Identification and evaluation of a new entomopathogenic fungal strain against Riptortus pedestris (Hemiptera: Alydidae) and its two egg parasitoids

A strain (ARP14) of the entomopathogenic fungus Beauveria bassiana (Balsamo) Vuillemin was isolated from field-collected Riptortus pedestris (Fabricius) (Hemiptera: Alydidae). The lethal median concentration of the ARP14 strain was compared with that of a commercialized strain (GHA) of the same fungus against R. pedestris and its two egg parasitoids, Ooencyrtus nezarae Ishii (Hymenoptera: Encyrtidae) and Gryon japonicum (Ashmead) (Hymenoptera: Platygastridae). Mortality and mycosis rates were evaluated after exposure to five concentrations of the fungus, i.e., 1×109, 1×108, 1×107, 1×106, and 1×105 conidia/mL, using a glass scintillation vial as an exposure arena in 25.0 ± 0.5°C and 93.7 ± 2.9% RH. The lethal median concentrations (LC50) for 2nd and 4th instar nymphs, and adults of R. pedestris were not significantly different between the two strains of B. bassiana. However, the mycosis rate of ARP14 was 1.3 and 1.8 times higher than that of the GHA strain in 4th instar nymphs and adult females of R. pedestris, respectively, at the 1×108 conidia/mL concentration. More interestingly, the mycosis rates at 1×108 conidia/mL concentration in the parasitoids G. japonicum and O. nezarae were much lower in the ARP14 strain (15.0 and 0%) than in the GHA strain (73.3 and 66.0%), respectively, suggesting that the B. bassiana strain ARP14 is less virulent to these parasitoids than the commercially available strain. Our results suggest that B. bassiana ARP14 may be a potential new biopesticide against R. pedestris with fewer negative effects on beneficial parasitoids than currently available options.


Introduction
Pod sucking bugs, including Riptortus pedestris (F.) (Hemiptera: Alydidae), Halyomorpha halys (Stål) (Hemiptera: Pentatomidae), and Nezara antennata Scott (Hemiptera: Pentatomidae), are major pests of soybean by reducing seed quality and yield [1,2]. Among these bugs, PLOS  oviposition substrate) that was hung in the corners of the cage. Riptortus pedestris nymphs were reared in a separate cage with potted kidney bean (Phaseolus vulgaris L.) plants, dry soybean seeds, and cotton soaked with a 2% vitamin C water solution. The egg parasitoids of R. pedestris were reared on non-viable host eggs (refrigerated for 25-30 d) according to Alim and Lim [27] in a centrifugal tube (50 mL) with a streak of honey on the wall. Mated females (3 d old) of the egg parasitoids were released for 24 h in the centrifugal tubes containing the nonviable host eggs, and the parasitized eggs were incubated at 27.2 ± 1.0˚C, 41.7 ± 8.7% RH, and a 16:8 h L: D photoperiod. Emerged parasitoid adults were collected in the centrifugal tube for bioassays and further rearing.

Morphological and molecular identification of B. bassiana strains
The morphology of the fungal pathogen's synnema was studied under scanning electron microscopy (MIRA3, Tescan Orsay Holding, Brno-Kohoutovice, Czech Republic) according to the taxonomic description of Rehner et al. [8]. DNA of each of the two fungal strains was extracted following the methodology described by Chi et al. [28]. About 20 mg of fungal mycelia (2-3 d old grown in SDA media) was harvested with a sterilized dissection blade and put into an Eppendorf tube (1.5 mL) containing 300 μL of extraction buffer [1 M KCl.100 mM Tris-HCl (pH 8.0), 10 mM EDTA]. Mycelia tissues were thoroughly ground using a pestle, followed by centrifugation at 11,000×g for 1 min. The supernatant was transferred to a sterile Eppendorf tube, and isopropanol (200 μL) was added. The tube was well mixed before additional centrifugation at 11,000×g for 10 min. The supernatant was discarded, 300 μL of ethanol was added, and the tube was gently inverted to wash the pellet three times followed by a final centrifugation at 11,000×g for 1 min. The supernatant was discarded, and the Eppendorf tube was left open at room temperature to allow excess ethanol to evaporate. After 10 min, the DNA pellet was gently dissolved in 50 μL of 1×TE buffer by tapping the tube. The ITS-rDNA region of the collected DNA sample was amplified using primer pairs-ITS1 (5'-TCC GTA GGT GAA CCT GCG G-3') and ITS4 (5'-TCC TCC GCT TAT TGA TAT GC-3') [29] in a SimpliAmp Thermal Cycler (Life Technologies Holding Pte Ltd, Singapore). The PCR extraction was done by preheating the sample at 95˚C for 5 min, followed by 35 incubation cycles at 94˚C for 45 sec, 55˚C for 30 sec, and 72˚C for 45 sec followed by a final extension at 72˚C for 5 min. The PCR product was purified using a PCR purification kit (Biofact Co., Ltd., Daejeon, Republic of Korea) and sequenced using ABI PRISM 3730XL analyzer by Macrogen Korea (Seoul, Republic of Korea). The nucleotide sequence of the APR14 (Accession No. MG952537.1) strain was compared with that of the other B. bassiana strain using a Blast search of sequences from the NCBI Genbank database. The nucleotide sequences most similar to ARP14 and that of the fungal species most closely related to B. bassiana, Isaria spp., and Metarhizium spp. were downloaded from the Genbank, and phylogenetic analysis of these taxa was conducted using MEGA7 software (Biodesign Institute, Tempe, Arizona).  [25]. Based on the count, we set the suspension to the concentration of 1×10 9 conidia/mL, and prepared other solutions in different concentration by serial dilution: 1×10 9 , 1×10 8 , 1×10 7 , 1×10 6 , and 1×10 5 conidia/mL.

Beauveria bassiana toxicity in a glass vial assay
Three different life stages of R. pedestris (<24 h old 2 nd instar nymphs, <36 h old 4 th instar nymphs, and <48 h old adult females) and adult females of its two egg parasitoids, G. japonicum and O. nezarae (both 5-7 d old), were tested at five different concentrations (1×10 9 , 1×10 8 , 1×10 7 , 1×10 6 , 1×10 5 conidia/mL) of ARP14 and GHA, using 0.1% Triton X-100 ddH 2 O as a control. The 20 mL LS vial was used for the bioassay, and we coated the inside of each vial with 100 μL of the test solution for each concentration and air dried in room temperature. For each replicate of each species or stage, five insects were exposed for 12 h in the fungus-coated vials. Exposed insects were then transferred to clean 2 mL Eppendorf tubes with a small hole in lid after 12 h of exposure and kept in desiccators (4202-0000, Bel-Art Products, Pequannck, NJ) at 25.0 ± 0.5˚C and 93.7 ± 2.9% RH inside a growth chamber (DS-50CPL, Dasol Scientific Co., Ltd, Suwon, Republic of Korea) to determine the fungal mycosis development rate over a 14 d period following exposure. RH inside desiccators was maintained using saturated Potassium Sulfate (K 2 SO 4 ) solution [30]. Water and food were not provided for the insects to remove the compounding effects on pathogens. Temperature and RH during the experiment was measured using a data logger (H8-003-02, Onset Computer Corporation Bourne, MA) inside the desiccators. Mortality of insects was observed at 12 h intervals from exposure until death. Insect was categorized as death when there was no movement during three times touch with a camel brush under stereoscopic microscope. Insects categorized as mycosis with B. bassiana when fungus mycelia were visible on insects' integument through a stereoscopic microscope.

Statistical analysis
Mortality from the various concentrations of the ARP14 and GHA strains were subjected to log-probit regression analysis to calculated lethal median time (LT 50 ), based on observations every 12 h after exposure. The mortality data from trials with each fungal strain and concentration for both R. pedestris and its parasitoids were also used to calculate the lethal median concentration (LC 50 ) [31]. Significant differences among treatments were determined based on the 95% confidence interval (CI). The toxicity index at different concentration levels was calculated by dividing the LT 50 of each control with that of the treatment [32]. The fungal mycosis development rates of the ARP14 and GHA strains were analyzed with normal approximation of the chi-square test, and Tukey type multiple comparison test was followed (α = 0.05) [33]. Comparison of data for the mortality and fungal mycosis rates between ARP14 and GHA strains at each concentration and insect stage or species was conducted using two proportion Z-tests [33].

Beauveria bassiana toxicity to R. pedestris in a glass-vial assay
Riptortus pedestris nymphs and adults died faster at higher conidial concentrations (e.g., 1×10 9 and 1×10 8 conidia/mL) for both B. bassiana strains tested (ARP14 and GHA) ( Table 1  Identification and evaluation of Beauveria bassiana ARP14 against R. pedestris than at lower conidial concentrations. The LC 50 values for bug either nymphs or adults were not significantly different between the two strains (ARP14 and GHA) ( Table 2). The mortality rates of 2 nd instar nymphs at either 48 (Z = 0.71, P = 0.476) or 72 h (Z = 1.42, P = 0.155) were not significantly different between the ARP14 and GHA strains at the 1×10 8 conidia/mL concentration. The 2 nd instar nymphs of R. pedestris showed 100% mortality 108 h after exposure  mortality reached 100% at 120 and 168 h after the exposure to 1×10 8 conidia/mL concentration for the ARP14 and GHA strains, respectively (Fig 3).
The LT 50 values of both strains, at concentrations higher than 1×10 6 conidia/mL, were all lower than that of the buffer control in all the life stages tested ( Table 1). The toxicity index presented in Table 1 illustrates the different survivorship among the concentrations of the two strains. Compared to the buffer control, the toxicity index at 1×10 8 conidia/mL was 1.3, 1.5, and 1.9 times higher in 2 nd instar nymphs, 4 th instar nymphs, and adult females of R. pedestris exposed in B. bassiana ARP14 strain whereas it was 1.4, 1.5, and 1.6 times higher for the B. bassiana GHA strain, respectively.

Beauveria bassiana toxicity to R. pedestris egg parasitoids in a glass-vial assay
The LT 50 values for both parasitoids of the B. bassiana ARP14 strain were not significantly different from the control for any of the five conidial concentrations except for mortality of O. nezarae at 1×10 9 conidia/mL (Table 3). However, for the GHA strain, parasitoid mortality significantly higher at 1×10 9 conidia/mL in both species and at 1×10 8 conidia/mL for O. nezarae, only compared to the controls ( Table 3). The mortality of G. japonicum at 48 (Z = 0.25, P = 0.806) and 72 h (Z = 1.02, P = 0.309) and that of O. nezarae at 48 (Z = 0.81, P = 0.417) and 72 h (Z = 0.21, P = 0.836) were not significantly different between APR14 and GHA strain at 1×10 8 conidia/mL concentration. In both parasitoids, G. japonicum and O. nezarae, 100% mortality occurred at 120 and 108 h after exposure to 1×10 8 conidia/mL concentration of B. bassiana ARP14 and GHA, respectively. Nevertheless, in control, 100% mortality occurred 120 h after exposure in both strains (Fig 4).
The toxicity index for G. japonicum and O. nezarae was 1.2 in each species at 1×10 8 conidia/mL exposed to the ARP14. Similarly, this index was 1.3 and 1.2 for the GHA strain at 1×10 8 conidia/mL for G. japonicum and O. nezarae, respectively.

Discussion
The new entomopathogenic fungal isolate collected from R. pedestris was identified as B. bassiana and designated as strain ARP14, based on morphology [2,34] and intraspecies and interspecies divergence rate with different Beauveria species and strains [35]. Strain ARP14 showed high virulence to R. pedestris in the glass-vial assay, and mortality rates of the tested life stages of R. pedestris increased with conidial concentration. The LC 50 of strain ARP14 was not significantly different from that of GHA in any of the tested life stages of R. pedestris. Nevertheless, both ARP14 and GHA strains were found to be more effective against nymphal stages than the adult stage of R. pedestris. In a study conducted on Riptortus linearis (L.), B. bassiana CH1 was Fig 5. Mycosis rate of Beauveria bassiana ARP14 and GHA strains on 2 nd instar nymph, 4 th instar nymph and adult stage of Riptortus pedestris exposed in a glass-vial assay for 12 h. Mean mycosis rates followed same letter are not significantly different among the different concentrations of Beauveria bassiana ARP14 (small letters) or GHA (capital letters) (χ 2 , P > 0.05). Significance of differences between mycosis rates within conidial concentrations, between the ARP14 and GHA strains are denoted as follows: Ã 0.01<P 0.05, ÃÃ 0.001<P 0.01, ÃÃÃ P 0.001 and ns indicates non-significance.
https://doi.org/10.1371/journal.pone.0195848.g005 Fig 6. Mycosis rate of the Beauveria bassiana strains ARP14 and GHA in adults of Ooencyrtus nezarae and Gryon japonicum when exposed in conidia-coated glass-vials for 12 h. Mean mycosis rates followed same letter are not significantly different among the different concentrations in Beauveria bassiana ARP14 (small letters) or GHA (capital letters) (χ 2 , P > 0.05). Significance of differences between mycosis rates within conidial concentrations, between the ARP14 and GHA strains are denoted as follows: Ã 0.01<P 0.05, ÃÃ 0.001<P 0.01, ÃÃÃ P 0.001 and ns indicates non-significance. https://doi.org/10.1371/journal.pone.0195848.g006 Identification and evaluation of Beauveria bassiana ARP14 against R. pedestris also more virulent to nymphal stages than adults [36]. A similar result was also found in other B. bassiana isolates, as for example where the larval stage of Alphitobius diaperinus (Panzer) (Coleoptera: Tenebrionidae) was more susceptible to B. bassiana than adults [37]. However, the LC 50 of B. bassiana CPD9 strain in Clavigralla tomentosicollis Stål. (Hemiptera: Coreidae) was not different between 5 th instar nymphs and adults [38]. Similarly, other B. bassiana isolates/strains showed similar virulence to nymphs and adults of several hemipteran bugs [23,36,39,40]. The efficacy of EPF is known to vary, depending upon the host's physiological state (i.e., weakened, ill, or low-immune condition) [41]. However, the mycosis rate of ARP14 in different life stages of R. pedestris was comparatively higher than the rates caused by the GHA strain, probably because ARP14 was isolated from R. pedestris. EPF are known to be more virulent on their natal host species than on novel species [40].
EPF, including B. bassiana, often have a wide physiological and ecological host ranges. Therefore, the development of an ecologically selective strain is needed for them to be an effective mycoinsecticide. In our study, B. bassiana ARP14 caused lower rates of mycosis in the pest's two-egg parasitoids, G. japonicum and O. nezarae, and thus may be a selective mycoinsecticide for control of R. pedestris. Among commercial formulations of B. bassiana, Natura-lis1-O is known to be relatively safe to the natural enemies of whiteflies, such as Encarsia formosa Gahan (Hymenoptera: Aphelinidae) and Orius insidiosus (Say) (Hemiptera: Anthocoridae), and Phytoseiulus persimilis Athias-Henriot (Mesostigmata: Phytoseiidae), while it lacks selectivity for the aphid parasitoid Aphidius colemani Viereck (Hymenoptera: Braconidae) [42]. Similarly, Beauveria brongniartii (Saccas) Petch when used to suppress larvae of Melolontha melolontha L. (Coleoptera: Scarabaeidae) in a forest habitat were less infectious to the natural enemies of these chafers [43]. Although the underlying mechanism of the selectivity of ARP14 against natural enemies is unknown, the virulence of EPF is known to vary interspecifically due to differences in toxin production, chemical composition of the host's epicuticle, host cleaning behavior of the host (which removes conidia), and the method used to apply the conidia [41,42,44,45]. Exact effects of how such fungi may or may not differentially affect the target pest versus its natural enemies cannot be easily predicted, and studies are required in each system to determine if a product will have beneficial ecological selectivity.
In conclusion, as a mycoinsecticide with a low negative effect on key non-target egg parasitoids that could be used in a compatible manner with natural enemies in IPM [24], B. bassiana ARP14 appears to be a good candidate for use against R. pedestris while having minimal effect on the pest's egg parasitoids. Nevertheless, development of formulation and verification of the efficacy in fields should be preceded before the application.