Doubts raised about CRISPR gene-editing study in human embryos | Hacker News Reader

Doubts raised about CRISPR gene-editing study in human embryos

nature.com

48 pointsokket9 comments

In summary, there are questions about one particular study out of Portland, with some plausible explanations for why the results/interpretation of this study were incorrect. This is not a concern about the utility / potential / feasibility of CRISPR itself but rather a critique of data interpretation and techniques by a specific study in human gametes.

The buried lead for me in this story is the use of pre-publishing services that are allowing this open review of biological papers without the delays traditionally associated with publishing. My understanding is that other fields like mathematics and CS have been doing this for a while, but to my knowledge, this is a fairly new phenomenon in medical biosci. (It's been a while since I've been through the publishing process).

>'Instead, the male gene copied the healthy sequence from the female gene. The authors don’t know why it happened. [...] “We were so surprised that we just couldn’t get this template that we made to be used,” said Shoukhrat Mitalipov, director of the Center for Embryonic Cell and Gene Therapy at Oregon Health and Science University and senior author of the study. “It was very new and unusual.”' https://www.nytimes.com/2017/08/02/science/gene-editing-huma...

>"We did not find any evidence of HDR using exogenous ssODN, suggesting that HDR is guided exclusively by the wild-type maternal allele." https://www.nature.com/nature/journal/vaop/ncurrent/full/nat...

I really don't see why this is so hard to understand. You added an enzyme that damages the DNA when it contains a certain sequence. The cells containing that sequence died more often, leaving behind a larger proportion of the ones without (it started out at ~50/50 in this case). This technique will always be doing this selection process to some extent (imo it could account for the majority of published "edits").

I haven't read the original study... However, I suspect that selective lethality of targeted cells would not account for this outcome. The process involves a single cell at the outset.

They started with 70 cells, 54 made it to 4-8 cell stage, and 36 of these were wild type (figure 2): http://www.nature.com/nature/journal/v548/n7668/fig_tab/natu...

In other words they started out with ~35/70 wt cells, and ended with 36/54. All 36 contained the WT sequence (rather than mutant) and not the template they tried to introduce. It is just as if the process killed ~16/35 mutants.

I've now scanned the original article.

As I understand, they injected ova with the CRISPR/CAS9 plasmids and then fertilized them with mutant sperm. The goal was to see if they could achieve CRISPR-mediated repair of the mutant allele provided by the sperm, guided by the genomic sequence of the wild-type ova allele.

My knowledge of CRISPR doesn't extent to knowing whether this is a smart idea or not... Though it seems to be an issue raised by Robin Lovell-Badge in the OP:

>"Robin Lovell-Badge, a developmental biologist at the Francis Crick Institute in London, says that it is possible that there is a “novel or unsuspected” biological mechanism at work in the very early human embryo that could explain how Mitalipov’s team corrected the embryos’ genomes in the manner claimed."

They appear to claim that they achieved somewhat mixed results, i.e. the repair of sperm-derived allele guided by the ova-derived allele was not 100% efficient.

>Among 54 CRISPR–Cas9-injected embryos, 36 (66.7%) were uniformly homozygous for the wild-type allele with each blastomere containing MYBPC3WT/WT, while 18 (33.3%) were uniform or mosaic heterozygous (Fig. 2a and Supplementary Table 3). In this group of 18, 5 embryos were uniformly heterozygous with each blastomere containing the intact wild-type and intact mutant allele (MYBPC3WT/∆GAGT) and 13 were mosaic,"

It doesn't appear (to me) that selective lethality would account for this data.

I would need to spend more time with the manuscript than I have right now to be more definitive, though...

> "It doesn't appear (to me) that selective lethality would account for this data."

Please go into more detail about what selective lethality is not explaining. Under that explanation

1) They started with ~35 wild type and ~35 mutant.

2) The treatment killed about half the mutants (but no/few wt) via inducing toxic double stranded breaks.

3) They end up with the original ~35 wt plus ~16 mutants, thus increasing the proportion of wt.

Regarding heterozygosity:

4) In some of the surviving mutants the cas9 damaged the DNA during the 2-cell phase or later, thus knocking out the mutant allele in some cells but not others.

As I said, I do not see what is difficult to understand. It seems to me guaranteed that a process like this accounts for at least some of the "editing" (ie, it is impossible to not be happening at all). Really, it is only a matter of what percent of the results are due to this.

It isn't clear to me where you're getting these numbers from... where do you find 35 WT and 35 mutant? Could you point me to the data for WT vs mutant embryos in this experiment?

Edit: in case you're referring to the table you linked to, the top line is unmanipulated embryos, and the second and third to experimental embryos, control and test. The survival rate of the second two lines is reported as similar.

The man who donated the sperm was heterozygous, which would mean about half the sperm would contain the mutant. You can also see this in their figure 2:

>"Sequencing of 83 individual blastomeres collected from 19 control embryos revealed that 9 (47.4%) were homozygous wild type (MYBPC3WT/WT) and 10 (52.6%) were heterozygous, carrying the wild-type maternal and mutant paternal alleles (MYBPC3WT/ΔGAGT) (Fig. 2a and Supplementary Table 2). This distribution was expected and confirms that the heterozygous patient sperm sample contained equal numbers of wild-type and mutant spermatozoa with similar motility and fertilization efficiency."

>"Edit: in case you're referring to the table you linked to, the top line is unmanipulated embryos, and the second and third to experimental embryos, control and test. The survival rate of the second two lines is reported as similar."

Also, not completely correct. The second line is not controls. Those are the 70 - 16 = 54 cells for which data is shown in figure 2A. The third line is for additional experiments which I didn't really look at.

Reply on news.ycombinator.com