Gao et al. published data in Nature Biotechnology (Nat Biotechnol. 2016 May 2) showing that DNA-guided genome editing using the Natronobacterium gregoryi Argonaute (NgAgo) protein targeted 47 mammalian genomic loci with a 100% success rate and an efficiency of 21.3%–41.3% at various targets. This report led us to test NgAgo’s utility in various cells and organisms such as mouse and zebrafish for gene editing. In most cases, a codon-optimized NgAgo for vertebrate animals was first synthesized and tested with appropriate guide oligos targeting specific genes using techniques similar to what has been utilized for the CRISPR/Cas9 system. After failing to confirm any NgAgo induced genomic DNA editing in any experiments, some of us switched to use an NgAgo expression vector (CMV-NLS-NgAgo-SK) used and provided by Han, the senior author of this paper, available from Addgene (#78253) since June or directly from his lab. Again, no success editing endogenous genomic DNA was achieved. As controls, the ability of this construct to induce indels was tested, targeting the same genes in cultured human 293T cells as those reported in Fig. 4 of Gao et al. Several researchers in different laboratories independently performed the experiments but no indels were observed at targeted loci, as assayed by T7E1 digestion, PAGE and/or sequencing. Representative data that directly repeat Fig. 4 of Gao et al from eight laboratories are shown in Fig. 1 and protocols used are detailed in supplementary information. We also include additional results from testing NgAgo in various systems by laboratories of signees of this letter in supplementary information. None of these studies proves that NgAgo has any genome editing activities.
Han issued public statements suggesting that the reported findings require “superb experimental skills” and one needs to be able to repeat the result of Fig. 3C, which is the inhibition of GFP expression in plasmid DNA transfected cells. Indeed, plasmid GFP expression reduction by co-transfection of NgAgo and its targeting DNA oligo is reproducible in our hands. However, we cannot demonstrate by sequencing this reduction is a result of DNA mutation. Many factors can affect this type of GFP expression, including NgAgo’s ability to target RNA as well as non-specific stress induced by oligo and DNA transfection. More recently, Han added that the activity of NgAgo is very sensitive to mycoplasma or bacteria in the culture. However, it seems unlikely that independent laboratories would all have their cells contaminated, resulting in consistently negative results for DNA editing activity. In fact, several of the signees of this letter have made sure that our cells are free of mycoplasma by first testing them before performing replication experiments.
The key point of paper by Gao et al is that DNA-guided NgAgo’s can efficiently target 47 genomic loci with a 100% success rate and a ≥20% efficiency. Neither the originally published protocol nor the newly released information on Addgene’s website involves any steps that seem to require “superb experimental skills”. To gain insights into NgAgo’s utility, some of us have even sent visiting researchers to Han’s laboratory but they were not allowed to perform genome editing experiments involving mammalian cells when they were there. Consequently, none of them returned with any information confirming Han’s data. Discussions on NgAgo have been frenzied in online forums, which cited some of the informal discussions in support of Han’s experimental data. Han also quoted David Cyranoski’s report (Nature, 2016 August 09) as evidence that NgAgo’s genome editing function had been confirmed. This further creates confusion because information in online forums is not accessible by the broader scientific community. We therefore urge the authors of the original paper to clarify the uncertainty surrounding NgAgo and provide all the necessary details for replicating the initial, very important results.
We thank the following researchers who participated in performing experiments listed in this correspondence: Gaurav Varshney, Haihua Xie, Xiaolin Qiu, Haigen Huang, Yihan Zhang, Guanrong Yan, Haipeng Bai, Xiukun Wang, Wei Qin, XiaoChan Lu, Yuexin Zhou, Ping Xu, Guanghai Xiang, Wei Mu, Qing Wu, Lu Gao.
All authors listed here declare that they have no conflict of interest. All institutional and national guidelines for the care and use of laboratory animals were followed.
An erratum to this article is available at http://dx.doi.org/10.1007/s13238-016-0349-3.
Electronic supplementary material
Below is the link to the electronic supplementary material.
About this article
Cite this article
Burgess, S., Cheng, L., Gu, F. et al. Questions about NgAgo. Protein Cell 7, 913–915 (2016). https://doi.org/10.1007/s13238-016-0343-9
- Supplementary Information
- Genomic Locus
- Genome Editing
- Gene Editing
- Online Forum