Isolation, purification and identification of biological compounds from Beauveria sp. and their evaluation as insecticidal effectiveness against Bemisia tabaci

Bemisia tabaci is one of the most notorious agricultural pests in the world. A vicious circle among insect resistance, dose increased, environment and human body impaired as the overuse of synthetic pesticides are becoming increasingly evident. Entomopathogenic Beauveria sp. is known as an effective natural enemy to control B. tabaci. Therefore, this study aimed to purify and identify the biological compounds from Beauveria sp. LY2 via extensive chromatographic techniques, NMR and MS and evaluated for their insecticidal activities against B. tabaci via contact and feeding assay. The outcome identified that one new cerebroside, cerebroside F (1), nine known compounds, cerebroside B (2), bassiatin (3), methyl 1,4-dihydro-4-oxo-2-quinolinecarboxylate (4), cerevisterol (5), 9-hydroxycerevisterol (6), 6-dehydrocerevisterol (7), (22E,24R)-ergosta-8(14),22-diene-3β,5α,6β,7α-tetrol (8), melithasterol B (9) and ergosterol peroxide (10) were isolated. Among the known compounds, methyl 1,4-dihydro-4-oxo- 2-quinolinecarboxylate (4) was isolated from natural origin for the first time. It is demonstrable from the results that compounds 3, 4 and 7 strongly featured insecticidal activities against B. tabaci, being the LC50 value as 10.59, 19.05, 26.59 μg/mL respectively in contact as well as 11.42, 5.66, 5.65 μg/mL respectively in feeding experiment. Moreover, no adverse effect on plant growth/height or phytotoxicity was observed on pepper, cucumber, tomato and cotton. The data from the current study has provided the foundation for the use of newly purified compounds against Bemisia tabaci as an alternative to synthetic chemical compounds.

The whitefly, Bemisia tabaci (Hemiptera: Aleyrodidae) is one of the world's top 100 invasive species which can attack more than 600 agricultural plant-host species under field and greenhouse conditions 1,2 . Due to its piercingsucking mouthpart, B. tabaci can directly cause plant weakness and indirectly transmit approximately 111 plant viruses 3 . Both nymphs and adults can also secrete honeydew, which can induce coal pollution. Besides B biotype (MEAM1) is the dominant species which significantly threatened many agricultural commodities in numerous countries [4][5][6] . Moreover, for the excessive use of synthetic pesticides have contributed to resistance development of B. tabaci. So far, 64 active ingredients have been reported for resistance from B. tabaci 7 .
The injudicious and excessive use of broad-spectrum synthetic pesticide can do harm to human health and environment 8 , and it can also result in the development of heritable resistance, pest resurgence and secondary pest problems. In contrast, bio-insecticides are safer for humans and more environmentally friendly. Moreover, plant and fungi are the sources of thousands of secondary metabolites 9 . Those secondary metabolites are ideal substitutes of those chemicals which play a significant role and possess a great potential on the field of biopesticides 10 .
Beauveria sp. (Ascomycota: Hypocreales) is a facultative entomopathogen with an extremely broad spectrum which is used as a commercial biopesticide against many agriculturally important insect pests 11 (Table 1) exhibited an α,β-unsaturated carbonyl carbon at δ C 203.0 (s, C-8), an amide acyl carbon at δ C 177.2 (s, C-1′), four olefinic carbon signals, a set of characteristic carbons corresponding to a β-D-glucopyranoside moiety appeared at δ C 104.7 (d, C-1″), 75.0 (d, C-2″), 77.9 (d, C-3″), 71.6 (d, C-4″), 78.0 (d, C-5″) and 62.7 (t, C-6″), other three oxygenated carbons at δ C 73.0 (d), 72.7 (d) and 69.7 (t), as well as a series of long-chain aliphatic carbon signals. The above NMR features were similar to those of cerebroside B (2) 20 , also isolated from the current study, and the obvious difference between them only came from the olefinic methyl part of the main chain, in which the olefinic methyl signal was removed and replaced by a terminal olefinic methylene in 1, in addition, a keto carbonyl at δ C 203.0 was newly detected. In the HMBC spectrum ( Figure S3), the correlations from H-19 [δ H 5.79, 6.09 (each 1H, br s)] to the keto carbonyl carbon confirmed the presence of α, β-unsaturated carbonyl moiety. Furthermore, the correlations from H-7 [2.82 (2H, br t, J = 7.4 Hz)] to C-5 [133.4 (d)] and C-8 [203.0 (s)] revealed the keto carbonyl at C-8. In view of their consistency of the chemical shifts and coupling constants, the configurations of C-2, C-3 and C-2′ were deduced to be the same as those of 2. The interpretation of the MS fragment ions (see Figure S8 in supplementary material) allowed to determine the specific length of the two chains. Therefore, the structure of compound 1 was established as illustrated in Fig. 2 13      www.nature.com/scientificreports/ In-vitro insecticidal activity of isolated compounds. After the isolation, purification and identification, compounds 1-10 were evaluated for the insecticidal activity against B. tabaci. The results of the mean mortality via both contact and feeding toxicity was exhibited in Fig. 3. Futhermore, the results of the LC 50 value at 72 h were presented in Table 2. However, detailed data was placed in supplementary material as Tables S1 to S10. According to the mortality results of compound 1, 41.67% mortality was reported at 72 h at 50 μg/mL via contact assay. Similarly, 38.33% mortality was displayed at the same concentration and time exposure via feeding assay (Fig. 3a). Similar results for insecticidal activity were afforded by compound 2 by both bioassays i.e. 47.50% and 44.17% mortality was recorded at 50 μg/mL via contact as well as feeding assay respectively (Fig. 3a). Results apparent from Fig. 3b represent that compound 3 has higher mortality as 97.50% and 79.17% at 50 μg/mL with exposure of 72 h and 48 h by feeding assay respectively. Similarly, by contact toxicity assay 79.17% and 54.17% mortality were exerted at the same concentration and time. Moreover, significant mortality was also afforded at 25 μg/mL by feeding as well as contact assay i.e. 80.83%; 66.67% and 66.67%; 46.67% mortality at 72 h and 48 h respectively. Similar results for mortality were also displayed by compound 4, where, 100% and 85.00% mortality were reported at 50 μg/mL at exposure of 72 h and 48 h respectively. Similarly, 92.50%, 76.66% and 66.67% mortality was displayed at 25 μg/mL, 12.5 μg/mL and 6.5 μg/mL at 72 h respectively. However, via contact assay maximum 87.50% mortality was recorded at 72 h and 50 μg/mL concentration (Fig. 3b). Results for insecticidal activity of compound 5 were presented in Fig. 3c which showed that 70.83% and 66.67% mortality were displayed at 72 h at the concentration of 50 μg/mL and 25 μg/mL respectively by feeding assay. Similarly, by contact assay 53.33% and 50.83% mortality were recorded at the same concentration and time exposure respectively. Same trend of mortality was exerted by compound 6 where 41.67% and 36.67% mortality were recorded at 50 μg/ mL and time exposure of 72 h (Fig. 3c). Alternatively, compound 7 was the best compound that presented significantly high mortality i.e. 100% and 85.00% mortality via feeding assay at 50 μg/mL concentration with 72 h exposure. Similarly, at 25 μg/mL 92.50% and 71.67% mortality were recorded at the same exposure period. Interestingly, this compound displayed 66.67% mortality at 72 h exposure even at low concentration 6.5 μg/mL. Whereas, relatively lower mortality i.e. 71.67% and 65.83% were recorded by contact assay at 50 μg/mL and 25 μg/ mL and 72 h exposure respectively (Fig. 3d). Results presented in Fig. 3d for compound 8 described that maximum mortality was afforded at 50 μg/mL and 25 μg/mL with 72 h and 48 h as 90.00% and 81.67% respectively by feeding toxicity assay, whereas, at the same concentration and exposure period 61.67% and 57.56% mortality were recorded by contact assay. Interestingly, compound 9 produced higher mortality on using contact assay rather than feeding assay like other described compounds. However, this compound produced moderate mortality in both bioassays i.e. 66.67%; 51.67% and 64.17%; 50.83% at 50 μg/mL at 72 h and 48 h exposure respectively (Fig. 3e). On the other hand, compound 10 recorded relatively higher mortality as 80.83% and 68.33% by contact toxicity assay at 50 μg/mL and 25 μg/mL and 72 h exposure period respectively as compared to contact toxicity assay which afforded 56.67% and 46.83% mortality at the same concentration and exposure period.
In essence, compound 3 exhibited an extraordinary mortality against B. tabaci on both contact and feeding experiments. Compound 4 and 7 highlighted the significant insecticidal activities on feeding bioassay and strong insecticidal activities on contact bioassay while compound 5, 8 and 10 demonstrated their strong feeding toxicity effectiveness while their contact insecitidal activities were moderate. Comparing the values above, compounds 1, 2, 6 and 9 ranged from moderate to fair effectiveness to B. tabaci. However, the organic solvents methanol and dichloromethane used as diluting agents in the contact assay did not cause any mortality.
Probability analysis showed the LC 50 values, slope value, Chi-square and fiducial limits at 95% confidence limit. Lowest LC 50 values displayed by compounds 3, 4 and 7 as 11.42, 5.66 and 5.65 µg/mL and 10.59, 19.05 and 26.59 µg/mL respectively via feeding and contact assay which illustrates their extraordinary insecticidal activity against B. tabaci. Whereas, other compounds displayed higher LC 50 values via feeding as well as contact assay respectively that showed their less toxicity against said pest (Table 2).

Safety evaluation of active compounds.
It is also important to note here that no phytotoxicity was observed during the whole experiment. The data presented in Table 3 shows the negligible effects of different compounds on height of pepper, cucumber, tomato and cotton were observed. Moreover, the data of pepper and cucumber presented in the Table 3 are highly significant while the data of tomato and cotton in Table 3 are moderatly to highly significant. Similarly, no phytotoxicity on leaves or stem was observed. After treatment on the 7th and 14th days were found healthy without any abnormalities like discoloration, necrosis, growth retardation, wilting and deformity.

Discussion
Several problems are associated with the use of synthetic chemicals for pest management, the introduction of natural products for this purpose is the primary concern. Essential oils, extracts and biologically active compounds are commonly used due to their effectiveness and safety for the environment as well as for humans. Secondary metabolites or natural products isolated from botanical source and range of microorganisms displayed bioactivity such as insecticidal, microbial, fungicidal and cytotoxic properties and contribute to their survival in several ways 29   www.nature.com/scientificreports/ against Spodoptera litura larvae via topical application at an interval of one week with 90-120 µg/insect, whereas, our research results showed 38.33% mortality at 72 h exposure at concentration of 50 μg/mL on B. tabaci. Additional biological activities of compounds ergosterol peroxide include antioxidant activities 32 , antimycobacterial activities 33 inflammatory activities 34 . Furthermore, the research study did not find any published material with the same compounds that was isolated on insecticidal activity study. This suggests there is still great potential of prospects on finding new active molecules from known entomopathogenic fungi.  www.nature.com/scientificreports/ B. tabaci is a potent agricultural pest with strong tendency to develop resistance against known pesticides, therefore new controlling agents are in constant need. In this study methyl 1,4-dihydro-4-oxo-2-quinolinecarboxylate (4), bassiatin (3), 6-dehydrocerevisterol (7) have shown promising insecticidal activity against B. tabaci. Compounds 1, 2, 4-9 are discovered in Beauveria sp. for the first time and compound 4 was discovered in natural products for the first time. Compound 4 was previously synthesized by Mazzoni 22 , even this is a known compound, we still use 2D NMR to fix the structure, due to no 13 C NMR data was reported in the literature. However, our findings on phytotoxicity and growth parameter/ height displayed that no phytotocxic effects on plants leaves and plant heights were observed which showed the safety profile of the isolated compounds.
Although different biological control approaches have been employed in current agriculture systems to control pest on crops and vegetables, however, the use of fungal based biological compounds against B. tabaci is limited.
The outcomes from this current study offered that isolated compounds from Beauveria sp. useful for the control of sucking pest especially B. tabaci. Hence, the introduction of these biologically active compounds could be an effective alternative and potential means to control such pests.

Materials and methods
General experimental procedures. NMR spectra were captured on an Avance-600 NMR spectrometer (Bruker, 57 Karlsruhe, Germany) at room temperature. High-resolution electrospray ionization mass spectrometry (HRESIMS) spectra data were recorded on a 6500 series quadrupole-time-of-flight (Q-TOF) mass spectrometer (Agilent, Santa Clara, CA). Mass spectrometry also recorded on LCMS 8050 (Shimadzu, Tokyo, Japan). High-performance liquid chromatography (HPLC) analysis was performed on a 1260 Infinity LC system (Agilent, Santa Clara, CA), and the column used was a 250 mm × 4.6 mm i.d., 5 µm, ZORBAX Eclipse XDB (Agilent, Santa Clara, CA). Semipreparative HPLC was performed on a 1260 series system (Agilent), and the column used was a 250 mm × 9.4 mm i.d., 5 μm, ZORBAX Eclipse XDB (Agilent). Column chromatography was performed using silica gel (100−200 mesh) (Qingdao Ocean Chemical Co. Ltd., Qingdao, China) and Sephadex LH-20 (GE Healthcare, Uppsala, Sweden). All chemical reagents were purchased from a chemical reagent company (Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) and used without further purification. which was cultured on SDAY medium at 25 ℃ in an incubator. The fermentation has two stages where the first stage, SDY medium (1% peptone, 1% yeast extract, 4% glucose, pH = 7.2) was used. Put a 0.5 cm diameter fungi pancake with SDAY medium into a 250 mL Erlenmeyer flask which contains 40 mL of SDY medium, then incubated them at 25 ℃ with 180 rpm shaking speed for 3 days as to prepare the seed culture. In the second stage, Czapek-Dox medium (0.03% NaNO 3 , 0.01% K 2 HPO 4 ·3H 2 O, 0.005% MgSO 4 ·7H 2 O, 0.005% KCl, 0.0001% FeSO 4 ·7H 2 O, 3% sucrose, pH = 7.0) was used. 120 2L Erlenmeyer flasks, each contains 400 mL Czapek-Dox medium, were inoculated with 40 mL seed culture at 25 ℃ with 180 rpm shaking speed for 15 days. The fermentation cultures were centrifuged at 4 ℃ with 5000 rpm for 30 min to remove mycelia, then add 3% Amberlite XAD 16 resin into the broth at 25 ℃ with 180 rpm shaking speed for 4 h. Resin was collected by using 100 meshes gauze and extracted four times with methanol. The dried crude extract was harvested through reduced pressure concentration. The dried methanol extract was dissolved in 600 mL solution (50% CH 3 OH, 50% H 2 O), the solution was extracted four times by 600 mL CH 2 Cl 2 . Collecting the resulted extract CH 2 Cl 2 solution then concentrate it to produce 10.0 g solid brown residue.
Isolation and purification. The concentrated extract was purified by using silica gel chromatography to elute stepwise with CH 2 Cl 2 -MeOH (100:0, 50:1, 25:1, 10:1 and 0:100, 1.5 L each) as the mobile phase to afford six fractions, A to G. C was separated via silica gel chromatography with PE:EA (9:1, 4:1, 7:3, 6:4 and 1:1, 300 mL each ) as the mobile phase to yield compound 3 (110 mg), compound 9 (7.3 mg) and compound 10 (14.4 mg). Fraction E was subjected to gel chromatography on Sephadex LH-20 eluted with CH 2 Cl 2 -MeOH (1:1) to obtain compound 5 (16.7 mg). D, F, G was purified by reverse-phase semi-preparative HPLC applying a MeOH-H 2 O gradient (contain 0.1% HCOOH) respectively. Compound 6 (2 mg) was gained from D with 90% MeOH on 40 min (see Figure S26), compound 7 (4 mg) and compound 8 (4.2 mg) were also obtained from D with 85% MeOH on 18 min and 20 min (see Figures S30 and S34). Compound 4 (3.1 mg) was yielded from F with 77% MeOH on 28 min (see Figure S19), 1 (3.8 mg) and 2 (25.9 mg) was obtained from G with conditions with 98% MeOH on 10 min and 16 min respectively (see Figures S7 and S12). Bemisia  www.nature.com/scientificreports/ Pre-experiment container preparation. One side of a lightproof bi-pass glass tube (3 cm inside diameter and 6 cm height) was covered with a stretched Parafilm M membrane (the first layer). Another side sealed with one piece of 100 meshes gauze and a rubber band 36 . All items had to be completely sterile. Safety evaluation of active compounds. pepper (Capsicum annuum L., cultivar 'Wanhao A5' from SYAU, China), cucumber (Cucumis sativus L., cultivar 'Cuibao' from SYAU, China), Tomato (Lycopersicon esculentum, cultivar 'Dafen' from SYAU, China) and cotton (Gossypium hirsutum, cultivar 'Guoshenhan 284' from Xingyuan Co., China) were chosen as experimental objects and planted in nutrition pots from seeds for one month until they reached to specific leaves stage (tomato and pepper for 8-12 leaves, cucumber and cotton for 4-5 leaves) in greenhouse with the temperature scale between 10 and 35 ℃. The equivalent crop plants with similar heights and proper leave sizes were picked and transplanted into 7 cm 3 black plastic pots, then divided the crop plants into 8 groups, each group consists of 3 pots for each kind of crop plants. The insecticidal compounds 3-5, 7-10 were diluted with water containing 0.5% DMSO and 0.5% Tween 80 into the concentration of 100 μg/mL solutions respectively. A hand-held vacuum sprayer was used to foliar spray each group of plants with 15 mL of solution per group. Plants were monitored for heights and abnormalities (discoloration, necrosis, growth retardation, wilting, deformity) every 7 days for 2 weeks according to Chinese National Standard of Laboratory Test For Crop Safety Evaluation 37 and data was calculated.

Evaluation of insecticidal activity against B. tabaci.
All the studies on plants totally complied with relevant institutional, national, and international guidelines and legislation. Statistical analysis. All the calculated data on mortality and plant heights were analyzed via one way analysis of variance (ANOVA), the mean difference between treatments was envisioned for significance test by Duncan multiple range test DMRT with IBM-SPSS statistics 25.0 version software. Probit analysis was performed using EPA Probit analysis program version 1.5. 0.

Conclusions
The present research study indicated that the secondary metabolites from Beauveria sp. possess potential botanical agents. Results also demonstrated that B. tabaci produced significant sensitivity to isolated bassiatin (3) via both feeding and contact bioassay as well as methyl 1,4-dihydro-4-oxo-2-quinolinecarboxylate (4) and 6-dehydrocerevisterol (7) via feeding bioassay. cerevisterol (5), (22E,24R)-ergosta-8(14),22-diene-3β,5α,6β,7α-tetrol (8), melithasterol B (9) and ergosterol peroxide (10) displayed moderate toxicity against this pest. In contrast, cerebrosides F (1) and cerebrosides B (2) showed much lower mortality than other compounds. Moreover, all the compounds were found safe without affecting the plants growth with no phytotoxicity. Therefore, except compounds 1 and 2, these compounds were introduced as alternatives to synthetic chemical insecticides. However, more research on the purification and characterization of bioactive compounds from entomopathogenic fungi is needed to be compared against B. tabaci and other agricultural insect pests in potential future research studies.