Bacterial Canker Disease on Populus × euramericana Caused by Lonsdalea populi in Serbia

: Populus × euramericana (Dode) Guinier clone (cl.) “I-214” is a fast-growing interspeciﬁc hybrid between Eastern cottonwood ( P. deltoides Bartr. ex Marsh) and European black poplar ( Populus nigra L.). Populus × euramericana was introduced into Serbia in the 1950s and has become one of the most widely grown poplar species. In September 2019, cankers were observed on stems and branches of P. × euramericana cl. “I-214” trees in a two-year-old poplar plantation in the province of Vojvodina, Serbia. The canker tissue was soft and watery, and a colorless ﬂuid that smelled rotten ﬂowed from the cracks in the bark, suggesting possible bacterial disease. After two weeks, diseased trees experienced crown die-back and oozing of foamy, odorous exudates and this study aimed to identify the causal agent of the disease. Canker margins and exudates were collected from 20 symptomatic trees. The associated bacterium was isolated and identiﬁed using biochemical characteristics, phylogenetic analyses based on 16S rRNA gene sequences, and multilocus sequence analyses (MLSA) based on partial sequencing of three housekeeping genes ( gyrB , infB , and atpD ). The pathogen was identiﬁed as Lonsdalea populi . Pathogenicity tests were conducted on rooted cuttings of P. × euramericana cl. “I-214” in an environmental test chamber and demonstrated that the isolated bacterial strain was able to reproduce symptoms of softened, water-soaked cankers and exudation. To the best of our knowledge, this is the ﬁrst report of L. populi causing bacterial canker disease on P. × euramericana cl. “I-214” in Serbia and in southeastern Europe (SEE). It is also the ﬁrst report of a bacterial disease on hybrid poplars, including P. × euramericana in this country and in SEE. If the disease spreads into new areas, selection for L. populi resistance may need to be integrated into future poplar breeding programs.


Introduction
Canadian poplar (Populus × euramericana (Dode) Guinier, syn. Populus × canadensis Moench) is a fast-growing interspecific hybrid between North American Eastern cottonwood (Populus deltoides Bartr. ex Marsh ♀) and European black poplar (Populus nigra L. ♂). It is an important tree species in many European countries. Populus × euramericana is characterized by rapid growth rates, ease of clonal propagation, coppice regeneration, high biomass production and carbon sequestration, potential

Sample Collection, Isolation, and Biochemical Characterization
For pathogen isolation, stem and branch tissues showing symptoms of a bacterial canker and whitish creamy exudates were collected using sterile equipment from twenty symptomatic P. × euramericana cl. "I-214" trees grown in a poplar plantation near Glogonj, in Vojvodina, Serbia (N 44°59′; E 20°32′). Samples of cankers and exudates were placed in polyethylene bags and sterile 2 ml Eppendorf tubes, respectively, and kept at 4 °C until isolation was undertaken. Small pieces (3-5 mm diameter) of

Sample Collection, Isolation, and Biochemical Characterization
For pathogen isolation, stem and branch tissues showing symptoms of a bacterial canker and whitish creamy exudates were collected using sterile equipment from twenty symptomatic P. × euramericana cl. "I-214" trees grown in a poplar plantation near Glogonj, in Vojvodina, Serbia (N 44 • 59 ; E 20 • 32 ). Samples of cankers and exudates were placed in polyethylene bags and sterile 2 mL Eppendorf tubes, respectively, and kept at 4 • C until isolation was undertaken. Small pieces (3-5 mm diameter) of woody tissue were cut from the canker margins, surface sterilized using 70% (v/v) ethanol for 1 min. followed by a solution of 10% (v/v) sodium hypochlorite for 1 min., and then rinsed in sterile distilled water. The tissue was macerated in 1 mL of sterile distilled water using a sterile mortar and pestle; the resulting suspension was transferred to 2 mL Eppendorf tubes, and shaken for 2 min. using ZX3 advanced vortex mixer (VELP Scientifica, Milan, Italy). The suspension was serially diluted (10-fold dilutions to 10 −9 ) and 50 µL of each dilution was spread onto tryptone soya agar (TSA, Titan Biotech Ltd., New Delhi, India). Moreover, samples of exudates were diluted in the same manner and spread onto TSA. Petri dishes were incubated at 30 • C for 48 h (Heidolph incubator 1000, Heidolph co., Kelheim, Germany). In total, seven bacterial colony types were isolated. The prevalent colonies were similar in appearance to L. populi, i.e., white-ivory colored, slightly bluish on the underside, round, and slightly convex [8]. These colonies were purified by subculturing and subjected to Gram test which was performed using a non-staining method with a 3% KOH solution [13]. Gram negative bacterial strains were further examined with an Olympus BX53F light microscope (Olympus Co., Tokyo, Japan) at ×400 and ×1000 magnification using an Olympus SC50 digital camera and accompanying software. A Gram-negative strain with cell morphology like L. populi (cells 0.5-1 × 1-2 µm in size, short-rod-shaped, motile, occurring single, or aggregated in clumps, Figure S1) [8] was selected and named ILFE-LP1. The strain was stored in tryptone soya broth (TSB, Titan Biotech Ltd., New Delhi, India) containing 40% glycerol (v/v) at −80 • C, deposited in the culture collections of the Institute of Lowland Forestry and Environment (ILFE) and NARIC Fruitculture Research Institute and further used in this study to identify the bacterium and confirm its pathogenicity.
The analyses of biochemical characteristics of the bacterium were conducted using API 20E kit (Bio-Mérieux, Marcy L'Etoile, France) following the manufacturer's instructions and the test strip was incubated for 24 h. A type strain of L. populi (NY060, provided by the NARIC Forest Research Institute, Mátrafüred, Hungary) was used as a positive control and isolates were assessed twice.

DNA Extraction, PCR Amplification, and Sequencing
Total bacterial genomic DNA was isolated from cells harvested from culture grown for 24 h at 30 • C in TSB using a mericon DNA bacteria kit (Qiagen, Hilden, Germany) following the manufacturer's protocols. The DNA was quantified with a nanodrop (BioSpec-nano, Shimadzu Biotech, Kyoto, Japan), stored at −20 • C and diluted to the concentration of 20 ng/µL prior to use in PCR reactions. Partial 16S rRNA gene was amplified by PCR using the primers and conditions as published by [14] (Table 1). Three housekeeping genes, including gyrB, infB, and atpD were amplified using primers designed by [15] ( Table 1). The conditions for PCR amplification of the housekeeping genes were as previously determined by [16]. The PCR products were separated by electrophoresis on 1.5% (w/v) agarose gels in 1 x TBE buffer, stained with Roti-GelStain (Carl Roth, Karlsruhe, Germany) and visualized under UV illumination. The size of the products was estimated using O'gene ruler 100bp DNA ladder (Thermo Fisher Scientific Inc., Bremen, Germany). The PCR products were purified using a PCR purification kit (QIAquick, Qiagen, Hilden, Germany). Sanger sequencing was performed by Mycrosynth (Balgach, Switzerland) using primers designed by [14,15] (Table 1).

Phylogenetic Analyses
Raw sequence data were examined and combined into a consensus sequence using BioEdit version 7.2.5 [17] and MEGA X [18]. Sequences were compared to those of the other Lonsdalea strains available in the GenBank database using BLAST and related sequences were downloaded and included in the analyses (Table S1). Multiple sequence alignments were obtained with MAFFT version 7 (on-line version) [19], checked manually for alignment errors in MEGA X [18] and corrected where necessary. The phylogenetic analyses were performed using Maximum Parsimony (MP) and Maximum Likelihood (ML) analyses. ML analyses were conducted for both 16S rRNA and the combined data set of three housekeeping genes, whereas MP analyses were run only for the combined data set. The MP analyses and the partition homogeneity test (PHT) were conducted as described by [20] and they were performed in PAUP version 4.0b10 [21]. ML analyses were run using an online version of PhyML 3.0 [22] by applying smart model selection [23]. Bootstrap analysis was carried out using 1000 replicates [24]. Phylogenetic trees were visualized using MEGA X [19]. The DNA sequences of isolate ILFE-LP1 obtained in this study were deposited in GenBank (MT505705-16S, MT537174-atpD, MT559754-gyrB, and MT559753-infB, Table S1). Table 1. Primers used in this study to amplify and sequence 16S rRNA gene and housekeeping genes (gyrB, atpD and infB) from ILFE-LP1 bacterial strain isolated from a Populus × euramericana cl. "I-214" tree with symptoms of a bacterial canker in Serbia.

Pathogenicity Test
In January 2020, shoots were collected from stooled beds of P. × euramericana cl. "I-214" established at an experimental forest nursery "Kaćka forest" of ILFE in Kać, Novi Sad (N 45 • 17 ; E 19 • 53 ). Dormant cuttings (diameter: 14 ± 0.3 mm; length: 30 ± 0.2 cm) were prepared from the lower parts of the collected shoots and stored in polyethylene bags in a cold chamber at 4 • C for two months. In March 2020, the cuttings were first soaked in water for two days in the dark at room temperature (18 ± 2 • C). They were then surface sterilized using 70% ethanol (v/v) and their top ends were sealed using grafting wax (Savacoop, Novi Sad, Serbia) to prevent desiccation and contamination. The cuttings were placed in 3 L plastic pots containing loamy fluvisol soil [5] obtained from the "Kaćka forest" nursery. They were kept in a greenhouse at ILFE (23 ± 2 • C day temperature, 19 ± 2 • C night temperature, 60-70% humidity, 16/8h day/night cycle) for one month and watered every other day to field capacity.
Thirty cuttings that developed roots were transferred to an environmental test chamber (Sanyo, MLR-351H) for the pathogenicity test. They were arranged in a completely randomized design with ten replicates (poplar plants) per treatment. Plants were inoculated with Serbian strain ILFE-LP1 and a Hungarian type strain of L. populi NY060 to serve as a positive control. Negative controls were mock inoculated using sterile distilled water. Prior to inoculation, bacterial strains were cultured for 24 h at 30 • C on TSA. Inoculum was prepared in 20 mL TSB. Single bacterial colonies were transferred to Erlenmeyer flasks and incubated at 30 • C and 180rpm for 24 h in a shaker incubator (Unimax 1010, Heidolph co., Kelheim, Germany). The number of colony forming units (CFU)/ml was determined by spread-plate technique [25] on TSA incubated at 30 • C for 24 h. Bacterial suspension in TSB was transferred to 2 mL Eppendorf tubes, centrifuged at 13,200 rpm for 10 min. and the inoculum concentration was adjusted with sterile distilled water to 10 8 CFU/ml. Plants were first surface sterilized using 70% ethanol (v/v) and then a cork borer was used to create a 6 mm diameter wound in the middle of each plant. The bacterial suspension was injected into the wound (40 µL) using a pipette and the wound was sealed with Parafilm (Pechiney, Chicago, IL, USA) to retain moisture and protect the wound from contamination. The inoculated plants were maintained at 28 • C, with a relative humidity of 90% under a 16/8 day/night cycle [26] and watered as described above. The experiment was carried out for one month and plants were monitored every day for the appearance of symptoms. At the end of the experiment, the presence of external and internal symptoms was recorded, and the length of internal canker lesions was measured. The pathogenicity test was repeated once.

Statistical Analyses
Statistical analyses of pathogenicity experiment data were performed using Statistica 12.0 (StatSoft Inc., Tulsa, OK, USA). The data were checked for normality using Kolmogorov−Smirnov test and homogeneity of variances was tested with Levene's test. The analyses were further conducted using non-parametric Mann−Whitney U test (α = 0,05). Because there were no significant differences in lesion lengths produced by the same strain in the two subsequent pathogenicity trials, the data from a single strain were combined for further analyses.

Biochemical Characterization
Isolate ILFE-LP1 was positive for acetoin and citrate utilization, and negative for β-galactosidase, arginine dihydrolase, lysine decarboxylase, ornithine decarboxylase, H 2 S, urease, tryptophan deaminase, indole, and gelatinase production. Acid was produced from D-glucose, D-mannitol, D-sucrose, and amygdalin. Nitrates were not reduced to nitrites. Biochemical characteristics of ILFE-LP1 resembled those of the type strain NY060 of L. populi.

Phylogenetic Analyses
The 16S data set contained 12 sequences and 1351 characters (1282 parsimony informative, 46 parsimony uninformative, CI = 0.9, RI = 0.8, TL = 76) and the model HKY85+I was chosen for the ML analyses (I = 0.897). The topology of the ML and MP phylogenetic trees was similar, and the ML tree is presented (Table S1, Figure S2).
The combined dataset of three housekeeping genes contained 36 sequences with Brenneria nigrifluens as an outgroup (Table S1). The sequence dataset contained 1833 characters (195 parsimony informative, 1638 parsimony uninformative, CI = 0.8, RI = 0.9, TL = 320). The result of the PHT test was not significant and showed that three loci can be combined (P = 0.03). The model GTR+G+I was chosen for the ML analyses (G = 0.507, I = 0.451). The MP and ML analyses produced phylogenetic trees with the similar topology and therefore, only the ML tree is shown (Figure 2). Serbian strain ILFE-LP1 formed a monophyletic clade with L. populi strains from Hungary, Portugal, and China within Lonsdalea species in the phylogenetic analyses. The separation of L. populi from other Lonsdalea species was moderately supported in the phylogenetic analyses of 16S rRNA sequences (bootstrap support = 85% ML, MP) and strongly supported in the phylogenetic analyses of a combined atpD, gyrB and infB dataset (bootstrap support = 81% ML, 100% MP). Although the branch lengths indicate differences in the concatenated sequences, the scale bar represents 0.001 nucleotide changes per site.
Based on phylogenetic analyses, strain ILFE-LP1 isolated in this study was identified as L. populi (Figure 1, Figure S2).

Pathogenicity Test
Four days after inoculation, the oozing of a colorless fluid and a small sunken area around the inoculation site were evident on most stems inoculated with a bacterial suspension of ILFE-LP1 and on stems of a positive control inoculated with L. populi NY060. The lesion gradually expanded further in the following days and oozing continued. One month after inoculation, sunken, water-soaked external cankers were visible on each stem inoculated with isolates ILFE-LP1 and L. populi NY060 and after the bark was removed, water-soaked, necrotic lesions were observed around the inoculation points and measured from 1.4 to 3.6 mm (average length = 1.8 mm) and 1.4 to 3.8 mm (average length = 1.9 mm), respectively (Figure 1). There were no significant differences in lesion lengths between the two strains (p = 0.47). No cankers formed on the stems of the control plants. Lonsdalea populi was successfully (100%) re-isolated from canker margins on TSB and its identity was confirmed using morphology, Gram test, PCR, and sequencing of the atpD gene following the procedure described above ( Figure S3). Lonsdalea populi was never isolated from negative controls.

Discussion
This study provides the first record of L. populi on P. × euramericana cl. "I-214" in Serbia, and southeastern Europe (SEE). The geographic range of this Gram-negative bacterium has extended and its host association with P. × euramericana was confirmed. Lonsdalea populi was identified using a polyphasic approach, i.e., biochemical characteristics, phylogenetic analyses of the 16S rRNA, and MLSA of three housekeeping genes (gyrB, infB and atpD). The pathogenicity test confirmed that L. populi is the causal agent of the bacterial canker disease of P. × euramericana cl. "I-214" in Serbia.
The present study is also the first report of a bacterial disease on hybrid poplars, including P. × euramericana in Serbia, and SEE. Despite the importance of hybrid poplars, no previous research has been conducted on bacterial diseases of these trees in Serbia and SEE. Moreover, little research has

Pathogenicity Test
Four days after inoculation, the oozing of a colorless fluid and a small sunken area around the inoculation site were evident on most stems inoculated with a bacterial suspension of ILFE-LP1 and on stems of a positive control inoculated with L. populi NY060. The lesion gradually expanded further in the following days and oozing continued. One month after inoculation, sunken, water-soaked external cankers were visible on each stem inoculated with isolates ILFE-LP1 and L. populi NY060 and after the bark was removed, water-soaked, necrotic lesions were observed around the inoculation points and measured from 1.4 to 3.6 mm (average length = 1.8 mm) and 1.4 to 3.8 mm (average length = 1.9 mm), respectively ( Figure 1). There were no significant differences in lesion lengths between the two strains (p = 0.47). No cankers formed on the stems of the control plants. Lonsdalea populi was successfully (100%) re-isolated from canker margins on TSB and its identity was confirmed using morphology, Gram test, PCR, and sequencing of the atpD gene following the procedure described above ( Figure S3). Lonsdalea populi was never isolated from negative controls.

Discussion
This study provides the first record of L. populi on P. × euramericana cl. "I-214" in Serbia, and southeastern Europe (SEE). The geographic range of this Gram-negative bacterium has extended and its host association with P. × euramericana was confirmed. Lonsdalea populi was identified using a polyphasic approach, i.e., biochemical characteristics, phylogenetic analyses of the 16S rRNA, and MLSA of three housekeeping genes (gyrB, infB and atpD). The pathogenicity test confirmed that L. populi is the causal agent of the bacterial canker disease of P. × euramericana cl. "I-214" in Serbia.
The present study is also the first report of a bacterial disease on hybrid poplars, including P. × euramericana in Serbia, and SEE. Despite the importance of hybrid poplars, no previous research has been conducted on bacterial diseases of these trees in Serbia and SEE. Moreover, little research has been conducted on bacterial diseases of Populus spp. in SEE and only 'Candidatus Phytoplasma asteris'-related phytoplasmas (yellow disease phytoplasmas) were reported from Lombardy poplar (Populus nigra L. 'Italica') trees planted as ornamentals in Belgrade, Serbia and in Zagreb, Croatia [27,28].
The isolation of L. populi from P. × euramericana in Serbia is not surprising, given that it is a well-known pathogen that causes bacterial canker disease of hybrid poplars [12]. Lonsdalea populi has been isolated from P. × euramericana in previous studies in Hungary, Spain, Portugal, and China [8,[10][11][12]. It has also been found associated with P. × interamericana in Spain and recently reported as a pathogen of Chinese willow (Salix matsudana Koidz.) causing cankers with large amounts of white sour exudates in China [11,29].
The disease symptoms (oozing cankers with water-soaked, soft wood) caused by L. populi observed in this study are consistent with previous reports of Lonsdalea canker of poplars [8,[10][11][12]. In the current study, however, L. populi was isolated from two-year old P. × euramericana trees, whereas in other countries it was found on more than three-year-old trees [8,[10][11][12]. Symptoms of a bacterial canker disease on P. × euramericana have previously also been reported associated with Xanthomonas populi (Ridé) Ridé and Ridé. This bacterium was a major concern in poplar-growing regions of Europe in the 1950s [30]. However, symptoms observed in this study were not typical of those caused by X. populi and swollen cankers with deep cracks in the bark have not been observed in the field. Moreover, Neocosmospora solani sensu lato has been found associated with cankers of hardwood trees, including Populus spp. [31]. Likewise, apart from L. populi, Li et al. [12] isolated N. solani (Mart.) L. Lombard & Crous from diseased tissues of P. × euramericana trees experiencing stem cankers in China. The authors, however, reported that N. solani is not an aggressive pathogen of this tree species and concluded that L. populi is the causal agent of a canker disease of P. × euramericana in China. Nevertheless, because disease symptoms indicated a possible bacterial infection, fungal isolations were not performed in this study, but additional research is currently being conducted to see if L. populi alone is causing the canker symptoms observed in the field.
To prevent the spread of the disease into new areas and plantations, in this study, a pathogenicity test was conducted using P. × euramericana rooted cuttings in an environmental test chamber under controlled conditions as described in Hou et al. [26]. Due to the high humidity to which poplars were exposed during the test (90%) to promote bacterial activity, and the fast-growing nature of P. × euramericana the experiment lasted for one month and symptoms of water-soaked cankers with exudation were successfully reproduced. Moreover, cankers of a similar size were formed when P. × euramericana was inoculated with the strain type of L. populi that was used as a positive control. However, oozing was not as abundant, foamy, and creamy as seen on trees in the field in an advanced stage of the disease development. This may be due to the age of the plants used for inoculation, and the duration of the test. Similarly, in a study of Li et al. [12] L. populi induced canker symptoms when inoculated into water-cultured excised stems in an environmental test chamber, but abundant, white, sour exudates were observed only when the test was carried out under field conditions using 3-5-year old trees. Difficulties in reproducing disease symptoms in pathogenicity trials have also been reported for other plant pathogenic bacteria, including  [32][33][34].
The occurrence and pathogenicity of L. populi on P. × euramericana cl. "I-214" is of a major concern in Serbia because cl. "I-214" is the most widely grown and economically important poplar clone in the country. Intensively cultured plantations (even-aged, clonal stands) and monoclonality have already increased the vulnerability of this clone to various leaf and canker diseases [4,9]. Because the use of antibiotics for plant disease control in Serbia is prohibited [35], management options for bacterial disease problems in Serbian poplar plantations are limited. Therefore, genetic improvement programs that continuously screen new clones for disease resistance while assuring highest possible volume production could be the most promising strategy to combat Lonsdalea canker of poplars. Moreover, an integrated approach of disease prevention and control, focusing not only on selection and breeding for resistance, but also on biological control is needed to assure long-term sustainability of poplar plantations in Serbia.

Conclusions
To the best of our knowledge, this is the first report of L. populi causing bacterial canker disease on P. × euramericana cl. "I-214" in Serbia and in SEE. It is also the first record of a bacterial disease on P. × euramericana in SEE.
Lonsdalea populi is currently the most serious pathogen affecting P. × euramericana plantations in Europe. It is also a serious threat to Serbian poplar production. Therefore, there is a need for disease management strategies that are not only economically practical, efficient, and sustainable, but also likely to be accepted by poplar growers. If the disease spreads into new areas, selection and breeding for Lonsdalea canker disease resistance might be such a strategy.
Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4907/11/10/1080/s1. Figure S1: Primers used in this study. Figure S2: Maximum-likelihood (ML) tree resulting from ML analyses of the partial 16S rRNA gene sequences (1351 bp). The bootstrap support values (ML/MP ≥ 80%) are indicated at the nodes, and the scale bar represents the number of changes. The tree was rooted to Brenneria nigrifluens. Figure  S3: Lonsdalea-like colonies (marked blue) re-isolated from symptomatic tissue of Populus × euramericana clone "I-214" on tryptone soya agar. Petri dishes were incubated at 30 • C for 48 h. Table S1: Bacterial strains used for phylogenetic analyses.