Rothamsted Repository Download

Background CRISPR/Cas has recently become a widely used genome editing tool in various organisms, including plants. Applying CRISPR/Cas often requires delivering multiple expression units into plant and hence there is a need for a quick and easy cloning procedure. The modular cloning (MoClo), based on the Golden Gate (GG) method, has enabled development of cloning systems with standardised genetic parts, e.g. promoters, coding sequences or terminators, that can be easily interchanged and assembled into expression units, which in their own turn can be further assembled into higher order multigene constructs. Here we present an expanded cloning toolkit that contains ninety-nine modules encoding a variety of CRISPR/Cas-based nucleases and their corresponding guide RNA backbones. Among other components, the toolkit includes a number of promoters that allow expression of CRISPR/Cas nucleases (or any other coding sequences) and their guide RNAs in monocots and dicots. As part of the toolkit, we present a set of modules that enable quick and facile assembly of tRNA-sgRNA polycistronic units without a PCR step involved. We also demonstrate that our tRNA-sgRNA system is functional in wheat protoplasts. We believe the presented CRISPR/Cas toolkit is a great resource that will contribute towards wider of the CRISPR/Cas genome editing modular cloning by


2
The CRISPR/Cas technology has recently become an easily accessible genome editing tool for many organisms, including plants [1]. Generating gene knockouts has become a rather straightforward CRISPR/Cas application in many plant systems [2][3][4], while more sophisticated applications, such as allele replacements or targeted gene insertions, still remain a challenge due to low efficiency of homology-directed repair (HDR) in plants [5].
In its conventional form, the CRISPR/Cas system includes a DNA nuclease, such as Cas9, which is guided to a specific genomic location by the guide RNA. Therefore, in order to perform targeted mutagenesis in planta, one needs to co-express both the CRISPR/Cas nuclease and its cognate guide RNA. Usually, the gene encoding the CRISPR/Cas nuclease is expressed using an RNA polymerase II (Pol II) promoter (e.g. 35Sp), while the guide RNA is expressed under an RNA polymerase III (Pol III) promoter (e.g. U6p or U3p), which has a defined transcription start nucleotide ('G' for U6p or 'A' for U3p). One of the advantages of CRISPR/Cas is multiplexing i.e. one can target DNA at multiple genomic locations by co-expressing multiple guide RNAs specific to those loci. Guide RNAs can be expressed either as individual transcriptional units, each under its own Pol III promoter [4], or as a tRNA-sgRNA polycistronic transcript [6]. In the latter case, guide RNAs are interspaced with tRNAs in a single transcript that gets processed into individual guide RNAs by the highly conserved tRNA processing machinery inside the plant cell [6].
As genome editing applications in plants often rely on delivering multiple expression units into plant cells, including a selectable marker, a CRISPR/Cas nuclease-encoding gene and one or more guide RNAs, it is important to be able to assemble DNA constructs encoding such expression units easily and rapidly. The modular cloning (MoClo) system based on the Golden Gate (GG) cloning method [7] is highly flexible and versatile, and provides a means for quick and facile assembly of multi-expression unit constructs using standard genetic parts, such as promoters, terminators, coding sequences etc. The system has already been successfully used for genome editing applications in plants [3,4,[8][9][10] but lacks modules encoding many of the newest genome editing reagents. Here we report on an expanded GG cloning toolkit for genome editing applications in monocot and dicot plants. We believe the toolkit will become a valuable addition to already existing GG-based tools for plant genome editing and be widely used by plant researchers across the community.
Base editors are a rather recent addition to the range of available genome editing tools and allow targeted conversion of DNA base pairs as following: C-G to T-A [23] and A-T to G-C [24] without introducing a double-strand break (DSB). The former base editor is based on the cytidine deaminase while the latter -on the adenosine deaminase. Both base editors have now been shown to be functional in various plants, including wheat, rice and tomato [25][26][27][28][29]. We have therefore generated level 0 modules encoding cytidine deaminase (pFH55 and pFH79) and adenosine deaminase (pFH45 and pFH92) based base editors (Additional file 2: Table S5).
EvolvR CRISPR-guided error-prone DNA polymerases have recently been shown to be able to introduce random point mutations at a targeted genomic locus [30]. Based on the Halperin et al.  Table S5) that could prove to be a useful tool for reverse genetics in monocot plants. 4 We used the above mentioned CRISPR/Cas nuclease level 0 modules to assemble twentythree nuclease expression units inserted into level 1 GG vectors to be applied in monocot and dicot plant species (Additional file 2: Table S5).

Guide RNA modules
As CRISPR/Cas is an RNA-guided nuclease, guide RNA is its essential component that must be co-expressed with the nuclease in order to achieve on-target DNA cutting. Guide RNAs are usually expressed under Pol III promoters, such as U3p or U6p, that have a defined transcription start nucleotide ('A' and 'G', respectively). A number of genomic loci can be targeted simultaneously by CRISPR/Cas by co-expressing multiple guide RNAs and the modular cloning system is highly suitable for assembling constructs carrying multiple expression units, such as the CRISPR/Cas nuclease and guide RNAs.
As part of this study, we have generated a number of level 0 Pol III promoter modules (TaU3p, OsU3p, OsU6-2p and AtU6-26p; Additional file 2: Table S5). In addition, we have produced a number of guide RNA backbone level 0 constructs that can be used to assemble single or multiple guide RNA expression units without a PCR amplification step involved (Additional file 2: Table S5). The cloning system we present allows guide RNAs to be expressed either under Up to four guide RNAs under individual Pol III promoters can be assembled in using the former cloning procedure (Fig. 1b) and up to six sgRNAs per polycistronic construct -using the latter one ( Fig. 4b). It must be noted that the number of guide RNAs under individual promoters could be increased to five, if no selectable marker is needed (Fig. 1b), or many more if level M/ level P vectors are used [11,31]. As to tRNA-sgRNA polycistronic constructs, the total number of sgRNAs assembled into a level 2 destination vector could be up to twenty-four (six per level 1), with a selectable marker, and up to thirty, if no selectable marker is used. Again, it is possible to add more than thirty sgRNAs by using level M/ level P vectors [11,31]. Our GG toolkit enables the user to build such complex constructs within a week (Additional file 1: Figure S1).

Testing of the tRNA-sgRNA CRISPR/Cas constructs in wheat protoplasts
All rights reserved. No reuse allowed without permission.

5
The GG toolkit therefore allows rapid parallel assembly of constructs by streamlining the cloning process. Since building multiple CRISPR constructs using GG is a straightforward procedure, it becomes reasonable to compare the activity of several experimental CRISPR setups in a transient expression system, such as protoplast, before proceeding with stable plant transformation, which could be a highly laborious and time consuming process. Our CRISPR toolkit includes three wheat codon optimised SpCas9 versions (level 0 constructs pFH13, pFH24 and pFH25; Additional file 2: Table S5) and their respective level 1 transcription units (pFH23, pFH66 and pFH67; Additional file 2: Table S5). These SpCas9 variants differ by e.g.
nuclear localisation signal (NLS) versions or affinity tags. We have therefore compared the activity of the three Cas9 variants in wheat protoplasts by cotransforming each of the level 1 constructs (pFH23, pFH66 and pFH67) with the level 1 plasmid containing the six sgRNAs ( Fig.   6a) assembled into a tRNA-sgRNA array. This has allowed us to target three different wheat genes at once (Fig. 6a). We have targeted each gene by at least two sgRNAs with large deletions between Cas9 cut sites expected to be detectable by PCR due to DNA band shifts as previously described [4]. PCR amplification of the target genes has revealed clear additional bands corresponding to alleles carrying large CRISPR/Cas-induced deletions in protoplasts transformed with pFH66. In contrast, application of the other two Cas9 versions (pFH23 and pFH67) resulted in very faint bands of the size corresponding to amplicons carrying the deletions (Fig. 6b). Our results therefore suggest a significantly higher activity of the pFH66encoded SpCas9, as compared to the other two Cas9 variants, in wheat protoplasts.

Discussion
The modular cloning kit presented in the study enables quick and facile assembly of DNA constructs for genome editing applications in plants and is an addition to previously published collections of compatible GG modules [7,[9][10][11]. The kit includes modules encoding a number of CRISPR/Cas nucleases (SaCas9, StCas9, LbCas12a etc.) that could be used as an alternative to the most commonly utilised SpCas9. SaCas9, for instance, has proven to be an efficient tool for generating gene knockouts in a number of plant species [13][14][15][16] and, in addition, has been shown to increase HDR efficiencies in plants [15]. Due to SaCas9 and StCas9 having longer than SpCas9 PAM motifs they are also likely to be more specific when it comes to DNA target recognition.
Cas12a (Cpf1) generates a staggered cut in DNA [32], while Cas9 -a blunt cut [12]. Due to this reason, Cas12a (Cpf1) could be a preferred choice of a CRISPR/Cas nuclease when it comes to HDR-based genome editing applications, such as targeted gene insertion [33,34]. It is noteworthy that the modular cloning system is highly suitable for HDR-based applications as the All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint 6 DNA repair template could easily be cloned as a level 1 module into a level 2 destination vector.
As part of our study, we have generated a number of guide RNA backbone level 0 modules, which are compatible with respective Pol III promoters and CRISPR/Cas nucleases (Additional file 1: Table S1). The guide RNA backbone modules can be used for PCR-free assembly of guide RNA expression units by cloning in an annealed pair of complimentary oligos encoding the guide sequence (Figs. 2 and 3). These guide RNA backbones are to be used when one wishes to express guide RNAs under individual promoters and up to four guide RNAs can be assembled into a level 2 vector together with level 1 modules encoding a CRISPR/Cas nuclease and a selectable marker (e.g. BAR, NPTII etc.; Fig. 1, Additional file 2: Table S5 and Additional file 3: Table S6).
In addition to expressing each guide RNA under its own promoter, we have generated modules that allow assembly of polycistronic tRNA-sgRNA constructs with up to six sgRNAs expressed using a single Pol III promoter (Fig. 4). The tRNA-sgRNA system, originally described by Xie et al. (2015) in rice, was later successfully applied in wheat [35] and Arabidopsis [36].
Nevertheless, the previously reported tRNA-sgRNA system relies on a rather cumbersome DNA construct assembly process as it involves PCR amplification of DNA fragments carrying repeats. The level 0 GG modules we have generated ( Fig. 5 and Additional file 1: Table S3) enable straightforward and efficient assembly of tRNA-sgRNA arrays without a PCR step involved. The system offers a choice of five monocot and dicot Pol III promoters, with TaU6p being a published module [37], and two different SpCas9 sgRNA backbones (classic and improved; Additional file 1: Table S3).
As stable transformation continues to be a major bottleneck for genome editing applications in many plants, including a major crop like wheat [38], it is advantageous to be able to test CRISPR/Cas constructs for activity in a transient expression system, such as protoplasts, before initiating an often lengthy and labour intensive stable transformation procedure. Using the wheat protoplast system, we have compared activity of three different wheat codon optimised SpCas9 variants (Fig. 6), which mostly differ in their C-terminal NLSs. The fact that one of the constructs (pFH66) performed better than the other two (Fig. 6b) could be due to the histone H2B NLS, located at the C-terminus of the pFH66 variant, being more efficient at importing Cas9 into the nucleus as compared to the SV40 or nucleoplasmin NLS present in the other two constructs. A possible link between different NLS versions and Cas9 activity was previously reported in Arabidopsis [9]. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint 7 The tRNA-gRNA system for Cas9 multiplexing has proven to work in monocots [6,35] as well as in dicots [36]. In dicots, it was shown that fusing tRNAs with the optimised sgRNA backbone [39] increased editing efficiencies [36]. In monocots however, only the classic sgRNA backbone [40] was so far used in tRNA-sgRNA arrays. The wheat protoplast assay has allowed us to verify the functionality of the tRNA-sgRNA array, carrying the optimised sgRNA backbone, in a monocot species. Our results suggest that the tRNA-sgRNA array, assembled using the optimised sgRNA backbone, could also result in higher CRISPR/Cas efficiencies in stably transformed monocot plants, in particular, wheat.

Conclusions
We believe the presented modular cloning kit will become a valuable addition to the range of already available GG modules [7,9,10] and expect that plant researchers, working with both monocots and dicots, will find the presented molecular tools useful for various genome editing applications. We also believe our study will contribute towards wider adoption of the GG modular cloning system by plant researchers and consequently facilitate exchange of standardised molecular cloning parts across the research community.

DNA construct assembly
All PCR amplifications were performed using Q5® DNA Polymerase (New England Biolabs) according to the manufacturer's instruction. All GG cut-ligation reactions were performed according to the described protocol (Additional file 1: section 1).
All ligations were transformed into One Shot™ TOP10 chemically competent E. coli (Thermo Fisher Scientific) and constructs were verified by sequencing (Eurofins Genomics).
Specific details related to assembly of all GG modules reported in this study are provided (Additional file 1: section 4). Sequences of all PCR primers used in the study are provided (Additional file 1: Table S4). All DNA constructs generated as part of this study were deposited with Addgene (www.addgene.org) with Addgene IDs indicated for each construct (Additional file 2: Table S5). Sequence information of all constructs can be found in GenBank (.gb) files (Additional files 4-105).

Protoplast assay
All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint Protoplasts were isolated from 10 day old, etiolated wheat seedlings (cv. Cadenza) as previously described [41] with some modifications. Cellulase R10 and Macerozyme R10 were   The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint 1 0 inserted into a level 0 acceptor with the sgRNA backbone using BpiI. During the second step (b), sgRNA is fused with the respective Pol III promoter using BsaI.   Assembly of less than six tRNA-sgRNA modules into a level 1 vector requires respective endlinkers (c). pFH and pAK constructs shown in black font carry the improved sgRNA backbone [39], while the ones shown in blue font -the classic sgRNA backbone [40]. *This is a published module [37]. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint Fig. 2

A B
The gRNA module can be assembled with a Pol III promoter in a level 1 acceptor via BsaI

Target specific annealed primers can be integrated into a level 0 acceptor vectors via BpiI
Level 0 module Level 1 module All rights reserved. No reuse allowed without permission.

A B
The crRNA module can be assembled with a Pol III promoter in a level 1 vector via BsaI

Target specific annealed primers can be integrated into level 0 acceptor vectors via BpiI
Level 0 module Level 1 module All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint Fig. 4 A B Level 2 construct Level 2 construct All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint Fig. 5 A B Up to six sgRNA modules can be assembled into a level 1 acceptor vector via BsaI

Target specific annealed primers can be integrated into a level 0 acceptor vectors via BpiI
Level 0 module C Assembly of 1 to 5 tRNA-sgRNA modules into a level 1 vector requires endlinkers All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/738021 doi: bioRxiv preprint A B Fig. 6 Target gene 1 Target gene 2 Target gene 3 All rights reserved. No reuse allowed without permission.