Beauveria bassiana-based mycoinsecticides have been developed and registered worldwide for control of agricultural pests [17, 29], usually being applied in the fields as a conidial spray [19]. This fungus infects a wide range of insects such as thrips, beetles [13], flies [20], and several species of ticks [19, 21] and is commonly found in nature [7]. The tick Boophilus microplus is a bovine ectoparasite that causes significant economic losses in herds of tropical and subtropical areas. It transmits diseases and causes reduction in milk and meat yield and leather production. The necessity of tick control represents significant investment and the present technology is based on the use of synthetic chemical products. However, the ability of B. microplus to develop resistance to acaricides, the demands of consumers for chemical free foods, and the negative environmental effects of acaricides call for the development of alternative strategies. Therefore, efforts to develop alternative methods, such as biological control of ticks using filamentous fungi, chiefly Metarhizium anisopliae [12] and B. bassiana [23], have been pursued.

To transpose the cuticle, the main host barrier, entomopathogenic fungi utilize a combination of mechanical and enzymatic mechanisms, and secretion of proteases is believed to be an important pathogenic factor for fungal attack on cuticle [28].

The best-understood model of a fungal determinant of entomopathogenicity is based on M. anisopliae, the subtilisin-like endoprotease designated Pr1 [30]. This enzyme is adapted to extensively degrade insects’ cuticular proteins [28] and has been ultrastructurally located in the host cuticle during the early stages of penetration [15]. The occurrence of natural variability in the production of cuticle-degrading proteases among isolates of M. anisopliae after growth on cuticular and noncuticular substrates has being investigated [3, 22]. Field studies have demonstrated that B. bassiana also colonizes corn, endophytically [5].

Beauveria amorpha has both endophytic and entomopathogenic characteristics, but its capacity for biological control has not been studied. To help understand the role of proteases and chitinases in B. microplus cuticle penetration, it is desirable to determine how their synthesis is regulated in these two fungal models. Knowledge of how protease and chitinase production is regulated could be highly relevant to understanding the pathogenic process. Therefore, this work aims to analyze the production of both enzymes by B. bassiana and B. amorpha and to investigate the infection process in B. microplus by scanning electron microscopy (SEM), to identify possible variations that may be relevant for tick biocontrol and for the development of commercial formulations.

Materials and Methods

Organisms and culture conditions

Beauveria bassiana strain CG166, originally isolated from Schrius sp. (Curitiba/PR, Brazil), was supplied by Empresa Brasileira de Pesquisa Agropecuária (Embrapa/Cenargen, Brasília/DF, Brazil). Beauveria B95 was isolated from Zea mays leaves by Dr. Ida Chapaval Pimentel (Universidade Federal do Paraná/UFPR, Brazil). DNA from this endophytic isolate (Beauveria B95) was characterized by sequencing 5.8S rDNA, ITS-1, and ITS-2, using the primers ITS1-F (5′-TCCGTAGGTGAACCTGCGG-3′), which is specific for higher fungi, and ITS4 (5′-TCCTCCGCTTATTGA TATGC-3′), which is a universal primer. The purified PCR products (GFX PCR DNA and band purification kit; GE Healthcare) were then sequenced in both directions using the ITS-1 and ITS-4 primers and the sequences compared to the GenBank database using the BLASTn program [1], which showed that this isolate is closer to Beauveria amorpha than to Beauveria bassiana. Conidia suspensions were prepared in 0.01%. Tween 80 solution from fungi grown on Sabouraud dextrose agar plates. After washing in H2O, conidia suspensions were maintained in 10% glycerol at a concentration of 108 conidia · mL−1. For the experiments, spores (106 mL−1) were inoculated in 100 mL of liquid Cove’s medium (MC) supplemented with 0.05% yeast extract [9]. As a carbon source, crystalline chitin, tick cuticle (B. microplus), glucose, or N-acetylglucosamine (GIcNAc) was added to the medium at different concentrations. Alanine, glycine, methyonine, and valine (0.5%) were added together with the fungal inoculum and at each 24 h of growth. After 72 h of incubation on a rotating shaking platform (150 rpm) at 27°C, the mycelium was removed by filtration on Whatman No. 1 filter paper. Prior to use in enzymatic assays, the filtrates were dialyzed against 20 mM Tris–HCI buffer (pH 8.0). The total protein content was determined by the Bradford method [6], with a known concentration of BSA as the standard.

Analytical procedure

Subtilisin-like protease was determined with N-Suc-(Ala)2-Pro-Phe-p-nitroanilide (Sigma). The reaction mixture was 15 μL substrate (2 μM), 10 μL enzyme sample, and 75 μL 50 mM Tris–HCL, pH 8.0. The kinetic assay was done in a Spectra Max 250 and read in a Softmax Pro (405 nm/30 min) (Molecular Devices). Enzyme activity is expressed as nanomoles nitroanilide (NA) released per milliliter per minute at 37°C [22]. The specific activity is represented as units per microgram of protein. Assays were performed in three independent experiments, with four replicates for each sample. Statistical and data analyses were performed using SPSS for Windows (Release 8.0, 1997). Tukey HSD (p < 0.05) was used for comparisons of means.

Chitinolytic activity was determined using N,N′,N′′,N′′′-tetracetylchitotetraose (4 mM) to detect endochitinase. The reaction mixture was 40 μL 0.2 mM acetate buffer (pH 5.4)/10 μL substrate/120 μL sample. After 1 h of incubation at 37°C the amount of N-acetylglucosamine (GIcNAc) released was determined as described [22]. One unit of chitinase was defined as the amount of enzyme that releases 1 μmol of GIcNAc per minute at 37°C.

Scanning electron microscopy (SEM)

Groups of 12 engorged B. microplus females were immersed for 30 s in B. bassiana or B. amorpha conidial suspensions (106 conidia · mL−1). Sterile distilled water was applied to the control ticks. After treatment, ticks were maintained in petri dishes at 28°C and 85% relative humidity for up 4 days. For SEM analysis, ticks were fixed overnight at 4°C with 2% (v/v) glutaraldehyde, 2% (v/v) paraformaldehyde in 0.1 M sodium cacodylate buffer at pH 7.2. Postfixation was carried out in 1% (w/v) osmium tetroxide in the same buffer. The specimens were rinsed in buffer, dehydrated in a series of 30–100% acetone solutions, dried at critical point in CO2 (CPD 030 BALTEC), and coated with gold in a sputter-coater (SCD 050 BALTEC). The material was examined in a Jeol JSM 5800 scanning electron microscope at the Centro de Microscopia Electrônica da Universidade Federal do Rio Grande do SuI (CME/UFRGS, Porto Alegre/RS).

Results and Discussion

Fungal extracellular hydrolytic enzymes are important for degradation of the host cuticle during infection, facilitating penetration and providing nutrients for further growth. Like most fungal pathogens, B. amorpha and B. bassiana might use a combination of enzymes and mechanical force to penetrate the host cuticle and access the nutrient-rich host hemolymph. The effects of different carbon sources on chitinase and protease secretion by B. amorpha and B. bassiana were tested in medium supplemented with simple or complex carbon sources individually or in combination. As shown in Tables 1 and 2, B. amorpha and B. bassiana produced chitinases and proteases in all media tested; however, the amount of secreted enzymes varied. The highest levels of endochitinase activity were found in culture supernatants from tick cuticle and chitin for both fungi. Glucose (1%) and GIcNAc (1%) repressed enzyme secretion (Table 1). The effect of glucose repression was previously described for proteins utilized in carbohydrate degradation pathways. According to St Leger et al. [25–27] GIcNAc might cause catabolite repression of chitinases when in excess of the immediate growth requirements of the organisms. In M. anisopliae, GIcNAc shows a special dual regulation on chitinase production. It induced the production and secretion of the enzyme at low concentrations but repressed chitinase secretion at higher concentrations [3]. This effect was also observed for the extracellular endochitinase of B. amorpha and B. bassiana. When GIcNAc was added to media at a concentration of 0.5%, chitinase activity was 11- and 8-fold higher compared with 1% GIcNAC for B. amorpha and B. bassiana, respectively (Table 1). Even when 0.5% GIcNAc was added to a chitin-containing medium, similar results were observed, on a more moderate scale (Table 1).

Table 1 Effect of carbon sources and amino acids on secretion of chitinases from B. amorpha and B. bassiana
Full size table
Table 2 Effect of carbon sources and amino acids on secretion of subtilisin-like protease from B. amoropha and B. bassiana
Full size table

High levels of subtilisin activity were observed in culture supplemented with tick cuticle for both fungi (Table 2). Since arthropod cuticles comprise about 70% protein, this enzyme may play an important role in host penetration. The addition of alanine (0.5%) to the culture medium repressed subtilisin secretion (Table 2). St Leger et al. [28] verified that in M. anisopliae, alanine addition repressed both apressorium formation and the release of subtilisin-like proteases. Alanine is the major amino acid found in the cuticle of insects and also in B. microplus [16]. Moreover, comparison between pr1 cDNA cloned from B. bassiana [18] and pr1 cDNA cloned from M. anisopliae [28] showed significant similarity.

Entomopathogenic fungi evolved distinct strategies for their attachment to hosts, varying considerably in their modes of action, virulence, and degree of host specificity [8]. Direct penetration of intact cuticle is the normal mode of entry by most entomophatogenic fungi. B. amorpha and B. bassiana are not exceptions, and the conidia are capable of germination on the host surface and often differentiate to form apressoria. Beauveria comprises two main insect pathogenic species, B. bassiana and Beauveria brongniartii, which are mainly parasitic on Lepidoptera and Coleoptera [24]. Beauveria species are classified by the shape of their conidia and the placement of conidia on the conidiogeneous apparatus [14]. Traditionally, the main difference among the most common species is the shape and size of the conidia. SEM analysis of infected B. microplus showed that B. bassiana and B. amorpha conidia are capable of attaching to the epicuticle surface (Fig. 1). The conidial morphology of B. bassiana was generally spherical (Fig. 1A), whereas B. amorpha conidia on ticks were often flattened on one side as in the original description (Fig. 1B). Adherence of conidia to the host surface is probably mediated by hydrophobic interactions between conidia and the arthropode cuticle [4] and production of an adhesive mucous layer. The fungus produces a thin amorphous mucilage layer and it firmly adheres the conidia and germ tubes to the tick integument (Fig. 1B). The first sign of conidia germination is germ-tube extrusion. Each conidium from both species usually produced only one germ tube that differentiated in apressoria and penetrated the tick cuticle (Figs. 1A and B). B. bassiana presented conidiogenous cells formed in tightly clustered groups (Fig. 2), whereas B. amorpha sometimes had solitary conidiogenous cells.

Fig. 1
figure 1

Scanning electron microscopy of Booplhilus microplus cuticle infected by Beauveria bassiana (A) and Beauveria amorpha (B). Detail of conidia (CO), germ tube (GT), and appressoria (AP) formation and penetration on the tick epicuticle surface 3 days postinfection. Arrow in B indicates the mucilage layer. Bar = 2 μm.

Full size image
Fig. 2
figure 2

Detail of B. bassiana conidiogenous cells formed in tightly clustered groups. Bar = 5 μm.

Full size image

For most aspects of B. microplus infection by B. amorpha and B. bassiana, our observations are consistent with the commonly described sequence of events that characterizes other entomopathogenic fungal interactions [2, 8]. The penetration mode of entomopathogenic fungi is similar to that of plant pathogenic fungi and is suggested to be based on a combination of mechanical pressure and enzymatic degradation [28]. Appressoria adhere to the plant surface by secreting a potent glue [11]. The force is exerted vertically and might be efficiently directed to the cuticle [10].

We showed that B. amorpha and B. bassiana produce subtilisin-like proteases and chitinases in the presence of tick cuticle. The multiplicity of these enzymes provides a major challenge in determining the role played by each particular enzyme in adaptation to a new environment or in pathogenicity. The high capacity of the secretion machinery of these fungi is still to be exploited for biotechnological purposes. However, our knowledge of the fungal secretion pathway is still at an early stage.