Adapted Western Corn Rootworm (Diabrotica Virgifera Virgifera L.)

ARTICLE IN PRESS

Insect Biochemistry and Molecular Biology 38 (2008) 697– 704

Contents lists available at ScienceDirect

Insect Biochemistry and Molecular Biology

journal homepage: www.elsevier.com/locate/ibmb

Increased expression of a cGMP-dependent protein kinase in rotation- adapted western corn rootworm (Diabrotica virgifera virgifera L.)

Freydoun Garabagi a, B. Wade French c, Arthur W. Schaafsma b, K. Peter Pauls b,Ã a Department of Enviromental Biology, University of Guelph, Guelph, Ont., Canada N1G 2W1 b Department of Plant Agriculture, University of Guelph, Guelph, Ont., Canada N1G 2W1 c North Central Agricultural Research Laboratory, USDA, 2923 Medary Avenue, Brookings, SD 57006, USA article info abstract

Article history: A new ‘variant’ behavior in western corn rootworm (WCR) has resulted in egg-laying into non- Received 21 June 2007 cornfields, compared to ‘normal’ deposition of eggs in cornfields, allowing these insects to circumvent Received in revised form crop rotation. No morphological or genetic characteristics have been defined to differentiate between 3 March 2008 the normal and variant biotypes. Cyclic GMP-dependent protein kinases (PKG) have been implicated in Accepted 29 March 2008 the regulation of behaviors in vertebrates, insects, and nematodes, including foraging behavior in Drosophila. A cDNA with homology to the Drosophila melanogaster foraging gene (called Dvfor1)was Keywords: cloned from WCR. The deduced DvFOR1 protein is approximately 70% similar to FOR proteins in Real-time PCR Drosophila, silkworm (Bombyx mori) and honeybee (Apis mellifera). It contains a coiled-coil region, two PKG tandem cyclic nucleotide-binding domains, a serine/threonine kinase catalytic domain, and a serine/ Foraging gene Ovipositioning threonine kinase catalytic domain extension, which are all characteristically found in PKG proteins. Diabrotica virgifera Real-time PCR assays of foraging transcript levels in heads of normal and rotation adapted females of Variant WCR obtained from lab-reared insect colonies indicated that the variants had higher levels (25%) of PKG expression than normals. The magnitude of this increase is similar to that observed in Drosophila rover phenotypes compared to sitter phenotypes. However, Diabrotica contains at least two different foraging gene transcripts, which complicates establishing a direct link between the level of gene expression and insect behavior. & 2008 Elsevier Ltd. All rights reserved.

1. Introduction rotation was losing its effectiveness in certain counties of eastern Illinois and there were increasing reports of major WCR damage to Western corn rootworm (WCR), Diabrotica virgifera virgifera first-year corn (Levine and Oloumi-Sadeghi, 1996). This unex- LeConte, has been a major pest of corn for over 50 years. The pected damage occurred because of the emergence of a new annual pest management (mainly soil insecticides) and yield loss behavior in WCR that was prompting the insects to lay eggs costs for WCR have been estimated to approximate 1 billion outside of cornfields (O’Neal et al., 1999). Because the non-corn dollars annually in the United States (Sappington et al., 2006), and fields were rotated back to corn in the following year, the over 260 million dollars in Canada during the mid 1980s (Madder abnormal egg-laying behavior provided the emerging larvae, the et al., 1988). Corn rootworms have been controlled effectively corn roots they require in the following year. The conditions that through crop rotation since the early 20th century, reducing their might have selected for this altered behavior are not certain but economic importance significantly (Schaafsma et al., 1999). might include high population densities (Onstad et al., 1999), WCR pass through only one generation per year and usually shifts in corn phenology (Darnell et al., 2000), and restricted oviposit in cornfields in late summer and fall. The larvae that cultural practices such as a high percentage of fields in corn/ emerge the following spring are obligate corn root feeders, soybean rotation (Onstad et al., 2001). therefore, planting a non-host crop such as soybeans after corn There are no phenotypic or morphological characteristics that is an effective management tool (Krysan and Branson, 1983). Crop can be used to distinguish rotation-adapted WCRs from their rotation strategies are based on the assumption that WCRs normal counterparts. The lack of a simple and quick diagnostic oviposit solely in cornfields. However, in the early 1990s, crop tool for identifying the variant insects makes it difficult to monitor their field occurrence in areas that are known to contain the variant insect, as well as in areas where the insect is expected to Ã Corresponding author. Tel.: +1519 824 4120x52460; fax: +1519 763 8933. spread. Accurate identification of a population of variant WCR in E-mail address: [email protected] (K. Peter Pauls). cornfields, through a genetic-based system, would allow corn

698 F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704 producers to implement alternative strategies to rotation to 3 min and cooled to room temperature. The reaction was control the pest. In a series of behavioral assays, Knolhoff et al. incubated at 37 1C for 30 min with 200 units SuperscriptII Reverse (2006) showed that D. v. virgifera females from lab-reared insects Transcriptase, 100 mM DTT, 2 mM dNTPs, 4 units RNaseOUT, in a that were originally collected from regions where crop rotation is 1 RT buffer (50 mM Tris–HCl, pH 8.3, 75 mM KCl, 3 mM MgCl2; no longer effective are more active than females collected from Invitrogen, Carlsbad, CA), followed by 90 1C for 10 min for enzyme regions where rotation remains effective. Based on differences inactivation. observed between normal and rotation-resistant populations of A clustalW multi-nucleotide alignment of published insect PKG beetles in similar environments, the authors suggest a genetic genes identified a highly conserved serine/threonine domain in these component as the underlying cause of this change in behavior but sequences (http://clustalw.genome.jp/; data not shown). Primers, no specific gene or genes were identified. designed on the basis of the Drosophila serine/threonine kinase Cyclic GMP-dependent protein kinases (PKG) have been domain (Dm-FOR-F 50-CGCCAGTGCTTTCAGACCATCATGATG-30,Dm- implicated in the regulation of behavioral processes in the FOR-R 50-GCGTCTCCACAATCTGTGACTTTTTCATCTG-30)wereusedto vertebrates, insects, and nematodes (Schafer, 2002). One exten- amplify a 650 bp fragment of the foraging gene from WCR cDNA. The sively studied example is the Drosophila PKG that was first presence of multiple stretches of sequence containing greater than identified as the foraging gene. The foraging gene has two naturally three guanine or cytosine nucleotides (particularly in the 50 region of occurring alleles that manifest in two food search behavioral the gene) made it difficult to amplify fragments larger than 700 bp. As phenotypes, named rover and sitter (Pereira and Sokolowski, a result, the complete foraging gene cDNA was compiled from four 1993; Sokolowski et al., 1997; Osborne et al., 1997). When placed overlapping pieces. A second internal fragment was cloned in the 50 into a patch of food, both larvae and adult rovers have long direction, utilizing a B. mori-based forward primer and a reverse foraging paths and have a greater tendency to leave the patch primer based on the newly obtained WCR sequence. The BD SMART compared to sitters, which have short foraging paths and less RACE cDNA Amplification Kit (BD Biosciences, Palo Alto, CA) was tendency to leave the food patch. The foraging gene is alternately utilized, as described by the manufacturer, to obtain the full-length 50 spliced to create three major transcripts with varying lengths, and 30 ends of the transcript. The primary fragments amplified by the namely T1, T2, and T3, as well as several minor transcripts kit were very faint and needed to be subjected to an extra round of (Kalderon and Rubin, 1989). Small differences in the abundance of re-amplification after gel purification. The 50 RACE reaction generated the T1 message between rovers and sitters account for the a truncated fragment, which required a second round of 50 RACE for observed behavioral variation (Osborne et al., 1997). The sitters acquiring the full-length sequence. were shown to have slightly (11%) lower levels of foraging transcript as well as PKG protein levels and enzyme activity. Rover insects could be converted by mutation to sitters by down regulating the foraging gene, or by treating them with inhibitors 2.2. Southern hybridization of PKG. The involvement of PKG in food search behavior has been Genomic DNA was extracted and pooled from thoraxes of demonstrated in several insects, including honeybee (Apis melli- normal and variant WCR female and male adults to obtain a total fera L., Ben-Shahar et al., 2002), red harvester ant (Pogonomyrmex of 30 mg. Genomic DNA was digested overnight with 100 units of barbatus Smith, Ingram et al., 2005), silkworm (Bombyx mori L., EcoRI (Invitrogen) at 37 1C, separated on an 1% agarose gel, Tanoue and Nishioka, 2003) as well as in the nematode transferred to a nylon membrane, and probed at high stringency Caenorhabditis elegans, Maupas, Daniels et al., 2000). In honeybee, with the DIG non-radioactive detection system (Roche Diagnostics up regulation of Amfor expression is concurrent with a change in GmbH, Penzberg, Germany) Two DIG probes were used in this the behavior from an individual that does not leave the hive to one study. The first probe was a 250 bp PCR amplified fragment of the that moves away from the hive in search of food (Ben-Shahar distal region of the serine/threonine kinase domain generated et al., 2002). Unlike Drosophila where different expression levels of with primers Probe-F 50-GCGACAACAAAATTCTGCACTGCTTGTG-30 the foraging gene are associated with different alleles in different and Probe-R 50-CGAACCAGTAAAGGGAGGTGTGCCTGTG-30.Thesec- insects, up regulation of the expression of the foraging gene in ond probe was a 200 bp fragment for the proximal region of the honeybee occurs developmentally in the same insect. serine/threonine kinase domain (with primers 1.35ForRealF and In the current work, the role of PKG in defining a normal and 1.35ForRealR; see below for sequences). Blots were hybridized at variant food search behavior, which ultimately effects egg-laying 42 1C, with stringent wash of 0.5 SSC at 68 1C. site selection in Diabrotica, was tested by cloning the complementary DNA (cDNA) orthologs for the foraging gene and by measuring the levels of transcripts from the gene in the heads of variant and normal insects from laboratory reared colonies. 2.3. Western corn rootworm insect colonies

Assays were conducted on lab-reared normal and variant WCR 2. Materials and methods colonies, provided by the USDA-ARS North Central Agricultural Research Laboratory (NCARL), Brookings, SD. Normal and variant 2.1. cDNA cloning of a foraging gene ortholog from WCR insect colonies were reared separately from field-collected insects originally collected with emergence traps, in a maize field near Because no sequence information was available on a foraging Brookings, South Dakota during 1995, and soybean fields near gene (cyclic GMP-dependent protein kinase) ortholog from WCR, West Lafayette, Indiana during 1997, respectively. Two sets of the strategy that was used to clone this gene was to PCR amplify a insects were analyzed from the above-mentioned colonies. The conserved internal sequence and to expand the sequence in both first WCR set consisted of 48 mature insects of approximately 4 directions. weeks of age, in four groups of 12 normal and variant, males RNA was extracted from heads of WCR using an RNeasy kit and females. The second set consisted of a total of 72 normal (Qiagen, Valencia, CA). Complementary DNA was generated in a and variant WCRs in three developmental stages of 3rd instar reverse transcriptase (RT) reaction from total RNA as follows: 1 mg (18 unsexed larvae), 2-week-old (14 normal and 14 variant), and total RNA and 2 mg of random primers were heated to 85 1C for 4-week-old (14 normal and 14 variant) adult females. ARTICLE IN PRESS

F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704 699

2.4. Real-time PCR assays 40–103), two tandem cyclic nucleotide-binding domains, (between residues 165–278 and 283–407), a serine/threonine kinase Two real-time primer pairs were used in this study. The first catalytic domain in the C-terminus (between residues 421–680), amplified a 220 bp region that was common to all foraging transcripts, and a serine/threonine kinase catalytic domain extension (be- including the two expressed pseudo-foraging transcripts (ForRealF tween residues 681–732). These motifs are characteristic of PKG 50-TGTGATAGTGGATGATCCAGAGGG-30,ForRealR50-CGAAACCACC- family of proteins, as determined by NCBI’s CDART (http:// TACTCCGAGAGTTG-30). The second primer pair was designed to www.ncbi.nlm.nih.gov/BLAST/), and SMART-Simple Modular amplify a 200 bp region in the proximal region of the serine/threonine Architecture Research Tool (http://smart.embl-heidelberg.de/). kinase domain (1.35ForRealF 50-TAGGTGGTTTCGGCAGAGTAGAAC-30, A neighbor-joining tree (http://align.genome.jp/) showed that 1.35ForRealR50-CACAATTTGCTTCTCCCATGATCTCTT-30). Total RNA was the putative DvFOR1 protein is more similar to honeybee and extracted from individual WCR heads using the QIAGEN RNeasy Mini silkworm FOR proteins (73%) than to the Drosophila FOR A protein Kit (QIAGEN Inc., Valencia, CA) and quantitated with Ribogreen (69%) (data not shown). The foraging gene cDNA sequence was (Invitogen). RT reactions of 20 ml were set up as described above with identical between normal and variant insects except for two 100 ng total RNA to generate cDNA template. The reaction volume was nucleotides in the coiled-coil region that would result in a brought up to 200 ml (1:10 diluted), and frozen as 10 ml aliquots for glutamine to arginine (CAA to CGG) amino acid substitution in further analysis. cDNA aliquots were thawed out independently so the variants (Fig. 1). that samples were subjected to the same number of freeze–thaw cycles during the course of the assay. These cDNA aliquots, were used 3.2. Southern hybridization as template in for 25 ml real-time PCR with the iCycler (Bio-Rad, Hercules, CA). The PCR conditions were as follows: 10 ml(10ng)of To estimate the number of foraging gene copies in Diabrotica 1:10 RT dilution for template, 2 mM dNTPs, 10 mM primers, 3 mM genomic DNA samples (bulks from 10 insects each of normal and MgCl ,0.1 SYBR-Green I, and 0.65 units PlatinumTaq in 1 buffer 2 variant female and male insects) digested with EcoRI were probed containing 20 mM Tris–HCl, pH 8.4 and 50 mM KCl (Invitrogen). Two- with a fragment derived from the serine/threonine kinase domain step PCR was carried as follows: 1 cycle at 95 1C for 3 min, 50 cycles at of the gene that is common to all three transcripts cloned in the 95 1Cfor10s,and551Cfor45s. present study. All of the samples had a strongly hybridizing band The ITS2 region of Diabrotica (Internal Transcribed Spacer approximately 5 kb in size and a second, less intense band Sequence2; Genbank accession no. AF110464) was cloned (ITS2F approximately 3 kb in size (Fig. 2A). For the normal male sample, 50-AGCTAATTGCGCGTCAACTTG-30,ITS2R50-CACACGATAACATCCA- the second band was very weak. Hybridization of the blot with a ACACG-30) and used as a control gene for normalizing the assays. A second probe for the proximal stretch of the kinase domain, which 10-fold dilution series of the plasmids containing the Dvfor1 and ITS2 discerns between the authentic foraging message and Dvfor Ps1 fragments were used as template standards for the assay. All reactions (but not between Dvfor1 and Dvfor Ps2), showed a strong for the samples and standard points were carried out in triplicates. hybridization band at approximately 3.5 kb, and a second much Outliers with a C difference of 40.5 cycle within each triplicate were t less intense band of approximately 4.5 kb in size (Fig. 2B). removed from the analysis. The mean and standard error were calculated for the two populations with a 95% confidence level. The values obtained for the transcript levels were relative and could only 3.3. Real-time PCR assay of the WCR foraging genes be compared within each experimental run. Real-time PCR assays were carried out to determine the relative abundance of the foraging transcript in the brains of 3. Results normal and variant WCR. Our initial assay on the first set of insects (a total of 48 insects in four groups of 12 normal and 3.1. Cloning of the foraging gene ortholog variant, males and females) was carried out with primers targeted to a conserved region (data not shown) shared among all the An ortholog of the foraging gene, namely Dvfor1, was PCR foraging transcripts. This assay was performed in two rounds, amplified as four overlapping fragments from total RNA obtained testing six insects from each group at a time, since even though from heads of WCR adults (Fig. 1). The compiled message consists individual samples could be compared accurately within a real- of a 2196 bp ORF, 550 bp of 30 UTR sequence and 280 bp of 50 UTR time PCR set in a 96-well plate, the large number of samples sequence (accession DQ913742). The presence of a complete prevented direct comparisons among experimental replicates. message in heads of corn rootworm was confirmed by PCR Therefore, relative trends and not absolute values were compared. amplifying the complete 2196 bp ORF (data not shown). The A 10-fold dilution series of the plasmids containing the Dvfor1 and WCR cDNA sequence was 56% and 60% identical to the fruit fly ITS2 fragments were used as template standards for the assay. (D. melanogaster) and silkworm (B. mori) foraging genes, respec- Standard curves for both fragments were linear when plotted as tively. Analysis of the secondary structure of the Dvfor1 transcript log of starting quantity, with correlation coefficient values of revealed a hairpin in the distal region of the 30 UTR with a 65 base 0.994 and 0.903, slopes of 3.95 and 2.96, intercepts of 39.8 and pair stem-loop, as determined by Mfold (http://www.bioinfo.r- 43.4, and PCR efficiencies of 79.0% and 117.5% for ITS2 and Dvfor1, pi.edu/applications/mfold). In addition to the full-length Dvfor1 respectively. Melting curve analysis that was carried out following transcript, two processed transcribed pseudogenes, Dvfor Ps1 every real-time run showed one peak only, suggesting homo- (accession no. DQ913744) and Dvfor Ps2 (accession no. geneity of the amplification product. The results indicated a 19% DQ913743), were also cloned and sequenced. Sequence analysis increase in the relative abundance of the foraging message in revealed that Dvfor Ps1 was truncated at the serine/threonine variant females (Fig. 3A) compared with normal females. Normal kinase domain, fused to an unknown sequence, and contained and variant males had the same levels of transcript as the normal many stop codons. Dvfor Ps2 was truncated at the first cGMP females. domain, and fused to an unrelated sequence. The presence of several transcripts in Diabrotica that have Fig. 1 depicts the cDNA sequence of Dvfor1 along with its some homology to the foraging gene complicated measurements protein translation. The deduced DvFOR1 protein contains a of expression from this gene. Cloning and characterization of a coiled-coil region at the amino terminus (including residues larger portion of the foraging pseudo-message Dvfor Ps1 allowed ARTICLE IN PRESS

700 F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704

GTGTATTTCGGTAACTTTTACGGTATCTTCGGTATATTATATTCAAAATATTCAAATTTTGCATAGTTCCATTATTTTGTTGTGGATTCTGGTTTTGCATGC 102 TACAAAGGTTTAAATAGTTATTTTGCCGACGTGGTTGCTCTAAGGATTTGAAAAGTCGATAATAATTACTCTCGCTAATCTTCCTGTATCGAAGCGCCGACC 204 CAGGAGCACGTTTTTGGCAAGATTCCAGGAGTGGTTCGAGGGGGAAATCGCGAACCAAGTCACTTGAGGCCGATATATGCGCGTCTGTTTCGGTTCTCTCTG 306 M R V C F G S L C 9

CTTCTCGTCCCGGCTGAACAATTCCGCCGTTGACGAGGAAGCCGCGCATCACACGGTGACAAGTAACGGAGGCAGTCACGTGATTAGCAAGATGGCAACGAT 408 F S S R L N N S A V D E E A A H H T V T S N G G S H V I S K M A T I 43 Coiled Coil Domain TGGAGAATTACAGGCGCTACTTGTCCAGAAGGACGAAAGAATTGAAGAGTTGACACGTCAAATTCAGAGTTTTGATCGCGAAGTGATGCGGGAACAACGGCT 510 G E L Q A L L V Q K D E R I E E L T R Q I Q S F D R E V M R E Q R L 77 GG(R) GCAAGATCTGCAGCAGCAGCTGCAGATACGCGAAATCGAAATATCGGATTTACGGTCGCAGTTGGACAAATTTCAGAGTGTTATTCGTGTTCAAAATCCGAC 612 Q D L Q Q Q L Q I R E I E I S D L R S Q L D K F Q S V I R V Q N P T 111

GAGTCCTAAAGGTATGAGACCACGTAAACAAAGGGCGGGTATTTCGGCGGAACCGCAGAGTGAGGCGTCCATATTGGAGTTAAGCAAGCAGACCTTTACTGT 714 S P K G M R P R K Q R A G I S A E P Q S E A S I L E L S K Q T F T V 145

GCATCCTAAAGATGAAAGTTCTCGGGAACTTATCAAGTCGGCACTTTTGGATAATGATTTTATGAAGAATTTGGAGATAACTCAAATCAGAGAGATAGTGGA 816 H P K D E S S R E L I K S A L L D N D F M K N L E I T Q I R E I V D 179 cGMP-Binding Domain 1 CTGTATGTACCCAGAAGAATATAAGGCTAATGATATCATTATTCAGGAGGGAGACGTCGGTAGCATAGTGTATGTTTTAGAAGAGGGCAAAGTAGAAATTTC 918 C M Y P E E Y K A N D I I I Q E G D V G S I V Y V L E E G K V E I S 213

AAGAGAAAACAAAATATTGCATCATTTGGACCCTGGTAAAGTGCTTGGAGAATTGGCCATTTTGTACAATTGCCAGAGGACCGCTACTATTAAAGCTCATAC 1020 R E N K I L H H L D P G K V L G E L A I L Y N C Q R T A T I K A H T 247

GGACTGTAAATTATGGGCAATAGAACGACAATGCTTCCAGACCATTATGATGAGAACTGGTCTTATCCGACAAGCAGAATATACCGATTTCTTAAAGAGCGT 1122 D C K L W A I E R Q C F Q T I M M R T G L I R Q A E Y T D F L K S V 281

ACCAATCTTCAAAACACTTCCAGAAGACACTCTTATCAAGATTTCCGATGTTTTGGAAGAAACTATATACGCCAATGGAGACTATATTATCAGACAAGGAGC 1224 P I F K T L P E D T L I K I S D V L E E T I Y A N G D Y I I R Q G A 315 cGMP Binding Domain 2 CCGGGGTGATACGTTCTTCATTATAAGTAAAGGAAAAGTTAAGGTTACCAGAAAAATGCCAAATTCTAACACTGAGGAGTTTATAAGGACACTCGGTAAAGG 1326 R G D T F F I I S K G K V K V T R K M P N S N T E E F I R T L G K G 349

AGATTTCTTCGGGGAGAAGGCTTTGCAAGGGGATGATTTAAGAACTGCCAATGTGATAGTGGATGATCCAGAGGGCGTCACCACCCTGGTAATAGATAGAGA 1428 D F F G E K A L Q G D D L R T A N V I V D D P E G V T T L V I D R E 383

AACATTTAACCAACTGATATCAAATCTAGATGAGATTAGAACTAAATATAAGGATGAAAATATAGAAAGAAGAAGAGTAAACCAAGAATTCGAAGGCGTCAA 1530 T F N Q L I S N L D E I R T K Y K D E N I E R R R V N Q E F E G V K 417

ATTATCCGACTTAGTAATACTTACAACTCTCGGAGTAGGTGGTTTCGGCAGAGTAGAACTTGTTCAAATCAGAGGCCGCTCAAACAAATCCTATGCACTTAA 1632 L S D L V I L T T L G V G G F G R V E L V Q I R G R S N K S Y A L K 451 Serine/Threonine Kinase Domain ACAAATGAAGAAAGCACAAATCGTGGAAACTCGGCAACAGCAACATATCATGTCAGAGAAAGAGATCATGGGAGAAGCAAATTGTGATTTTATTGTTAAATT 1734 Q M K K A Q I V E T R Q Q Q H I M S E K E I M G E A N C D F I V K L 485

GCTTAGAACTTTTAAAGACGCTAAATATTTGTACATGTTAATGGAAAGCTGTTTGGGAGGAGAACTTTGGACTGTCCTAAGGGATAAAGGCCATTTTGATGA 1836 L R T F K D A K Y L Y M L M E S C L G G E L W T V L R D K G H F D D 519

CGCGACAACAAAATTCTGCACTGCTTGTGTCGTCGAAGCATTTGACTACTTACACTCCCGGAACATCATCTACAGGGATCTGAAACCTGAAAATCTTCTCCT 1938 A T T K F C T A C V V E A F D Y L H S R N I I Y R D L K P E N L L L 553

TGATAATAGCGGATATGTGAAATTAGTCGACTTTGGATTTGCGAAAAAATTGCAAACCGGGAGAAAAACTTGGACCTTCTGCGGAACCCCCGAATATGTAGC 2040 D N S G Y V K L V D F G F A K K L Q T G R K T W T F C G T P E Y V A 587

ACCCGAAGTTATCTTGAACAAAGGTCATGATATCAGCGCAGATTATTGGTCATTGGGAGTTTTAATGTTCGAATTACTCACAGGCACACCTCCCTTTACTGG 2142 P E V I L N K G H D I S A D Y W S L G V L M F E L L T G T P P F T G 621

TTCGGATCCAATGAGGACTTACAATATAATCTTAAAAGGAATAGATCAGATTGATTTCCCTAGAAGTATTACCAGAAATGCTCAGGCTCTGATTAAAAGGCT 2244 S D P M R T Y N I I L K G I D Q I D F P R S I T R N A Q A L I K R L 655

TTGTAGGGACAATCCTGCGGAACGTTTAGGATATCAAAAAGGAGGAATTAGTGACATCCAGAAACACAAATGGTTTGACGGATTCAATTGGGAGGGTCTCGT 2346 C R D N P A E R L G Y Q K G G I S D I Q K H K W F D G F N W E G L V 689 Serine/Threonine Kinase Extension TACCCGGACTCTTACACCGCCCATAATACCAACTGTGCAATGCGTAACGGATACGTCTAACTTTGATAACTATCCCCGTGATACGGACGACCCTCCACCGGA 2448 T R T L T P P I I P T V Q C V T D T S N F D N Y P R D T D D P P P D 723

TGATATATCTGGATGGGACAATAATTTCTAGATATTACTTTTTCAAATGAATGTAGGCTTTTTTAATTTTATAATTGAATGGGACATTTTTGGAAACAAGGC 2550 D I S G W D N N F 732

GTGTATTCAAAAAGCTTTATTAATAAAACTGCAAAGCTTAATGTTATAAGATTACAACTCCTTGCCCAAGGAAAATGTTTTGGATAATACCCAGTTTATTAC 2652 TTTTATAAGAACATGAATCTTGACGTTATGTCATATCTTAGGGAGTATTTATTGTCAGAACTTTTCTTTTTCCACATTATTTCACATAAAAAATAGACCGAG 2754 CTATATTGCTGTCGAAAACCTCTATTATCCAAATACGTGTCTTCTTAATAGAGCTTTTCATTCACCGTCATTTGTTTCGAGCTTCTGTCATATGTCGTATAA 2856 TCCGTGTATATTAATATTATACACACATTATACAGCATATGACAGAGGCTTGAAACAAATGACAATCGATGAAAAGCCCTATTTTTAATTATATTGTTTATG 2958 GATCGAATAAATTCGGGATGCCTACTGACTGTCAAATATACCAACAACTTTAAGATTTTATCTGTCTGTGCAAAAAAAAAAAAAAAAAAAAA

Fig. 1. Hypothetical translation of Dvfor1. Putative start and stop codons are underlined. The untranslated regions of the transcript are marked gray. The locations of characteristic PKG family proteins motifs are indicated, including: a coiled-coil region, two tandem cyclic nucleotide-binding domains, a serine/threonine kinase catalytic domain and a serine/threonine kinase catalytic domain extension. The sequence of the 30 stem-loop is marked white within the UTR. Boxed nucleotides indicate a two nucleotide difference between normal and variant sequence, leading to a glutamine to arginine amino acid substitution in position 78, within the coiled-coil domain.

us to design a primer pair that would discern between Dvfor1 and the latter primer pair was used to re-examine the relative foraging Dvfor Ps1 (but not between Dvfor1 and Dvfor Ps2, data not shown). transcript levels between the 12 variant and 12 normal female This primer pair was used for the remainder of this study. When insects, a statistically signiﬁcant (one-way ANOVA p-value ¼ 0.03) ARTICLE IN PRESS

F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704 701

Female Male Female Male

std N V N V std N V N V 6.5 6.5

4.3 4.3

2.3 2.3

2.0 2.0

Fig. 2. Restriction enzyme blot of western corn rootworm genomic DNA probed with foraging gene fragments. Thirty microgram samples of EcoRI digested genomic DNA from 10 female and 10 male, normal and variant behavioral biotypes, were probed with (A) a 250 bp and (B) a 200 bp digoxigenin (DIG) probe for the distal and proximal stretches of the serine/threonine kinase domain of the foraging gene, respectively. The sizes of the DNA markers in the left lane are shown in kilobases.

100 1.5 80

1.0 60

0.5 20 DvFor1 Mean Transcript No. Transcript DvFor1 Mean DvFor1 Relative Expression 0 0 NVNVNV NVN V 3rd Instar 2-W-Old 4-W-Old

Male Female Fig. 4. Changes in the expression of the foraging gene during development of the WCR. Absolute levels of the foraging transcript in heads of 3rd instar and 2- and 4-week-old adult WCRs determined with PCR primers that discerned between 1.5 Dvfor1 and Dvfor Ps1 (but not between Dvfor1 and Dvfor Ps2). The values are means of 16 3rd instars, 28 2-week-olds, and 28 4-week-old individual assays per developmental stage.

1.0 higher level (25%) of expression was observed in the variant females compared with normal females. Standard curve correlation coefﬁcient values were 0.968 and 0.917, with slopes of 3.90 and 2.92, intercepts of 43.4 and 41.0, and PCR efﬁciencies of 80.4% and 119.9% for ITS2 and Dvfor1, respectively (Fig. 3B). 0.5 In another real-time experiment, the foraging gene levels in heads of normal and variant insects in three developmental

DvFor1 Relative Expression stages, including 3rd instar larvae, 2- and 4-week-old female adults were compared. The study included a total of 72 insects (16 0 3rd instars, 28 2-week-old, and 28 4-week-old). Standard curves N V for both fragments were linear when plotted as log of starting quantity, with correlation coefﬁcient values of 0.998 and 0.999, Female slopes of 3.37 and 3.52, intercepts of 31.0 and 32.9, and PCR

Fig. 3. Relative expression of the foraging gene transcript in heads of adult WCR: efﬁciencies of 97.9% and 92.3% for ITS2 and Dvfor1, respectively. (A) real-time PCR assay of transcript levels in lab-reared, male and female WCRs Expression levels relative to the 3rd instar normal larvae appeared from normal and variant behavioral biotypes. The assay was carried out with a to change with development, with the highest levels occurring in general primer pair that spans the highly conserved region among the foraging the 2-week-old adults (Fig. 4). Also, the levels in the variant messages. The values are means of six individuals per group. (B) An analysis of insects were consistently higher. Because the expression data set foraging transcript levels in additional normal and variant female WCR using PCR primers that discerned between Dvfor1 and Dvfor Ps1 (but not between Dvfor1 and for the developmental study followed a normal distribution Dvfor Ps2). The values are means of 12 female insects per group. (Kolmogorov-Smirnov p-value ¼ 0.15), it was analyzed as a ARTICLE IN PRESS

702 F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704

Table 1 (coiled-coil-containing protein) also generated through alternate Western corn rootworm mean foraging transcript number at three stages of splicing, have suggested that the expressed pseudogene stabilizes development (Fig. 4), expressed per 10 ng total RNA Makorin1 expression through selective stabilization of the

Factor F-value p4F LSD Mean transcript authentic mRNA. We characterized one complete foraging tran- 0 no. per 10 ng total script and two pseudo-transcripts that had the same 3 kinase RNA sequence but varied in the 50 end. The first of the three real-time experiments described here was performed prior to finding Developmental stage 18.19 0.0001 9.9 o pseudotranscripts Dvfor Ps1&2 and the primer pair used in the 3rd instar 55.3a 2-week-old adult 79.7b assay was common to all three transcripts. In the second and third 4-week-old adult 55.1a experiments, a different primer pair was used that would discern between Dvfor1 and Dvfor Ps1, but not between Dvfor1 and Dvfor Behavior 7.10 0.0097 7.9 Normal 59.3a Ps2. Nonetheless, a greater difference was observed in the foraging Variant 70.2b transcript levels between the two behavioral types. Perhaps there Developmental 0.10 0.9038 would be an even greater difference between normal and variant stage behavior insects if the PCR assay were completely specific for the authentic

Means followed by the same letter are not significantly different from each other. transcript. Data was analyzed as a two-way factorial ANOVA with the main factors being In spite of our inability to sex the 3rd instars, a significant developmental stage and type of behavior. difference was observed between normal and variant larvae at this stage. Since normal and variant males did not show a two-way factorial ANOVA using the GLM procedure with the main difference in expression levels of the foraging gene (Fig. 3A), we factors being developmental stage and type of behavior (statistical speculate that an even greater difference could have been package: SAS). The analysis indicated that there was no significant observed if we were able to discern between the two sexes in interaction between the two factors (p-value ¼ 0.90, Table 1). this developmental stage. However, a significant difference in the expression of the foraging Another potential level of complexity in PKG gene expression is transcript occurred between the 2-week-old and the 4-week-old indicated by our Southern blots that show the presence of more and 3rd instar stages of development (F ¼ 18.19, po0.0001) with than one foraging gene locus in the Diabrotica genome. Similarly, the highest levels of expression occurring in the two-week-old Drosophila contains two PKG genes (dg1 and dg2, Kalderon and adult variant females. Additionally, there were significantly Rubin, 1989). The proteins from the two major transcripts of dg2 (F ¼ 7.10, p ¼ 0.0097) higher (18%) transcript levels in the variant are membrane associated and likely involved in signaling, compared with the normal insects across all three stages of whereas the dg1 transcript encodes a cytosolic protein (MacPher- development. son et al., 2004). We currently have no indication that a similar diversity of function exists in WCR but the multiple bands in the Southern blots suggest that potential for a similar range of cGMP 4. Discussion activities exists in Diabrotica. Also, bulked samples from the variants had an extra band in the Southern blots. Therefore, in The current results provide the first physical difference, based future studies, it would be interesting to compare the levels of on gene expression, between normal and rotation resistant WCR PKG protein and enzyme activity between normal and variant populations. Further, these results are among the first that insects. identify a relationship between gene expression and behavior in The sequence difference in the Dvfor1 cDNA from normal Coleoptera; the most diverse order of animals. versus variant insects that would cause replacement of a The increase in the level of the Dvfor1 transcript that we Glutamine with Arginine at position 78 within the coiled-coil observed in variant females of WCR was similar to reports of domain of the protein may have significant effects on the activity higher transcript levels for this gene in behavioral variants of of the protein. For example, point mutations in various positions Drosophila and Apis. In all cases, the variants which travel the of the coiled-coil region of c-Fes, a tyrosine kinase, affects the farthest, namely, the WCR variant that lays eggs outside of corn kinase activity of the protein positively or negatively, depending fields, the insects with the rover allele in Drosophila (Osborne on the position of the mutation (Cheng et al., 2001). It may be et al., 1997), and the older bees that leave the hive to collect nectar possible to exploit this nucleotide difference to distinguish in Apis (Ben-Shahar et al., 2002), have the higher levels of the between normal and variant insects using a simpler assay such foraging gene transcript. In Drosophila, there was approximately as PCR-based SNP detection. 25% less of one of the T1 PKG transcript in sitters compared to Currently, the only method for identifying rotation-adapted rovers (Osborne et al., 1997) and rovers have a 12% increase in PKG WCR is by evaluating their behavior through emergence trapping enzyme activity in their heads compared with sitters (Sokolowski, in first year corn fields. A clear indication that variant insects exist 2001). The change in the level of the foraging gene transcript is in a region can only be obtained from several years of field greater in honeybees, ranging from two- to eight-fold higher in monitoring (Knight et al., 2005). The current results are based on foragers than in normals (Ben-Shahar et al., 2002). However, insects from laboratory colonies established from areas that are Sokolowski (2001) suggested that small changes in transcript considered to be free of the variant (South Dakota) and from a levels in extracts from whole brain tissue may, in fact, reflect fairly field in a region that is known to have the variant (Indiana). large changes in levels in the cells that are involved in controlling However, the use of the colonies introduces two possible the behavior. confounding factors into the analysis, namely: a possible The complexity of the foraging gene and its transcripts also geographic effect and a possible artifact attributable to selection complicate assays of its activity. The functional significance of the in a closed colony. While these possibilities must be acknowl- multiple transcripts from PKG coding and psuedogenes has not edged, several molecular studies of genetic diversity have shown been determined but it seems to be a common feature of this gene that there is little or no geographic structuring in Diabrotica family. Multiple splicing versions of PKG transcripts were also (Szalanski et al., 1999; Kim and Sappington, 2005) and no found in the nematode C. elegans (Fujiwara et al., 2002). Hirotsune significant differences in genetic diversity between insects from et al. (2003), working with Makorin1 mRNA, a ring finger protein the laboratory colonies at the North Central Agricultural Research ARTICLE IN PRESS

F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704 703

Laboratory in Brookings, SD (which have been maintained for identification of a gene expression difference between normal and more than 20 generations) and insects collected from the wild rotation resistant WCR provides a starting point for a number of (Kim et al., 2007). studies, including analysis of Dvfor1 mutant dispersal and egg- All of the differences we observed between normal and variant laying behaviors as well as cellular and molecular studies of WCRs could potentially be used to develop assays to identify the DvFOR interactors. Ultimately, this information may be used to variant insects. The real-time PCR assay for the Dvfor1 transcript devise better monitoring methods for variant WCRs and perhaps could be used for this purpose, but it is clear from the current to devise better pest control strategies. results that it will be important to control for insect developmental stage, since the overall levels of foraging transcript varied between 2- and 4-week-old adults even though the difference between normals and variants were similar within each develop- Acknowledgments mental stage. An antibody-based quantitative ELISA can also be utilized with the advantage of high throughput processing. A PCR- The work was supported by funding from the Ontario based single nucleotide polymorphism (SNP) detection assay, Agricultural Genomics Centre of the University of Guelph, based on the sequence difference observed between the normal Monsanto Inc., the Ontario Corn Producers Association, and the and variant Dvfor1 cDNA, might have the greatest utility as a high- Canadian Agricultural Adaptation Council (CanAdapt). throughput detection system for identifying variant WCR because it would circumvent potential complications caused by measuring References transcript or enzyme levels that were shown to be affected by stage of development. The GC-rich nature of this region, however, Ben-Shahar, Y., Robichon, A., Sokolowski, M.B., Robinson, G.E., 2002. Influence might pose a challenge in developing an SNP diagnostic assay. of gene action across different time scales on behavior. Science 296, It is still not clear whether the feeding and egg-laying 741–744. Cheng, H.Y., Schiavone, A.P., Smithgall, T.E., 2001. A point mutation in the n- behaviors are coupled in WCR since there are conflicting reports terminal coiled-coil domain releases c-Fes tyrosine kinase activity and survival available in the literature. One group reports on adult WCRs signaling in myeloid leukemia cells. Mol. Cell Biol. 21, 6170–6180. counts rising in soybean fields adjacent to cornfields as the season Daniels, S.A., Ailion, M., Thomas, J.H., Sengupta, P., 2000. Egl-4 acts throught progresses in affected areas but not in regions with no variant growth factor-b/SMAD pathway in caenorhabditis elegans to regulate multiple neuronal circuits in response to sensory cues. Genetics 156, 123–141. WCR damage (Levine et al., 2002) suggesting that a selective loss Darnell, S.J., Meinke, L.J., Young, L.J., 2000. Influence of corn phenology on adult of feeding preference occurs in the variants compared to normal western corn rootworm (Coleoptera: Chrysomelidae) distribution. Environ. insects (Rondon and Gray, 2003, 2004). These reports conclude Entomol. 29, 587–595. Fujiwara, M., Sengupta, P., McIntire, S.L., 2002. Regulation of body size and that the lack of preference for corn allows the variant WCR to behavioral state of C. elegans by sensory perception and the EGL-4 cGMP- travel farther in search of food, which may result in incidental dependent protein kinase. Neuron 36, 1091–1110. oviposition in adjacent soybean, wheat or alfalfa fields due to Hirotsune, S., Yoshida, N., Chen, A., Garrett, L., Sugiyama, F., Takahashi, S., Yagami, K., Wynshaw-Boris, A., Yoshiki, A., 2003. An expressed pseudogene regulates favorable soil conditions. In contrast, Darnell et al. (2000) showed the messenger-RNA stability of its homologous coding gene. Nature 423, that WCR’s feeding preference is generally influenced by the age 91–96. of corn plants, and that insects tend to move to areas where Ingram, K.K., Oefner, P., Gordon, D.M., 2005. Task-specific expression of the foraging gene in harvester ants. Mol. Ecol. 14, 813–818. younger plant material is available for feeding, and O’Neal et al. Kalderon, D., Rubin, G.M., 1989. cGMP-dependent protein kinase genes in (2002) failed to find any difference in the feeding behavior of WCR Drosophila. J. Biol. Chem. 264, 10738–10748. on soybean leaves from affected and non-affected areas and Kim, K.S., Sappington, T.W., 2005. Genetic structuring of western corn rootworm (Coleoptera: Chrysomelidae) populations in the US based on microsatellite loci suggested that WCR’s acceptance for soybean leaves increases as analysis. Environ. Entomol. 34, 494–503. corn plants grow older. The findings of O’Neal et al. (2002) Kim, K.S., French, B.W., Sumerford, D.V., Sappington, T.W., 2007. Genetic diversity suggests that a loss of ovipositioning preference for cornfields, in laboratory colonies of western corn rootworm (Coleoptera: Chrysomelidae), and not a feeding preference, is what distinguish the variants from including a nondiapause colony. Environ. Entomol. 36, 637–645. Knight, C.C., Pauls, K.P., Sears, M.K., Schaafsma, A.W., 2005. Oviposition site normal WCRs. This implies that in areas with normal insect selected by the western corn rootworm (Diabrotica virgifera virgifera Leconte) populations, the insects return to cornfields for egg-laying, and in southern Ontario strip plots. Can. J. Plant Sci. 85, 949–954. that foraging and egg-laying are two separate behaviors, similar to Knolhoff, L.M., Onstad, D.W., Spencer, J.L., Levine, E., 2006. Behavioral differences between rotation-resistant and wild-type Diabrotica virgifera virgifera northern corn rootworms that move around to various non-corn (Coleoptera: Chrysomelidae). Environ. Entomol. 35, 1049–1057. fields for feeding but return to corn for laying eggs (Naranjo, Krysan, J.L., Branson, T.F., 1983. Biology, ecology, and distribution of Diabrotica. In: 1991). Gordon, D.T., Knoke, J.K., Nault, L.R., Ritter, R.M. (Eds.), Proceedings of the International Maize Virus Discussion Colloqium and Workshop, Wooster, OH, Although the question of whether feeding and egg-laying pp. 144–150. behaviors are coupled in WCR remains, it is interesting to note Levine, E., Oloumi-Sadeghi, H., 1996. Western corn rootworm (Coleoptera: that in the nematode (C. elegans) roaming and egg-laying are both Chrysomelidae) larval injury to corn grown for seed production following soybeans grown for seed production. J. Econ. Entomol. 89, 1010–1016. affected by the expression of the PKG gene EGL-4 in sensory Levine, E., Spencer, J.L., Isard, S.A., Onstad, D.W., Gray, M.E., 2002. Adaptation of the neurons (Trent et al., 1983; Fujiwara et al., 2002). Osborne et al. western corn rootworm to crop rotation: evolution of a new strain in response (1997 and references therein) make the point that behavioral to a management practice. Am. Entomol. 48, 94–107. MacPherson, M.R., Lohmann, S.M., Davies, S.-A., 2004. Analysis of drosophila cGMP- mutations often affect signal transduction components, like the dependent protein kinases and assessment of their in vivo roles by targeted PKG protein, and that PKG has a variety of pleiotropic cellular expression in a renal transporting epithelium. J. Biol. Chem. 279, regulatory functions typical of signal transduction components. 40026–40034. Based on our findings, and taking into consideration the Madder, D.L., Stemeroff, M., Gerber, G.H., Philips, H.G., Doane, J.F., Ellis, C.R., Becker, E.C., 1988. The economics of insect control on wheat, corn, and canola in physiological and behavioral effects of altered PKG levels in other Canada. Bull. Entomol. Soc. Can. Suppl. 20, 168. organisms, we suggest the change in ovipositioning site selection Naranjo, S.E., 1991. Movement of corn rootworm beetles, Diabrotica spp. observed in variant WCR is either directly or indirectly linked to (Coleoptera: Chrysomelidae), at cornfield boundaries in relation to sex, reproductive status, and crop phenology. Environ. Entomol. 20, 230–240. higher PKG levels, which coincides with a documented shift in the O’Neal, M.E., Gray, M.E., Smyth, C.A., 1999. Population characteristics of a western insects’ feeding grounds from aging corn to fresher non-corn plant corn rootworm (Coleoptera: Chrysomelidae) strain in east-central Illinois corn material such as soybean. and soybean fields. J. Econ. Entomol. 92, 1301–1310. O’Neal, M.E., DiFonzo, C.D., Landis, D.A., 2002. Western corn rootworm (Coleop- While many questions remain to be addressed regarding the tera: Chrysomelidae) feeding on corn and soybean leaves affected by corn genetic and physiological bases of WCR’s rotation-adaptation, the phenology. Environ. Entomol. 31 (2), 285–292. ARTICLE IN PRESS

704 F. Garabagi et al. / Insect Biochemistry and Molecular Biology 38 (2008) 697–704

Onstad, D.W., Joselyn, M.G., Isard, S.E., Levine, E., Spencer, J.L., Bledsoe, L.W., Sappington, T.W., Siegfried, B.D., Guillemaud, T., 2006. Coordinated diabrotica Edwards, C.R., Di Fonzo, C.D., Willson, H., 1999. Modeling the spread of western genetics research: accelerating progress on an urgent insect pest problem. Am. corn rootworm (Coleoptera: Chrysomelidae) population adapting to soybean- Entomol. 52 (2), 90–97. corn rotation. Environ. Entomol. 28, 188–194. Schaafsma, A.W., Baufeld, P., Ellis, C.R., 1999. Inﬂuence of cropping practices on Onstad, D.W., Spencer, J.L., Guse, C.A., Levine, E., Isard, S.A., 2001. Modeling corn rootworm in Canada as a basis for assessment of the potential impact of evolution of behavioral resistance by an insect to crop rotation. Entomol. Exp. Diabrotica virgifera in Germany. EPPO. Bull. 29, 145–154. Appl. 100, 195–201. Schafer, W.R., 2002. PKG and the neural basis for behavioral phenotypes. Neuron Osborne, K.A., Robichon, A., Burgess, E., Butland, S., Shaw, R.A., Coulthard, H.S., 36, 991–993. Pereira, H.S., Greenspan, R.J., Sokolowski, M.B., 1997. Natural behavior Sokolowski, M.B., 2001. Drosophila: genetics meets behaviour. Nat. Rev. 2, 879–890. polymorphism due to a cGMP-dependent protein kinase of Drosophila. Science Sokolowski, M.B., Pereira, H.S., Hughes, K., 1997. Evolution of foraging behavior in 277, 834–836. drosophila by density-dependent selection. Proc. Natl. Acad. Sci. USA 94, Pereira, H.S., Sokolowski, M.B., 1993. Mutations in the larval foraging gene affect 7373–7377. adult locomotory behavior after feeding in Drosophila melanogaster. Proc. Natl. Szalanski, A.L., Roehrdanz, R.L., Taylor, D.B., Chandler, L., 1999. Genetic variation in Acad. Sci. USA 90, 5044–5046. geographical populations of western and Mexican corn rootworm. Insect Mol. Rondon, S.I., Gray, M.E., 2003. Captures of western corn rootworm (Coleoptera: Biol. 8, 519–526. Chrysomelidae) in Pherocon AM and vial traps in four crops in East Central Tanoue, S., Nishioka, T., 2003. Molecular characterization of a membrane-bound Illinois. J. Econ. Entomol. 96, 737–747. cGMP dependent protein kinase from the silk moth Bombyx mori. Insect. Mol. Rondon, S.I., Gray, M.E., 2004. Ovarian development and ovipositional preference Biol. 12, 621–629. of the western corn rootworm (Coleoptera: Chrysomelidae) variant in east Trent, C., Tsung, N., Horvitz, H.R., 1983. Egg-laying defective mutants of the central illinois. J. Econ. Entomol. 97, 390–396. nematode Caenorhabditis elegans. Genetics 104, 619–647.