Shockwave-induced DNA-free genome editing in tobacco : targeting the 1 actin depolymerizing factor gene increases drought and salinity tolerance 2

150 words) 11 DNA-free genome editing involves the direct introduction of ribonucleoprotein (RNP) complexes into 12 cells, but this strategy has rarely been successful in plants. Here we describe a new technique for the 13 introduction of RNPs into plant cells involving the generation of cavitation bubbles using a pulsed laser. 14 The resulting shockwave achieves the efficient transfection of walled cells in tissue explants by the 15 creation of transient membrane pores. RNP-containing cells were rapidly identified by fluorescence 16 microscopy, followed by regeneration and the screening of mutant plants by high-resolution melt 17 analysis. We used this technique in tobacco to target the endogenous phytoene desaturase (pds) and 18 actin depolymerizing factor (adf) genes. Genome-edited plants were produced with an efficiency of 5.6– 19 8.7%. We also evaluated the effects of adf mutations in T2 mutant plants under drought and salinity 20 stress, showing that adf acts as a key regulator of osmotic stress tolerance in plants. 21


22
Genome editing in plants is widely used for the functional analysis of genes and the development of 23 improved crop varieties. The CRISPR/Cas9 system is the most popular genome-editing tool because it 24 is simple and adaptable, especially for the simultaneous mutation of multiple genes [1][2][3]. CRISPR/Cas9 25 has therefore been used to manipulate many traits of agricultural significance [4]. However, the major

34
We have previously shown that RNP-mediated genome-editing experiments targeting the tobacco pds 35 gene achieved a mutation efficiency of 19% when the RNP was delivered by particle bombardment, but 36 the absence of a selection marker resulted in a laborious and time-consuming screening process [7].

37
Furthermore, particle bombardment causes severe cell damage and the efficiency of regeneration is 38 2 therefore low [8][9][10]. PEG-mediated protoplast transfection is a gentler method for the delivery of 39 CRISPR/Cas9 components (as DNA, RNA or RNPs), but the efficient regeneration of many plant 40 species from protoplasts is not yet possible [11]. DNA-free genome editing therefore remains 41 challenging in species recalcitrant to transformation, such as wheat, sorghum and woody plants [12]. A 42 new method is required for the delivery of RNPs which is gentle and compatible with intact plant cells.

43
Shock waves induced by cavitation bubbles show great potential for the delivery of nucleic acids and 44 proteins because they create transient pores in the plasma membrane, thus increasing its permeability 45 [13,14]. Cavitation bubbles are usually induced by acoustic stimuli such as ultra-sonication, and the 46 resulting shock waves can promote the uptake of DNA by plant cells [15]. This non-invasive and 47 chemical-free method has been used to transfect individual plant cells, protoplasts, plant cell suspension 48 cultures, and intact tissue explants [16,17]. Although the generation and oscillation of gas bubbles in 49 liquid is usually achieved by acoustic cavitation [18,19], shock waves can also be produced by a pulsed 50 laser, an approach that has not yet been used for plants but was successful for the transfection of animal 51 cells [20]. Based on these findings, we adapted the pulsed laser method for the transfection of intact 52 plant cells with CRISPR/Cas9 RNPs.

53
The new pulsed laser technique has the potential to increase the efficiency of genome editing in a broad 54 range of plant species. The rapid selection of RNP-containing cells was achieved by incorporating a 55 labeled RNA that can be detected by fluorescence microscopy. This was followed by regeneration and 56 the screening of mutant plants by high-resolution melt analysis (HRMA). We targeted the endogenous 57 phytoene desaturase gene (pds) because mutations generate an easy detectable albino phenotype 58 allowing the calculation of mutation efficiency. We also targeted the actin depolymerizing factor (adf) 59 gene because mutations increase osmotic stress tolerance, allowing us to demonstrate the potential of

77
Having established a working procedure, we then used the pulsed laser-induced shockwave method for 78 the direct delivery of RNPs into the intact cells of tobacco leaf discs, which are much easier to prepare 79 and handle than protoplasts or zygotes. We introduced a preassembled RNP comprising the Cas9 80 protein, crispr RNA (crRNA) and ATTO-550-labeled trans-activating crispr RNA (tracrRNA) targeting 81 either the tobacco pds or adf genes. The fluorescent tracrRNA allowed the direct screening of 82 transfected cells so that a selectable marker gene was unnecessary ( Figure 2A′). The sample size and 83 experimental setup were the same as described above for the transfection with DsRed ( Figure 1A and 84 B). We screened for RNP-containing fluorescent cells 48 h after the laser pulse (Figure 2A and A′

104
We used the same HRMA method discussed above to screen 1011 plants from the adf experiment and 105 identified 57 homozygous/biallelic mutants and 110 heterozygous mutants. In the former group, Sanger 106 sequencing confirmed the positive HRMA results and resolved 48 homozygous mutants (e.g., lines A3, 107 A57, A61, B2 and C60) and 9 biallelic mutants (e.g., lines C66 and C72). The distinct melting curves of 108 adf homozygous mutant A57 and the wild-type control are compared as an example in Figure 3A Figure S4A). Furthermore, a standard test of cell membrane 116 thermostability in 10 randomly-selected adf T0 mutants and wild-type controls under normal irrigation 117 conditions revealed higher membrane thermostability among the mutants ( Figure 3I).

118
Stable inheritance of the adf mutation in generations T1 and T2

119
T0 adf homozygous mutant plants (48 events) were allowed to grow and set seed under controlled 120 conditions in the greenhouse, but germination and seed setting were inefficient. After selfing, 25% of 121 the T0 events did not set seeds, 31% set very few seeds, 16% set a few seeds but at a later time than   The relative water content (RWC) of the 10 T2 adf homozygous mutant plants lines was higher than that 155 of wild-type plants under drought stress. On day 0 before soil moisture stress was applied, the RWC of 156 the mutant and wild-type plants was similar at ~50%. After 5 days of drought, the RWC of wild-type 157 plants had fallen to ~15% but remained much higher in the mutant plants, ranging from ~40% in lines 158 A57, A67 and C60 to ~30% in line C6. Furthermore, the RWC of the mutant plants recovered quickly 159 after normal irrigation was reinstated whereas the wild-type plants still showed a low RWC of ~10% after 160 5 days of normal irrigation ( Figure 4E).    234 Consistent with the above study, we also observed higher membrane stability followed by actin

305
The target sites for genome editing were amplified by PCR using 50 ng genomic DNA, 1.5 mM dNTPs, 306 0.25 µM of each primer, and one unit of Q5 high fidelity DNA polymerase for each 20-µL reaction. For 307 the pds gene, the reaction was heated to 98°C for 30 s, followed by 30 cycles of 98°C for 10 s, 56°C for 308 30 s, and 72°C for 30 s, then a final extension step at 72°C for 7 min before cooling to 4°C. For the adf 309 gene, the reaction was heated to 98°C for 30 s, followed by 30 cycles of 98°C for 10 s, 53°C for 30 s, 310 and 72°C for 30 s, then a final extension step at 72°C for 7 min before cooling to 4°C. The PCR products 311 were purified from agarose gels using the NucleoSpin gel and PCR clean-up kit and sequenced using