Abstract
Background and goals Grapevine red blotch virus (GRBV) is a major pathogen of Vitis vinifera L. that causes the eponymous grapevine red blotch disease (GRBD). This disease is characterized by reduced leaf carbon assimilation and delayed berry ripening, resulting in diminished fruit quality. Recent work suggests that the virus alters leaf carbon metabolism prior to emergence of visible symptoms. However, the physiological manifestation of this altered metabolism is currently unknown. Therefore, an experiment was conducted to unravel the progression of GRBD symptoms in field-grown Pinot noir grapevines.
Methods and key findings Diurnal and seasonal measurements were conducted to quantify changes in leaf carbon balance and chronologize symptom progression in leaves and fruit. Leaf-level physiology was measured as a function of water relations, photosynthesis, and nonstructural carbohydrates in a commercial vineyard containing both healthy and infected vines. Additionally, fruit sugar and anthocyanin accumulation were monitored during ripening. Virus infection reduced carbon assimilation pre- and postveraison, but the effect was more pronounced postveraison and during the afternoon. Similarly, virus infection significantly increased leaf starch concentration pre- and postveraison. Starch granules accumulated in the chloroplasts and caused apparent damage to their structural and functional integrity. The virus had no sustained impact on chlorophyll fluorescence, suggesting that changes in fluorescence were a transient response to reduced carbon assimilation and export.
Conclusions and significance This study provides evidence that GRBV infection causes starch accumulation in infected leaves, which may have downstream effects on leaf carbon metabolism and ripening prior to the appearance of visible symptoms.
Introduction
Grapevine red blotch virus (GRBV) is a phloem-limited monopartite virus of grapevine (Vitis vinifera L.) in the family Geminiviridae (Sudarshana et al. 2015) and is the causal agent of grapevine red blotch disease (GRBD) (Yepes et al. 2018). GRBD was first documented in California vineyards in 2008 (Calvi 2011), and GRBV has been subsequently identified in numerous viticultural regions around the world (Rumbaugh et al. 2021). GRBD significantly reduces fruit quality, primarily by impairing sugar accumulation in the berry and, in the case of red-fruited cultivars, by reducing skin anthocyanin accumulation (Girardello et al. 2019, Martínez-Lüscher et al. 2019, Levin and KC 2020, Kurtural et al. 2023). With the global world trade of wine valued at more than $40 billion (OIV 2023), the physiological impacts of this deleterious disease necessitate continued study to improve management strategies.
The primary visible symptoms of GRBD in the plant consist of the eponymous foliar reddening (red blotches) and are accompanied by impaired carbon assimilation (Anet) and fruit ripening (Martínez-Lüscher et al. 2019, Levin and KC 2020). Blanco-Ulate et al. (2017) reported that GRBV affects ripening by hindering transcriptional regulation of anthocyanin biosynthesis and impairing hormonal signaling pathways for abscisic acid, ethylene, and auxin. Martínez-Lüscher et al. (2019) measured increased concentrations of foliar nonstructural carbohydrates (NSC) together with lower total soluble solids (TSS) at harvest and concluded that the reduction in berry sugar accumulation was due to impaired sugar translocation from source to sink.
The coincidence of increased foliar NSC and increased foliar anthocyanins suggests feedback inhibition of Anet, whereby excess NSC is a signal to downregulate Anet and synthesize photoprotective flavanoids (Lee and Gould 2002). Such a mechanism has been implicated in other grapevine-virus pathosystems, like grapevine leafroll virus (Halldorson and Keller 2018), and with other phloem-limited pathogens of woody perennial crops such as Liberibacter spp., the causal agent of huanglongbing (HLB; citrus greening) in citrus (Etxeberria et al. 2009). Girdling experiments have shown that mechanical injury to the phloem causes an accumulation of foliar NSC and subsequent reduction of Anet (Roper and Williams 1989). In response to biotic stress from pathogens, host defense responses such as callose formation around plasmodesmata are triggered in the phloem to prevent systemic infection (Lewis et al. 2022) and could also cause an excess accumulation of foliar NSC. An increase in foliar NSC—primarily the starch accumulating in the chloroplast stroma as insoluble semicrystalline granules—also occurs under nutrient deficiencies, such as with nitrogen (Bondada and Syvertsen 2005) and calcium (Duan et al. 2020). It has been further suggested that reduced Anet and translocation rates contribute to other reported GRBD symptoms such as reduced stomatal conductance (gs) and increased stem water potential (Ψstem) (Martínez-Lüscher et al. 2019, Levin and KC 2020). However, these associations have been inferred only from seasonal changes to Anet and late-season NSC measurements, without quantifying changes in sugar translocation.
Recent studies of red oak (Quercus rubra) and grapevine presented a mass balance approach to estimating carbon export that can be similarly used to quantify inhibition of sugar translocation in GRBV-infected vines (Gersony et al. 2020, Dayer et al. 2021). In healthy vines, there is diurnal variation in both the composition and synthesis of foliar NSC, a disruption of which may constitute the impact of GRBV on carbohydrate production and export throughout the day (Chaumont et al. 1994, Yu et al. 2009). Dayer et al. (2021) and Gersony et al. (2020) integrated these parameters to show that there is considerable diurnal variation in carbon export rate in grapevine and red oak, respectively.
Chlorophyll fluorescence (ChlF) could further elucidate the effect of GRBV on photosynthesis. ChlF parameters exhibit similar diurnal variation to Anet (Ding et al. 2006) that could be altered by GRBV. Synthesis of photoprotective flavonoids like anthocyanin in infected leaves may indicate an effect of GRBV on photosystem II (PSII) (Feild et al. 2001). Additionally, specific ChlF parameters such as maximum photochemical efficiency of PSII (Fv/Fm) indicate whether there is sustained damage to the photosystem attributable to biotic or irreversible abiotic damage (Maxwell and Johnson 2000, Gallé and Flexas 2010).
The present study was conducted to quantify changes in leaf carbon balance, to test the hypothesis that GRBV reduces leaf carbon export, and to provide evidence for the role of foliar NSC accumulation in the downstream effects of GRBD. Additionally, we hypothesized that there is diurnal variation with respect to the impact of GRBV on leaf carbon metabolism and ChlF. Leaf water potential (Ψleaf), gas exchange, ChlF, and NSC were quantified in field-grown healthy and infected vines. Leaf carbon export was estimated using a mass balance approach integrating Anet and changes in NSC. Fruit sugar and anthocyanin accumulation were monitored in healthy and infected vines throughout the ripening period. Finally, leaf samples from infected vines were also used for microscopic imaging of the leaf mesophyll and chloroplast ultrastructure related to NSC accumulation. This study aimed to further chronologize the effects of GRBV on vine physiology and fruit ripening while providing a physiological basis for potential early detection of the virus for disease management.
Materials and Methods
Vineyard site and plant material
The study was conducted in a 3.9-ha commercial block of V. vinifera L. cv. Pinot noir (clone 777) grafted to 3309 Couderc (Vitis riparia Michx. × Vitis rupestris Scheele) rootstock. The vineyard site was in the Rogue Valley American Viticultural Area in southern Oregon (42°19′N; 122°56′W; 434 m asl). The vineyard was established in 2010 with a north-south row orientation, and 2.0 m × 1.5 m of row × vine spacing. The soil was comprised of both Medford and Gregory silty clay loams with less than 3% slope. Vines were trained on a vertical shoot-positioned trellis and spur pruned with bilateral cordons. The fruiting wire was located 1.0 m above the soil surface, with two sets of foliage catch wires at ~1.3 m and 1.6 m above the soil surface. All vineyard management practices were conducted according to industry standards for the region.
Determination of vine infection status and classification
Grapevines were sampled and tested for GRBV at commercial harvest of the 2018 growing season. Tissue samples consisted of four whole basal leaves (leaf blade and petiole), two from each cordon. Petioles were cut into ~1 mm slices using a sterile razor blade, of which 100 mg per sample was transferred to microcentrifuge tubes for DNA extraction. DNA extractions were performed using a modified CTAB DNA extraction protocol (Richards et al. 1994). The diagnostic primers used in PCR amplification were CPfor/CPrev for the GRBV coat protein gene fragment, REPfor/REPrev for the GRBV replication-associated gene fragment, and 16Sfor/16Srev as a grapevine internal control for the 16S ribosomal DNA gene fragment (Krenz et al. 2014).
Disease symptom severity was quantified weekly starting on 27 July and continued through 31 Aug in 2020. Severity was estimated as the percentage of symptomatic leaves exhibiting any foliar reddening. Raw percentage data were converted into midpoint percentage values for analysis (Horsfall and Barratt 1945). All data vines were retested using dormant cane samples at the end of each growing season and recategorized as needed based on test results, but the total vine number was not adjusted.
Experimental design and measurement procedure
In 2019, leaf samples for NSC analysis were collected from previously tested grapevines of two virus statuses, GRBV-negative (RB−) and GRBV-positive (RB+), on 6 Aug (preveraison) and 29 Aug (postveraison). Preveraison, 13 RB− and 15 RB+ vines were sampled. Postveraison, nine previously sampled vines of each virus status were sampled again. Two leaves per vine were excised and brought to the lab for NSC analysis.
Diurnal measurements were conducted in 2020 on 23 July (preveraison) and 3 Sept (postveraison). Measurements and samples were collected every 2 hr starting before dawn (0400 to 0500 hr) from RB− and RB+ vines, for a total of 10 vines across eight sampling times per date. At each sampling time, one leaf per vine was analyzed for Anet, gs, and ChlF, then excised for measurement of Ψleaf and stored at ~5°C for later lab NSC analysis. Due to the appearance of symptoms in several data vines previously classified as RB−, new RB-vines were identified prior to the postveraison data collection period. Vines that were previously classified as RB− in the preveraison period and in 2019, but later displayed GRBD symptoms and subsequently tested positive for GRBV via PCR, were subsequently reclassified as RB+ prior to analysis.
Apart from the pre- and postveraison diurnal measurements, seasonal measurements were also conducted on the same vines in 2020. At regular intervals throughout the growing season, the same data were collected as in diurnal measurements described above, but at a single time point in the day (solar noon). In addition, berry samples were collected for maturity analyses.
Environmental conditions
Air temperature and relative humidity data were accessed from a nearby weather station (MDFO, AgriMet, United States Bureau of Reclamation) located ~1.8 km from the study site (42°19′N; 122°55′W). Vapor pressure deficit (VPD) was calculated from air temperature and relative humidity. Photosynthetically active radiation data were collected using a photosynthetic photon flux density (PPFD) sensor attached to a portable photosynthesis system (LI-6400XT, LI-COR Biosciences) and determined at the time of each gas exchange measurement.
Chloroplast ultrastructure imaging
Leaves were sampled from previously identified RB− and RB+ vines from the same vineyard block as the field experiment on 7 Aug 2020 and immediately placed on ice. After 24 hr, the leaf samples (2 × 2 mm pieces) were fixed with 2% glutaraldehyde, 2% paraformaldehyde, 0.05 M cacodylate, and 0.1 M phosphate buffer. Following washing with a 0.1 M phosphate buffer and postfixing with 1% osmium tetroxide, samples were placed in a 4.4°C fridge overnight. After ~12 hr, samples were washed with double-distilled water (ddH2O) and dehydrated in a series of ethanol steps starting at 10%, in 10% intervals up to 90% concentration, for 10 min each. Samples were submerged in 100% ethanol for 10 min. Thereafter, they were placed in glass vials, submerged in an ice bath, and microwaved in a laboratory microwave (Pelco BioWave Pro 36500, Ted Pella, Inc.) for 1 min at 500 W, 0.06 MPa vacuum, 30°C, and load cooler at 23°C. Samples rested for 2 min before increasing ethanol concentration, then were placed in 100% propylene oxide.
The infiltration steps proceeded, starting with a ratio of Spurr’s epoxy resin to propylene oxide starting at 1:1, 1:1, 2:1, and finally, 100%. To ensure proper infiltration, each step was infiltrated overnight in a flow hood on a rotational agitator. Samples were then embedded in molds with 100% Spurr’s epoxy resin in a 21°C oven overnight. The samples were sectioned to 2 × 2 mm at 60 nm using an ultramicrotome (Reichert Ultracut R Ultramicrotome, Leica Microsystems). Sections were placed on polyvinyl formal resin-covered carbon-coated slot grids and air dried. After drying, the grids were placed on a grid stick and stained. The staining procedure consisted of suspending the grids in a 2% aqueous uranyl acetate solution (1 to 2 mL) for 20 min, then washing the grids 30 times in distilled water. Samples were then stained for 8 min with lead citrate (1 to 2 mL) and washed 30 times in ddH2O. Samples were imaged with a transmission electron microscope (FEI Tecnai G2 20 Twin, FEI Company) at various magnifications.
Leaf gas exchange, ChlF, and water status
Leaf gas exchange was measured as Anet and gs at each sampling point, using the portable photosynthesis system described above. Measurements were made only on cloudless days. Two fully expanded, sunlit (when sun was present) midcanopy leaves per vine were used for measurement. Relative humidity, temperature, and PPFD were set in the leaf chamber to match ambient conditions. Flow rate was set at 400 μmol/sec, and chamber CO2 concentration was set in the reference cell at 400 μmol/mol. Measurements were made once Anet and gs had stabilized, which took 60 to 90 sec per leaf. Infrared gas analyzers were matched every 20 min during measurement periods.
ChlF was measured simultaneously with gas exchange, using a pulse-amplitude modulated fluorometer (6400-40, LI-COR Biosciences). Fv/Fm was determined only for dark-acclimated leaves at the predawn time points on each measurement date. Subsequent analyses showed no differences between healthy and infected vines at each measurement date, thus predawn Fm values were used in the calculation of light-acclimated parameters. For light-acclimated leaves, the quantum yield of PSII (ΦPSII), linear electron transport rate (J), proportion of open PSII reaction centers (qP), and non-photochemical quenching (NPQ) were calculated. ΦPSII was calculated using the following equation, in which Fm’ is maximal fluorescence of the light-acclimated leaf and Fs is the steady-state fluorescence of the light-acclimated leaf:
NPQ was calculated using the following equation, in which Fm was substituted from the predawn measurements averaged per GRBV status:
J was calculated using the following equation, in which f is the fraction of absorbed quanta used by PSII (assumed 0.5), PPFD is PPFD at the time of measurement, and αleaf is leaf light absorptance recorded by the system:
Ψleaf was determined using a pressure chamber (Model 615, PMS Instruments) according to Levin (2019) for the same fully expanded, sunlit leaves (except prior to sunrise) as used for gas exchange and fluorescence measurements.
Extraction and quantification of NSCs
Leaves that were used in the gas exchange, fluorescence, and Ψleaf measurements were removed immediately after Ψleaf determination, placed in a paper bag, returned to the laboratory, and microwaved for 90 sec at 600 W to halt metabolic activity (Landhäusser et al. 2018). Samples were stored at −20°C until oven-dried at 70°C for 48 hr. Dried leaves were ground using a Mini Wiley Mill (Thomas Scientific) and passed through a 40-mesh sieve.
NSCs were extracted as described (Chow and Landhäusser 2004, Landhäusser et al. 2018), with some modifications. Twenty-five mg of ground leaf tissue were mixed with 1 mL of 80% ethanol in a 2-mL screwcap microcentrifuge tube, vortexed for 10 sec, and incubated at 90°C for 10 min. The ground leaf tissue suspension was then centrifuged at 13,000 × g for 1 min, and 1 mL of supernatant was removed for soluble carbohydrate quantification. The remaining starch pellet was washed twice more, using the same volume of ethanol, and all supernatant was discarded. After leaving the starch pellet to air-dry for ~18 hr, the pellet was resuspended in a 600 U/mL α-amylase solution and incubated at 85°C for 30 min and then centrifuged at 13,000 × g for 1 min. One hundred μL of supernatant was then mixed with 500 μL of 12 U/mL amyloglucosidase solution and incubated at 55°C for 30 min. The mixture was brought to room temperature before quantifying the glucose hydrolysate.
Glucose quantifications were performed using an anthrone-sulfuric acid method at 620 nm, in accordance with Leyva et al. (2008). Dilutions of a 180 mg/L glucose solution were used as a standard curve. Five mg of potato starch was used as a starch control for determining the starch digestion efficacy (SDE). Samples and glucose controls were analyzed in technical duplicates, and starch controls were analyzed in biological duplicates. Glucose and starch were calculated as mg/g of sample dry weight (mg/g DW) using the following formulae:
where A620mean sample is the mean absorbance of the unknown sample duplicates, A620mean blank is the mean absorbance of the reagent blank duplicates, Vextract is the volume of the extract expressed as L made with either ethanol or α-amylase solution, DF1 is the dilution factor of the soluble carbohydrate sample, DF2 is the dilution factor of the glucose hydrolysate sample multiplied by six or the dilution factor for the dilution of α-amylase solution in amyloglucosidase solution, 0.9 is the weight conversion factor of starch to glucose, ϵ is the absorption coefficient of the glucose standard, Wsample is the measured weight expressed as g of dry tissue sample in the tube, and SDE is calculated using the starch formula with a Wsample of 0.005 g.
Estimation of leaf carbon export
Leaf carbon export was estimated using a mass balance approach as outlined previously (Gersony et al. 2020, Dayer et al. 2021), with the following formula:
where export is the difference between average Anet and the change in carbon concentration (C) between two time points (t and t−1). C is the total quantity of NSC per unit leaf area converted into molar units compatible with Anet using the molecular weight of glucose. The export rate is expressed as μmol C/m2/sec and is plotted in figures at the midpoint between t and t−1.
Berry sugar content and skin anthocyanins
Sugar per berry and anthocyanin content were quantified at six and five time points, respectively, throughout ripening, starting at ~10 days prior to the onset of ripening (veraison). For each replicate vine at each time point, 20 berries were sampled and weighed, and 10 berries were subsampled to be stored at −20°C for later anthocyanin analysis. The remaining 10 berries were crushed, and the juice was centrifuged at 15,000 × g for 5 min before being analyzed for TSS using a digital refractometer (AR200, Reichert Analytical Instruments). Sugar per berry was estimated as the product of berry mass and TSS, as in Krasnow et al. (2009). The 10-berry subsamples were peeled, and the skins were then dried and extracted in 70% acetone for 24 hr at 100 rpm on an orbital shaker (VWR). Acetone was removed from skin extracts by vacuum distillation (Syncore Analyst Polyvap, BUCHI Corporation). Anthocyanins were then quantified from the distilled extracts using a comprehensive red wine phenolics assay for microplate spectrophotometry (Harbertson et al. 2002, Heredia et al. 2006).
Statistical analyses
Statistical analyses were conducted using R software for statistical computing (v. 4.0.3; www.R-project.org). For diurnal and seasonal data sets from 2020, two-way analyses of variance (ANOVA) were conducted on data fit with linear models corresponding to a split-plot design in which viral infection status was the main plot factor and sample time (or date) was the subplot factor. Models were fitted with the lmer() function from the ‘lmerTest’ package (Kuznetsova et al. 2017). For diurnal data, pre- and postveraison data sets were analyzed separately due to the spread of the viral disease as previously mentioned. At the end of the season, n = 3 and 7 for RB− and RB+ preveraison, respectively, while n = 5 for both statuses postveraison.
For 2019 data, two-way ANOVAs were conducted on data fit with linear models corresponding to a 2 × 2 factorial in completely randomized design in which viral infection status and sample date were the two factors. Models were fitted with the lm() function from the ‘R base’ package. Not all vines were the same across both sampling dates. Preveraison, n = 13 and 15 for RB− and RB+, respectively, while postveraison, n = 9 for both RB− and RB+. Means were estimated and compared using the ‘emmeans’ package (v.1.6.3; Lenth 2022). Figures were generated with the ‘ggplot2’ package (v.3.3.5; Wickham 2016).
Results
Environmental conditions
Environmental conditions during the diurnal measurements in 2020 were typical of the growing region, though air temperature (Tair) and VPD were higher compared to the historical average during the postveraison sample date (Figure 1A to 1D). Preveraison, Tair increased steadily until a maximum of ~31°C was recorded at 1800 hr. Postveraison, the maximum Tair of 38°C was recorded at 1600 hr. Following a similar trend, VPD reached maximum values during the same time periods as Tair, with values peaking at 3.8 and 5.8 kPa for preveraison and postveraison, respectively. Saturating PPFD for grapevine leaves (>1500 μmol/m2/sec; Smart 1974) was observed between 1000 and 1600 hr for both sampling periods (Figure 1E and 1F); however, this period was slightly shorter postveraison, as typical for early September compared to late July.
Diurnal courses of ambient air temperature (Tair; A, B), vapor pressure deficit (VPD; C, D), and photosynthetic photon flux density (PPFD; E, F) at the study site on 23 July (preveraison; A, C, E) and 3 Sept 2020 (postveraison; B, D, F). Tair and VPD data were 15-min averages recorded by the nearby weather station, while PPFD data were recorded by the LI-6400XT sensor head during photosynthesis measurements. Colors represent data collected during one sampling round. Smoothed lines show a 1-hr running average ± 95% confidence intervals.
Ψleaf and gas exchange
There were no significant effects of virus status on Ψleaf at any time of day, except for the measurement taken at 2000 hr for preveraison sampling, at which point RB+ vines had a higher Ψleaf compared to RB− vines (Figure 2A). At both sampling dates, Ψleaf fell steadily throughout the course of the day and reached a minimum between 1500 and 1700 hr. Despite similar predawn values for Ψleaf among the two sampling dates, the minimum values were lower postveraison (−0.99 MPa) compared to preveraison (−0.73 MPa).
Responses of leaf water potential (Ψleaf; A, B), stomatal conductance to water vapor (gs; C, D), and net carbon assimilation (Anet; E, F) between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines over the course of the day on 23 July (preveraison; A, C, E) and 3 Sept 2020 (postveraison; B, D, F). Data are means ± 1 SE (n = 3 to 7). *, **, and *** represent statistically significant differences between means at p < 0.05, 0.01, and 0.001, respectively.
Virus status affected gas exchange with respect to both gs and Anet, especially during the postveraison sampling period (Figure 2D and 2F). During this period, RB+ vines exhibited consistently lower gs and Anet values. Notably, the decreased postveraison Anet in RB+ vines was more pronounced during the afternoon than in the morning. Though there were statistically significant interactions among virus status and measurement time for gs and Anet during both sampling dates, during the preveraison sampling date, significant differences between individual means were limited to early morning and late evening. When averaged across all measurement times during which light levels were saturating (between 1000 and 1600 hr), gs and Anet responses between RB+ and RB− vines were not significantly different preveraison (p = 0.773 and 0.229 for gs and Anet, respectively); however, these responses were significantly different postveraison (p = 0.033 and <0.0001 for gs and Anet, respectively).
ChlF
There were no significant differences in Fv/Fm between RB+ and RB− vines on either sampling date (Table 1). Averaged across virus status, Fv/Fm was significantly higher preveraison compared to postveraison (p = 0.011), however, all values ranged between 0.75 and 0.85 (Table 1). Virus status altered light-acclimated fluorescence parameters such as ΦPSII, J, and qP during the postveraison sampling date (Figure 3A to 3F). These parameters exhibited significantly lower values in RB+ vines compared to RB− vines between 1200 and 1600 hr on the postveraison sampling date. The impact of GRBV on ΦPSII, J, and qP followed a similar trend to Anet, whereby reductions in RB+ vines were limited preveraison but pronounced postveraison (Figure 2E and 2F, Figure 3). Significant reductions in preveraison responses were observed at the end of the day (e.g., increasing reduction in J in late afternoon and early evening). Averaged across all measurement times during which light levels were saturating (1000 to 1600 hr), ΦPSII, J, and qP responses were not significantly different between RB+ and RB− vines preveraison (p = 0.479, 0.476, and 0.644 for ΦPSII, J, and qP, respectively) (data not shown). However, these responses were significantly different during postveraison sampling date (p = 0.008, <0.0001, and 0.005 for ΦPSII, J, and qP, respectively). NPQ was higher by an average of 15% during the postveraison sampling day in RB+ vines compared to RB− vines, but it was not statistically significant (Figure 3H).
Response of maximum quantum efficiency of dark-acclimated leaves (Fv/Fm) between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines taken predawn on 23 July 2020 (preveraison) and 3 Sept 2020 (postveraison). Data are means ± 1 SE (n = 3 to 7). ANOVA, analysis of variance.
Responses of quantum yield of photosystem II (PSII; ΦPSII; A, B), electron transport rate (J; C, D), proportion of open PSII (qP; E, F), and non-photochemical quenching (NPQ; G, H) between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines over the course of the day on 23 July (preveraison; A, C, E, G) and 3 Sept 2020 (postveraison; B, D, F, H). Data are means ± 1 SE (n = 3 to 7). *, **, and *** represent statistically significant differences between means at p < 0.05, 0.01, and 0.001, respectively.
Chloroplast ultrastructure
Increased starch content in the RB+ leaves corresponded with a massive accumulation of starch granules in the whole chloroplasts (Figure 4). As a result, chloroplasts were swollen and distended with starch, which perturbed the configuration of granal stacks and stroma lamellae, showing no evidence of stroma and thylakoid membrane assembly (Figure 4C and 4D). The starch granules displaced the photosynthetic membrane system toward the periphery of the organelle, destroying the grana and stroma lamellae and resulting in a loss of structural integrity. The membrane system remained intact in the healthy vines (Figure 4B).
Scanning electron micrograph of palisade (Pa) and spongy parenchyma (Sp) (A), chloroplast with intact membrane (M) system and grana (G) from healthy vines (B), chloroplasts (Ch) with massive accumulation of starch granules (St) in Sp (C), and Pa cells of grapevine red blotch virus-infected vines (D). Scale bars: 25 (A), 250 (B), 1 (C), and 1 μm (D).
NSCs
No differences were found in soluble sugar concentrations between RB+ and RB− leaves either pre- or postveraison in 2019 (Table 2). Soluble sugar concentrations were significantly higher in leaves postveraison than preveraison, irrespective of virus status. Starch, however, was significantly higher (16%) in RB+ leaves both pre- and postveraison. Unlike soluble sugar, the starch concentrations in RB+ and RB− leaves remained stable from the preveraison sampling date to the postveraison sampling date.
Responses of leaf starch and soluble sugars between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines sampled on 6 Aug (preveraison) and 29 Aug 2019 (postveraison). Data are means ± 1 SE (n = 9 to 16). ANOVA, analysis of variance.
No differences were found in soluble sugar concentrations between RB+ and RB− leaves during either pre- or postveraison sampling dates in 2020 (Figure 5A and 5B). Foliar soluble sugar concentration declined from predawn to midmorning (~1000 hr), then increased to a peak in the afternoon (between 1400 and 1500 hr), before dropping again in late afternoon.
Responses of leaf soluble sugars (A, B) and starch (C, D) between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines over the course of the day on 23 July (preveraison; A, C) and 3 Sept 2020 (postveraison; B, D). Data are means ± 1 SE (n = 3 to 7). •, *, and ** represent statistically significant differences between means at p < 0.10, 0.05, and 0.01, respectively.
Starch concentration, however, was higher in RB+ leaves compared to RB− leaves across measurement times during both sampling dates (Figure 5C and 5D). The leaf starch concentration was higher (ranging from 14 to 91%) in RB+ vines during late afternoon (past 1600 hr) compared to RB− vines at preveraison; however, the difference was statistically significant only after sunset (2000 hr). Average leaf starch concentration preveraison was 20% higher (p = 0.115) in RB+ vines (17.4 ± 0.9 mg/g DW) compared to RB− vines (14.5 ± 1.4 mg/g DW). Postveraison, the starch concentration in RB+ leaves was higher at all time points throughout the day (Figure 5D). On average, leaf starch concentration was ~32% higher (p = 0.029) in RB+ vines (21.3 ± 1.4 mg/g DW) compared to RB− vines (16.1 ± 1.4 mg/g DW). However, significant differences between individual means were observed only during early morning and late afternoon.
Carbon export
Like on Anet, GRBV had a noteworthy impact on carbon export, especially during late afternoon postveraison (Figure 6B and 6D). Although there were no significant differences in carbon export between RB− and RB+ vines preveraison, afternoon export in RB+ trended lower compared to RB−. Averaged across the afternoon (1300 to 1900 hr) measurements, postveraison carbon export was ~46% lower (p = 0.033) in RB+ leaves compared to RB− leaves. By comparison, postveraison Anet was ~37% lower in RB+ leaves compared to RB− leaves in the afternoon. Average Anet for RB+ leaves (8.6 μmol/m2/sec) was slightly higher than export rate (7.1 μmol/m2/sec) during the afternoon period.
Response of leaf carbon export rate between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines over the course of the day (A, B) and averaged during morning and afternoon (C, D) on 23 July (preveraison; A, C) and 3 Sept 2020 (postveraison; B, D). Data are means ± 1 SE (n = 3 to 7). * represents statistically significant differences between means at p < 0.05.
Chronology of disease symptoms
Over the course of the season, starch accumulation in RB+ leaves preceded delayed ripening in the fruit, which in turn preceded the expression of red leaves (Figure 7). The starch concentration of RB+ leaves was consistently—if not significantly—higher than that of RB− leaves as soon as 14 days before veraison (Figure 7A). Fruit ripening (i.e., berry sugar accumulation) in RB+ vines did not lag behind RB− vines until 10 days postveraison, at which point sugar per berry was significantly lower in RB+ vines compared to RB− vines (Figure 7B). Similarly, berry anthocyanin content was significantly lower in RB+ vines compared to RB− vines at 10 days postveraison (Figure 7C). Following the delay in fruit ripening, expression of red leaf symptoms (i.e., disease severity) increased significantly in RB+ vines (Figure 7D). The chronology of symptoms described here provides evidence for a cascade of physiological events that unfolds in RB+ vines and implicates carbohydrate signaling in RB+ leaves that precedes the appearance of the eponymous red blotches.
Responses of leaf starch concentration (A), berry sugar content (B), berry anthocyanin content (C), and disease severity (D) between healthy (RB−) and grapevine red blotch virus-infected (RB+) vines over the course of the 2020 growing season. The vertical dashed line indicates the date of veraison. Data are means ± 1 SE (n = 3 to 7). •, *, **, and *** represent statistically significant differences between means at p < 0.10, 0.05, 0.01, and 0.001, respectively.
Discussion
This study utilized seasonal and diurnal sampling strategies in combination with a mass balance approach to test the hypothesis that GRBV infection directly reduces leaf carbon export that leads to starch accumulation, which contributes to inhibition of photosynthesis. The results support this hypothesis and further demonstrate that the impacts of GRBV infection on leaf carbon metabolism vary seasonally and diurnally, with the strongest adverse effects occurring postveraison and in the afternoon, when water and heat stress are greatest. Additional microscopic evidence confirms that starch accumulates in the chloroplasts of RB+ vines and causes, at a minimum, physical damage to the photosynthetic apparatus. Finally, the results suggest a chronology of symptom development that begins with the early accumulation of starch in leaves and is followed by a reduction in leaf carbon export, Anet, berry sugar accumulation, anthocyanin synthesis in berry skins, and lastly, anthocyanin synthesis in leaves.
The impact of GRBV on carbon assimilation and export varies diurnally
The diurnal pattern of GRBV impact on gas exchange correlates with the degree of environmental stress insofar as Anet and carbon export are reduced by a greater magnitude later in the afternoon, when air temperature and VPD are higher. This is the same period of the day when Ψleaf reaches a daily minimum. Prior research studying the effects of GRBV on water relations report that disease symptoms are exacerbated by water deficits (Levin and KC 2020, Copp and Levin 2021). In contrast, maintaining a high plant water status can improve Anet and sugar accumulation in the fruit of GRBV-infected leaves (Copp and Levin 2021). Following the pressure-flow hypothesis of phloem function (Münch 1927), Copp and Levin (2021) hypothesized that maintaining a higher water status in RB+ leaves would increase the gradient of water flow in the phloem, subsequently improving sugar translocation out of the leaf. Thus, lower Ψleaf later in the day may contribute to the feedback loop of reduced phloem loading, accumulation of NSC, and consequent inhibition of photosynthesis. The mechanism, however, of inhibited translocation in the phloem in RB+ vines remains an open question but may be linked to physical blockages at the phloem sieve plates, as demonstrated in other plant pathosystems (Welker et al. 2022).
The higher average Anet for RB+ leaves compared to their carbon export rate during the afternoon period indicates a temporal gap between carbon export and assimilation. This study shows that GRBV affects Anet and carbon export rate differentially throughout the day, even postveraison when effects of the disease are assumed most severe (Levin and KC 2020). The late afternoon reductions in Anet and carbon export have obvious consequences for fruit ripening in RB+ vines, but also significant implications for gas exchange sampling and study of GRBV-infected vines. Gas exchange is often measured in plants when irradiance is saturating and stable to capture steady-state Anet and/or gs (e.g., at solar noon). However, the maximum impact of GRBV on Anet observed in this study may be mismatched with the windows for steady-state irradiance or Anet. This field study was instigated, in small part, to evaluate whether the standard sampling window for gas exchange measurements (~1100 to 1300 hr) could fail to capture the full impact of GRBV on gas exchange. Though the present study does not contain a continuum of diurnal response of gas exchange to GRBV throughout the season, it clearly demonstrates that standard sampling may not always capture the most significant impacts of GRBV on gas exchange.
Starch accumulates in GRBV-infected leaves
The present study observed only higher leaf starch in RB+ vines compared to RB− vines, in contrast to other studies of foliar NSC in RB+ vines that reported elevated soluble sugars (Wallis and Sudarshana 2016, Halldorson and Keller 2018, Martínez-Lüscher et al. 2019). The diurnal changes in soluble sugar concentration observed in our study match the previously reported pattern of sucrose concentration in grapevine leaves, increasing in the morning before saturating in the afternoon (Chaumont et al. 1994). At the time of day preveraison when this study observed saturation of soluble sugars in both RB+ and RB− vines, starch concentration in RB+ vines began to increase, whereas in RB− vines, starch decreased. Postveraison however, the saturation in soluble sugars in RB- vines was achieved at least 4 hr prior (1200 hr) to RB+ vines (1600 hr). This different temporal dynamic may explain why other studies reported differences in soluble sugars between RB+ and RB− vines. The time of day when leaves are sampled will consequently affect the NSC concentrations and relative differences between RB+ and RB− leaves. Nevertheless, leaf starch trended downward for both RB+ and RB− vines during the day postveraison, except for a brief but significant increase after the leaf soluble sugar saturated in RB+ vines. Thus, the excess leaf carbohydrates beyond this saturation point in RB+ vines may have been converted to starch, as previously shown in C3 plants (van Bel and Hafke 2005).
This study implicates starch accumulation in the reduction of photosynthesis in RB+ leaves, but the exact nature of this response remains unclear. In RB+ vines, the starch was mainly confined to the chloroplast, filling its whole volume and inhibiting Anet by physically disrupting the chloroplast and decreasing CO2 diffusion from the intercellular spaces to the stroma (Stitt 1991, Sawada et al. 2001). Other scenarios leading to such starch accumulation include fruit removal (Syvertsen et al. 2003), girdling (Roper and Williams 1989), nutrient deficiencies (Bondada and Syvertsen 2005, Duan et al. 2020), and high CO2 (Sawada et al. 2001). The data herein corroborate previous reports of starch driving end-product inhibition of photosynthesis across many plant species (van Bel and Hafke 2005). Goldschmidt and Huber (1992) showed that elevated starch and sucrose concentrations from girdling had a strong relationship with inhibition of Anet across species. Turgeon (2010) reported that a low NSC concentration (including sugars, sugar alcohols, and starch) in leaves of woody plants is associated with high efficiency in carbohydrate synthesis through photosynthesis. Halldorson and Keller (2018) showed an association of high leaf starch and total NSC with feedback inhibition of carbon assimilation in leafroll virus-affected grapevines. Although the current study did not measure leaf sucrose concentration, it is apparent that leaf starch accumulation was partly responsible for inhibition of Anet. The concentrations of sucrose, fructose, and glucose in response to vascular blockage, such as girdling or phloem-inhibiting viruses, vary across species and by invertase activity, but these carbohydrates are more transient forms than starch. Because chloroplasts are simultaneously the site of carbon assimilation and starch storage (Hummel et al. 2010), there seems to be a strong relationship between end-product regulation of photosynthesis and starch accumulation in GRBV-infected grapevines. Whether starch drives feedback inhibition of Anet in RB+ leaves or is merely an artifact of elevated soluble carbohydrates converted to starch, starch is a far more robust and consistent indicator than soluble carbohydrates.
The accumulation of starch in RB+ leaves may be detectable before other changes to carbon metabolism. In the measurements conducted over the course of the season, increases in foliar starch in RB+ vines relative to RB− vines were observed prior to veraison and were at least 2 wk prior to the first observation of red leaf symptoms. Detecting starch accumulation in RB+ leaves may be useful to guide sampling related to the effects of GRBV on vine physiology and carbon metabolism, but it has even greater potential to inform early detection of the disease using remote sensing techniques (Tanner et al. 2022). For example, hyperspectral imaging has been used for early detection of HLB through changes to foliar NSC concentrations, which tend to be higher in leaves from trees with HLB (Weng et al. 2018). Thus, the elevated foliar starch levels shown here have the potential to serve as biomarkers to guide the development of hyperspectral imaging protocols for early identification of GRBV without intensive lab-based virus detection methods (DeShields and KC 2023).
GRBV-infected vines do not exhibit sustained impairment of photochemical efficiency
The similar Fv/Fm values among RB− and RB+ vines observed in this study suggest there is no sustained impairment of photochemical efficiency due to GRBV infection. The measured values ranged from 0.79 to 0.81 and indicated a non-stressed status (Gallé and Flexas 2010) even in RB+ vines. Interestingly, GRBV may have little sustained impact on the efficiency of PSII. Light-adapted fluorescence (ΦPSII) was used to approximate photochemical efficiency during the day and was reduced in RB+ leaves relative to RB− leaves. This trend followed those of Anet and carbon export, to the degree that ΦPSII values in RB+ leaves were only significantly lower than those in RB− leaves in the afternoon during the postveraison sampling date. In tandem with transient reductions in J and qP, the reduction in ΦPSII suggests that the quanta of light absorbed exceeded that required for carbon assimilation (Gallé and Flexas 2010). This transient reduction in ΦPSII, J, and qP, and similar Fv/Fm for RB+ vines compared to RB− vines may support an end-product inhibition of photosynthesis, as opposed to sustained damage or alteration of photochemical efficiency (Pammenter et al. 1993).
NPQ did not vary with virus status, which was curious considering the reduction in photochemical quenching in RB+ vines inferred from reductions in ΦPSII, J, and qP. NPQ calculation usually requires Fm values from dark-adapted leaves, but predawn Fm values were substituted here in the calculation of NPQ, so interpretation of NPQ values may be limited. In leaves with reduced qP, elevated NPQ values are expected alongside the initiation of the xanthophyll cycle, to protect the leaf from light damage (Gallé and Flexas 2010). For RB+ vines, foliar anthocyanin synthesis may in part fulfill this role (Liakopoulos et al. 2006). In this study, all ChlF measurements were made on asymptomatic (i.e., green) leaf portions and thus dissipation of excess light energy would require another mechanism. Additional leaf phytochemical analysis may explain the NPQ capacity and photoprotective strategies of RB+ leaves both before and after anthocyanin synthesis.
Expression of disease symptoms is temporally linked to carbon metabolism
The measurements conducted during ripening indicated that reduced berry sugar accumulation preceded berry anthocyanin accumulation, which ultimately preceded anthocyanin synthesis in leaves. While Martínez-Lüscher et al. (2019) first linked the reduction in anthocyanins to lower sugar in the fruit, this study elucidates the temporal nature of this relationship. In the leaves, accumulated NSC may be involved directly and indirectly with anthocyanin synthesis. The role of carbohydrates in foliar and fruit anthocyanin synthesis is well documented, as sucrose is both the signal and substrate for anthocyanin synthesis (Pirie and Mullins 1976, Lecourieux et al. 2014). Halldorson and Keller (2018) demonstrated that the relationship between carbohydrate accumulation and anthocyanin synthesis in leaves is unaltered in leafroll-infected leaves and suggested that this can explain foliar anthocyanin synthesis in leafroll-infected vines. It is reasonable to surmise that this same phenomenon could be observed with GRBV, another phloem-limited virus, and future study may confirm this. While presently unclear if or how accumulated starch contributes to anthocyanin synthesis in RB-infected vines, these photoprotective pigments may also aid in scavenging excess light as Anet is reduced (Feild et al. 2001). The coordination of these processes has not been reported previously in RB+ vines.
Despite an emerging chronology, onset and severity of disease symptoms are likely influenced by the interaction of the virus and other biotic and abiotic factors. The association between severity of GRBD symptoms and stress suggests that foliar anthocyanin synthesis and accumulation of sugar and anthocyanins in the berries of RB+ vines relative to RB− vines could be accelerated. In this study, visible symptoms (i.e., red leaves) were first observed ~10 days postveraison at a berry TSS range of 16 to 17 Brix. By contrast, Levin and KC (2020) observed earlier onset of visible symptoms with respect to veraison (~5 days pre- and 1 day postveraison) and TSS (6 to 7 and 12 to 13 Brix [unpublished data]) over 2 years. That study, however, investigated the impact of deficit irrigation on RB+ vines, which may in part explain why visible symptoms appeared earlier. Indeed, the variability of symptom onset and severity across seasons and growing locations has been previously reported (Girardello et al. 2019, Levin and KC 2020, Rumbaugh et al. 2021) and is one of the most challenging aspects of studying this plant pathosystem.
Conclusion
This study delineates a cascade of GRBV-mediated physiological symptoms beginning with an accumulation of foliar starch, followed by reduction of Anet, impaired carbon export, reductions in sugar accumulation and anthocyanin synthesis in the fruit, and finally, synthesis of foliar anthocyanin. To the best of our knowledge, this is the first report of such chronological progression of GRBV in grapevines. This study also demonstrates that GRBV infection causes considerable diurnal variation in carbon assimilation and export, revealing the transient nature of the impact of GRBV on whole-plant carbon metabolism. The negative impact of the virus (and subsequent symptom expression) is likely caused by end-product feedback inhibition of photosynthesis that is modulated by abiotic environmental stress. In addition to the broadened understanding of GRBV, this study provides insights into the development of “invisible” symptoms (e.g., starch accumulation), which could guide future hyperspectral detection of disease prior to the appearance of visible foliar symptoms.
Footnotes
The authors thank Kathryn E. Lundquist for helping with the editing and presentation of figures. The authors also thank the hosting vineyard for providing the study site and for field plot maintenance, and the Rogue Valley Winegrowers Association for its continued support of the viticulture research program at the Southern Oregon Research and Extension Center. This work was supported, in part, by the Specialty Crop Research Initiative, project award number 2019-51181-30020, and, in part, by the Oregon Agricultural Experiment Station with funding from the Hatch Act capacity funding program, both from the United States Department of Agriculture’s National Institute of Food and Agriculture. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and should not be construed to represent any official USDA or U.S. Government determination or policy.
Copp CR, DeShields JB, Kar S, Clark RW, KC AN, Hallwachs B et al. 2025. Foliar starch accumulation precedes the cascade of grapevine red blotch disease symptoms. Am J Enol Vitic 76:0760001. DOI: 10.53444/ajev.2024.24045
By downloading and/or receiving this article, you agree to the Disclaimer of Warranties and Liability. If you do not agree to the Disclaimers, do not download and/or accept this article.
The data underlying this study are available on request from the corresponding author.
- Received August 2024.
- Accepted October 2024.
- Published online January 2025
This is an open access article distributed under the CC BY 4.0 license.