Growth Inhibition of Beauveria bassiana by Bacteria Isolated from the Cuticular Surface of the Corn Leafhopper, Dalbulus maidis and the Planthopper, Delphacodes kuscheli, Two Important Vectors of Maize Pathogens

The phytosanitary importance of the corn leafhopper, Dalbulus maidis (De Long and Wolcott) (Hemiptera: Cicadellidae) and the planthopper, Delphacodes kuscheli Fennah (Hemiptera: Delphacidae) lies in their ability to transmit phloem-associated plant pathogens, mainly viruses and mollicutes, and to cause considerable mechanical damage to corn plants during feeding and oviposition. Fungi, particularly some members of the Ascomycota, are likely candidates for biocontrol agents against these insect pests, but several studies revealed their failure to invade the insect cuticle possibly because of the presence of inhibitory compounds such as phenols, quinones, and lipids and also by the antibiosis effect of the microbiota living on the cuticular surface of the host. The present work aims to understand interactions between the entomopathogenic fungus Beauveria bassiana (Balsamao-Crivelli) Vuillemin (Hypocreales: Cordycipitaceae) and bacterial antagonists isolated from the cuticular surface of D. maidis and D. kuscheli. A total of 155 bacterial isolates were recovered from the insect's cuticle and tested against B. bassiana. Ninety-one out of 155 strains inhibited the growth of B. bassiana. Bacterial strains isolated from D. maidis were significantly more antagonistic against B. bassiana than those isolates from D. kuscheli. Among the most effective antagonistic strains, six isolates of Bacillus thuringiensis Berliner (Bacillales: Bacillaeae (after B. subtilis)), one isolate of B. mycoides Flügge, eight isolates of B. megaterium de Bary, five isolates of B.pumilus Meyer and Gottheil, one isolate of B. licheniformis (Weigmann) Chester, and four isolates of B. subtilis (Ehrenberg) Cohn were identified.


Introduction
Argentina is a leading maize-producing country, with an annual production of 15,500,000 tons. Several auchenorrynchan (Hemiptera: Auchenorryncha) species belonging to the Cicadellidae (leafhoppers) or the Delphacidae (planthoppers) families can reduce the yield and quality of maize grains because they transmit different plant pathogens, mainly viruses and mollicutes, and cause considerable mechanical damages during feeding and oviposition (Nault and Ammar 1989). The mentioned families include the largest number of vector species, with worldwide representatives (Nault and Ammar 1989). The corn leafhopper Dalbulus maidis (De Long and Wolcott) (Hemiptera: Cicadellidae) is widely distributed in tropical areas of the Americas, from southern USA to temperate zones of Argentina (Virla et al. 1990(Virla et al. /1991Giménez Pecci et al. 2002), and it is considered one of the most damaging species to corn due to its role as a vector of maize rayado fino virus, Spiroplasma kunkelii, and maize bushy stunt mycoplasm. These three pathogens, alone or in combination, are the ethiological agents of corn stunt, a disease that causes economic losses to corn crops in Mexico and Central and South America, and has been detected in restricted areas in the north of Argentina in 1990(Giménez Pecci et al. 2000. Among the delphacid pests of maize is the planthopper, Delphacodes kuscheli Fennah (Hemiptera: Delphacidae) a native species of Argentina that has been reported as a vector of Mal de Río Cuarto virus (Remes Lenicov et al. 1985;Remes Lenicov and Virla 1999). Due to high incidence and severity of damages, the Mal de Río Cuarto is the most important disease of corn crops in Argentina (Laguna et al. 2000;. Entomopathogenic fungi are widespread in agroecosystems and belong to a group of microorganisms extensively studied with more than 700 species within 100 genera (Lecuona 1996). These fungi infect a great number of arthropods and hence can be used as pest control agents in an Integrated Pest Management approach (Lecuona 1996). Fungi, particularly some members of Ascomycota, are attractive candidates as biocontrol agents against leafhoppers and planthoppers (Rice and Choo 2000;Toledo et al. 2007), but several studies revealed its failure to invade the insect cuticle, possibly due to the presence of inhibitory compounds such as phenols, quinones, and lipids (Smith and Grula 1981;Szafranek et al. 2001;Howard and Lord 2003;James et al. 2003;Lord and Howard 2004) and also by antibiotic effect of the microbiota living on the cuticular surface of the host (Hubner 1958;Walstad et al. 1970;Schabel 1978). Several works reported antagonistic interactions among microorganisms, such as fungi and bacteria (Currie et al. 1999;Ansari et al. 2005;Alippi and Reynaldi 2006), but there are no studies about the interactions between the microbiota found on the cuticle of hemipterous species and entomopathogenic fungi.
The purpose of the present work was to investigate the interactions among the entomopathogenic fungus Beauveria bassiana (Balsamo-Crivelli) Vuillemin (Hypocreales: Cordycipitaceae) and several bacterial antagonists isolated from the cuticular surfaces of D. maidis and D. kuscheli.

Insect culture
Dalbulus maidis and Delphacodes kuscheli were obtained from colonies reared on corn (Zea mays L.) and oat (Avena sativa L.) respectively in 24 x 9 cm polyethylene terephthalate plastic cages, in a greenhouse at the Facultad de Ciencias Agrarias y Forestales, UNLP (35º S -57º W), La Plata, Buenos Aires, Argentina.

Fungal isolate and preservation
The B. bassiana strain used in this study was isolated from one adult of Cycloneda sanguinea L. (Coleoptera: Coccinellidae) associated from corn at El Manantial, Tucumán, Argentina (26º 49' 50.2'' S -65º 16' 59.4'' W). This isolate has been previously reported as pathogenic to planthoppers and leafhoppers (Toledo et al. 2007), and was deposited as ARSEF 8372 in the Collection of Entomopathogenic Fungal Cultures, Agricultural Research Service, Ithaca, New York, USA, and as CEP 147 in the Collection at Centro de Estudios Parasitológicos y de Vectores, La Plata, Buenos Aires, Argentina.

Isolation and preservation of bacterial strains
Ten adults of each D. maidis and D. kuscheli (approximately 10 d old) were collected from April 2007 to April 2008 at monthly intervals. Using mouth aspirators, insects were placed into glass vials and transported to the laboratory. A total of 120 living insects were evaluated by placing them individually into sterile vials containing 300 μl of sterile distilled water. Each vial was vortex-mixed for 1 min, from which 25 μl were pippeted out and streaked on plates of tryptic soy agar (TSA) (Britania) using a Drigalski spatula. Plates were incubated at 30º C in aerobiosis and examined for bacterial growth every 24 h for and up to 10 days. Potential bacterial antagonists were primarily identified on the basis of Gram reaction and colony morphology. Microscopic examination of bacterial smears stained using the Schaeffer-Fulton technique was made to determine presence and location of spores within cells as well as the size and shape of vegetative cells. The isolates were maintained on sterile mineral water (Glaciar, www.glacierwater.com) at 4º C and stored in tryptic soy broth plus 20 % glycerol (v/v) at -80º C.

Preliminary screening for bacterial strains with antagonistic activity and statistical analysis
A total of 155 bacterial isolates were recovered from the cuticular surface throughout the sampling period, and tested for antagonistic effect against B. bassiana. A first screening to evaluate the effect of potential antagonists on the fungal growth was carried out by a central disk test assay (Reynaldi et al. 2004). Briefly, the fungal strain was cultured on malt extract agar for 7 days at 25º C in darkness and a 7 mm mycelium disk from the sporulating area was cut and transferred to the centre of a TSA plate. At the same time, three 7 mm disks containing each bacterial strain from a 48 h culture on TSA were transferred to each plate in the same way and placed at three equidistant points from the central disk. For controls, only a central disk of fungal growth was used. There were 3 replicate plates for each bacterial strain and for each control group (making a total of 65 controls). Treated and control plates were incubated at 30º C and the evaluation was performed by measuring the diameter of the fungal colony at 7 and 10 days, respectively. The percentage of mycelial growth inhibition (MGI) was calculated according to the formula proposed by Michereff et al. (1994). Only those treatments in which the fungal growth in the presence of bacteria was smaller than that of the controls were included in the statistical analysis. Differences in inhibition growth levels among treatments were analyzed by Kruskal-Wallis test, and means were compared by Fisher's least significant difference (LSD) multiple range test option (P 0.05) using Statgraphics statistical software (STSC, 1994(STSC, -2001. Bacterial strains isolated from D. maidis and D. kuscheli were analyzed separately. Differences in biological activity Table 1. Lifetable, fecundity and rate of natural increase of Commom Hoopae louse (Upupicola upupae).
Values followed by the same letters do not differ significantly according to LSD test (P 0.05). NV= Negative values, where fungal growth in the presence of bacteria was greater than that of the controls, were not included in the statistical analysis. Table only shown the 91 strains that inhibited the growth of B. bassiana between bacteria isolated from D. maidis and those isolated from D. kuscheli were analyzed by analysis of variance (ANOVA), and their means were compared by LSD test (P 0.05) using the Statgraphics software.

Identification of selected bacterial antagonists
Twenty-four bacterial isolates that showed the most effective antagonist effect against B. bassiana (MGI values between 40% and 83%) were further characterized to identify them at species level. Tests performed include catalase reaction, oxidase activity, motility, lipid globule staining, production of lecithinase, haemolytic activity, reduction of nitrate, anaerobic utilization of glucose, mannitol and arabinose utilization, and starch and gelatin hydrolysis according to standard protocols (Gordon et al. 1973). When necessary, API 20E and API 50CH strips plus API 50CHB medium and data base Apiweb (Biomerieux, www.biomerieux.com) were used.

Antagonistic activity against conidial germination of B. bassiana
According to the results obtained in the preliminary screening, ten bacterial strains were selected for testing their antagonistic activity on the conidial germination of B. bassiana ARSEF 8372 by means of a paired suspension assay. B. bassiana was cultured on malt extract agar and incubated for 8 days at 26º C in darkness. Conidia were harvested with a sterile loop and placed into test tubes containing 5 ml of Tween 80 (0.1 % v/v) (Sigma, www.sigmaaldrich.com). The suspensions were vortex-mixed for 1 min, filtered through a sterile muslin layer, and adjusted to a concentration of 1 x 10 8 conidia/ml after determination of conidial concentration using a Neubauer hemacytometer.
Each test bacterium suspension (vegetative cells after 24 h incubation or spores after 7 days incubation on TSA) was adjusted to a concentration of 0.5 Mc Farland. Bacterial suspensions were prepared in Tween 80 (0.1 % v/v).
Five l of each conidial suspension and 5 l of each bacterial suspension (vegetative cells or spores) were deposited on the surface of a microscopic slide containing 100 l of water agar medium as a substrate. The slides were placed over moist filter paper inside sterile 90 mm-diameter Petri dishes and incubated in darkness at 30º C. After 24 h germinated conidia were counted under a light microscope (400 X) by counting 3 times 100 conidia for each fungus-bacterial combination and each control taking into account that germinated conidia are those exhibiting a germ tube greater than the conidial diameter (usually once or twice). There were 3 replicates and one control per treatment. The whole assay was run twice over time in the same conditions mentioned above. The differences in inhibition growth levels among treatments were analyzed by Kruskal-Wallis analysis, and means were separated by LSD test (P 0.05) using Statgraphics statistical software. In addition, the abnormalities of the conidial germ tubes, if any, were registered.
Values followed by the same letters do not differ significantly according to LSD test (P 0.05). Table 3. Effects of bacterial antagonist preparations obtained from vegetative cells or spores, upon the growth of germinative tubes of B. bassiana (μm mean ± SE).
Values followed by the same letters do not differ significantly according to LSD test (P 0.05).

Inhibition of conidial germination
After 24 h of incubation no significant differences were found among treatments for vegetative cells (K= 13.4; P= 0.2) or spores (K= 16.3; P= 0.1), although B. pumilus Dm-B23 showed a higher inhibitory activity in both cases (Table 2). In addition, significant differences in conidial germ tube lengths were found for both vegetative cells (K= 88.3; P= 0.00) and spores (K= 45.4; P= 0.00), with B. pumilus Dm-B23 found to be the most effective antagonist (Table  3). No abnormalities of the conidial germ tubes were found in any of the bacterial-fungus combinations tested.

Discussion
According to the results presented here, Gram-positive aerobic spore forming bacteria belonging to the species B. megaterium, B. mycoides, B. pumilus, B. licheniformis, B. subtilis, and B. thuringiensis showed the most effective antagonistic effect on B. bassiana mycelial growth and conidial germination.
Bacillus pumilus, B. licheniformis, and B. subtilis along with Bacillus atrophaeus and B. amyloliquefaciens are closely related species that comprise the Bacillus subtilis group (Wattiau et al. 2001). The ability of this group of bacteria to inhibit fungal and bacterial growth by secreting antibiotics, antibiotic-like compounds, bacteriocins, or antifungal compounds has been well documented (Thimon et al. 1992;Feignier et al. 1995;Leifert et al. 1995;Gálvez et al. 1993;Martinari et al. 2002). These substances could play an important role in antagonistic interactions between microorganisms, which may be based on parasitism, direct competition, or antibiosis (Singh and Faull 1988).
Bacterial strains isolated from D. maidis were significantly more antagonistic, or at least produced large amounts of antagonistic compounds against B. bassiana than those isolates from D. kuscheli. The failure of B. bassiana to invade the D. maidis cuticle and the greater mortality rates previously observed in D. kuscheli could be related to a less antagonistic activity of the bacteria living in the same ecological niche. This hypothesis might explain those results observed in previous studies (Toledo et al. 2007) reporting that D. maidis mortality caused by B. bassiana was 23% less than that of D. kuscheli after 14 d post-inoculation. Although, further bioassays will be necessary to clarify this hypothesis by testing insects previously treated with antibiotics and inoculated separately with each bacterial strain and then with the pathogenic fungus.  (Galvez et al. 1993;Lebbadi et al. 1994;Alippi et al. 2000). Similar results have been reported for B. megaterium on A. apis (Reynaldi et al. 2004;Gilliam 1993), on S. tritici (Kildea et al. 2008), and on Phytophtora capsici (Peronosporales: Pythiaceae) (Akgul and Mirik 2008). It is interesting to point out that previous reports showed that symbiotic bacteria of entomopathogenic nematodes as Xenorhabdus nematophilus, X. bovienii, and Photorhabdus luminescens were antagonistic to the entomopathogenic fungi B. bassiana and M. anisopliae (Hypocreales: Clavicipitaceae) by inhibiting their growth and conidial production (Barbercheck and Kaya 1990;Chen et al. 1994;Ansari et al. 2005).
Our findings suggest the existence of a kind of antimicrobial activity possibly due to antibiosis effect and/or direct competition of spore-forming bacteria associated with Cicadellidae and Delphacidae that can reduce or inhibit the growth of B. bassiana. The presence of bacteria belonging to Bacillus cereus sensu lato group, B. megaterium, B. subtilis and closely related species in the cuticle of hemipterous insects could be an obstacle for the optimization and promotion of the use of entomopathogenic fungi in an integrated pest management approach in corn crops. Further studies are needed in order to clarify these microbial interactions and to characterize the chemical nature of the compounds involved in the inhibitory activities.