Concerns surface on Chinese paper on genetic modification of human embryos

The paper that came out Wednesday from a research group in China reporting the first genetic modification of human embryos has sparked a lot of discussion. Some concerns about this paper have surfaced.

GM human embryo review

1-day review? The paper (HT to John Borghi) was in review only from March 30-April 1 — so at most 24 hours. Really? That certainly raises a red flag of inadequate or absent peer review. That kind of “review” in the past with high-profile papers has been associated with a high risk of errors being found later in such papers. I’m not saying that will happen here, but it wouldn’t shock me.

The journal? The paper was also published in the journal, Protein & Cell, which Buzzfeed reports is partially owned by the Chinese government. Could there be some kind of COI there? Also, why was the paper reportedly rejected at more rigorous journals? There have been suggestions that reviewers raised ethical issues, but this remains unclear.

Rushed as evidenced by striking typo. The paper’s abstract has a pretty bad typo in the abstract (emphasis mine) suggesting issues with preparation and peer review again:

Taken together, our work highlights the pressing need to further improve the fidelity and specificity of the CRISPR/Cas9 platform, a prerequisite for any clinical applications of CRSIPR/Cas9-mediated editing.

Of course anyone can make a typo and I certainly do on a regular basis in emails or draft blog posts, but in a published paper abstract?

Unnecessary and premature? Another question to ask here is whether doing these studies specifically in human embryos was at all necessary or provided novel insights specifically because it was done in human embryos (as opposed to limiting the work to say just 293 cells as they did in part of the paper). So far, I don’t see much if anything that has been gained from using human embryos here other than maybe a hint of unique DNA repair. Jennifer Doudna has raised that kind of concern with the paper:

“I don’t see the value in working with human embryos right now. There’s a lot to be learned by working in other systems,” she says. In her view, the Huang paper provided little new scientific insight and seemed intended to “attract attention.”

The bottom line seems to be a final question of whether publishing this paper now and including human embryos was prudent given all the circumstances. I’m on the fence.

What do you think?

The big blind spot on CRISPR for human embryo editing: PGD

blind spotIt seems there is a big old blind spot in the discussion over germline gene editing in humans.

There’s been a lot of talk in 2015 about worries over how gene editing technology such as CRISPR might be used prematurely in the clinic in an unsafe or unethical manner in humans in the germline to try to prevent genetic disease.

This is a very serious concern and I share it.

However, in a way the dialogue on this usually misses a crucial, more basic question.

Why would anyone even try human gene editing in the germline given the existence of the very powerful, already proven safe technology of preimplantation genetic diagnosis (PGD)?

In a match up of say CRISPR versus PGD, PGD would win almost 100% of the time as by far the best choice to tackle genetic disease. It would be both safer and more effective. Again, we are talking about germline editing here, not gene therapy in children or adults where gene correction is a logical option.


Using PGD, anywhere from one to a few cells (depending on embryo stage) are plucked from early human embryos for genetic analysis. In this way, almost any genetic disease imaginable in principle and a huge number already in practice today can be detected. Right here in Davis and Sacramento, for example, the local fertility clinic called California IVF offers PGD for testing for a huge number of genetic diseases (see image).

Embryos lacking genetic disease can then be used for pregnancies. PGD can even be used to detect mutations in genes that do not always cause disease, but lead to disease predisposition such as BRCA1 mutations.

PGD also has the advantage of picking up other, random genetic problems.

This week has been abuzz with two papers related to the use of gene editing technology to prevent human genetic diseases. We saw the paper out of China on the use of CRISPR to genetically modify human embryos for a hypothetical future path to treat beta thalassemia. Then there was the Salk paper on the use of TALENs to prevent human mitochondrial disease.

The authors in both cases talked about potential future clinical applications of gene editing, but realistically why go that route?

If you think about it, as awesome as CRISPR-Cas9 is as a tool it seems like it would nearly always be far riskier and less effective than PGD for clinical applications. Same for use of TALENs for gene editing. PGD would win out as the choice by a mile.

Sure, PGD is not perfect.

It does not always work and some biopsied embryos fail due to the cell(s) being plucked off. There are also ethical concerns over the use of PGD for sex selection or potentially in ways that are eugenic.

Still it’s a generally safe and effective technology.

In fact it works for the diseases focused on in this week’s two gene editing papers: beta thalassemia and mitochondrial diseases.

If given all the facts, why would parents who are genetic disease carriers pick gene editing over PGD?

Perhaps if both parents-to-be were carriers of mutations? Then it would be only a rare instance that PGD could effectively find an embryo lacking a mutation so instead try gene editing? Seems like some pretty unlikely scenarios would have to be invoked.

And even then, it still would come back to PGD even if one chose a gene editing route because you would need to do PGD in order to validate that you achieved the mutation correction and hopefully there was no off-target activity in the embryos.

In the end, even in a hypothetical future scenario with an essentially perfectly accurate gene editing technology, going with PGD instead is going to be the wiser choice for parents and doctors almost every time.

Why is almost nobody mentioning PGD?

Neither the Baltimore, et al. nor Lanphier, et al. commentary pieces on germline editing mentioned PGD. Also, one of the top reviews on the potential therapeutic use of gene editing never mentioned PGD even once either. In my own ABCD plan for managing human germline editing, admittedly I also didn’t mention PGD. I should have.

The reality of PGD as a powerful, more often preferable option for prevention of human genetic disease needs to be an integral part of the therapeutic gene editing discussion today. Its inclusion in the dialogue would also further temper imprudent consideration of rapid clinical use of CRISPR or other gene editing technologies in the germline in humans.

Update: Hank Greely rightly pointed out in the comments that his recent blog post did mention the important element of PGD in the discussions. Hat tip to him.

Views on Izpisua Belmonte Cell Paper on Gene Editing for Mitochondrial Disease

mtDNA edit paper abstractJuan Carlos Izpisua Belmonte’s group published a Cell paper today on using gene editing to reverse mutations associated with human mitochondrial disease.

The paper is Reddy, et al. and is entitled, “Selective Elimination of Mitochondrial Mutations in the Germline by Genome Editing”.

The authors report success using TALEN-based gene editing or mitochondrial-direct restriction enzyme (mito-ApaLI) to reduce the burden of mutant mitochondrial DNA (mtDNA).

Their work was done primarily in mice, but also using chimeras made with murine oocytes fused with human cells bearing mitochondrial mutations.

They sum up their work in this way:

“Using mitochondria-targeted nucleases, mtDNA mutations are specifically eliminated in the germline to prevent their transgenerational transmission. This strategy represents a potential therapeutic avenue for treating human mitochondrial disease.”

The graphical abstract is above.

I am of two minds about this paper.

On the one hand, I think technically it is intriguing that they could take a gene editing approach to mitochondrial mutations, but on the other hand the notion that this approach could be safely and effectively used clinically in a human context in the germline as a way to prevent mitochondrial disease is somewhat concerning. This also resonates more strongly because of yesterday’s report out of China of gene editing using CRISPR of human embryos.

There is a growing trend of scientists heading the direction of human genetic modification.

Could it be safe?

What would be the ethical considerations?

Focusing first on the technical side of this new paper, the data look convincing to me after a quick read. One model was heteroplasmic mice (NZB/BALB) that contain two distinct mtDNA mutation types. In this model they were able via gene editing-based targeting one of the two mutation types (BALB) to introduce “heteroplasmic shift” demonstrating reduced abundance of the mutation. They have published this kind of work in the past on somatic cells and others have shown the same so I suppose that makes this work a bit less novel.Reddy, et al. Figure 6B

They did most of the work with mito-ApaLI targeting a unique cutting site in BALB. The gene edited one-cell embryo was also able to produce healthy mice (Figs. 3-4), but one caveat there is that it wasn’t clear to me from the paper and a quick Google search whether the NZB/BALB heteroplasmic mice normally have any phenotype due to the mitochondrial mutations or not. I suppose at least the gene editing did not seem to affect normal development.

The other model system here (Fig. 6) was the fusion of mouse oocytes with human cells (see Fig. 6B) containing mtDNA mutations. Using mito-TALEN technology they could again observe a heteroplasmic shift in this context as well. Depending on the human mtDNA mutation that was target, the efficiency of the gene editing to remove the mutant form ranged from a modest just over two-fold (NARP) to an impressive near complete elimination (LHOND).

What about potential off-target effects?

They reported that CGH (Figure S3C) and exomic sequencing (I believe this was not shown) did not indicate significant off-target effects. I would have liked to have seen more data on this however. It’s difficult to imagine no off-target effects were created and if true that’d be great, but it seems like they should have shown more data on this.

Shifting now to the bigger picture, what about possible clinical application of this technology in humans?

The authors are fairly gung-ho about the translational potential of their approach in humans, especially relative to so-called “3-person IVF” or “mitochondrial transfer” technology, recently approved for human use in the UK, but not permitted in the US:

“Mitochondrial replacement techniques involve a series of complex technical manipulations of nuclear genome between patient and donor oocytes that will result in the generation of embryos carrying genetic material from three different origins. For these reasons, mitochondrial replacement techniques have raised some biological, medical, and ethical concerns (Hayden, 2013; Reinhardt et al., 2013). Despite their great potential, more studies are still required to show that these techniques are safe in human oocytes. The approach presented here relies on a single injection of mRNA into patient oocytes, which is technically simpler and less traumatic to the oocyte compared to mitochondrial replacement techniques (Craven et al., 2010; Paull et al., 2013; Tachibana et al., 2013; Wang et al., 2014). Importantly, it does not require healthy donor oocytes, thus avoiding ethical issues related to the presence of donor mtDNA.”

The idea that a gene editing approach to mitochondrial disease in humans could be superior to the so-called “3-person IVF” or “mitochondrial transfer” approach to mitochondrial disease is kind of provocative. The Scientist provides a quote suggestive of a possible tension between gene editing and “mitochondrial replacement” approaches to mitochondrial disease prevention:

“While mitochondrial replacement advocate Doug Turnbull of Newcastle University, U.K., praised this latest advance in an e-mail as “elegant and exciting,” he cautioned that the technique “may be of limited value for those women whose oocytes have either large amounts of mutated mitochondrial DNA or all mutated mitochondrial DNA.”

Some dispute the notion that alteration of the mitochondrial genome is a “genetic modification”, but frankly I think that kind of argument is more political that scientific. I don’t buy it. We would be talking about genetically modified human beings whether with 3-person IVF or the gene editing approach described in this new paper.

For many families facing mitochondrial disease and wanting to have a genetically related baby, a better and far safer approach is preimplantation genetic diagnosis (PGD), although admittedly that is not an option for 100% of families.

It is also notable that multiple groups and I myself on this blog have called for restrictions on clinical application of gene editing technology in humans.

Clearly there is a lot more to learn. Great care should be exercised in discussion of potential human clinical applications of human germline modifications from this kind of research.

From Nature News, it seems that despite ethical concerns, researchers want to forge ahead into human embryos and ultimately the clinic with gene editing:

“Nevertheless, Ocampo and Izpisua Belmonte say that they are in the process of acquiring discarded human eggs and embryos from a fertility clinic in California, and waiting for approval from an ethics board. They plan to develop a line of stem cells from these modified cells, but say that they will not implant embryos into mothers or allow them to grow.”

Are we really ready?

Scientists in China create genetically modified human embryos: ‘A cautionary tale’

Update: apparently this paper (HT to @JohnBorghi) was only reviewed for 1 day (see image at bottom of post), raising major concerns about the depth of peer review.

Rumors have been flying for months that researchers in China and possibly elsewhere were shopping papers around at high-profile journals that reported gene editing and genetic modification of human embryos.

The rumors were right.

Today, one of the Chinese teams of researchers published their paper on genetically modified (GM) human embryos in the journal Protein & Cell.

The paper is open access so that’s good.

According to an excellent news piece in NatureNews by David Cyranoski and Sara Reardon, the paper had been submitted and rejected at top journals such as Nature and Science due at least in part to ethical issues. I had heard the same thing.

NatureNews quoted George Daley on this development:

“I believe this is the first report of CRISPR/Cas9 applied to human pre-implantation embryos and as such the study is a landmark, as well as a cautionary tale,” says George Daley, a stem-cell biologist at Harvard Medical School in Boston. “Their study should be a stern warning to any practitioner who thinks the technology is ready for testing to eradicate disease genes.”

The paper, entitled “CRISPR/Cas9-mediated gene editing in human tripronuclear zygotes” came from the lab of Junjiu Huang and is Liang, P. et al.

The authors apparently sought what they thought would be a relatively more ethically acceptable way to go by using the abnormal human 3PN embryos (that cannot develop normally because they have two sperm genomes) as a basis to create GM human embryos.

Liang, P. et al.

The team reported that the CRISPR gene editing in the human embryos didn’t go well (see Figure 2A above from the paper summarizing the basic numbers):

“We found that CRISPR/Cas9 could effectively cleave the endogenous β-globin gene (HBB). However, the efficiency of homologous recombination directed repair (HDR) of HBB was low and the edited embryos were mosaic.”

Specific mutations in the HBB gene can cause beta thalassemia.

To make matters worse technically, there were high levels of off-target activity:

“These data demonstrate that CRISPR/Cas9 has notable off-target effects in human 3PN embryos.”

According to the NatureNews piece the issue was raised by some that the problems with the CRISPR-Cas9 targeting reported in this paper could have been due to the embryos being abnormal.

While formally possible, I think that is unlikely to be the whole explanation. I’m with George Daley on this being a cautionary tale.

It is worth noting that the current study had institutional ethical approval according to a statement in the paper:

“This study conformed to ethical standards of Helsinki Declaration and national legislation and was approved by the Medical Ethical Committee of the First Affiliated Hospital, Sun Yat-sen University. The patients donated their tripronuclear (3PN) zygotes for research and signed informed consent forms.”

Would an institutional review board in another country such as the US have given the green light to making GM human embryos? I don’t know.

This study reaffirms the reasons that a pause is needed on in vivo human gene editing studies.

Even though in principle I could support some kinds of in vitro work on gene targeting in human germ cells and even early embryos (see my ABCD plan), I have to admit that this kind of work and the outcomes reported here, where we now can see this in the real world as a paper and not just hypothetically, make me very uncomfortable from an ethical perspective.

I feel like I need to have more time to read this paper carefully and to learn more about how the work was done before coming to more concrete conclusions as to whether it is acceptable from an ethical perspective. In addition, it would be helpful to learn more specifically about why other journals rejected the manuscript and what ethical concerns were raised.

Several other groups in China (and perhaps elsewhere) are conducting similar research and rumor has it that at least one is using normal or near-normal human embryos that have only specific disease-associated mutations.

GM human embryo review

Who deserves the patent on using CRISPR-Cas9 in human cells? Take our poll

There is no hotter technology than CRISPR-Cas9 gene editing tools.

Perhaps it is not surprising then that there is a patent dispute over it, which falls into two camps:

(1) Jennifer Doudna & Emmanuelle Charpentier, and (2) Feng Zhang.

Who deserves the intellectual property for use of CRISPR-Cas9 in human cells?

Take our poll.