Fungal pathogens associated with black foot of grapevine in China

Copyright: © 2021 Q. Ye, W. Zhang, J. Jia, X. Li, Y. Zhou, C. Han, X. Wu, J. Yan. This is an open access, peerreviewed article published by Firenze University Press (http://www.fupress. com/pm) and distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


INTRODUCTION
Grapevine (Vitis vinifera L.) is an economically important fruit crop, with global cultivation area of 7,449,000 hectares in 2018, and China is ranked the second in the world grapevine cultivation area (2019 OIV). More than 70 diseases have been reported in grapevines, most of which are caused by fungi or oomycetes (Wilcox et al., 2006), and among these, at least 27 diseases have been reported in China. Esca complex, Botryosphaeria dieback, black foot (BF), Eutypa dieback, and Phomopsis dieback are major fungal grapevine trunk diseases (GTDs) worldwide. These diseases have been reported in almost all the main grapegrowing countries (Gramaje et al., 2018). GTDs are complexes that affect grape yields, wine quality and lifespan of plants in many grape-growing regions. The global financial losses attributed to GTDs are estimated to be more than $US 1.5 billion per year (Hofstetter et al., 2012).

Vineyard surveys
Surveys were carried out in ten vineyards, located in Ningxia, Hebei, Shanxi, Guangxi and Xinjiang provinces of China, during 2017 and 2019 ( Figure 1a). These provinces belong to the top ten grape cultivated grapevine areas in China, and Xinjiang province ranked the first, followed by Hebei province. The training systems used in the surveyed vineyards was mini "J". The vineyards were of similar age, from 5 to 6 years old. Typical symptoms associated with diseased vines were shortened shoot internodes, chlorotic leaves, and trunk and root necroses (Figure 1, b-h). Initial disease symptoms included root necroses (especially small roots). As the disease progressed, the above-ground plant parts developed shoot shortened internodes and chlorotic leaves in severe cases. Some grapevines were grafted (rootstock Fercal), and some others were self-rooted (Personal communication, some of the grape growers).

Sample collection, fungus isolation and morphology of the pathogens
Samples were collected from V. rotundifolia Michx., and V. vinifera cvs Marselan, Cabernet Franc or Cabernet Sauvignon. Typical symptoms were recorded by taking appropriate photographs. The samples were kept at 4°C for further study, and the presence of spores or structures on the surfaces of trunks or roots were detected using a microscope. Isolations were made from symptomatic trunks and roots. Necrotic root and trunk samples were debarked and cut into small pieces (4-5 mm 2 ). These small pieces were then surface-sterilized in 75% ethanol for 30 s, rinsed three times with sterilized water, dried, and cultured on potato dextrose agar (PDA; 20% potatoes, 2% dextrose, 1.5 to 2% agar) in Petri plates. The plates were incubated at 25°C. Fungi growing from tissue pieces were transferred onto new PDA plates after 7 d, and pure cultures were obtained by isolating single spores. Pure cultures were grown on PDA and malt extract agar (MEA) and incubated at 25°C in the dark for 7 d. Conidia and colonies on the MEA plates were observed and photographed using the Axio Imager Z2 photographic microscope (Carl Zeiss Microscopy).

DNA amplification and phylogenetic analyses
Single-spore purification were done for all the isolates before DNA extractions. Total genomic DNA was extracted from 50-100 mg of mycelium after 14 d of incubation on PDA (Guo et al., 2000). For initial genus identification, the internal transcribed spacer and intervening 5.8S gene regions (ITS) were amplified and sequenced for all the isolates, and the resulting sequences were searched using BLASTN within GenBank/NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi), as described by Manawasinghe et al. (2019). All the isolates in the present study belonged to Cylindrocladiella, Neonectria or Dactylonectria.
For species confirmation, phylogenetic analyses were conducted using multigene phylogenies. For Cylindrocladiella, histone H3 (his3), β-tubulin (tub2), and partial translation elongation factor 1-alpha (tef1) were sequenced (Marin-Felix et al., 2019). ITS, tub2, his3, and tef1 gene regions were sequenced for Dactylonectria and Neonectria species (Berlanas et al., 2020). The primer pairs and amplification protocols used in the present study are summarized in Table 1. Each PCR mixture comprised 1.0 µL of genomic DNA, 0.6 µL of TaKaRa ExTaq DNA polymerase, 5.0 µL of 10 × ExTaq DNA polymerase buffer, 4.0 µL of dNTPs, and 1.0 µL of each primer, and was adjusted with sterilized double-distilled water to a final volume of 50.0 µL. The PCR reactions were carried out in a thermal cycler (Bio-Rad, model C1000). Amplification products were visualized on 1% agarose electrophoresis gels under UV light using a Gel DocTM XR+ Molecular Imager (Bio-Rad). All positive bands obtained by PCR amplification were sequenced by Tsingke Company, Beijing, China, and the sequence data obtained were deposited in Gen-Bank (Table 2).
Reference sequences of related taxa were obtained from GenBank (Marin-Felix et al., 2019;Berlanas et al., 2020). The sequence data generated in the present study were included, and individual gene regions were aligned using the MAFFT v. 7 webserver (Kuraku et al., 2013;Katoh et al., 2019) (https://mafft.cbrc. jp/alignment/server/). The alignments were checked and edited manually, where necessary using BioEdit v7.0.9 (Hall, 1999). Phylogenetic trees were generated using Maximum Likelihood (ML) in RAxML (Silvestro and Michalak, 2016) and Maximum Parsimony (MP) in PAUP (v4.0) (Swofford, 2002). The ML and MP trees were constructed using the methods described by Manawasinghe et al. (2019). For MP, heuristic searches were conducted with 1000 bootstrap replicates by random addition. All characters were unordered and equally weighted. Gaps were treated as missing data, and the steepest descent option not in effect, whereas the MulTrees option was used. The Tree Length (TL), Consistency Index (CI), Retention Index (RI), Relative Consistency Index (RC), and Homoplasy Index (HI) were calculated in PAUP. All the resulting trees were saved and checked using Kishino-Hasegawa tests (Kishino and Hasegawa, 1989). The ML analyses of single genes and combined multiple genes were accomplished using the RAxML-HPC2 on XSEDE (8.2.8) in the CIPRES Science Gateway (https://www.phylo.org/ portal2/createTask!create.action). Phylogenetic trees were visualized using FigTree v1.4.4 (Rambaut, 2018) and were annotated in Microsoft PowerPoint 2016.

Pathogenicity tests
Pathogenicity tests of potential BF pathogens were conducted on detached green shoots or potted 3-month-  Neonectria coccinea  old healthy plants of grapevine cv. 'Summer Black'. Five isolates (JZB3320001, JZB3310007, JZB3310008, JZB33100011 and JZB3210004) were selected randomly for pathogenicity tests. Mycelium discs (4 mm diam.) were obtained from the edges of PDA colonies which were grown for 10 d at 25°C. Detached shoots were surface-disinfected in 75% ethanol and then dried, and each shoot was then wounded (4 mm) using a sterilized scalpel. The mycelium discs were placed onto the wound sites and covered with parafilm (Bemis). Non-colonized sterile PDA plugs were used as negative controls. The shoots were then inserted into moist soil and kept at 25°C. Each experiment included ten shoots for each fungus isolate, with a total of three parallel experiments conducted.
The lengths of the lesions were measured after 7 d, and meanwhile photos were taken.
Pathogenicity tests of BF fungal agents were further conducted on the 3-month old grapevine cuttings which were inoculated in a manner similar to the detached green shoots. The experiment was performed in six cuttings for each tested isolate and the negative controls. The plants were kept in a greenhouse maintained at 25°C, and the trial was conducted twice. Shoots were collected, and lesion lengths were measured upward and downward from the points of inoculation after 80 d.
Fungi were re-isolated from necroses on the test plants in all pathogenicity tests, and fungus identifications were based on cultural and morphological characters. The lesion dimension data were statistically ana-  lyzed with IBM SPSS Statistics 21.0 (IBM Corp.) using a one-way analysis of variance (at P = 0.05) to determine differences in shoot lesion dimensions resulting from different fungus isolate inoculations.

Fungus isolation and initial species identifications
Incidence of BF-like symptoms in the investigated vineyards was 0.1% to 1%.
Colony morphology of all the isolates distinguished after 14 d of growth on PDA. In total, 50 isolates were obtained from the symptomatic grapevine tissues. For genus confirmation of the isolates, the ITS regions were amplified for all the isolates. The products of the ITS regions were approx. 0.5 kb. All sequences obtained were compared to those deposited in GenBank, and the isolates possessed 95%-99% similarity with sequences from the genera Cylindrocladiella, Dactylonectria or Neonectria. One or two isolates were selected from each of these three genera for pathogenicity tests.
The ML MP trees had similar topologies, so only the ML tree is presented in this study, with ML and MP bootstrap support values.

Morphological characteristics
Morphological observations for the five identified species are outlined below.
Colony reverse sides were burnt umber to raw sienna or brownish yellow on PDA (Figure 4f (c and d). Conidia of D. macrodidyma (g and h), and D. torresensis (k and l). All the fungi were grown on PDA for 14 d. Bars = 10 μm (c and d) or 20 μm (g, h, k and l).

Lombard & Crous
Pathogenic on trunks and rootstocks of Vitis vinifera. Asexual morph: The isolates rarely formed chlamydospores and microconidia, producing some macroconidia on MEA. Macroconidia straight or minutely curved, cylindrical, one to four septate. Microconidia zero to one septate, ellipsoid to ovoid. Sexual morph: undetermined. Culture characteristics: Colonies on PDA reached 55.5 ± 3.6 mm diam. after 9 d at 25°C in the dark, and were pale buff to chestnut (Figure 4, i and j). Colony reverse sides were buff to umber to chestnut on PDA. isolates did not produce macroconidia, microconidia, or chlamydospores on MEA. Sexual morph: undetermined. Culture characteristics: Colonies on PDA reached 49.3 ± 2.2 mm diam. eafter 9 d at 25°C in the dark, and were felty to slightly cottony (Figure 4 m-

Pathogenicity tests
In the pathogenicity tests conducted with detached green shoots, the non-inoculated shoots did not develop any symptoms (Figure 5a). In contrast, shoots inoculated with mycelium discs resulted in necroses. The lesions were brown to black, and the mean lesion lengths differed among the different inoculated fungi (P < 0.05) ( Figure 5, b to h). Dactylonectria macrodidyma was the most aggressive pathogen (mean lesion length = 1.18 cm) among the five species ( Figure 5). The re-isolation rates of isolates C. lageniformis,D. torresensis,D. macrodidyma,D. alcacerensis,and Neonectria sp. 1. were between 70% with 100% from the lesions.
Pathogenicity tests on 3-month-old grapevine cuttings showed different results for the different inoculated pathogens, as well. The non-inoculated controls showed no symptoms on the shoots (Figure 6a). Dacty-lonectria macrodidyma caused brown to black necrotic lesions on the shoots (mean lesion length = 1.95 cm) ( Figure 6, d and e). Less necrosis was observed in the cuttings inoculated in C. lageniformis, D. torresensis, Neonectria sp. 1., or D. alcacerensis (Figure 6, b to c', f to g'). The re-isolation rates of the different inoculated fungi from the respective lesions were between 70% with 100%. This is the first report of C. lageniformis, D. torresensis, D. macrodidyma, D. alcacerensis and Neonectria sp. 1 associated with BF of grapevines in China.

DISCUSSION
Grapevines can be affected by several diseases throughout each year, especially during fruit production. In the present study, 50 isolates obtained from diseased grapevine samples in five provinces of China were identified as C. lageniformis, D. torresensis, D. macrodidyma, D. alcacerensis, or Neonectria sp. 1. To date, D. torresen- sis has been reported as grapevine pathogen in Australia, Canada, Czech Republic France, Italy, New Zealand, Portugal, South Africa, Spain and USA Carlucci et al., 2017;Pecenka et al., 2018;Pintos et al., 2019), and Cylindrocladiella lageniformis has been reported mainly from California and South Africa (Van Coller et al., 2005;Koike et al., 2016).
The BF pathogens are soil-borne organisms that affect roots and the basal ends of rootstock vines, and most of the fungi reported in the present study were also reported in California and Spain (Koike et al., 2016;Berlanas et al., 2020). However, whole grapevines in the north grape-growing regions of China are routinely buried under the soil in winter, to allow survival during low temperature winter conditions. It has been proposed that the soil-borne pathogens could infect plants through wounds (to roots, rootstocks, trunks, canes and shoots). Pathogenicity tests of BF pathogens were carried out with canes or roots in previous studies (Koike et al., 2016;Berlanas et al., 2020). Nevertheless, pathogenicity tests of Cylindrocladiella lageniformis have also been conducted with green shoots (Van Coller et al., 2005;Koike et al., 2016). Therefore, in the present study, it was important to determine whether the soil-borne fungi could infect host canes or shoots. The fungi were inoculated by wounding between two nodes of each cutting. Among the tested fungi, D. macrodidyma produced the longest lesions in the pathogenicity tests. However, Berlanas et al. (2020) reported that virulence of D. alcacerensis was greater than that of D. macrodidyma and Neonectria sp. 1., while D. macrodidyma was found to be more virulent than D. alcacerensis in the present study. Differences in virulence of D. alcacerensis or the other species could be attributed to: (1) strain origins (Probst et al., 2019), (2) host genotype susceptibility to black foot fungus infections (Berlanas et al., 2020), (3) methods of inoculation (Alaniz et al., 2009b;Probst et al., 2019;Berlanas et al., 2020), or (4) inoculum dose.
The distribution of BF fungal pathogens in the present study may have been influenced by climate. The climate system of China is diverse due to the varied topography and vast area, including climates of Tibetan plateau, temperate continental, subtropical monsoon, and tropical monsoon (Yan et al., 2013). The characteristics of BF pathogens are likely to vary due to the diverse temperature of China. Based on the present study (data not shown), colony diameter on PDA after 6 d of C. lageniformis from the south of China reached up to 60.4 ± 3.3 mm at 30°C while the other fungi (D. torresensis, D. macrodidyma, D. alcacerensis and Neonectria sp. 1.) from the north of China hardly grew at 30°C. Most of these fungi could grow below 5°C in PDA, except for C. lageniformis.
Although the incidence of diseased plants with BF symptoms was about 1% in the surveyed vineyards in China, which is much less than in France (losses of 50%: Larignon et al., 1999), BF pathogens can infect grapevine roots and trunks in young nurseries and plantations, and the pathogenic fungi can be transmitted to new vineyards by cuttings (De la Fuente et al., 2016). The fungi C. lageniformis, D. torresensis, D. macrodidyma, D. alcacerensis, and Neonectria sp. 1. are all soilborne, and can infect hosts through the soil (Halleen et al., 2003). In the north of China, grapevines need to be buried under the soil for survival during cold weather, resulting in small wounds that are likely to be susceptible to infection by soilborne fungi, so more attention should be paid to BF in China in future.
Grapevine BF is prevalent in nurseries and new plantations (De la Fuente et al., 2016), and the current strategies for controlling this disease include good hygiene or sanitation, which are the most important means of obtaining healthy vines (Gramaje and Armengol 2011), including treatments with hot water, (Gramaje et al., 2010;Halleen and Fourie 2016), fungicides (Halleen et al., 2007;Rego et al., 2009;Alaniz et al., 2011) and biological control agents Martínez-Diz et al., 2021;van Jaarsveld et al., 2020van Jaarsveld et al., , 2021. Chemical treatments during propagation processes in nurseries for control of BF pathogens have been evaluated, including treating cutting prior to cold storage, cutting prior to callusing, rooting pre-and post-grafting, and pre-planting fungicide treatments of rooted cuttings, to eliminate or reduce potential fungal agents before planting (Halleen et al., 2007;Rego et al., 2009, Alaniz et al., 2011, Gramaje et al., 2018. Based on previous research, benomyl was effective for elimination or reducing Cylindrocarpon destructans infections . Reductions of D. torresensis and D. macrodidyma incidence and disease severity on the bases of 2-year-old plants have been reported from applications of Streptomyces sp. E1+R4 before preplanting (Martínez-Diz et al., 2021).
Some practices, such as hot water treatments, are useful for sanitizing commercially produced plants. Generally, this practice entails treating the plants at 50°C for 30 min. However, this is stressful for the plants (Waite et al., 2013). Despite treated with these practices, diseases in symptomless plants can still be transmitted to non-infested areas (De la Fuente et al., 2016). The detection of BF fungi in soils or vines is essential for controlling the disease in nurseries and new plantations. Alaniz et al. (2009a) reported a multiplex PCR system for specific and early detection of Ilyonectria liriodendri (=Cylindrocarpon liriodendra), Dactylonectria macro-didyma (=Cylindrocarpon macrodidymum), and Dactylonectria pauciseptata (=Cylindrocarpon pauciseptatum) from pure fungus cultures or diseased plants. Martínez-Diz et al. (2020) attempted to detect I. liriodendri in bulk soils, rhizosphere soils, and grapevine endorhizospheres using Droplet Digital PCR (ddPCR) and real-time PCR (qPCR) techniques. They showed that ddPCR was more sensitive than qPCR to lower target concentrations. Nevertheless, the ddPCR technique has not been used for detection of C. lageniformis, D. torresensis, D. macrodidyma, D. alcacerensis or Neonectria sp.1, and this technology could be useful for detection of BF in China. Further study should also be conducted to develop specific protocols for effective BF management.