Reflective blog assignment

PERSPECTIVE

Gene-edited babies: What went wrong and

what could go wrong

Haoyi WangID 1,2,3*, Hui YangID

4,5,6*

1 State Key Laboratory of Stem Cell and Reproductive Biology, Institute of Zoology, Chinese Academy of

Sciences, Beijing, China, 2 Institute for Stem Cell and Regeneration, Chinese Academy of Sciences, Beijing,

China, 3 University of Chinese Academy of Sciences, Beijing, China, 4 Institute of Neuroscience, State Key

Laboratory of Neuroscience, Key Laboratory of Primate Neurobiology, Shanghai Institutes for Biological

Sciences, Chinese Academy of Sciences, Shanghai, China, 5 CAS Center for Excellence in Brain Science

and Intelligence Technology, Shanghai, China, 6 Shanghai Research Center for Brain Science and Brain-

Inspired Intelligence, Shanghai, China

* [email protected] (HW); [email protected] (HY)

Abstract

During the second World Summit of Human Gene Editing, Jiankui He presented the gene-

editing project that led to the birth of two baby girls with man-made C-C chemokine receptor

type 5 (CCR5) mutations. This extremely irresponsible behavior violated the ethical consen-

sus of scientists all over the world. His presentation revealed a troubling lack not only of

basic medical ethics but also of the requisite understanding of genetics and gene editing.

Here, we review the rationale and design of his experiment along with the presented data,

and provide our scientific criticism of this misconduct.

On November 25, 2018, Jiankui He, an associate professor from Southern University of Sci-

ence and Technology, announced that two babies with edited C-C chemokine receptor type 5

(CCR5) genes had been born in China. This genetic modification, he claimed, would render these babies immune to HIV infection. On November 28, He presented the experimental data

of this project at the second World Summit of Human Gene Editing. While solid evidence of

this experiment remains to be disclosed and the veracity of such claims ascertained, the experi-

mental design and data presented at the summit revealed serious misconduct on both the sci-

entific and ethical levels. As researchers working in the gene-editing field in China, we were

completely shocked by this news. It would appear that He had been doing this work in secret.

As far as we know, He has not published noteworthy scientific papers in the gene-editing field

and was not actively involved in the gene editing community in China. We were enraged by

this extremely irresponsible misconduct, which clearly violated the regulatory and medical eth-

ics of China and nations all over the world. Here, we focus on the pitfalls of the scientific

aspects, assuming the data He presented were true, because we believe that responsible scru-

tiny and discussion of this event requires a good understanding of the scientific facts.

First, we would like to criticize his overall rationale. He claimed that he edited the CCR5 gene to prevent HIV infection in the babies, whose father is an HIV carrier (the mother does

PLOS Biology | https://doi.org/10.1371/journal.pbio.3000224 April 30, 2019 1 / 5

a1111111111

a1111111111

a1111111111

a1111111111

a1111111111

OPEN ACCESS

Citation: Wang H, Yang H (2019) Gene-edited

babies: What went wrong and what could go

wrong. PLoS Biol 17(4): e3000224. https://doi.org/

10.1371/journal.pbio.3000224

Published: April 30, 2019

Copyright: © 2019 Wang, Yang. This is an open access article distributed under the terms of the

Creative Commons Attribution License, which

permits unrestricted use, distribution, and

reproduction in any medium, provided the original

author and source are credited.

Funding: H.W. is supported by National Key

Research and Development Program of China

(2018YFA0107703), Strategic Priority Research

Program of the Chinese Academy of Sciences(No.

XDA16010503), and National Natural Science

Foundation of China (No. 31722036). H.Y. is

supported by R&D Program of China

(2018YFC2000100 and 2017YFC1001302), CAS

Strategic Priority Research Program

(XDB32060000), National Natural Science

Foundation of China (31871502), Shanghai

Municipal Science and Technology Major Project

(2018SHZDZX05), and Shanghai City Committee of

science and technology project (18411953700,

18JC1410100). The funders had no role in study

design, data collection and analysis, decision to

publish, or preparation of the manuscript.

Competing interests: The authors have declared

that no competing interests exist.

not carry the virus). Yet gene editing in embryos is completely unnecessary to prevent HIV

transmission to the fetus. It is possible for an HIV-positive father to generate healthy babies

using established Assisted Reproductive Technology (ART) with an extraordinarily high suc-

cess rate [1]. As for considering future immunity to HIV infection, simply avoiding potential

risk of HIV exposure suffices for most people. Therefore, editing early embryos does not pro-

vide benefits for the babies, while posing potentially serious risks on multiple fronts, which we

will discuss next.

The CCR5 gene encodes a receptor on white blood cells that HIV-1 uses, along with another receptor, to infect human cells. A naturally occurring CCR5Δ32 allele is present in certain European populations. Although both heterozygous and homozygous individuals have slower

progression or resistance to HIV infections [2, 3], even homozygous CCR5Δ32 individuals can still be infected by certain HIV strains [4]. Individuals carrying the CCR5Δ32 allele are in gen- eral healthy; however, this allele exists in very low frequency in non-European populations,

and no homozygous mutant has been identified in Chinese populations [5, 6]. Therefore, it is

very difficult to predict the risk of introducing the CCR5Δ32 allele or other CCR5 mutant alleles into a Chinese genetic background. While He claimed that there was a long-term health

follow-up plan, there are no details on who will fund this or assume responsibility in the event

that any medical issues arise.

Next, we’ll address his data. He first presented the data in Ccr5 knockout (KO) mice in order to evaluate, “would loss of CCR5 at the embryos stage by CRISPR/Cas9 gene editing cause undesirable genetic, physiological, or behavioral consequences?” (all contents in the quo-

tation marks are quoted directly from He’s presentation slides). This is absurd. It is not possi-

ble to answer that question simply by comparing histology staining of four different tissues

without any quantification and by doing two simple behavior tests in mice. The quality of the

science is very poor and superficial. For example, the data from the novel object investigation

behavior test suggested that there was a difference between the wild-type (WT) and Ccr5 KO mice, although the P value is above 0.05. Further evaluation using a larger sample size is neces- sary before claiming Ccr5 KO did not cause any behavioral phenotype. A cursory literature search would have revealed that CCR5 has normal immune functions as a receptor of chemo- kines, and CCR5 KO mice have natural killer cell (NK)-related phenotypes leading to higher risks for various viral infections [7–9].

Next, He designed multiple single-guide RNAs (sgRNAs) and tested their efficiency in

human cell lines and monkey embryos. These are very routine procedures used for gene-edit-

ing experiments. After Clustered Regularly Interspaced Short Palindromic Repeats and the

CRISPR-associated protein 9 (CRISPR-Cas9) components are delivered into cells, a DNA dou-

ble-strand break (DSB) will be generated at the target genomic locus. Either the nonhomolo-

gous end joining (NHEJ) repair process or the homology-directed repair (HDR) pathway is

employed to repair this DNA DSB. NHEJ repair often leads to small insertions or deletions

(indels), while HDR results in perfect repair or precise genetic modification at the targeted

site. He’s presentation showed only the characterization of indel mutation rates via NHEJ

repair; no experiment designed to introduce the CCR5Δ32 allele via HDR repair was shown, suggesting that He had no intention of generating the CCR5Δ32 allele. As far as we know, CCR5 mutant indel alleles other than the CCR5Δ32 allele do not exist in human populations at a high frequency. Previous studies suggested that expression and stability of the truncated

CCR5Δ32 protein in CCR5−/− individuals could also contribute to the HIV-resistance pheno- type [10]. Therefore, other CCR5 null alleles cannot simply be equated to the CCR5Δ32 allele when considering potential benefits and risks. Moreover, in-frame indel mutations could

potentially generate gain-of-function mutations, the risks of which are even more difficult to

predict.

PLOS Biology | https://doi.org/10.1371/journal.pbio.3000224 April 30, 2019 2 / 5

Abbreviations: ART, Assisted Reproductive

Technology; CCR5, C-C chemokine receptor type 5;

CRISPR-Cas9, clustered regularly interspaced

short palindromic repeats and the CRISPR-

associated protein 9; DSB, double-strand break;

HDR, homology-directed repair; hESC, human

embryonic stem cell; KO, knockout; MII,

metaphase II; NHEJ, nonhomologous end joining;

NK, natural killer cell; PGD, preimplantation genetic

diagnosis; sgRNA, single guide RNA; WGS, whole-

genome sequencing; WT, wild-type.

Provenance: Commissioned; externally peer

reviewed.

He attempted to optimize the microinjection procedure using monkey zygotes, and per-

formed sequencing to evaluate the gene-editing efficiency and level of mosaicism. Because his

data have not been published on any platform as a research paper, the information shown on

PowerPoint slides is insufficient for vetting. From what we can tell from his presentation,

despite various attempts, mosaicism remains a problem in the monkey embryo experiments.

He then translated his microinjection protocol to human embryos. As illustrated by He’s

data and previous studies, injecting embryos at the metaphase II (MII) stage and using Cas9

protein instead of Cas9 mRNA may reduce the mosaicism but does not eliminate it [11, 12].

Moreover, this strategy only works on NHEJ-mediated gene knockout, not on HDR-mediated

precise gene repair. While many strategies for increasing HDR have been reported in cell lines

[13], whether they are applicable in human embryos remains an open question. Work done by

the Mitalipov’s group suggests that the maternal allele could serve as a template for gene repair

to achieve correction of pathogenic mutation [12], but other groups have argued that Cas9

may induce large-scale deletions or rearrangements that lead to false positive results using

PCR-based genotyping [14]. These scientific debates reveal our incomplete understanding of

the DNA-repair mechanisms and outcomes associated with gene editing in human early

embryos and suggest that He probably underestimated the rate of mosaicism and the risk of

introducing harmful genetic alterations.

To assess mutations caused by off-target editing, He established one human embryonic

stem cell (hESC) line from the edited human embryos. Here, again, the quality of the science is

substandard. Only one hESC line was derived from one edited human embryo, which was

then used for whole-genome sequencing (WGS) to detect potential off-target mutations. Dur-

ing the process of hESCs’ derivation and expansion, many genetic alterations will occur [15].

Therefore, to identify the true off-target mutations caused by gene editing, multiple hESC lines

need to be established from edited and unedited embryos and characterized by deep sequenc-

ing and extensive bioinformatic analysis.

He further claimed that he performed so-called single cell–based WGS on preimplantation

genetic diagnosis (PGD) samples from 19 edited human blastocysts to assess on-target and off-

target editing events, before choosing the ones to transfer into recipients. Twelve out of nine-

teen embryos contained WT alleles, indicating the CCR5 gene was not completely edited in these embryos. Importantly, there is no mature and reliable technique for single cell–based

WGS to address the off-target mutations [16]. The whole genome amplification process, which

amplifies the single copy of the genome to a large enough quantity for WGS, introduces many

artificial mutations [16]. In addition, mosaicism is a major concern that cannot be addressed

by PGD, as we cannot sequence all cells in an embryo [17]. This means that even if the tested

cells are correctly edited, there is still a non-negligible risk that other cells in the embryo

remain unedited or carry unwanted mutations that may have unpredictable consequences.

Thus, He’s claim is unreliable.

In addition to potential off-target effects, it has been reported that DSBs generated by

CRISPR-Cas9 may also lead to on-target mutagenesis effects [18, 19]. Besides the types of

insertions, deletions, translocations, and rearrangements, on-target effects include large chro-

mosome deletions, chromosome truncations, and homozygosis of the genome by inter-homol-

ogy repair. Currently, no single method could detect all these types of off-target mutations,

especially when they occur at a very low frequency.

After the two baby girls were born, He’s team collected DNA from their cord blood, umbili-

cal cord, and placenta and performed WGS to confirm the success of CCR5 editing. The WGS results suggested that only two different CCR5 alleles existed in these samples, each one repre- sented by approximately half of all sequencing reads. For Lulu, one allele remained WT, and

the other allele had an in-frame deletion (−15 bp). For Nana, the two CCR5 mutant alleles

PLOS Biology | https://doi.org/10.1371/journal.pbio.3000224 April 30, 2019 3 / 5

represented 100% of all sequencing reads at the CCR5 target region, which suggests, quite sur- prisingly, that none of the mother’s tissue (containing WT CCR5 allele) contaminated any of Nana’s samples. Yet, because the details of sample collection and data analysis are lacking, we

cannot draw a robust conclusion. We strongly suggest that the authorities conduct a thorough

examination of all the original data and disclose the facts to the scientific community and gen-

eral public.

In conclusion, based on currently available information, we believe there is no sound scien-

tific reason to perform this type of gene editing on the human germline, and that the behavior

of He and his team represents a gross violation of both the Chinese regulations and the consen-

sus reached by the international science community. We strongly condemn their actions as

extremely irresponsible, both scientifically and ethically. We strongly urge the international

community of scientists and regulators to initiate a comprehensive discussion as soon as possi-

ble to develop the criteria and standards for genome editing in the human germline for repro-

ductive purposes. After reaching a clear consensus, clear and strict laws need to be passed,

implemented, and enforced at an international level. We also believe, however, that it is neces-

sary to further develop and improve the technologies for introducing precise genetic modifica-

tions into the human germline, including early embryos, sperm, and oocytes, using in vitro

experimental setups. These improved technologies may provide solutions for genetic diseases—

but only when consensus has been met and a regulatory framework has been put in place for

treating specific medical implications.

References 1. Zafer M, Horvath H, Mmeje O, van der Poel S, Semprini AE, Rutherford G, et al. Effectiveness of

semen washing to prevent human immunodeficiency virus (HIV) transmission and assist pregnancy in

HIV-discordant couples: a systematic review and meta-analysis. Fertil Steril. 2016; 105(3):645–55 e2.

Epub 2015/12/22. https://doi.org/10.1016/j.fertnstert.2015.11.028 PMID: 26688556; PubMed Central

PMCID: PMCPMC4779710.

2. Samson M, Libert F, Doranz BJ, Rucker J, Liesnard C, Farber CM, et al. Resistance to HIV-1 infection

in caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Nature. 1996;

382(6593):722–5. Epub 1996/08/22. https://doi.org/10.1038/382722a0 PMID: 8751444.

3. Marmor M, Sheppard HW, Donnell D, Bozeman S, Celum C, Buchbinder S, et al. Homozygous and het-

erozygous CCR5-Delta32 genotypes are associated with resistance to HIV infection. J Acquir Immune

Defic Syndr. 2001; 27(5):472–81. Epub 2001/08/21. PMID: 11511825.

4. Lopalco L. CCR5: From Natural Resistance to a New Anti-HIV Strategy. Viruses. 2010; 2(2):574–600.

Epub 2010/02/01. https://doi.org/10.3390/v2020574 PMID: 21994649; PubMed Central PMCID:

PMCPMC3185609.

5. Martinson JJ, Chapman NH, Rees DC, Liu YT, Clegg JB. Global distribution of the CCR5 gene 32-base-

pair deletion. Nat Genet. 1997; 16(1):100–3. Epub 1997/05/01. https://doi.org/10.1038/ng0597-100

PMID: 9140404.

6. Zhang C, Fu S, Xue Y, Wang Q, Huang X, Wang B, et al. Distribution of the CCR5 gene 32-basepair

deletion in 11 Chinese populations. Anthropol Anz. 2002; 60(3):267–71. Epub 2002/10/16. PMID:

12378793.

7. Glass WG, Lim JK, Cholera R, Pletnev AG, Gao JL, Murphy PM. Chemokine receptor CCR5 promotes

leukocyte trafficking to the brain and survival in West Nile virus infection. J Exp Med. 2005; 202

(8):1087–98. Epub 2005/10/19. https://doi.org/10.1084/jem.20042530 PMID: 16230476; PubMed Cen-

tral PMCID: PMCPMC2213214.

8. Khan IA, Thomas SY, Moretto MM, Lee FS, Islam SA, Combe C, et al. CCR5 is essential for NK cell traf-

ficking and host survival following Toxoplasma gondii infection. PLoS Pathog. 2006; 2(6):e49. Epub

2006/06/23. https://doi.org/10.1371/journal.ppat.0020049 PMID: 16789839; PubMed Central PMCID:

PMCPMC1475660.

9. Thapa M, Kuziel WA, Carr DJ. Susceptibility of CCR5-deficient mice to genital herpes simplex virus

type 2 is linked to NK cell mobilization. J Virol. 2007; 81(8):3704–13. Epub 2007/02/03. https://doi.org/

10.1128/JVI.02626-06 PMID: 17267483; PubMed Central PMCID: PMCPMC1866094.

PLOS Biology | https://doi.org/10.1371/journal.pbio.3000224 April 30, 2019 4 / 5

10. Alkhatib G. The biology of CCR5 and CXCR4. Curr Opin HIV AIDS. 2009; 4(2):96–103. Epub 2009/04/

03. https://doi.org/10.1097/COH.0b013e328324bbec PMID: 19339947; PubMed Central PMCID:

PMCPMC2718543.

11. Suzuki T, Asami M, Perry AC. Asymmetric parental genome engineering by Cas9 during mouse meiotic

exit. Sci Rep. 2014; 4:7621. Epub 2014/12/24. https://doi.org/10.1038/srep07621 PMID: 25532495;

PubMed Central PMCID: PMCPMC4274505.

12. Ma H, Marti-Gutierrez N, Park SW, Wu J, Lee Y, Suzuki K, et al. Correction of a pathogenic gene muta-

tion in human embryos. Nature. 2017; 548(7668):413–9. Epub 2017/08/08. https://doi.org/10.1038/

nature23305 PMID: 28783728.

13. Bollen Y, Post J, Koo BK, Snippert HJG. How to create state-of-the-art genetic model systems: strate-

gies for optimal CRISPR-mediated genome editing. Nucleic Acids Res. 2018; 46(13):6435–54. Epub

2018/06/30. https://doi.org/10.1093/nar/gky571 PMID: 29955892; PubMed Central PMCID:

PMCPMC6061873.

14. Egli D, Zuccaro MV, Kosicki M, Church GM, Bradley A, Jasin M. Inter-homologue repair in fertilized

human eggs? Nature. 2018; 560(7717):E5–E7. Epub 2018/08/10. https://doi.org/10.1038/s41586-018-

0379-5 PMID: 30089924.

15. Lund RJ, Narva E, Lahesmaa R. Genetic and epigenetic stability of human pluripotent stem cells. Nat

Rev Genet. 2012; 13(10):732–44. Epub 2012/09/12. https://doi.org/10.1038/nrg3271 PMID: 22965355.

16. Gawad C, Koh W, Quake SR. Single-cell genome sequencing: current state of the science. Nat Rev

Genet. 2016; 17(3):175–88. Epub 2016/01/26. https://doi.org/10.1038/nrg.2015.16 PMID: 26806412.

17. Steffann J, Jouannet P, Bonnefont JP, Chneiweiss H, Frydman N. Could Failure in Preimplantation

Genetic Diagnosis Justify Editing the Human Embryo Genome? Cell Stem Cell. 2018; 22(4):481–2.

Epub 2018/02/13. https://doi.org/10.1016/j.stem.2018.01.004 PMID: 29429945.

18. Shin HY, Wang C, Lee HK, Yoo KH, Zeng X, Kuhns T, et al. CRISPR/Cas9 targeting events cause com-

plex deletions and insertions at 17 sites in the mouse genome. Nat Commun. 2017; 8:15464. Epub

2017/06/01. https://doi.org/10.1038/ncomms15464 PMID: 28561021; PubMed Central PMCID:

PMCPMC5460021.

19. Kosicki M, Tomberg K, Bradley A. Repair of double-strand breaks induced by CRISPR-Cas9 leads to

large deletions and complex rearrangements. Nat Biotechnol. 2018; 36(8):765–71. Epub 2018/07/17.

https://doi.org/10.1038/nbt.4192 PMID: 30010673.

PLOS Biology | https://doi.org/10.1371/journal.pbio.3000224 April 30, 2019 5 / 5