Genotypic and phenotypic uniformity among the population of Pectobacterium atrosepticum strains isolated during three growing seasons from potato fields in Poland

Pectobacterium atrosepticum (Pba) are phytopathogenic, pectinolytic, non-sporulating, gram-negative rods causing blackleg and soft rot diseases in potato. They are ubiquitously present under temperate climate conditions and contribute to significant economic losses. Currently, there are no efficient control methods against soft rot Pectobacteriaceae (SRP). As limitation of the spread of SRP is solely prevention-based, we undertook broad characterization on both genotypic and phenotypic levels of the P. atrosepticum isolates acquired during monitoring of potato fields in Poland in 2013, 2014 and 2016. Of REP, ERIC and BOX-based genomic fingerprinting, the BOX-based approach turned out to be the most informative and allowed for classification of 118 Pba strains into six groups. The present study indicated that IV and I BOX profiles dominated among the investigated population. Finally, 23 Pba strains representing all BOX groups and originating from various growing seasons were selected for genotypic, phylogenetic and phenotypic studies. rpoS-based phylogeny revealed intraspecies variation (16 SNP sites) among the studied Pba strains, in contrast to the analyses relying on gyrA (1 SNP) and recA (0 SNP) sequences. Pba strains showed higher potency to macerate potato tissue at 20 °C than 28 °C. These isolates exhibited rather uniform production of plant cell wall degrading enzymes (pectinases, cellulases and proteases), lipases, siderophores and biofilm, however in the majority of cases inferior in comparison to Dickeya solani and Pectobacterium carotovorum strains. In conclusion, this study revealed genotypic and phenotypic uniformity in addition to high virulence potential of Pba strains isolated from potato fields during several growing seasons in Poland.


Introduction
Pectobacterium atrosepticum is a Gram(-), nonsporulating necrotrophic plant pathogenic bacterium responsible for blackleg and soft rot on potato (Perombelon & Kelman, 1980). It was isolated for the first time nearly 120 years ago by C. J. J. van Hall and assigned to Bacillus atrosepticus species (van Gijsegem et al., 2021a, b). Subsequently, due to improvement of characterization and differentiation methods, the taxonomic position of this phytopathogen was changed to Erwinia carotovora subsp. atroseptica, then Pectobacterium carotovorum subsp. atrosepticum to be finally classified as P. atrosepticum within the family Pectobacteriaceae, order Enterobacterales (Adeolu et al., 2016;Gardan et al., 2003;Hauben et al., 1998;Winslow et al., 1920).
Until the 1970s, P. atrosepticum was stated as a unique cause of potato blackleg and main pathogen responsible for potato soft rot disease in Europe (van Gijsegem et al., 2021a, b), putatively due to strict association with temperate climate conditions (Toth et al., 2003). Later on, other soft rot Pectobacteriaceae (SRP) species such as Dickeya solani, Pectobacterium parmentieri or Pectobacterium brasiliense started being detected on this continent, though P. atrosepticum was still frequently isolated in potato-growing countries and significantly contributed to the local SRP populations (Dees et al., 2017;Motyka-Pomagruk et al., 2021). To underline the importance of P. atrosepticum, this species was placed by Mansfield et al. (2012) on the top ten list of plant pathogenic bacteria. Moreover, the firstly published full genomic sequence of any SRP was that of P. atrosepticum SCRI1043 (Bell et al., 2004). Importantly, P. atrosepticum in contrast to some other SRP, like e.g. Pectobacterium carotovorum, P. brasiliense or Dickeya dadantii, is regarded to have a narrow host range (Ma et al., 2007). P. atrosepticum causes blackleg and soft rot on potato, though strains from this species were also previously isolated from Brassica rapa, Helianthus annuus, Solanum melongena, Zantedeschia aethiopica or asymptomatic unspecified weeds originating from potato fields Toth et al., 2021). The disease symptoms on potato include maceration of the inner tissue of the tubers known as soft rot, in addition to blackening of the stem base leading to wilting of the whole plant referred to as blackleg (Perombelon & Kelman, 1980). Blackleg symptoms are restricted to the field conditions, while soft rot is recognized in the fields, transportation, storage and during marketing of the tubers. However, P. atrosepticum has been isolated more frequently to the present day from affected potato stems than tubers Śledź et al., 2000).
To observe the typical disease symptoms, a virulent pathogen needs to be present, as well as a susceptible host and environmental conditions favoring bacterial multiplication (Barrett et al., 2009). At first, P. atrosepticum cells infiltrate susceptible plant host via wounds or natural openings such as lenticels (Perombelon & Kelman, 1980). Afterwards, bacterial cells remain in the vascular tissue and plant intercellular spaces until the infection-favoring conditions occur (Toth et al., 2003). In the meantime, the pathogen produces extracellular polysaccharide (EPS), a component of the 3D biofilm structure, which is essential for facing water flow and plant defense mechanisms during colonization of the plant xylem (Gorshkov et al., 2016). In SRP, including the cells of P. atrosepticum, the quorum sensing mechanism strictly controls the transcription factorsdependent regulatory network, which is activated after reaching a required cell density threshold (Liu et al., 2008). As a result, efficient production of major and minor virulence factors begins (Liu et al., 2008). P. atrosepticum secretes plant cell wall degrading enzymes (PCWDEs) extracellularly, via Type I or Type II secretion systems. Pectinases, cellulases and proteases depolymerize main components of the plant cell wall, triggering leakage of the protoplast content, which is rich in nutrients highly valuable for the pathogen (Perombelon, 2002). Of the PCWDEs listed, pectinases, including pectate lyases, pectin lyases, pectin methyl esterases and polygalacturonases (Barras et al., 1994), contribute to the highest extent to the virulence of this pest. Moving to the additional PCWDEs, two cellulases (CelV/CelB in addition to CelS) and up to six putative proteases have been identified in P. atrosepticum (van Gijsegem et al., 2021a, b). Even though in a closely related SRP species inactivation of a single or several genes coding for the above-listed enzymes did not end up in a complete loss of bacterial pathogenicity, reduction in the resultant disease symptoms was noted (van Gijsegem et al., 2021a, b). Access to the nutrients is furthermore provided by the action of lipases and nucleases, releasing readily available fatty acids and oligonucleotides, respectively (van Gijsegem et al., 2021a, b). Recently, a notable impact on the overall virulence of P. atrosepticum was attributed to the production of an iron-scavenging siderophore, namely enterobactin (Gorshkov et al., 2021).
Interestingly, previous genome-oriented studies based on polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) or amplified fragments length polymorphism (AFLP) suggested high genomic homogeneity within the P. atrosepticum species in contrast to, for instance, P. carotovorum (Darrasse et al., 1994;Ngadze et al., 2012;Waleron et al., 2002). The uniformity of the genomes of distinct P. atrosepticum strains is especially puzzling taking into consideration that these strains have been isolated for many years from all continents where potato plants are grown (van der Wolf et al., 2021a, b). Ubiquitous presence of P. atrosepticum strains results from efficient spread by various artificial and environmental routes. These disease-causing agents may be dispersed by shipping of latently infected plant material or transfer of plant debris containing SRP cells . P. atrosepticum cells may also originate from alternative plant hosts or be spread by various vectors such as insects, maggots or nematodes (Rossmann et al., 2018;Toth et al., 2021). As other sources of these pathogens, there are listed waterways (McCarter-Zorner et al., 1984), soil (Peltzer & Sivasithamparam, 1988), air, aerosols and surfaces of field machinery or storage containers (Toth et al., 2021). Despite the high economic importance and common occurrence of P. atrosepticum, the currently applied methods to combat SRP, are solely prevention-based (Czajkowski et al., 2011). Monitoring of the presence of P. atrosepticum on potato fields followed by a detailed characterization of the acquired isolates seems crucial in finding a way to manage this pathogen.
In order to broaden the current knowledge on this phytopathogen, we extensively characterized the P. atrosepticum isolates acquired in years 2013-2016 during monitoring of potato fields in Poland for the presence of SRP . We applied Repetitive Extragenic Palindromic (REP), Enterobacterial Repetitive Intergenic Consensus (ERIC) and BOX-based repetitive sequencebased PCRs (rep-PCR) to reveal genomic variability among 118 P. atrosepticum strains. Subsequently, 23 representative P. atrosepticum strains exhibiting the 6 assigned BOX profiles and preferably originating from various growing seasons were selected for further genotypic and phenotypic analyses. rpoS, gyrA and recA-based phylogenies showed extremely high homogeneity among all 23 studied P. atrosepticum strains. In terms of phenotypic traits, the abilities of P. atrosepticum strains to macerate potato tissue, produce plant cell wall degrading enzymes (pectinases, cellulases and proteases), lipases, siderophores and biofilm were investigated in comparison to 5 reference strains belonging to the other SRP species.

Materials and methods
Bacterial isolation and identification P. atrosepticum strains were isolated in 2013, 2014 or 2016 during monitoring of potato fields in Poland for the presence of blackleg-and soft rot-causing bacteria as described by Potrykus et al., 2016;Zoledowska et al., 2018a).
Briefly, inspectors from the Regional Inspectorates of Plant Health and Seed Inspection Service in Poland (PH & SIS) and the employees of the collaborating plant breeding institutions collected samples of blackleg-or soft rot-affected potato plants or tubers (for more information on the studied samples see Motyka-Pomagruk et al., 2021). Shortly after arrival, 1 g of each plant or tuber tissue was manually homogenized in 10 mL phosphate buffer (pH = 7.2) in an extraction bag (Bioreba, Reinach, Switzerland). The obtained suspension was serially diluted to 10 −6 in 0.85% NaCl solution. 100 μl of 10 -5 and 10 -6 dilutions were plated on Crystal Violet Pectate (CVP) medium (Hélias et al., 2012). After 48 h incubation at 28 °C, the observed cavity-forming colonies of pectinolytic bacteria, were picked and replated several times either on CVP or Trypticase Soy Agar (TSA) media to acquire axenic cultures.
Preliminary identification of pectinolytic isolates was performed with multiplex PCR (Potrykus et al., 2014a). Assignment to the P. atrosepticum species was based on acquisition of a specific 420 bp amplicon in a single PCR reaction conduced with Y45 and Y46 starters according to Frechon et al. (1998). The identified P. atrosepticum strains were frozen in 40% Genomic DNA of the 23 selected P. atrosepticum isolates and 6 reference strains (Table 1) was isolated from single overnight cultures in Trypticase Soy Broth (TSB) (BTL, Łodz, Poland) with Genomic Mini AX Bacterial Kit (A&A Biotechnology, Gdansk, Poland) according to the manufacturer's instructions. The quality and concentration of the isolated DNA was assessed spectrophotometrically with a NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA) and stored at 4 °C for further use.
The resultant PCR products of 920 bp for rpoS, 737 bp for recA and 1018 bp for gyrA, respectively, were sequenced from both ends by a commercial company (Genomed, Warsaw, Poland), manually edited and subjected to multiple sequence alignment by the ClustalW algorithm in the MEGA X (Pennsylvania State University; www. megas oftwa re. net; (Kumar et al., 2018)) software. The rpoS, recA and gyrA-based phylogenetic trees were also constructed in MEGA X (Kumar et al., 2018). A maximum likelihood algorithm with the Tamura-Nei model was used. Bootstrap was set at the level of 1000 replicates. The lengths of the tree branches were associated with the number of substitutions per site. The corresponding gene sequences of D. solani IFB0099 were applied as the outgroups. The sequences of rpoS, recA and gyrA genes of P. atrosepticum isolates were submitted to the GenBank database and are freely available under the accession numbers listed in Table 2.

Phenotypic analyses
Overnight cultures in TSB of the 23 selected P. atrosepticum strains and the reference strains (Table 1) were centrifuged (10 min, 4779 g) and the cell pellet was washed twice in sterile 0.85% NaCl. Subsequently, either 0.5 McF or OD 600 = 0.1 bacterial suspensions in 0.85% NaCl were prepared with the use of densitometer DEN-1B (Biosan,

Potato maceration assay
Potato slice maceration assay was performed as described previously (Zoledowska, et al., 2018a) with some modifications. Briefly, potato tubers cv. Lord were acquired from a local market, washed under tap water and surface-sterilized by immersing for 10 min in 10% NaOCl. After washing-off disinfectant residuals, the tubers were dried in a laminar flow cabinet. Additionally, 15 min UV-exposure was applied to both sides of the tubers. Surface-sterilized potato tubers were cut in 1 cm slices. Then, up to 3 holes (of 5 mm in diameter) were drilled in each potato slice. Potato slices were placed on moistened Whatman filter paper in sterile plastic boxes. 50 µl of OD 600 = 0.1 bacterial suspension of the tested P. atrosepticum strain (Table 1) or the reference SRP strain (Table 1) were introduced into one hole. The control samples, i.e. 50 µl of 0.85% NaCl solution-inoculated holes were included. After covering the boxes with the lids, potato slices were incubated under anaerobic, humid conditions for 48 h at either 28 °C or 20 °C depending on the experiment. Subsequently, the diameters (in millimeters) of the resultant rotten spots were measured. The collected results corresponded to the potato tissue maceration capacity of the studied strains. Each experiment was repeated twice with seven replicates per tested strain.

Growth at different temperature conditions
Bacterial ability to grow at diverse temperatures, i.e. at 20 °C, 28 °C and 37 °C (Golanowska et al., 2017) was evaluated as follows: 5 µl of 0.5 McF bacterial suspension of the tested P. atrosepticum strain (Table 1) or the reference SRP strain (Table 1) were placed on the surface of TSA medium. The abilities of bacterial strains to form colonies after 72 h incubation at 20 °C, 28 °C or 37 °C were recorded based on visual observations. The experiment was repeated three times with two replicates each. A plus sign was assigned to strains exhibiting the ability to grow at the examined temperature, a minus sign was attributed to strains incapable of multiplying. If discrepancies were observed between the replicates the outcome was marked as inconclusive.

Growth on medium with 5% NaCl
The bacterial ability to grow at an elevated salinity level (5% NaCl) (Sarfraz et al., 2020) was evaluated as follows. 5 µl of 0.5 McF bacterial suspension of the tested P. atrosepticum strain (Table 1) or the reference SRP strain (Table 1) was placed on the surface of TSA + 5% NaCl medium. Incubation of the plates at 28 °C for 72 h followed. Bacterial ability to form visible colonies was assessed. The experiment was repeated three times with two replicates each.

Pectinases activity
Pectinases activity of bacterial strains was investigated on M63 + PGA medium as described previously (Reverchon et al., 1986) with some modifications. 1.5 µl of OD 600 = 0.1 bacterial suspension of the tested P. atrosepticum strain (Table 1) or the reference SRP strain (Table 1) was placed on the surface of M63 + PGA medium. Incubation at 28 °C for 48 h followed. Subsequently, the medium was flooded with 10% Cu(CH 3 COO) 2 for 10 min. The plates then stained blue, while white halo zones around bacterial colonies resulted from PGA degradation. The diameters of the halo zones were

Proteases activity
Protease activity of bacterial strains was investigated on skimed milk agar medium as described previously (Ji et al., 1987) with some modifications. 1.5 µl of OD 600 = 0.1 bacterial suspension of the tested P. atrosepticum strain (Table 1) or the reference SRP strain (Table 1) was placed on the surface of skim milk agar medium. Incubation at 28 °C for 48 h followed. The diameters of the halo zones around bacterial colonies were measured. The experiment was repeated three times with four replicates each.

Lipases activity
Lipases activities of bacterial strains were investigated on rhodamine medium as described previously (Samad et al., 1989) with some modifications. 1.5 µl of OD 600 = 0.1 bacterial suspension of the tested P. atrosepticum strain (

Siderophores production
Siderophores production of bacterial strains was investigated on a chrome azurol S (CAS) agar medium as described previously (Schwyn & Neilands, 1987) with some modifications. 1.5 µl of OD 600 = 0.1 bacterial suspension of the tested P. atrosepticum strain (Table 1) or the reference SRP strains (Table 1) was placed on the surface of CAS medium. Incubation at 28 °C for 48 h followed. The diameters of the orange halo zones around bacterial colonies were measured. The experiment was repeated three times with two replicates each.

Biofilm formation assay
Biofilm formation capacity of bacterial strains was investigated as described previously (Babinska et al., 2021) with some modifications. 10 µl of an overnight culture in TSB medium of the tested P. atrosepticum strain ( water and air-dried. For quantitative evaluation 600 µl of 96% ethanol was added to each biofilm-containing microcentrifuge tube. The absorbance of the extracted crystal violet, which corresponded to the biofilmformation capacity was assessed spectrophotometrically at 565 nm with Victor2 I420 Multilabel Counter (Wallac, Finland) in a 96-well microplate (Sarstedt, Germany). P. fluorescens CCM2115 and P. aeruginosa PAO1 were included in this analysis as positive controls of potent biofilm producers. The experiment was repeated twice with four replicates each.

Statistical analysis and data visualization
The gathered data were analysed in R (R Core Team, 2021) version 3.1.3 with the use of the agricolae package. The equality of variances and the data normality were determined by Levene's and Shapiro-Wilk's tests, respectively. If the requirements for parametric analysis were fulfilled, the analysis of variance (ANOVA) followed by Tukey's honest significant difference test was performed (Fig. 4a) 1998) as P. atrosepticum. This pool of strains was subjected to rep-PCR-based genomic profiling with the BOX (Fig. 1), ERIC (Supplementary Fig. 1) and REP primers ( Supplementary Fig. 2). Of the above-listed tests, BOX-PCR turned out to be the most effective for differentiation of P. atrosepticum isolates (Fig. 1). It allowed for classification of 118 collected P. atrosepticum strains to 6 diverse BOX profiles I-VI (Table 3). BOX-PCR-based genomic profile IV (Fig. 1) was the one most frequently observed ( Table 3). The second most commonly detected (Table 3) genomic pattern I (Fig. 1) was characteristic for about 30% of the Polish strains and was also attributed to P. atrosepticum type strain LMG 2386 TS . The other genomic profiles (II, III, V and VI) have been assigned solely to several strains (Tables 2, 3). On the basis of the outcomes of the BOX-based genomic profiling, 23 P. atrospeticum strains (Tables 1, 2) exhibiting diverse genomic patterns and isolated in different years were selected as a representative pool for further phylogenyoriented and phenotypic analyses.

Phylogenetic analysis of P. atrosepticum strains originating from potato fields in Poland
Of the performed recA ( Supplementary Fig. 3), gyrA ( Supplementary Fig. 4) and rpoS-based phylogenetic analyses, the one relying on rpoS (Fig. 2) revealed intraspecies variability within the Polish population of P. atrosepticum. The sequence of recA was identical among all the tested P. atrosepticum strains ( Supplementary Fig. 3), while in terms of gyrA, there was only one SNP identified that differentiated IFB5665 (of BOX profile I) from the other P. atrosepticum strains ( Supplementary  Fig. 4). The rpoS-based analysis yielded 16 SNPs identified in the alignment generated for the studied P. atrosepticum strains. Regarding the phylogeny computed on rpoS sequences, the vast majority i.e. 19 of the studied P. atrosepticum strains in addition to P. atrosepticum type strain LMG 2386 TS grouped into a single clade (Fig. 2). This taxonomic group was most closely related to P. atrosepticum IFB5654 (of BOX profile II) and IFB5672 (of BOX profile IV) strains, which clustered together. Interestingly, the next clade consisted solely of P. atrosepticum IFB5660 and IFB5661 (Fig. 2), being the only strains in this study attributed with BOX profile V ( Table 1). The observed tree branching in the part of Fig. 2 enclosing the Pectobacteriaceae reference strains reflected the established evolutionary relations between the included species, i.e. P. brasiliense LMG 21371 was located in the proximity to P. carotovorum IPO 200, while P. wasabiae CFBP 3304 was placed next to P. parmentieri SCC 3193. The tree was rooted on D. solani IFB0099.
Phenotypic analysis of P. atrosepticum strains originating from potato fields in Poland At first, we estimated the capacity of the representative P. atrosepticum strains to macerate potato tissue at different temperatures, i.e. at 20 °C or 28 °C (Fig. 3) in relation to the included Pectobacteriaceae reference strains (Table 1). It is worth to notice that P. atrosepticum strains of Polish origin showed higher capacity to macerate potato slices at the lower (Fig. 3A) than the higher temperature (Fig. 3B), as demonstrated by differences in the average rotten spot diameter (17.0 vs. 16.1 mm). Despite the fact that P. atrosepticum strains exhibited diverse BOX profiles, their ability to decay potato tissue was rather homogenous. Interesting observations concern the Pectobacteriaceae reference strains as P. atrosepticum type strain LMG 2386 TS and P. wasabiae CFBP 3304 showed also higher ability to macerate potato tissue at 20 °C than at 28 °C (Fig. 3). P. brasiliense LMG 21371 and P. carotovorum IPO 200 degraded potato more efficiently at 28 °C (Fig. 3).
Moving to the bacterial capacity to grow at different temperatures, all strains of P. atrosepticum and the other Pectobacteriaceae reference strains  (Table 2). In terms of P. parmentieri SCC 3193, the results of the conducted experiments were inconclusive ( Table 2). The bacterial ability to multiply in spite of higher salinity in the medium was examined. All the studied P. atrosepticum strains, similarly to P. wasabiae CFBP 3304, P. parmentieri SCC 3193 and D. solani IFB0099, were unable to grow in the medium supplemented with 5% NaCl (Table 2). Only P. brasiliense LMG 21371 and P. carotovorum IPO 200 reference strains were capable of multiplying in the medium of elevated salinity (Table 2).
Subsequently, the activities of PCWDEs being major virulence factors of soft rot Pectobacteriaceae were investigated. Concerning pectinase activity, the representative P. atrosepticum strains, independently of the assigned BOX profiles, revealed slight deviations in this feature (Fig. 4a). Notably, the mean pectinase activity of the Polish P. atrosepticum strains, which corresponded with the mean halo diameter of 18.52 mm, was higher than that of P. atrosepticum type strain LMG 2386 TS . There were just two reference strains, D. solani IFB0099 and P. carotovorum IPO 200, whose pectinase activity was either higher or comparable to that of the Polish P. atrosepticum strains (Fig. 4a).
Regarding the cellulase activity, the vast majority of the isolates of P. atrosepticum in addition to P. atrosepticum type strain LMG 2386 TS exhibited rather low, but homogeneous cellulase activity with no relation to the assigned BOX profiles (Fig. 4b).
The mean halo diameter measured for P. atrosepticum strains from Poland was 5.37 mm. There were just three strains from the pool of P. atrosepticum of a slightly higher cellulase production, namely IFB5669, IFB5671 and IFB5665. However, the reference strains of D. solani IFB0099, P. carotovorum IPO 200 and P. wasabiae CFBP 3304 showed a notably elevated cellulase activity in contrast to the P. atrosepticum strains (Fig. 4b).
In terms of protease production, there was some variability among the investigated P. atrosepticum strains (Fig. 4c). Primarily, one-third of the P. atrosepticum strains, i.e. IFB5655, IFB5658, IFB5664, IFB5668, IFB5653, IFB5666 and IFB5667, showed low capacity to produce proteases and the intensity of this trait was similar to that of P. parmentieri SCC 3193 reference strain. We observed also P. atrosepticum strains, i.e. IFB5669, IFB5660 and IFB5661, of high proteases activity, which was comparable to that of D. solani IFB0099 and P. wasabiae CFBP 3304. Notably, the above-listed P. atrosepticum strains of either elevated or diminished protease activities had been assigned to diverse BOX genomic profiles. Interestingly, P. atrosepticum type strain LMG 2386 TS turned out not to possess any proteases activity under the applied experimental conditions, contrary to all the other strains belonging to this species yielding the mean halo diameter of 11.8 mm. Undoubtedly, P. carotovorum IPO 200 and P. brasiliense LMG 21371 indicated the highest protease production in this analysis (Fig. 4c).
The ability of P. atrosepticum strains to produce lipases, which are involved in distortion of the plant cellular membranes, was studied next. It turned out that the representative P. atrosepticum strains isolated in Poland together with P. atrosepticum type strain LMG 2386 TS revealed similar activities of this group of enzymes (Fig. 5a) with a mean halo diameter of 12.0 mm. Moreover, no correlation between the BOX-based genomic patterns attributed to P. atrosepticum strains and their capacity to secrete lipases was observed. Interestingly, reference strains belonging to the other species of the Pectobacteriaceae family, i.e. P. brasiliense LMG 21371, P. carotovorum IPO 200, D. solani IFB0099, P. wasabiae CFBP 3304 and P. parmentieri SCC 3193, showed higher lipase activity than the tested P. atrosepticum strains (Fig. 5a).
In terms of the siderophore production efficacy, that is essential for iron scavenging under the limiting environmental conditions, there were several strains of P. atrosepticum, e.g. IFB5656, IFB5662, and IFB5655 and P. atrosepticum type strain LMG 2386 TS , of diminished production of these molecules (Fig. 5b). However, the observed differences were not statistically significant in relation to majority of the other P. atrosepticum strains. Curiously, the above-mentioned P. atrosepticum strains of lower siderophore production belong to the groups attributed with either BOX profile I or IV. In general, the mean halo diameter, corresponding to the siderophore production of the representative P. atrosepticum strains isolated in Poland, equaled 10.1 mm (Fig. 5b).
Regarding the bacterial ability to form biofilm, also high homogeneity among the studied P. atrosepticum strains was revealed (Fig. 5c). There was one strain, IFB5660 of BOX profile V, which showed elevated capacity (in a statistically significant manner) to generate biofilm. Overall, the mean absorbance value measured for the representative P. atrosepticum strains isolated in Poland was 0.034. Notably, the included reference strains classified to the other species of soft rot Pectobacteriaceae formed biofilm more efficiently than the investigated P. atrosepticum strains. In more detail, P. brasiliense LMG 21371, P. parmentieri SCC 3193, P. carotovorum IPO 200 and P. wasabiae CFBP 3304 generated this 3D structure more potently than even one of the included positive controls of efficient biofilm producers i.e. P. fluorescens CCM2115 (Fig. 5c).

Discussion
High dominance of Pectobacterium spp. over Dickeya spp. on the territory of Poland has been demonstrated in our previous study . Among the detected species of soft rot Pectobacteriaceae, the contribution of P. atrosepticum strains exceeded 26% in the period from 1996 until 2014 . It is worth to note that the majority of the SRP-oriented studies focused on emerging or newlyestablished Pectobacteriaceae species (Sarfraz et al., 2018;van der Wolf et al., 2014;Waleron et al., 2018Waleron et al., , 2019Zoledowska et al. 2018a, b). We examined the biodiversity and virulence of an important constituent of the SRP population in Poland being P. atrosepticum. Several studies on this species were also conducted in other countries due to the major contribution of P. atrosepticum to development of blackleg symptoms (van der Wolf et al., 2021a, b) and its status of a well-established pathogen of significant economic impact. So far, USA (Ma et al., 2018), Scotland (Toth et al., 2011), Canada (de Boer et al., 2012, Norway (Dees et al., 2017), Finland (Degefu, 2021), Northern Ireland (Zaczek-Moczydłowska et al., 2019), Pakistan (Sarfraz et al., 2020), Serbia (Arsenijević et al., 1994), Turkey (Ozturk et al., 2018), New Zealand (Pitman et al., 2008) or Nigeria (Monilola & Abiola, 2011) reported the occurrence and importance of this pathogen.
The diversity among P. atrosepticum strains isolated in different countries has been evaluated by using serological analyses, phage typing, PCR-RFLP and Random Amplification of Polymorphic DNA (RAPD) fingerprinting (Darrasse et al., 1994;Gross et al., 1991;Maki-Valkama & Karjalainen, 1994;Toth et al., 1999;Waleron et al., 2002). Application of the above-listed approaches, which differed in the number and types of markers used, resulted in the detection of 9 serogroups, 4 phage-typing groups, 2 PCR-RFLP profiles and nine RAPD profiles (Darrasse et al., 1994;Gross et al., 1991;Maki-Valkama & Karjalainen, 1994;Toth et al., 1999;Waleron et al., 2002). The first report on the phenotypic diversity of P. atrosepticum strains isolated in Poland was provided by Śledź et al. (2000). In that research, 8 P. atrosepticum biovars have been differentiated among approx. 750 isolates collected from potato fields in 1996 and 1997. Waleron et al. (2002) performed single gene (recA) PCR-RFLP-based genotyping on pectinolytic Erwinia (among them 13 strains of P. atrosepticum isolated in different countries were analysed) by digesting 730 bp fragments of the recA gene with each of the following 4 endonucleases: AluI, HinfI, TasI and Tru1I (Waleron et al., 2002). This study indicated 2 diverse PCR-RFLP profiles among P. atrosepticum strains, emphasizing significantly lower intraspecies variation in this taxon than in e.g. P. carotovorum (18 distinctive profiles) or Dickeya chrysanthemi (15 diverse profiles) (Waleron et al., 2002). In the current study variability among the studied 118 P. atrosepticum strains collected in 2013, 2014 or 2016 in Poland has been shown by applying rep-PCR and distinction of 6 BOX profiles. In addition, the analyses of the sequences of 3 single housekeeping genes, i.e. rpoS, gyrA and recA, indicated the presence of 16 SNPs in the rpoS gene, 1 in the gyrA gene and no SNPs in the recA gene fragment. Besides, we have shown that the rpoS-based phylogeny was discriminative enough to reveal biodiversity within the P. atrosepticum species. From the agronomical point of view, low genetic variability in P. atrosepticum species could be beneficial in terms of studies focused on development of eradication or detection methods for this pathogen. The association between a restricted host range of this bacterium and the observed minor genetic diversity can be hypothesized. In addition, our research revealed that the analysis of the repeated BOX sequences had higher discriminatory power for genotyping P. atrosepticum than the ERIC-and REPbased fingerprinting approaches.
Interestingly, we have demonstrated that the Polish P. atrosepticum strains multiply and macerate potato tissue under laboratory conditions more efficiently at 20 °C than 28 °C, which correlates with the research conducted earlier (Hasegawa et al., 2005). The herein investigated P. atrosepticum strains grew very well at 20 °C and 28 °C, while they were unable to multiply at 37 °C, which agrees with the published findings (Gardan et al., 2003;van der Wolf et al., 2021a, b). These observations are in accordance with the previously stated strict association between P. atrosepticum and the temperate climate conditions (Perombelon & Kelman, 1980;Toth et al., 2003). Even at subarctic and boreal regions of Greenland, the causal agents of soft rot and blackleg are identified as P. atrosepticum (de Neergaard et al., 2020). Further support for high virulence of P. atrosepticum at relatively low air temperatures was given by the fact that this species is able to infect earlier in summer (i.e. June or July) in comparison to other SRP Tsuyama & Sakamoto, 1952).
In terms of bacterial resistance to elevated salinity, the Polish P. atrosepticum strains were unable to grow at 5% NaCl, which contrasts with the previously published findings (Gardan et al., 2003;van der Wolf et al., 2021a, b). Nonetheless, one of the reference strains belonging to the P. brasiliense species did show capacity to multiply when subjected to an elevated salinity agreeing with what was reported by Li et al. (2019) for another P. brasiliense strain. Likewise, the reference P. carotovorum strain did grow at 5% NaCl, agreeing with de Haan et al. (2008). Bacterial tolerance to salinity has been associated with potent production of exopolysaccharides and biofilm formation capacity that limit the uptake of Na + ions (Shultana et al., 2020). The occurrence of salt-intolerant P. atrosepticum seems agronomically relevant in a view of the saline-related abiotic stress, which affects plant cultivation on 6% of the total and approx. 20% of the irrigated land (Samant & Jawali, 2016).
In general, the ability of SRP to macerate plant tissue results from efficient production of numerous PCWDEs. Also proper synchronization of gene expression and release of the resultant virulence factors is crucial for the pathogenesis of P. atrosepticum. Pagel and Heitefuss (1990) revealed that polygalacturonase, pectate lyase, cellulase, protease and xylanase were detected in P. atrosepticum-inoculated potato tubers sequentially. It was shown that development of soft rot and blackleg symptoms relies on pectinase activity. In P. atrosepticum, these enzymes are encoded by the following genes: pelB/C/Z (pel1/2/3), pelI, pelL /N, rhiE, pnl, pel, pehK, pehN (pehX), pehA, pehA2, paeY, pemA and pemB (van Gijsegem et al., 2021a, b). In our study, the P. atrosepticum strains isolated in Poland showed high, rather uniform pectinase production. Similarly, the cellulase activity of these strains was homogeneous, with the exception of P. atrosepticum IFB5669. These outcomes agree with the work of Almomani et al. (2013), who described low variation in cellulase production between Jordanian P. atrosepticum isolates in relation to P. carotovorum strains. Here, the included reference strains of P. carotovorum and D. solani had higher cellulase activity than the P. atrosepticum strains, which is in agreement with the study of Ozturk et al. (2018). Moving to protease production, all Polish P. atrosepticum strains showed notable activity of this group of enzymes, which contradicts with the reference P. atrosepticum strain LMG 2386 TS . This observation matches the results of Potrykus et al. (2014b) for a closely related SRP species, D. solani. In that research D. solani IFB0223 strain showed no protease activity. Possibly the proteases of SRP might either degrade proteins associated with plant defense mechanisms or enable acquisition of amino acids for building up bacterial proteins (Heilbronn & Lyon, 1990). However, it should be taken into consideration that according to the current knowledge the contribution of cellulases and proteases to the overall virulence of SRP seems minor in comparison to pectinases (Marits et al., 1999).
Likewise, the studied P. atrosepticum strains turned out to secrete lipases less efficiently than the other SRPs included. Similar observations on the lowest lipolytic activity of P. atrosepticum among the members of the Pectobacteriaceae family were previously reported by Babinska et al. (2021). This phenotypic feature was also studied by Zoledowska, et al. (2018a), who found small variation in lipase production between diverse P. parmentieri strains of Polish origin. Here, we also included one reference strain of P. parmentieri that excreted more fatty acids than all the investigated P. atrosepticum strains. In spite of obvious benefits for SRP resulting from digestion of host lipids for nutrient acquisition, the free fatty acids stimulate bacterial adhesion and therefore contribute to colonization and persistence of the pathogen (Stehr et al., 2003).
Additional molecules broadly engaged in plantmicrobe interactions are siderophores. Gorshkov et al. (2018) noted significant upregulation of the enterobactin-and TonB transporter-related genes during plant colonization by P. atrosepticum. Subsequent studies on ∆entA P. atrosepticum knockout mutants, deprived of the enterobactin biosynthesis gene entA, revealed impaired virulence of this strain on salicylic acid-primed plants (Gorshkov et al., 2021). In spite of solely carrying an essential trace element required as a cofactor for multiple enzymes, siderophores were shown to be involved in the transfer of other metals than iron (also sequestration of toxic metals). Thus, these molecules also play a signaling role and assure protection from oxidative stress (Gorshkov et al., 2021;Kramer et al., 2020). All the P. atrosepticum strains studied exhibited siderophore production, though some variation in the intensity of this trait was noted. Interestingly, in the current research D. solani being a chrysobactin and achromobactin producer showed higher siderophore activity than the enterobactin-synthesizing P. atrosepticum strains.
Furthermore, we provided insights into the biofilm formation of P. atrosepticum strains from Poland. All these strains generated biofilms similarly, though less efficiently than the other reference SRP strains. The herein reported outcomes are in agreement with Pérez-Mendoza et al. (2011) who disclosed low biofilm-formation ability of P. atrosepticum SCRI1043 under standard laboratory conditions. In that study the ability to form biofilms by P. atrosepticum was correlated with significant levels of c-di-GMP, a signaling molecule whose concentration is dependent on an interplay between the pathogen and the plant host or changes due to environmental or physiological stimuli (Pérez-Mendoza et al., 2011). Moreover, P. atrosepticum SCRI1043 was reported to be capable of forming multicellular biofilm-like arrangements named bacterial emboli in the plant xylem vessels (Gorshkov et al., 2014). It seems intriguing that the chemical structure of major component of P. atrosepticum SCRI1043 EPS, which are polymers providing resistance to external factors and involved in incorporation of bacterial cells within the biofilm 3D matrix, was similar to the structure of P. atrosepticum SCRI1043 O-specific polysaccharide (OPS), a constituent of the lipopolysaccharide of this bacterium (Gorshkov et al., 2017).
In conclusion, characterization of P. atrosepticum strains isolated in Poland during three growing seasons provided not only scientific proofs for the significance of this phytopathogen for potato cultivation under temperate climate conditions, but also pointed to notable uniformity among the bacteria from this taxonomic group. As the current approaches to control SRP are solely prevention-based, proper identification of the disease-causing agent in addition to gathering knowledge on its biodiversity, virulenceassociated traits and background mechanisms of interactions with the plant host, is crucial to build up a future perspective for development of procedures to combat the pest.

Declarations
Human and/or animals rights Not applied.
Informed consent Not applied.

Conflict of interest
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.