Developing an Efficient DNA-free, Non-transgenic Genome Editing Methodology in Grapevine

Genome editing is a plant breeding innovation that allows rapid targeted modification of plant genome, like natural mutations, to improve traits such as yield and disease tolerance. Loss-of-function mutations in one or more of ‘Mildew resistance Locus O’ (MLO) genes impart protection to plants from powdery mildew fungi and confer durable broad-spectrum resistance. This mlo-based resistance, detected initially as a natural mutation in barley, has been successfully employed for nearly four decades. Recently, mlo-based resistance was found to be engineered or identified, from natural mutants, in many plant species of economic relevance. Genome editing in grapevine is currently performed through conventional Agrobacterium[1]mediated plasmid delivery, which integrates foreign genes into the genome and is labeled GMO plants. A more recent CRISPR RiboNucleoProtein (RNP) delivery in protoplasts is a DNA-free approach but makes it difficult to regenerate plantlets and identify the gene-edited mutants. This project aims to establish an alternative gene-editing approach for the grapevine model to generate powdery mildew-resistant mutant vines free of foreign genes. We proposed to edit the MLO susceptibility genes through a combination of plasmid- and RNP-delivered CRISPR/Cas9 in microvine to produce DNA-free powdery mildew-resistant plants. The project is currently in the 3rd quarter of the final year, and the first phase of generating mlo mutants through conventional Agrobacterium-mediated gene editing has been concluded. MLOs 3, 4, 13, and 17 in grapevine have been identified as the closest orthologs of MLO genes responsible for powdery mildew resistance in other species. We obtained single mutants of candidate MLOs and the combinations of double and quadruple mlo mutants. The mutant combinations will help identify the appropriate MLO genes for mildew resistance in grapevine. Currently, the genome-edited mlo mutant embryos are at various stages of plantlet regeneration. Preliminary genotyping analyses of the transformants showed mutations in the predicted areas of the MLO genes targeted by the CRISPR system. So far, we have identified over 200 potential mlo mutants. The next phase of the project is the removal of the transgene cassette, integrated into the genome of mlo mutants by employing CRISPR RNP delivery to obtain transgene-free mlo vines. Towards this goal, we concluded the experimental work. We established 1) the methodology to purify the modified Cas9 protein, 2) the RNP complexations and in vitro cleavage tests, 3) the excision of the transgene cassette using in vitro RNPs, 4) the use Cell-Penetrating Peptides (CPP) to facilitate the entry of the CRISPR RNP into regenerable embryogenic cells, and 5) the optimization of the CPP-RNP delivery conditions. The editing of the target genes through the CPP-conjugated RNP delivery into the embryogenic cells was also tested on GFP-expressing embryogenic cell lines. We expect the regenerating mlo-edited plantlets to be ready for genotyping and powdery mildew resistance assays in 3 months, after which the second phase of the project (in vivo excision of the transgenic cassette) will begin to obtain transgene-free powdery mildew resistant vines.

Evaluating Grape Rootstocks for Nematode Susceptibility

In California grape production is one of the perennial high-value commodities. Because of the traditional long time period a vineyard is cultivated, soil-borne pests and diseases have ample time and opportunity to build-up and be major yield constraints. Once introduced in vineyard soil, these soil-dwelling parasites can reproduce on susceptible plant roots copiously. Typically multiple species including Meloidogyne spp. (root-knot nematodes), Mescocriconema xenoplax (ring nematode), Xiphinema americanum or X. index (dagger nematode), and Tylenchulus semipenetrans (citrus nematode) and others occur in various combinations in vineyard soil and damage the crop. Once they are established in an existing vineyard they are difficult to manage because they colonize at least the upper 5-ft soil layers. Genetic resistance in rootstocks can offer protection from these culprits, and offer sustainable and efficient protection of the risk of nematode damage. The frequent co-occurrence of multiple plant-parasitic nematode species makes attaining highest levels of resistance against such nematode assemblies tedious. Typically, the resistance towards one species has limited to no effect on a different species. In the current project, selections of grape rootstock genotypes selected foremost under greenhouse conditions were exposed to field populations of nematode assemblies. In 2019, one experimental vineyard was planted following a >30 year old planting of Thompson seedless grape. A total of seventeen experimental rootstocks developed in Dr. Andrew Walkers program were planted alongside repeat entries from prior experimentation and longer-tested GRN1, GRN2, GRN4 (A. Walker) and 1017A. In addition, Zinfandel, St. George, Harmony, Flame, and Salt Creek were planted as commercial controls. In mid-June 2019, rootstocks were planted in randomized complete block design with five replications into this nematode-infested site. A month after planting, each planting site receive additional root-knot nematode-infested soil to boost the infestation levels. In 2020, another nine experimental rootstocks with similar controls were planted, and additionally inoculated with infested soil. First nematode assessments were done in the dormant season. Field testing these rootstock candidates will provide important information on the nematological and horticultural status of them. These activities will allow to identify elites that warrant further characterization, and that could add new so urgently needed rootstocks for the industry.


Developing an Efficient DNA-Free, Non-Transgenic Genome Editing Methodology in Grapevine

Genome editing technology is a plant breeding innovation that allows rapid targeted modification of plant genome, similar to natural mutations, to improve crop traits such as yield and disease tolerance. Objective of this project is to apply CRISPR/Cas9 gene-editing to produce genome-edited non-transgenic (devoid of foreign genes) grapevine. We proposed to edit the susceptibility genes for powdery mildew (PM) in grapevine named ‘mildew locus O’ genes (MLO) through a combination of plasmid- and ribonucleoprotein (RNP)-delivered CRISPR/Cas9 into the intact embryonic cells to produce DNA-free PM-resistant grapevines. We adopted this two-phase gene-editing approach rather than RNP-delivery alone to overcome the large-scale screening logistics to identify mutated plants in the absence of selection marker genes.

* At the end of the 2nd quarter of Year 2, the conventional Agrobacterium-mediated transformation of microvine to generate mlo mutants has been concluded. Single, double and quadruple mutants of MLOs 3, 4, 13, and 17, involved in grapevine PM-susceptibility, are expected from the first phase gene-editing experiments. The CRISPR-Cas9 plasmid used for MLO gene editing was prepared with an alternative strategy which is explained in the following main text of the report. Gene edited mlo mutants are expected in 3-4 months after which characterization and identification of the specific MLO genes involved in grapevine PM-resistance and assessment of the level of resistance will be performed. As part of the second phase, removal of the transgene T-DNA cassette, integrated into the plant genome, by employing Cas9 RNP delivery has been proposed to generate transgene-free mlo mutant microvines. For that purpose, we plan to use Cell Penetrating Peptides (CPP) fused to Cas9 to facilitate the entry of the RNP into the embryogenic cells. Experiments to optimize the RNP delivery using CPP have continued during the 2nd year. After confirming the efficacy of the candidate CPP for cell penetration and internalization using fluorescent dye, we proceeded to test the delivery of CPP-conjugated Cas9 RNPsinto embryogenic cells. For these trials, GFP-expressing embryogenic callus has been used and the gene-editing rate of the GFP target gene has been assessed. Lower than expected editing rate was observed in these preliminary experiments and experiments to improve the editing rate and to optimize the methods to regenerate gene-edited embryos have been continuing in the year 2. When the MLO gene-edited microvines from first phase are ready, the RNP-delivery methods that are being optimized during the 2nd year will be employed to get the final transgene-free mlo mutant microvines.

Evaluating Traits To Improve Grapevine Water-Use Efficiency and Drought Tolerance

The goal of this multi-year project is to identify leaf and root traits that improve drought tolerance and water-use efficiency (WUE) in grape cultivars and rootstocks. Growers will need more drought tolerant and water-efficient vines to avoid declines in berry yield and quality under the hotter, drier conditions expected for California’s wine regions. This project has focused on leaf and root values for the drought tolerance traits the turgor loss point, measured in waterstressed conditions (TLPdry), and the adjustment in turgor loss point between well-watered and water-stressed conditions (DTLP, or TLPwet – TLPdry). TLP measures the water potential threshold that causes cell collapse, which impedes water transport and growth in leaves and roots. A more negative TLP indicates the leaves and roots can undergo more severe water stress before losing turgor. Plants can also change their cellular biochemistry to make TLP more negative under water stress, with a greater adjustment indicated as a larger DTLP. Previous studies found an important role for these traits in drought tolerance in other crops, but these traits have never been assessed for their impact on grapevine performance under water stress.

Our first year, we conducted a greenhouse experiment to evaluate whether rootstocks with a more negative root TLPdry and a larger DTLP allowed vines to maintain greater gas exchange, growth, and WUE under water stress. This year, we have reanalyzed these data and found significant correlations between a more negative TLPdry and larger root system, and between greater vine gas exchange under drought and lower values of a second trait, the root capacitance (CAPdry). CAP measures the volume of water lost from the root as the root dehydrates to TLP. A lower CAP indicates the root loses less volume, with potentially important effects on belowground water transport and root-to-shoot chemical signaling. However, TLPdry was significantly less negative in the rootstocks that field trials have classified as drought tolerant, suggesting that a less negative TLP in dry roots could benefit drought tolerance by redirecting resources to increase root growth in deeper, wetter soil. Together, these findings suggest that root TLP and CAP are promising traits to improve rootstock drought tolerance, though more work is needed to clarify their function.

Our second year, we used a vineyard experiment with 7 wine grape cultivars to test the impacts of leaf TLP and DTLP, and the biochemistry underlying variation in these traits, on scion drought tolerance, including gas exchange, vigor, and berry yield and quality, in realistic vineyard conditions. All 7 cultivars significantly adjusted TLP to improve leaf drought tolerance over the growing season, but cultivar rankings in TLP were highly consistent. However, contrary to expectation, leaf TLP was not related to gas exchange or plant water stress in this study, suggesting leaf biochemistry was responding to other stresses in these vines. Further, this study is the first to report a significant accumulation in mannitol as part of DTLP in wine grapes, which has been shown to ameliorate heat and pathogen stress in other plant species.

Breeding, Genetics, and Germplasm Evaluation

The overall objective of this research project is to develop and use a genome editing technology for trait improvement of elite grape cultivars. In our 2018-2019 work, we successfully demonstrated the feasibility of editing a grape color gene, VvMybA1, by delivering a VvMybA1editing construct, p201H-MybA1Double1(1176/1180), into V. vinifera ‘Chardonnay’ embryogenic callus through stable Agrobacterium transformation. Subsequently in this project year, we focused on editing the grape color gene using a non-transgenic approach. Our non-transgenic approach is based on the facts that genome editing components (i.e. Cas9 and gRNA) on the T-DNA of a construct in Agrobacterium can transiently be expressed in target cells before degraded in or stably integrated into the host cell genome and the transiently expressed gene editing components, when taking place at a right time with appropriate strength, can result in successful editing of a gene in the host cell. By using GUS as a reporter gene, we observed that the transient expression of GUS in ‘Chardonnay’ embryogenic callus reached the peak after 2-3 days of co-incubation with Agrobacterium carrying the GUS construct. Based on the GUS observations, we co-incubated ‘Chardonnay’ embryogenic callus with Agrobacterium carrying the same construct p201H-MybA1Double1(1176/1180) we used in the 2018-2019 stable transformation for editing the VvMyb1A color gene. The co-incubation experiments were conducted for 2 and 3 days, respectively. High throughput sequencing revealed that editing took place in both 2- and 3-day incubations. Various types of editing were found, including deletions, insertions, and substitutions. Among the deletion events, single bp deletion was most frequent, as expected. Deletions involving multiple bps (2-15 bps) were also observed. Many of these events were likely resulted from transient expression of the CRISPR-Cas9 and VvMyb1A gRNAs. While the overall editing frequencies were very low (<0.015%), we demonstrated the success of editing a grape gene through transient Agrobacterium transformation. Now the key issue for grape gene editing is not so much about whether we can edit a grape gene in a non-transgenic manner, but about how to select the edited events from millions of non-edited cells. Solving this issue will be the focus of the proposed 2020-2021 project.

As a part of this continuing project, we induced some transgenic vines from the embryogenic callus transformed with the p201H-MybA1Double1(1176/1180) construct in the previous project year 2018-2019. Unfortunately, the induced vines showed extremely low vigor in tissue boxes, likely due to some undesirable clonal variation in the batch of ‘Chardonnay’ callus used. New ‘Chardonnay’ callus transformed with the same construct was produced in this project year and transgenic vines from the callus will be induced. The purpose of generating stable transgenic vines from the experiment is to demonstrate the phenotypic effect after the transposon Gret1 is removed from the VvMybA1 gene in ‘Chardonnay’.

Development of next generation rootstocks for California vineyards

Since my last report (June 2019) Nina Romero has made excellent improvements to
our rootstock screening and is currently re-vamping our ring nematode resistance screening, and she has replaced three technicians who departed over the past year. There are 444 genotypes in testing for resistance to nematodes, salt or both. Our 2019 crosses again focused on using fertile and tetraploid VR hybrids to get rotundifolia forms of resistance into better rooting backgrounds, and mostly failed (due to the genetic distance between Vitis grape species and rotundifolia. We have been successful with crosses to two VR hybrids both of which resist phylloxera and combined with rootstocks. Seed from the 2018 crosses are mostly in storage for the next grape breeder, except for the 18113 (GRZN3 x V. acerifolia 9018. Chris Chen is working on his PhD with this population which brings excellent and broad nematode
resistance to our best form of salt tolerance (which also has strong root-knot resistance). We have improved our phylloxera screening in the greenhouse and have verified a number of fertile VR hybrids also have strong phylloxera resistance. A new post-doc (Erin Galarneau from the Baumgartner lab) was hired to direct examinations of phenolic compounds responsible for phylloxera and nematode resistance. They will also assist our efforts to determine how O39-16 induces fanleaf degeneration resistance. We are also making good progress on identifying the basis of red leaf virus tolerance. These efforts are being directed by a visiting scholar from China. We have rootstock examples of strong tolerance (St. George and AXR1) and very sensitive (Freedom and 101-14) and rapid tissue-culture and greenhouse-based
screens that are rapid and mimic field tests.

Evaluation of Grapevine Rootstock Selections

The purpose of this project is to identify selections from a USDA rootstock breeding program that might warrant release as commercial stocks, and to develop useful data on the performance of recently released rootstocks from other breeding programs to aid growers in selecting appropriate stocks for their vineyards. The initial plantings from the USDA rootstock breeding program number over 700 selections. This initial group was grown until maturity, and then evaluated for their ability to be good mother vines. This evaluation identified nearly 150 selections that were good to moderate mother vines. Selections with good mother vines qualities were tested by Dr. Andreas Westphal and Dr. Andrew Walker for resistance to aggressive root-knot nematodes. Selections that performed well in both assessments have been grafted to scions for in field evaluations. Currently six selections (PC0333-5, PC0349-11, PC0349-30, PC04153-4, PC0495-51, and PC0597-13) have been planted in a replicated trial in a commercial winegrape vineyard in Merced. These vines should start production during the 2020 growing season. Five selections had been grafted to a table grape and planted in a replicated trial in a commercial table grape vineyard in Delano, CA. Unfortunately, this trial has run into problems and only limited data will be collected. For the past several years Dr. Gan-Yuan Zhong, USDA-ARS, screened
seedling selections for resistance to aggressive strains of root knot nematodes and shipped cuttings of resistant selections to the UC Kearney Agricultural Center in Parlier, CA, where they were rooted and planted into a vineyard for observation. Since 2016 a total of 165 new selections have been planted. These vines will need to undergo the same testing as the previous rootstocks once they are mature. It is hoped that a new table grape vineyard can be established with these advanced selections. The Merced trial should resolve questions about the potential, if any, of these selections as young grafted vines. The Merced trial is adjacent to another rootstock trial planted by former UCCE advisor Lindsay Jordan in 2016. The second Merced trial includes full rows of 1103P, and more recently released stocks including RS3, RS9, GRN2, GRN3, GRN4, and GRN5 grafted to Malbec, and replicated four times. However, most vines on GRN5 failed, so GRN5 was eliminated from the trial. Vine training was completed during the 2019 growing season.

Developing an efficient DNA-free, non-transgenic Genome editing methodology in grapevine

Genome editing technology is a plant breeding innovation that allows rapid targeted-modification of plant genome, similar to natural mutations, to improve crop traits such as yield and disease tolerance. Objective of this project is to apply CRISPR/Cas9 gene-editing to produce genome-edited non-transgenic (devoid of foreign genes) grapevine. We proposed to edit the susceptibility genes for powdery mildew (PM) in grapevine named ‘mildew locus O’ genes (MLO) through a combination of plasmid- and ribonucleoprotein (RNP)-delivered CRISPR/Cas9 into the intact embryonic cells to produce DNA-free PM-resistant grapevines. We adopted this two-phase gene-editing approach rather than RNP-delivery alone to overcome the large-scale screening logistics to identify mutated plants in the absence of bacterial selection markers.

Since the start of project in July 2019, the CRISPR plasmid to be used in the first phase of MLO gene editing is being built. To deliver CRISPR-Cas9 RNP in the second phase, we planned to use cell penetrating peptides (CPP) fused to Cas9 to facilitate Cas9 entry into the embryogenic cells. Experiments to optimize the methods for RNP delivery using CPP and to identify the optimum stage of embryogenic callus, efficiency of the CPP to penetrate the cells, and the treatment methods have been undertaken. The CPP has been synthesized and the efficiency of cell penetration and internalization were assessed using fluorescent dye attached to the CPP. Based on the detection of the dye in majority of the cells in the treated callus we confirmed that Cas9 protein can be delivered into embryogenic cells using this CPP. Methods for the bacterial expression and purification Cas9 protein and its conjugation with CPP have been optimized. Cleavage activity of the CPP-conjugated Cas9 protein is assessed through in vitro cleavage experiments using the guide RNA targeting GFP gene. In the optimization process, GFP-expressing embryogenic callus has been used to treat with the RNP in order to assess the efficiency and accuracy of RNP-mediated gene-editing. Data from these experiments will be used to streamline the process of MLO gene editing in microvine.

Determining the role of Auxin-Response Factor 4 in the timing of ripening initiation in Vitis vinifera

The timing of ripening initiation is a major trait for wine grape production. Rapidly evolving climactic conditions will affect the ripening process of major cultivars in several growing regions of U.S, which may result in loss of fruit and wine quality. While the development of improved viticulture practices to mitigate effects of climate change is critical, the use of complementary approaches must be investigated. The identification of regulatory genes controlling the timing of ripening initiation is a research avenue to consider for molecular breeding programs in order to identify clones/cultivars more suitable to evolving climatic conditions. Innovative molecular practices in the field such as Spray Induced Gene Silencing have recently received scientific attention and its potential for rapid industrial application can be seen as alternate solution to traditional and molecular breeding. Yet, all these translational tools need scientific validation up front to characterize the cause-to-effect relationship between the gene and/or several genes and the trait of interest (fruit composition, disease resistance, etc.). The current research project aims to validate the contribution of a regulatory protein, VitviARF4, to the ripening initiation in grape berry. Three objectives were designed to achieve this goal; 1) the characterization of VitviARF4 via genetic engineering and the identification of its potential partners during the ripening process, 2) the identification of ripening-related genes that are targeted by VitviARF4, and 3) the evaluation of fruit composition on genetically engineered berries.

After the establishment of the microvine model at OSU to conduct the genetic engineering experiments, our research efforts were focused this year on several milestones of the project. For the objective 1, we conducted all the microvine transformations (four in total, control, two over-expression, and one knock – down) to either turn on or off the expression of VitviARF4. At least 20 independent transformed plants were selected per transformation event. We also confirmed protein-protein interactions of VitviARF4 with other proteins that play a major role in various physiological process of fruit ripening (sugar, brassinosteroids, ethylene, and epigenetic mechanism). For the objective 2, we demonstrated that we can control the expression of the transgene in transgenic microvines during the development of the plants. This result was critical to the success of the research outcomes of the objective 2. We have generated about 15 to 20 GFP positive transformed lines per construct and few of them were transferred to the greenhouse for being characterized. In parallel, we also conducted a Spray Induced Gene Silencing (SIGS) experiment on pre-véraison berries of V6 microvine. Our preliminary findings seem to support the delaying effect of VitviARF4 on the timing of berry ripening. We observed a faster ripening on berries treated with dsRNA aiming at silencing the expression of the endogenous VitviARF4. However, this experiment is currently repeated to confirm this first results. If confirmed, it might lead to a potential direct application of the SIGS technology in the field to manipulate trait. For the objective 3, we developed a
new analytical method to measure organic, amino, and phenolic acids, different types of carbohydrates, polyols, and three classes of flavonoids (anthocyanins, flavonols, and monomer and dimer of tannins). We have built an in-house library of 95 analytes that were tested against berry extracts from pericarp samples collected at different stages of grape berry ripening. We are currently testing the method on mature berries of the microvine and we have identified around 30 analytes covering the major families of compounds existing in grape berry. Dr. Tomasino has optimized the volatile and aroma analyses on mature fruits of regular microvines.

Integrating systems biology with marker assisted selection to guide the stacking of powdery mildew resistance genes

The long-term goal of this project is to develop grape varieties that possess effective and durable resistance to powdery mildew (PM). Stacking resistance genes from multiple resistant genetic backgrounds and with the least functional redundancy is a proven breeding strategy to improve both durability and level of resistance. This strategy requires (a) the identification of multiple sources of resistance, (b) the functional characterization of the mechanisms of resistance to prioritize optimal genetic combination, and, finally, (c) marker assisted breeding to introduce the selected genes into elite varieties.

In continuation of our multi-year breeding effort, with the awarded budget in the 2017-2018 funding period we have continued the functional characterization of resistance responses activated in presence of known PM resistance loci. Analysis of the data generated by the experiments for the functional characterization of Ren2, Ren3, Ren4, Ren6, Ren7, Run1, Run1.1, Run2.1 and Run2.2 showed differences of gene expression between accessions in response to PM. Deep-learning analysis to predict the best R gene combination for gene stacking by marker assisted breeding is