Selection of Endophytic Beauveria bassiana as a Dual Biocontrol Agent of Tomato Pathogens and Pests

Endophytic fungi (EF) can enhance both plant growth and defense barriers against pests and pathogens, contributing to the reduction of chemical pesticides and fertilizers use in agriculture. Beauveria bassiana is an entomopathogenic fungus showing endophytism in several crops, often associated with a good capacity to limit the development of pests and disease agents. However, the diversity of the protective efficacy and plant response to different strains can be remarkable and needs to be carefully assessed for the successful and predictable use of these beneficial microorganisms. This study aims to select B. bassiana strains able to colonize tomato plants as endophytes as well as to control two important disease agents, Botrytis cinerea and Alternaria alternata, and the pest aphid, Macrosiphum euphorbiae. Nine wild-type isolates and one commercial strain were screened for endophytism, then further characterized for plant-growth promotion plus inhibition of disease development and pest infestation. Four isolates proved to have a good control activity against the biotic stressors tested, but only Bb716 was also able to promote plant growth. This work provides a simple workflow for the selection of beneficial EF, paving the way towards more effective use of B. bassiana in Integrate Pest Management (IPM) of tomato.


Introduction
Feeding the growing world population while decreasing the environmental impact of agriculture is one of the biggest challenges of our time that must be confronted in the present and future [1][2][3]. A major concern for the scientific community is to provide new strategies to guarantee food security plus the safety of agricultural products as well as applications to reduce chemical pesticides and fertilizers [4].
Among the explored alternatives, the use of beneficial microorganisms (BMs) is one of the main pillars to provide a green turn in agriculture farming systems worldwide due to their noteworthy potential to increase crop health and fitness plus limiting negative impacts on the environment [5,6]. The most relevant plant-benefit effects demonstrated by BMs are the control and induced resistance to pest and pathogen attack, increased tolerance to abiotic stress, improved plant nutrition, plus growth promotion [7][8][9][10][11].

Biocontrol of Tomato Foliar Pathogens-Test In Vitro
The plate confrontation assays visibly showed that all tested B. bassiana strains were able to inhibit the growth of both A. alternata and B. cinerea in vitro, demonstrating a clearing zone between the pathogen and the BCA colonies in all the experimental plates ( Figure  2).
In the case of B. cinerea, the mycelial radial growth of the pathogen was significantly lower in the presence of all 10 tested Beauveria strains (one-way ANOVA: F (  The colonization frequency of Beauveria in the different tissues collected from the same plant differed significantly among the tested strains and resulted higher in the roots than in the stems and greater in the stems compared to the leaves ( Figure 1B). The colonization frequency observed in the roots and in the stems was significantly higher in tomato plants treated with Bb716, Bb74040, and Bb688 strains (one-way ANOVA: roots, F (9, 50) = 2.249, p = 0.033; stems, F (9, 50) = 2.658, p = 0.013). In the leaves, it resulted significantly higher in Bb716-treated plants (one-way ANOVA: F (9, 50) = 3.598; p = 0.001). Taking into account both the colonization rate and frequency, the isolates Bb74040, Bb688, and Bb716 showed the best ability to endophytically colonize the tomato plants.

Biocontrol of Tomato Foliar Pathogens-Test In Vitro
The plate confrontation assays visibly showed that all tested B. bassiana strains were able to inhibit the growth of both A. alternata and B. cinerea in vitro, demonstrating a clearing zone between the pathogen and the BCA colonies in all the experimental plates ( Figure 2).
In the case of B. cinerea, the mycelial radial growth of the pathogen was significantly lower in the presence of all 10 tested Beauveria strains (one-way ANOVA: F (10,33) = 35.13, p < 0.0001) (Figure 2A), with the width of the inhibition zone variable depending upon the strain, ranging from 2.1 mm with Bb632 to 7.5 mm with Bb688 (one-way ANOVA: F (9,30) = 42.49, p < 0.0001) ( Figure 2B). The percent of growth inhibition (PGI%) was also highly variable depending on the tested strain; indeed, it ranged from 21% for Bb709, which was the least effective strain, to 65% for Bb688, which showed the highest biocontrol potential against B. cinerea along with Bb762 and Bb632, characterized by growth inhibition of 57% and 52%, respectively (one-way ANOVA: F (9, 30) = 16.05, p < 0.0001) ( Figure 2C).
Additionally, in the case of A. alternata, the radial growth of the pathogen mycelium was always significantly lower when placed in dual culture with B. bassiana (one-way ANOVA: F (10, 33) = 72,72, p < 0.0001) (Figure 2A). The clearing zone of growth inhibition was always evident except for the Bb632 strain, where it was barely noticeable, but highly variable depending on the tested strain, as it ranged from 1.6 mm for Bb709 to 9.3 mm for Bb716 (one-way ANOVA: F (9, 30) = 82.02, p < 0.0001) ( Figure 2B). The PGI was also strain-dependent, but it was reduced and less variable in comparison to that observed with B. cinerea, ranging from 22 to 40% for all the experimental strains (one-way ANOVA: F (9, 30) = 16.05, p < 0.0001) ( Figure 2C). The most effective strains in limiting the mycelial growth of A. alternata were Bb688 and Bb762, both with a PGI of 40%.

Selection of Potential EF and Biocontrol B. bassiana Strains
The selection of B. bassiana strains was based firstly on the ability of the tested isolates to endophytically colonize the tomato plant, then secondly on the biocontrol potential assessed in vitro with dual-confrontation tests with B. cinerea and A. alternata. Strains demonstrating 100% colonization of the plant roots were selected for further evaluation, and three strains with lower root colonization, such as Bb672, Bb682, and Bb758, were no longer considered ( Figure 1A).
Secondly, the most effective strains in counteracting pathogen growth in vitro were selected. The bioassay against A. alternata did not show noteworthy differences between the isolates; therefore, the biocontrol of B. cinerea was used to exclude the strain Bb709 with the lowest efficacy. Thus, the six most promising strains used thereafter were: Bb74040, Bb632, Bb633, Bb688, Bb716, and Bb762.
Pathogens 2021, 10, x FOR PEER REVIEW 5 cinerea, ranging from 22 to 40% for all the experimental strains (one-way ANOVA: F = 16.05, p < 0.0001) ( Figure 2C). The most effective strains in limiting the mycelial grow of A. alternata were Bb688 and Bb762, both with a PGI of 40%.

Selection of Potential EF and Biocontrol B. bassiana Strains
The selection of B. bassiana strains was based firstly on the ability of the tested isol to endophytically colonize the tomato plant, then secondly on the biocontrol potential sessed in vitro with dual-confrontation tests with B. cinerea and A. alternata. Stra demonstrating 100% colonization of the plant roots were selected for further evaluat

Plant Growth Promotion Assay
The overall effect of the B. bassiana endophytic colonization on the plant biometric parameters was highly strain-dependent and produced opposing effects. Only the Bb632and Bb716-treated plants were significantly greater than the untreated control, whereas a negative effect was noted for plants colonized by isolate Bb633 (one-way ANOVA: F (6, 98) = 4.227, p = 0.0008) ( Figure 3A). significant differences among five measured strains, whereas a decrease was reported in the Bb633-treated plants in comparison to the control (one-way ANOVA: F (6, 126) = 11.11, p < 0.0001) ( Figure 3B).
The plant dry weight was significantly higher in Bb632-and Bb716-treated plants, as noted for plant height, while no differences were demonstrated for all the other treatments (one-way ANOVA: F (6, 98) = 5.003, p = 0.0002) ( Figure 3C). Since strain Bb633 negatively affected plant growth, it was excluded from the subsequent biocontrol assays.

Biocontrol of Tomato Foliar Pathogens-Tests in Planta
The efficiency of endophytic B. bassiana to counteract the plant pathogen infections in vivo was variable depending upon the endophytic fungal isolate and the specific phytopathogen. Better effects were observed with the endophytic B. bassiana treatments on the biocontrol of B. cinerea infection in comparison with A. alternata. All strains had similar biocontrol activity, significantly reducing B. cinerea infection by 35% compared to the control (one-way ANOVA: F (6, 84) = 8.955, p < 0.0001) ( Figure 4A). The control of A. alternata disease symptoms was variable among the strains, whereby four out of five B. bassiana treatments (Bb74040, Bb762, Bb688, and Bb716) produced a substantial reduction in symptoms ( Figure 3B), and the best reduction in the tomato leaf spot attack by almost 40% was noted with Bb74040 (one-way ANOVA: F (6, 84) = 6.955, p < 0.0001) ( Figure 4B). The mean leaf area of the third, fourth, and fifth true leaf from each plant showed no significant differences among five measured strains, whereas a decrease was reported in the Bb633-treated plants in comparison to the control (one-way ANOVA: F (6, 126) = 11.11, p < 0.0001) ( Figure 3B).
The plant dry weight was significantly higher in Bb632-and Bb716-treated plants, as noted for plant height, while no differences were demonstrated for all the other treatments (one-way ANOVA: F (6, 98) = 5.003, p = 0.0002) ( Figure 3C). Since strain Bb633 negatively affected plant growth, it was excluded from the subsequent biocontrol assays.

Biocontrol of Tomato Foliar Pathogens-Tests in Planta
The efficiency of endophytic B. bassiana to counteract the plant pathogen infections in vivo was variable depending upon the endophytic fungal isolate and the specific phytopathogen. Better effects were observed with the endophytic B. bassiana treatments on the biocontrol of B. cinerea infection in comparison with A. alternata. All strains had similar biocontrol activity, significantly reducing B. cinerea infection by 35% compared to the control (one-way ANOVA: F (6, 84) = 8.955, p < 0.0001) ( Figure 4A). The control of A. alternata disease symptoms was variable among the strains, whereby four out of five B. bassiana treatments (Bb74040, Bb762, Bb688, and Bb716) produced a substantial reduction in symptoms ( Figure 3B), and the best reduction in the tomato leaf spot attack by almost 40% was noted with Bb74040 (one-way ANOVA: F (6, 84) = 6.955, p < 0.0001) ( Figure 4B).
The area of the necrotic lesions caused by the fungal pathogen infection was substantially reduced in the plants colonized with the B. bassiana endophyte. B. cinerea symptoms were significantly lower than the control in all instances, with strain Bb762 showing the greatest reduction in disease, almost 70%, relative to the untreated plants (one-way ANOVA: F (6, 76) = 5.954, p < 0.0001) ( Figure 4C). In the case of A. alternata, the foliar damage was significantly lower for all the strains except for Bb632 when compared to the control, with strain Bb74040 showing the strongest biocontrol effect (one-way ANOVA: F (6, 91) = 5.851, p < 0.0001) ( Figure 4D).
were significantly lower than the control in all instances, with strain Bb762 showing th greatest reduction in disease, almost 70%, relative to the untreated plants (one-wa ANOVA: F (6, 76) = 5.954, p < 0.0001) ( Figure 4C). In the case of A. alternata, the foliar damag was significantly lower for all the strains except for Bb632 when compared to the contro with strain Bb74040 showing the strongest biocontrol effect (one-way ANOVA: F (6,91) 5.851, p < 0.0001) ( Figure 4D).

Biocontrol of the Aphid M. euphorbiae-Test in Planta
Aphids that were reared on B. bassiana-treated plants for their entire lifespan were negatively affected in terms of survival and fertility in comparison to the control cohorts.
Adults of M. euphorbiae, feeding on tomato plants endophytically colonized by B. bassiana, showed a significantly lower survival compared to the adults fed on control plants regardless of the EF strain used (log-rank test: χ 2 = 64.27; p = 0.0001; dF = 5) ( Figure 5A). There was a clear significant reduction in survival when aphids were reared on plants treated with any of the five Beauveria strains in comparison to the control, (logrank test: Bb74040 − χ 2 = 42.47, p < 0.0001, dF = 1; Bb632 − χ 2 = 33.38, p < 0.0001, dF = 1; Bb762 − χ 2 = 27.21, p < 0.0001, dF = 1; Bb688 − χ 2 = 32.55, p < 0.0001, dF = 1; Bb716 − χ 2 = 29.93, p < 0.0001, dF = 1), which were not significantly different among themselves (log-rank test: p > 0.05) ( Figure 5A). spring of the initial 10 females reared on each plant), was detected in plants treated with Bb74040, in which a decrease of over 70% was shown. This treatment resulted as the most effective EF strain in terms of reducing fertility, and it was significantly different from Bb716, Bb762, and Bb688, but similar to strain Bb632. Dead adult specimens were collected throughout the experiment and among the over 200 cadavers none showed the white muscardine mycosis that is typically provoked by B. bassiana infection of the fungus-parasitized insects.

Discussion
In this work, nine wild-type isolates and one commercial strain of B. bassiana were tested for their capacity to colonize tomato cv. Dwarf San Marzano, act as biocontrol agents against Botrytis and Alternaria pathogens plus the Macrosiphum pest, as well as perform as a plant-growth promoter. The procedure developed to select promising BM strains of the entomopathogenic fungus Beauveria bassiana as endophytic dual biocontrol agents and plant biostimulators is presented in the experimental workflow design ( Figure  6). Initially, the capacity of the fungal strains to internally colonize the plant was assessed. Simultaneously, the ability to inhibit pathogens in vitro was tested. The most effective strains were subjected to in vivo bioassays that simulated the real-life plant-soil-microbe interactions in controlled conditions to determine the effect on plant growth, the control of the two foliar fungal pathogens and the aphid pest in vase experiments. In the end, among the 10 B. bassiana strains screened, only four isolates proved to have a noteworthy capacity as dual biocontrol agents providing promising protection to tomato from the harmful biotic agents tested. Similar effects were also noted in aphid fertility, with a significant reduction observed on B. bassiana-treated plants compared to control (one-way ANOVA: F (5, 72) = 16.02, p < 0.001) ( Figure 5B). Among the strains, the lowest number of newly laid nymphs (offspring of the initial 10 females reared on each plant), was detected in plants treated with Bb74040, in which a decrease of over 70% was shown. This treatment resulted as the most effective EF strain in terms of reducing fertility, and it was significantly different from Bb716, Bb762, and Bb688, but similar to strain Bb632. Dead adult specimens were collected throughout the experiment and among the over 200 cadavers none showed the white muscardine mycosis that is typically provoked by B. bassiana infection of the fungus-parasitized insects.

Discussion
In this work, nine wild-type isolates and one commercial strain of B. bassiana were tested for their capacity to colonize tomato cv. Dwarf San Marzano, act as biocontrol agents against Botrytis and Alternaria pathogens plus the Macrosiphum pest, as well as perform as a plant-growth promoter. The procedure developed to select promising BM strains of the entomopathogenic fungus Beauveria bassiana as endophytic dual biocontrol agents and plant biostimulators is presented in the experimental workflow design ( Figure 6). Initially, the capacity of the fungal strains to internally colonize the plant was assessed. Simultaneously, the ability to inhibit pathogens in vitro was tested. The most effective strains were subjected to in vivo bioassays that simulated the real-life plant-soil-microbe interactions in controlled conditions to determine the effect on plant growth, the control of the two foliar fungal pathogens and the aphid pest in vase experiments. In the end, among the 10 B. bassiana strains screened, only four isolates proved to have a noteworthy capacity as dual biocontrol agents providing promising protection to tomato from the harmful biotic agents tested.
The selected strains were Bb74040, isolated from a commercial product, plus three wild-type isolates (Bb762, Bb716, and Bb688), were able to negatively affect the survival and fertility of the sucking insect pest M. euphorbiae and to significantly decrease the symptoms caused by two foliar fungal pathogens: A. alternata and B. cinerea. Moreover, Bb716 strain, in addition to biocontrol, also was able to promote plant growth, significantly increasing plant height and dry weight, demonstrating to be an overall beneficial EF in a global sense (Table 1). These results underlined that strain screening is crucial in order to select valuable and efficient BMs since the diversity of the protective efficacy and plant response The selected strains were Bb74040, isolated from a commercial product, plus three wild-type isolates (Bb762, Bb716, and Bb688), were able to negatively affect the survival and fertility of the sucking insect pest M. euphorbiae and to significantly decrease the symptoms caused by two foliar fungal pathogens: A. alternata and B. cinerea. Moreover, Bb716 strain, in addition to biocontrol, also was able to promote plant growth, significantly increasing plant height and dry weight, demonstrating to be an overall beneficial EF in a global sense (Table 1). These results underlined that strain screening is crucial in order to select valuable and efficient BMs since the diversity of the protective efficacy and plant response to different strains can be remarkable and needs to be carefully assessed for successful and predictable use of B. bassiana in agriculture. Table 1. Summary of the overall biological activity of the five B. bassiana isolates: biocontrol effect against the pest M. euphorbiae and the pathogens A. alternata and B. cinerea and stimulation effect on plant growth. Biocontrol was evaluated on a scale of four, whereby "0" = no effect, "++" = decrease of the infestation or symptoms of infection, to "+++" = strong decrease of the infestation or symptoms of infection; and the plant-growth effect was evaluated, whereby "0" = no effect and "+" = positive stimulation effect. The results indicated that tomato plant tissues were endophytically colonized by B. bassiana following a repeated soil-drench treatment with a conidia suspension. All of the 10 experimental isolates were able to successfully infect and establish themselves in tomato roots, stems, and leaves. Endophytic colonization observed within the plant tissues  Table 1. Summary of the overall biological activity of the five B. bassiana isolates: biocontrol effect against the pest M. euphorbiae and the pathogens A. alternata and B. cinerea and stimulation effect on plant growth. Biocontrol was evaluated on a scale of four, whereby "0" = no effect, "++" = decrease of the infestation or symptoms of infection, to "+++" = strong decrease of the infestation or symptoms of infection; and the plant-growth effect was evaluated, whereby "0" = no effect and "+" = positive stimulation effect. The results indicated that tomato plant tissues were endophytically colonized by B. bassiana following a repeated soil-drench treatment with a conidia suspension. All of the 10 experimental isolates were able to successfully infect and establish themselves in tomato roots, stems, and leaves. Endophytic colonization observed within the plant tissues was much higher in the hypogeal tissues of the roots, where the colonization rate was close to 100% compared to the leaves, where it ranged from 10 to 50%. Recent studies have revealed that the colonization of B. bassiana is not uniform, as it remains concentrated in the zones where the inoculum is applied, even if the translocation towards more distal zones may occur but to a lesser extent [36,41]. In fact, the endophyte spread mainly relies on passive transmission, as it migrates within plant tissues due to water translocation [54][55][56][57]. This pattern was confirmed in the present study, where translocation from roots to leaves was observed for all the tested strains 14 days post-inoculum (dpi). This supports the hypothesis of systemic colonization since the inoculum was at the root level, and the tested fungal strains were isolated from stem and leaves, as reported by numerous studies [23,44,[58][59][60]. Further studies should be directed towards defining the endophytic colonization window, in order to understand if by using the current protocols of infection, B. bassiana is established as an endophyte until the end of the plant life cycle or if this relationship is just transient.

B. bassiana
Furthermore, the results presented have shown that several B. bassiana strains had a noteworthy capacity for biocontrol against the grey mold B. cinerea and the tomato leaf spot agent A. alternata both in vitro and in vivo as endophytes. In vitro, using plate dual-culture test, the presence of a clear zone of growth inhibition between the pathogen and the BCA colonies suggests that the type of interaction between the pathogens and B. bassiana may be of antibiosis [61,62]. This effect was already reported against B. cinerea, Cladosporium herbarum (Pers.) Link, Fusarium spp. Link, Gaeumannomyces graminis var. tritici J. Walker, and R. solani, but to our knowledge, this is the first report regarding A. alternata [39,[63][64][65][66][67][68][69]. The remarkable rate of pathogen growth inhibition implemented by some B. bassiana strains used in this study could be linked to the ability of these fungi to produce an abundant variety of bioactive metabolites with antimicrobial properties [70]. This was also confirmed in a recent study demonstrating the antifungal properties of the culture filtrate of B. bassiana against various phytopathogens, including Alternaria tenuis [71]. The most recognized metabolites produced by B. bassiana are oosporein, beauvericin, bassianolide, bassianin, beauveriolide, bassiacridin, and cyclosporine, and among them, beauvericin and oosporein evidenced remarkable antibiotic and antifungal properties and are probably responsible for the fungal growth inhibition observed in the bioassay presented here [39,[72][73][74][75][76][77][78].
In vivo, four strains, Bb74040, Bb762, Bb716, and Bb688, significantly reduced the symptoms caused by B. cinerea and A. alternata-with a decrease in the infection rate, respectively ranging from 32 to 40% and 21 to 41% with a reduction of the necrotic area ranging from 35% to 73% and 45% to 63%-and, to our knowledge, this is the first evidence of endophytic B. bassiana biocontrol against A. alternata. The reduction of disease symptoms in B. bassiana-treated plants could be explained by direct or indirect mechanisms linked to the presence of the endophyte within the plant tissues. As already mentioned, the plant colonization was highly concentrated at the root level, whereas the pathogens were inoculated on leaves, thus, the direct interaction between the BCA and the pathogens appears to be unlikely. In particular, one of the most effective strains, Bb762, colonized the leaf tissue only in 30% of the inoculated plants with a frequency of about 20%, but the treated plants were remarkably less affected by both pathogens, thus suggesting that the biocontrol mechanism underlying this effect could be indirect. One important indirect mode of action may be due to the above-mentioned secondary metabolites, which may be active in the plant tissues at a distance from the point that they are produced [79]. This action is well-known to occur in the genus Trichoderma, another biocontrol fungus, whereby the application of its active secondary metabolites leads to improvement in plant fitness, growth, and protection from harmful biotic agents [31,[80][81][82][83][84]. Another important mechanism of action in the biocontrol effect of endophytic B. bassiana may be the induction of plant resistance by ISR or SAR or a combination of both. ISR is an essential process known to be triggered by endophytic BM, by which the plant is primed for improved defense against a wide range of pests and pathogens [25]. One proof for ISR is the reduction of disease incidence in plant parts distant from the location of the inoculated beneficial agent, as observed in the bioassay reported in this work [18]. ISR in tomato plants by endophytic B. bassiana has been well documented in the literature [18,41,53,66,85]. Thus, it may be hypothesized that one of the diverse indirect mechanisms of action or a synergistic committance responsible for the observed biocontrol. Nonetheless, the biological processes underlying the biocontrol effect here described remains to be further investigated.
The activity of endophytic B. bassiana against the aphid M. euphorbiae was noted in all the five strains used in the bioassay, which equally acted as valuable biocontrol agents. Aphids reared for their entire lifespan on B. bassiana-colonized plants showed a strong decrease in survival and fertility. Since none of the adult aphid cadavers stored for three weeks showed symptoms of mycosis, it was assumed that this fitness reduction is linked to an indirect effect of the fungal endophytic colonization. One of the possible explanations is that of a toxic or antifeeding effect of the treated plants. In fact, the deterrent and toxic effect of colonized plants against insects was reported more than once in literature and could be again associated with the production of fungal secondary metabolites [47,53,56]. This is thought to be one of the modes by which biocontrol is triggered by endophytic entomopathogenic fungi against pests [18]. On the other hand, another mechanism that several authors reported is once again ISR in which the endophyte induces the activation of plant defenses that suppress insects [14,40]. As evidenced by a recent study, the resistance of tomato plants root-colonized by Trichoderma against M. euphorbiae is related to an increase of plant endogenous defense processes attested by the up-regulation of the transcripts coding for ET, JA, SA, and PR proteins that are likely responsible for a primed state of the plants, as described for ISR and SAR [86]. It can be hypothesized that a similar mechanism underlies the biocontrol effect that was observed in B. bassiana-colonized plants, but further investigation need to be undertaken to shed light on this important aspect. Endophytic B. bassiana was reported to control various species of aphids on different crops, such as Myzus persicae Sulzer infesting chili pepper [87], A. gossypii infesting cotton [88], or Sitobion avenae Fabricius infesting maize [89], but to our knowledge, this is the first evidence of endophytic B. bassiana controlling M. euphorbiae infestation.
Bb74040 strain, isolated from a commercial product, has been already introduced as an endophyte in numerous crops including tomato, but to our knowledge, this is the first report in which the tomato endophytic colonization with this strain was achieved with a soil watering of spore suspension [90]. Its application as an endophyte determines multiple beneficial effects on various crops, such as plant-growth promotion in wheat and fava bean [64,91], reduction of disease incidence by Fusarium spp. in sweet pepper and wheat [64,92], Zucchini yellow mosaic virus (ZYMV) in squash [93], and Plasmopara viticola (Berkeley and Curtis) Berlese and de Toni in grapevine [94]; it also decreased the infestation of insect pests, such as P. ficus, E. vitis, and O. sulcatus on grapevine [95,96]; Delia radicum L. on cabbage [97]; and M. persicae on pepper [98,99]. In recent work, Klieber and Reineke pointed out the endophytic capacity of this strain applied to tomato plants by spraying the solution of the fungal-based commercial product directly on the leaf surface and its activity against the tomato leaf-miner Tuta absoluta (Lepidoptera: Gelechiidae) [90]. In conformity with this study, the results presented showed that endophytism by strain Bb74040 induced resistance against pests without affecting plant growth, suggesting that products based on B. bassiana are highly versatile and well fit in the context of integrated pest/pathogen management.
Endophytic B. bassiana was proposed as a dual control agent in a recent review by Jaber and Ownley that underlined the efficiency of this BM to control both pests and pathogens and to induce systemic resistance in numerous crops [18]. This is corroborated by the findings presented here, indicating that two consecutive soil drenches with B. bassiana may protect tomato plants from foliar pathogens and aphid pest attacks; thus, the selected isolates represent amenable candidates for a dual biocontrol strategy. Moreover, a soil-drench treatment may have a direct effect on pests and pathogens inhabiting the rhizosphere, providing extra-protection to plants, and be beneficial also from a nutritional perspective, due to the known capability of endophytic entomopathogens, such as Metharizium, to provide nitrogen transfer from insects to plants [100], although these beneficial side-effects of the treatment need further investigation for tomato and Beauveria. These aspects are particularly relevant in the framework of integrated pest and pathogen management programs. The fungus may colonize plants and provide protection from a wide range of biotic stress; thus, the use of B. bassiana could be considered as a preventive control measure, meeting the needs for the implementation of food safety and environmental protection and conservation. Nonetheless, since B. bassiana is a known producer of many secondary metabolites that are possible mycotoxins, it is essential to assess the risk that the use of this fungus as an endophyte may represent for the environment and the health of humans and animals.
Based on the four endophytic-biocontrol fungal isolates selected in this study, future investigations could focus on developing bioformulations containing a combination of the best B. bassiana performers for biocontrol of pathogens plus the best biocontrol of pests plus the best plant biostimulant. Furthermore, a microbial consortia consisting of different BMs, such as entomopathogens, other biocontrol agents, mycorrhizae fungi, and plant-growth promoting rhizobacteria with various crop protection and production properties, could be considered by mixing compatible microorganisms to produce a single, multi-purpose agricultural product [101].

Conclusions
This investigation demonstrated that the endophytic and entomopathogenic fungus B. bassiana is a valuable, prospective biocontrol agent against both insect pests or fungal pathogens of tomato and a potential plant biostimulant. This paves the way for the development of a crop-improvement strategy with a single BM application that offers multiple beneficial effects to the plant. The screening process that was developed showed that the biocontrol efficiency and plant response to different strains can be remarkably diverse; thus, a solid workflow procedure for strain selection is highly necessary to carefully chose the most promising plant BM. However, it should be mentioned that the biocontrol efficacy of pathogens and/or pests plus the plant benefits observed in this study may be influenced by the controlled conditions imposed in the experimental design. In fact, abiotic factors, such as temperature, have been noted to significantly affect the important favorable characteristics of diverse strains of Trichoderma tested as the biological control agents as well as their culturability [86]. The next phase in the present BM selection process should involve testing in the field, in the "real-world" cultivation environment to determine if the selected isolates maintain their capacity as useful agricultural products. Overall, this work suggests that dual biocontrol exerted by endophytic B. bassiana represents a promising tool worthy of consideration when defining new IPM strategies. This application may represent a valuable alternative to the use of chemical pesticides in agriculture that aids in maintaining the biological equilibrium of the agroecosystem.

Fungal Isolates: Origins, Identification, and Culture Methods
Ten different strains of B. bassiana (Bb) were used. Nine wild isolates (Bb716, Bb633, Bb672, Bb682, Bb688, Bb709, Bb632, Bb758, Bb762) were received from the Agricultural University of Plovdiv (Bulgaria), while one commercial strain, ATCC 74,040 (hereby referred to as Bb74040), was isolated from the commercial product Naturalis ® (CBC S.r.l., Grassobbio, Italy). All the wild isolates were identified as B. bassiana from morphological characteristics using the taxonomy keys of Humber [102]. Furthermore, molecular characterization was performed using the transcription elongation factor (TEF-1α) DNA fragment gene and Internal Transcriber Spacer region (ITS). The sequences obtained were compared with those already present in the GenBank database by applying the BLAST software on the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov/BLAST/, accessed on 25 July 2021) and gave 99-100% similarity with B. bassiana as the first result.
Botrytis cinerea and Alternaria alternata strains were isolated from tomato-infested leaves and morphologically identified [103]. These fungal diseases agents were selected for the bioassays since they are common to tomato, and in particular, the two strains are highly pathogenic to tomato cv. Dwarf San Marzano, and the protocol for infection is already well consolidated [104,105]. All the isolates were maintained in 90mm Petri dishes containing Potato Dextrose Agar medium (PDA, HiMedia, Mumbai, India) in the dark at 25 • C.

B. bassiana Conidia Production on Rice
B. bassiana conidia of each isolate were obtained as follows: 500 g of parboiled rice were placed in autoclavable bags equipped with breathable filter bands for air exchange (SacO2, Flanders, Belgium) and then autoclaved for 20 min at 121 • C. A starter liquid culture was prepared by adding 10 mycelial plugs (6 mm 2 ) from a fresh PDA culture to a 250-mL Erlenmeyer Flask containing 100 mL of sterile Potato Dextrose Broth (PDB, HiMedia, Mumbai, India) and maintained for 72 h at 25 • C in orbital agitation at 120 rpm. Successively, the autoclaved rice was inoculated with the starter culture in a laminar flow hood. The sealed bag containing the inoculated rice was incubated at 25 • C with light for 5-7 days, until white conidia covered the rice. Conidia were harvested by washing the rice with 500 mL of sterile water under sterile conditions. Conidia concentration was determined by making a dilution series and counting the number of conidia in a hemocytometer (Neubauer counting chamber) under a microscope.

Induction and Assessment of Endophytic Colonization
Tomato seeds were surface sterilized in 1% NaOCl for 5 min, then rinsed 3 times with sterile distilled water to get rid of any surface fungal contamination. Seeds were germinated on Whatman ® sterile filter paper (Sigma-Aldrich, Darmstadt, Germany) soaked with sterile distilled water in the dark in an environmental chamber at 25 • C. Seedlings were individually transplanted to 8-cm diameter pots containing 200 mL of sterile commercial soil (Universal Potting Soil, Floragard, Oldenburg, Germany) and kept in a growth chamber at 25 ± 2 • C, 70 ± 10% RH, and photoperiod of 14:10 h light:dark and arranged in a randomized design with 15 plants divided in 3 replicates for each isolate. After one week, the emerged tomato seedlings were watered with a 20 mL of the B. bassiana conidial suspension (~1 × 10 7 conidia/mL) in order to obtain a final concentration of 1 × 10 6 conidia/mL of soil volume. Control plants were watered with sterile water. The same treatment was repeated after one week.
To determine the endophytic colonization, 5 tomato plants for each treatment were randomly chosen, uprooted, and dissected 2 weeks after the second watering treatment (one-month-old plants), and the surface-sterilized tissues were plated. The roots were carefully washed under tap water before the sterilization process in order to remove the soil. Three leaves for each plant and the stem were surface-sterilized in 1% NaOCl for 3 min, while the roots were maintained in the sterilizing solution for 5 min; after that, all the plant tissues were rinsed 3 times with sterile distilled water. The success of the disinfection procedure was assessed by plating three replicates of 100 µL each of the residual rinse water on PDA medium plates. Each plant tissue was dried on sterile paper, cut into 5 pieces with a sterile scalpel, and pieces were placed on 90 mm Petri plates containing PDA supplied with 1% lactic acid to avoid bacterial contamination. Plates were incubated at 25 • C in the dark and daily monitored to verify the fungal growth emerging from the cut plant tissues.
The endophytic colonization rate for each strain was calculated as the percentage of colonized plants on the total number of screened plants, and the frequency of colonization, as a measure of colonization intensity, was calculated as the percentage of colonized tissue section per plant on the total number of tissue sections [106]. Colonization rate and frequency were assessed separately for root, leaf, and stem tissues.
The fungal mycelia were isolated from the substrate close to the plant tissue and transferred on new plates containing PDA in order to obtain pure cultures for the morphological identification of the isolated fungus. Endophytic B. bassiana colonization was confirmed within 14 days after the tissue plating. The experiment was conducted twice.

Biocontrol of Tomato Foliar Pathogens-Tests In Vitro
Plate confrontation bioassays were performed with all experimental B. bassiana isolates to determine their potential as BCA against B. cinerea and A. alternata and to select the most effective ones. A fungal plug, about 6-mm diameter, was obtained from the periphery of fresh mycelial plate cultures for each plant pathogen and beneficial fungi. One pathogen and one B. bassiana isolate plugs were placed at approximately 4 cm of distance, in the middle of 90-mm diameter Petri dishes containing one-fifth strength PDA. The B. bassiana plugs were inoculated 2 days before the pathogens since its growth is considerably slower in comparison with one of two pathogens, and the simultaneous inoculum may hide the significance of the inhibition effect. Control consisted of plates inoculated only with one of the pathogens. The radial growth of the pathogen colonies, calculated as the mean distance between the inoculation point (center of the colony) and the edge in the four main directions, was measured at 7 dpi.
The biological activity of the beneficial fungi was assessed by comparing the growth of the pathogens with and without the presence of the B. bassiana isolates and reporting the presence of an antagonist/inhibition clearing zone formed between the two confronted fungal colonies. The distance of this zone between the pathogen and the beneficial fungi was measured. Fungal mycelial growth was determined by measuring colony radius and the percentage growth inhibition (PGI) of the pathogens was calculated according to the formula: PGI = (C − T) × 100, where C is the radial growth of the pathogen in control plates (mm), and T is radial growth of the pathogen in presence of the antagonist (mm) [107]. Three plates for each combination were prepared, and the bioassay was replicated twice.

B. bassiana Strain Choice for In Vivo Assay
All the assays described below were carried out on 30-da-old tomato plants endophytically colonized as described above, except for the insect bioassay, which was realized using 15-day-old plants.
The B. bassiana strains, selected for their ability to endophytically colonize the tomato plant and to counteract the growth of pathogens in vitro, were used for plants colonization, with water served as negative control. The six strains were: Bb74040, Bb632, Bb633, Bb762, Bb716, and Bb688.

Plant-Growth Promotion Assay
To determine the effect of the endophytic colonization with B. bassiana on plant growth, a survey of biometric parameters was carried out on Bb-colonized plants. Five plants per treatment were used to evaluate the effect of the endophytic infection on plant growth. The third, fourth, and fifth true leaf of every plant was gently removed, laid on a white paper sheet beside a measuring stick, and shot. The pictures were analyzed using ImageJ, an open-source image processing software [108], to measure the leaf area for each plant. To record the plant weight, each plant was uprooted, and the roots were gently washed under running water to remove the soil. Successively, the entire plant was placed into a paper bag already containing the removed leaves used to measure the leaf area and the bags were placed into an oven at 70 • C for 72 h. When completely dried, the plant was weighed. The experiment was repeated three times.

Biocontrol of Tomato Foliar Pathogens-Tests in Planta
To determine the effect of Bb endophytic colonization on foliar pathogens infection, biocontrol assays against A. alternata and B. cinerea were conducted. Ten plants per treatment were placed in a high-humidity chamber (80-90% RH, 20 • C) and inoculated with B. cinerea, while an additional subset of plants, again 10 per treatment, were placed in a high-humidity chamber and inoculated with A. alternata (80-90% RH, 27 • C). Three true leaves for each plant were inoculated with the pathogen conidial suspension laying on the leaf surface a 10 µL-drop of spore suspension at a concentration of 5 × 10 6 spores/mL. The inoculation point was marked with a permanent marker. The severity of the disease symptoms and the infection progression were evaluated 7 days dpi. The infection rate was calculated as the number of necrosis observed near the inoculation point over the total number of drops. Furthermore, the necrotic area provoked by the pathogens near the inoculation point was measured as follows: the leaves marked with the permanent marker were removed from the plant, laid on a white paper sheet beside a measuring stick, and shot. The pictures were analyzed using the ImageJ software to measure the mean necrotic area for each plant. The experiment was repeated three times.

Insect Rearing
The aphid Macrosiphum euphorbiae was collected from an infested field of tomato crops (Battipaglia, Salerno, Italy) and is permanently reared on S. lycopersicum plants (cv. Dwarf San Marzano) placed a growth chamber at 23 ± 2 • C, 70 ± 10% RH, and photoperiod of 14:10 h light:dark. This insect pest was chosen since it has been successfully utilized in previous studies in which it was found to readily feed on the tomato cv. Dwarf San Marzano, and the protocol of infestation and monitoring was already well defined [86,109].

Biocontrol of the Aphid M. euphorbiae-Tests in Planta
To assess the effects of endophytic colonization of tomato with B. bassiana on sucking insects infestation, a bioassay on the aphid pest M. euphorbiae was carried out. The biocontrol assays against M. euphorbiae were conducted in a growth chamber at 25 ± 1 • C, 70 ± 10% RH, and photoperiod of 14:10 h light:dark. For each treatment, four plants at the stage of the second true fully expanded leaf were placed in an entomological cage of 35 × 35 × 60 cm in transparent synthetic canvas (Omnes ed Artes s.a.s., Bergamo, Italy) and infested with M. euphorbiae as follows: 15 apterous adults were placed on each plant using a fine-tipped paintbrush and allowed to give birth; after 6 h, the adults were removed, and the first instar nymphs were counted. Nymphs' development was monitored until reaching the adult stage, and at that moment, only 10 adults were left per plant, removing the exceeding ones.
The survival and the fertility of these adult specimens, reared for their entire lifespan on the experimental plants, were observed for 14 days. Three times a week, the number of alive adults on each plant was recorded; moreover, the laid nymphs were counted and removed. When observed, the adult cadavers were removed, stored in groups of 30 specimens for 30 days in 50-mL falcon tubes with holes for air exchange, and checked weekly to observe putative mycosis occurrence under a microscope. The experiment was replicated three times.

Data Analyses
Data were analyzed by one-way Analysis of Variance (one-way ANOVA) and aposteriori LSD Fisher test using Minitab18 ® software (Minitab, State College, PA, USA). Data were tested for normality with the Shapiro-Wilk test and homoscedasticity with Levene's test using PAST 3 ® software (Softpedia, Carlsbad, CA, USA). Survival curves of M. euphorbiae were compared by using Kaplan-Meier and log-rank analysis using GraphPad Prism 5.0 software (GraphPad Software, La Jolla, CA, USA). The graphs were made using GraphPad Prism 5.0 software.