Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Cell. Author manuscript; available in PMC 2017 April 7.
Published in final edited form as:
PMCID: PMC4826288

Programmable RNA tracking in Live Cells with CRISPR/Cas9


RNA-programmed genome editing using CRISPR/Cas9 from Streptococcus pyogenes has enabled rapid and accessible alteration of specific genomic loci in many organisms. A flexible means to target RNA would allow alteration and imaging of endogenous RNA transcripts analogous to CRISPR/Cas-based genomic tools, but most RNA targeting methods rely on incorporation of exogenous tags. Here we demonstrate that nuclease-inactive S. pyogenes CRISPR/Cas9 can bind RNA in a nucleic acid-programmed manner and allow endogenous RNA tracking in living cells. We show that nuclear-localized RNA-targeting Cas9 (RCas9) is exported to the cytoplasm only in the presence of sgRNAs targeting mRNA and observe accumulation of ACTB, CCNA2 and TFRC mRNAs in RNA granules that correlate with fluorescence in situ hybridization. We also demonstrate time-resolved measurements of ACTB mRNA trafficking to stress granules. Our results establish RCas9 as a means to track RNA in living cells in a programmable manner without genetically encoded tags.

Graphical abstract

An external file that holds a picture, illustration, etc.
Object name is nihms764827u1.jpg


Clustered Regularly-Interspaced Short Palindromic Repeats (CRISPRs) form the basis of adaptive immune systems in bacteria and archaea by encoding CRISPR RNAs that guide CRISPR-associated (Cas) nucleases to invading genetic material (Wiedenheft et al., 2012). Cas9 from the type II CRISPR system of S. pyogenes has been repurposed for genome engineering in eukaryotic organisms (Hwang et al., 2013; Li et al., 2013a; Mali et al., 2013; Nakayama et al., 2013; Sander and Joung, 2014; Yang et al., 2014) and is rapidly proving to be an efficient means of DNA targeting for other applications such as gene expression modulation (Qi et al., 2013) and imaging (Chen et al., 2013). Cas9 and its associated single guide RNA (sgRNA) require two critical features to target DNA: a short DNA sequence of the form 5′-NGG-3′ (where ‘N’ = any nucleotide) known as the protospacer adjacent motif (PAM) and an adjacent sequence on the opposite DNA strand that is antisense to the sgRNA. By supporting DNA recognition with specificity determined entirely by a short spacer sequence within the sgRNA, CRISPR/Cas9 provides uniquely flexible and accessible manipulation of the genome. Manipulating cellular RNA content, in contrast, remains problematic. While there exist robust means of attenuating gene expression via RNA interference and antisense oligonucleotides, other critical aspects of post-transcriptional gene expression regulation such as subcellular trafficking, alternative splicing or polyadenylation, and spatiotemporally-restricted translation are difficult to measure in living cells and are largely intractable.

Analogous to the assembly of zinc finger nucleases (Urnov et al., 2010) and transcription activator-like effector nucleases (TALEN) to recognize specific DNA sequences, efforts to recognize specific RNA sequences have focused on engineered RNA binding domains. Pumilio and FBF homology (PUF) proteins carry well-defined modules capable of recognizing a single base each and have supported successful targeting of a handful of transcripts for imaging and other manipulations (Filipovska et al., 2011; Ozawa et al., 2007; Wang et al., 2009). PUF proteins can be fused to arbitrary effector domains to alter or tag target RNAs, but PUFs must be redesigned and validated for each RNA target and can only recognize 8 contiguous bases which does not allow unique discrimination in the transcriptome. Molecular beacons are self-quenched synthetic oligonucleotides that fluoresce upon binding to target RNAs and allow RNA detection without construction of a target-specific protein (Sokol et al., 1998). But molecular beacons must be microinjected to avoid the generation of excessive background signal associated with endosome-trapped probes and are limited to imaging applications. An alternative approach to recognition of RNA substrates is to introduce RNA aptamers into target RNAs, enabling specific and strong association of cognate aptamer binding proteins such as the MS2 coat protein (Fouts et al., 1997). This approach has enabled tracking of RNA localization in living cells over time with high sensitivity (Bertrand et al., 1998) but relies upon laborious genetic manipulation of the target RNA and is not suitable for recognition of arbitrary RNA sequences. Furthermore, insertion of exogenous aptamer sequence has the potential to interfere with endogenous RNA functions. Analogous to CRISPR/Cas9-based recognition of DNA, programmable RNA recognition based on nucleic acid specificity alone without the need for genetic manipulation or libraries of RNA binding proteins would greatly expand researchers’ ability to modify the mammalian transcriptome and enable transcriptome engineering.

Although the CRISPR/Cas9 system has evolved to recognize double-stranded DNA, recent in vitro work has demonstrated that programmable targeting of RNAs with Cas9 is possible by providing the PAM as part of an oligonucleotide (PAMmer) that hybridizes to the target RNA (O’Connell et al., 2014). By taking advantage of the Cas9 target search mechanism that relies on PAM sequences (Sternberg et al., 2014), a mismatched PAM sequence in the PAMmer/RNA hybrid allows exclusive targeting of RNA and not the encoding DNA. The high affinity and specificity of RNA recognition by Cas9 in cell-free extracts and the success of genome targeting with Cas9 indicate the potential of CRISPR/Cas9 to support programmable RNA targeting in living cells.

To assess the potential of Cas9 as a programmable RNA binding protein in live cells, we used a modified sgRNA scaffold with improved expression and Cas9 association (Chen et al., 2013) with a stabilized PAMmer oligonucleotide that does not form a substrate for RNAse H. We measured the degree of nuclear export of a nuclear localization signal-tagged nuclease-deficient Cas9-GFP fusion and demonstrate that the sgRNA alone is sufficient to promote nuclear export of Cas9 without influencing the abundance of the targeted mRNA or encoded protein. In order to evaluate whether RNA-targeted Cas9 (RCas9) signal patterns correspond with an established untagged RNA labeling method, we compared distributions of RCas9 and fluorescence in situ hybridization (FISH) targeting ACTB mRNA. We observed high correlation among FISH and RCas9 signal that was dependent on the presence of a PAMmer, indicating the importance of the PAM for efficient RNA targeting. RNA trafficking and subcellular localization are critical to gene expression regulation and reaction to stimuli such as cellular stress. To address whether RCas9 allows tracking of RNA to oxidative stress-induced RNA/protein accumulations called stress granules, we measured ACTB, TFRC, and CCNA2 mRNA association with stress granules in cells subjected to sodium arsenite. Finally we demonstrated the ability of RCas9 to track trafficking of ACTB mRNA to stress granules over time in living cells. This work establishes the ability of RCas9 to bind RNA in live cells and sets the foundation for manipulation of the transcriptome in addition to the genome by CRISPR/Cas9.


RNA-targeted Cas9 Export from Nucleus in Presence of sgRNA targeting GAPDH mRNA

We initially assessed the ability of RCas9 to recognize specific mRNA substrates in human cells by evaluating the degree of nuclear export of a nuclear-localized RCas9 system. Specifically, we tested if mCherry-tagged Cas9 containing a nuclear localization signal (NLS) can be co-exported from the nucleus with an mRNA in the presence of a cognate sgRNA and PAMmer designed to recognize that mRNA (Figure 1A). Nuclease-null Cas9 (dCas9) was fused to two SV40 NLS sequences at the C-terminus with the coding sequence for mCherry and cloned into a mammalian expression vector (dCas9-2xNLS-mCherry, abbreviated as dCas9-mCherry). In a separate expression vector, a modified sgRNA scaffold with an extended stem-loop structure that improves association with Cas9 and mutations that eliminate a partial transcription termination sequence (Chen et al., 2013) was driven by the U6 snRNA polymerase III promoter. The PAMmer was synthesized as a mixed DNA and 2′-O-methyl (2′OMe) RNA oligonucleotide using standard phosphoramidite chemistry and purified using HPLC (see Tables S1 and S2 for target mRNA, sgRNA and PAMmer sequences). As a proof-of-concept, we designed an sgRNA-PAMmer pair to target the 3′ untranslated region (3′UTR) of GAPDH mRNA (Figure 1B). As a negative control, we designed an sgRNA-PAMmer pair targeting a sequence in the λ bacteriophage that is absent in human cells (“N/A” sgRNA and PAMmer, “λ2” sgRNA and PAMmer sequences in Table S1–2). We observed that transiently transfected dCas9-mCherry co-transfected with the negative control sgRNA and PAMmer is almost exclusively nuclear with only 12% of cells containing greater mCherry signal in the cytoplasm compared to the nucleus (Figures 1B and 1C). When the negative control PAMmer was replaced with the GAPDH-targeting PAMmer, the results were identical. Importantly, upon co-transfection of GAPDH-targeting sgRNA plasmid, we observed that 84% of cells had greater mCherry signal in the cytoplasm. Interestingly, even with a non-targeting PAMmer, 79% of cells had predominantly cytoplasmic mCherry signal, suggesting that the sgRNA is the primary determinant of RNA substrate recognition. Indeed, the sgRNA targeting GAPDH resulted in a significant increase in the fraction of cells with cytoplasmic mCherry signal compared to a non-targeting sgRNA (Figure 1C). Overall, these results are consistent with previous in vitro RNA pull-down experiments showing that RNA binding by Cas9:sgRNA is independent of but strengthened by the PAMmer (O’Connell et al., 2014). Thus, we demonstrate that RCas9 RNA recognition is programmable with an sgRNA and a PAMmer designed to target a specific, abundant mRNA in live cells.

Figure 1
Targeting mRNA in living cells with RNA-targeted Cas9 (RCas9). (A) Components required for RCas9 targeting of mRNA include a nuclear localization signal-tagged nuclease-inactive Cas9 fused to a fluorescent protein such as GFP, a modified sgRNA with expression ...

Recognition of an mRNA with RNA-targeted Cas9 does not alter RNA abundance or amount of translated protein

To further characterize the interaction between RCas9 and a target mRNA, we directed RCas9 to the 3′UTR of Renilla luciferase carrying a commonly-used RNA tag for RNA tracking from the MS2 bacteriophage (Fouts et al., 1997) and a sequence targeted by a previously validated sgRNA:PAMmer pair (“λ2”, see Table S2) (O’Connell et al., 2014) (Figure 1D). RNA immunoprecipitation with an antibody recognizing EGFP revealed a four-fold greater association of luciferase mRNA to dCas9-EGFP in the presence of a cognate sgRNA and PAMmer compared to non-targeting sgRNA (sense to the λ2 RNA sequence) or scrambled PAMmer or to EGFP protein alone (Figure 1E). We next measured the effect of RCas9 targeting on luciferase mRNA abundance by quantitative RT-PCR. We observed no significant difference in the abundance of MS2-tagged luciferase mRNA in the presence of the targeting or non-targeting RCas9 system or EGFP alone. In contrast, co-expression of EGFP fused to the MS2 coat protein (MCP) recognizing the MS2 aptamer had a significant stabilizing effect (Figure 1F). We also considered potential effects of RCas9 targeting on translation of luciferase (Figure 1G) and observed that the presence of the targeting sgRNA and PAMmer caused no significant changes in protein levels compared to non-targeting RCas9. To confirm that the RCas9/sgRNA/PAMmer complex does not perturb its endogenous RNA substrates, we evaluated the influence of RCas9 targeting on ACTB and GAPDH mRNAs. We transiently transfected the RCas9 system and isolated transfected cells using fluorescence-activated cell sorting (FACS) gated on cells positive for both Cas9 and sgRNA transfection. We observed no significant differences in GAPDH and ACTB mRNA or protein levels as measured by western blot and quantitative RT-PCR among samples with RCas9 targeting ACTB mRNA, GAPDH mRNA, cells transfected with GFP instead of dCas9-GFP, and RCas9 targeting sequence from λ bacteriophage (Figure S1, Table S1–2). Our results demonstrate that RCas9 recognition of RNA with an sgRNA and PAMmer does not perturb RNA and encoded protein levels.

Correlation of RNA-targeted Cas9 signal distributions with an established untagged RNA localization measurement

To assess whether RCas9 signal distributions correlate with an orthogonal method to measure RNA localization, we targeted the 3′UTR of ACTB (“+” sgRNA and “+” PAMmer) and compared dCas9-GFP signal to RNA fluorescence in situ hybridization (FISH) for ACTB mRNA (Figure 2A) and non-targeting sgRNA and PAMmer (“−“ sgRNA” and “−“ PAMmer with sequences corresponding to λ bacteriophage). By comparing the Mander’s overlap coefficients that describe pixel-by-pixel overlap among FISH and RCas9 (Manders et al., 1992) (Figures 2B and S2), we determined that the sgRNA primarily accounts for co-localization among FISH and RCas9 with maximal overlap in the presence of both sgRNA and PAMmer targeting ACTB mRNA. A non-targeting PAMmer results in a significantly less overlap (Figure 2B) (p=0.035, Mann-Whitney U test) and produces a diffuse pattern of RCas9 signal in the cytoplasm that contrasts with the highly localized pattern revealed by FISH (Figure 2A). This result is consistent with weaker binding of RCas9 with a non-targeting PAMmer observed in cell-free systems (O’Connell et al., 2014). A non-targeting sgRNA results in largely nuclear retention of RCas9 signal with low correlation between cytoplasmic RCas9 signal and FISH (Figures 2A and 2B). We conclude that the localization patterns of single EGFP-fused RCas9 in live cells correlates surprisingly well with localization obtained by FISH using fifty probes per mRNA in fixed cells.

Figure 2
ACTB mRNA localization with RCas9 compared to FISH. (A) The RCas9 system was delivered to U2OS cells and the cells were subjected to FISH for ACTB mRNA. RCas9 with sgRNA and PAMmer targeting ACTB mRNA was compared to non-targeting sgRNA and PAMmer antisense ...

Tracking RNA trafficking to stress granules over time

Many polyadenylated RNAs including ACTB mRNA are known to localize to stress granules (Unsworth et al., 2010) during oxidative stress. We simultaneously tracked ACTB mRNA using RCas9 and RFP fused to the Ras GTPase-activating protein-binding protein 1 (G3BP1) protein, a well-described marker for stress granules (Tourriere et al., 2003). Application of sodium arsenite to induce cellular stress results in an accumulation of mRNAs, including ACTB mRNA, into G3BP1-positive granules. We observed accumulation of RCas9 signal to G3BP1-positive foci in the presence of the RCas9 system targeting ACTB mRNA. To verify that this accretion of RCas9 signal required specific RNA recognition, we utilized three different sgRNAs and PAMmers targeting sequences from λ bacteriophage (NTC1-3) (Figures 3A and 3B). Indeed, in these negative controls, increased signal in G3BP1-positive foci was not observed. We also evaluated the ability of RCas9 to track less-abundant transcripts CCNA2 and TFRC mRNA to stress granules. Notably, we observed statistically significantly higher fractions of stress granules that accumulate RCas9 signal (39% and 23%, respectively) compared to non-targeting controls (Figures 3A and 3C). Next we tracked RCas9 signal in stressed, live cells over time (Figure 3B). We observed accumulation of RCas9 signal in G3BP1-positive foci in a manner dependent on the presence of sgRNA and PAMmer targeting ACTB mRNA. We also observed that the rate and degree of RCas9 signal accumulation in stress granules is dependent on dosage of the stressor sodium arsenite (Figure 3D). These results indicate the potential of RCas9 as a means to generate time-resolved RNA localization measurements.

Figure 3
Tracking of mRNA trafficking to stress granules with RCas9. (A) The RCas9 system targeting ACTB, TFRC, or CCNA2 mRNAs or one of three non-targeting controls (NTC) was delivered to HEK293T cells expressing G3BP1, a protein known to be trafficked to stress ...


Effective RNA recognition by Cas9 in living cells while avoiding perturbation of the target transcript relies on careful design of the PAMmer and delivery of Cas9 and its cognate guide RNA to the appropriate cellular compartments. Binding of Cas9 to nucleic acids requires two critical features: a PAM DNA sequence and an adjacent spacer sequence antisense to the Cas9-associated sgRNA. By separating the PAM and sgRNA target among two molecules (the PAMmer oligonucleotide and the target mRNA) that only associate in the presence of a target mRNA, RCas9 allows recognition of RNA while avoiding the encoding DNA. To avoid unwanted degradation of the target RNA, the PAMmer is composed of a mixed 2′OMe RNA and DNA that does not form a substrate for RNAse H. Further, the sgRNA features a modified scaffold that removes partial transcription termination sequences and a modified structure that promotes association with Cas9 (Chen et al., 2013). Other CRISPR/Cas systems have demonstrated RNA binding in bacteria (Hale et al., 2009; Sampson et al., 2013) or eukaryotes (Price et al., 2015), although these systems cannot discriminate RNA from DNA targets, feature RNA targeting rules that remain unclear, or rely on large protein complexes that may be difficult to reconstitute in mammalian cells.

In this work, we demonstrate RCas9-based recognition of GAPDH, ACTB, CCNA2 and TFRC mRNAs in live cells. Since the U6-driven sgRNA is largely restricted to the nucleus, the NLS-tagged dCas9 allows association with its sgRNA and subsequent interaction with the target mRNA before nuclear co-export with the target mRNA. As an initial experiment, we evaluated the potential of RNA recognition with Cas9 by targeting GAPDH mRNA and evaluating degree of nuclear export of dCas9-mCherry (Figure 1B). Robust cytoplasmic localization of dCas9-mCherry in the presence of an sgRNA targeting GAPDH mRNA compared to nuclear retention in the presence of a non-targeting sgRNA indicated that Cas9 association with the mRNA was sufficiently stable to support co-export from the nucleus.

RCas9 as an RNA imaging reagent requires that RNA recognition by RCas9 does not interfere with normal RNA metabolism. Here, we show that RCas9 binding within the 3′UTR of Renilla luciferase does not affect its mRNA abundance and translation (Figures 1F and 1G). The utility of RCas9 for imaging and other applications hinges on the recognition of endogenous transcripts, so we evaluated the influence of RCas9 targeting on GAPDH and ACTB mRNAs and observed no significant differences among the mRNA and protein abundances by western blot analysis and qRT-PCR (Figure S1). These results indicate that RCas9 targeting these 3′UTRs does not perturb the levels of mRNA or encoded protein.

We also evaluated the ability of RCas9 to reveal RNA localization by comparing RCas9 signal patterns to FISH. We utilized a FISH probe set composed of 50 singly-labeled probes targeting ACTB mRNA and compared FISH signal distributions to a single dCas9-GFP/sgRNA/PAMmer that recognizes the ACTB mRNA. Our findings indicate that the sgRNA primarily determines the degree of overlap among the FISH and RCas9 signals while the PAMmer plays a significant but secondary role. Importantly, in contrast to other untagged RNA localization determination methods such as FISH and molecular beacons, RCas9 is compatible with tracking untagged RNA localization in living cells and can be delivered rapidly to cells using established transfection methods. We also note that the distribution of ACTB mRNA was visualized using a single EGFP tag per transcript and higher-sensitivity RNA tracking or single endogenous RNA molecule visualization may be possible in the future with RCas9 targeting multiple sites in a transcript or with a multiply-tagged dCas9 protein.

Stress granules are translationally-silent mRNA and protein accumulations that form in response to cellular stress and are increasingly thought to be involved with neurodegeneration (Li et al., 2013b). There are limited means that can track the movement of endogenous RNA to these structures in live cells (Bertrand et al., 1998). In addition to ACTB mRNA, we demonstrate that RCas9 is capable of measuring association of CCNA2 and TFRC mRNA trafficking to stress granules (Figure 3A). Upon stress induction with sodium arsenite, we observed that 50%, 39% and 23% of stress granules featured overlapping RCas9 foci when targeting ACTB, TFRC, and CCNA2 mRNAs, respectively (Figure 3C). This result correlates with the expression levels of these transcripts (Figure S3) as ACTB is expressed about 8 and 11 times more highly than CCNA2 and TFRC, respectively. We also observed that RCas9 is capable of tracking RNA localization over time as ACTB mRNA is trafficked to stress granules over a period of 30 minutes (Figure 3B). We noted a dependence of RCas9 signal accumulation in stress granules on stressor concentration (Figure 3D). This approach for live cell RNA tracking stands in contrast to molecular beacons and aptamer-based RNA tracking methods which suffer from delivery issues and/or require alteration of the target RNA sequence via incorporation of RNA tags.

Future applications of RCas9 could allow the measurement or alteration of RNA splicing via recruitment of split fluorescent proteins or splicing factors adjacent to alternatively spliced exons. Further, the nucleic acid-programmable nature of RCas9 lends itself to multiplexed targeting (Cong et al., 2013) and the use of Cas9 proteins that bind orthogonal sgRNAs (Esvelt et al., 2013) could support distinct activities on multiple target RNAs simultaneously. It is possible that the simple RNA targeting afforded by RCas9 could support the development of sensors that recognize specific healthy or disease-related gene expression patterns and reprogram cell behavior via alteration of gene expression or concatenation of enzymes on a target RNA (Delebecque et al., 2011; Sachdeva et al., 2014). Efforts towards Cas9 delivery in vivo are underway (Dow et al., 2015; Swiech et al., 2015; Zuris et al., 2015), and these efforts combined with existing oligonucleotide chemistries (Bennett and Swayze, 2010) could support in vivo delivery of the RCas9 system for targeted modulation of many features of RNA processing in living organisms.

RNA is subject to processing steps that include alternative splicing, nuclear export, subcellular transport, and base or backbone modifications that work in concert to regulate gene expression. The development of a programmable means of RNA recognition in order to measure and manipulate these processes has been sought after in biotechnology for decades. This work is, to our knowledge, the first demonstration of nucleic acid-programmed RNA recognition in living cells with CRISPR/Cas9. By relying upon an sgRNA and PAMmer to determine target specificity, RCas9 supports versatile and unambiguous RNA recognition analogous to DNA recognition afforded by CRISPR/Cas9. The diverse applications supported by DNA-targeted CRISPR/Cas9 range from directed cleavage, imaging, transcription modulation, and targeted methylation, indicating the utility of both the native nucleolytic activity of Cas9 as well as the range of activities supported by Cas9-fused effectors. In addition to providing a flexible means to track this RNA in live cells, future developments of RCas9 could include effectors that modulate a variety of RNA processing steps with applications in synthetic biology and disease modeling or treatment.


Plasmid construction, PAMmer synthesis, and target site choice

The dCas9-2xNLS sequence was amplified from pHR-SFFV-dCas9-BFP-KRAB (a gift from Stanley Qi & Jonathan Weissman, Addgene plasmid #46911), tagged with two SV40 nuclear localization signals (NLS) on the C-terminus, and fused to EGFP or mCherry in pCDNA 3.1 (Life Technologies, Carlsbad CA) using Gibson assembly. To construct the sgRNA scaffold construct, the human U6 polymerase III promoter with the modified sgRNA scaffold (Chen et al., 2013) was purchased as a gBlock from IDT with two BbsI restriction sites at the 5′ end of the sgRNA scaffold (see sequence in Table S2) and cloned into the multiple cloning site of pBlueScript II SK (+) (Agilent, Santa Clara, CA) using Gibson assembly. Phosphorylated oligonucleotides encoding the sgRNA sequences (with overhangs 5′CACC on the RNA antisense strand and 5′AAAC on the sense strand) were ligated into BbsI-digested sgRNA scaffold construct to produce sgRNAs targeting the 3′UTR of GAPDH, ACTB, and Renilla luciferase mRNAs (see Tables S1 and S2). The luciferase-PEST construct for pull-down and abundance experiments was modified from plasmid pRLuc (gift from Jens Lykke-Andersen, UCSD). pCMV-Renilla luciferase is a version of the same construct lacking MS2 and RCas9 target sites.

RCas9 target sites were chosen with a combination of the IDT antisense oligonucleotide design tool and the microarray probe design tools Picky (Chou et al., 2004) and OligoWiz (Wernersson and Nielsen, 2005). We designed PAMmers against high-confidence sites with 8 bases on the 5′ end beyond the PAM sequence. PAMmers were composed of mixed 2′OMe RNA and DNA bases and purified by HPLC (Integrated DNA Technologies, Coralville IA).

Cell lines

U2OS and HEK293T cells were grown in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum, glutamax, penicillin/streptomycin and non-essential amino acids (Life Technologies). Cells were passaged every 3–4 days with TrypLE EXPRESS (Life Technologies) using standard methods and maintained in a humidified incubator at 37 °C with 5% CO2.

GAPDH and ACTB mRNA targeting with RCas9

U2OS cells cultured as described above were passaged at ~80% confluency. Glass-bottom 96-well plates or chamber slides were coated with 20 μg/mL fibronectin in PBS for 2h at 37 °C, then the fibronectin solution was aspirated and 20,000 cells were plated in each well. 16 hours later, cells were transfected with the sgRNA and dCas9-2xNLS-EGFP plasmids using Lipofectamine 3000 (Life Technologies) according to the manufacturer’s instructions. pCMV-Renilla luciferase was co-transfected in these experiments so that total transfected protein load was the similar among various dosages of sgRNA and dCas9. The mass ratio of sgRNA and dCas9-EGFP plasmids was 5:1 in experiments described in this work where the amount of dCas9-EGFP was fixed at 10% of total transfected material. Immediately after plasmid transfection, PAMmers were transfected using Lipofectamine RNAiMax (Life Technologies) according to manufacturer’s instructions. 24 hours after transfection, cells were washed with PBS and fixed with 3.7% paraformaldehyde in PBS, permeabilized with 70% ethanol at 4 °C for one hour, and mounted using Prolong Gold Antifade mounting medium with DAPI (Life Technologies). Confocal microscopy was conducted using a Zeiss LSM 810 confocal microscope.

Nuclear export of RCas9 in the presence of sgRNA and PAMmer targeting the 3′UTR of GAPDH mRNA was analyzed by measuring the average signal in the nuclei and cytoplasm of individual cells. Cells with average cytoplasmic signal greater than the average nuclear signal were considered positive and counted towards the fraction of cells with cytoplasmic RCas9 signal.

RNA immunoprecipitation

HEK293T cells cultured as described above were passaged at 80% confluency and 600,000 cells were seeded in each well of 6-well tissue culture plates coated with poly-L-lysine. 16 hours later, cells were co-transfected with the RCas9 system as described above or plasmids encoding MS2-EGFP or EGFP along with a plasmid encoding the model Renilla luciferase mRNA driven by a CMV promoter (Figure 1D). 24 hours later, the growth media was aspirated and the cells were washed with PBS. 1% paraformaldehyde in PBS was applied to the cells, incubated for 10 minutes at room temperature, then the solution was aspirated and the cells washed twice with cold PBS. Next, the cells were scraped from the wells in cold PBS and the cell suspension was centrifuged at 800×g for 4 minutes to pellet the cells. The cells were washed once more, then resuspended in RIPA buffer with protease inhibitors (Roche) and sonicated for 5 minutes in a Bioruptor sonicator (50% duty cycle, 1 minute period). Insoluble material was pelleted after a high-speed centrifugation and the supernatant was applied to protein G DynaBeads (Life Technologies) coated with mouse anti-GFP antibody (Roche). After overnight incubation at 4 °C, the be ad supernatant was retained and beads washed 3 times with RIPA buffer containing 0.02% Tween-20 and once with DNase buffer (350mM Tris-HCl, pH 6.5; 50mM MgCl2; 5mM DTT). The beads were resuspended in DNase buffer and TURBO DNase (Life Technologies) was added to 0.08 units/μL. The beads were incubated at 37 °C for 30 minutes, then proteinase K (NEB) was added to 0.1 U/μL and incubated with shaking at 37 °C for 30 minutes. Next, urea was added to 2.5 M and the beads were incubated with shaking at 37 °C for 30 minutes. The bead supernatant was collected and subjected to a two sequential phenol:chloroform:isoamyl alcohol (25:24:1) extractions followed by three chloroform extractions. The RNA was precipitated and reverse transcribed using SuperScript III (Life Technologies) using random hexamer primers, and relative abundance of Renilla luciferase RNA on the beads was compared to the supernatant using RT-PCR (see Table S3 for primer sequences).

Measurements of influence of RCas9 on RNA stability and translation

HEK293T cells were cultured as described above, passaged and plated in 96- or 12- well tissue culture plates, and were co-transfected 24h later with the RCas9 system as described above and the Renilla luciferase construct carrying MS2 and RCas9 binding sites in the 3′UTR. In the protein abundance measurements, a small amount of CMV-driven firefly luciferase vector (5% of total transfected plasmid) was co-transfected as a transfection control. For RNA stability measurements, RNA was isolated 24h after transfection, DNase treated, reverse transcribed with Superscript III (Life Technologies) using dT(20) primers according the manufacturer’s instructions. The amount of Renilla luciferase cDNA relative to GAPDH was then measured using RT-PCR (see Table S3 for primer sequences). For the translation studies, Renilla and firefly luciferase protein were measured with the Dual Luciferase Kit (Promega) according to the manufacturer’s instructions.

For measuring the influence of RCas9 on GAPDH and ACTB mRNA and protein abundance, the RCas9 system targeting the 3′UTR of each transcript, GFP alone, or RCas9 with the λ2-targeting sgRNA and PAMmer was transfected. Modified sgRNA vectors carrying pGK-driven blue fluorescent protein (BFP) were used in this experiment and transfected cells were isolated using FACS gated on cells positive for GFP and BFP expression. Transfected cells were lysed subjected to western blotting and RT-PCR for ACTB, GAPDH, and TUBA1A transcripts.

Fluorescence in situ hybridization

Stellaris FISH Probes recognizing human ACTB, CCNA2, and TFRC mRNAs and labeled with Quasar 670 (VSMF-2003-5, Biosearch Technologies, Inc., Petaluma, CA) were hybridized to cells 24 hours after transfection with the RCas9 system. Hybridization was conducted according to the manufacturer’s instructions. Confocal microcscopy was conducted using an Olympus FV1000 confocal microscope or Zeiss LSM 810 confocal microscope.

Overlap analysis for fluorescence in situ hybridization and RCas9

Colocalization analysis among FISH and RCas9 targeting ACTB mRNA was conducted using the Coloc 2 plugin from the image analysis software FIJI (Schindelin et al., 2012). The cytoplasm of individual cells with similar dCas9-EGFP transfection levels was selected and the Coloc 2 analysis was conducted using default parameters. The Mander’s overlap coefficient describing degree of overlap of the FISH signal with RCas9 for more than 60 cells in each condition was compiled and p-values were calculated with the two-tailed Mann-Whitney U test.

Tracking ACTB mRNA trafficking to stress granules

A HEK293T cell line was genetically modified with a fusion of RFP to the C-terminus of Ras GTPase-activating protein-binding protein 1 (G3BP1) using CRISPR/Cas9. Briefly, a donor plasmid was constructed consisting of the RFP ORF, a puromycin selection cassette, and flanking 1.5kb homology arms directed at the G3BP1 locus. An sgRNA sequence targeting the C-terminus of G3BP1 was cloned into pSpCas9(BB)-2A-GFP (pX458) (gift from Feng Zhang, Addgene plasmid #48138) and co-transfected with the donor plasmid using Fugene HD (Roche) following the manufacturer’s instructions. 48 hours after transfection, cells were selected with 1 μg/mL puromycin in growth medium for 14 days and RFP-positive clones were selected and screened by PCR.

A clone with at least one modified allele was plated on glass chamber slides coated with fibronectin and transfected with the RCas9 system targeting the 3′UTR of ACTB, CCNA2, or TFRC mRNA or one of three non-targeting sgRNA and PAMmer combinations (λ2-λ4) as described above (see Table S2 for sgRNA and PAMmer sequences). 24 hours after transfection, cells were subjected to 200 μM sodium arsenite for 30 minutes, fixed, and subjected to FISH and imaged with a Zeiss LSM 810 confocal microscope. In a separate experiment, cells were stressed with varying concentrations of sodium arsenite and imaged with a Zeiss LSM 810 confocal microscope with a stage incubator over 32 minutes. Cells were maintained at 37 °C in a humidified atmosphere and 5% CO2 and imaged every 8 minutes.

Analysis of ACTB mRNA trafficking to stress granules

Stress granules with average RCas9 signal at least 2-fold greater than surrounding cytoplasm were determined to be positive for RCas9. The average RCas9 signal in stress granules was calculated using the ‘measure’ function in Fiji and compared to the average cytoplasmic RCas9 signal in the surrounding cytoplasm. RCas9-positive foci were visualized by dividing the value of each pixel by the average cytoplasmic signal which reduced cytoplasmic signal to a value of ‘1’ and RCas9-positive foci to a value of ‘2’ or more. Next, a Gaussian blur was applied to improve foci definition. These images are reported in Figure 3A. The accumulation of RCas9 signal over time in stress granules was calculated by recording average signal intensity in the RCas9 channel in areas with overlapping G3BP1-RFP foci at each time point. Average signal intensity in the RCas9 channel surrounding G3BP1-RFP foci was recorded as background and subtracted from the previous value to produce the background-adjusted RCas9 signal in stress granules.


  • RNA-targeted Cas9 (RCas9) enabled recognition of endogenous, unmodified mRNAs
  • RCas9 did not influence mRNA abundance or amount of translated protein
  • Subcellular distribution of RCas9 was highly correlated with RNA-FISH
  • RCas9 revealed trafficking of mRNAs to stress granules in live cells

Supplementary Material


We acknowledge members of the Yeo lab, particularly Ashleigh Schaffer, Stefan Aigner and Ron Batra for critical comments. This work was supported by grants from the National Institute of Health (HG004659 and NS075449) and from the California Institute of Regenerative Medicine (RB3-05009 and RB4-06045) to G.W.Y. D.A.N. is supported by the National Science Foundation Graduate Research Fellowship (DGE-1144086) and Powell-Focht graduate fellowship. M.R.C is supported by a CJ Martin fellowship from the National Health and Medical Research Council (Australia). J.A.D. is a Howard Hughes Medical Institute Investigator. G.W.Y. is an Alfred P. Sloan Research Fellow.


Author contributions

Conceptualization, D.A.N., M.R.O., J.A.D., G.W.Y.; Methodology, D.A.N. and M.R.O.; Investigation, D.A.N., M.Y.F., J.L.X. S.J.M.; Writing – Original Draft, D.A.N., G.W.Y.; Writing – Review & Editing, D.A.N., G.W.Y., M.R.O.; Funding Acquisition, G.W.Y.; Supervision, G.W.Y.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Bennett CF, Swayze EE. RNA targeting therapeutics: molecular mechanisms of antisense oligonucleotides as a therapeutic platform. Annu Rev Pharmacol Toxicol. 2010;50:259–293. [PubMed]
  • Bertrand E, Chartrand P, Schaefer M, Shenoy SM, Singer RH, Long RM. Localization of ASH1 mRNA particles in living yeast. Mol Cell. 1998;2:437–445. [PubMed]
  • Chen B, Gilbert LA, Cimini BA, Schnitzbauer J, Zhang W, Li GW, Park J, Blackburn EH, Weissman JS, Qi LS, et al. Dynamic imaging of genomic loci in living human cells by an optimized CRISPR/Cas system. Cell. 2013;155:1479–1491. [PMC free article] [PubMed]
  • Chou HH, Hsia AP, Mooney DL, Schnable PS. Picky: oligo microarray design for large genomes. Bioinformatics. 2004;20:2893–2902. [PubMed]
  • Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, Hsu PD, Wu X, Jiang W, Marraffini LA, et al. Multiplex genome engineering using CRISPR/Cas systems. Science. 2013;339:819–823. [PMC free article] [PubMed]
  • Delebecque CJ, Lindner AB, Silver PA, Aldaye FA. Organization of intracellular reactions with rationally designed RNA assemblies. Science. 2011;333:470–474. [PubMed]
  • Dow LE, Fisher J, O’Rourke KP, Muley A, Kastenhuber ER, Livshits G, Tschaharganeh DF, Socci ND, Lowe SW. Inducible in vivo genome editing with CRISPR-Cas9. Nat Biotechnol. 2015;33:390–394. [PMC free article] [PubMed]
  • Esvelt KM, Mali P, Braff JL, Moosburner M, Yaung SJ, Church GM. Orthogonal Cas9 proteins for RNA-guided gene regulation and editing. Nat Methods. 2013;10:1116–1121. [PMC free article] [PubMed]
  • Filipovska A, Razif MF, Nygard KK, Rackham O. A universal code for RNA recognition by PUF proteins. Nat Chem Biol. 2011;7:425–427. [PubMed]
  • Fouts DE, True HL, Celander DW. Functional recognition of fragmented operator sites by R17/MS2 coat protein, a translational repressor. Nucleic Acids Res. 1997;25:4464–4473. [PMC free article] [PubMed]
  • Hale CR, Zhao P, Olson S, Duff MO, Graveley BR, Wells L, Terns RM, Terns MP. RNA-guided RNA cleavage by a CRISPR RNA-Cas protein complex. Cell. 2009;139:945–956. [PMC free article] [PubMed]
  • Hwang WY, Fu Y, Reyon D, Maeder ML, Tsai SQ, Sander JD, Peterson RT, Yeh JR, Joung JK. Efficient genome editing in zebrafish using a CRISPR-Cas system. Nat Biotechnol. 2013;31:227–229. [PMC free article] [PubMed]
  • Li D, Qiu Z, Shao Y, Chen Y, Guan Y, Liu M, Li Y, Gao N, Wang L, Lu X, et al. Heritable gene targeting in the mouse and rat using a CRISPR-Cas system. Nat Biotechnol. 2013a;31:681–683. [PubMed]
  • Li YR, King OD, Shorter J, Gitler AD. Stress granules as crucibles of ALS pathogenesis. J Cell Biol. 2013b;201:361–372. [PMC free article] [PubMed]
  • Mali P, Yang L, Esvelt KM, Aach J, Guell M, DiCarlo JE, Norville JE, Church GM. RNA-guided human genome engineering via Cas9. Science. 2013;339:823–826. [PMC free article] [PubMed]
  • Manders EM, Stap J, Brakenhoff GJ, van Driel R, Aten JA. Dynamics of three-dimensional replication patterns during the S-phase, analysed by double labelling of DNA and confocal microscopy. J Cell Sci. 1992;103(Pt 3):857–862. [PubMed]
  • Nakayama T, Fish MB, Fisher M, Oomen-Hajagos J, Thomsen GH, Grainger RM. Simple and efficient CRISPR/Cas9-mediated targeted mutagenesis in Xenopus tropicalis. Genesis. 2013;51:835–843. [PMC free article] [PubMed]
  • O’Connell MR, Oakes BL, Sternberg SH, East-Seletsky A, Kaplan M, Doudna JA. Programmable RNA recognition and cleavage by CRISPR/Cas9. Nature. 2014;516:263–266. [PMC free article] [PubMed]
  • Ozawa T, Natori Y, Sato M, Umezawa Y. Imaging dynamics of endogenous mitochondrial RNA in single living cells. Nat Methods. 2007;4:413–419. [PubMed]
  • Price AA, Sampson TR, Ratner HK, Grakoui A, Weiss DS. Cas9-mediated targeting of viral RNA in eukaryotic cells. Proc Natl Acad Sci U S A. 2015;112:6164–6169. [PubMed]
  • Qi LS, Larson MH, Gilbert LA, Doudna JA, Weissman JS, Arkin AP, Lim WA. Repurposing CRISPR as an RNA-guided platform for sequence-specific control of gene expression. Cell. 2013;152:1173–1183. [PMC free article] [PubMed]
  • Sachdeva G, Garg A, Godding D, Way JC, Silver PA. In vivo co-localization of enzymes on RNA scaffolds increases metabolic production in a geometrically dependent manner. Nucleic Acids Res. 2014;42:9493–9503. [PMC free article] [PubMed]
  • Sampson TR, Saroj SD, Llewellyn AC, Tzeng YL, Weiss DS. A CRISPR/Cas system mediates bacterial innate immune evasion and virulence. Nature. 2013;497:254–257. [PMC free article] [PubMed]
  • Sander JD, Joung JK. CRISPR-Cas systems for editing, regulating and targeting genomes. Nat Biotechnol. 2014;32:347–355. [PMC free article] [PubMed]
  • Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012;9:676–682. [PMC free article] [PubMed]
  • Sokol DL, Zhang X, Lu P, Gewirtz AM. Real time detection of DNA.RNA hybridization in living cells. Proc Natl Acad Sci U S A. 1998;95:11538–11543. [PubMed]
  • Sternberg SH, Redding S, Jinek M, Greene EC, Doudna JA. DNA interrogation by the CRISPR RNA-guided endonuclease Cas9. Nature. 2014;507:62–67. [PMC free article] [PubMed]
  • Swiech L, Heidenreich M, Banerjee A, Habib N, Li Y, Trombetta J, Sur M, Zhang F. In vivo interrogation of gene function in the mammalian brain using CRISPR-Cas9. Nat Biotechnol. 2015;33:102–106. [PMC free article] [PubMed]
  • Tourriere H, Chebli K, Zekri L, Courselaud B, Blanchard JM, Bertrand E, Tazi J. The RasGAP-associated endoribonuclease G3BP assembles stress granules. J Cell Biol. 2003;160:823–831. [PMC free article] [PubMed]
  • Unsworth H, Raguz S, Edwards HJ, Higgins CF, Yague E. mRNA escape from stress granule sequestration is dictated by localization to the endoplasmic reticulum. FASEB J. 2010;24:3370–3380. [PubMed]
  • Urnov FD, Rebar EJ, Holmes MC, Zhang HS, Gregory PD. Genome editing with engineered zinc finger nucleases. Nat Rev Genet. 2010;11:636–646. [PubMed]
  • Wang Y, Cheong CG, Hall TM, Wang Z. Engineering splicing factors with designed specificities. Nat Methods. 2009;6:825–830. [PMC free article] [PubMed]
  • Wernersson R, Nielsen HB. OligoWiz 2.0--integrating sequence feature annotation into the design of microarray probes. Nucleic Acids Res. 2005;33:W611–615. [PMC free article] [PubMed]
  • Wiedenheft B, Sternberg SH, Doudna JA. RNA-guided genetic silencing systems in bacteria and archaea. Nature. 2012;482:331–338. [PubMed]
  • Yang D, Xu J, Zhu T, Fan J, Lai L, Zhang J, Chen YE. Effective gene targeting in rabbits using RNA-guided Cas9 nucleases. J Mol Cell Biol. 2014;6:97–99. [PMC free article] [PubMed]
  • Zuris JA, Thompson DB, Shu Y, Guilinger JP, Bessen JL, Hu JH, Maeder ML, Joung JK, Chen ZY, Liu DR. Cationic lipid-mediated delivery of proteins enables efficient protein-based genome editing in vitro and in vivo. Nat Biotechnol. 2015;33:73–80. [PMC free article] [PubMed]