How to knock-out plant genes using RNA-guided Cas9
RNA-guided Cas9 can be used to create a lesion in ‘your favourite gene’ by inducing double strand break(s) (DSBs) at a specified DNA target. Plant cells are most likely to mend induced DSBs using the endogenous DNA repair mechanism of non-homologous end-joining (NHEJ). To create a knock-out it is sometimes easiest to create a pair of double strand breaks, deleting a fragment of the gene because a deletion is easier to screen for than a small mutation. This protocol describes how to assemble plasmid vectors for Cas9-induced targeted mutagenesis using TypeIIS mediated assembly. If you are not familiar with this technique then please take our online course (& if you’re at TSL, please email or drop in and see Mark Youles if you need advice or help).
The Cas9 and sgRNAs genes are introduced into the plant genome as a transgene cassette. Stable integration increases the likelihood of recovering a mutation because the integrated transgene will be continuously making DSBs: eventually the repair will be imperfect. To obtain a transgene-free mutated plant, the first generation of transgenic plants in which mutations are identified are ‘selfed’. The transgene cassette and edited target will most likely be inherited as two unlinked traits. Therefore, as long as the cas9 transgene integrates as a single (or low) copy, a portion of the next generation will not inherit the transgene but will inherit the deletion in the target gene. In some cases the initial mutation may not take place early in development resulting in a chimeric plant. In this case the next generation will need to be carefully screened. In Arabidopsis, in which floral tissues are transformed, some labs have reported difficulties with inheritance because certain promoters do not express in germ-line cells. For any new species, you might need to try more than one promoter for success.
In barley we have successfully used the promoter and 5′ UTR from Ubiquitin (Zea mays) to express Cas9.
In Brassica oleracea we have successfully used the promoter and 5′ UTR from CSVMV to express Cas9
In tomato and Nicotiana we have successfully used the CaMV-35s promoter with the TMV omega sequence to express Cas9.
STEP 1
Select targets for cas9 within your gene(s) of interest:
The target consists of 20 bp followed by an “NGG” PAM (protospacer adjacent motif) sequence. In plants, sgRNAs have mostly been expressed from promoters that use RNA polymerase dependent III such as U6 and U3 because they have a strict transcriptional start. The first transcribed base is also the first base of the spacer that will recognise the target (see figure 1). For U6 promoters the start of transcription is a ‘G’. Ideally, your target should begin with a G so that the spacer is positioned at the start of the transcribed sgRNA : 5’ GNNNN NNNNN NNNNN NNNNN NGG 3’
However, if no suitable targets beginning with a G are identified, then 5’ NNNNN NNNNN NNNNN NNNNN NGG 3’ can be used instead. An additional G will be included before the first base to maintain the native start of transcription. This will mean that the sgRNA has one additional base before the spacer. Additional 5′ bases have been shown not to interfere with recognition or cleavage of the target and are likely to be accessible to exonucleases as they will protrude from the cas9 complex (See Figure 1).
Evidence from mammals suggests that it may be best to avoid pyrimidines (C or T) in the last 5-10 bases before the PAM. Some software (e.g. Vector NTI, CLC) allows a degenerate sequence search. In this case case search your sequence for 5’ GNNNN NNNNN RRRRR RRRRR NGG 3’. Note that this has not been systematically verified in plants and successful cleavage of pyrimidine-rich targets have also been published.
If there is concern about ‘off-target’ activity, then search the rest of the genome with the potential target sequence. An exact match to 3′ half of the sgRNA (known as the seed region) followed by a PAM motif, may be a target for off-target activity. If there are mismatches in the 3′ half of the guide RNA, or if the PAM is absent, then the sequence is unlikely to be cleaved. Do note, however, that this is extrapolation from research in mammalian systems where cleavage efficiency is much higher.
There are several online-tools that are free-for-academics that will choose targets in your GOI indicating, where a complete genome sequence is available, potential off-target sites. Try:
STEP 2
Design primers to makeLevel 1 plasmid vectors to express your sgRNAs
You will introduce your chosen target into the sgRNA scaffold by PCR. You will then combine this PCR product with a U6 or U3 promoter in a one-step one-pot reaction.
The forward primer will contain the 20 bp guide sequence specific to your chosen target as a 5′ tail. Note that the PAM is not part of the sgRNA (see Figure above). If the Arabidopsis U6-26 (pICSL9002) promoter is being used then the forward primer should look like this:
tgtggtctca ATTG NNNN NNNNN NNNNN NNNNN gttttagagctagaaatagcaag
(BsaI site is in blue, the 20 bp guide sequence is in red, the 3′ end in black, lower-case anneals to the sgRNA target)
Your guide sequence will replace the red ‘N’s. Do not alter the black, lower-case sequence – this is the part of the primer that sticks to the template!
If the wheat U6 promoter (pICSL90003) is being used then the foreward primer will look like this:
tgtggtctca CTTG NNNN NNNNN NNNNN NNNNN gttttagagctagaaatagcaag
(BsaI site is in blue, the 20 bp guide sequence is in red, the 3′ end in black lower-case anneals to the sgRNA target)
Your guide sequence will replace the red ‘N’s. Do not alter the black, lower-case sequence – this is the part of the primer that sticks to the template!
If in step 1 you did not choose a target starting with a G then an additional nucleotide should be included in the forward primer. The G must be included as this is (a) the start of transcription and (b) part of the 4 bp GoldenGate overhang by which the promoter and the sgRNA will join together. The primer (and resulting guide) will be one base pair longer. The G will protrude from the sgRNA-cas9 duplex.
tgtggtctca ATTG N NNNN NNNNN NNNNN NNNNN gttttagagctagaaatagcaag
(BsaI site is in blue, the 20 bp guide sequence is in red, the 3′ end in black lower-case anneals to the sgRNA target)
Your guide sequence will replace the red ‘N’s. Do not alter the black, lower-case sequence – this is the part of the primer that sticks to the template!
The reverse primer will always look exactly like this if you are using the template described below as an sgRNA scaffold:
tgtggtctca AGCG TAATGCCAACTTTGTAC
(BsaI site is in blue, the 3′ end in black lower-case anneals to the sgRNA target)
STEP 3
Use your pair of primer pairs to amplify a unique sgRNA with proofreading polymerase:
The vector pICSL70001 (Addgene #46966) contains the sgRNA scaffold. Amplification with the primers pair from the previous step will result in a PCR product like the one below, with your target replacing the 19 (or 20 if your target does not begin with a G) red ‘N’s:
tgtggtctcaATTG(N)NNNNNNNNNNNNNNNNNNNGTTTTAGAGCTAGAAAT
AGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGA
GTCGGTGCTTTTTTTCTAGACCCAGCTTTCTTGTACAAAGTTGGCATTACGC
Ttgagaccaca
STEP 4.
Clean up your PCR amplicon(s) and quantify the amount of DNA that you have.
STEP 5
Set up dig-lig reactions with each PCR product, a suitable level 1 acceptor and a plasmid containing the small RNA promoter (e.g. pICSL0001, 2 or 3).
The exact level 1 acceptor (i.e. position) depends on how many sgRNAs you are making and what else you are including in your final construct (see step 6 below). You will need to plan your whole cloning strategy in advance and then return to this step when you know which position level 1 acceptor to choose.
Each amplicon is combined with a plasmid containing the chosen U6 or U3 promoter and the chosen level 1 acceptor. A standard GoldenGate digestion-ligation reaction will assemble the promoter and sgRNA into a complete transcriptional unit in the acceptor. The digestion-ligations reaction transformed into E. coli and white colonies are selected on agar (with X-Gal and carbenicillin). This will result in Level 1 sgRNA expression cassettes (see Figure 2).
6. There are three main ways in which a multi-gene construct containing a selection marker, cas9 and custom guide RNAs can be assembled. The choice depends on (a) how many guide RNAs are required and (b) if the experiment and/or species demands a custom selectable marker or cas9 cassette (e.g. promoters/selection specific to the species).
a) The simplest strategy is to use a level 2 vector that already contains a plant selection cassette and cas9 cassette in the backbone (e.g pICSL002207 or pICSL002208 contain cassettes for kanamycin resistance and and 35s:cas9). Note that these acceptor plasmids are unpublished and therefore only available to those at TSL. These vectors will accept up to 6 guide RNAs cloned into level 1 position 1-6 (see Option 1 in figure 3, below). You will always need to include the correct end-linker to join the last level 1 module to the GGGA overhang of the level 2 acceptor in the final reaction.
b) A second strategy is to use an empty level 2 vector (e.g. pAGM4723 /Addgene #48015) and to include any (existing or novel) plant selectable-marker cassette in position 1 and any (existing or novel) Cas9 cassette (e.g. pICSL11021/ AddGene #49771) position 2. This option is suitable when the selection and cas9 cassettes available in the option a, above, are not suitable for your experiment or species. To use this strategy the guide RNAs must be assembled into Level 1 acceptors from position 3 onwards. A maximum of 4 sgRNAs can be used. Option 1 in figure 3, below, shows the introduction of two sgRNAs into an empty level 2 vector. You will always need to include the correct end-linker to join the last level 1 module to the GGGA overhang of the level 2 acceptor in the final reaction.
(c) The last option applies when more than 6 sgRNAs are to be assembled. In this case it will be necessary to use the Level M/P cloning system (see Figure 4, below). Selection and cas9 cassettes to suit the experiment/species are assembled into the Level M Position 1 plasmid (in pAGM8031). The sgRNAs are assembled cloned into the correct level 1 acceptors to fit into the Level M position 2+ acceptors (pAGM8043 onwards). The position of each acceptor will depend on how many transcriptional units are in each level M acceptor. The multiple level Ms are then joined together in a level P Position 1 acceptor (If at TSL/JIC use pAGM4723-P). It is possible to assemble a very large number of transcriptional units using level M and P. Remember to include the correct end-linker to join the last level 1 module to the GGGA overhang of the level M acceptor. Also, remember to include the correct end-linker to join the last module to the GGGA overhang of the level P acceptor in the final reaction. NB Users at TSL should check the database for additional suitable Level 1/2/M modules or come and discuss their project in person.
Step 7.
Transform your final level 2/M/P construct into Agrobacterium and proceed with plant transformation as normal.
Step 8.
Apply suitable selection to recover primary (T0) transgenic plants from tissue culture. When the plantlets are big enough, take leaf tissue samples and extract DNA.
Step 9.
If a pair of sgRNAs were used to create a deletion, amplify this region with a pair of primer that flank the putatively deleted sequence.
Evidence of a deletion can be seen in the form of amplicons smaller than those obtained from a wild type control (Figure 5 a, below). The absence of the wild-type amplicon may indicate that the deletion was homozygous (lane 3). The direct sequencing of this PCR product may confirm if both sister chromatids were repaired in the same way or if the plant is bialleic. If an amplicon corresponding to the wild-type is also present (lane 2) then the deletion may be heterozygous or, alternatively, the transgenes may be expressed in somatic tissues with different cells in the sample showing multiple genotypes. In all cases the seeds of these plants should be collected. In the next generation, null-segregent progeny, which have not inherited the transgene, and progeny that have inherited the transgene analysed in the same way. The mutation can be classified as heritable and stable when progeny with the same mutant genotype as the parent are recovered and the transgene has been segregated out. This may take more than one generation and, depending on your aims, you may wish to progress though multiple generations to confirm the heritability and stability of the mutation.
If there is no evidence of smaller band indicating a deletion then two experiments are possible:
The first is to digest the purified genomic DNA with a restriction endonuclease with one or more recognition sites between the targets and to PCR amplify the locus with oligonucleotide primers designed to the flanking regions (Figure 5 b) . This pre-digestion will remove any wild-type sequence enabling the detection of deletions even if they occurred in just a few cells in your sample (lane 4). Such plants are highly likely to be chimeric and will need to be progressed to a second generation.
The second method allows the detection of small insertion-deletion events at the cut-site rather than a deletion. A DSB is most likely to occur three base pairs before the PAM in the seed-region of the target (Figure 5c, below). If there is a restriction endonuclease recognition site that would be disrupted by imperfect repair of the DSB, a PCR amplicon of the target locus can be digested with this enzyme. Any amplicon showing resistance to digestion with this enzyme can be sequenced (Figure 5c, Lane 4 upper band). It would be prudent to design your target regions to contain RE sites that could be used for such screening. Again, the mutation can be classified as heritable and stable when progeny with the same mutant genotype as the parent are recovered and the transgene has been segregated out. This may take more than one generation and, depending on your aims, you may wish to progress though multiple generations to confirm the heritability and stability of the mutation.