VIRULENCE OF METARHIZIUM ANISOPLIAE TO EMBRYOS OF THE GRASS SHRIMP PALAEMONETES PUGIO
Fred J. Genthner, Steven S. Foss and Patricia S. Glas
U.S. Environmental Protection Agency and National Research Council, Environmental Research Laboratory, 1 Sabine Island Drive, Gulf Breeze, Florida 32561.
SUMMARY
Experiments were performed in which developing embryos of the grass shrimp, Palaemonetes pugio, were exposed to conidiospores of the insect pathogenic fungus, Metar-hizium anisopliae. Responses were variable with significant (p 0.05) adverse effects observed in 5 out of 6 experiments conducted. Dead embryos and larvae with visible growth of Metarhizium anisopliae were observed in all experiments. Growth of Metarhizium anisopliae was occasionally observed on embryos and larvae prior to death. Delayed hatch was also observed. In one of the initial experiments an increase in N-acetyl--D-glucosaminidase, EC 3.2.1.30, (NAGase) activity accompanied by an increase in virulence toward shrimp embryos was observed. Additional experiments in which conidiospores were produced on homogenized caterpillars suggested a positive correlation between virulence of M. anisopliae to P. pugio embryos and activity of spore-associated NAGase. M. anisopliae was an invasive pathogen of grass shrimp embryos, and the growth substrates on which their spores develop can "condition" them for enhanced virulence toward nontargets.
Key words: Palaemonetes pugio, Metarhizium anisopliae, virulence, biological control, N-acetyl--D-glucosaminidase
INTRODUCTION
Metarhizium anisopliae is an imperfect, entomopathogenic fungus found in soils throughout the world. It was first recognized as a biocontrol agent in the 1880's. Four groups of insect pests (termites, locusts, spittlebugs and beetles) are currently being targeted for control by M. anisopliae (Zimmermann, 1993). M. anisopliae is applied as spores or mycelia in vahous formulations. Control is achieved through the induction of a fungal epizootic where new spores and vegetative cells produced in infected insects are spread to healthy members of the population. The dynamic nature of this process instructs that safety remain a major factor in the development and use of M. anisopliae and other microbial pesticides.
Zimmermann (1993) summarized the safety data of M. anisopliae and reported that when the fungus was applied by different methods to birds, fish, mice, rats, guinea pigs or rabbits, no toxicological or pathological symptoms were observed. Recently; however, Genthner and Middaugh (1995) reported that when developing embryos of the inland silverside fish, Menidia beryllina, were exposed to conidiospores of M. anisopliae, several adverse effects were observed in both embryos and newly-hatched larvae. In a follow-up study designed to validate embryo tests for determining adverse effects of fungal pest control agents, Genthner et al. (1995) presented data from a single experiment that suggested M. anisopliae was an invasive pathogen of embryos of the grass shrimp, Palaemonetes pugio.
To expand upon this preliminary finding, we report the results of several additional experiments where grass shrimp embryos were exposed to conidiospores of M. anisopliae. Data are presented which correlate virulence to enzymatic activity associated with ungerminated conidiospores. In addition, the effect of the growth substrate on enzymatic activity and, ultimately, virulence is also investigated.
MATERIALS AND METHODS
Cultivation of Fungus and Recovery of Spores. Metarhizium anisopliae 1080 was obtained from the USDA-ARS collection (Ithaca, NY) entomopathogenic fungi. The fungus was received on an agar plate as a sporulated culture. Conidiospores from this plate served as inocula for all experiments. Conidiospores were produced at 25C on either glucose-yeast extract-basal salts (GYBS) agar medium (Boucias et al., 1988) or on a homogenate of corn earworm, Helicoverpa zea, previously infected with M. anisopliae. To prepare the homogenate of corn earworm previously infected with M. anisopliae, second instar caterpillars were dipped into a suspension of M. anisopliae conidiospores (ca. 1 x 107 ml-1). Exposed caterpillars were then placed in a Petri dish containing approximately 3 g of sterile diet consisting of ground raw pinto beans, 14 g; wheat germ, 10g; torula yeast, 6.3 g; casein, 5.0 g; agar, 2.3 g; and distilled water, 135 ml. After five days, dead caterpillars displaying external conidiogenesis were stored frozen at -80C.
Approximately one week before conidiospores were needed for the shrimp embryo tests, either 0.3 ml of the M. anisopliae-infected, caterpillar homogenate caterpillars were spread upon the surface of an MM agar plate (0.1 % KH2PO4, 0.05% MgSO4, 1.5% agar) or a loopful of conidiospores, removed from the agar plate received from the USDA-ARS collection, was spread upon the surface of a GYBS agar plate. On the first day of testing, conidiospores were scraped from the surface of sporulating cultures and suspended in filtered (0.22 µm) sea water at a salinity of 20 ± 0.5 (FSW) by gentle aspiration in a hand-held tissue homogenizer. Densities of the conidiospores were determined using a hemocytometer. Viable spore counts were performed by diluting spores in sterile distilled water containing 0.03% Triton X-100 (Union Carbide, Indianapolis, IN), spreading the dilutions onto GYBS agar plates, and counting the colonies which appearing after a 6-day incubation. Conidiospores produced on GYBS agar were used in Expts 1-4 and 6 while conidiospcres produced on the corn earworm homogenate were used in Expts 5 and 6.
M. anisopliae was recovered from exposed embryos and larvae in the following manner. Infected animals were first extensively washed in FSW to eliminate spores carried over from the test water. Washed embryos and larvae were then homogenized and dilutions were spread onto the surface of GYBS agar plates. Fungal colonies arising from these plates were identified as M. anisopliae by the size, shape, color and arrangement of the spores.
Enzyme assays. At the beginning of each shrimp embryo test, activities of several hydrolytic enzymes on the surface of the ungerminated conidiospores were analyzed using the API ZYM® system (Analytab Products, Plainview, NY). Sixty-five microliters of conidiospores, suspended in FSW at a density of 1 x 107 ml-1, were added to each microtube in the API ZYM® test strip. After inoculation the strips were incubated for 4 h at 37°C. Strips were read, and the reactions were recorded according to the manufacturer's instructions. N-acetylglucosaminidase (NAGase) activity on the conidiospores was also assayed separately according to the procedures of Bowers et al. (1980) and Borooah et al. (1961), as adapted by the Sigma Chemical Co. (St. Louis, MO). Concentrations in a 1.10 ml reaction mix were: citric acid, 10 mM; p-nitrophenyl-N-acetyl--D-glucosamine (Sigma), 0.91 mM; bovine serum albumin, 0.05% (w/v). Formation of the p-nitrophenol product was determined at 405 nm.
P. pugio embryo tests. Developing embryos of the grass shrimp, P. pugio, were used in assessment of virulence of M. anisopliae conidiospores to nontarget aquatic arthropods. P. pugio collection, maintenance, and shrimp embryo testing were performed as described by Fisher and Foss (1993). Briefly, single embryos, at the blastopore/tissue cap stage (age 2-4 days after oviposition), were placed into Leighton® culture tubes containing 50 µl of FSW. Six ml of FSW at a salinity of 20 ± 0.5 . were added to each "no exposure" control tube, and six ml of conidiosidores suspended in FSW were added to each exposure tube. Experimental treatments consisted of a "no exposure" control, a heat-killed control, and either three conidiospore suspensions at densities of approximately 104, 105, and 106 ml-1, or two conidiospore suspensions (ca. 106 ml-1) produced on different growth media. A "heat-kiiled" control was prepared by sterilizing a conidiospore suspension (ca. 1 x 106 ml-1) in an autoclave (20 min. 15 lb/in2) . Each treatment consisted of three replicates of 10 embryos each. Tubes were incubated at 27C with slow shaking (60 rpm). Using an inverted microscope, individual embryos were examined daily for stage of development, viability, and extent of fungal growth on the extra-embryonic coat. Microbiological or histological analyses to recover or identify M. anisopliae was performed following death. Any embryo that failed to hatch 48 h after completion of the normal hatching period was scored as a delayed hatch.
Statistical analysis of data. At 15 days post-extrusion, numbers of live embryos and larvae and dead embryos and larvae in each treatment group were counted and percent alive and normal calculated. Each treatment group (30 embryos) was randomly divided into thirds to permit statistical analysis. A one-way analysis of variance (SAS Institute, Inc.,1985) was performed. The analysis of variance model included differences among replicates within treatments. Dunnetts test was used compare all treatments against the "heat-killed" control.
RESULTS
Developing P. pugio embryos were exposed to conidiospores of M. anisopliae in 6 separate experiments run consecutively between May 26, 1994, and November 27, 1994. Responses were variable with significant (p < 0.05) adverse effects observed in 5 of 6 experiments conducted (Table 1, Figure 1). No significant (p > 0.05) differences were found among replicates within treatments. In all experiments, the "viable spore" treatment groups harbored at least one dead embryo or larvae with visible fungal growth. Fungal mycelia in or on the animals was cultured and identified as M. anisopliae. Growth of M. anisopliae was occasionally observed on embryos and larvae prior to death. In the control treatments fungal growth was never observed on embryos or larvae, although a few deaths occurred.
Expts 1-4 (Table 1, Figure 1) were replicated with freshly prepared conidiospore suspensions. Conidiospores, produced on GYBS, were approximately the same age (6- 8 d) and were started with an inoculum from the same source. There was substantial variability between experiments in the severity and effects on developing shrimp embryos (Figure 1).
Variability was also observed in the activities of several conidiospare-associated enzymes measured in the conidiospore suspensions prepared for each replicated experiment (Table 2). Activities of 18 enzymes associated with ungerminated spore suspensions were measured semi-quantitatively with the API ZYM® system. Although API ZYM® test strips were not replicated in each experiment, further studies showed that when the same conidiospore preparation was analyzed by the system identical results were obtained.
The activity of NAGase was also measured quantitatively using the chromogenic substrate p-nitrophenyl--D-glucosamine. A semi-quantitative NAGase activity of 30 nanomoles of substrate cleaved was equivalent to a quantitative activity of 0.46 units of NAGase/ml where a unit of NAGase was defined as the hydrolysis 1.0µM of p-nitrophenyl--D-glucosamine to p-nitrophenol and N-acetyl--D-glucosamine per minute at pH 4.8 at 37C.
In Expt 1 all control shrimp remained alive and normal with moribund or dead larvae the predominant adverse responses in the viable conidiospore treatment groups. Growth of M. anisopliae was observed on two of the three dead embryos at the medium spore density treatment. M. ansiopliae mycelia were seen emerging from the exoskeleton on >90% of dead larvae (Figure 1). Significant adverse responses compared against the controls were obtained in the low and medium conidiospore treatments (Table 1). No clear dose-response was observed in the experimental treatments.
Delayed hatch or failure to hatch, was first observed in Expt 2 (Figure 1). The highest incidence of delayed hatch was observed at the low spore density where 6 out of 30 embryos demonstrated this effect. Delayed hatch was not limited to the experimental treatments, as a single delayed hatch was also observed in the control. Despite the fact that dead embryos with fungal growth were observed in both the medium and high conidiospore treatments, the adverse effects were not significant in Expt 2 (Table 1).
An unanticipated increase in NAGase activity was measured in the conidiospore suspension used for Expt 3 (Table 2). Of the 18 hydrolytic enzymes measured with the API ZYM® system, NAGase was the only enzymatic activity which increased between Expts 1, 2 and 3 (Table 2). The NAGase activity in the conidiospore suspension used in Expt 3 was 3 times greater than in the conidiospore suspensions used in Expts 1 and 2, a difference of 20 nanomoles of cleaved substrate after 4 h at 37C. This enzymatic increase was accompanied by an increase in virulence of M. anisopliae conidiospores to developing grass shrimp embryos. A distinct dose-response was also observed; the percentage of shrimp showing adverse effects increased from 50 to 77, and then to 90 for the low, medium and high conidiospore treatments, respectively (Figure 1). Dead embryos with growth of M. anisopliae was the most prevalent adverse effect. At the high spore density treatment, 20 embryos died, displaying M. anisopliae growth. Compared against the control, significant adverse effects were obtained in all viable conidiospore treatments, with adverse effects at the high conidiospore density treatment being significantly greater than the adverse effects at the low conidiospore density treatment (Table 1).
Compared with Expt 3, virulence of the M. anisopliae conidiospores to developing grass shrimp embryos was greatly reduced in Expt 4 (Figure 1). There were, however, 4 dead embryos and 3 dead larvae displaying growth of M. anisopliae at the low conidiospore density and 3 dead embryos and 1 dead larvae with growth of M. anisopliae at the high conidiospore density. When compared against the control, significant adverse effects were obtained at the low and high conidiospore density treatments. Examination of the enzymatic activities of the conidiospore suspension used for Expt 3 revealed that the greatest difference was in NAGase activity that had been reduced to 20 nanomoles of cleaved substrate.
Conidiospores in Expt 5 were produced by growing M. anisopliae on homogenized corn earworm tissue. Results of this test were very similar to the results in Expt 3 with high virulence, an obvious dose-response, and high NAGase activity (30) in the conidiospore suspension (Figure 1, Table 2).
Expt 6 was designed to compare effects when embryos obtained from a single female were exposed to conidiospores produced on either GYBS or homogenized corn earworm. As shown in Figure 1, conidiospores produced on homogenized corn earworm were more virulent toward developing shrimp embryos than were conidiospores produced on GYBS. The percentage of shrimp "alive and normal" at day 15 was 80 for the GYBS-produced conidiospore treatment compared to 60 for the corn earworm produced conidiospore treatment. When compared against the control, a significant adverse effect on the developing embryos was obtained only with conidiospores produced on homogenized corn earworm (Table 1). Examination of the enzymatic activities of the conidiospore suspensions revealed that NAGase activity was higher in the corn earworm-produced conidiospore suspension. Trypsin, acid phosphatase, -glucosidase and phosphoamidase activities in the com earworm-produced conidiospore suspension were also higher than in the GYBS-produced conidiospore suspension, and also higher than in the com earworm-produced conidiospore suspension used in Expt 3 (Table 2).
DISCUSSION
Infection of insects by M. anisopliae requires adhesion, pre-penetration growth, penetration into the host (for a review see St. Leger, 1993) and establishment of the pathogen in the host (Charnley, 1989; Samoson et al., 1988). The M. anisopliae infection process in developing shrimp embryos shared characteristics similar to those in insects. In all experiments., conidiospores, both viable and heat-killed, were frequently observed attached to the outer envelope of shrimp embryos. Large numbers of attached conidiospores were not observed until 2 days after exposure. This attachment delay was in agreement with St. Leger (1993) who reported that enhanced adhesion of Metarhizium spores occurs with the secretion of a mucilaginous coat during conidial hydration. No direct relationship between conidiospore attachment to the envelope and infection was observed.
Delayed hatch was observed in Expts 2 and 3. The highest incidences of delayed hatch occurred at the low conidiospore densities, with 20% in Expt 2 and 13% in Expt 3. In all controls, only a single delayed hatch was observed. In other shrimp embryo toxicity tests performed at this laboratory, delayed hatch was either an infrequent, rare event (Fisher and Foss, 1993; Genthner et al., 1994) or was observed as a sublethal effect in embryos exposed to other toxicants (Rayburn et al., 1995). In our research, delayed hatch was probably an adverse effect caused by exposure of the embryos to conidiospores of M. anisopliae.
The M. anisopliae infection process in developing shrimp embryos was first observed 2 to 3 days after exposure as an area on the outer envelope where several conidiospores had attached. Over the next 3 days, conidiospores germinated on the surface of the outer envelope and produced hyphae which formed a "foci of infection" or "infection court" (see Fig. 2A). Appressoria were not observed on the surface of the outer envelope. Either Metarhizium, strain 1080, did not produce appressoria or extensive mycelial growth over the site of penetration prohibited this observation. Dead or moribund embryos showing internal growth of M. anisopliae through the envelope were not observed until the sixth or seventh day of exposure. Hyphae would re-emerge once extensive fungal growth had occurred inside the embryo.
Extensive internal growth of M. anisopliae preceded embryo death. Larval death followed by emergence of hyphae from the shrimp exoskeleton shortly after hatching suggests M. anisopliae was inside the embryo prior to hatching. These observations imply that M. anisopliae strain 1080 does not produce large amounts of toxic secondary metabolites during infection. If large amounts of toxin were produced, extensive fungal growth might not have been observed inside living embryos, and embryos infected with M. anisopliae might not have successfully hatched.
Expts 1-4 were replicates conducted on different days. Inocula were from the same source; conidiospores were of the same age and were produced on the same medium, yet differences in virulence and the enzymatic activities associated with each conidiospore preparation were observed in each experiment. Samsinakova and Kalalova (1993) reported that there was a high likelihood of variability in genetically nonuniform, multispore fungal strains and that these variations may be responsible for alterations in virulence. Changes in virulence have been attributed to the enzymatic activity in the fungal strains (Charnley and St. Leger, 1991).
We observed that changes in the enzyme activities of ungerminated conidiospores occurred spontaneously over time from culture to culture and that changes in virulence accompanied changes in some of the enzymatic activities. For example, the sudden increase in virulence in Expt 3 was accompanied by a large increase in conidiospore-associated NAGase activity. Indeed, in Expts 3, 5 and 6 where the highest percentages of dead embryos were observed, the conidiospore-associated NAGase activity was consistently measured at the highes" activity of 30 nanomoles of cleaved substrate (Table 2). However, since the activities of several other conidiospore-associated enzymes varied among experiments, a single enzyme may not be controlling virulence.
By exposing embryos from a single brood to conidiospores produced on two different substrates, environmental conditions that might pre-adapt M. anisopliae to a pathogenic lifestyle were investigated. Conidiospores produced on GYBS possessed lower NAGase activity that the conidiospores that were produced on homogenized corn earworm. The corn earworm-produced conidiospores showed significantly higher virulence toward shrimp embryos (Figure 1). If a chitin-like oligosaccharide was a component of the outer extra-embryonic coat of P. pugio embryos, then a direct argument could be made regarding NAGase as a contributing virulence factor. Whether the outer coat contains chitin remains to be learned. However, lectin binding to the extra-embryonic coat of P. pugio detected carbohydrate moieties (mannose and N-acetylgulcosamine) that may be substrates for conidiospore associated enzymes (Glas et al., 1995).
Embryos used in Expts 1-5 were all from separate broods. Raybum et al.
(1995), working with the water soluble fraction of #2 fuel oil, reported that embryos from different broods were a source of variability in the shrimp embryo toxicity test. Our results in Expt 6, where embryos from the same brood were exposed to conidiospores produced on two different substrates, suggest that if there were variability among broods, it was not great enough to conceal the differences in virulence caused by the different conidiospore preparations.
Variable responses were not unique to this study. Genthner and Middaugh
(1992, 1995) and Middaugh and Genthner (1994) exposed developing embryos of the inland silverside fish, Menidia beryllina, to conidiospores of Beauveria bassiana or M. anisopliae and reported highly variable results. In these studies the activities of enzymes on the ungerminated spores were not measured. Thus, the conclusion was made that unknown changes in the conidiospores contributed to the variability. The results of this present study suggest that some of these changes are in the activities of conidiospore associated enzymes.
The results from Expt 5 and 6 in which the conidiospores were produced on homogenized corn earworm support the results of St. Leger et al. (1991) who found that levels of enzymes on conidiospores from infected Manduca sexta larvae were higher than those harvested from Sabauraud dextrose agar, indicating that environmental conditions in which conidiospores of M. anisopliae develop can pre-adapt them for the pathogenic life style. Our measurements of the enzymatic activities on the conidiospores, however, go on to suggest that changes in activities of these cuticle-degrading enzymes may correlate with changes in virulence of this fungus toward a nontarget organism. In addition, the substrate upon which conidiospores are produced does not always govern the activities of conidiospore-asscciated enzymes, because sudden changes in enzymatic activities and virulence can occur.
Lastly, in terms of risk assessment, two important questions wait to be addressed. First, can predictions on nontarget effects be made on the basis of enzyme activity profiles? Second, can nontarget effects be minimized by producing conidiospores possessing certain enzymes at specific activities?
REFERENCES
Borooah, J., Leaback, D. H., and Walker, P. G. 1961. Studies on glucosaminidase 2. Substrates for N-acetyl--glucosaminidase. Biochem J. 78, 106-1 10.
Boucias, D. G., Pendiand, J. C., and Latgé, J. P. 1988. Nonspecific factcrs involved in attachment of entomopathogenic deuteromycetes in host insect cuticle. Appl. Environ. Microbiol. 54, 795-1805.
Bowers, G. N. Jr., McComb, R. B., Christensen, R. G. Jr., and Schaffer, R. 1980. Highpurity 4-nitrophenol: Purification, characterization, and specifications for use as a spectrophotometric reference material. Clin. Chem. 26, 724-729.
Charnley, A. K. 1989. Mechanisms of fungal pathogenesis in insects. In: The Biotechnology of Fungi for Improving Plant Growth. (J. M. Whipps and R. D. Lumsden, Eds.), pp. 85-125. Cambridge University, London.
Chamley, A. K., and St. Leger, R. J. 1991. The role of cuticle-degrading enzymes in fungal pathogenesis in insects. In: The fungal spore and disease initiation in plants and animals" (G. T. Cole and H.C. Hoch, Eds.), pp. 267-286, Plenum, New York.
Fisher, W. S., and Foss, S.S. 1993. A simple test for toxicity of number 2 fuel oil and oil dispersants to embryos of grass shrimp, Palaemonetes pugio. Mar Pollut. Bull. 26, 385-391.
Genthner, F. J., Foss, S. S., and Fisher, W. S. 1994. Testing of the insect pest control fungus Beauveria bassiana in grass shrimp Palaemonetes pugio. Dis. Aquat. Org. 20, 49-57.
Genthner, F. J., and Middaugh, D. P. 1995. Nontarget testing of an insect control fungus: effects of Metarhizium anisopliae on developing embryos of the inland silverside fish Menidia betyllina. Dis. Aquat. Org. 22,. 163-171.
Genthner, F. J., Middaugh, D. P., and Foss, S. S. 1995. Validation of embryo tests for determining effects of fungal pest control agents on nontarget aquatic animals. Arch. Environ. Contam.Toxicol. 29, 540-544.
Glas, P. S., Genthner, F. J., and Fisher, W. S. 1995. Extra-embryonic coats of
Palaemonetes pugio: morphology and composition. Am. Zool. 35, 70A.
Rayburn, J. R., Glas, P. S., Foss, S. S., and Fisher, W. S. 1996. Characterization of grass shrimp (Palaemonetes pugio) embryo toxicity tests using the water soluble fraction of number 2 fuel oil. (Accepted, Mar. Poll. Bull.).
St. Leger, R. J., Goettel, M., Roberts, D. W., and Staples, R. C. 1991. Pre-penetration events during infection of host cuticle by Metarhizium anisopliae. J. Invertbr. Pathol. 58, 168-179.
Sampson, R. A., Evans, H.C., and Latgé, J. P. 1988. "Atlas of Entomopathogenic
Fungi" Springer-Verlag, Berlin.
Samináková, A., and Kálalová, S. 1983. The influence of a single-spore isolate and repeated subculturing on the pathogenicity of conidia of the entomophagous fungus Beauveria bassiana. J. Invert. Pathol. 42, 156-161.
SAS Institute, Inc. 1985. SAS user's guide: Statistics, version 5 edition. SAS Institute Inc., Cary, NC.
St. Leger, R. J. 1993. Biology and mechanisms of invasion of deuteromycete fungal pathogens. In:Parasites and Pathogens of Insects. Vol. 2 (N. C. Beckage, S. N. Thompson, and B. A. Federici, Eds.), pp. 211-229. Academic Press, NY.
Zimmermann, G. 1993. The entomopathogenic fungus Metarhizium anisopliae and its potential as a biocontrol agent. Pestic. Sci. 37, 375-379.
Figure 1. Responses of embryonic and larval grass shrimp Palaemonetes pugio to
conidiospores of Metarhizium ansiopliae in Expts 1-6. Ctl: no conidicspcre control;
KCtl: heat-killed conidiospore control; Low: approximately 1 x 104 conidiospores ml-1; Med: approximately 1 x 105 conidiospores mi-1; High: approximately 1 x 106
conidiospores ml-1; -C: conidiospores produced on homogenized corn earworm; -G: conidiospores produced on GYBS agar.