And Corn Rootworm (Coleoptera: Chrysomelidae)

Home , Pentatomidae

BIOLOGICAL AND MICROBIAL CONTROL Toxicity of Chromobacterium subtsugae to Southern Green Stink Bug (Heteroptera: Pentatomidae) and Corn Rootworm (Coleoptera: Chrysomelidae)

1 2,3 2 PHYLLIS A. W. MARTIN, EDSON HIROSE, AND JEFFREY R. ALDRICH

USDAÐARS, Insect Biocontrol Laboratory and Chemicals Affecting Insect Behavior Laboratory, Beltsville, MD 20705-2350

J. Econ. Entomol. 100(3): 680Ð684 (2007) ABSTRACT Diabrotica spp. (Coleoptera: Chrysomelidae) beetles and southern green stink bugs, Nezara viridula (L.) (Heteroptera: Pentatomidae), are pests on corn, Zea mays L., and soybean, Glycine max (L.) Merr., as well as on cucurbits. Control of these insects has depended on chemicals. An alternative to chemical control is the use of biologicals. Use of bacteria, fungi, viruses, pheromones, and metabolites to control these insects can potentially improve resistance management and reduce pesticide use. Other than Bacillus thuringiensis Berliner, few bacteria have been discovered that are lethal to either of these pests. Chromobacterium subtsugae Martin et al., a newly described bacterium that is known to be toxic to Colorado potato beetle, Leptinotarsa decemlineata (Say), larvae, was found to be toxic to both diabroticite adult beetles and southern green stink bug adults. In laboratory assays, toxins produced by these bacteria kill 80Ð100% of the adults of two species of diabroticite beetles, Diabrotica undecimpunctata howardi Barber and Diabrotica virgifera virgifera LeConte, and 100% of southern green stink bug adults within 6 d. For green stink bug, live bacteria were not needed for toxicity.

KEY WORDS Diabrotica, Nezara viridula, biocontrol

Corn rootworm beetles are principal pests of corn, (Balsamo) Vuillemin and Metarhizium anisopliae although certain species of diabroticite beetles also (Metschnikoff), have been suggested for control of feed on soybean, Glycine max (L.)Merr.; melons (Cu- these pests (Krueger and Roberts 1997, Pereira and cumis spp.), cucumbers, Cucumis sativus L., and pea- Roberts 1991, Sosa-Gomez and Moscardi 1998). Bac- nuts, Arachis hypogea L. In addition to damaging the teria, other than Bacillus thuringiensis Berliner strains crop, the adult beetles transmit bacterial wilt and viral for corn rootworm (Diabrotica spp.) (Baum et al. diseases (York 1992). The southern green stink bug, 2004), are not used for control of these insects. Other Nezara viridula (L.) (Heteroptera: Pentatomidae), is bacteria such as pigmented Serratia marcescens Bizo a pest of solanaceous crops, but it also has more gen- have been isolated from dead insects and implicated as eral tastes that extend to soybean, other legumes, the causative disease agent (Grimont and Grimont cotton (Gossypium spp.), and various vegetables, in- 1978). Pathogens have traditionally been identiÞed cluding tomatoes (Lycopersicum spp.) (McKinlay et using KochÕs postulates (as per Black 1996) in which al. 1992). These insects are presently controlled by the the Þrst step in identifying the causal agent is isolation use of insecticides, either prophylactically or in an from diseased insects. integrated pest management program. The propensity Violet-pigmented bacteria such as Chromobacte- of insects to develop resistance to insecticides has rium violaceum Bergonzini have infrequently been encouraged the use of other means of control. isolated from insects, and they have not been previ- An alternative to traditional chemical control is the ously considered harmful to insects (Bucher 1981). In use of pathogens. Fungi, such as Beauveria bassiana the larger grain borer, Prostephanus truncatus (Horn), these bacteria may be involved in cellulose digestion Mention of a trademark or proprietary product does not constitute (Vazquez-Artista et al. 1997), forming a symbiotic a guarantee or warranty of the product by the United States Depart- rather than a pathogenic association. The genome of ment of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable. C. violaceum has been sequenced. In addition to the 1 Corresponding author: USDAÐARS Insect Biocontrol Laboratory, violet pigment violacein, which has antimicrobial ac- Beltsville, MD 20705-2350 (e-mail:[email protected]). tivity against gram-positive and gram-negative bacte- 2 Chemicals Affecting Insect Behavior Laboratory, Beltsville, MD ria (Duran et al. 1983) and Trypanosoma cruzi Chagas 20705-2350. 3 Current address: Universidade Federal do Parana´, Departamento (Duran et al. 1994), it contains a gene similar to an de Zoologia, Caixa Postal 19020, Curitiba, PR 81531-990, Brazil. insecticidal gene found in Photorhabdus luminescens June 2007 MARTIN ET AL.: TOXICITY OF C. subtsugae TO PEST INSECTS 681

Thomas and Poinar and in Xenorhabdus nematophilia versity. Western corn rootworms, Diabrotica virgifera Poinar and Thomas (Brazilian National Genome virgifera LeConte, were obtained from C. Neilson Project Consortium 2003). (USDAÐARSÐNGIRL, Brookings, SD) and maintained A new species of Chromobacterium, Chromobacte- under the same conditions as described for southern rium subtsugae Martin et al., was found to be toxic to corn rootworms. Colorado potato beetle, Leptinotarsa decemlineata Bioassays. Southern green stink bugs were incu- (Say) Martin et al. 2004). Here, we describe the tox- bated 48 h in plastic cages without water. Twenty icity of these bacteria in laboratory assays to other pest adults were used per treatment with Þve stink bugs per insects: two corn rootworm species and southern cage. Males and females were tested separately. The green stink bug. suspensions of C. subtsugae were delivered to the adults in 1 ml of liquid composed of 0.5 ml of distilled water and 0.5 ml of bacterial suspension. The micro- Materials and Methods centrifuge tubes containing the test material or water Bacterial Strains and Media. Violet colonies of bac- as a control were plugged with cotton and replaced terial strain PRAA4-1 were isolated in summer 2000, every 3 d. Mortality was recorded daily for 6 d. Survival from Maryland forest soil plated on L-agar, by S. Stone times were determined with LIFEREG procedure by (Martin et al. 2004). The violet bacteria were Gram- using the Weibull distribution with 95% conÞdence rods, preliminarily identiÞed as C. violaceum by com- intervals (SAS Institute 2004), because of limitation of parison with descriptions in BergeyÕs manual (Sneath numbers of insects (Preisler and Robertson 1989). We 1984). When the Þrst 500 bp of the 16S ribosomal DNA attempted feeding C. subtsugae to nymphs, but the were sequenced, the identiÞcation was only to the nymphs did not feed consistently in a liquid delivery genus Chromobacterium (Accugenix, Newark, DE). system. This bacterial strain has been subcultured weekly at The toxicity of C. subtsugae to green stick bug adults room temperature on L-agar (Atlas 1997). For bacte- also was tested by injection of killed cultures (auto- rial counts, we used RM-agar (Martin et al. 1998), claved at 121ЊC for 20 min). One microliter of undi- containing half the nutrients of L-agar. Because viable luted culture was injected ventrally into the abdomen bacterial counts were not correlated with toxicity, we of 10 males and 10 females. Controls of 10 males and used optical density as a measure of cell density. The 10 females also were injected with sterile distilled correlation between optical density and toxicity for water. Mortality was recorded daily for 5 d. Colorado potato beetle was 0.77 (P.A.W.M., unpub- Corn root worm assays were performed with adult lished data). beetles of either southern corn rootworm or western Preliminary Toxin Characterization. Bacteria were corn rootworm of mixed sexes as described by Schro- grown on L-agar plates at 25ЊC for 48 h to test whether der et al. (2001). C. subtsugae was harvested at 48 h bacteria could be recovered from dead insects, thus into 20 ml of watermelon juice containing 0.08% cu- fulÞlling KochÕs postulates to identify a pathogen; or curbitacin-E glycoside (Matsuo et al. 1999) mixed for 120 h for increased stability upon storage. Bacteria with 3% Mirasperse starch (A.E. Staley Manufacturing were harvested into 15 ml of sterile distilled water. Co., Decatur, IL). Five 25-␮l droplets were added to Ten-milliliter volumes were autoclaved at 121ЊC for each 100-mm petri dish. Five beetles were added to 10 min to test for stability. Cells were removed from each dish, and controls (without the bacteria) and suspension by centrifugation and Þltered through a treatments were repeated Þve times. At 48 h, addi- 0.45-␮m Millipore Þlter (Millipore Corporation, Bil- tional moisture was introduced by a premoistened lerica, MA) to yield a cell-free supernatant. The pellet dental wick and additional food was added (either fraction was formed by resuspending the pellet to the western corn rootworm diet or squash). Mortality was original volume. recorded 5 d after treatment was initiated. Insects. The southern green stink bug colony orig- For corn rootworm larval assays, eggs were hatched inated from specimens collected near Stoneville, MS, and larvae were fed on corn roots germinated in sand and insects were reared on sunßower, Helianthus an- at room temperature. Roots were kept moist by daily nuus L., seeds and green beans, Phaseolus vulgaris L., watering. Larvae (16) were removed after 3 wk days at 28ЊC, 65% RH, and a photoperiod of 16:8 (L:D) h and transferred to freeze-dried diet pellets (Martin (Aldrich et al. 1993). 2004a) made with corn rootworm diet (BioServ, Initially, southern corn rootworm adults, Diabrotica Frenchtown, NJ), which had been rehydrated with undecimpunctata howardi Barber (Coleoptera: Chry- water or C. subtsugae suspension. The larvae in sealed somelidae), were obtained from French Agricultural trays were incubated at 25ЊC and 50% RH in the dark. Research, Inc. (Lamberton, MN). Adults were held Mortality was recorded every 24 h. Surviving larvae for 3 wk before testing at 24ЊC, 30% RH, and a pho- were weighed after 4 d. Differences in weights were toperiod of 16:8 (L:D) h. Beetles were fed artiÞcial compared using PROC MIXED (SAS Institute 2004). diet (Branson and Jackson 1988). Later, a southern Recovery of C. subtsugae from Dead Insects. Be- corn rootworm colony was established from Þeld- cause viability of C. subtsugae declines after 72 h of collected insects from Beltsville, MD, in 2003. The growth on media, recovery from insects was done colony has been maintained on corn roots for larvae, using cultures that were grown on solid media for 48 h. and on squash and water for the adults. Field-collected Insects that had died after ingesting C. subtsugae cells insects are introduced yearly to maintain genetic di- were ground in 0.5 ml of sterile water by using a 682 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 100, no. 3

Fig. 1. Survival times of 48- and 120-h cultures of C. subtsugae against the southern green stink bug. Dark bars, Fig. 2. Survival times for whole cultures and separated males; light bars, females. Error bars represent 95% CL. Con- cultures of C. subtsugae against the southern green stink bug. trol mortality for females was 0% (0/20) and for males was Dark bars, males; light bars, females. Error bars represent 95% 15% (3/20). CL. Control mortality for females was 5% (1/20) and for males was 10% (2/20). miniature mortar and pestle. Large particles were al- lowed to settle and liquid was plated directly on L-agar and incubated for 48 h at 25ЊC and scored for the violet Upon injection, only the males that received the C. colonies typical of the bacteria. subtsugae injection died. All mortality (70%) occurred within 2 d. No females receiving C. subtsugae injec- Results tions and no controls died. Corn Rootworm. Approximately 80% of both south- Southern Green Stink Bug. Male and female southern and western corn rootworm adults died when fed ern green stink bugs are known to differ in their C. subtsugae in a bait formulation (Fig. 4). The beetles response to insecticides (Panizzi and Hirose 1995), so consumed the entire control bait droplets, whereas they were tested against C. subtsugae separately. For the violet droplets containing C. subtsugae were still treatment with undiluted cultures, mortality for both visible at the end of the assay. In additional assays, sexes usually reached 100% by 6 d, and the main 100% of southern corn rootworm adults died within difference between males and females was how 120 h, whereas control mortality was 13.3%. quickly they died. In most assays, female stink bugs did For larvae the mortality reached only 40% when fed not begin to die until day 4, whereas mortality often ϭ whole cultures at4d(OD600 2.31). However, there occurred in the treated male stink bugs by the second was signiÞcant a signiÞcant difference (df ϭ 24, P ϭ day. This difference is best expressed by differences in 0.0057) in weight between surviving controls (15.02 Ϯ survival times with cultures that were 48 h old. 1.24 mg) and those treated with C. subtsugae (9.81 Ϯ For undiluted cultures at the same optical density 0.86 mg). The control larvae burrowed into the diet ϭ (OD600 2.3), the males died 29% faster than the pellets and produced frass. For the C. subtsugae- females for cultures that were harvested at 48 h but treated larvae, there was no evidence of feeding and only 15% faster for cultures harvested at 5 d (Fig. 1). no frass. Final mortality for females was 90% and for the males Recovery of C. subtsugae from Dead Insects. Eleven it was 95%. However, the males died quicker only southern green stink bug adults that died were ground when the cell-associated fraction was used. When the after being fed C. subtsugae. Although living bacteria cell-free supernatant was used for treatment, the sur- were recovered from all of them, with titers ranging vival time was essentially identical for both males and from 103 to 105 cells per insect, no violet colonies were females, although there was more variability for females (Fig. 2). In these assays, all the males were dead at 6 d after treatment with the whole culture, the supernatant or the resuspended pellet. However, the female mortality was 100% only in the whole culture and the supernatant treatments. The Þnal mortality for the females fed the resuspended pellet was only 70%. To test for stability, the autoclaved whole cultures were tested. There were little differences in the survival times between the whole cultures and the autoclaved suspension for either males or females (Fig. 3). In treatments with the autoclaved suspension, the difference in survival times between males and females also was absent, as determined by overlapping Fig. 3. Survival times for autoclaved versus nonauto- 95% conÞdence intervals. Final mortality for both claved cultures of C. subtsugae against the southern green males and females in the autoclaved cultures was 95%. stink bug. Dark bars, males; light bars, females. Error bars A 1:3 dilution of supernatants killed 50% of the fe- represent 95% CL. Control mortality for females was 15% males, and a 1:6 dilution killed 50% of the males. (3/20) and for males 20% (4/20). June 2007 MARTIN ET AL.: TOXICITY OF C. subtsugae TO PEST INSECTS 683

low (Ͻ40%), but feeding was inhibited as evidenced by lack of weight gain. What remains is to determine whether attractant or bait combinations can be effective under in the Þeld and whether the nymphs of N. viridula will feed on and be killed by C. subtsugae. It is unusual that an orally active compound produced by bacteria is Þrst discovered as lethal to adults. The ␦-endotoxin of B. thuringiensis kills the larval stage of most insects (Schnepf et al. 1998). Even the toxin of C. subtsugae did not kill adult Colorado potato beetles, but it killed in all instars Fig. 4. Percentage of mortality of southern (dark bars) of feeding larvae (Martin et al. 2004). and western (light bars) corn rootworm adults when fed whole cultures of C. subtsugae. Error bars represent standard error. Control mortality for western corn rootworm is shown Acknowledgments by thick line, and control mortality for southern corn rootworm is shown by thin line. We thank the following Eleanor Roosevelt High School students; S. Stone for isolation of PRAA from soil, and H. Romani for some bioassays; and USDA employees L. Liska recovered on RM agar. Five southern corn rootworms for rearing insects, and A. Shropshire for coordinating bio- were ground, and no violet colonies were recovered assays and general technical assistance. on RM agar. References Cited Discussion Aldrich, J. E., H. Numata, M. Borges, F. Bin, G. K. Waite, and The deaths of both southern green stink bug and W. R. Lusby. 1993. Artifacts and pheromone blends corn rootworm adults after consuming C. subtsugae in from Nezara spp. and other stink bugs (Heteroptera: water or in a bait formulation were similar to results Pentatomidae) A. Naturforsh. C. 48: 73Ð79. with Colorado potato beetle (Martin et al. 2004). Re- Atlas, R. M. 1997. Handbook of microbiological media. duction in weight or feeding inhibition with corn CRC, Boca Raton, FL. rootworm larvae was similar to that seen with Lyman- Baum, J. A., C.-R. Chu, M. Rupar, G. R. Brown, W. P. Don- ovan, J. E. Huesing, O. Ilagan, T. M. Malvar, M. Pleau, M. tria dispar (L.) (Martin 2004b). Although whole bac- Walters, and T. Vaughn. 2004. Binary toxins from Bacil- terial cultures were used in most experiments, live lus thuringiensis are active against the western corn root- bacteria do not seem to be necessary for lethality. worm, Diabrotica virgifera virgifera LeConte. Appl. En- Several lines of evidence support this hypothesis. viron. Microbiol. 70: 4889Ð4898. First, violet pigment-producing bacteria could not be Black, J. G. 1996. Microbiology: principles and applications, isolated from either stink bugs or corn rootworms that 3rd ed. Prentice Hall, Upper Saddle River, NJ. had died after feeding on C. subtsugae cells. Thus, the Branson, T. F., and J. J. Jackson. 1988. An improved diet for Þrst step in identifying a pathogen could not be ful- adult Diabrotica virgifera virgifera (Coleoptera: Chry- Þlled. Second, for stink bugs, cell-free Þltrates alone somelidae). J. Kans. Entomol. Soc. 61: 353Ð355. Brazilian National Genome Project Consortium. 2003. The caused mortality. Third, cells killed by autoclaving also complete sequence of Chromobacterium violaceum re- killed stink bugs in a manner similar to viable whole veals remarkable and exploitable bacterial adaptability. cultures. Fourth, cells killed by autoclaving killed 70% Proc. Natl. Acad. Sci. U.S.A. 100: 11660Ð11665. of male stink bugs in 48 h by injection. The injection Bucher, G. E. 1981. IdentiÞcation of bacteria found in in- data showed the most obvious difference in suscep- sects, pp. 7Ð33. In H. D. Burges [ed.], Microbial control tibility between males and females (Panizzi and of pests and plant diseases 1970Ð1980. Academic, New Hirose 1995). This suggests a heat-stable toxin(s) is York. responsible for the mortality. Heat-stable oral in- Duran, N., R. V. Antonio, M. Haun, and R. A. Pilli. 1994. secticidal toxins, such as the ␤-exotoxin, are known to Biosynthesis of a trypanocide by Chromobacterium vio- ˇ laceum. World J. Microbiol. Biotechnol. 10: 685Ð690. be produced by strains of B. thuringiensis (Sebesta et Duran, N., S. Erazo, and V. Campos. 1983. Bacterial chem- al. 1981). Attempts to localize the toxicity in the sol- istry II. Antimicrobial photoproduct from pigment of uble (supernatant) or particulate (pellet) fractions Chromobacterium violaceum. An. Acad. Bras. Cie´n. 55: were unsuccessful, because signiÞcant toxicity was 231Ð234. noted with both fractions. Grimont, A. D., and F. Grimont. 1978. The genus Serratia. Although undiluted bacterial cultures were used to Annu. Rev. Microbiol. 32: 221Ð248. kill these insects, bait formulations with attractants for Krueger, S. R., and D. W. Roberts. 1997. Soil treatment with N. viridula (Aldrich et al. 1993) or feeding stimulants entomopathic fungi for corn rootworm (Diabrotica spp.) for diabroticite beetles (Schroder et al. 2001) could larval control. Biol. Control Theor. Appl. Pest Manag. 9: 67Ð74. increase the effective concentration of the toxins in Martin, P.A.W., S. Mischke, and R.F.W. Schroder. 1998. the Þeld. For corn rootworm adults, the compulsive Compatibility of photoactive dyes with insect biocontrol feeding on formulations containing cucurbitacin en- agents. Biocontrol Sci. Technol. 8: 501Ð508. sures the ingestion of a lethal dose (Schroder et al. Martin, P.A.W. 2004a. A freeze-dried diet to test pathogens 1998). Southern corn rootworm larval mortality was of Colorado potato beetle. Biol. Control 29: 109Ð114. 684 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 100, no. 3

Martin, P.A.W. 2004b. A stilbene optical brightener can en- Schroder, R.F.W., A. B. DeMilo, C.-J. Lee, and P.A.W. Martin. hance bacterial pathogenicity to gypsy moth (Lepidop- 1998. Evaluation of a water-soluble bait for corn rootworm tera: Lymantriidae) and Colorado potato beetle (Co- (Coleoptera: Chrysomelidae) control. J. Entomol. Sci. 33: leoptera: Chrysomelidae). Biocontrol Sci. Technol. 14: 355Ð364. 275Ð383. Schroder, R.F.W., P.A.W. Martin, and M. M. Athanas. 2001. Martin, P.A.W., M. Blackburn, and A. S. Shropshire. 2004. Effect of a phloxine B-cucurbitacin bait on diabroticite Two new bacterial pathogens of Colorado potato beetle beetles (Coleoptera: Chrysomelidae). J. Econ. Entomol. (Coleoptera: Chrysomelidae). J. Econ. Entomol. 97: 774Ð 94: 892Ð897. 780. Sˇebesta, K. J. Farkasˇ, K. Horska´, and J. Van˜ ova´. 1981. Thu- Matsuo, K., A. B. DeMilo, R.F.W. Schroder, and P.A.W. ringiensis, the beta exotoxin of Bacillus thuringiensis. Martin. 1999. Rapid high performance liquid chroma- pp. 249Ð282. In H. D. Burges [ed.], Microbial control of tography method to quantitate elaterinide in juice and pests and plant diseases 1970Ð1980. Academic, London, reconstituted residues from a bitter mutant of Hawkes- United Kingdom. bury watermelon. J. Agric. Food Chem. 47: 2755Ð2759. Sneath, P.H.A. 1984. Genus Chromobacterium Bergonzini, McKinlay, R. G., A. M Spaull, and R. W. Staub. 1992. Pest of pp. 580Ð582. In N. R. Krieg and J. G. Holt [eds.], BergeyÕs Solaneceous crops, pp. 263Ð326. In R. G. McKinlay [ed.], manual of systematic bacteriology, vol. 1. Williams & Vegetable crop pests. CRC, Boca Raton, FL. Wilkins, Baltimore, MD. Panizzi, A. R., and E. Hirose. 1995. Survival, reproduction, Sosa-Gomez, D. R., and F. Moscardi. 1998. Laboratory and starvation resistance of adults southern green stink and Þeld studies on the infection of stink bugs, Nezara bug (Heteroptera: Pentatomidae) reared on sesame or viridula, Piezodorus guildinii, and Euschistus heros soybean. Ann. Entomol. Soc. Am. 88: 661Ð665. (Hemiptera: Pentatomidae) with Metarhizium anisop- Pereira, R. M., and D. W. Roberts. 1991. Alginate and con- liae and Beauveria bassiana in Brazil. J. Invertebr. starch mycelial formulations of entomopathogenic fungi, Pathol. 71: 115Ð120. Beauveria bassiana and Metarhizium anisopliae. J. Econ. Vazquez-Arista, M., R. H. Smith, V. Olalde-Portugal, R. E. Entomol. 84: 1657Ð1661. Hinojosa, R. Hernandez-Delgadillo, and A. Blanco-Labra. Preisler, H. K., and J. L. Robertson. 1989. Analysis of time- 1997. Cellulolytic bacteria in the digestive system of Pro- dose mortality data. J. Econ. Entomol. 82: 1534Ð1542. stephanus truncatus (Coleoptera: Bostrichidae). J. Econ. SAS Institute. 2004. SAS OnlineDoc7, version 9.1. SAS In- Entomol. 90: 1371Ð1376. stitute, Cary, NC. York, A. 1992. Pests of cucurbit crops: marrow, pumpkins, Schnepf, E. N. Crickmore, J. Van Rie, D. Lereclus, J. Baum, squash melon, and cucumber, pp. 144Ð147. In R. G. Mc- J. Feitelson, D. R. Zeigler and D. H. Dean. 1998. Bacillus Kinlay [ed.], Vegetable crop pests. CRC, Boca Raton, FL. thuringiensis and its pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62: 775Ð806. Received 30 December 2004; accepted 19 January 2007.