Assessing Fungicide Resistance of Grape Powdery Mildew in Wine, Table and Raisin Grapes

Throughout 2017, a field scouting campaign that covered Central and Northern California, Western Oregon, and Southern Washington yielded over 850 field samples and 64 isolates of Erysiphe necator. Analysis of these samples for resistance to quinone outside inhibitor (QoI) fungicides (FRAC Group 11) using the G143A qPCR assay indicated widespread resistance throughout all grape growing regions scouted (90% resistance among ToughSpot kit samples). These results were confirmed when isolate and field sample DNA underwent genotyping by sequencing analysis of the cytb gene. These results are similar to the QoI resistance observed throughout Oregon in 2015 and 2016. Analysis of these samples using various molecular techniques and fungicide resistance bioassays to determine resistance to demethylation inhibitor (DMI) fungicides (FRAC group 3) and succinate dehydrogenase inhibitor (SDHI) fungicides (FRAC group 7) is ongoing. A qPCR assay was developed to target a point mutation (Y136F) of the CYP51 gene that is a contributing factor to DMI resistance in E. necator; the mutation is present in 79.1% of tested 2017 samples (n=43). Analysis by sequencing of the SDHI gene complex has yielded 5 point mutations within two of the subunits in the SDH complex. Two of those mutations induce amino acid changes in their respective proteins and are being analyzed for potential contribution to SDHI resistance.

Durable Powdery Mildew Disease Resistance in Vitis Vinifera: Proof-of-Function and Discovery of Genetic Variation

For the past sixty years, mlo resistance to barley powdery mildew has remained durable and is now incorporated into over 50{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} of the European barley acreage. Related mlo genes in Arabidopsis, peach, pea, and tomato have recently been discovered and also confer resistance to powdery mildew via the same penetration resistance mechanism. Previously, we identified four strong candidates for mlo resistance in grapevine. To determine whether this gene could confer durable resistance in grapevine, we are pursuing a transgenic gene silencing approach. In 2009, we designed, cloned and confirmed 17 new artificial microRNA (amiRNA) silencing constructs. In 2010, we initiated a collaboration with Dr. Bruce Reisch to develop stably transformed Chardonnay grapevines. Plans are to silence four VvMlo candidate resistance genes individually, in pairs, and all at once in a thorough effort to confirm resistance gene function. In addition, we implemented a similar strategy to silence two candidates for a second resistance gene from Arabidopsis, Pmr6, individually and in pairs. Typical results for biolistic transformation experiments yielded hundreds of transiently-transformed Chardonnay cells per Petri dish. To-date we have regenerated a total of 58 transgenic plants that are ready for functional characterization and have over 600 total embryos after selection that may germinate for transfer to plant growth medium. While the transgenic assays described above could efficiently result in the development of transgenic V. vinifera cultivars with powdery mildew resistance, the information obtained could also be harnessed to identify V. vinifera breeding germplasm with previously undetected resistance alleles.

Revising the Gubler-Thomas Model for Powdery Mildew – Adding a Revised High Temperature Threshold, Varietal Susceptibility, and Expanding Public Access

The Gubler Thomas (GT) risk assessment model is a powerful tool that has been used to successfully control grape powdery mildew throughout California, as well as many other grape growing regions of the world. Proper use of the model has significantly reduced the number of fungicide applications significantly in most years while maintaining thorough disease control. Powdery mildew, caused by the fungus Erysiphe necator, can quickly spread throughout a vineyard. Controlling and reducing powdery mildew in grape crops can lead to higher yields and quality in subsequent grapes and wine. In California, powdery mildew germination, growth and reproduction is largely dependent on temperature. If the temperature remains above the pathogens optimal range for too long, the pathogens growth is inhibited. In previous years, our lab has shown that the pathogen can survive higher temperatures than previously thought (Backup, 2009). This led us to begin field trials testing whether a new high temperature threshold could be successfully incorporated into the model, with the goal of reducing fungicide sprays and improving disease control. We report here our finding on our second year of field trials and our continuing efforts in the laboratory to better understand the effects of heat, including a continuing study of the effects of multiple consecutive exposures to lethal and sub-lethal temperatures. In the last two years, we created two field trials that integrated our controlled environment laboratory studies with a vineyard and fungicide application regiment. These field sites were developed to test alternate high temperature thresholds for the GT model. In response to last years data, we have increased the duration that the pathogen must be exposed to 34?C, the lowest high temperature threshold. In addition, we have also included this year a new high temperature threshold modified by a cooperator in Oregon. All of the experimental high temperature models showed significant control in one of our vineyards, although none of the new models were significantly different from the original GT model. Our other vineyard had significantly greater disease pressure. All of the experimental models performed just as well in reducing disease incidence and severity compared to the GT original model. In the 2010 growing season, the unusual mild weather at our field sites made it difficult to test alterations to the high temperature threshold. The alternate models did not differ from one another in their subsequent fungicide spray schedules until the end of the season. Without higher temperatures, it is challenging to distinguish between these alterations. As part of the Western Weather Work Group (http://uspest.org/wea/index.html), we are collaborating with public and private scientists to help control disease in the western US by providing access to real time, virtual and forecasted weather data. We have completed 2 years of field data, and we are adjusting how the GT index accounts for observed delay in fungal growth. These modifications will ultimately be used to assist both public and private end users of the GT model.

A Microbiome-Based Approach to the Management of Grape Powdery Mildew

This report describes the progress we made during the first year of a two-year project that aims to describe the microbial communities (?microbiomes?) that are naturally associated with grape leaves and berries in vineyards. We are interested in these communities because we think they contain strains of bacteria, fungi and/or yeast that can be exploited to delay or predict the establishment in the vineyard of the fungal pathogen Erysiphe necator, which causes grape powdery mildew. Such strains are relevant to grape growers, because they can be incorporated into existing management practices or forecast models for powdery mildew in an effort to reduce the use of fungicides. To identify said strains, we are using a novel DNA-based method called pyrosequencing that is superior over older methods in revealing the identity and relative abundance of members that make up these microbial communities. So far, we have collected over 350 leaf and berry samples from 7 different vineyards in 2 states (California and Oregon), from 5 different clones of Pinot Noir and Chardonnay, and from 6 different stages of grape growth, ranging from flowering to harvest. This large number of samples is necessary to increase our chances of being able to pick out, from the huge number of microorganisms that are naturally present on grape leaves and berries, those that show a negative or positive correlation with the occurrence of E. necator and thus have true potential as challengers or foretellers of the establishment of this fungus in the vineyard. From the surfaces of all leaves and berries we collected, we have isolated bacteria, fungi, and yeasts and have quantified their abundances. We also extracted their DNA for pyrosequencing analysis which is currently underway. A preliminary analysis of a subset of DNA samples provided us with an early view into the bacterial diversity that is associated with Chardonnay leaves and berries. Clear from this analysis is the fact that leaves and berries differ in the types of bacteria that they carry on their surfaces. Does this difference correlate with levels of E. necator on the same leaves and berries? If so, what bacterial types underlie this difference, and can they be used for our purpose? These are some questions that will be addressed in the second year of the project. Our preliminary data set from the first year is the start of a database that will grow during the course of the second year into a depository of DNA-based information which we can use to identify strains with protective or predictive value against grape powdery mildew. The long-term goal of this research is to present grape growers with ?microbiome?-based options for management decisions relating to disease in their vineyards and quality of their product.

Development of Grower Performed LAMP PCR for Detection-Based Management Programs for Grapevine Powdery Mildew in Vineyards

Inoculum detection for timing fungicide applications against grape powdery mildew has been shown to work using qualitative and quantitative PCR approaches. However, these approaches require expensive equipment, specialized skills, and labor costs that impede implementation by viticulturist. Loop mediated isothermic PCR (LAMP) is a robust method for the detection of DNA that can be performed with minimal equipment and skill. A set of LAMP primers were designed against the ITS2 segment of the ribosomal DNA region of Erysiphe necator that are specific and can detect less than one spore or less than 5 copies of target DNA in a purified plasmid. Spores were trapped from vineyard air using by continuously running an impaction trap with 40 ? 1.5mm stainless steel rods coated with vacuum grease and replacing sample rods every 3 to4 days. DNA extraction was accomplished by placing rods in 100 ?l of TE buffer, centrifuging for 1 min, boiling for 5 min, vortexing for 10 sec and then placing 5 ?l DNA extract in PCR tube with mastermix. The PCR tube was then placed at 65?C for 45 min followed by 80?C for 5 min. Positive detection was determined by the formation of white precipitate. Grower implementation was tested by placing 3 traps at each vineyard with one processed by the grower using LAMP and the others processed in the lab for LAMP and quantitative PCR (qPCR). The results of the grower implementation was that participating growers were more than 50{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} accurate in detecting 1 or 10 spores and 100{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} accurate in detecting 100 spores in spiked samples. They had 74{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} agreement in detecting E. necator compared to our LAMP-PCR results, and our LAMP-PCR results were 96{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} in agreement with our qPCR results.

Durable powdery mildew disease resistance in Vitis vinifera: proof-of-function and discovery of genetic variation

For the past sixty years, mlo resistance to barley powdery mildew has remained durable and is now incorporated into over 50{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} of the European barley acreage. Related mlo genes in Arabidopsis and tomato have recently been discovered and also confer resistance to powdery mildew via the same penetration resistance mechanism. Previously, we identified four strong candidates for mlo resistance in grapevine. To determine whether this gene could confer durable resistance in grapevine, we are pursuing a transgenic gene silencing approach and are screening Vitis vinifera germplasm for natural and induced mutations. In 2009, we designed, cloned and confirmed 17 new artificial microRNA (amiRNA) silencing constructs. After showing that Agroinfiltration and protoplast transfection were inefficient transient assays for detecting gene silencing, we initiated a collaboration with Dr. Bruce Reisch to develop stably transformed Chardonnay grapevines. Plans are to silence four VvMlo candidate resistance genes individually, in pairs, and all at once in a thorough effort to confirm resistance gene function. In addition, we implemented a similar strategy to silence two candidates for a second resistance gene from Arabidopsis, Pmr6, individually and in pairs. Thus far, typical results for biolistic transformation experiments have yielded 1000 to 3000 transiently-transformed Chardonnay cells per Petri dish, with regeneration of in vitro plants scheduled for 2010. While the transgenic assays described above could efficiently result in the development of transgenic V. vinifera cultivars with powdery mildew resistance, the information obtained could also be harnessed to identify V. vinifera breeding germplasm with previously undetected resistance alleles. Having previously discovered a surprising absence of functional genetic variation in preliminary investigations of Mlo DNA sequences across diverse Vitis spp., we successfully adapted and tested capillary electrophoresis ecoTILLING for grapevine in 2009. However, we documented the inefficiency of the technology when searching for known mutations and finding that peaks for known mutations were often hidden by background noise and that most peaks were false positives ? artifacts of the technology (86{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23}). Therefore, we developed and implemented a strategy to sequence 9 Mb (million base pairs) of our Mlo and Pmr6 candidate genes in a Chardonnay M2 seedling population segregating for random, induced point mutations. This will allow us to identify V. vinifera seedlings with functional mutations for use in grape breeding programs along with molecular markers perfectly linked with the resistance gene.

Revising the Gubler-Thomas Model for Powdery Mildew – Adding Real Time PCR and a Revised High Temperature Threshold

Use of the Gubler-Thomas Model for powdery mildew risk assessment by California grape growers has already achieved goals of better disease control by targeting fungicide applications to high risk conditions over many of the temperature ranges seen in California grape growing areas. Use of the model in some years has reduced fungicide applications significantly. We report here our results to extend the predictive power of this model to higher temperature regimes. Better control of powdery mildew of grape has the potential to improve crop yield and quality as well as sustainability. Extending the high temperature range of the Gubler-Thomas Model would potentially allow for even fewer fungicide applications per season. With global climate change, the need for better understanding of the role of high temperatures on disease may increase in importance. We have conducted 2.years of controlled environment studies which have enabled us to understand the influence of high temperatures on grape powdery mildew (Backup 2009). Under controlled laboratory conditions, temperature and duration were carefully tested and their effects characterized on germination, growth and sporulation of the fungus. Our work shows that E. necator continues to germinate, infect, grow, and sporulate in the lab at higher temperatures than previously thought. In 2009, after controlled work with single heat exposures, we tested how multiple heat spikes affect fungal growth and reproduction. We found that temperature is increasingly lethal to the pathogen, and slows or reduces colony survival, delays spore production, and reduces the number of spores produced. The higher temperatures, such as 36 and 38° C for 4 and 2 hours, respectively, had a more pronounced effect than 34º C or the room temperature control, as did the higher number of consecutive heat treatments (1, 2 or 3), although to a lesser degree than temperature increases alone. However, repeated exposures to 4 hours of 36 and 38 C up to 3 times, which would total 12 hours, did not result in the same colony death and lack of spore production as one longer exposure of 12 hours straight had resulted. It appears that the fungus can and does recover with these shorter high temperature intervals. We have completed one year of testing the results of our integrating the lab work with our field studies. We are focusing on 36 °C and 38° C as important high temperatures and at 2 to 4 hours duration for the new high temperature threshold for the GT model. We are also adjusting how the index accounts for observed delays in fungal growth and reproduction due to the sublethal conditions experienced in the vineyard including not adding points to the index for several days after a high temperature spike to mimic the delays in growth observed in the lab. We are using the results from this work to assist both public and private end users of weather data and the model. With funding from the CDFA Specialty Crops Block grant only, we are developing vineyard spore trapping and molecular diagnostic techniques for the pathogen.

Revising the Gubler-Thomas Model for Powdery Mildew

Laboratory studies were developed to measure the effects of high temperature (above known optimal temperatures for the fungus) on spore germination, colony expansion and spore production on 4 isolates of E. necator from California vineyards. Spore germination, colony size, and spore production decreased as a result of either increasing high temperature or duration. Regression analysis showed increasingly negative slopes with shorter (in hours) x-axis intercepts as temperature increased from 30° to 42° C. High temperature-dependent reductions in spore viability, colony growth, and rate of sporulation can reduce the rate of disease development in the field.

At 30° and 32° C, colonies continued to grow, sporulate, and germinate after 24-hr heat treatment. At these temperatures, decreases in pathogen fecundity compared to 22.5º C appeared to be negligible or inconsistent, and not biologically relevant for most growing areas in California. Treatment at 34º C decreased biological activity of E. necator, but at 36º C between 12 and 24 hours was required to kill the pathogen. For all 3 assays, the duration required to extinguish biological activity reduced as the temperature increased in 2º increments. At 44º C, only ½ to 1 hr was required to extinguish all or most activity. In addition to locating the lethal conditions from 30 to 44º C, and from ¼ hr to 24 hr, we also detected sub-lethal conditions that delayed colony growth by up to 6 days. Using multiple regression analysis, we derived equations for each assay to predict cessation of biological function in the pathogen as a function of temperature and duration. The three-factor ANOVA resulted in highly significant main effects and interactions for temperature and its duration, but isolate was not significant, nor were interactions of isolate with any other factor. This information then is potentially relevant to a number of California’s grape growing regions, because under field conditions, vines may be repeatedly exposed to sub-lethal temperatures.

Although our controlled environment results are compelling and internally consistent among the 3 assays, they differ from previous field data used to establish the high temperature threshold of the Gubler-Thomas powdery mildew risk index, and with some published studies. In general, our results indicate pathogen survival at higher temperatures and durations than previously reported. Some of the discrepancy is because our laboratory experiments do not account for effects of uV radiation, sub-optimal humidity, ontogenetic resistance, or resistance to infection induced by high temperatures. But our data strongly suggest that the powdery mildew risk index should be revised for higher temperatures.

We established the second year of a field trail in unsprayed plots in a Sacramento County Chardonnay vineyard to study variations in canopy microclimate and powdery mildew. We installed a Metos weather station with real time satellite data collection and web based access to hourly temperature data and the powdery mildew risk index. The weather station had a sheltered ambient air temperature sensor at 2.2m height above the canopy, and 16 specially designed thin wire thermocouples attached to the underside of leaves for intensive characterization of the vine canopy microclimate. We found that fully exposed external leaves had higher temperatures on average about 2° F than internal shaded leaves, with maximum differences at times of 10° F. We found that internal shaded leaves were closer in temperature to above the canopy sheltered ambient air temperatures. Exposed, external leaves experienced 3 times more hours above the current high temperature threshold for the Gubler Thomas model than internal leaves or the sheltered ambient air temperatures. We found greater disease earlier in the season on the shaded leaves, although eventually a similar incidence but lower severity was observed on the exposed leaves inoculated mid June. Internal leaves inoculated mid July had consistently greater disease than the external leaves through mid August. A handheld IR leaf surface temperature sensor measured about 10° F lower temperatures than the thermocouple sensors attached to the bottom of the same leaves. We found the IR temperature sensor to be quite variable and did not allow for a consistent differentiation of leaf surface temperatures by canopy location.

We will continue with our analysis of both the laboratory and field data collected in 2008, to complete this project. Results from the controlled environment studies and the field studies will be combined to refine the Gubler-Thomas model at high temperatures. Our goal will be to adjust the lethal and sub-lethal ranges predicted in the lab to conditions found in the canopy. By correlating disease outbreaks with canopy microclimate conditions recorded by environmental sensors, we will be able to identify which temperatures at what durations stop, reduce, or promote powdery mildew epidemics.

Revising the Gubler-Thomas Model for Powdery Mildew Control

Laboratory studies were developed to measure the effects of temperature on spore germination, colony expansion and spore production on 4 isolates of E. necator from California vineyards. Spore germination was decreased as a result of either increasing temperature or duration. Higher temperatures were eventually lethal to conidia. Preliminary regression analysis showed relationships between temperature and spore germination had increasingly negative slopes with shorter (in hours) x-axis intercepts as temperature increased from 30° to 42° C. A majority of the temperatures studied prevented spore germination at shorter durations than for the other 2 assays. Reductions in spore viability due to high temperatures can reduce the rate of new infections in the field and are important in managing the pathogen, even if already established colonies are still viable.

At 30° and 32° C, colonies continued to grow, but at a slower rate than at the optimum growth temperature of 24° C. Higher temperature treatments produced increasingly negative regression slopes. Exposure to 40°, 42°, and 44° C stopped colony growth and was capable of eradicating mature colonies at sufficient durations (3 hr, 1.5 hr, and 1 hr, respectively). Growth rates increased from 9 to 19 mm2/day and sporulation increased from 13 to 41 conidia/ mm2 at 24° C as colonies matured over the durations studied. Colony expansion and spore production rates increased at the lowest temperature studied (30° C), but were dramatically reduced at higher temperatures studied.

Results of the first year laboratory studies are important in developing and refining the 3 assays used to assess high temperature effects on E. necator sporulation, growth, and infection. We established a field trail in unsprayed plots in a Sacramento County Chardonnay vineyard to study variations in canopy microclimate and powdery mildew. A research weather station (Campbell Scientific CR10) with 4 thin-wire thermocouples and light intensity (PAR) sensors was installed. Higher south-side leaf surface temperatures were observed (21.3° C) compared to north (20.7° C) (averaged for 1.5 months), with greatest differences at highest temperatures. Maximum south-side temperatures reached 36.7°C and north-side 35° C, with a maximum difference of 3.5° C. Ambient sheltered temperatures at 2 M were higher than canopy leaf surface temperatures, with greatest difference 10° C higher than north-side of the vines. First detected on May 23, 2007, powdery mildew disease incidence (DI) and severity (DS) were rated weekly. DS north-side was not significantly different from DS south-side. Interestingly there was slightly more disease early season south-side as compared to north-side (60{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} vs. 51{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} DI) and (4{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} vs. 2{aed9a53339cdfc54d53cc0c4af03c96668ab007d9c364a7466e3349a91bf0a23} DS). Subsequently, DI and DS north-side were greater (although not significantly). The Gubler-Thomas PM risk index showed that environmental conditions were conducive to disease beginning about May 12, 2007 and continued into July.

We intend to conduct further analysis of both the laboratory and field data collected in 2007, and to use the results and additional data gathered in the second year to complete this project. Results from the controlled environment studies and the field studies will be combined to refine the Gubler-Thomas model at high temperatures.

Accelerating the Development of Powdery Mildew Resistant Grapevines through Marker Assisted Selection

Summary of year one results: The aim of this project is to harness molecular biology in the selection and advancement of improved cultivars having resistance to powdery mildew. Segregating populations from three sources of significant powdery mildew resistance (Vitis davidii, V. rotundifolia, and V. aestivalis), each backcrossed to V. vinifera, were previously generated by David Ramming. The first objective of this proposal is to characterize the plant-pathogen interactions, in terms of race-specificity and microscopic analysis, for each of the three resistance sources in order to inform the second objective, which is the development of molecular markers that co-segregate with powdery mildew resistance in each of these populations for use by grape breeding programs.

In 2006, powdery mildew resistance was assessed in 182 progeny from the three populations using three separate pathogen sources in California and New York. The resulting data suggests the presence of multiple race-specific resistance genes segregating in two of the populations. A subset of rotundifolia and aestivalis progeny resistant in the vineyard were susceptible to one or both pathogen sources. Therefore, we hypothesize and have data suggesting that these resistant parents contribute two (or more) resistance genes that segregate independently in their progeny and that some of the resistance genes would be rapidly overcome if inappropriately deployed. In contrast, some progeny were resistant regardless of the pathogen source, suggesting the presence of all parental resistance alleles (as a resistance gene pyramid). These results will be confirmed using powdery mildew isolates with differential interactions with the resistance genes. These two populations underscore the value of marker assisted selection, with which we will be able to monitor and pyramid all functional resistance genes using a simple molecular assay rather than assaying resistance and durability by complex inoculation studies with multiple pathogen sources.
We have not identified race-specificity in the davidii resistance source, which has the added intrigue of providing resistance against the penetration of the fungus (i.e., the pathogen is unable to access the epidermal cells where it must obtain sustenance to survive). Most powdery mildew penetration resistance genes are effective against all races of powdery mildew.

To address the second objective, we require molecular markers that are polymorphic (appear different between the two parents), so that we can track groups of genes, or regions of the genome, that were contributed by the resistant or the susceptible parent to individual progeny. For the first phase of the project, we are using Simple Sequence Repeat markers, or SSRs. In the first year, we were able to test 160 SSRs against each of the parents in these three populations, and we successfully identified polymorphic markers for each linkage group in each population. In each population over 100 SSRs were informative for mapping. We will initiate a second round of marker exploration to enhance marker coverage and will identify which polymorphic markers predict disease resistance to target further marker saturation for each population.