaps on the way to the two-cell stage, but that doesn’t seem compatible with no reported mosaicism. I’ll be curious if Dieter sees this phenomenon in his own CRISPR of human embryos.
      See here for some recent thoughts on Mitalipov’s 2nd paper on this:

  3. Paul, if someone has a question about the technical challenges of CRISPR-Cas, George Daley, George Church, Robin Lovell-Badge, would NOT be the people to ask. Nor would I be, because I have no experience myself in designing or carrying out gene editing with CRISPR. They don’t either.

    I would ask younger scientists, those who are in the lab experimenting with these techniques, what they find to be the difficulties…do they see off-target effects? Do they do whole genome sequencing to look for other CRISPR-associated abnormalities, like the large deletions that have been reported in some cases in which people actually looked? How about mosaicism? We need to see the actual data, but those data are not generated by us. Invite the postdocs.

    This is not a matter of being pessimistic or optimistic…it’s just being a scientist. We have opinions, but we also have facts.

    1. Hi Jeanne,
      It’s a good suggestion.
      From a technical point of view, safely and effectively gene editing a human embryo with clinical intent in a truly rigorous manner is going to be way harder than some seem to be assuming. When I hear a few folks saying off-targets maybe won’t be such an issue any more for CRISPR babies it seems unbelievable. The genome is a big place and it’s very hard to do WGS on single cells plucked from blastocysts and as I wrote, that can’t rule out mosaicism anyway.
      CRISPR-Cas9 also seems sometimes (we don’t know how often) to work differently in human embryos than in standard in vitro cultured human cells as well so I’d say the field knows relatively little specifically about gene editing human embryos.

Leave a Reply