One-step dual CRISPR/Cas9 guide RNA cloning protocol

Existing protocols for dual guide RNA cloning rely on synthesised DNA oligonucleotides of >100 bp that contain both guide RNA sequences, and are therefore not reusable in alternative experimental designs. Here, we describe a single-step protocol to rapidly and inexpensively generate vectors expressing two guide RNAs (gRNAs) simultaneously, which allows re-usage of gRNAs oligonucleotides from one experimental design to another. This protocol is applicable to cloning gRNAs into virtually any CRISPR/Cas9 backbone that allows cloning by Golden Gate, by adapting the primer design. Here, we provide details for cloning gRNAs into vectors with BbsI and BsmBI sites, two of the most frequently found enzymes in CRISPR/Cas9 gRNA expression cassettes. This protocol has been successfully applied to delete pancreatic islet enhancers that harbour type 2 diabetes variants and to validate enhancer-promoter interactions (Miguel-Escalada et al., Nature Genetics 2019). In the future, we foresee that this simple protocol may also be applied to target coding sequences, as well as to target other important kinds of noncoding regulatory elements, including lncRNAs, miRNAs, and chromatin structural anchor points.


Introduction
The CRISPR/Cas9 system is now a well-established lab tool to edit virtually any sequence in the genomes of human cells and model systems. CRISPR/Cas9 is most frequently applied to introduce random indels in target DNA sequences to generate frameshift mutations in coding sequences. This strategy however is not applicable to non-coding sequences. Non-coding sequences harbour a large pairs of gRNAs into expression vectors, namely for fast generation of CRISPR/Cas9 deletion libraries 8,9 . However, these methods rely on synthesis of long oligonucleotides (>100 nt) that already contain two gRNAs, which does not allow repurposing of oligonucleotides in alternative experimental designs. To target individual genomic loci, repurposing of gRNAs to generate multiple deletions of the same element, may prove a cost-effective experimental design to yield robust perturbation data, following a similar logic to the delivery of multiple siRNAs against a gene to demonstrate on-target rather than off-target effects 10 . Here, we provide an inexpensive single-step dual gRNA cloning protocol method, in which each oligonucleotide contains only one gRNA (Figure 1). As a result, oligonucleotides can be repurposed to clone different pairs of gRNAs by simply setting up a PCR using different forward and reverse primer combinations.
Similarly to previously developed protocols for dual gRNA cloning 8,9 , we deploy a strategy in which the expression of the two gRNAs is driven by different promoter sequences, avoiding potential plasmid recombination events due to sequence repeats in the final vector 9 . Specifically, this strategy yields a vector that contains the hU6 promoter driving the expression of one gRNA and the H1 promoter driving the expression of the other gRNA (Figure 1a). To this end, we generated a plasmid that can be used as PCR template for the cloning strategy outlined in Figure 1a. This has the advantage of providing an inexpensive source of PCR template, which can be propagated and easily shared between laboratories (pScaffold-H1 vector, Addgene #118152, see attached map in Supplementary Information). The pScaffold-H1 vector was generated by insertion of a PCRamplified fragment that contained a sgRNA scaffold followed by the H1 promoter (amplified from the pDECKO-GFP, Addgene #72619) into the pCR Blunt II-TOPO vector (Invitrogen), which contains a Kanamycin resistance cassette. The fact that the pScaffold-H1 vector does not contain an Ampicillin resistance cassette (present in most CRISPR/Cas9 sgRNA expression vectors) eliminates the formation of colonies containing it due to plasmid carryover and maximises the efficiency of the protocol.
The primer design has been developed to not require optimisation between reactions, since the primers will anneal to constant sequences in the template DNA plasmid: the forward primer anneals to a 22 nt sequence at the 5' end of the sgRNA scaffold; and the reverse primer anneals to a 17 nt sequence at the 3' end of the H1 promoter (see Figure 1 and refer to the Supplementary Tables "BbsI" and "BsmBI" for details and a design template).
The cloning strategy we present here has the potential to be applied with many common sgRNA expression vectors. We provide details for two common types of backbone for expression of the two gRNAs and SpCas9: plasmid (such as the plasmids developed by the Zhang lab, pX458 11 , pX459 11 , and our own SpCas9-Hygro 7 ), and lentiviral (such as the lentiCRISPR v2 12 ).
We have successfully applied this protocol to clone gRNA pairs for deletion of pancreatic beta cell transcriptional enhancers 7 . Moreover, this protocol is amenable to be deployed to clone gRNA pairs for targeting of other genomic sequences, including lncRNAs, miRNAs, promoters, CTCF sites and even coding sequences, if the desired outcome is a defined deletion. We also anticipate that the applications of this cloning strategy will be expanded to other CRISPR/Cas9-based perturbations such as CRISPR interference (CRISPRi) or CRISPR activation (CRISPRa) with two gRNAs against the same target gene/regulatory element to boost CRISPRi/a efficiency, and even to combined CRISPRa-CRISPRi to study complex gene-gene interactions 13

Procedure
We provide two alternatives for this step, depending on the backbone used for cloning: Version A allows cloning into CRISPR vectors with BbsI (BpiI) sites (such as pX458); and Version B allows cloning into CRISPR vectors with BsmBI (Esp3I) sites (such as lentiCRISPR v2).  Technical note: The cohesive ends generated by the BbsI/BpiI and the BsmbI/Esp3I enzymes are identical. Therefore, primers containing a different restriction site than the vector (e.g. BbsI primers and BsmBI vector) can be used to clone the pair of guide RNAs in this vector using an additional digestion step. This allowed the repurposing of a single pair of primers for both types of vector. The additional step is performed after step 2d. 500 ng of the purified PCR product is digested following the same procedure as step 3 with the restriction enzyme corresponding to the primer used, omitting the vector, ATP and T7 ligase and incubated 1h at 37ºC. The digested product is purified using QIAquick PCR Purification Kit and diluted at 10 ng/µL. The protocol can then be continued at step 3 using the protocol corresponding of the vector's restriction sites. Technical note 2: We strongly advice using an E. coli strain suitable for cloning unstable DNA constructs (e.g. RecA, RecA1, or RecA13 strain), such as One Shot Stbl3 Chemically Competent E. coli (Invitrogen) or NEB Stable (New England Biolabs).

Confirmation of successful cloning
The pScaffold-H1 vector used as PCR template contains a Kanamycin resistance cassette and the most commonly used CRISPR vectors contain a Ampicillin resistance cassette. Therefore, after transformation and spreading onto LB-agar plates with ampicillin, we do not detect colonies containing the pScaffold-H1 vector.
Using this protocol, we usually observe an efficiency of >95% for most backbones, for this reason, we do not perform colony PCR to screen for positive colonies ahead of sequencing. a) Pick one colony from the plate and inoculate 2 mL of LB Broth. b) Incubate overnight at 225 rpm in an orbital shaking incubator.
c) The next day, analyse by plasmid isolation with a miniprep kit, followed by Sanger sequencing using primer LKO or U6.

Anticipated Results
After PCR, a clear 353bp band should be visible by agarose gel migration (Figure 2). We advise running PCR products on a 1.5% agarose gel as an intermediate QC step when cloning many gRNA pairs.
After cloning into the desired backbone, Sanger sequencing with either one of the suggested primers (U6-F and LKO.1) should allow reading of the two gRNAs and the H1 promoter.
Considerations on genome editing efficiency: Deletion efficiency of the target sequence will depend on the individual efficiencies of the two gRNAs in each gRNA pair. We suggest testing at least four gRNA combinations, which can be achieved by designing two gRNAs on each side of the target region (since our protocol allows repurposing of gRNA in different gRNA pairs). In our experience, higher Cas9 levels yield higher deletion efficiencies. Therefore it is important to work with a system that yields high Cas9 expression in the targeted cell type.     Supplementary Files This is a list of supplementary files associated with this preprint. Click to download.