Protocols for sgRNA design, RNA preparation, injection and high resolution melt analysis
Requirements of the system
sgRNAs are designed to match a 20 nt target sequence in the DNA followed by a protospacer adjacent motif (PAM) sequence of NGG. This NGG is absolutely required for cleavage, although it does not appear in the sgRNA.
Since we use a T7 promoter to drive transcription of the sgRNA, the first two bases of the sgRNA are required to be a GG dinucleotide as well, making the final target sequence in the DNA as follows
20 nt Target site PAM
Recently, it has been found that extensions at the 5' end of the sgRNA may be tolerated (link1, link2, link3), making it possible to relax the requirement for the GG at the beginning of the guide sequence, simply by appending a GG to any 20 nt sequence.
T7 20 nt Target site PAM
GG NNNNNNNNNNNNNNNNNNNN NGG
Off target effects
A perfect match to the final 12 nt of the target site (the "seed") and PAM sequence (NGG) are required for efficient cleavage, but mismatches in the remaining sequence are tolerated, and do not strongly affect cleavage efficiency (Cong et al. 2013, figure 3). In our recent paper, we have also shown that even a 1 nt mismatch within this 12 nt seed sequence is sufficient to abolish cleavage.
12 nt "seed" PAM
NNNNNNNN NNNNNNNNNNNN NGG
When designing sgRNAs, efforts should therefore be made to avoid perfect matches to the "seed" sequence, followed by NGG at other places in the genome. Those sites with minimal off target effects have been identified by Feng Zhang and colleagues, and can be viewed on the UCSC genome browser (link). The ZiFit target design software also can help in designing CRISPR target sites. Other websites providing more information can be found on the CRISPR community page.
Recent results in human cells show that these rules for off target effects are oversimplified, and that it is safer to avoid sequences with fewer than 3-4 nt mismatches to the entire 20 nt seed sequence. (Fu et al. 2013)
RNA preparation for injection
We use two overlapping oligonucleotides to produce the template for sgRNA
1. CRISPRF - specific for the desired target site, containing the GGN18-20 guide sequence and T7 promoter
2. sgRNAR - A common oligo containing the sgRNA stem loop structures necessary for incorporation into Cas9
Overlap between the two sequences is indicated in bold, and the sequence to replace with the target site is GGN18-20. The PCR product is then used as a template for synthesis of the sgRNA using the Megascript kit (Ambion).
Cas9 mRNA is made from Pme I linearised MLM3613 (Joung lab), with the addition of a 5' cap and 3' poly A tail using the mMESSAGEmMACHINE and poly(A) tailing kits (Ambion).
Injection of Drosophila embryos
We inject Drosophila embryos through the chorion essentially as described in the following protocols (from the Carroll lab, or the following embryo collection and embryo injection CSH protocols). We use Femtotip II needles (Eppendorf) that are broken before use, and a 1:1 mixture of oxygenated halocarbon 700 and halocarbon 27 oils (Sigma) for injection.
High resolution melt analysis (HRMA) of indels
We perform high resolution melt analysis HRMA on PCR products of 100-200 bp across the predicted CRISPR target site in individual mosaic flies of the injected generation (after crossing!) to look for insertions and deletions created by imperfect non-homologous end joining (NHEJ) repair. This needs to be done with saturating DNA dyes such as LCGreen to obtain melt curve resolution (NOT Sybr green commonly used for quantitiative PCR).
The PCR needs to be optimised to give a single product, by adjusting annealing temperatures and DMSO concentration. The products generated can then the analysed by dedicated HRMA equipment such as the LightScanner, or a quantitative PCR machines with appropriate software (available for most models).
This allows analysis not only of mosaic flies of the injected (G0) generation, but also of heterozygous backcrossed lines to identify those ones that contain a mutation at the target site(s) of interest.