Occurrence of cherry viruses in South Tyrol (Italy) by comparing growth periods in two consecutive years

Summary. Sweet cherries ( Prunus avium L.) are important as fruit crops, and can be affected by numerous viruses. An investigation on the occurrence of the three most common viruses of sweet cherry was carried out in commercially managed orchards in South Tyrol (Italy). The incidence of apple chlorotic leaf spot virus (ACLSV), Prunus necrotic ringspot virus (PNRSV) and prune dwarf virus (PDV) was investigated using enzyme-linked immunosorbent assays (ELISA) and the reverse transcriptase-polymerase chain reaction technique (RT-PCR) in spring 2018 and 2020, and during the summer and autumn of 2020. All three viruses were detected in the surveyed orchards. Comparative analyses showed that detection was more effective with RT-PCR than with ELISA, especially for detecting PNRSV and PDV. Mixed infections were detected in all the surveyed orchards. The results also showed clear differences between and during host growth periods, likely due to a variable virus concentration in the host trees.


INTRODUCTION
Italy has among the worlds' greatest growing areas of sweet cherries (Prunus avium L.). In 2018, this country ranked fifth, with a growing area of 29,010 ha, after Syria Arab Republic (30,317 ha), United States of America (34,400 ha), Chile (39,645 ha), and Turkey (82,729 ha) (FAO, 2022). Almost 90% of Italian sweet cherries are cultivated in the regions of Puglia, Campania, Emilia Romagna, and Veneto. However, sweet cherry production does not have particular temperature requirements for fruit ripening, so cultivation in altitudes up to 1,300 m a.s.l. is possible (Zago, 2003;Gamper, 2010). Hence, cherry cultivation can be an economically important supplementary crop for small farmsteads in Alpine regions.
New cherry plantings require substantial investments for protective nets against hail, birds, and insects, as well as for rain covers (Gamper, 2013). Nevertheless, good returns can be achieved for premium quality cherries produced in South Tyrol. These result from favourable market situations in late summer, when cherries grown in warmer cultivation areas are no longer being offered, and growers from the alpine areas can supply markets at profitable prices (Zago and Ropelato, 2009;Gamper, 2010). Therefore, cherry production has increased in the last decade in South Tyrol, the northernmost province of Italy (Pirazzoli and Palmieri, 2019), and is currently producing sweet cherries on approx. 100 ha (Catalano, 2013;Martini, 2021).
ACLSV is a member of Trichovirus (Betaflexiviridae, Martelli et al., 1994;Adams et al., 2004;Watpade et al., 2013), and is known to infect pome and stone fruits. The concentration of ACLSV in infected trees is low and irregular, hence reliable detection is difficult (Polák and Svoboda, 2006;Katsiani et al., 2014). Generally, ACLSVinfected trees are symptomless, but can be responsible for cherry necrosis (Liu et al., 2014), fruit deformation and discoloration, as well as graft incompatibilities in nurseries (Rana et al., 2011;Sánchez et al., 2015). However, symptoms and severity depend largely on the infected Prunus species and/or the infecting virus strain (Osman et al., 2016;Rubio et al., 2017). PNRSV belongs to Ilarvirus (Bromoviridae), and may cause infections of all Prunus species (Mekuria et al., 2003;Fiore et al., 2008). Cherry trees infected with PNRSV can exhibit a wide range of symptoms, which may include foliar mosaics, ring-shaped or spotted chlorotic areas, but symptoms may also be latent (Oliver et al., 2009;Sánchez et al., 2015). Furthermore, infections with PNRSV may lead to reduced numbers of flower buds, which leads to yield losses of 20 to 56% (Wang et al., 2018), and culminates in death of infected trees (Song et al., 2013). Like PNRSV, PDV belongs to Ilarvirus (Bromoviridae) (Öztürk and Çevik, 2015), and can infect different Prunus spp. (Kozieł et al., 2020). Depending on environmental conditions, virus strain, and plant host, PDVinfected trees often remain symptomless. This virus may cause chlorosis, yellowing, mosaic, ringspot, necrosis, and malformations on leaves, as well as reduction numbers of fruits (Predajňa et al., 2017). Viruses such as ACLSV, PNRSV, and PDV can occur as mixed infections in cherry orchards, and may result in decline, low yields, and unusual fruit disorders, with crop losses of up to 57% (Mandic et al., 2007;Yardimci and Culal-Klllc, 2011). ACLSV is only known to be transmitted by grafting and other vegetative propagation techniques. In contrast, PNRSV and PDV can also be transmitted horizontally in pollen and vertically in seeds (Rubio et al., 2017).
Phytopathological alterations observed in many Italian cherry cultivation areas in the last two decades have led to detailed studies of the presence of fruit tree viruses (Aparicio et al., 1999;Myrta et al., 2003;Matic et al., 2007;Babini et al., 2014). None of these studies included cherry plantings in the South Tyrol province, there is no information on the virus phytosanitary status of this small cultivation area. However, due to the expansion of cherry cultivation and the associated high investment costs in this mountainous area, it is important to obtain information on virus occurrence, to avoid yield losses caused by viruses. Reliable diagnoses of these viruses are also important for maintaining sustainable plant production. This will assist in selection of pathogen-free propagative plant material in nurseries and in orchards used for commercial production.
Routine virus diagnoses in cherry trees are based on enzyme-linked immunosorbent assays (ELISA) and the reverse transcriptase-polymerase chain reaction (RT-PCR) (Barba et al., 2015). There are issues with both of these methods that have been outlined elsewhere (Mekuria et al., 2003;Noorani et al., 2013;Hu et al., 2014;Rubio et al., 2017). Furthermore, the seasonal dynamics of virus concentrations throughout growing seasons may affect assay reliability (Tsai et al., 2012).
The aims of the present study were: (a) to obtain an overview of virus incidence in commercially managed cherry orchards in South Tyrol; (b) to compare ELISA and RT-PCR for routine diagnoses in cherry virus detection; and (c) to assess annual and seasonal fluctuations of concentrations of ACLSV, PNRSV, and PDV in South Tyrolean sweet cherry orchards.

Sampling procedure
Eight commercially managed orchards were selected in the Val Venosta/Vinschgau valley (South Tyrol, Italy) ( Table 1). The first sampling was carried out on each site in spring 2018, during the full bloom host stage (end of April until early May). Flowers were manually sampled from 30 randomly selected trees at each site (n = 240). All samples were tested immediately after sampling for the presence of three viruses (ACLSV, PNRSV, and PDV) using ELISA (Table S2). Nine trees per orchard, including those which were positive in the ELISA results 2018 (except for four trees tested positive for one of the viruses had to be replaced by negative trees, due to agronomical reasons: one tree tested positive for PDV in each of orchards A and D, and one tree tested positive for ACLSV in each of orchards D and F), were used for RNA extractions and were again sampled in spring of 2020 during full bloom at the end of April until early May, depending on the sampling site. To determine seasonal dynamics of the virus distribution in commercially managed orchards, the same nine trees of the field sites D and F, which were already tested in spring (2018 and 2020), were sampled again in summer (August) and autumn (October) 2020.
All samples (flowers, buds, and leaves) were taken randomly from different branches at different heights all around each assessed tree. Each individual sample consisted of either ten flowers in spring 2018/2020 or ten buds and ten leaf discs in summer and autumn 2020. Each sample was pooled, placed in a plastic bag, put into a cooling box, and brought to the laboratory for further processing.

Laboratory analyses
Enzyme-linked immunosorbent assays (ELISA). 0.5 g of single petals per flower were homogenised together with 4.75 mL of extraction buffer in a universal bag (Bioreba AG,) using an automatic homogenizer (Bioreba AG). The ELISA tests were carried out according to the manufacturer's instructions, with the commercially available Double Antibody Sandwich Assay (DAS-ELI-SA) (LOEWE® Biochemica GmbH). Absorbance values were detected at 405 nm with FLUOstar OPTIMA microplate reader (BMG Labtech). Data analyses were carried out according to the Bioreba AG technical information (BIOREBA, 2011). The cut-off was set-up individually for each plate. Absorbance values were sorted in ascending order and a histogram for each plate was created. In the resulting histograms, negative and background values could be easily distinguished from potential positive values, which were characterized by an abrupt increase in the OD value. The mean value and the standard deviation calculated from values before this abrupt increase were used to calculate the cut-off as: Total RNA isolations Immediately after sampling, petals (spring 2018 and 2020) or buds together with leaf discs (August 2020 and October 2020) were cut into pieces with a scalpel and mixed homogeneously. Approx. 0.2 g of the blended tissues was homogenized using a TissueLyser II (Qiagen), for which adaptors were pre-cooled in liquid nitrogen. RNA was extracted with RNeasy Plant Mini Kits (Qiagen) according to manufacturer's instruction. Extraction vials were cooled on ice between the extraction steps.

RT-PCR assays
One-step multiplex RT-PCR was carried out using the SuperScript™ III One-Step RT-PCR System with the Platinum™ Taq DNA Polymerase kit (Invitrogen by Life Technologies), and the primer pairs ACLSV_s/ACLSV_a, PNRSV_s/PNRSV_a, and PDV_s/PDV_a (Sanchez-Navarro et al., 2005), at final concentration of 0.25 pmol µL -1 for each primer. A region of the chloroplast gene rbcL, which encodes the large subunit of ribulose bisphosphate carboxylase, was used as internal control. For the rbcL gene, primers Rbcl_s/Rbcl_a (Sanchez-Navarro et al., 2005) were used, at final concentrations of 0.05 pmol µL -1 . Thermal cycling (Sanchez-Navarro et al., 2005) was: an initial cycle of 50°C for 30 min for cDNA synthesis, followed by a denaturation step of the RT enzyme at 94°C for 2 min, 40 cycles each at 94°C for 15 s, 50°C for 30 s, 68°C for 1 min, and a final incubation at 68°C for 7 min. The amplified PCR products were separated on 2% agarose gels stained with GelRed® Nucleic Acid Gel Stain (Biotium Inc.), and were visually checked using Chemi-Doc™ MP with Image Lab™ v.4.0.1 (Bio-Rad Laboratories Inc.). On each gel, an artificial ladder was included, made from a mixture of amplicons of each target. This mix was prepared by cDNA amplification of reference gene fragments for ACLSV, PDV, PNRSV, and rbcL, with the same primers used for the RT-PCR from formerly tested positive cherry petals ( Figure S1). Amplicons were cloned into the pJET1.2/blunt Cloning Vector (Thermo Fisher Scientific s.p.a.) according to the manufacturer's instructions. Inserts were amplified and sequenced with vector-specific forward and reverse primers to confirm amplification of the gene fragments by LGC Biosearch Technologies. The quality of the sequencing data was controlled with the software Geneious v.11.1.5 (Biomatters Ltd.), and sequence identity was confirmed by BLASTn search. The sequences of the three viruses were deposited in the NCBI GenBank under the accession numbers OM585596 for ACLSV, OM585598 for PNRSV, and OM585597 for PDV. Appropriate amplicons were purified with the QiaQuick PCR Purification Kit according to the manufacturer's instructions (Qiagen), and were mixed at approximately equal concentrations for the internal controls.

ELISA detection of viruses in 30 randomly selected cherry plants at each orchard site
All three assessed viruses were detected in the sampled cherry orchards in spring 2018 (Table 2). ACLSV was detected in four orchards, and PDV in five orchards, while PNRSV was detected in two orchards. Only orchard H showed no virus infection. The greatest number of ACLSV-positive trees were detected in orchard D in 11 samples. However, no obvious symptoms were observed on cherry trees during the orchard monitoring, except for samples I29 and I30 which were manifesting scattered chlorotic areas on the leaves. Those trees were tested positive for ACLSV (Table S1). In general, 15 (6.3 %) out of the 240 trees tested positive by ELISA for ACLSV, two (0.8 %) for PNRSV, and 11 (4.6 %) tested positive for PDV.
The following results presented focus on selected samples (n = 9 per orchard), as described in the Material and Methods (above).

Comparison of ELISA and RT-PCR detection methods in 2018
When the results of ELISA and RT-PCR were compared, not all samples positively detected by ELISA were positive by RT-PCR, and vice versa (Tables 3 and  4). In general, ELISA was less sensitive for detection of the viruses than RT-PCR, regardless of the virus type. However, the largest difference between the results from both techniques was for PDV, followed by PNRSV and ACLSV.

Occurrence of mixed virus infections in 2018 and 2020 or both years based on RT-PCR results
Single infections with either ACLSV, PNRSV, or PDV were more predominant than mixed infections by the three viruses. (Table 5). Based on RT-PCR, the most common co-infection was by PNRSV and PDV (16.7%). Mixed infections by ACLSV and PDV were found in two samples (C18 and D08). Mixed infection by ACLSV and PNRSV was detected in only one sample (D13). Mixed infections by ACLSV, PNRSV and PDV were detected twice (A13 and H15).

Annual and seasonal fluctuations of viruses in cherry plants based on RT-PCR
Not all infections determined in 2018 were confirmed in 2020, and vice versa. ACLSV was detected in 11 identical samples of both years, PNRSV in four, and PDV in seven (Table S2). Generally, the number of positive samples during full bloom was greater in 2018 than in 2020 in all orchards (Table S2).
Detailed results of the infection rates of orchard D and F for all four sampling times (two years, three different seasons) are shown in Figure 1

DISCUSSION
This study, using both ELISA and RT-PCR, confirmed the presence of ACLSV, PNRSV, and PDV in commercial sweet cherry orchards in South Tyrol, the northernmost province of Italy (Tables 2 and 3). Previous studies have reported of several Prunus species infected with these viruses in Europe and in Italy. These viruses have also been detected in plums, apricots, almonds, peaches, and sweet cherries (Barba et al., 1985;Savino et al., 1991;Myrta et al., 2003;Myrta and Savino, 2005;Paduch-Cichal et al., 2007). Recent investigations showed that over 40% of cherries in the Emilia Romagna region of Italy tested positive for ACLSV, PNRSV, and PDV (Babini et al., 2014).
Targeted detection of plant viruses relies on diagnosis method and choosing the correct host growth stage for sampling. Routine diagnosis for virus detection in cherry fruit trees is generally performed by ELISA and RT-PCR (Barba et al., 2015). In the present study, comparison of ELISA and RT-PCR as diagnostic methods for the detection of cherry viruses resulted in considerable differences, especially for detection of PNRSV and PDV, but less for ACLSV (Table 3). These results were in line with previous studies, which have reported greater infection rates for PNRSV and PDV using RT-PCR than using ELISA (Sánchez-Navarro et al., 1998;Yardimci and Culal-Klllc, 2011;Gospodaryk et al., 2013). For example, Gospodaryk et al. (2013) reported a nearly twofold greater infection rate for PNRSV and PDV obtained by RT-PCR than by ELISA. These results suggested greater sensitivity for RT-PCR than ELISA, especially for the detection of PNRSV and PDV. Furthermore, Hadidi et al. (2011) suggested that Ilarvirus particles are unstable in ELISA buffers. However, a few samples detected by ELISA tests were not detected by RT-PCR (Table 4). Nevertheless, as extensively discussed in previous reports, ELISA remains the most widely used approach for detection of viruses in large numbers of samples (Mekuria et al., 2003;Noorani et al., 2013;Hu et al., 2014;Rubio et al., 2017). This is probably because the development of a multiplex RT-PCR system can be laborious and time consuming, but is a rapid, reliable, and cost-effective method (Wei et al., 2008). RT-PCR has been successfully used for simultaneous detection of several stone fruit viruses, even with low virus titre and in the presence of inhibitors (Bariana et al., 1994;Kummert et al., 2001). Therefore, focus in the present study was placed on the results obtained with RT-PCR.
Based on RT-PCR results of selected trees (n = 9 per orchard), PNRSV and PDV were both detected in all eight of the surveyed South Tyrolean commercial sweet cherry orchards, while ACLSV was not detected  in two of the orchards (B and F, Table 3). Considering the relatively high number of trees infected with ACLSV, especially at the sampling sites C and D, it is likely that virus-infected plant material was used when establishing these orchards (Figure 1). However, the certified status of the propagation material at purchase is known only for orchards A, F, and H. These orchards were Conformitas Agraria Communitatis (CAC) certified at the time of planting. Unlike ACLSV, PNRSV and PDV can be transmitted by grafting or vegetative propagation techniques and in pollen (Barba et al., 2015). Hence, pollen and seed transmission of PNRSV and PDV may have contributed to the occurrence of these in all the surveyed orchards. Viruses with these transmission modes have high distribution potential.
Most surveyed trees (n = 9 per orchard) were infected either with ACLSV, PNRSV, PDV, or with combinations of these viruses (Table 5). Mixed infections of these viruses were reported for cherry orchards in previous studies (Myrta et al., 2003;Yardimci and Culal-Klllc, 2011;Gospodaryk et al., 2013). In the present study, the combination of PNRSV and PDV occurred most frequently. As suggested by Gospodaryk et al. (2013), mixed infections may be the result of the different transmission modes, geographical origins, and grafting, which may contribute to wide distribution of these viruses.
In the present study, virus detection using both ELISA and RT-PCR was found to be variable, with clear differences between and during host growth periods. Detections were maximum during full flowering in spring of both years (2018 and 2020; Table S2 and Figure 1). Uneven distribution of virus in individual trees is a likely explanation for the detection discrepancies between and during growth periods. (Knapp et al., 1995;Spiegel et al., 1997;Marbot et al., 2003). The same phenomenon could also explain differences between sampling years, and influences of weather conditions are also likely to have contributed to the seasonal and annual discrepancies experienced in the present study. Honjo et al. (2020) suggested that in natural systems, the concentration of viruses within hosts may change greatly during growth periods, depending on virus replication and host growth. In the present study, the uneven virus distribution of PNRSV and PDV within trees varied particularly between the two years. For example, only three samples tested positive in spring 2018 and 2020, while for six trees a positive infection was detected at least once, either in spring 2018 or 2020. However, samples F10 (for PNRSV), and D03 and D24 (for ACLSV) showed constant infection status throughout the growth periods. It is likely that the virus concentration was high in these trees, and/or the viruses were uniformly spread throughout the host trees. To obtain reliable results and to prevent false negatives, samples consisted of pooled homogenized tissues, taken from around each tree at different heights. However, even pooling the collected samples from different parts of each same tree, may not sufficiently overcome the effect of uneven virus distribution. Further studies are required to take account of the discrepancies observed. Greater numbers of samples per tree should be examined. Although care was taken to avoid errors during processing samples, human error can never be excluded. Furthermore, no differences in virus prevalence were found for different altitudes or ages of trees, although more studies are required to confirm these results.

CONCLUSIONS
Virus infections can seriously compromise the phytosanitary status of fruit tree nurseries, and commercially managed orchards. In the present study, all three assessed viruses (ACLSV, PNRSV and PDV) were detected in cherry orchards, either by ELISA or RT-PCR. However, RT-PCR was more sensitive than ELISA for detection of these viruses, and especially for PNRSV and PDV. Virus detection was variable and was different during different host growth stages. Future research should include host sampling from different altitudes, orchard ages, host cultivars, and propagation material at time of acquisition, to better understand the observed phenomena.

ACKNOWLEDGMENTS
The South Tyrolean Extension Service for Fruit and Wine Growing, especially Gregor Trafojer and Michael Gamper, supported this study during the sampling campaigns. Dr Valeria Gualandri and Paola Bragagna of Technology Transfer Centre, Fondazione E. Mach, provided information on technical feasibility. South Tyrolean cherry growers gave excellent cooperation within this study. Dr. Markus Gorfer, Austrian Institute of Technology, provided guidance during laboratory work and in proofreading of the manuscript of this paper.