• Home /
  • Blog /
  • Optimization of CRISPR Delivery in Cultured Cells & Single Cell Embryos

Optimization of CRISPR Delivery in Cultured Cells & Single Cell Embryos

Jun 13, 2016 11:22:03 AM No Comments

The CRISPR/Cas9 system has been rapidly adapted to practically every model system for its ease to generate and high efficiencies to cleave target DNA. But unlike our experience with Zinc Finger Nucleases, in the human, rat and mouse cell lines we tried successful co-transfection of Cas9 mRNA and sgRNA was cell-line dependent, and often resulted in either very low or no cleavage activities.

However, sequential transfection of cells with Cas9 DNA first, and sgRNA followed 24 hrs later, reliably produced good level of activity, indicating the requirement of Cas9 presence at the time of introduction of sgRNA. Not surprisingly, creation of a cell line stably expressing Cas9 led to consistently high cleavage activities upon transfection of sgRNAs. Transfection of recombinant Cas9 protein pre-complexed with sgRNA (ribonucleoprotein particles, or RNPs) led to efficient cleavage as well.

On the other hand, when Cas9 mRNA and sgRNAs are co-microinjected into single cell embryos, it produces target cleavage as efficiently as RNPs to produce straight KOs and large deletions between two target sites, again raising a question of local concentrations of Cas9 protein and sgRNA.

Below we summarize some of the work we've done optimizing delivery of CRISPR-Cas9, which which can be read in full publication form in Human Gene Therapy here.

Co-transfection of sgRNA and Cas9 mRNA does not lead to consistent cleavage activity in cells

Figure 1 shows the relatively low cleavage activity observed when using Cas9 mRNA and sgRNA. Cleavage in the rat C6 cell line was rarely observed (Fig 1A), with slight improvements in the mouse Neuro-2A cell line (Fig. 1B).

Cotransfection of Cas9 mRNA and sgRNA

Figure 1:  A Two sgRNAs, one targeting the rat ApoE gene (Lanes 1-6) and the other, the Rosa26 locus (Lanes 8-16), were nucleofected along with Cas9 mRNA (R) or Cas9 expression plasmid (D) into the rat C6 cell line. Cells were collected at 8, 24, and 48 hours post nucleofection, and target sites were PCR amplified and analyzed by Cel-I assay. Transfection of ZFN mRNA (ZR) targeting the Rosa26 locus was used as a positive control. It introduced in/dels and generated cleaved bands of expected 234 bp and 161 bp in Cel-I assay (Lanes 17-19). Lanes 7 and 20 are transfection controls. The positions of expected but absent cleaved bands are marked with arrowheads.

B The same test in the mouse Neuro-2a cell line: Lanes 1-7, cells transfected with sgRNA against Ptsg1 gene, and Lanes 8-16, sgRNA mRosa26, targeting the mouse Rosa26 locus, and Lanes 17-19, ZFN mRNA (ZR) against Rosa26 locus.

Sequential transfection, Cas9 stable lines and RNPs are comparably active

When Cas9 expression plasmid was first transfected, and followed by transfection of in vitro-transcribed sgRNA 24h later, we observed consistent and efficient cleavage at target sites in both rat C6 cells (Figure 2A) and mouse Neuro-2A cells (not shown). These results imply that presence of Cas9 protein at the time of introducing sgRNAs into the cells helps CRISPR activity, and thus we created stable Cas9 cell lines to test this further. NHEJ events were observed at 8h post transfection (Fig 2B &2C). 

With the Cas9 protein expressed constitutively and consistently, sgRNA transfection into Cas9 stable cell lines eliminates the need for the extra transfection step using Cas9 plasmid, reduces experimental variability and provides consistent results, again suggesting that the presence of the Cas9 protein at the time of sgRNA introduction helps complex formation.

We believe that using Cas9 stable cell lines is the most efficient and reliable means to validate sgRNA activity.


sgRNA Transfection

Figure. 2 Alternative formats of CRISPR delivery led to efficient modification of the target sites.

A The rat C6 cells were first transfected with the Cas9 expression plasmid and followed by transfection of ApoE sgRNA 24 hours later. Cells were then collected at different time points after transfection for analysis. Expected bands are marked with asterisks, and sizes in number of base pairs. The bands below the 231 and 224 bp doublet are nonspecific amplifications.

B sgRNAs for ApoE and rRosa26 were transfected into a rat C6 stable line expressing Cas9 protein, and Cel-I assay was done at different time post transfection.

C sgRNAs for mPtsg1 and mRosa26 were tested in a mouse Neuro-2a cell line stably expressing Cas9

D A more detailed time course of nuclease activity detected in sgRNA mRosa26 transfected into a Neuro-2a Cas9 stable cell line.

Microinjection of RNA mixture and RNPs

Keeping Cas9 levels constant (5ng/ul) we observed a reduction in cleavage rate with lowered sgRNA concentration. Conversely, keeping sgRNA level constant (10ng/ul), we found that Cas9 mRNA concentration was not critical in the range we tested, and at all concentrations cleavage rate was >50%.

Cas9 mRNA (ng/ul) sgRNA (ng/ul) Cleavage Rate (%)
60 10 75
20 10 69.2
10 10 83.3
5 15 40.9
5 2.5 28

Table 1 Titration of Cas9 mRNA and sgRNA for microinjection into single-cell embryos. Each embryo was analyzed for NHEJ separately. Concentrations of Cas9 mRNA tested did not prove to be critical.

Cas9 protein vs mRNA for gene replacement

We compared Cas9 protein and Cas9 mRNA for the generation of rat pups with gene replacement at the ApoE locus (Fig 3A). Following coinjection of Cas9 mRNA with sgRNA and donor (Fig 3B), no founders was identified amount 56 pups. Following one injection session with Cas9 RNP a founder was obtained from 8 pups.

Relative efficiencies of other genomic modifications can be found in Table 2.

Gene replacement ApoE rat

Humanized ApoE

Figure 3 Gene replacement at the rat ApoE locus.

A. Schematic of the rat ApoE locus. The 1.8 kb rat ApoE coding sequence between the translational start and stop codons is replaced by the human counterpart, including introns. sgRNAs were designed and validated to target at the start and stop codons.

B. Schematic of the donor plasmid. The human ApoE coding sequence (marked as human ApoE gene) is flanked by 800 bp homology arms (HA-L and HA-R). 


% of founders among live births

Cas9 mRNA

Cas9 Protein

# of Projects



# of Projects



Mouse, KI







Mouse point mutation







Mouse,  KO including large deletion, 1-20kb







Mouse, floxing







Rat, point mutation







Rat, KI cDNA







Rat,  KO including large deletions, 1-569kb







Table 2 Comparison of targeted integration rates among mice or rats born to coinjection of sgRNA, donor and Cas9 in either mRNA or protein format. Cas9 mRNA was injected at 60 ng/ul, sgRNA at 25 to 75 ng/ul, Cas9 protein at 50-100 ng/ul. Single stranded oligo donors were used at 100 ng/ul each, while plasmid plasmid donors were injected at 1-2 ng/ul.

Analysis of sgRNA competition

The low toxicity of CRISPR in cells and embryos alows multiplexing in targeting. One natural concern is whether there is competition between sgRNAs for binding the Cas9 protein and form the complex. To test this we formed RNPs by combining Cas9 protein and a target sgRNA with a competing sgRNA in one tube, Cas9 protein to total sgRNA at 1:1 mass ratio to maximize potential competition. Two of the four sgRNAs against rat targets, Rosa26, ApoE #3, #7, and C2 were co-complexed with Cas9 protein.

The three more active sgRNAs, Rosa26, #3 and #7 were not detectably affected by the presence of other sgRNAs in the reaction, and all their target sites were cleaved as efficiently as by single RNPs. We do observe some competition against the relatively weak sgRNA at low concentrations, but the competition is not predictable nor universal, even at protein limiting concentrations. For more on competition analysis please refer to our publication.

In Vitro assay competition

Figure 4. Competition assay in cells. RNPs were formed by combining Cas9 protein with one or two sgRNAs, as indicated above the panels. Target sites are marked on top of the panels. Each RNP complex was tested for cleaving each target site by transfection into the rat C6 cells.


Co-transfection of Cas9 mRNA/DNA and sgRNA into cell lines does not reliably result in cleavage, with efficiency depending on cell line, method of transfection, transfection efficiency, cell line passage number, etc.

We have found that sequential transfection, Cas9 DNA first and sgRNA 24 hours later can produce more consistent activity in cells, suggesting that local concentrations of Cas9 protein and sgRNA are critical to nuclease activity. This is further supported by data from our Cas9-expressing stable cell lines, which we have found to be the fastest way to screen sgRNA for activity. Finally we found that Cas9 protein/sgRNA (RNP) complex works comparably as well as Cas9 stable cell lines.

For embryos, co-injection of Cas9 mRNA/sgRNA mixture works very well, although RNPs seem to have advantages over mRNAs at certain target sites for KIs. Titration of CRISPR reagents for microinjection showed that concentration range of Cas9 mRNA tested (20-60 ng/ul) did not prove to be critical, and produced NHEJ activity >50%. Lower concentration (5ng/ul) still produced NHEJ much higher than in cell transfections.

If you would like to read more about this work, or access the methods please take a look at the complete paper published in Human Gene Therapy. 

Download the complete paper 

#Cell lines, #Gene editing, #Microinjection

Subscribe to Email Updates