Cultures of CFF were initiated from a freezer stock of an isolate obtained from a symptomatic dry bean plant field in southern Alberta in 2005 (Fig. 2). The bacterium was identified as CFF as it was a small, Gram-positive rod that formed yellow pigmented colonies on modified CNS agar and was not soluble in 3% KOH. The frozen stock was revived by streaking onto brain heart infusion (BHI) agar (BD Biosciences, San Jose, CA, USA) and incubating for 48-72 h at 22 ± 2°C at 75% relative humidity. Colonies were used to prepare inoculum on BHI agar plates or in BHI broth (BD Biosciences).

Dry bean seed (cv. AC Polaris) was obtained from fields with a history of bacterial wilt symptoms. The seed lot used in this study was colonized by a yellow variant, and the seed lot was divided into four categories based on visual appearance. To confirm the negative impact of seed infection with emergence, the symptomatic seed was sorted according to the following categories: (1) yellow discoloration and shrivelled, (2) yellow discoloration, not shrivelled, (3) brown discoloration, and (4) no discoloration. Emergence bioassays were performed in 500 ml moistened potting medium, Pro-Mix (Premier Tech Home and Garden, QC, Canada) dispensed into 15 cm diameter pots. Five seeds were sown in each pot with five biological replicates for each category. The 20 pots were arranged in a completely randomized design in a greenhouse set at 25 ± 2°C for 18 h with light and 18 ± 2°C for 6 h in the dark. Emergence counts were taken weekly, and the final emergence (after 4 weeks) was reported as the percent of seedlings emerged. This experiment was repeated two more times with similar results.

Five seeds from category 2 (yellow discoloration, not shrivelled) were selected for scanning electron microscopy (SEM). Additionally, 6-week-old plants grown from category 2 seeds were evaluated by excising 1-4 cm² samples of infected plant tissue with a scalpel or razor blade. Sample fixation was performed using a standard approach (Bozzola and Russell, 1992) as follows; sample tissues were covered with primary fixative (3% glutaraldehyde + 1.6% paraformaldehyde in 0.2 M sodium cacodylate buffer, pH = 7.5) and incubated at room temperature for 2 to 4 h. Tissues were rinsed three times (15 min each) in 0.2 M sodium cacodylate buffer. Tissues were then added to 1-3 ml (or just enough to cover plant tissues) of 1% osmium tetroxide and incubated at room temperature for at least 1 h. Rinsing was repeated and tissues were dehydrated through a graded ethanol series (serial rinses for 10 min each at the following concentrations: 35%, 50%, 70%, 85%, 95%, and 100%). Samples were post-fixed with osmium tetroxide and critical point dried. All samples were mounted on aluminium stubs, sputter coated and scanned in a Philips C-60 ESEM (Philips/FEI Co., Hillsboro, OR, USA).

Culturing of biofilms and planktonic cells, and disinfectant efficacy testing, was performed in accordance to the methods and procedures outlined in ASTM E2799-17 (Standard Test Method for Testing Disinfectant Efficacy against Pseudomonas aeruginosa Biofilm) (ASTM International, 2017), using the MBEC assay device, with modifications as previously described in (Harding et al., 2011). Briefly, the bacterium was grown in BHI broth and biofilms were cultured by inoculating each well of a 96-well MBEC plate with 150 μl of CFF inoculum made from a broth culture of 10⁵ colony forming units per ml (cfu/ml) in OSB and incubated on a humidified gyrorotary shaker at 110 rpm at 25 ± 2°C for 24 h. The biofilms were exposed to a 96-well biocide challenge plate prepared separately for each biocide. Each challenge plate included 10 replicates at each of eight concentrations of one biocide (80 wells), sterility control (3 wells), neutralizer function control (2 wells), neutralizer toxicity control (2 wells), non-treated control (4 wells), and growth control (5 wells). The MBEC assay plate lids, with biofilms attached, were placed into the biocide challenge plates for 30 min. They were then serially transfered to a sterile distilled water rinse plate (200 μl per well) for 1-2 min, a neutrlizer plate (200 μl per well) for 24 h, and then placed into sonicator for 30 min to disaggregate the biofilms. A serial dilution and spread plate on OSA were performed from each well to determine cfu/ml. For minimum biocidal concentrations (MBC) calculations, 20 μl volumes from each well of the challenge plates were transferred to a new plate containing neutralizer and enumerated as described as described above with the exception that sonication to disagregate the cells was not required. All broth, agar and media ingredients were obtained from (BD Biosciences) with the exception of the universal neutralizer solution which was prepared as follows: 1.0 g L-histidine (Sigma, St. Louis, MO, USA), 1.0 g L-cysteine (Sigma), 2.0 g reduced glutathione (Sigma) in 20 ml double distilled water filtration sterilized through 0.22 μm diameter pore size filter (Corning Inc., Kaiserslautern, Germany) and stored at −20°C. MBEC Assay plates were provided by Innovotech Inc. (Edmonton, AB, Canada). The concentrations of bactericidal chemicals were based on preliminary efficacy results (not shown). The chemicals were selected primarily to evaluate the efficacy of oxidized silver compounds against CFF biofilms with copper compounds and streptomycin as standards for comparison. Stock solutions were prepared by adding the raw product into sterile water to reach the desired concentration as follows: copper hydroxide (2,000 μg/ml), copper sulfate (2,000 μg/ml), oxysilver bisulfate (1,000 μg/ml), oxysilver nitrate (1,000 μg/ml), and streptomycin (1,024 μg/ml). Solutions were mixed thoroughly and stored at 4°C until used. Stock solutions were serially diluted by 50% by combining 0.4 ml of solution with 0.4 ml of sterile water.

Copper hydroxide, copper sulfate and streptomycin were obtained from Sigma Aldrich (Oakville, ON, Canada), oxysilver nitrate and oxysilver bisulfate were provided by Innovotech Inc.

The effective concentrations to cause 50% inhibition (EC₅₀) and 95% inhibition (EC₉₅) were calculated using Quest Graph EC50 Calculator (https://www.aatbio.com/tools/ec50-calculator) (AAT Bioquest, 2019). Student t-tests (P ≤ 0.01) were performed in Microsoft Excel 2016 (Mississauga, ON, Canada). ANOVA was performed using Statistics Kingdom online calculator (https://www.statskingdom.com/180Anova1way.html, 2022).

Seedlings emerged from the soil 5 to 10 days after sowing. The category 1 seed (completely yellow and shrivelled) had the lowest emergence (0%), followed by category 2 (some yellowing), which had 20% emergence, while category 3 (brown discoloration) had 44% emergence, and category 4 (no disoloration or shrivelling) had the highest emergence (100%) (Fig. 3). The differences between the means of each category were statistically significant (P ≤ 0.01) according to pair-wise comparisons using a Student’s t-test (Table 1).

Bacterial growth in planta was visualized using SEM. Bacteria were clearly seen attached to the seed coat of discolored bean seeds. The bacteria were rod-shaped, and were not observed as solitary cells, but were dramatically clustered in biofilms (Fig. 4). Cells were clustered in multi-layer colonies and encased in what appeared to be the remnants of exopolymeric substances (EPS) (Fig. 4). Matrix-associated clustered layers of bacterial cells were not always densely packed, but had numerous openings and channels (Fig. 4). Surface-associated biofilm colonies were seen in categories 1 and 2, but were absent in categories 3 and 4 (not shown).

Leaf tissues from plants grown from category 2 seeds were also visualized via SEM. Occasionally, modest aggregates of small rod-shaped bacteria were seen on the surfaces of leaf tissues along with what appeared to be an EPS (Fig. 5). A number of internal vessels showed aggregates of rod-shaped bacteria within them (Fig. 5). Furthermore, colonization and occlusion of xylem elements with small rod-shaped bacteria within presumptive EPS was observed in category 2. Thus, evidence of biofilm, or pre-biofilm, formation by CFF cells was visible on seed surfaces, leaf surfaces and within vascular elements. No evidence of bacterial biofilms were observed in categories 3 and 4. Leaf tissues for category 1 could not be evaluated for biofilms on/in plants because emergence levels were zero in greenhouse pots.

The growth of CFF in the MBEC Assay reactor was evaluated and the bacterium was able to achieve 5 log₁₀ colony forming units per mL (cfu/ml) within biofilms in 48 h. When challenged with bactericides, planktonic cultures were reduced to zero (i.e., total kill was achieved), with MBCs of less than 10 μg/ml (Fig. 6). ANOVA of the MBC values revealed no significant differences (P ≤ 0.05) between treatments (f-ratio value = 0.12059, and P = 0.974343). The counterpart CFF cells within biofilms were not so easily killed as MBEC curves consistently had higher numbers of survivors than MBC curves, and much higher concentrations were required to reduce the populations to zero or near zero (Fig. 6). For example, all MBC values were less than 10 μg/ml, but some MBEC values were more than 100 times greater. The bactericides with the largest gap between the MBC and MBEC curves were copper hydroxide (CuOH), oxysilver bisulfate (OSBS), and oxysilver nitrate (OSN). For these three bactericides, a Student’s t-test confirmed that the MBEC curve diverged significantly from the MBC curve between zero and 200 μg/ml (Table 2). ANOVA of the MBEC values revealed no significant differences (P ≤ 0.05) between treatments (f-ratio value = 0.30369, and P = 0.87378).

To further assess the bactericidal treatments’ efficacies, the EC₅₀ was calculated for each. Copper hydroxide (Cu(OH)₂) (0.32 μg/ml) and OSBS (0.49 μg/ml) had the lowest EC₅₀ values followed by OSN (6.42 μg/ml). Streptomycin (15.51 μg/ml) and CuSO₄ had the highest EC₅₀ values (68.63 μg/ml), as seen in Fig. 7.

The EC₅₀ results are valuable for comparing efficacies of the bactericides, however the portion of the curve being evaluated at 50% reduction is not very revealing when viewed from a practical perspective. For example, when dealing with bacterial wilt, one would like to reduce populations by ≥95%, because this is the level of reduction necessary to prevent disease spread. For this reason, the EC₉₅ value was also calculated for each bactericide in order to evaluate the MBEC curves within the window that reductions in populations become more practically meaningful. The EC₉₅ values show that OSBS (26.89 μg/ml) and streptomycin (95.35 μg/ml) were the most practically effective bactericides for reducing CFF biofilms by ≥95%, at the lowest concentrations, whereas the other three required concentrations from nearly 500 to over 1,300 μg/ml (Fig. 8).