Grapevine vein clearing virus Is Prevalent and Genetically Variable in Grape Aphid (Aphis illinoisensis Shimer) Populations

The highest number of emergent diseases are viral diseases (Elena et al. 2014; Fargette et al. 2006). One factor for numerous emergences is the change of the agroecological interface in which new crops have been introduced. This change provides opportunities for plant viruses to diversify and invade introduced agricultural crops (Fargette et al. 2006; Gibbs et al. 2008; Jones 2009; Webster et al. 2007). Intensive cultivation accelerates the speed of a new virus emerging, especially when a crop happens to be a susceptible host to native viruses. Three emerging viruses of grapevine are Grapevine vein clearing virus (GVCV), Grapevine Pinot gris virus, and Grapevine red blotch virus (Cieniewicz et al. 2020). GVCV infection has resulted in the removal of seven vineyards in the state of Missouri since it was first reported in grape cultivar Chardonnay in 2004 (Qiu and Schoelz 2017).

GVCV belongs to the genus Badnavirus in the family Caulimoviridae (Zhang et al. 2011). The reference genome of GVCV-CHA, the first isolate identified and sequenced, is a circular, double-stranded DNA molecule of 7,753 bp. GVCV infects multiple grape cultivars (Guo et al. 2014; Qiu and Schoelz 2017) and is prevalent in Midwestern vineyards. The most susceptible cultivars are white-berried Chardonel and Vidal blanc. Most GVCV-infected grapevines exhibit symptoms of translucent foliar vein clearing, deformed leaves, shortened internodes, and severe dwarfing that gradually progress until vine death. Hence, it is one of the most destructive viruses affecting grape production in the Midwest (Qiu and Schoelz 2017). Reducing GVCV incidence is urgent; this would be expedited by finding sources and transmission routes of the virus.

Our previous studies discovered possible sources of GVCV in vineyard ecosystems (Beach et al. 2017; Petersen et al. 2019). The virus was found in wild plants in the Vitaceae family. Two isolates have been sequenced from wild Vitis rupestris and given the variant names of GVCV-VRU1 and GVCV-VRU2. Both share 92% identical nucleotides with the reference GVCV-CHA isolate (Beach et al. 2017). Up to 34% of Ampelopsis cordata plants in Missouri, a close relative of Vitis species, are infected with GVCV (Petersen et al. 2019). Wild Vitis and Ampelopsis species grow across a wide range of native habitats in the Midwest and are perennial components of a vast natural flora. Once infected by GVCV, they become natural reservoirs that are potential sources of GVCV inoculum to cultivated Vitis spp. during each growing season. However, GVCV cannot reach vineyards without a mobile vector.

Grape aphids can transmit GVCV from a native A. cordata to cultivated Chardonel under greenhouse conditions (Petersen et al. 2019). Two central questions are as follows: What is the proportion of grape aphids in native habitats that carry GVCV? What is the genetic relationship of GVCV isolates among native hosts, aphids, and cultivated grapevines? Knowing the prevalence of GVCV in grape aphids will reveal virus-carrying aphid populations that are acquiring GVCV from infected native plants and can potentially transmit the virus onto new plants during each growing season. This knowledge will contribute to the understanding of GVCV epidemiology among native plants and cultivated grapevines. It will also fill a missing link between virus reservoirs and production vineyards in this historically and ecologically important agroecological system. The link is crucial to designing a sustainable disease management strategy. Furthermore, conducting genetic analysis of GVCV variants in grape aphids and host plants will uncover the genetic structure of GVCV among mobile vectors and immobile hosts.

Virus epidemics in a host and its insect vector are intertwined (Roossinck 2015). An insect vector moves a plant virus between native and cultivated hosts, and thus is a major player in the ecology and evolution of a plant virus (Roossinck 2013, 2015; Roossinck and García-Arenal 2015). Insect transmission links sources of viruses with new hosts, and movement of virus-bearing insects shapes the temporal and spatial patterns of disease spread in an agroecological system (Jeger et al. 2018). Understanding virus–vector–plant interactions in an agroecological system will help manage viral disease (Malmstrom et al. 2011). Viral diversity analysis in its transmission vector helps to understand viral evolution, ecology, and epidemiology (García-Arenal et al. 2001; Moury et al. 2006; Pagán 2018), and thus can be used to trace dispersal patterns of a plant virus over geographical landscapes (Picard et al. 2017).

GVCV, its grape aphid vector, native hosts, and cultivated grapevines provide a good model for studying virus ecology and epidemiology. In this study, we present a comprehensive survey of GVCV among aphid populations. We acquired open reading frame (ORF) II sequences of GVCV isolates from groups of 10 aphids and from single aphids, and we comparatively analyzed them with the corresponding sequences of GVCV isolates from cultivated grapes, wild Vitis spp., and A. cordata. We also analyzed genome numbers of GVCV in individual aphids. The results will help vineyard managers to plan, design, and implement strategies for preventing GVCV spread into vineyards.

Materials and Methods

Collection of grape aphids and plant samples.

Grape aphid samples were collected from May to September in the 2018 and 2019 seasons. Vitaceae plants were inspected for grape aphid, and both alate and apterous individuals were collected. We refer to all aphids inhabiting a single vine as a colony, while a community of aphids is a distinct group found on a single shoot of a vine. The communities contained populations ranging in size from 20 to hundreds of aphid individuals. The aphids found on Vitaceae plants were collected and labeled with specific colony and community numbers. One hundred and five grape aphid communities and their host leaf tissue samples were collected in 2018, and 192 additional single aphids were assayed in 2019 from several locations as listed in Supplementary Table S1. Coordinates of each sample were recorded using the Gaia GPS navigation app (Trailbehind Inc.) and are included in Supplementary Table S1. This study also includes GVCV isolates from Vitaceae samples collected in 2018 and 2019 and all previously collected isolates since 2009. A map showing locations of GVCV isolates in aphids and hosts is presented in Supplementary Figure S1.

Fifteen adults and 26 immature aphids were sent to the Florida Department of Agriculture and Consumer Services for morphological identification. The identity of grape aphids was further verified by the amplification and sequencing of a fragment of the mitochondrial cytochrome oxidase subunit I (COI) gene from DNA extracts obtained from five randomly chosen aphid communities using primers LCO 1490 and HCO 2198 (Folmer et al. 1994) (Supplementary Table S2).

DNA extraction from plants, aphid communities, and single aphids for diagnostic and quantitative PCR.

DNA was extracted from the plant samples as described by Petersen et al. (2019). Briefly, 80 to 100 mg of leaf tissue was processed for DNA extraction using the Synergy 2.0 Plant DNA Extraction Kit following the manufacturer’s protocol (OPS Diagnostics, Lebanon, NJ). DNA was eluted in 75 µl of autoclaved, deionized water.

Total DNA was also extracted from a group of 10 aphids from each community using an Insect DNA E.Z.N.A Kit (Omega Bio-tek Inc., Norcross, GA), according to the manufacturer’s protocol with two modifications. During the first step of the protocol, aphids were crushed with tip-sealed 1,000-µl pipette tips in CTL buffer provided with the kit instead of grinding in liquid nitrogen. DNA was eluted in 30 µl of autoclaved, deionized water.

Extraction of total DNA from single aphids (n = 525) was performed with STE buffer (0.1 mM of NaCl, 1 mM of EDTA, pH 8, and 10 mM of Tris-HCl, pH 8) using a 10% Chelex-100 as in a published protocol (Wang and Wang 2012). The grape aphids were examined under a microscope to record size, color, and life stage of each specimen prior to extraction. A single grape aphid was ground in 100 µl of Chelex DNA extraction solution. The homogenate was incubated at 65°C for 20 min, and then boiled at 100°C for 10 minutes. The samples were centrifuged at 15,700 × g for 5 min and the supernatant containing the DNA was transferred into sterile tubes. The quality of the extracted DNA was analyzed on a NanoDrop 1000 instrument (Thermo Fisher Scientific, Waltham, MA). The final DNA concentration was adjusted to 50 ng/µl.

To analyze GVCV genome numbers in the stylet and body of individual aphids, 20 single aphids were selected from the remaining aphids of three previously tested communities: five and 10 aphids from two communities that had tested positive for GVCV and five aphids from one community in which GVCV was not detected. Each grape aphid was examined under a microscope (4.4:1 zoom ratio, Leica EZ4) and the stylet was severed from the body using a razorblade. To prevent cross-contamination, the razorblades were dipped in a 10% bleach solution and wiped dry between samples. The head was included with the body for DNA extraction in the first set of 10 aphids but was removed from the body and discarded in the second set of 10 aphids, with the stylet and the body placed into separate tubes. DNA was extracted using the Insect DNA E.Z.N.A kit (Omega Bio-tek Inc.) with two modifications: the aphid stylets were not ground to prevent accidental loss, and DNA was eluted in 40 µl of nuclease-free water.

Diagnostic PCR assay of GVCV.

To check for GVCV in grape aphid communities and single aphids and to ensure that high-quality DNA was extracted, a duplex PCR was used. The first set of primers, 5044F and 5387R (Supplementary Table S2), amplified a 344-bp fragment within the ORF III region of GVCV. The second set of primers, EFF and EFR (Supplementary Table S2), amplified a 200-bp fragment within the grape aphid elongation factor 1‐α (EF1) (GenBank accession KC897260) gene to verify the quality of the extracted DNA and eliminate false-negative results. When testing grape aphids using PCR, 1 µl of undiluted DNA from communities and 50 ng of DNA from single grape aphids were used as template with 1.25 units of GoTaq DNA polymerase (Promega Corporation, Madison, WI), 1× buffer, 0.2 mM of dNTPs, 0.4 mM of GVCV primers, 0.07 mM of EF primers, and autoclaved water to 25 µl. The thermocycling conditions were as follows: initial denaturation at 95°C for 1 min, followed by 40 cycles of 95°C for 30 s, 58°C for 30 s, and 72°C for 30 s, then a final extension of 72°C for 10 min.

Plant samples were tested for GVCV with primers 5044F and 5387R. To ensure quality DNA from the plant samples, primers 16SF and 16SR (Supplementary Table S2) were used to amplify a 105-bp fragment of the 16S ribosomal RNA gene of Vitis spp. The thermocycling conditions were as follows: 1 min of initial denaturation at 95°C, followed by 35 cycles of 95°C for 15 s, 57°C for 15 s, and 72°C for 25 s, and then a final extension of 72°C for 5 min.

Purification and sequencing of GVCV ORF II.

Grape aphid and plant tissue samples positive for GVCV were tested separately by PCR with primer pairs 877F and 1866R or 963F and 1634R that flank the ORF II region of the virus (Supplementary Table S2). The amplified DNA fragments were removed from a 1% agarose gel after electrophoresis and purified through a Wizard SV Gel and PCR Clean-Up System (Promega Corporation). The resulting DNA concentration was measured using a NanoDrop 1000 instrument (Thermo Fisher Scientific). For DNA samples with low concentrations, we applied nested PCR with primer set 963F and 1866R using 1 µl of the initial 990-bp PCR product with 0.5 units of Platinum Taq DNA polymerase high fidelity (Thermo Fisher Scientific), 1× buffer, 2 mM of MgSO₄, 0.2 mM of dNTPs, 0.2 mM of primers, and water to 25 µl. The thermocycling conditions for the nested PCR reaction were as follows: initial denaturation of 94°C for 1 min, 40 cycles of 94°C for 15 s, 61°C for 30 s, and 68°C for 1 min, and then 7 min at 68°C.

Two picomoles of the appropriate forward and reverse primers was added to 20 ng of each purified DNA and sent to the Nevada Genomics Center at the University of Nevada for Sanger sequencing. The resulting sequencing chromatograms were loaded into CodonCode Aligner (CodonCode Corporation) and aligned using the GVCV-CHA reference genome. Sequences consisting of nucleotides with Phred quality scores of ≥20 were used for phylogenetic analysis.

Cladogram and matrix construction.

A total of 174 GVCV ORF II sequences derived from wild and cultivated grapes, A. cordata, and grape aphids were imported into MEGA X software (Kumar et al. 2018) and aligned using the default settings in the MUSCLE algorithm (Edgar 2004). A maximum likelihood goodness-of-fit test using 24 different substitution models was used on the aligned sequences. The GTR + G + I substitution model was found to have the lowest Bayesian information criterion score and was chosen. An analysis of the 174 aligned sequences was conducted using the maximum likelihood method with a bootstrap value of 1,000. Initial tree(s) for the heuristic search were obtained automatically by applying neighbor-join and BioNJ algorithms to a matrix of pairwise distances estimated using the maximum composite likelihood approach, and then the topology with superior log likelihood value was selected. To model evolutionary rate differences among sites, a discrete gamma distribution was used (five categories [+G, parameter = 0.6158]). The rate variation model allowed a portion of sites to be evolutionarily invariable ([+I], 42.60% sites). Codon positions included were first, second, third, and noncoding. There was a total of 394 positions in the final dataset.

Using 25 GVCV ORF II sequences from grape aphids and nine sequences from their infected host plants, a percent identity matrix was created using Clustal2.1. This matrix was imported into Excel to generate Figure 1.

Download as PowerPoint

Statistical tests of population differentiation.

Statistical tests of genetic differentiation were conducted on subpopulations according to three host plants and aphid, to two geographical locations and to the three groups from the phylogenetic analysis. Because the 174 GVCV ORF II sequences were not of the same length, they were aligned in MEGA X (Kumar et al. 2018). The alignment file was then imported to DnaSP version 6 (Rozas et al. 2017). The statistical parameters Ks, Kst, Z, and Fst were calculated and permutation tests with 1,000 replicates each were used to determine their P values. Explanations of Ks, Kst, Z, and 5Fst and interpretations of the null hypothesis and significance of each test followed a previous publication (Tsompana et al. 2005).

Quantitative PCR assay.

To follow the guidelines and fulfill the minimum information for publication of quantitative real-time PCR experiments (Bustin et al. 2009), we performed the following assays to optimize quantitative PCR (qPCR): PCR efficiency, linear dynamic range, limit of detection, and precision. Two primer sites were used, one in the intergenic region and one in ORF I, which are conserved among all seven GVCV genomes sequenced (Supplementary Table S2).

The qPCR assay was performed on a Stratagene MX3005p Real-Time PCR System (Agilent Technologies Inc.). Five microliters of DNA was added as template to 25-µl reactions that contained Brilliant SYBR Green QPCR Master Mix, ROX (Agilent Technologies Inc.), nuclease-free water, and 400 nM of both the forward and reverse primers. The standard curve and each sample were processed in three technical replicates. The thermocycling parameters were as follows: an initial denaturation of 95°C for 10 min, followed by 45 cycles of 95°C for 30 s and 62°C for 1 min; finally, the dissociation curve ran at 95°C for 1 min, 55°C for 30 s, and 95°C for 30 s. Fluorescence intensity of both ROX and SYBR was measured at the end of each extension cycle, and at 1°C intervals during the ramp from 55 to 95°C during the dissociation curve. Cq values were determined by MxPro software (Agilent Technologies Inc.) using the standard curve method. The slope of the standard curve was used to calculate the copy number, efficiency, and R² value. Identity of qPCR amplified DNA was verified to be GVCV specific by Sanger sequencing.

Results

GVCV is prevalent in aphid communities and single aphids.

Knowing the prevalence of viruses in insect vectors in an agroecological system will help manage a viral disease (Malmstrom et al. 2011). After grape aphids were found to be an insect vector of GVCV (Petersen et al. 2019), an immediate question we asked is whether grape aphids in native habitats carry GVCV. Since we did not know the sensitivity of detecting GVCV in aphids by PCR, we extracted DNA from a pool of 10 aphids that were randomly sampled from a single community and subjected the DNA to PCR (Table 1). We analyzed a total of 105 communities and 91 (87%) tested positive for GVCV. We isolated PCR-amplified DNA fragments covering the ORF II from representative communities and sequenced the DNA fragment using Sanger’s method. Chromatography showed mixed peaks at some bases (data not shown), indicating that diverse GVCV variants exist in discrete aphid communities.

Table 1. Prevalence of Grapevine vein clearing virus (GVCV) in 105 communities of grape aphids (Aphis illinoisensis) that were collected from native hosts in six locations across the U.S. state of Missouri in 2018

Sample location (Missouri town)	Samples (n)	GVCV positive (n)	Percentage
Springfield	52	47	90
Battlefield	26	23	88
Willard	2	2	100
Saddlebrook	2	2	100
Mountain Grove	1	1	100
Plattin	22	16	73
Total	105	91	87

View as image HTML

We then investigated the prevalence of GVCV among native individual grape aphids; these aphids were collected from nine locations in Missouri (Supplementary Fig. S1; Supplementary Table S1). We used Cochran’s equation to determine that 385 samples were needed, given a large unknown population size, an unknown population proportion, and a 95% confidence level with a ±5% margin of error (Israel 1992). GVCV was detected in 212 of 525 single aphids, a 40% incidence (Table 2). Therefore, the 40% incidence is at a 95% confidence level.

Table 2. Incidence of Grapevine vein clearing virus (GVCV) in single grape aphids (Aphis illinoisensis) on native and cultivated Vitaceae that were collected across the U.S. state of Missouri in 2018 and 2019

Sample location (Missouri town)a	Samples (n)	GVCV positive (n)	Percentage
Springfield	192	80	42
Battlefield	54	20	37
Willard	10	3	30
Saddlebrook	1	1	100
Mountain Grove	6	3	50
Plattin	107	39	36
Hermann	14	7	50
Augusta	30	10	33
Coffman	111	49	44
Total	525b	212	40

^a Bold denotes locations where aphids were collected within a radius of 500 m around a vineyard.

^b Total consists of 493 aphids from native Vitis and Ampelopsis plants and 32 aphids from cultivated grapevines.

View as image HTML

To investigate the prevalence of GVCV in grape aphids near a commercial vineyard block, we assayed 111 aphids and detected GVCV in 49 of them, a 44% incidence (Table 2). In comparison, we assayed 192 single aphids around Springfield, Missouri, where there are no nearby vineyards. GVCV was detected in 80 aphids, a 42% incidence. Thus, incidence of GVCV is similar between grape aphids near or far from commercial vineyard blocks based on populations at two sites.

GVCV isolates are diverse among grape aphids.

Nucleotide sequences of ORF II are the most variable among GVCV genomes, and they are thus used as criteria for delineating GVCV variants (Beach et al. 2017). We found that if the ORF II nucleotides are identical, the entire genome of the two GVCV isolates is identical (Petersen et al. 2019). Therefore, ORF II is a good candidate for differentiating GVCV variants. We acquired 25 ORF II sequences of GVCV isolates from grape aphids and nine ORF II sequences of GVCV isolates in their host plants. We then generated a percent identity matrix to compare the 34 ORF II sequences. Identity of the GVCV ORF II sequences ranged from 85.42 to 100% (Fig. 1).

ORF II sequences of aphid GVCV isolates with ≥99% identity are shown in green in Figure 1. Fifteen GVCV isolates from aphids are arranged into five distinct groups (numbered; Fig. 1; Table 3; Supplementary Fig. S1). In group 1, four aphids and their host A. cordata all contained the same GVCV variant (Fig. 1, lines 3 to 7). Group 2 shows the same GVCV variant in two aphids (Fig. 1, lines 9 and 10), one from a GVCV-positive A. cordata and the other from a GVCV-negative Vitis host, 288 km apart (Table 3). The GVCV variant in group 3 was found in aphids colonizing both Vitis plants and A. cordata. Four were collected proximally to each other (Fig. 1, lines 11 to 14); however, one was found 277 km away (Fig. 1, line 17; Table 3). In group 4, the same GVCV variant is from two grape aphids collected from GVCV-positive A. cordata 280 km apart (Fig. 1, lines 20 and 22; Table 3). The GVCV sequence from one of the A. cordata matched that of the GVCV isolated from the two aphids (Fig. 1, line 21), whereas the GVCV from the second A. cordata had only 90 to 91% identity with GVCV from the two aphid samples (Fig. 1, line 7, 19Amp006). In group 5, the two aphids carried the same variant that were collected from GVCV-negative A. cordata and Vitis hosts separated by 280 km (Fig. 1, lines 27 and 28; Table 3).

Table 3. Five groups of Grapevine vein clearing virus (GVCV) variants in grape aphids (Aphis illinoisensis) that share >99% identical nucleotides in open reading frame (ORF) II, location collected, and status of GVCV in host plants Ampelopsis cordata and Vitis spp.

Groupa	GVCV isolateb	Town of Missouri	Distance (km)	Host plant	GVCV in host plant
1	19Aph018-12	Coffman	0	19Amp006	+
	19Aph018-13	Coffman		19Amp006	+
	19Aph018-19	Coffman		19Amp006	+
	19Aph018-25	Coffman		19Amp006	+
2	18Aph032-4	Plattin	288.8	18Amp016	+
	18Aph076-5	Battlefield		18Vit056	−
3	18Aph011-1	Springfield	277.8	18Vit023	−
	18Aph018-1	Springfield		18Vit030	−
	18Aphs055c	Springfield		18Vit053	+
	18Aph015-7	Springfield		18Amp010	+
	18Aph026-1	Plattin		18Vit038	−
4	19Aph018-21	Coffman	280.7	19Amp006	+
	18Aph105-7	Springfield		18Amp031	+
5	18Aph46-7	Battlefield	283.1	18Amp21-5	−
	18Aph23-3	Plattin		18Vit035	−

^a Locations of the five groups are presented in Supplementary Figure S1.

^b Sample codes that begin with 18 and 19 denote the years 2018 and 2019, respectively. Numerals after the genus abbreviations indicate the series of sample numbers. Aph = Aphis illinoisensis, Amp = host plant Ampelopsis cordata, and Vit = Vitis spp.

^c Indicates that the ORF II sequence was acquired from a community of aphids.

View as image HTML

Sixteen ORF II sequences were derived from GVCV isolates in grape aphids whose host Vitaceae was GVCV positive (blue or underlined, Fig. 1). In nine cases, the ORF II was identical between GVCV isolates from both aphids and host plants (underlined, Fig. 1). In seven cases, the sequence was not identical between GVCV isolates from aphids and their host plants (blue, Fig. 1). These results indicate that grape aphids probably acquired GVCV from multiple source plants or the GVCV population in the host plants consisted of a mixture of GVCV variants.

In five cases, diverse GVCV variants were found within the same aphid community. In one community, four sequences were identical to each other and the host plant, but GVCV isolate 19Aph18-21, in the same community, shared only 90.53% identical nucleotides (red, Fig. 1). Interestingly, GVCV isolate 19Aph18-21 is identical to 18Aph105-7, which was collected in 2018, and the two collection sites are >270 km apart (group 4, Table 3).

In 11 cases, the ORF IIs of GVCV isolates in grape aphids were ≥99% identical to those of GVCV isolates in plants other than those from which the aphids were collected (italicized, Fig. 1). The GVCV-infected Vitaceae are 18Vit054, 18Vit053, 18Amp01, and 18Amp031, and these plants are each >270 km from aphids carrying the same variant.

This detailed analysis shows that aphids carry a diverse array of GVCV variants, with aphids feeding side by side on host plants carrying the same variant in a few instances, but different variants in most cases.

Aphid GVCV variants disperse among host plants.

To reveal the genetic structure of GVCV populations, we first analyzed the genetic relationship among 149 GVCV isolates from Vitaceae. One key finding is that three GVCV isolates from host plants shared 100% identity, but they were collected from regions that are 290 to 460 km apart. The presence of the same GVCV variant in two host plant species at different locations indicates possible long-distance transmission. We then constructed a cladogram of a total of 174 GVCV isolates by adding 25 GVCV ORF II sequences from grape aphids to the 149 GVCV ORF II sequences from wild Vitis, A. cordata, and cultivated grapevines (Supplementary Fig. S2). When all 174 sequences were analyzed using the maximum likelihood method with a 95% bootstrap value, they were not distinctly resolved (Supplementary Fig. S2). We then conducted statistical analysis of genetic differentiation of the 174 ORF II sequences according to host species and two geographical regions. The Ks, Kst, and Z test statistics for genetic differentiation and the Fst values for estimating the extent of genetic differentiation did not support significant genetic differentiation of the subpopulations according to host species and aphid (Supplementary Table S3). The Fst value for two geographical populations was 0.08, indicating a moderate genetic differentiation (Supplementary Table S3). Therefore, both analyses suggested that GVCV variants cannot be genetically differentiated by hosts and aphid based on the ORF II sequences. Dispersal of diverse GVCV isolates by aphid and rapid changes of ORF II sequences might be the two major factors that make the genetic differentiation of GVCV variants undiscernible.

There were three noticeable groupings in the phylogenetic tree (boxed in Supplementary Fig. S2). Each group includes GVCV isolates from aphids and from cultivated grapevines or wild Vitaceae. The Ks, Kst, and Z tests indicated that there is significant genetic differentiation among three groups (Supplementary Table S3). The values of Fst in pairwise comparison of ORF II sequences were 0.48, 0.62, and 0.68, respectively, between groups A, B, and C (Supplementary Table S3). This statistical analysis supports three defined groupings of GVCV isolates based on the ORF II sequences. These three groups that include GVCV isolates of nearly identical ORF II sequences provide a glimpse of GVCV spread in an agroecosystem (Fig. 2).

Download as PowerPoint

Group A (Fig. 2A) includes one GVCV isolate (18Aph023-2) from aphid, one isolate (VRU1) from wild Vitis, and six isolates from A. cordata. These isolates were not only spatially separated but were also collected 3 years apart (searchable in Supplementary Fig. S1). Group B (Fig. 2B) has nine GVCV isolates from grape aphids, wild Vitis plants, and cultivated grapevines. In one lineage of GVCV isolates with 99.9% identity, three are from grape aphids in one location; one was sampled about 300 km away; two are from Vitis plants in one location. Two identical GVCV isolates, GVCV-CHA and VitCVV5IIIa, are from cultivated grapevines at a distance of >200 km. The reference GVCV-CHA variant shares >99.47% identical nucleotides to isolates from grape aphids and wild Vitis. Group C (Fig. 2C) contains 11 GVCV isolates that are from grape aphids, wild Vitis, and A. cordata. The distance at which these samples were collected ranges from <5 to >300 km. These results suggest that grape aphids carry diverse GVCV isolates and likely transmit them among wild plants and cultivated grapevines at different locations across Missouri.

GVCV genome number varies largely in individual grape aphids.

Viral genome quantity in insect vectors has been measured in groups or individual aphids by qPCR (Khelifa 2019; Liu et al. 2019), but those experiments were done under greenhouse conditions. We applied qPCR to measure GVCV genomes in the stylet and body of 20 individual aphids from natural populations on the Vitaceae plants in their native habitats. The stylets contained 14 to 260,571 GVCV viral genomes, and the bodies contained 136 to 1,713,143 GVCV genomes (Table 4).

Table 4. Genome number of Grapevine vein clearing virus (GVCV) in the stylet and body of 20 single aphids measured by quantitative PCR

Sample numbera	Genome numberb
	Stylet		Body
	Actual	Extrapolated	Actual	Extrapolated
1	7	43	107	609
2	23	131	124	707
3c	74	422	89	510
4	29	163	137	783
5	2.46*	14.06*	97	557
6c	7	41	383	2,188
7c	11	61	234	1,338
8	127	723	23,380	133,600
9	8,993	51,389	299,800	1,713,143
10	45,600	260,571	85,620	489,257
11	13	134	71	705
12	25	245	225	2,245
13	76	762	26	258
14	8	80	37	372
15	47	467	83	825
16	14	141	1.05*	10.5*
17	67	666	113	1,127
18	66	664	23	233
19	7	74	32	323
20	20	197	14	136
Average	2,906	16,683	21,610	123,627

^a Bold denotes samples in which the head was removed from the body.

^b The average of three replicated reactions from each sample was used to determine initial template quantity. The initial quantity was transformed to the extrapolated GVCV genome number by accounting for the total amount of DNA extracted and the efficiency of the DNA extraction kit. An asterisk indicates that the GVCV genome number is below the threshold of detection.

^c Winged aphids.

View as image HTML

Two of these aphids contained very high GVCV genome numbers in their stylets, ranging from 51,388 in aphid sample 9 to 260,571 in sample 10. Interestingly, both aphids were wingless. In contrast, the number of GVCV genomes was low in two of the three stylets of the winged aphids. In the majority of aphids (16 of 20), the body contained more GVCV genomes than the stylet.

Discussion

In this study, we found a 87% prevalence of GVCV in grape aphid communities (Table 1) and a 40% incidence among individual grape aphids of the natural population (Table 2). We noticed that prevalence of GVCV is higher among communities than in single aphids, which is logical considering that even one GVCV-bearing aphid would deem the whole community as positive. High prevalence of virus in a transmission vector has been found in another virus-vector-perennial plant system. Citrus tristeza virus was found in 35.4% of A. gossypii and in 28.8% of A. spiraecola in Morocco, which creates a high inoculum pressure for virus spread (Elhaddad et al. 2016). However, the presence of GVCV in aphids is merely an indicator of ingestion or acquisition of the virus. It does not indicate that GVCV will be transmitted to plants. In other words, not every GVCV-bearing aphid will be able to transmit the virus to a host. Despite this fact, prevalence of GVCV in grape aphids constitutes a mobile reservoir of the virus and increases the risk for its spread.

GVCV is present in 34% of A. cordata sampled (Petersen et al. 2019), in 10% of wild Vitis plants sampled (Beach et al. 2017), and in 8% of cultivated grapevines sampled (J. Schoelz, personal communication), suggesting that there is a gradient decrease of GVCV incidence from A. cordata to grape cultivars. One explanation is that grape aphids prefer to feed on A. cordata rather than on Vitis plants. It is proposed that an intricate relationship exists between aphids and virus-infected plants (Donnelly et al. 2019). Another factor is that more A. cordata are available for aphids to feed on, since A. cordata has a higher density than Vitis plants in native flora. It is also possible that A. cordata is more susceptible to GVCV than Vitis spp. The lower prevalence of GVCV in vineyards could be explained by the fact that seasonal management of insect pests in vineyards substantially reduces aphid population size compared with the pristine native habitats, thus reducing opportunities of GVCV spread and infection in vineyards.

Among the 25 GVCV isolates in aphids whose ORF IIs are sequenced, only five share identical ORF IIs. The remaining 20 GVCV isolates share 85 to 99% identical nucleotides of ORF II. In six cases, the same GVCV variants in the aphids are also present in the plants from which the aphids were collected (Table 3), indicating that the aphids acquired GVCV from their host plant. In other cases, however, GVCV was detected in aphids collected from plants that were negative for GVCV (Fig. 1; Table 3). Our data also suggest that GVCV isolates are genetically more diverse among grape aphids on GVCV-negative plants than on GVCV-infected plants (Fig. 1). Thus, it is clear that genetically diverse GVCV variants are acquired and retained in aphids that are widely dispersed across Missouri. The identity and genetic structure analyses showed that the same GVCV variants were found in grape aphids and wild and cultivated Vitaceae that are geographically separated (Figs. 1 and 2; Table 3; Supplementary Fig. S1), indicating that aphids can transmit GVCV from wild hosts to cultivated grapevines. Our analysis of the 174 GVCV isolates using ORF II genetic information only allows us to infer genetic structure at spatial scale, not at temporal scale. Sampling of grape aphids over multiple years and analyzing GVCV genetic composition in viruliferous aphids will reveal temporal dynamics of GVCV populations.

GVCV was detected in both the stylet and body of single aphids. Overall, more GVCV genomes were detected in the body than in the stylet (Table 4). In samples where the head was included with the body, there was a 53% increase in the average number of GVCV genomes after outliers were removed (Table 4). It is still not clear whether the epidemiological significance of GVCV virions is in the head or if GVCV virions concentrate in the stylet tip. Although GVCV is present in the body, it cannot be discerned if GVCV only exists in the gut and/or in the hemolymph. It was reported that Rubus yellow net virus, a closely related badnavirus, was detected in heads and bodies of aphids and in headless bodies during the transmission access period, suggesting that Rubus yellow net virus is transmitted likely in a semipersistent mode (Jones et al. 2002). Presence of GVCV in both the stylet and body may indicate a semipersistent or a circulative, nonpropagative transmission mode. More research is required to distinguish the transmission mode.

The number of GVCV genomes varies largely among single grape aphids. A similar finding was also documented in the Potato virus Y-aphid (Khelifa 2019), Plum pox virus-aphid (Olmos et al. 2005), and Citrus tristeza virus-aphid (Liu et al. 2019) systems. Vast variation likely is attributable to the following factors: aphids acquire GVCV virions from plants that have different virus titers, acquisition access periods differ among aphids, and the developmental stage of aphids affects acquisition and retention of GVCV virions. The largely variable genome numbers in single aphids propelled us to ask questions that have not been raised before: Why does the body of aphid contain such high viral copies, and do they contribute to transmission and how? The epidemiological significance of these discoveries merits further investigations on what is occurring in an agroecosystem. This vast variation in GVCV genome numbers has motivated us to collect more grape aphid colonies from native habitats and investigate the transmission rate of GVCV from native grape aphids to cultivated grapes.

Wild susceptible Vitis species and A. cordata are ubiquitous along the edge of vineyard blocks in the Midwest. A daunting challenge for preventing GVCV infection is reservoirs of GVCV in native wild plants. The presence of GVCV in grape aphid populations also forms a moving reservoir. At the beginning of each growing season, newly hatched aphids, winged and wingless, may be blown long distances over native Vitis species and A. cordata or migrate among these wild plants at short distances. They may feed on GVCV-infected native plants in massive populations. Assisted by ever-changing wind strength and directions, aphids migrate from plant to plant locally, too. Aphids’ flight directions and distances are drastically influenced by the wind. They may carry GVCV variants from original sources and alight randomly on hosts in their native habitats and in vineyards. Wide dispersal and long-distance movement of aphids creates endless chances for viruses to spread into cultivated crops (Fereres et al. 2017). Abundant aphids and wide dispersal provide ambient opportunities for vectors to spread diverse GVCV variants. Furthermore, annual cycles of dormancy and new growth maintain continuous pathways for multiple variants of genetically diverse GVCV to spread through wild natural corridors of Vitaceae habitats to cultivated grapevines, and back to native flora, although less frequently. The seasonal dynamic flow of GVCV via grape aphids forms a channel for GVCV persistence among its hosts. Therefore, multiple sources of GVCV and migrating vectors create opportunities for numerous spillovers of GVCV variants into vineyards during each growing season.

This study reveals the prevalence and genetic and quantitative variabilities of GVCV variants among grape aphids. Virus epidemics on a crop are largely influenced by dispersal pattern, population density and dynamics, transmission mode and efficiency, and probing and feeding behaviors of insect vectors (Pagán 2018; Picard et al. 2017). Complex interactions of multiple factors such as these determine the outcome of disease incidence (Jeger et al. 2018). It is very hard to curtail aphid migration. The best management of a viral disease on grapevines is to implement preventative strategies by planting virus-tested clean grapevines in a newly planted vineyard (Golino et al. 2017). Given that wild Vitis species and A. cordata contain 10 and 34% GVCV infection in their native habitats, removing wild Vitaceae plants around vineyards or before planting a new vineyard is a good practice that can impede and reduce the movement of grape aphids among a wild host reservoir and cultivated hosts. GenBank accession numbers of the 133 ORF II sequences from this study are MT515625 to MT515705, MT536146 to MT536154, MT725446 to MT725449, MT597428, MT587668 to MT587677, MT474151 to MT474157, and MW219113 to MW219135.

Acknowledgments

We are grateful to Dr. James Schoelz for critical review and editing of the manuscript. We thank Hunter Blalock for assisting us in diagnostic analysis and submitting the ORF II sequences.