Evaluation of Pseudomonas fulva PS9.1 and Bacillus velezensis NWUMFkBS10.5 as Candidate Plant Growth Promoters during Maize-Fusarium Interaction

Based on in vitro assessments, molecular and chemical analysis, Pseudomonas fulva PS9.1 and Bacillus velezensis NWUMFkBS10.5 are candidate biocontrol agents for plant disease management including maize fusariosis, a disease caused by members of the Fusarium species. This in vivo study evaluated the bio-protective potential of the aforementioned rhizobacteria strains on maize against the proliferation of the pathogenic fungus Fusarium graminearum (Fg). The study results show that the bacterized plants were not susceptible to Fg aggression and the antagonists displayed the capability to proliferate in the presence of other likely competing microflora. The screen-house data also suggest that the presence of resident soil microbiota impacted the activity of antagonists (PS9.1 and NWUMFkBS10.5). This variation was recorded in the soil treatments (sterilized and unsterilized soil). In all the experimental periods, bacterized maize plants with or without Fg inoculation significantly (p = 0.05) grew better in unsterilized soil. Besides, during the experimental periods, all the consortia treatments with or without Fg infection regardless of the soil used demonstrated appreciable performance. The result of this study suggests that the microbial agents can actively colonize the surface of their maize plant host, improve plant growth, and suppress the growth of phytopathogens. Considering their overall performance in this screen-house evaluation, P. fulva PS9.1 and B. velezensis NWUMFkBS10.5 have potential for field applications. All safety issues regarding their use under field conditions and risks associated with their extended-release into the environmental will, however, be assessed prior to further bioformulation, field investigation, and scale-up.


Introduction
Microflora that inhabit or colonize the rhizosphere can be classified based on the effects they have on plants, and plant roots can serve as a portal of entry for both beneficial and pathogenic micro-organisms that influence plant growth and development. This zone where root activity influences the biological interaction taking place between plant, soil, and resident flora significantly is referred to as the rhizosphere [1]. The influence of the beneficial micro-organisms present in the root zone on plant growth has been investigated for decades and these beneficial influences are exerted through direct and indirect mechanisms. The activities of these rhizobacteria could either result in the stimulation of plant growth or protection of the plant against pathogen attack [2][3][4]. Plant growth promotion involves the direct secretion of plant growth-regulators such as auxin, while biocontrol involves the production of metabolites such as siderophores, antibiotics, and hydrogen cyanide, respectively [5,6].

Compatibility of Test Antagonists, Seed Germination Test, and Seed Treatment Preparations
In the compatibility tests, there was no area of inhibition between the two test antagonist-bacteria (PS9. 1 and NWUMFkBS10.5). Consequently, the test antagonist-bacteria were considered compatible with one another. During the seed germination test, all the seedlings tested sprouted healthily. Two hundred seeds were grown and they all sprouted. Growth parameters were, however, not recorded for the seed germination test ( Figure S1). All the screen-house treatment combinations and soils used (sterile and unsterile) are shown in Table 1. Table 1. Pot experiment treatment combinations for soils used (sterile and unsterile).

Harvest of Screen-House Pot Experiments Conducted over Three Experimental Periods
For the first experiment harvested at the V4-V5 stage, although plant growth was significantly (p = 0.05) retarded in the maize pots treated with only Fg spores and roots lacked vigor (Ms + P; Mus + P), no rot or wilting was observed ( Figure 1; Table 2). In all the pots with non-bacterized seeds (Ms + P; Mus + P; Ms; Mus), the primary and lateral roots were not fibrous despite watering and the mesocotyl were unhealthy. Overall, better plant growth and pathogen suppression were recorded in the unsterilized soil. Although maize pots having treatments with NWUMFkBS10.5 (B + Ms; B + Mus) had better shoot length both in sterile and unsterilized soil, the fresh plant weight of the untreated plants (Ms; Mus) was significantly (p = 0.05) higher than all the other treatments (Table 2). Apart from the maize pots treated with only Fg spores, the consortia treatments with Fg spores (AB + P + Mus; AB + P + Ms) had the lowest shoot length. In addition, consortia treatments (AB + Mus; AB + Ms) had longer roots than all the other treatments and, from visual observation, plants treated with PS9.1 had more fibrous roots and root hairs ( Figure 1).

Harvest of Screen-House Pot Experiments Conducted over Three Experimental Periods
For the first experiment harvested at the V4-V5 stage, although plant growth was significantly (p = 0.05) retarded in the maize pots treated with only Fg spores and roots lacked vigor (Ms + P; Mus + P), no rot or wilting was observed ( Figure 1; Table 2). In all the pots with non-bacterized seeds (Ms + P; Mus + P; Ms; Mus), the primary and lateral roots were not fibrous despite watering and the mesocotyl were unhealthy. Overall, better plant growth and pathogen suppression were recorded in the unsterilized soil. Although maize pots having treatments with NWUMFkBS10.5 (B + Ms; B + Mus) had better shoot length both in sterile and unsterilized soil, the fresh plant weight of the untreated plants (Ms; Mus) was significantly (p = 0.05) higher than all the other treatments (Table 2). Apart from the maize pots treated with only Fg spores, the consortia treatments with Fg spores (AB + P + Mus; AB + P + Ms) had the lowest shoot length. In addition, consortia treatments (AB + Mus; AB + Ms) had longer roots than all the other treatments and, from visual observation, plants treated with PS9.1 had more fibrous roots and root hairs ( Figure 1).   The harvest of the second experiment showed that overall, the bacterized seed performed better in unsterilized soil; treatments in unsterilized soil without Fg showed a higher increase (p = 0.05) in plant vigor than plants in sterilized soil (Figure 3a(i,ii),b). The controls (maize pots treated with only Fg spores) had retarded growth as expected. Growth was also retarded in the non-bacterized maize seedlings in sterilized soil. The harvest of the second experiment showed that overall, the bacterized seed performed better in unsterilized soil; treatments in unsterilized soil without Fg showed a higher increase (p = 0.05) in plant vigor than plants in sterilized soil (Figure 2a(i,ii),b). The controls (maize pots treated with only Fg spores) had retarded growth as expected. Growth was also retarded in the non-bacterized maize seedlings in sterilized soil.
(a) During the harvest of the third experimental period (Table 3), plant growth and fungal suppression after 90 days of infecting maize with Fg spores were recorded. Overall, all the treatments in unsterilized soil showed the best performance. NWUMFkBS10.5 treatments in unsterilized soil (B + Mus) without Fg inoculation, had the highest performance in all the parameters evaluated. The consortia (AB + Mus + P) with Fg inoculation had the best performance when compared with all other consortia treatments (B + Ms + P; B + Mus + P; A + Mus + P; A + Ms + P; AB + Ms + P) with Fg During the harvest of the third experimental period (Table 3), plant growth and fungal suppression after 90 days of infecting maize with Fg spores were recorded. Overall, all the treatments in unsterilized soil showed the best performance. NWUMFkBS10.5 treatments in unsterilized soil (B + Mus) without Fg inoculation, had the highest performance in all the parameters evaluated. The consortia (AB + Mus + P) with Fg inoculation had the best performance when compared with all other consortia treatments (B + Ms + P; B + Mus + P; A + Mus + P; A + Ms + P; AB + Ms + P) with Fg inoculation. As observed, the plants with only Fg treatments (Mus + P; Ms + P) without any bacterization dried and died off three weeks after seeding (Figure 4a). The primary and lateral roots were not fibrous despite watering in the pots with non-bacterized seeds (Figure 4b). The bioprotective effect of both antagonists on the root system and tassel development was, however, observed in other treatments (including seeds bacterized alone, seeds bacterized with Fg, and consortia) (Figure 4c).
inoculation. As observed, the plants with only Fg treatments (Mus + P; Ms + P) without any bacterization dried and died off three weeks after seeding (Figure 3a). The primary and lateral roots were not fibrous despite watering in the pots with non-bacterized seeds (Figure 3b). The bioprotective effect of both antagonists on the root system and tassel development was, however, observed in other treatments (including seeds bacterized alone, seeds bacterized with Fg, and consortia) (Figure 3c).

Discussion
Management of grain fusariosis remains unalleviated and it might be necessary to integrate multiple plant disease approaches including efficient cultural practices, use of resistant cultivars for cultivation, or addition of a low concentration of fungicide or a combination along with the biocontrol agents to see how efficient they would be. A combination of several disease management practices has become the popular approach to managing the continued incidence of cereal grain fusariosis [25][26][27]. In this study, the influence of two candidate biocontrol rhizobacteria (P. fulva PS9.1 and B. velezensis NWUMFkBS10.5) on maize development in the presence of a phytopathogen was observed. However, to ascertain that the bacteria antagonist treatments were the only source of nutrients received by the plants during their germination period, no additional external fertilization or fungicide treatment was applied during the planting periods. Our results demonstrate that inoculation of the bacteria strains (PS9.1 and NWUMFkBS10.5) independently or in consortia did not only improve the maize plant growth under controlled conditions, but also ameliorated the detrimental effects of the Fg pathogen on the maize plant growth. This bioprotection against Fg aggression may be attributed to the antibiotic secreting potential of the bacteria antagonists or possible induction of systemic resistance in maize by the isolates [28,29]. This might be part of a further study to identify the responses elicited by maize as a result of treatments with these antagonists.
From the results, treatments in unsterilized soil with or without Fg pathogen showed a higher increase (p = 0.05) in plant vigor than plants in sterilized soil., This could be due to the influence or presence of transient microflora. Resident microflora has been implicated in plant growth promotion and biocontrol of phytopathogens in disease suppressive soils [30,31]. Although the antagonists exhibited compatibility in vitro, the consortia treatments (where Fg infection was excluded) had a slightly lower influence on plant growth and Fg suppression when compared to the singular bacteria mix treatments (B + Ms+; B + Mus; A + Ms; A + Mus + P/A + Mus + P). The consortia, however, enhanced plant growth and suppressed Fg activity better where Fg infection was included in the treatments. Whilst antagonist PS9.1 possesses some plant-growth-promoting and biocontrol compounds, rhamnolipids, pyoverdine, and rhizomide [22,32], the antagonist NWUM-FkBS10.5 performed significantly (p = 0.05) better than PS9.1 in all the treatments. This could be attributed to the action of its multiple plant-growth-promoting and biocontrol biosynthetic compounds, macrolactin, bacillibactin, mersacidin, bacilysin, surfactin, difficidin, iturin, and fengycin [21,33]. A large number of the Bacillus and Pseudomonas spp. (e.g., Bacillus velezensis and Pseudomonas putida) harbor multiple beneficial genes in their genome [34] that confer on them better proliferative potential. The results of this study agree with previous studies seen under greenhouse and field trials in which species within the genera Pseudomonas and Bacillus suppressed the growth and aggression of Fusarium pathogens in cereal cultivars [35][36][37][38][39][40][41] Despite the study not evaluating disease incidence and severity, the results of this investigation show that the bacterial treatments enhanced maize plant growth compared to the Fg inoculated controls and the untreated and non-bacterized (Mus; Ms) controls. In the non-bacterized Fg inoculated treatments, evidence of pathogen aggression was observed-there were no improvements in plant growth parameters in comparison to bacterized plants. Wilting was observed in the non-bacterized-Fg inoculated treatments-some of the plants died off (Figure 4a). The unhealthiness of the maize pots treated with only Fg spores (without any bacterization) could be attributed to Fg systemic infection [9]. The observable survival of the bacterized maize plants is not unexpected since they are no longer distressed, and they possess a non-diseased root system as shown in Figure 4b. This observation correlates with the report of Pandey, et al. [42]. Treatments with PS9.1 also had better root systems which were seen throughout all the pot experiments. We observed some discolorations in the tassels from the plants harvested from sterilized soils. Tassels from NWUMFkBS10.5 treated plants were significantly larger than the control and other treatments (Figure 4c). Besides, in the pot treatments with non-bacterized seeds, the primary and lateral roots were not fibrous despite watering and the mesocotyl were unhealthy.
Biocontrol inoculant formulations and their mode of delivery are important for their success under field conditions [1]. From the result we gathered, the root dip treatment of maize seedlings appears to be effective in conferring bioprotective ability on the maize. This correlates with the reports of Pal et al. [43] and Pereira et al. [44], in which pre-sowing of seedlings enhanced the activities of BCAs. The root dip treatment involves an inoculated release approach-microbial treatments influence plant development as they proliferate on and within plant parts. The root dip approach ensures that the microbial agents do not just actively colonize the surface of the host but may also become endophytic [45], possibly leading to the activation of the plant's defense system. Although this study neither characterized soil microbiota nor evaluated bacteria persistence, the result from the unsterilized soil treatments suggests the possibility of transient soil microbiota playing a crucial role in the maize plant's growth. Growth was significantly (p = 0.05) retarded in the non-bacterized maize seedlings in sterilized soil. Besides, the bioactivities of the antagonistic bacteria (P. fulva PS9.1 and B. velezensis NWUMFkBS10.5) in unsterilized soil are suggestive of their ability to persist and compete in the environment against resident microflora if they are to be considered as field biocontrol candidates.
In recent years the biosafety status of candidate plant growth-promoting organisms (PGPO) and biocontrol strains have become a topic of major concern [46][47][48]. Despite strain PS9.1's beneficial properties [22,32,49,50], some of its genetically related strains have also been recognized as opportunistic pathogens in previous studies [51][52][53][54][55][56]. Hence, PS9.1 can be considered a strain with overlapping biological impact, which poses a challenge for its use in planta. Because of the safety concerns regarding the release of candidate PGPO with possible overlapping (beneficial and deleterious) attributes [48], we recommend the cautionary use of strain P. fulva PS9.1. According to Keswani, et al. [48], strains of several bacteria genera exhibiting clinical and beneficial overlaps are still utilized in vitro as indicators of plant growth promotion and biocontrol, even though they are not commercially utilized because of the risk of pathogen outbreaks. We consider PS9.1 a prime candidate for the production of microbial synthesized bioinoculants without viable organisms since in silico genome analysis shows the bacterium also harbors biosynthetic genes for synthesis of pyoverdin, lankacidin C, and rhizomides (A, B, and C). Additionally, the candidate biocontroller (PS9.1) will likely be employed strictly for invitro investigative purposes under controlled environments. For example, studies involving the synthesis of agricultural metabolite additives lacking viable organisms will be conducted. We also intend to conduct further bioinformatic analysis of the PS9.1 sequenced genome [32], this should provide additional insights into its beneficial roles and limitations. Information gathered thereafter should be useful in evaluating the biotechnological importance of other beneficial plant-microbes with clinical overlaps.

Determination of Compatibility of Test Antagonists
The bacteria antagonists used in this study, P. fulva PS9.1 and B. velezensis NWUM-FkBS10.5, were previously isolated from the maize rhizosphere, identified based on the 16S rDNA gene sequence analysis (accession numbers MF098600 and KX353617.1), and analyzed by genome sequencing [21,22,32,33]. The Fg pathogen was provided as a gift. Bacterial cultures were preserved in 15% glycerol at -70 • C, Luria Bertani (LB) broth (Sigma Aldrich L3522) at 4 • C, and maintained on LB agar (Sigma Aldrich L3147) at 4 • C until needed. Fg was maintained on potato dextrose agar (PDA) (Sigma Aldrich P2182) at 4 • C until needed.
Before the pot cultivation (treatment of planting seeds and seeding), the compatibility of the rhizobacteria antagonists (PS9.1 and NWUMFkBS10.5) were determined through a dual-culture interaction test described previously by Dubey et al. [57] and Kumar, et al. [24], with slight modifications. An overnight pure culture of each test antagonist was subcultured on fresh LB agar plates. One hundred microliters of the suspension of each antagonist (OD 600 nm = 0.5) were then prepared and spot inoculated or streaked opposite each other on a LB agar for 48 h at 30 • C. Additionally, the PS9.1 antagonist was streaked on an overnight LB spread plate of antagonist NWUMFkBS10.5 and vice versa. The assay was incubated for 48 h at 30 • C. After incubation, plates were observed for overlapped growth of the antagonists and possible inhibition zones. The absence of inhibition zones and overlapping of growth indicated compatibility of the test antagonists while the presence of inhibition zone (if applicable) signified incompatibility.

Surface Sterilization of Maize Seeds and Seed Germination Test
Maize seeds variety DKC 73-72 (200 g) used in this study were obtained from NWK Limited (Econobuild) Mafikeng industrial area, North West Province, South Africa. To ensure the removal of fungicides from the seeds, they were washed in sterile distilled water. The seeds were further soaked in 0.75% Sodium hypochlorite solution for 5 min, followed by rinsing five times with sterile distilled water, and a final soaking on the fifth wash. The fifth wash was then inoculated on sterile nutrient agar plates to determine the sterilization efficiency. The surface disinfection was recorded as the absence of a colony-forming unit on the Nutrient agar (NA) plate (Millipore 70116).
The disinfected maize seeds were then subjected to a seed germination test based on the paper towel method (ISTA 2003) while the percentage germination (%) was calculated according to Abdul-Baki and Anderson [58]. One hundred maize seeds already lined with a paper towel (moistened with 10 mL of sterile distilled water) were arranged in three 1 L beakers (11.8 cm of side). The seeds were again covered with a paper towel (earlier moistened with 10 mL of sterile distilled H 2 O). The beaker was covered and incubated at room temperature for 4 days. Thereafter, the number of germinated seeds per beaker was counted to ascertain the germination percentage using the formula: where GP = germination percentage; NG = Number of Germinated seed; TNS = total number of seeds.

Collection of Soil for Pot Experiments
Planting soil was collected from the North-West University Animal Science Department agricultural planting area. Sterilization of a portion of the soil was done by dry heat for 1 week at 120 • C. Sterilized soil was plated on NA to confirm sterilization, and sterilization was continued until no growth was observed on the NA plates. Planting pots with dimensions 13 cm (diameter) × 10 cm (depth) were filled with 80 kg of both sterilized soil and unsterilized soil up to water-holding capacity. The sterile and unsterile soils were utilized to compare the persistence and competitiveness of the two rhizobacteria antagonists.

Pre-Germination of Maize Grains for Pot Experiments
Two hundred grams of disinfected maize grains presoaked in sterile distilled water were placed in a 1 L beaker previously wallpapered with a sterile paper towel moistened with 10 mL of sterile distilled water. The seeds were covered again with the moistened (10 mL sterile distilled water) sterile paper towel. The flask was covered and incubated at 30 • C for 5 days after which only pre-germinated seeds with 2 cm callus were used for the seed-root dip pot experiment.

Seed Treatments Preparations
A modified seed-root dip pot experiment by Cook, Bruckart, Coulson, Goettel, Humber, Lumsden, Maddox, McManus, Moore, Meyer, Quimby, Stack and Vaughn [45] was employed for seed bacterization. From an overnight LB broth culture of PS9.1 and NWUM-FkBS10.5, 20 µL of each isolate was transferred into 100 mL of LB broth in a 250 mL Erlenmeyer flask, and cultured for 3 days (28 • C) with continuous shaking at 150 g. Bacteria cells were recovered by centrifugation at 8000× g for 20 min and the supernatant was discarded. The pellet of each isolate was re-suspended in 100 mL sterile LB broth and optical density (OD) was adjusted to 0.5:600 nm. For single bacterization, 120 sterile pregerminated maize seeds were submerged in the 100 mL bacteria inoculum (OD 0.5:600 nm) of each treatment, and the bacterized mixture consisted of 120 pre-germinated seeds sub-merged in 100 mL of co-inoculated bacteria broth (50:50 v/v) at OD 0.5:600 nm. This was incubated for 2-4 h with continuous shaking at 100 g for homogenization and adherence of bacteria to seeds. Sixty sterile pre-germinated maize seeds were also infected with 10 7 spores mL −1 of the Fg pathogen. The overnight pre-germinated (120) bacterized seeds above were also air-dried and 60 grains were aseptically removed and submerged in the spore suspension of the Fg pathogen (10 7 spores mL −1 ). All the treatments (Table 1) were incubated overnight to allow for adherence of inoculum to the seeds.

Seed Cultivation and Planting Experimental Periods
A randomized complete block experimental design (with four replications per pot treatment) based on a modified protocol by Bacon and Hinton [59] was employed in this section. The germinated bacterized seeds, Fg inoculated seeds, and consortia were transferred to both soils (sterile and unsterile), with four seeds in one pot (pots in quadruplicate). Uninoculated seeds were used as a control and pots were watered with sterilized water. Furthermore, treatments consisting of 10 mL bacteria antagonist, 10 mL mixtures of both antagonists, and 10 mL of pathogen spore suspensions (10 7 spores mL −1 ) were applied to the plants using a sterile syringe after one week of seeding in pots according to the treatments listed above. Planting was conducted in three experimental periods. The first experiment to determine the bioprotective capability of PS9.1 and NWUMFkBS10.5 during maize germination lasted 2.5 weeks in January 2017 after which the plants were harvested at the V4−V5 stage. The second experimental period occurred after the second inoculation of plants with Fg, and was for 40 days during February 2017−March 2017 and the third experimental period was for 3 months July 2017−September 2017. However, during the third experimental period, treatments consisting of 10 mL bacteria antagonist, 10 mL mixtures of both antagonists, and 10 mL of pathogen spore suspensions (10 5 spores mL −1 ) were applied (following the experimental design and treatments) after 1 week of seeding, using a sterile syringe. During the third experimental period, the plants were harvested at the tasseling stage (VT). At the end of each experimental period, harvested plants were evaluated for growth and survival and growth parameters were recorded. The parameters measured at harvest were wet plant weight, shoot length, root length, and dry plant weight.

Statistical Analysis
Multivariate general linear model, Duncan Multiple Range Test, and Tukey test (highest significant different test (HSD)) were used to analyze and compare observed treatment means, pathogen-antagonist relationship, treatment effects, and interactions in SPSS statistical software (version 22) at the significance level of 5%. Prism 9.1.0 was used for further data analysis and visualization.

Conclusions
Integrating biocontrol approaches into the current cereal disease management practices will provide novel alternatives (in form of biofungicides) for cereal farmers and the grain industry for the control of inherent Fusarium pathogens. The significant findings of the study are the in planta potential of the P. fulva PS9.1 and B. velezensis NWUMFkBS10.5 andtheir capability to bioprotect maize plants against Fg fusariosis and their effectiveness in unsterilized soil. Considering biosafety issues, strain NWUMFkBS10.5 is the preferred candidate for the development of a biofertilizer, biostimulant, and biofungicide product. In as much as the maize germination was protected against Fg aggression, further work to determine the disease severity, reduction of mycotoxin contamination, and level of nutrient uptake by treated seeds after application of the rhizobacteria duo will be required to ascertain the complete bioprotective potential of these rhizobacteria strains.