韩春雨事件梳理# Biology - 生物学
a*n
1 楼
2016年5月，韩春雨团队在nature biotechnology上发表文章，报道了一种新的基因编
辑技术，当时成为NgAgo。这篇报道轰动一时，被称为诺贝尔级别的发现，国内多家媒
体和自媒体争相报道，韩春雨被作为三流高校土博也能做一流科研的典型。
根据韩春雨的报道，通过将AGO蛋白与DNA片段结合，知道AGO剪切目标DNA，基因编辑的
效率超过当时风头正劲的CRISPR系统。而且，由于不熟PAM序列的限制，NgAgo的应用前
景明显优于CRISPR。
2016年6月2日，韩春雨在浙江大学医学院报告厅内介绍NgAgo时说道：“实验可控性非
常高，重复率在我的实验室达到了90％。”
然而，在韩春雨的文章发表后仅仅一个月，便出现了实验无法重复的传言，方舟子开始
在新语丝发表一系列文章提出多点质疑，包括对韩春雨声称的已经有多个实验室重复出
图4结果的声明表示质疑等等。
对此，韩春雨立即做出了回应如下：
我目前仍只有一个做实验的小团队，无法做好服务性工作，一天十几个个电话我都是
重复的以下内容：----所谓高超的技巧，其实也很简单，细胞做好检测，不要有寄生菌
污染，不要有支原体污染（Ago系统对污染特别敏感---特别是单转guide假阳性---我非
常确定在没有污染的情况下，单转guid不会影响GFP表达，假阴性也大多因为细胞污染
，假阳性和假阴性我都遇到过，国内买的细胞我就不吐槽了），转染用的质粒质量要好
（可用30ngGFP转细胞，若是均一且效率高表明质量好，强弱不均一则表明质量差，越
不均一表明质量越差），guide磷酸化完全（没有磷酸化不能load和切割---谢谢印证我
的Fig2结果，至于怎么检测，自己跑个PAGE看看---），特别的，对于Fig4，也是几乎
惟一算是有难度的，就是控制转染时的细胞密度和用血清控制细胞生长，时间稍长，毕
竟NgAgo未经过系统的密码子优化---照着online method上protocol for genome
editing and T7E1做（小经验，PCR最好不要有引物二聚体，如果引物二聚体多请重设
引物；PCR产物直接回收最好不跑胶回收，可能跟ago切24bp有关，设好对照！）---银
染分辨率高，可以排除T7E1假阳性，能做出预期条带后请用PCR产物去测序--------欢
迎大家广泛使用此系统，并分享成功经验和失败教训。附，addgen上的质粒，可以直接
用作基因组编辑。再次强调，不加NgAgo，就能抑制GFP，是假阳性，是细胞出问题了（
谢天谢地不但我自己而且审稿人也有这个对照）出现假阳性的细胞切基因组就没戏了。
对于Fig4，若是不习惯做银染，也可以直接做KI：doner 用GFPcoden+polyA signal，
前后加几个保护性的碱基---从GFP-N1扩出来连到T-vector上，再从此doner-Tvector上
扩（防止GFP-N1污染的假阳性）出来纯化，500-800ng共转（for each well of 24
well plate）。
随后的一个月里，多家实验室表示NgAgo有效。2017年7月，Gaetan Brugio在Twitter表
示，初步PCR结果显示NgAgo实验在小鼠身上有效。几乎同时，有消息称新德里的基因与
综合生物研究所的Debojyoti Chakraborty博士已经成功重复实验。20日下午，北京交
通大学计算机与信息技术学院教授刘峰在科学网发表博文称：“我刚和上海神经所@仇
子龙研究员确认，他重复了河北科技大学韩春雨老师的试验。” 而韩春雨则在各类报
告上继续声称，需要超高的实验技巧才能重复实验。同时，韩春雨也表示NgAgo系统不
够稳定，需要进一步完善，将会继续开发优化的NgAgo2.0版本。
然而，仅仅在同一个月的7月29日，Gaetan Brugio撰写长文描述自己的研究经历，表示
自己得出结论韩春雨的实验无法重复，而且韩在论文中所说自相矛盾，国际转基因技术
协会创建者Lluis Montoliu根据Gaetan Brugio和自己实验室的结果，向国际转基因技
术协会会员发出一封公开信，建议停止继续试图验证韩春雨的实验，不要再浪费时间、
金钱、人员和项目。信件原文如下：
From: [email protected] on behalf of Lluis
Montoliu To: ISTT List
Subject: [ISTT_list] Great disappointment with Ago: Long life to
CRISPR !
Date: Friday, July 29, 2016 9:49:22 AM
Dear colleagues,
the publication by Gao et al in May in Nature Biotechnology
http://www.ncbi.nlm.nih.gov/pubmed/27136078 triggered an enormous
expectation. This Chinese team led by Chunyu Huan reported that the
Argonaute (Ago) protein from a rare haloarchaea, Natronobacterium
gregoryi, (NgAgo) would efficiently work for gene editing purposes in
human cells. Ago had been described as an DNA-guided endonucleases two
years before, through a Dutch-Spanish microbiologist collaborating
team (Swarts et al. 2014, Nature:
http://www.ncbi.nlm.nih.gov/pubmed/24531762).
On paper, the new (fourth) Gene Editing system looked great. An
endonuclease, using ssDNA guides (5' phosphorylated though) and not
RNA guides, without a PAM, requiring 24 nucleotides (and not 20nt),
hence with higher specificity, and apparently with fewer off-target
issues, since modifications in just one position of the DNA guide
resulted in >90% decrease of the protein activity.
On paper.
I must confess we read the Huan paper in my lab with some
disappointment, after two years battling, unsuccessfully, with Ago
from Thermus, through a collaboration with my friend and colleague J.
Berenguer, from the neighbouring reserch centre CBMSO, and one of the
co-authors of the Nature 2014 paper. We had been scooped. We
repeteadly failed to find any gene editing activity using Ago from
Thermus thermophilus (TtAgo) in mammalian cells, through a variety of
conditions and we didn't understand why, though we always suspected
that these proteins would not be too comfortable at "too cold"
temperatures as physiological +37C. After reading the Gao paper we
concluded we simply missed the right bug and congratulated them for
being smarter and lucky and for finding this archaea. Perhaps the
trick was in using NgAgo instead of TtAgo.
Shortly after NgAgo was released from Addgene, beginning of June, many
labs, including mine, jumped onto it to try experiencing the
anticipated great expectations and joy associated with this new tool
of prokaryotic origin. But soon it was clear that something wasn't
quite right. Rumours began spreading during June and July at congresses,
through social networks, list emails and discussions groups that NgAgo
didn't appear to work as reported. Actually, didn't work at all. Some
colleagues that I absolutely trust at scientitic and technological
levels started to indicate that they could not reproduce Huan's paper
results.
At the recent TAGC meeting (where IMGS was contributing to, merging in
along with other Genetics Societies) Gaetan Burgio, from ANU, Camberra,
Australia, presented some very preliminary data with a gel with some
intermediate bands that would suggest NgAgo would be working and
editing at the expected places. But, shortly thereafter, Gaetan
engaged his lab in an OpenScience project, tried to characterize all
these bands and.... found nothing. So, again, another evidence
confirming NgAgo is not working as a gene-editing tool.
Gaeatan just released today his experience using NgAgo, openly sharing
his failures and providing details and some explanations for them.
My experience with Natronobacterium gregoryi Argonaute (NgAgo) Gaetan
Burgio Group leader at JCSMR, ANU
https://medium.com/@GaetanBurgio/my-experience-with-natronobacterium-g
regoryi-argonaute-ngago-3ed8909b410c#.bo9y6mf9u
At first, KUDOS to Gaetan. Many thanks to him for sharing their
efforts trying to confirm some gene-editing activity associated with
NgAgo. There is apparently none. In his view, NgAgo might be working
as a ligase at physiological conditions. Similar to our negative
results using TtAgo it would appear that NgAgo requires some higher
temperatures to work as initially reported. This of course seeds some
doubts on the Gao et al. publication and Gaetan, among other, is
requesting to Nature Biotechnology to request the Huan's lab to reveal
and share their raw data. We will see this part of the history how it
develops...
But, now, the most important message to convey is: NgAgo does not work
for gene editing in mammalian cells. Be aware and do not waste your
time, your money, your peoople and projects. If anyone has any
positive hint suggesting Ago is indeed working as a genomic editor,
please share the results, for the sake of Open Science, as Gaetan
beautifully and most generously did. Many thanks to Gaetan!
Unfortunately, this is a great disappointment. But, it also highlights
the uniqueness and the robustness of the CRISPR-Cas systems.
Long life to CRISPR!
Lluis phone: +49621- 1703 6210/6204
___________________________
Dr. Lluis Montoliu
Investigador Cientifico - Research Scientist
CSIC Centro Nacional de Biotecnologia (CNB-CSIC)
Campus de Cantoblanco
C/ Darwin, 3
28049 Madrid (Spain)
Tel. +34-91-5854844 / Fax +34-91-5854506
e-mail: [email protected]
WEB: http://www.cnb.csic.es/~montoliu/
U756 CIBERER: http://www.ciberer.es
Spanish EMMA node: http://www.infrafrontier.eu
ISTT: http://www.transtechsociety.org
At present, I would recommend everyone abandoning any project
involving the use of NgAgo. And avoid wasting time, money, animals and
people. Results are clear in many labs. The results we have in Madrid
are that TtAgo does not work in mammalian cells. The results that
Gaetan has in Australia are that NgAgo does not work in mammalian cells.
I know of many other colleagues who also tried and failed, but have
not reported these failures publicly. This is why I posted this message.
I think Ago might have some potential but we don't seem to have found
yet the adequate version of it (or the right bug).
As a side history, Francis Mojica struggled to find Grants to support
his pioneer experiments with CRISPR and got a couple of projects
turned down. That is why, instead of using Streptococcus pyogenes
(Doudna and Charpentier) or thermophilus (Siksnys), buga more
difficult and expensive to grow, he chose to work in Escherichia coli
CRISPR systems, and faced many difficulties, only to find out, many
years later that E.coli CRISPR-Cas system was of type I (not type II,
as Cas9) and, furthermore, was mostly inactive.
Please, let's focus in CRISPR-Cas9, Cpf1, C2c2,... et al... and leave
alone Ago while microbiologists don't find out the right one.
Lluis
辑技术，当时成为NgAgo。这篇报道轰动一时，被称为诺贝尔级别的发现，国内多家媒
体和自媒体争相报道，韩春雨被作为三流高校土博也能做一流科研的典型。
根据韩春雨的报道，通过将AGO蛋白与DNA片段结合，知道AGO剪切目标DNA，基因编辑的
效率超过当时风头正劲的CRISPR系统。而且，由于不熟PAM序列的限制，NgAgo的应用前
景明显优于CRISPR。
2016年6月2日，韩春雨在浙江大学医学院报告厅内介绍NgAgo时说道：“实验可控性非
常高，重复率在我的实验室达到了90％。”
然而，在韩春雨的文章发表后仅仅一个月，便出现了实验无法重复的传言，方舟子开始
在新语丝发表一系列文章提出多点质疑，包括对韩春雨声称的已经有多个实验室重复出
图4结果的声明表示质疑等等。
对此，韩春雨立即做出了回应如下：
我目前仍只有一个做实验的小团队，无法做好服务性工作，一天十几个个电话我都是
重复的以下内容：----所谓高超的技巧，其实也很简单，细胞做好检测，不要有寄生菌
污染，不要有支原体污染（Ago系统对污染特别敏感---特别是单转guide假阳性---我非
常确定在没有污染的情况下，单转guid不会影响GFP表达，假阴性也大多因为细胞污染
，假阳性和假阴性我都遇到过，国内买的细胞我就不吐槽了），转染用的质粒质量要好
（可用30ngGFP转细胞，若是均一且效率高表明质量好，强弱不均一则表明质量差，越
不均一表明质量越差），guide磷酸化完全（没有磷酸化不能load和切割---谢谢印证我
的Fig2结果，至于怎么检测，自己跑个PAGE看看---），特别的，对于Fig4，也是几乎
惟一算是有难度的，就是控制转染时的细胞密度和用血清控制细胞生长，时间稍长，毕
竟NgAgo未经过系统的密码子优化---照着online method上protocol for genome
editing and T7E1做（小经验，PCR最好不要有引物二聚体，如果引物二聚体多请重设
引物；PCR产物直接回收最好不跑胶回收，可能跟ago切24bp有关，设好对照！）---银
染分辨率高，可以排除T7E1假阳性，能做出预期条带后请用PCR产物去测序--------欢
迎大家广泛使用此系统，并分享成功经验和失败教训。附，addgen上的质粒，可以直接
用作基因组编辑。再次强调，不加NgAgo，就能抑制GFP，是假阳性，是细胞出问题了（
谢天谢地不但我自己而且审稿人也有这个对照）出现假阳性的细胞切基因组就没戏了。
对于Fig4，若是不习惯做银染，也可以直接做KI：doner 用GFPcoden+polyA signal，
前后加几个保护性的碱基---从GFP-N1扩出来连到T-vector上，再从此doner-Tvector上
扩（防止GFP-N1污染的假阳性）出来纯化，500-800ng共转（for each well of 24
well plate）。
随后的一个月里，多家实验室表示NgAgo有效。2017年7月，Gaetan Brugio在Twitter表
示，初步PCR结果显示NgAgo实验在小鼠身上有效。几乎同时，有消息称新德里的基因与
综合生物研究所的Debojyoti Chakraborty博士已经成功重复实验。20日下午，北京交
通大学计算机与信息技术学院教授刘峰在科学网发表博文称：“我刚和上海神经所@仇
子龙研究员确认，他重复了河北科技大学韩春雨老师的试验。” 而韩春雨则在各类报
告上继续声称，需要超高的实验技巧才能重复实验。同时，韩春雨也表示NgAgo系统不
够稳定，需要进一步完善，将会继续开发优化的NgAgo2.0版本。
然而，仅仅在同一个月的7月29日，Gaetan Brugio撰写长文描述自己的研究经历，表示
自己得出结论韩春雨的实验无法重复，而且韩在论文中所说自相矛盾，国际转基因技术
协会创建者Lluis Montoliu根据Gaetan Brugio和自己实验室的结果，向国际转基因技
术协会会员发出一封公开信，建议停止继续试图验证韩春雨的实验，不要再浪费时间、
金钱、人员和项目。信件原文如下：
From: [email protected] on behalf of Lluis
Montoliu To: ISTT List
Subject: [ISTT_list] Great disappointment with Ago: Long life to
CRISPR !
Date: Friday, July 29, 2016 9:49:22 AM
Dear colleagues,
the publication by Gao et al in May in Nature Biotechnology
http://www.ncbi.nlm.nih.gov/pubmed/27136078 triggered an enormous
expectation. This Chinese team led by Chunyu Huan reported that the
Argonaute (Ago) protein from a rare haloarchaea, Natronobacterium
gregoryi, (NgAgo) would efficiently work for gene editing purposes in
human cells. Ago had been described as an DNA-guided endonucleases two
years before, through a Dutch-Spanish microbiologist collaborating
team (Swarts et al. 2014, Nature:
http://www.ncbi.nlm.nih.gov/pubmed/24531762).
On paper, the new (fourth) Gene Editing system looked great. An
endonuclease, using ssDNA guides (5' phosphorylated though) and not
RNA guides, without a PAM, requiring 24 nucleotides (and not 20nt),
hence with higher specificity, and apparently with fewer off-target
issues, since modifications in just one position of the DNA guide
resulted in >90% decrease of the protein activity.
On paper.
I must confess we read the Huan paper in my lab with some
disappointment, after two years battling, unsuccessfully, with Ago
from Thermus, through a collaboration with my friend and colleague J.
Berenguer, from the neighbouring reserch centre CBMSO, and one of the
co-authors of the Nature 2014 paper. We had been scooped. We
repeteadly failed to find any gene editing activity using Ago from
Thermus thermophilus (TtAgo) in mammalian cells, through a variety of
conditions and we didn't understand why, though we always suspected
that these proteins would not be too comfortable at "too cold"
temperatures as physiological +37C. After reading the Gao paper we
concluded we simply missed the right bug and congratulated them for
being smarter and lucky and for finding this archaea. Perhaps the
trick was in using NgAgo instead of TtAgo.
Shortly after NgAgo was released from Addgene, beginning of June, many
labs, including mine, jumped onto it to try experiencing the
anticipated great expectations and joy associated with this new tool
of prokaryotic origin. But soon it was clear that something wasn't
quite right. Rumours began spreading during June and July at congresses,
through social networks, list emails and discussions groups that NgAgo
didn't appear to work as reported. Actually, didn't work at all. Some
colleagues that I absolutely trust at scientitic and technological
levels started to indicate that they could not reproduce Huan's paper
results.
At the recent TAGC meeting (where IMGS was contributing to, merging in
along with other Genetics Societies) Gaetan Burgio, from ANU, Camberra,
Australia, presented some very preliminary data with a gel with some
intermediate bands that would suggest NgAgo would be working and
editing at the expected places. But, shortly thereafter, Gaetan
engaged his lab in an OpenScience project, tried to characterize all
these bands and.... found nothing. So, again, another evidence
confirming NgAgo is not working as a gene-editing tool.
Gaeatan just released today his experience using NgAgo, openly sharing
his failures and providing details and some explanations for them.
My experience with Natronobacterium gregoryi Argonaute (NgAgo) Gaetan
Burgio Group leader at JCSMR, ANU
https://medium.com/@GaetanBurgio/my-experience-with-natronobacterium-g
regoryi-argonaute-ngago-3ed8909b410c#.bo9y6mf9u
At first, KUDOS to Gaetan. Many thanks to him for sharing their
efforts trying to confirm some gene-editing activity associated with
NgAgo. There is apparently none. In his view, NgAgo might be working
as a ligase at physiological conditions. Similar to our negative
results using TtAgo it would appear that NgAgo requires some higher
temperatures to work as initially reported. This of course seeds some
doubts on the Gao et al. publication and Gaetan, among other, is
requesting to Nature Biotechnology to request the Huan's lab to reveal
and share their raw data. We will see this part of the history how it
develops...
But, now, the most important message to convey is: NgAgo does not work
for gene editing in mammalian cells. Be aware and do not waste your
time, your money, your peoople and projects. If anyone has any
positive hint suggesting Ago is indeed working as a genomic editor,
please share the results, for the sake of Open Science, as Gaetan
beautifully and most generously did. Many thanks to Gaetan!
Unfortunately, this is a great disappointment. But, it also highlights
the uniqueness and the robustness of the CRISPR-Cas systems.
Long life to CRISPR!
Lluis phone: +49621- 1703 6210/6204
___________________________
Dr. Lluis Montoliu
Investigador Cientifico - Research Scientist
CSIC Centro Nacional de Biotecnologia (CNB-CSIC)
Campus de Cantoblanco
C/ Darwin, 3
28049 Madrid (Spain)
Tel. +34-91-5854844 / Fax +34-91-5854506
e-mail: [email protected]
WEB: http://www.cnb.csic.es/~montoliu/
U756 CIBERER: http://www.ciberer.es
Spanish EMMA node: http://www.infrafrontier.eu
ISTT: http://www.transtechsociety.org
At present, I would recommend everyone abandoning any project
involving the use of NgAgo. And avoid wasting time, money, animals and
people. Results are clear in many labs. The results we have in Madrid
are that TtAgo does not work in mammalian cells. The results that
Gaetan has in Australia are that NgAgo does not work in mammalian cells.
I know of many other colleagues who also tried and failed, but have
not reported these failures publicly. This is why I posted this message.
I think Ago might have some potential but we don't seem to have found
yet the adequate version of it (or the right bug).
As a side history, Francis Mojica struggled to find Grants to support
his pioneer experiments with CRISPR and got a couple of projects
turned down. That is why, instead of using Streptococcus pyogenes
(Doudna and Charpentier) or thermophilus (Siksnys), buga more
difficult and expensive to grow, he chose to work in Escherichia coli
CRISPR systems, and faced many difficulties, only to find out, many
years later that E.coli CRISPR-Cas system was of type I (not type II,
as Cas9) and, furthermore, was mostly inactive.
Please, let's focus in CRISPR-Cas9, Cpf1, C2c2,... et al... and leave
alone Ago while microbiologists don't find out the right one.
Lluis
