Optimization of Cas9 RNA sequence to reduce its unexpected effects as a microRNA sponge
Molecular Cancer volume 21, Article number: 136 (2022)
Cas9 RNA functions as a miRNA sponge.
Let-7 is the dominant regulated miRNA by Cas9 RNA.
RNA sequence optimization of Cas9 by synonymous mutation improves its safety.
CRISPR–Cas9 gene-editing technology has a wide range of potential future applications, including cancer treatment . Currently, the main safety concerns about CRISPR–Cas9 technology are the host immune response to its components [2, 3] and off-target modifications . However, safety concerns at the RNA level have not been evaluated.
A wide range of interactions between microRNAs (miRNAs) and mRNAs occur, not only in the 3′ untranslated regions but also in the amino acid coding sequences (CDSs) of mRNA [5, 6]. An increasing number of research groups [7, 8], including our team [9,10,11], recently found that a long RNA can bind and adsorb partially complementary miRNAs through a “miRNA sponge” mechanism, upregulating other target genes inhibited by these miRNAs. In this study, we investigated whether Cas9 RNA, an exogenous long-chain nucleic acid substance, could affect host cell gene expression through a miRNA sponge mechanism and developed an RNA sequence optimization strategy to improve the safety of this gene-editing system.
Cas9 RNA could affect intracellular gene expression through a miRNA sponge mechanism
To analyze the regulatory effect of Cas9 on its predicted binding microRNAs and their target genes, we first mined a database with the transcriptome characteristics of 331 cell lines overexpressing a Cas9 RNA and their parental control counterparts . We also analyzed the binding possibilities between this Cas9 RNA and all the known human miRNAs using miRanda software . The 51 miRNAs that were predicted to have the highest likelihood of binding Cas9 RNA were named “Cas9-miRNAs” (Fig. 1A-B and Table S1), while the 69 miRNAs predicted not to bind Cas9 were named “non-Cas9-miRNAs” (Fig. 1A and Table S2). According to the microRNA sponge hypothesis, the expression levels of a certain miRNA’s targeting genes should be increased in cells into which Cas9 is introduced, and these cells could be identified as this miRNA related Cas9-sensitive cells, and vice versa. For most Cas9-miRNAs in most cell types, the change trends of their target genes were consistent with this hypothesis (Fig. 1C, Fig. S1). For example, more of the target genes of let-7i were upregulated after Cas9 introduction in 186 cell types, which should be identified as let-7i-related Cas9-sensitive cells, such as the T98G, TE-5 and 786-O cell lines (Fig. 1E, G, Fig. S1). Noteworthy, for the overall feature of 51 Cas9-miRNAs, their related Cas9-sensitive cells were significantly more than the Cas9-insensitive cells (p = 3.9e-11) (Fig. 1C). In contrast, the non-Cas9-miRNAs, such as miR-1-3p (Fig. 1D, F, H), miR-215-5p, and miR-340-5p (Fig. S1), did not show such a trend. We therefore suggest that Cas9 RNA may act as a microRNA sponge, specifically influencing Cas9-miRNA target genes.
Among all the Cas9-miRNAs, let-7i was the mainly expressed Cas9-miRNA in most cells (Fig. S2). As let-7 family members are well-known tumor suppressor miRNAs [13, 14], the possible “miRNA sponge” effect of Cas9 on this family is worthy of further study. We also found by analysis with miRanda software that Cas9 RNA can bind to many other let-7 family members (Fig. S3). Moreover, through qPCR and Western Blot experiments, we confirmed that some validated let-7 target genes were significantly upregulated after the introduction of Cas9 in more types of cells (U251, 786-O, T98G v.s. MCF7) (Fig. 1I and J, Fig. S4).
We then investigated why the microRNA sponge effect of Cas9 varies greatly among different cell lines. Again, we used let-7 as a representative microRNA. Based on a joint analysis of the transcriptome data of 331 cell lines and the basic miRNA and mRNA expression data and gene mutation data in the Cancer Cell Line Encyclopedia (CCLE) dataset, we identified the top 35 types of cells in which Cas9 induced the greatest increase in let-7i target genes as “let-7i-related Cas9-sensitive cells” and vice versa. We found that the basal expression of let-7 family members, especially let-7i, was lower in “let-7i-related Cas9-sensitive cells” rather than the control cells (Fig. 1K, Fig.S5). At the same time, the basal expression levels of let-7 target genes were also lower in “let-7i-related Cas9-sensitive cells” (Fig. 1L, Fig. S6). We also investigated the gene mutation frequency in the two groups of cells. Interestingly, we found almost all high-frequency mutations of let-7 target genes are distributed in the “let-7i-related Cas9-insensitive” cells (Fig. 1M, Fig. S7). These results suggested that, if some key let-7 target genes are mutated, the effect of Cas9 on up-regulating let-7 target genes might be weakened. All these data indicate that the sensitivity of cells to Cas9-mediated miRNA sponge effect may be determined by complicating factors, including the appropriate expression levels of Cas9-miRNAs and their target genes and the tumor mutation burden levels, especially for the miRNA target genes.
Next, we analyzed the miRNA sponge effect for more types of Cas9 RNA. We found that another popular Cas9 RNA used in our lab also had a very high possibility of binding the let-7 family, especially let-7b (Fig. S8A). Moreover, RNA pulldown experiments confirmed that let-7 could bind this Cas9 RNA directly (Fig. 1N). Moreover, Cas9 RNA was significantly decreased when let-7 mimics were introduced into HEK293-Cas9 cells stably expressing Cas9 (Fig. S8B). After Gene Expression Omnibus (GEO) analysis of the public dataset GSE84534 , we found that target genes of the let-7 family were positively enriched after introducing Cas9, which has the same mRNA sequence as the Cas9 in our laboratory (Fig. S8C-D). Several reported let-7 target genes were found to be upregulated in some tumor cells after this Cas9 introduction, such as CDK6, KRAS, FN1, and HMGA2 (Fig. 1O, S8E and S9A-B). However, the corresponding changes were not obvious after the introduction of excess let-7 (Fig. 1P), suggesting that Cas9 affects let-7 target genes through a “miRNA sponge” mechanism depending on the appropriate let-7 expression level, which is consistent with the results of previous bioinformatics analysis (Fig. 1K, Fig. S5). Furthermore, we found that dCas9-VP64, which was used for gene activation , could also increase the expression of some let-7 target genes in U251 cells (Fig. S9C).
Moreover, in Cas9-transgenic mice, KRAS and CDK6 expression levels were also found to be mildly increased in the testis tissues (Fig. 1Q), although they were not significantly changed in most of the tissues we tested (Fig. S10), which might also be associated with the basic expression levels of let-7 and its target genes.
These results suggest that Cas9 RNA could regulate let-7 target genes through the miRNA sponge mechanism in some types of cells, indicating that Cas9 itself may affect cells through mechanisms other than DNA cleavage.
RNA sequence optimization of Cas9 to reduce its effect on cell proliferation and let-7 downstream genes
The majority of let-7 downstream target genes are oncogenes [13, 14]. Since Cas9 can regulate the expression of some let-7 target genes, we speculate that exogenously introduced Cas9 may affect the biological characteristics of cells. We transduced the Cas9 viral vector, as well as the control virus, into the human prostate cancer cell line DU145 and found that Cas9 slightly promoted cell proliferation in a Cell Counting Kit-8 (CCK-8) experiment (Fig. S11A), cell cycle assay (Fig. S11B), and cell colony formation assay (Fig. 2A). In an in vivo experiment, we also found that the tumor growth rate and tumor volume were slightly higher in the group with stable Cas9 expression (Fig. 2B).
In addition, we introduced Cas9 into several other types of tumor cells, such as the human renal carcinoma cell line 786–O and the human colon carcinoma lines HT-29 and HCT116. Cas9 had no significant effect on HT-29 cells (data not shown) but slightly promoted the proliferation of HCT116 and 786–O cells (Fig. 2C, Fig. S11C). These results indicate that Cas9 does not have a significant effect on all cells but could slightly promote the proliferation of DU145 cells, 786–O cells and HCT116 cells.
Moreover, Cas9 could also increase some let-7 target genes in normal human cultured primary bone marrow mesenchymal stem cells (bMSCs) (Fig. S12 A-E) and the human normal epithelial cell line HaCaT (Fig. S12F-G) and could promote the proliferation of HaCaT cells (Fig. 2D-E).
In summary, we found that Cas9 RNA could slightly promote proliferation in some cells through a miRNA sponge mechanism. Although the effect is weak, especially in vivo, our observations suggested that it is necessary to improve the safety of CRISPR–Cas9 technology by modifying Cas9 itself in addition to reducing the off-target effect of the CRISPR–Cas9 system.
To further confirm the presence of this miRNA sponge mechanism and to optimize the Cas9 sequence at the RNA level, we constructed a synonymous mutant plasmid named Cas9-Mut. Cas9-Mut had mutations at the three let-7 binding sites; however, the mRNA transcript from Cas9-Mut could be translated into the same amino acid sequence as Cas9 (Fig. 2F). In fact, the RNA pulldown results in HEK293 cells showed that the binding of Cas9-Mut to let-7 was significantly reduced compared to that of Cas9 (Fig. 2G). Moreover, although Cas9-Mut somewhat promoted colony formation in DU145 cells compared with the control cells, its activity was weaker than that of Cas9 (Fig. S13). EdU assays showed that the proliferation ability of Cas9-Mut-transfected DU145 cells was weaker than that of Cas9-transfected cells (Fig. 2H). The percentage of G1 phase cells increased and that of G2 phase cells decreased in the Cas9-Mut group compared with the Cas9 group (Fig. 2I). At the molecular level, we found that the ability of Cas9-Mut to upregulate the let-7 target gene was also significantly weaker than that of Cas9, not only in DU145 cells (Fig. 2J-K) but also in U251 cells (Fig. 2L). Additionally, in hepatic carcinoma cell line HepG2, the cell proliferation ability of Cas9-Mut-transfected cells was weaker than that of Cas9-transfected cells (Fig. 2M). Although the current mutants cannot completely prevent the binding of let-7 and Cas9 RNA at other sites with higher binding free energy nor can they prevent the binding of other miRNAs that may affect proliferation, such as miR-145-5p, the partial effectiveness of the mutation suggests that Cas9 RNA can be further optimized in the future by the method of synonymous mutation.
Theoretically, because the amino acid sequence in Cas9-Mut is not changed, the ability to cleave DNA is not changed (Fig. 2F). To determine whether the optimized Cas9-Mut still has a gene-editing function, we first tested the ability of Cas9-Mut to knock out the EGFP reporter gene. After the introduction of Cas9-Mut and the designed EGFP-gRNA into HEK293 cell line cells stably expressing EGFP, some EGFP-negative cells appeared, indicating that Cas9-Mut still has a gene-editing function (Fig. 2N). We further tested the gene-editing ability of Cas9-Mut by transducing the Cas9-Mut plasmid and gRNA targeting the UBXN4 gene into DU145 cells (Fig. 2O). Genomic DNA was extracted, and the sequence around the predicted cutting site was amplified by PCR (Fig. 2O). Sanger sequencing of this PCR product showed double peaks around the predicted cutting site, which suggested successful DNA cutting (Fig. 2O). Furthermore, UBXN4 protein was not expressed in the selected single clones of DU145 cells by Western blotting (Fig. 2P). Therefore, the gene-editing function of Cas9-Mut was confirmed, providing support for its potential use in practical applications.
Our bioinformatic analysis and experimental results suggest that Cas9 RNA could interact with endogenous miRNA through the miRNA sponge mechanism, and it is valuable to optimize CRISPR–Cas9 technology at the RNA level for safer application in the future. Specifically, the current widely used Cas9 RNA could bind and sequester let-7, thus upregulating some let-7 target genes and slightly promoting cell proliferation in some cell types. Through synonymous mutation, the RNA sequence optimization of Cas9, introducing modified let-7 binding sites, weakened the effect on the promotion of cell proliferation and the expression of some let-7 downstream genes (Fig. 2N).
Clustered regularly interspaced short palindromic repeats
Gene set enrichment analysis
Normalized enrichment score
quantitative real-time PCR
Cell Counting Kit-8
Human normal epithelial cell line
Bone marrow mesenchymal stem cells
Enhanced Green Fluorescent Protein
Polymerase Chain Reaction
Bovine Serum Albumin
Cancer Cell Line Encyclopedia
Zhang H, et al. Application of the CRISPR/Cas9-based gene editing technique in basic research, diagnosis, and therapy of cancer. Mol Cancer. 2021;20(1):126.
Chew WL, et al. A multifunctional AAV-CRISPR-Cas9 and its host response. Nat Methods. 2016;13(10):868–74.
Charlesworth CT, et al. Identification of preexisting adaptive immunity to Cas9 proteins in humans. Nat Med. 2019;25(2):249–54.
Kleinstiver BP, et al. High-fidelity CRISPR-Cas9 nucleases with no detectable genome-wide off-target effects. Nature. 2016;529(7587):490–5.
Tay Y, et al. MicroRNAs to Nanog, Oct4 and Sox2 coding regions modulate embryonic stem cell differentiation. Nature. 2008;455(7216):1124–8.
Zhang K, et al. A novel class of microRNA-recognition elements that function only within open reading frames. Nat Struct Mol Biol. 2018;25(11):1019–27.
Hansen TB, et al. Natural RNA circles function as efficient microRNA sponges. Nature. 2013;495(7441):384–8.
Chen D-L, et al. The circular RNA circDLG1 promotes gastric cancer progression and anti-PD-1 resistance through the regulation of CXCL12 by sponging miR-141-3p. Mol Cancer. 2021;20(1):166. https://doi.org/10.1186/s12943-021-01475-8.
Wang Y, et al. Endogenous miRNA sponge lincRNA-RoR regulates Oct4, Nanog, and Sox2 in human embryonic stem cell self-renewal. Dev Cell. 2013;25(1):69–80.
Xiao G, et al. The long noncoding RNA TTTY15, which is located on the Y chromosome, promotes prostate Cancer progression by sponging let-7. Eur Urol. 2019;76(3):315–26.
Xu C, et al. Long non-coding RNA GAS5 controls human embryonic stem cell self-renewal by maintaining NODAL signalling. Nat Commun. 2016;7:13287.
Enache OM, et al. Cas9 activates the p53 pathway and selects for p53-inactivating mutations. Nat Genet. 2020;52(7):662–8.
Powers JT, et al. Multiple mechanisms disrupt the let-7 microRNA family in neuroblastoma. Nature. 2016;535(7611):246–51.
Busch B, et al. The oncogenic triangle of HMGA2, LIN28B and IGF2BP1 antagonizes tumor-suppressive actions of the let-7 family. Nucleic Acids Res. 2016;44(8):3845–64.
Dai W, et al. Cancer therapy with a CRISPR-assisted telomerase-activating gene expression system. Oncogene. 2019;38(21):4110–24.
We thank Novel Bioinformatics Co., Ltd. (Shanghai, China) and Xin Fu for bioinformatic analysis.
We thank Shurong Li (from Naval Medical University, Shanghai, China.) for her help in cell experiments.
This research was funded by the China National Key Research and Development Program Stem Cell and Translational Research Key Projects (2018YFA0108300), the National Natural Science Foundation of China (81972397,31971109, 31471390, and 81600926), the Shanghai Key Laboratory of Cell Engineering (14DZ2272300), Shanghai Rising-Star Program (22QA1411500, 17QA1405400) and Shanghai Health and family planning system program (2017YQ028).
Ethics approval and consent to participate
The whole study design and protocols were approved by the Ethics Committee of Naval Medical University, China, in accordance with the instruction about Care and Use of Laboratory Animals published by the USNIH.
Consent for publication
The authors approve the publication of this manuscript.
The authors declare that they have no competing financial interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Genome Editing and the Future of Personalized and Targeted Cancer Medicine
. MiRNA List of “Cas9-miRNAs”. Table S2. MiRNA List of “Non-Cas9-miRNAs”. Figure S1. Cas9 RNA showed the trend of disturbing endogenous genes through “miRNA sponge” mechanism. Figure S2. Basic expression levels of 12 Cas9-miRNAs in different cell lines. Figure S3. The let-7 family predicted to bind the RNA of Cas9 that has been introduced into 331 cell lines. Figure S4. Experimental validation of the expression levels of some let-7 target genes in MCF7. Figure S5. Appropriate expression level of let-7 family, especially let-7i-5p, might be related to whether Cas9 regulates the target genes of let-7i-5p. Figure S6. Basic expression level of let-7 target genes might be related to whether Cas9 regulates these genes by miRNA sponge mechanism. Figure S7. Mutation of let-7 target genes were related to whether Cas9 regulates let-7 target genes by miRNA sponge mechanism. Figure S8. Cas9 regulates the targets of let-7 through a sponge mechanism. Figure S9. Cas9 and dCas9-VP64 promote the target genes of let-7 after transduction through adenovirus. Figure S10. Cas9 slightly upregulated the expression of let-7 downstream genes in limited tissue samples from Cas9-transgenic mice. Figure S11. Cas9 slightly promoted the proliferation of DU145 and 786-O cells. Figure S12. Cas9 regulates the target genes of let-7 through a sponge mechanism in bMSC and Hacat cells. Figure S13. RNA sequence optimization of Cas9 could reduce its effect on cell proliferation in DU145.
About this article
Cite this article
Jiang, J., Zeng, T., Zhang, L. et al. Optimization of Cas9 RNA sequence to reduce its unexpected effects as a microRNA sponge. Mol Cancer 21, 136 (2022). https://doi.org/10.1186/s12943-022-01604-x
- miRNA sponge
- RNA sequence optimization