SELECTION AND CHARACTERIZATION OF Bacillus thuringiensis ISOLATES WITH A HIGH INSECTICIDAL ACTIVITY AGAINST Spodoptera frugiperda ( Lepidoptera : Noctuidae )

Spodoptera frugiperda (SMITH, 1797) (Lepidoptera: Noctuidae) affects diverse crops of great economic interest, for instance, it can cause severe yield losses in maize, rice and sorghum. In this study, a selection and characterization of Bacillus thuringiensis (BERLINER, 1911) isolates with a high insecticidal activity against S. frugiperda was performed. Fifty-two crystal-forming B. thuringiensis isolates that were identified from 3384 Bacillus-like colonies were examined and screened by PCR for the presence cry genes (cry1, cry1Aa, cry1Ab, cry1Ac, cry1D, cry2 and cry2Ab). Four isolates that showed high toxicity towards S. frugiperda were shown to harbor cry2 genes. The crystals were analyzed by electron microscopy and showed bipyramidal and cuboidal shapes. Furthermore, these four isolates had lethal concentration (LC50) values of 44.5 ng/cm 2 (SUFT01), 74.0 ng/cm (SUFT02), 89.0 ng/cm (SUFT03) and 108 ng/cm (SUFT 04) to neonate S. frugiperda larvae. An ultrastructural analysis of midgut cells from S. frugiperda incubated with the SUFT01 spore-crystal complex showed disruptions in cellular integrity and in the microvilli of the midgut columnar cells. The isolates characterized in this work are good candidates for the control of S. frugiperda, and could be used for the formulation of new bioinsecticides.


INTRODUCTION
The Lepidopteran insect Spodoptera frugiperda (Smith, 1797) (Lepidoptera: Noctuidae) affects diverse crops of great economic interest, for instance, it can cause severe yield losses in maize, rice and sorghum (Gallo et al., 2002).Since 35% of crops are lost to pest damage, an efficient pest control program is an important component in any effort to increase crop yields.The main trouble with the control of this pest is due to this insect behavior, which remains within the plant cartridges, reducing the contact with insecticides applied for its control (BRAGA MAIA et al., 2013).Moreover, in the conventional agricultural systems, the effects caused by changes in biodiversity and instability between trophic levels make the control of S. frugiperda increasingly difficult and costly (SANTOS-AMAYA et al., 2016;BOREGAS et al., 2013).The chemical control of insects in agriculture has been estimated to cost US$ 1391 miillion dollars annually in the united states (PIMENTEL, 2005), thus sustainable insect pest control is of vital importance.
Applications of insecticides and growing resistant cultivars are considered effective control methods.However, insecticides are considered generally, harmful to the environment (TODOROVA; KOZHUHAROVA, 2010) and can result in insecticide resistance in the insect pest population.There is an urgent demand to supplement existing management strategies with new methods that improve the insect pest control.
Biological control through the application of bioactive agents or microorganisms that are entomopathogenic to insect pests is an attractive alternative and a sustainable strategy for plant protection (MASSON et al., 1998).The biological control of S. frugiperda has been attempted using various microorganisms.Bacillus thuringiensis (BERLINER, 1911) (Bt) produces an array of Cry proteins with are potentially toxic to S. frugiperda (SANTOS et al., 2009;VALICENTE et al., 2010).The interest in this microorganism stems from its potential as an economical, effective, speciesspecific and environmentally safe pesticide (ARRIETA; ESPINOSA, 2006).
The accumulation of Cry proteins in a Bt cell can constitute 20-30% of the dry weight of the sporulated cell and each crystal protein has in its own insecticidal spectrum (AGAISSE AND LERECLUS, 1995).Some Cry proteins display toxicity to more than one insect order.For example, Cry1I is active against both Lepidoptera and Coleoptera (TAILOR et al., 1992), whereas Cry1B shows toxicity against Lepidoptera, Coleoptera and Diptera (HONGYU et al., 2000).Therefore, Cry proteins have been classified on the basis of their host specificity and their amino acid composition (SCHNEPF et al., 1998;ASOKAN et al., 2012).
The diversity and distribution of Bt and its cry genes have been described in several reports (CHAK et al., 1994;BRAVO et al., 1998;WANG et al., 2003).Depending on the soil type, different microbial communities are found.Several studies have demonstrated the toxicity of B. thuringiensis against S. frugiperda obtained from different types of environment.According to observations by Praça et al. (2004) of the 300 Bt isolated from soil samples, only one was effective against S. frugiperda with an LC 50 of 90.24 ng / cm 2 .In contrast, have found a larger number of Bt isolates with toxicity against S. frugiperda from the surface of leaves (phylloplane) (65%) and fewer were obtained soil samples (4.7%) in Colombia (Jara et al (2006).According to observations by Armengol et al. (2007), from 445 Bt isolates of different geographic locations from Colombia only 9.7% were toxic against S. frugiperda.The selection of new isolates of Bt that are toxic against S. frugiperda and could serve as a source of new cry genes, which could be introduced into economically important plant genomes that need protection against this important pest.
In the present study, soil samples from different regions of the state of Tocantis, Brazil, were used for the screening of Bt isolates toxic to S. frugiperda.New Bt isolates were characterizated and shown potential for further research that may lead to the development of new biopesticides.

Insects
S. frugiperda population with three generations was maintained at the Integrated Pest Management Laboratory of the Federal University of Tocantins.The insects were reared on an artificial diet based on beans, wheat germ and agar, in a controlled room under the following conditions: 26 ± 1°C, 70 ± 10% relative humidity and a light: dark period of 12:12 h.The adults were fed on an artificial diet according to the procedures described by Martins et al. (2008).

Insect bioassays
For selective bioassays, a Bt suspension was spread on top of the insect diet in 24-well plates (TPP, Techno Plastic Products AG, Switzerland) and one neonate S. frugiperda larvae was introduced in each well.Sterile distilled water was used as control, and the mortality was recorded 24 and 48 h after inoculation.Insect mortality rate was determined by Abbott's formula: %M= (T-I)/T × 100, where %M = percent insect mortality, T = the number of insects in the control treatment without the application of Bt, and I = the number of insects treated with Bt (ABBOTT, 1925).Dose-response bioassays for the determination of lethal concentration 50 (LC 50 ) were performed with 24 neonate S. frugiperda larvae as described by Monnerat et al. (2007), using different doses of dry spores and crystals (2,23,43,63,83,103,123,143,163,183,203,243 and 263 ng / cm 2 ).The insects were kept in an incubator with a photoperiod of 12 / 12 h (light/dark) at 26°C.Mortality was recorded at 24 and 48 h, and the LC 50 was obtained by a Probit analysis, (FINNEY, 1971), using Polo Plus software (LeOra Software Berkeley, CA, USA).Bacillus thuriengiensis subsp.Kurstaki HD1 (Btk) was used as reference strain and positive control.

Cry gene analysis
DNA from different isolates of Bt was purified and used as the template for polymerase chain reaction (PCR) with oligonucleotide primer pairs designed to amplify the cry1 and cry2 genes (Cerón et al., 1995;Ben-Dov et al., 1997;Bravo et al., 1998;Lima et al., 2008) (Table 1).PCR was carried out using Taq DNA polymerase (Invitrogen), and the PCR program was performed as follows: 1 min at 95°C; 30 cycles at 95°C for 1 min, annealing temperature (established as described in Table 1) and 72°C for 1 min, and a final step at 72°C for 5 min.The PCR products were analyzed by electrophoresis on 1, 2% agarose gels in TBE buffer (400 mM Tris, 10 mM boric acid, and 100 mM EDTA; pH 8.0) at 120 V for 30 min.

Protein analysis by SDS-PAGE
Proteins from spore/crystal mixtures were obtained according to the protocol described by Lecadet et al. (1992).Proteins were suspended in a small volume of phosphate-buffered saline (136 mM NaCl, 1.4 mM KH 2 PO 4 , 2.6 mM KCl, 8 mM Na 2 HPO 4 and 4.2 mL H 2 O, pH 7.4), and fractionated by electrophoresis in 12% SDS-PAGE gels ( Sambrook et al., 2001).

Ultrastructural characterization of spores and Cry proteins
The ultrastructural characterization of the spores and Cry proteins from B. thuringiensis isolates was performed by scanning electron microscopy.The isolates were cultivated in NYSM agar medium at 30 o C for 72 h and then a loop of the isolate was collected and diluted in sterile water.A volume of 100 µL of this dilution was deposited over metallic supports to be dried for 24 h at 37°C, covered with gold for 180 s using an Emitech apparatus (model K550), and observed in a Zeiss scanning electron microscope (model DSM 962) at 10 or 20 Kv.

Analysis of S. frugiperda midguts
Neonate S. frugiperda larvae were fed an artificial diet containing a suspension of Bt.Inoculated larvae were collected 24 h after diet ingestion, and the midguts were removed and processed for transmission electron microscopy, according to the protocol described by Martins et al. (2008).Ultra-thin sections were cut in an ultramicrotome (Leika ultracut UCT) and observed in a JEOL 1011 at 80 kV.

Colony forming units (CFUs) and δ-endotoxin production
The CFU and δ-endotoxin production of the Bt isolates were assessed in CCY medium (Stewart et al., 1981) after 72 h at 30°C in a rotary shaker set at 200 rpm (ZOUARI et al., 2002;AZZOUZ et al., 2014).The number of spores and the δ-endotoxin production were determined as described by Ghribi et al. (2007).

Isolation of B. thuringiensis isolates
In total, 52 crystal-forming B. thuringiensis isolates were identified from 3384 Bacillus-like colonies isolated from soil samples (Table 2).Only 37 showed any toxicity to S. frugiperda larvae (Table 3).The occurrence of Bt having higher levels of toxicity towards S. frugiperda was more prevalent in soil samples from the south-west region of Tocantins State (Araguaçu and Sandolândia), including four isolates that induced more than an 80% mortality in neonate S. frugiperda larvae.Isolates obtained from Palmas, Guaraí were less toxic to S. frugiperda, with induced mortalities of less than 15% (Table 3).In this study, 71.2% of Bt isolates showed toxicity against S. frugiperda.As can be seen from these results, the Bt isolates toxic to S. frugiperda can be found in different sample types and regions.These isolates have the potential to encode new combinations of cry genes or new cry genes.In this work, Bt isolates were found in both clay and sandy soils.Our data also indicated the occurrence of Bt isolates with higher toxicities to S. frugiperda in clay soil samples (Table 3).Intensive screening programs have isolated and characterized new isolates with different combinations of crystal proteins resulting in the discovery of new toxins and toxins with a broader activity spectrum (BRAVO et al., 1998).

Insect bioassays
The four isolates that produced 80% or more larval mortality in the selective assay were selected for analysis using the dose-response bioassay.SUFT01 showed the highest toxicity with an LC 50 of 44.5 ng / cm 2 , while isolates SUFT02, SUFT03 and SUFT04 had LC 50 values of 74.0, 89.0 and 108 ng / cm 2 , respectively (Table 4).S. frugiperda has shown a variable susceptibility to different Bt toxins, probably related to the genetic variability among different populations of the insect (MONNERAT et al., 2007).
At the same time, some Cry proteins were shown to be non-toxic to S. frugiperda (ARANDA et al., 1996;LUTTRELL et al., 1999).Santos et al. (2009) evaluated the effects of one B. thuringiensis subsp.aizawai and several B. thuringiensis subsp.kurstaki isolates against S. frugiperda.They found that, in general, most isolates of Bt were not toxic or showed low toxicity towards S. frugiperda larvae.Dias et al. (1999) also found that out of 25 Bt isolates, only eight had a high toxicity towards S. frugiperda larvae.
Isolates that harbored cry1, cry1Aa, cry1Ab, cry1Ac, cry1D, cry2 and cry2Ab were more toxic to S. frugiperda (Table 3).Valicente et al. (2010) suggested that the toxicity of different Brazilian B. thuringiensis isolates to S. frugiperda was related to their cry1 gene content.The cry genes frequencies and profiles differ among worldwide Bt collections based on region, moreover, they reported that the genes most frequently found in Minas Gerais, Amazonas, Paraná and São Paulo State, Brazil were cry1D, cry1G, cry1B and cry1E, respectively.
Since the toxic activity of any Bt isolate is due to the combination of different Cry proteins, we attempted to identify some common cry1 and cry2 genes present in a collection of Bt isolates, which had any level of toxicity to S. frugiperda, from Tocantins State (Brazil).The high toxicity of the four isolates of Bt may be associated with the presence of both the cry1 and cry2 genes (Table 3).In addition, several factors could be implicated in the insecticidal activity of the four isolates such as: different expression levels of cry1A and cry2 genes and the specific proportions of delta-endotoxins present in the crystal; mutations in cry genes and the presence of unknown Cry toxins in the crystal (CHENG et al., 1999;TOUNSI et al., 2006;HIRE et al., 2008).Moreover, the toxicity of Bt isolates against S. frugiperda is associated combinations of cry genes.Bravo et al. (1998), showed the combination of cry1D and cry1C genes exhibit high toxicity LC 50 against S. frugiperda below 35 ng / cm 2 .Similar results were observed by Jara et al. (2006) with combinations of cry1Aa, cry1Ac, cry1D and cry1b genes showing LC 50 of 29 ng / cm 2 .In the present study we observed that Bt isolated with higher toxicity (SUFT01) against S. frugiperda had combinations of cry1Aa, cry1Ab, cry1Ac, cry1D, cry2 and cry2Ab genes and LC 50 of 44.5 ng / cm 2 (Table 4).

Crystal protein patterns and morphologies
The protein profiles of purified crystals from isolates SUFT01, SUFT02, SUFT03 and SUFT04 were analyzed using SDS-PAGE and showed polypeptides of ~ 70 and 140 kDa (Table 3, Figure 3).An ultrastructural analysis of thespore/crystal mixtures from SUFT01, SUFT02 and SUFT03 isolates showed the presence of bipyramidal and cuboidal crystals (Figure 4A, B and   C).The SUFT04 isolate showed only cuboidal crystals (Figure 4D).The different profiles could be the result of different toxins being produced in these isolates.The analysis of the Bt isolates revealed that from 37 isolates analyzed, 12 formed bipyramidal crystals (Table 3), similar to isolates active against Lepidoptera (PRAÇA et al., 2004).However, 5 isolates that exhibited toxicity to S. frugiperda showed cuboidal and bipyramidal crystal inclusions (Table 3).The presence of a particular crystal type is not always associated with the toxicity towards an insect order.For example, Cry1I is active against both Lepidoptera and Coleoptera (TAILOR et al., 1992), whereas Cry1B shows toxicity against Lepidoptera, Coleoptera and Diptera (HONGYU et al., 2000).SUFT04 isolate formed cuboidal crystals and the presence of ~ 140 kDa band in SDS-PAGE.The ~140kDa band is usually associated with the formation of bipyramidal crystals.This absence of bipyramidal crystals could be due to the expression of a mutated version of a Cry protein or interaction of more than one Cry protein to produce the cuboidal crystal or even to the presence of a different Cry or other unrelated protein of the same size.
Several Bt toxins have already been identified and their toxicity to S. frugiperda has been demonstrated.However, new toxins are still being discovered.That diversity of these toxins includes activity against protozoans, mites, sheep lice, bark beetles, cockroaches, grasshoppers, tephritid fruit flies, moth flies and delphacids (FRANKENHUYZEN, 2009).Specific toxins responsible for those activities have not yet been identified or are still being characterized (SONG et al., 2008).
Additionally, Azzouz et al. ( 2014) showed that δ-endotoxin production by two Bt isolates that had high insecticidal activity against Spodoptera littoralis (BOISDUVAL, 1833) (Lepidoptera: Noctuidae) produced 43.71 and 80.81% more δendotoxin than HD1.The spore and δ-endotoxin production of Bt isolates obtained in this work indicates the potential for using formulations, at proper concentrations, to control S. frugiperda (GONZÁLEZ-CABRERA et al., 2011).

Analysis of the integrity of S. frugiperda midguts
Isolates from the SUFT01 spore-crystal complex were highly toxic to S. frugiperda midgut cells, causing disruptions to the cellular integrity and the microvilli of the midgut columnar cells (Figure 5A).After 24 h, an extensive disintegration in the columnar cells of the S. frugiperda midgut could be observed by transmission electron microscopy (Figure 5B).The disorganization of the cell cytoplasm caused by loss the microvilli in cells treated with Bt crystal and spore suspensions were visualized (Figure 5B).Endomembrane dilation, plasma membrane blebbing and cell fragmentation were observed, and they became more frequent as the post-inoculation time increased.Martins et al. (2008) described the effects of recombinant Cry1Ia protein action in the midgut of cotton boll weevil ,Anthonomus grandis (Boheman) (Coleoptera: Curculionidae) larvae, reporting morphologic changes in the midgut cells that were similar to those described in this work.Monnerat et al. (2007) obtained 27 isolates of Bt that killed 100% of S. frugiperda, Anticarsia gemmatalis (HÜBNER, 1818) and Plutella xylostella (LINNAEUS, 1758) larvae after 5 days in selective bioassays, and 19 isolates that exhibited LC 50 values that were lower than those of the standard strain Btk HD-1.From our results, four isolates showed promise in controlling S. frugiperda larvae with LC 50 values of 44.5, 74.0, 89.0 and 108 ng / cm 2 .However, the isolates SUFT01, SUFT02 and SUFT03 showed LC 50 values that were less than the 285 ng / cm 2 value for the Btk HD-1 strain reported in the work of (MONNERAT et al., 2007).The search for alternative methods to control insect pests is necessary, mainly to reduce the use of conventional chemical control methods.Thus, new Bt isolates represent a source of new toxin genes with the potential to be introduced into the genome of plants of economic interest.Currently, numerous studies have been reported in several transgenic plants containing the cry genes of Bt, such as potato (VALDERRAMA et al., 2007;KUMAR et al., 2010), tomato (LI et al., 2007), cotton (WU et al., 2005), rice (YE et al., 2003;YOUNG-JUN et al., 2004;GAO et al., 2010), corn (BELTAGI, 2008) and soybean (WALKER et al., 2000).Bt containing transgenic crop plants covered over 179,7 million hectares (corn, soybeans, cotton and potatoes) in worldwide and the highest increase was in Brazil, with 2 million hectares (James, 2015), this signifies the importance of this technology in controlling insect pests (ZHAO et al., 2005).

CONCLUSION
The high level of insecticidal activity of the isolates described in this work makes them excellent candidates for the control of S. frugiperda, and could provide alternatives in controlling insect pest populations that have developed resistance to chemical insecticides.Moreover, screening for new Bt isolates and their cry genes is important for the construction of cry gene databases for a possible future use in economically important transgenic crops.RESUMO: Spodoptera frugiperda (SMITH, 1797) (Lepidoptera: Noctuidae) afeta diversas culturas de grande interesse econômico, por exemplo, pode causar severas perdas em milho, arroz e sorgo.Neste estudo, foi realizada uma seleção e caracterização de isolados de Bacillus thuringiensis (BERLINER, 1911) com elevada atividade inseticida contra S. frugiperda.Cinquenta e dois isolados formadores de cristal B. thuringiensis que foram identificados a partir de 3384 colônias foram examinados e testados por PCR para a presença dos genes cry (cry1, cry1Aa, cry1Ab, cry1Ac, cry1D, cry2 e cry2Ab).Quatro isolados que apresentaram alta toxicidade contra S. frugiperda foram mostrados para abrigar os genes cry2.Os cristais foram analisados por microscopia eletrônica e mostraram formas bipiramidais e cúbicas.Os valores da

Figure 5 .
Figure 5. Toxicity effect of SUFT01 isolate suspensions in insect diets on Spodoptera frugiperda larvae midgut columnar cells.A. Midgut columnar cell of a S. frugiperda larva without the Bt treatment.Arrows indicate intact microvillus (MV).B. Midgut columnar cell of a larva fed an artificial diet containing the Bt suspension (24 h after inoculation).Arrows indicate the disorganization of the cell cytoplasm caused by loss of the microvillus in cell.

Table 2 .
Bacillus thuringiensis (Bt) isolates from soil samples from Tocantins State, Brazil.

Table 3 .
List of Bacillus thuringiensis (Bt) isolates that showed toxic activity against Spodoptera frugiperda.The data show the polypeptide and gene (cry1 and cry2) profiles, crystal morphology and % mortality rate for the spore/crystal mixture of each isolate in a selective bioassay.Selection and characterization of Bacillus thuringiensis… CERQUEIRA, F. B. et al.

Table 4 .
LC 50 values of Bacillus thuringiensis (Bt) isolates that have activity against Spodoptera frugiperda larvae.The total number of larvae tested in bioassays; b FL = Fiducial Limits; c LC 50 and FL calculated by Probit analysis; d χ 2 = chisquare test; P=; The results of three different bioassays.

Table 5 .
δ-endotoxin production of selected Bacillus thuringiensis (Bt) isolates on a CCY medium at 72 hours.Results are expressed as means ± standard error.Data represent the means of six replicates.Means in the same column followed by different letters are significantly different (p < 0.05).Analysis of variance by Tukey's test.