ENTOMOLOGY L.A. Lacey et al. (2007) Phytoparasitica 35(5):479-489 Gut Bacteria Associated with the Pacific Coast Wireworm, Limonius canus, Inferred from 16s rDNA Sequences and Their Implications for Control L.A. Lacey, * T.R. Unruh, H. Simkins and K. Thomsen-Archer A multitude of bacteria have been isolated from the guts of several insect species. Some of these have been modified to interfere with the development of the host insect or with the development and transmission of plant and animal pathogens transmitted by the host insect. We surveyed the gut flora of the Pacific Coast wireworm. Limonius canus LeConte. a serious pest of potato, at two sites in Oregon and Washington. Isolates were obtained from surface-sterilized triturated larvae by dilution plating on standard media. A rich diversity of species was found in 86 isolates, including spore-formers, non-spore-formers and aerobic and facultatively anaerobic species collected on four sampling dates at each location. Twenty-one of the isolates were identified to species based on rDNA sequence (nine distinct species). An additional 34 isolates were identified to genus from the sequence data while six isolates could be assigned only to family based on sequence comparisons. Twenty-seven additional isolates were identified to species (9). genus (17) or family (1) based on side-by-side morphological comparisons with isolates identified from rDNA sequence. The most frequently isolated bacterium was Bacillus inegareriuni, followed by Rahnella aquarius. A naturally occurring bacterium found in the gut and/or environment of a targeted insect that is modified to express toxins or other detrimental substances could provide certain advantages (such as persistence and recycling) over inundatively applied microbial control agents, particularly within soil habitats. The hypothesis that these species or others from the survey represent candidates for genetic modification to provide control options for L. canus is discussed. KEY WORDS: Bacillus megateriuln; midgut flora: Rahnella aquatilis: spore formers: non- spore formers. INTRODUCTION A wide variety of bacteria are associated with insects, ranging from obligate endosym- biotes and pathogens. to facultative pathogens and others (9.24,3 1). These comprise a broad spectrum of anaerobic and aerobic spore forming and non-spore forming species including short- and long-term residents of the gut. The gut microbial flora of insects has been surveyed in several taxa of insects (7,9.25). Some bacteria isolated from pest species have been considered as candidates for genetic manipulation to create paratransgenic biological control agents (20,25.28). The Pacific coast wireworm. Li,nonius canus LeConte, is found in irrigated habitats in western North America, where it is a pest of grain and vegetable crops, especially potato tubers in the Pacific Northwest of the USA and elsewhere. In addition to tuber Received May 30. 2007: accepted July 29, 2007: http://www.phyt0parasit1Ca.org posting Sept. 10. 2007. Yakima Agricultural Research Laboratory, USDA-ARS, Wapato, WA 98951. USA. Corresponding author [e- mail: [email protected]]. 479 Phvtoparasitica 35:5, 2007 damage through tunneling and feeding, the presence of wireworm larvae in processed potatoes is highly undesirable. Control of wireworms traditionally relies on broad spectrum insecticides despite the low efficacy of this approach (18.26). Reduction of the use of broad spectrum insecticides will require development of alternative control strategies, such as microbial control agents, that are as efficacious and affordable as chemical pesticides. Microbial based insecticides have shown promise for some insect pests of potato (35) but their efficacy against wireworms has been limited. Engineering bacteria that are normally found in the midgut to become pathogenic for wireworm larvae may provide an opportunity for microbial control. With this ultimate objective, a survey of bacteria in L. canus was undertaken in the eastern part of Washington State and Oregon. MATERIALS AND METHODS Collection of wireworms Surveys were made in the summer of 2004 at two locations, the Hermiston Agricultural Research and Extension Center in Hermiston, Oregon (elevation 183 m) and the USDA Agricultural Experiment Research Farm near Moxee, Washington (elevation 472 m). Soil at the Hermiston site is an Adkins fine, sandy loam, pH 6.7-7.1. with organic matter 0.7%-1%. Soil at the Moxec site is Cleman very fine sandy loam, pH 7.4, with organic matter <1%. In late spring and early summer (May 5—June 28), wireworms were collected on four dates from the Hermiston site. Later in the summer (July 19—Aug. 30), as wireworms became difficult to collect at the Hermiston site, L. canus larvae were collected on four dates from the Moxee site. Collections at the Moxee site were terminated after August 30, when wireworms became difficult to collect. Bait balls composed of one-third dry oatmeal and two-thirds soil held together with organdy netting were used as lures, as described previously by Horton and Landolt (14). Twenty bait balls were buried at a depth of 10-15 cm with a separation distance of 1-1.5 m for each collection date. Colored string tied to the baits facilitated retrieval. The baits were collected 3 to 4 d after placement and brought back to the laboratory for recovery of wireworm larvae. Isolation of enteric bacteria Whole larvae were used for isolating gut flora because of the difficulty of removing intact guts. Bacteria isolated from whole larvae were considered enteric rather than hemocoelic in origin for two reasons. Internal bacterial symbiotes, if any, would have required more complex media and handling to isolate. The non-spore- forming pathogens of insects have low pathogenicity, but once they gain entry into the hemocoel, multiply rapidly and cause death of the insect by septicemia (24,31). Only living active wireworms were used for isolations. Furthermore, no spore-forming pathogens are known from wireworms. The larvae were first surface-sterilized as described by Lacey and Brooks (21) by submersing them in 70% ethanol followed by dilute NaCIO (15% Clorox) and rinsing with sterile distilled water. Groups often randomly selected late instar larvae were then ground in 10 ml of sterile water using a tissue grinder (Pyrex model 7725- 19: Fisher Scientific. New Lawn, NJ, USA). A drop of Tween 80 (Fisher Scientific) was added to facilitate homogenization of the suspension. The stock suspension and 10and 10-2 dilutions of the suspension were plated onto Brain Heart Infusion A gar, Tryptic Soy Agar (Soybean-Casein Digest Agar) and Nutrient Agar (Hardy Diagnostics. Santa Maria, CA, USA). When available (on 4.v.04, l.vi.04, 21.vi.04, 19.vii.04728.vii.04, 18.viii.04), a second lot of ten larvae on each collection date was prepared in the same manner and subsequently pasteurized in sterile 15 ml screw top test tubes in an 80°C water bath for 12 min and then cooled rapidly in crushed ice. The pasteurized stock suspension was plated 480 L.A. Lacey et al. onto the three media as above. This procedure allowed selective isolation of spore-forming species. On the dates when fewer than 20 larvae were available, the 10 ml suspension with ten or fewer larvae was divided into two batches of 5 ml; then one of them was diluted and plated and the other was pasteurized and plated. When numbers of larvae retrieved from the field were fewer than ten (28.vi.04. 30.viii.04), they were all used for the isolation procedure without pasteurization. The plates were incubated at 25°C for 24- 72 h. depending on growth. On selected dates (19.vii. 28.vii. 18.viii. 30.viii), the stock solution was also plated on the three media under anaerobic conditions usin g gas packs (AnaeroGen Compact sachets: Fisher Scientific). Clean plates of the same medium upon which initial isolations were made were streaked with each resulting colony that exhibited a distinct morphology. Bacteria from each colony were observed microscopically before and after Gram staining. Representative distinct colonies for each date were then chosen for identification based on 16S rDNA sequence: additional colonies of like type were identified by morphological traits alone based on side-by-side comparisons with isolates identified by sequencing and the use of descriptions in Bergeys Manual (13). Characteristics used were colony color and appearance, cell morphologies, presence or absence of spores. and Gram staining. Each bacterial isolate was grown in nutrient broth (Dilco Amplification of 16S rDNA Laboratories, Detroit, Ml. USA) and the genomic DNA was extracted using a QIAGEN DNeasy Tissue Kit (QIAGEN Sciences, Germantown. MD, USA). The 16S rDNA of each isolate was amplified by PCR using universal eubacterial primers (Sigma Genosys. The ccgaattcgtcgacaacagagtttgatcCtggct.-3 and reverse Woodlands. TX, USA; forward 5- cttgttacgactt 3) which correspond to fD- 1 and rP-2 reported 5 cccgggatccaagcttacggctaC by Weisburg et al. (34) as suitable for most Eubacteria. The PCR was carried out in an MJ- 100 thermal cycler (Mi Research. Watertown, MA, USA) as follows: one cycle at 95°C for 2 mm, 30 cycles at 94°C for 30 s. 45°C for 30 s and 72°C for 2 mm, followed by 45°C for 1 min and 72°C for 2 mm. The PCR products were then purified with QlAquick PCR Purification Kits (QIAGEN Sciences) in preparation for cycle sequencing. Sequencing and identification Seven of the PCR products were sequenced commer- cially (Macrogen Inc., Seoul, Korea) and the remaining products were sequenced in house. Sequencing reactions consisted of 1 or 2 tl of the ABI, Big-dye v3.1 sequencing kit (Applied Biosystems. Foster City, CA, USA), with 3 pl of Big Dye sequencing buffer. I fil of template (50 ng PCR product) and I il (10 no) of primer with water to make 10 Pl.