Scientist make knockout human embryos with CRISPR?
Today we see a new Nature paper (Fogarty, et al.) on CRISPR “gene editing” of human embryos, this time from the UK from Kathy Niakan’s group.
Niakan got UK permission about 18 months ago to CRISPR healthy human embryos so they’ve been hard at work since. Because Fredrik Lanner of Sweden (see my interview here) also has governmental permission to CRISPR healthy human embryos, I’m guessing we’ll see a paper from his lab soon too. If I had to predict I would bet that their paper will also induced targeting of OCT4, but perhaps also with another key pluripotency gene being CRISPR’d as well such as NANOG.
What’s the scoop on this Fogarty, et al. paper from Niakan’s team?
How does it relate to the Mitalipov group Ma, et al. paper that has stirred so much debate?
This new paper is entitled, “Genome editing reveals a role for OCT4 in human embryogenesis”. It comes shortly after the big Mitalipov group Ma, et al. paper also in the same journal that has left many wondering in that case if CRISPR gene correction even happened.
In the new Fogarty paper, the team introduced CRISPR-Cas9 into human zygotes to target the key pluripotent gene OCT4/POU5F1. They injected a sgRNA–Cas9 ribonucleoprotein complex for the targeting and did so in S phase embryos approximately 5 hours prior to cytokinesis. That’s a fairly late beginning to gene targeting in terms of trying to prevent mosaicism and in fact they mention that a subgroup of embryos were at a later stage at thawing and so gene editing likely occurred developmentally later in those. In both cases it appears gene editing likely happened during mitosis, not earlier.
The Niakan group team found that they could efficiently target OCT4/POU5F1 in this way for disruption creating knockout human embryos that lack expression of that gene. In turn they report that loss of OCT4/POU5F1 leads to early developmental problems for the human embryo, somewhat earlier than in what happens in OCT4/POU5F1 knockout mice.
One of the first things that jumps to mind is that OCT4/POU5F1 (Oct4/Pou5f1 in mice) has been studied a ton in cultured stem cells, reprogramming and in mouse and other embryos. For this reason, as I was reading the Fogarty paper I kept being sure to ask myself, “What’s clearly new here besides the human context?” The earlier human phenotype is one reported difference and they also found evidence of OCT4 function beyond strictly pluripotent cells of the ICM.
They did the CRISPR-Cas9 targeting of OCT4/POU5F1 first in human ES cells and found interestingly that they could predict from the ES cell context many of the Indels that they found in the human embryos by sequencing. Notably, they also report an absence of detectable off-target sites, which is encouraging. Off targets may still be there, but not be detected due to limitations on sequencing of individual cells in human embryos. Still again this apparent accuracy is a positive sign that the technology can have good accuracy.
The Fogarty, et al. team found that the control method of injection gene editing machinery without guides (no genetic change) still led about 1/2 of embryos to fail to develop, but that this is roughly the same rate observed in published studies for unperturbed human embryos. In other words, about 1/2 of human embryos fail even without any poking or prodding in the lab. The CRISPR-Cas9 gene targeted embryos largely lacking OCT4 had a far lower rate of development though.
The Fogarty team reported that loss of OCT4/POU5F1 led to a number of important transcriptomic changes including many that one would expect given what is already known about pluripotency regulatory machinery. For instance, they found that knockout of OCT4/POU5F1 in the human embryos led to increased expression of some differentiation-associated genes. In addition, they report more heterogeneity in cellular gene expression in the knockouts, suggesting early embryo cells kind of lose their way and controlled cellular identity without OCT4.
How is Fogarty different than the Ma paper from Mitalipov?
First, in the Fogarty paper, the team used leftover embryos from preoccuring IVF procedures, whereas the Ma, et al. team made embryos themselves via IVF expressly for use in research. This is a very important distinction at a bioethics level. So the starting materials had a different origin.
Second, in Fogarty, the goal was to disrupt a gene via Indels, whereas in Ma the goal was to repair a mutant gene, which is a huge difference. In Ma, Indels were to be avoided. In Fogarty no repair templates were introduced because again generation of Indels was the only goal so it is difficult to compare results in some ways to the Ma paper.
Third, on a higher level, the intent of the research in these papers is also different in that Mitalipov clearly says he hopes this technology can be used in a reproductive, heritable fashion in humans, whereas Fogarty, et al. are primarily saying their work is about advancing knowledge, although they too mention potential clinically-relevant outcomes in the future.
Fourth, it is interesting that Fogarty did not report any evidence of inter-homologue repair, which is the hot button issue in the crosshairs related to the Ma paper. But Fogarty also didn’t seem to find very large Indels of the kind that would have made detection of Indels in the Ma paper definitely fail to work, as had been put out there as a possible alternative explanation for the Ma claim of inter-homologue repair. However, the nature of Indels will vary by gene so the jury’s out.
In addition, finally Fogarty reported plenty of mosaicism, whereas the Ma paper made a point of how little (only one mosaic embryo) that was detected. This difference may be due to technical distinctions between the studies.
Overall, this new paper is interesting and I expect more such papers are coming. We need to evaluate all such papers where there is use of CRISPR in human embryos both at scientific as well as ethical and policy levels.
In the abstract of the Fogarty paper, it’s striking that their main conclusion doesn’t even mention OCT4: “We conclude that CRISPR–Cas9-mediated genome editing is a powerful method for investigating gene function in the context of human development.” In a sense, this paper is more about CRISPR use in human embryos than about OCT4.
I have a feeling in a few years it’ll seem less of a big deal and such papers will not routinely be in journals like Nature. Maybe it’ll go through phases as it did with IPS cells. In the very beginning many IPS cell papers were in top journals, but then people got kinda used to it. The buzz wore off.
Still now it’s not everyday that a lab uses to CRISPR to genetically modifies human embryos. Yet.
Paul, thanks for this extremely valuable summary. We are moving towards another marathon societal debate relating to hESCR, with a host of new issues and concerns. Dr. Sherley’s comment is noted.
How do you think Niakans work influences the dispute between Mitalipov and Egli?
Dear Admin:
If “we” let scientists do experiments with selected children that resulted in the children’s disfigurement or death, some scientist would do so…for sure. They would justify it by the quest for knowledge and potential benefits for society on the whole, even if only selected individuals were actually likely to gain. Which of us would then enjoy discussing such experiments, vicariously or perhaps even as voyeurs? The activity reported here and its discussion differs only in the age of the humans exploited and destroyed.
Everyone, please, reach out to your various professional scientific organizations and ask them to strongly denounce such activities, under the justification of “Science,” as destructive not only to the harmed nascent humans, but also to humanity on the whole. If the ideal for our quest for knowledge is not being morally and ethically beyond reproach, in the end, it will engender more pain, suffering, and ill will than good. This world needs more goodness, love, the one thing that benefits everyone of us.
James L. Sherley
Reading the methods, they clearly state that they co-incubate the guide and Cas9 before injection. This is in contrast with the Mitalipov paper where this is not done, which might explain some of their results.
This is paper seems technically sound, but apart from the “modifying embryos” idea, I don’t particularly think it’s that interesting. To me, it’s just yet another CRISPR/Cas9 paper. Now if they did paired deletions or TALENs in addition or even compared shRNA-mediated knockdown to editing, this would be much more interesting.