5′ capped and 3′ polyA-tailed sgRNAs enhance the efficiency of CRISPR-Cas9 system
The clustered regularly interspaced short palindromic repeats (CRISPR)-associated (Cas) system is an adaptive immune system in a variety of bacteria and archaea (Terns and Terns, 2011). The most commonly used Streptococcus pyogenes type II CRISPR-Cas9 system consists of Cas9 nuclease and two short RNAs, crRNA and tracrRNA, which can be linked together forming one chimeric single guide RNA (sgRNA) (Jinek et al., 2012). Guided by sgRNA, Cas9-sgRNA complex can generate DNA double strand breaks (DSB) at specific genomic loci (Jinek et al., 2012; Cong et al., 2013; Mali et al., 2013). Cas9 with mutations at the catalytic RuvC and HNH domains loses the endonuclease activity (indicated “dCas9”), while maintains its DNA binding capability (Gilbert et al., 2013). DCas9 fused with transcription effectors, such as VP64, P65-HSF1 and KRAB can be guided to specific promoters by sgRNA, enabling regulation of gene transcription (Gilbert et al., 2013, 2014).
RNA structures and sequences that can be recognized by specific RNA binding proteins were incorporated into sgRNA backbone to enable more efficient and versatile effector recruiting such as synergistic activation mediator (SAM) system (Konermann et al., 2014), Casilio system (Cheng et al., 2016) and CRISPR-Display (Shechner et al., 2015). In addition, chemical modifications such as 2′-O-methyl 3′-phosphorothioate (MS) or 2′-O-methyl 3′ H-thioPACE (MSP), which have been shown to enhance siRNA stability (Deleavey and Damha, 2012; Eckstein, 2014), were applied to sgRNA as well as crRNA and tracrRNA (Hendel et al., 2015; Rahdar et al., 2015), leading to improved gene editing efficiency. Despite all these advances, chemical synthesis of long RNA oligos with these modifications is challenging and expensive. With current length limitation of RNA chemical synthesis technologies, it is difficult to generate sgRNA with extra sequences or structures. Here, we established a modification method to improve the stability of in vitro transcribed (IVT) sgRNA and therefore the efficiency of CRISPR-Cas9 system, which can be adapted by any lab with basic molecular biology experience.
Next, we evaluated whether CT modification on sgRNA led to a higher genome editing efficiency. We delivered each AAVS1 sgRNA structure together with in vitro transcribed Cas9 mRNA into K562 cells by electroporation and analyzed genome disruption frequencies by Surveyor assay (Fig. S2A). Consistent with the improved stability, CTsgRNA led to higher insertion or/and deletion (indel) frequencies (26.90%) compared to control sgRNA (15.29%), while the sgRNAs with other structures had an editing efficiency lower than control (Fig. S2A). Improved gene editing efficiency using CTsgRNAs was also obtained at the VEGFA, EMX1, HBB and PD1 loci in K562 cells, with indel frequency quantified using tracking of indels by decomposition (TIDE) analysis (Figs. 1C and S3), Surveyor assay (Fig. S2B) and chip-based QuantStudio 3D digital PCR (Fig. S4).
As CT modification on sgRNA led to higher on-target editing efficiency, we evaluated whether it also affected the off-target activity. We tested three sgRNAs with well-defined off-target sites, and found that the CTsgRNAs induced similar level of off-target indel frequencies when compared to the unmodified sgRNA (Fig. 1C). However, we cannot exclude the possibility that CTsgRNA caused higher mutation rates at other off-target sites. When sgRNAs with different structures and Cas9 mRNA were delivered into human primary T cells, genome editing induced by unmodified AAVS1 sgRNA was almost undetectable, while CTsgRNAs resulted in editing frequencies of 15.23% (Fig. 1D). Neither 5′ cap nor the 3′ polyA tail modification alone was able to induce indels more efficiently than unmodified sgRNAs in primary T cells (Fig. 1D). Similar results were obtained at the HBB loci (Fig. 1E). Taken together, these results showed that the CT modification enhanced sgRNA intracellular stability, improved genome editing efficiency in K562 and human primary T cells.
We next tested whether CTsgRNA can be applied to efficiently activate endogenous genes in human primary T cells. We chose OCT4 and FOXP3 as target genes. Delivery of dCas9-P65HSF1 mRNA with the unmodified sgRNA pool barely activated target genes, while CT modified sgRNA pool induced 22-fold and 7-fold improvement at OCT4 and FOXP3 mRNA level, respectively (Fig. 2D). Based on flow cytometry analysis using FOXP3 antibody, the mean fluorescence intensity increased more than 2-fold (Fig. S5B). Since the gene activation system was expressed transiently, the FOXP3 positive cells decreased with time (Fig. S5C).
To further improve the performance of CT modified sgRNA and establish an efficient endogenous gene activation platform in primary T cells, we applied the CT modification to two sgRNA structures (sgRNA1.1 and sgRNA2.0) descried previously (Konermann et al., 2014), (1) sgRNA1.1: one copy of MS2 hairpin aptamer was incorporated into the tetraloop of the sgRNA backbone, indicated as Tetra sgRNA; (2) sgRNA2.0: two copies of MS2 hairpin aptamers were incorporated into the tetraloop and stem loop 2 of sgRNA backbone, indicated as 2×MS2 sgRNA (Fig. 2D). We delivered dCas9-P65HSF1 mRNA, MS2-P65HSF1 mRNA and unmodified or CT modified sgRNAs with different structures: OCT4 Tetra or 2×MS2 sgRNA pool simultaneously into primary CD3+ T cells. Schematic structures of these two complexes were shown in Fig. 2E. Unmodified OCT4 Tetra and 2×MS2 sgRNA pools did not activate endogenous OCT4 efficiently. Remarkably, both activated OCT4 expression up to 60-fold after CT modification (Fig. 2F). To further improve the endogenous gene activation in human primary T cells, we optimized the proportion of different components. DCas9-P65HSF1 mRNA, MS2-P65 mRNA and CT modified Tetra sgRNA pool at the ratio of 1:1:2 led to the best activation efficiency in CD3+ T cells (Fig. S6A). We further optimized the total amount of reagents used, and achieved more than 300 fold activation of OCT4 gene in primary T cells using CT modified Tetra sgRNA pool (Fig. 2G), with reasonable cell viability (Fig. S6B). Thus, the CT modified Tetra sgRNA system consisted of three components further improved the endogenous gene activation in primary T cells.
In this study, we demonstrated that: (I) sgRNA stability was enhanced by adding 5′ cap and 3′ polyA tail; (II) CT-modified sgRNA mediated higher level of genome disruption than unmodified sgRNA both in human cell line and primary T cells; (III) CT modified sgRNA with MS2 binding sites mediated efficient activation of endogenous gene expression.
By combining CT modification with sgRNA containing MS2 binding sites, we achieved efficient activation of endogenous genes in human primary T cells. The capability of controlling the expression of endogenous genes will help us better study the functions of different genes and the transcription networks in various primary cells. We have shown that genes important to T cell function, such as FOXP3, can be efficiently activated using our method, providing exciting possibilities of changing cell fate by regulating gene expression. Efficient activation of endogenous gene in primary cells holds great potential for both basic research and therapeutic applications.
We would like to thank Junning Wei and Yi Yang (Beijing Cord Blood Bank) for their help in preparing the cord blood samples. This work was supported by the National Natural Science Foundation of China (Grant No. 31471215), Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDA16010205), National Key Research and Development Program of China (No. 2016YFA0101402), and National High-tech R&D Program (863 Program) (No. 2015AA020307). H. Wang is supported by the “Young Thousand Talents Plan”.
W. Mu concept and design, collection and assembly of data, date analysis and interpretation, manuscript writing; Flow cytometry were conducted by Y. Zhang. X. Xue: collection of data; Templates for in vitro transcription was made by L. Liu. X. Wei prepared and provided essential reagents to the experiments. H. Wang: concept and design, manuscript writing, and final approval of the manuscript.
Wei Mu, Yongping Zhang, XuTong Xue, Lei Liu, Xiaofei Wei and Haoyi Wang declare that they have no conflict of interest. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000 (5). All donors of umbilical cord blood (UCB) units have provided informed consent.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.