Materials

Insect samples used in this study are listed in Table 1. Upon collection, some insects were dissected with fine forceps under a binocular microscope to isolate bacteriomes. Other insects were immediately placed in acetone-filled glass vials and preserved at room temperature until molecular analyses (Fukatsu, 1999).

Molecular biological procedures

Each sample was crushed and digested in a 1.5-ml plastic tube with a lysis buffer (10mM Tris-HCl [pH 8.0], 1 mM EDTA, 0.1 M NaCl, 0.5% sodium dodecyl sulfate, 0.2 mg/ml proteinase K) at 56°C overnight. DNA was extracted with phenol-chloroform, precipitated with ethanol, dried, and dissolved in TE buffer (10 mM TrisHCl [pH 8.0], 0.1 mM EDTA). Bacterial 16S rRNA gene fragments were amplified from the DNA samples by PCR using the following primer sets and conditions. A 1.5-kb fragment was amplified with primers 16SA1 (5′-AGA GTT TGA TCM TGG CTC AG-3′) and 16SB1 (5′-TAC GGY TAC CTT GTT ACG ACT T-3′) under conditions of 95°C for 10 min followed by 35 cycles of 95°C for 30 sec, 52°C for 30 sec, and 72°C for 2 min (Fukatsu and Nikoh, 1998). A 1.1-kb fragment was amplified with primers EUB338F (5′-ACT CCT ACG GGA GGC AGC-3′), a sequence complementary to oligonucleotide probe EUB338 used to detect eubacterial 16S rRNA (Amann et al., 1990), and 16SB1 under conditions of 95°C for 10 min followed by 35 cycles of 95°C for 20 sec, 61°C for 20 sec, and 72°C for 2 min. A 1.5-kb fragment was amplified with primers 16SA1 and gma16S-R1 (5′-GTG ATT CAT GAC TGG GGT G-3′) under conditions of 95'C for 10 min followed by 35 cycles of 95°C for 10 sec, 58°C for 30 sec, and 72°C for 2 min. A 0.6-kb fragment of the wsp gene from Wolbachia was amplified with primers wspF (5′-GGG TCC AAT AAG TGA TGA AGA AAC-3′) and wspR (5′-TTA AAA CGC TAC TCC AGC TTC TGC3′) under conditions of 95°C for 10 min followed by 35 cycles of 95°C for 20 sec, 55°C for 20 sec, and 72°C for 1 min (Kondo et al., 2002). PCR products were cloned, genotyped by means of restriction fragment length polymorphisms, and sequenced essentially as previously described (Fukatsu and Nikoh, 1998).

Molecular phylogenetic analysis

Multiple alignments of nucleotide sequences were generated with the Vector NTI Advance 10.3.1 (Invitrogen). The alignments were then inspected and corrected manually with the GeneDoc 2.6.002 (Nicholas et al., 1997) to remove ambiguously aligned nucleotide sites. Phylogenetic analyses were conducted by three methods: maximum parsimony (MP), maximum likelihood (ML), and Bayesian analysis (BA). MP trees were constructed with the MEGA 4.0 (Tamura et al., 2007). In the analysis, all sites and character changes were weighed equally. ML and BA trees were created with PhyML 3.0 (Guindon and Gascuel, 2003) and MrBayes v3.1.2 (Ronquist and Huelsenbeck, 2003), respectively. We selected the GTR+I+G model for phylogenetic analyses with 16S, and the HKY+I model for analysis with wsp, on the basis of the Akaike information criterion estimated with Modeltest 3.7 (Posada and Crandall, 1998). Bootstrap values were determined from 100 replicates in the MP and ML analyses. In the BA analysis, 3750 trees were obtained for each analysis (ngen=500,000, samplefreq=100, burn in=1250) and used to construct a 50% majority-rule consensus tree.

Relative rate test

A relative rate test was performed with RRTree (RobinsonRechavi and Huchon, 2000) on the basis of genetic distances estimated under Kimura's two-parameter model (Kimura, 1980). For the flavobacterial and enterobacterial 16S sequences, 1147 and 1349 unambiguously aligned nucleotide sites were subjected to the analysis, respectively.

Diagnostic PCR

A 0.2-kb fragment of the 16S rRNA gene from the flavobacterial symbiont was amplified with primers DcFlv1 079F (5′-AGG TGT TGG GTT AAG TCC TGA-3′) and DcFlv1 248R (5′-CCA GTG GCT TCT CTC TGT A-3′). A 0.4-kb fragment of 16S from the enterobacterial symbiont was amplified with primers DcEnt628F (5′-AAC TGC ATT CGA GAC TGG T-3′) and DcEnt1017R (5′-CCC GAA GGC ACC AAA GG-3′). A 0.6-kb fragment of wsp from the Wolbachia symbiont was amplified with the primers wspF and wspR. These PCR reactions were performed under conditions of 95°C for 10 min followed by 35 cycles of 95°C for 15 sec, 55°C for 15 sec, and 72°C for 1 min, with negative and positive control samples. To confirm the quality of template DNA in the samples, a 0.65-kb fragment of the insect 18S rRNA gene was amplified with primers 2880 (5′-CTG GTT GAT CCT GCC AGT AG-3′) (Tautz et al., 1988) and B (5′-CCG CGG CTG CTG GCA CCA GA-3′) (von Dohlen and Moran, 1995) under conditions of 95°C for 10 min followed by 30 cycles of 95°C for 15 sec, 55°C for 15 sec, and 72°C for 1 min.

Fluorescence in-situ hybridization

Fluorescence in-situ hybridization was performed essentially as previously described (Koga et al., 2009) with oligonucleotide probe Cy5-DcFlv1 410 (5′-Cy5-ATA CCT CCG ACT TCC AGG A-3′), which targeted 16S rRNA of the flavobacterial symbiont, and Cy3DcEnt1248 (5′-Cy3-GAG GTC GCT TCT CTT TG-3′), which targeted 16S rRNA of the enterobacterial symbiont.

For whole-mount in-situ hybridization, female adult insects were dissected and thoroughly washed in phosphate-buffered saline (PBS; 137 mM NaCl, 8.1 mM Na₂HPO₄, 2.7 mM KCl, 1.5 mM KH₂PO₄ [pH 7.4]), and preserved in acetone until use. Male adult insects were decapitated to facilitate the infiltration of reagents, and preserved in acetone. The preserved insect samples were dissected and thoroughly washed in 80% ethanol, and fixed in Carnoy's solution (6:3:1 ethanol:chloroform:acetic acid). After an overnight fixation, the tissues were treated with 6% hydrogen peroxide in 80% ethanol for several weeks to quench autofluorescence from the tissues (Koga et al., 2009). After thorough washing with absolute ethanol and PBST buffer (137 mM NaCl, 8.1 mM Na₂HPO₄, 2.7 mM KCl, 1.5 mM KH₂PO₄ [pH 7.4], 0.2% Tween20), the samples were incubated with hybridization buffer (20 mM TrisHCl [pH 8.0], 0.9 M NaCl, 0.01% sodium dodecyl sulfate, 30% formamide) three times for 5 min each. The prehybrized samples were hybridized in 100 nM solutions of each of the probes Cy5DcFlv1410 and Cy3-DcEnt1248 in hybridization buffer overnight. Nuclei of the host cells were counterstained with 4 µM 4′, 6diamidino-2-phynylindole (DAPI). After thorough washing with PBST, the samples were mounted in Slowfade antifade solution (Invitrogen) and observed under an Axiophot epifluorescence microscope (Carl Zeiss) or a Pascal 5 laser confocal microscope (Carl Zeiss). The specificity of the hybridization signals was confirmed by the following control experiments: no-probe control, RNase digestion control, and competitive suppression control with excess unlabelled probes.

For in-situ hybridization of sectioned tissues, the preserved samples were washed with 80% ethanol and treated with 6% hydrogen peroxide in 80% ethanol to quench autofluorescence. The samples were thoroughly washed with absolute ethanol, dehydrated and cleared through an ethanol-xylene series, embedded in paraffin, and processed into 4-µm serial tissue sections on a rotary microtome. The tissue sections were mounted on silane-coated glass slides, dewaxed through a xylene-ethanol series, and dried. The tissue preparations were incubated in about 150 µl of hybridization buffer containing 100 nM each probe, covered with a coverslip, and incubated in a humidified chamber at room temperature overnight. After thorough washing with Tris-buffered saline (20 mM Tris-HCl [pH 7.4], 0.15 M NaCl), the samples were mounted in Slowfade antifade solution (Invitrogen) supplemented with either 4µM DAPl or 0.25 µM SYTOX Green (Invitrogen). The rest of the procedure was performed in the same manner as the whole-mount in-situ hybridization.

Electron microscopy

Fresh insects were dissected with fine forceps in 0.1 M sodium cacodylate buffer (pH 7.4) containing 2.5% glutaraldehyde, and isolated bacteriomes were prefixed in the fixative at 4°C overnight and postfixed in 2% osmium tetroxide at 4°C for 90 min. After dehydration through an ethanol series, the materials were embedded in Spurr resin (Nisshin-EM). Ultrathin sections were made on an Ultracat-N ultramicrotome (Leichert-Nissei), mounted on collodioncoated copper meshes, stained with uranyl acetate and lead citrate, and observed under a Model H-7000 transmission electron microscope (Hitachi).

Nucleotide sequence accession numbers

The nucleotide sequences determined in this study have been deposited in the DDBJ/EMBL/GenBank nucleotide sequence databases under accession numbers AB491196-AB491203 and AB492154.

General observation of the symbiotic organ

Fig. 2A shows the internal organs of a dissected adult female D. corpulenta. In the central abdomen, there were large white ovaries and red Malphigian tubules. On both sides of the abdominal body cavity, there were symbiotic organs consisting of around six large lobes, the bacteriomes. The bacteriome lobes were connected anteroposteriorly to each other and often reached 5 mm in length. Light microscopy of tissue sections revealed that the bacteriomes consisted of three cellular components: large, syncytial central bacteriocytes located at the center of each lobe; smaller, uninucleate peripheral bacteriocytes surrounding the central bacteriocytes; and thin sheath cells located on the outer surface of the bacteriomes as well as intercalating the bacteriocytes (data not shown). Symbiontlike particles were observed in the cytoplasm of the central and peripheral bacteriocytes (data not shown). The same configuration of the symbiotic organs was observed in D. pinicola, with the bacteriomes consisting of around five lobes and reaching 2 mm in length (data not shown).

Fig. 2.

Internal tissues and organs of Drosicha giant scales, and localization of the flavobacterlal (green) and enterobacterial (red) symbionts visualized by fluorescence in-situ hybridization. (A) Dorsal view of the dissected whole body of an adult female D. corpulenta. A pair of large lobed bacteriomes (3–5 mm in length) is located on both sides of the abdomen. (B) Localization of the flavobacterial symbiont and the enterobacterial symbiont in the bacteriome lobes of D. corpulente. (C) Co-localization of the flavobacterial and enterobacterial symbionts in the bacteriome lobes of D. pinicola. (D) Localization of the flavobacterial and enterobacterial symbionts in the bacteriome lobes of D. corpulenta. (E) Enlarged image of a bacteriome lobe of D. corpulente harboring the flavobacterial and enterobacterial symbionts in the peripheral and central bacteriocytes, respectively. (F) Enlarged image of bacteriocytes of D. corpulenta. wherein symbiont cells are evident in the cytoplasm. (G) Enlarged image of a bacteriome lobe of D. pinicola. (H) Enlarged image of bacteriocytes of D. pinicola. (I) Localization of the flavobacterial and enterobacterial symbionts in an adult male D. pinicola. Arrows and arrowheads indicate signals from the flavobacterial and enterobacterial symbionts, respectively. (A–C) and (I) are images of whole mounts; (D–H) are images of sectioned tissue. Abbreviations: am, anterior midgut; bac, bacteriome; cbc, central bacteriocyte; mt, Malpighian tubule; ov, ovary; pbc, peripheral bacteriocyte; pm, posterior midgut; sc, sheath cell; sg, salivary gland; sog, suboesophageal ganglion; vg, vagina.

Bacterial 16S rRNA gene sequences from D.corpulenta

From an adult female of D. corpulenta from Tsukuba, 24 clones of the 16S rRNA gene fragment amplified with primers 16SA1 and 16SB1 were sequenced and yielded 22 identical flavobacterial sequences and two identical γproteobacterial sequences. These sequences were consistently identified from five more insect individuals from Tsukuba, three from Ueno, and three from Uozu. BLASTN searches against the DNA databases identified the highest hits for the flavobacterial sequence as a bacterial symbiont from the cushion scale Icerya brasiliensis (96.3% sequence similarity [1239/1286], accession number DQ133551), and for the γ-proteobacterial sequence, as an uncultured bacterium from the antlion Myrmeleon mobilis (98.5% [1421/ 1443], DQ068866).

Bacterial 16S rRNA gene sequences from D.pinicola

From an adult female of D. pinicola from Tsukuba, 24 clones of the 16S fragment amplified with the primers 16SA1 and 16SB1 were sequenced and yielded 18 identical flavobacterial sequences and six identical α-proteobacterial sequences. These sequences were consistently detected from six more insect individuals from Tsukuba. BLASTN searches against the DNA databases retrieved the highest hits for the flavobacterial sequence as symbiont from the cushion scale Icerya brasiliensis (96.3% [1239/1286], DQ133551) and for the α-proteobacterial sequence as a Wolbachia symbiont from the chrysomelid beetle Diabrotica lemniscata (99.2% [1416/1427], AY007547). PCR with primers EUB338F and 16SB1 amplified a γ-proteobacterial sequence very similar to that identified from D. corpulenta. All eight clones sequenced represented the same γproteobacterial sequence, but no flavobacterial sequence was obtained, probably because EUB338F contained mismatches to the flavobacterial sequence. On the basis of the sequence determined, we designed a new, specific reverse primer, gma16S-R1, upstream of 16SB1. PCR with primers 16SA1 and gma16S-R1 consistently yielded the γproteobacterial sequence from 10 individuals of D. pinicola. A BLASTN search against the DNA databases retrieved the highest hit for the γ-proteobacterial sequence as Pantoea agglomerans (98.2% [1417/1443], AM184290).

Molecular phylogenetic analysis of flavobacterial sequences

Fig. 3 shows the phylogenetic relationships among the flavobacterial 16S rRNA gene sequences from D. corpulenta and D. pinicola and the representative flavobacterial sequences. The sequences from the giant scales formed a highly supported clade with the symbiont sequences from the cushion scales (Icerya spp.). Note that Drosicha spp. and Icerya spp. belong to the same family Monophlebidae. The symbiont sequence from Cryptococcus ulmi of Family Eriococcidae was placed just outside the clade. These sequences constituted a sister clade to that containing the symbiont sequences from armored scale insects of Family Diaspididae (Uzinura diaspidicola). The monophlebideriococcid-diaspidid symbiont sequences further formed a sister clade to the clades comprising the male-killing symbiont sequences from lady beetles, the symbiont sequences from cockroaches and termites (Blattabacterium spp.), and the symbiont sequences from cicadas and planthoppers (Sulcia mulleri). Sequences from free-living flavobacteria were placed outside the large clade of flavobacterial insect symbionts. We designate this type of Drosicha symbionts as “flavobacterial symbionts”.

Fig. 3.

Molecular phylogeny including the flavobacterial symbionts from Drosicha giant scales and allied symbiotic and free-living flavobacteria, inferred from 16S rRNA gene sequences. In all, 1149 aligned nucleotide sites were analyzed. The Bayesian (BA) tree is shown, but the MP and ML analyses gave substantially the same results (data not shown). Support values are indicated near nodes, with posterior probabilities from the BA shown above bootstrap values from the MP/ ML analyses. Asterisks indicate support values lower than 50%; sequences from Drosicha giant scales are indicated by bold font; collection localities are in parentheses; sequence accession numbers are in brackets; and the AT content (%) of each sequence is given. F-symbiont, flavobacterial symbiont.

Molecular phylogenetic analysis of γ-proteobacterial sequences

Fig. 4 shows the phylogenetic relationships among the γproteobacterial 16S rRNA gene sequences from D. corpulenta and D. pinicola and representative γ-proteobacterial sequences. The symbiont sequences from the giant scales formed a robust clade, and were allied to free-living γ-proteobacteria such as Enterobacter aerogenes and Escherichia coli. No γproteobacterial insect symbiont sequences exhibited close phylogenetic affinity to the Drosicha symbiont sequences. We designate this type of Drosicha symbionts as “enterobacterial symbionts”.

Molecular phylogenetic analysis of Wolbachia sequence

Fig. 5 shows the phylogenetic placement of the wsp sequence from Wolbachia from D. pinicola. This sequence was closely related to the Wolbachia sequence from the sand fly Phlebotomus papatasi, and was placed in Wolbachia supergroup A.

Prevalence of flavobacterial, enterobacterial, and Wolbachia symbionts in host populations

Diagnostic PCR surveys revealed 100% infection frequencies for the flavobacterial and enterobacterial symbionts in populations of D. corpulenta and D. pinicola. The Wolbachia symbiont exhibited 0% and 100% infection frequencies in populations of D. corpulenta and D. pinicola, respectively (Table 1).

Localization of flavobacterial and enterobacterial symbionts

Whole-mount in-situ hybridization of adult females of D. corpulenta revealed that the flavobacterial and enterobacterial symbionts are localized peripherally and centrally in the bacteriome lobes, respectively (Fig. 2B). In-situ hybridization of tissue sections confirmed the peripheral and central locations of the flavobacterial and enterobacterial symbionts in the bacteriome lobes (Fig. 2D, E). The uninucleate peripheral bacteriocytes harbored the flavobacterial symbiont in their cytoplasm (Fig. 2E, F), whereas the syncytial central bacteriocytes housed the enterobacterial symbiont endocellularly (Fig. 2D, E). In adult females of D. pinicola, the same patterns of symbiont localization were observed (Fig. 2C, G, H). In adult males of D. pinicola, the flavobacterial and enterobacterial symbionts each exhibited a peculiar localization pattern in the abdomen (Fig. 2I).

Fig. 4.

Molecular phylogeny including the enterobacterial symbionts from Drosicha giant scales and allied symbiotic and free-living γ-proteobacteria, inferred from 16S rRNA gene sequences. In all, 1338 aligned nucleotide sites were analyzed. The BA tree is shown, but the MP and ML analyses gave substantially the same results (data not shown). Support values are indicated near nodes, with the posterior probability from the BA analysis shown above the bootstrap values from the MP/ML analyse. Asterisks indicate support values lower than 50%; sequences from Drosicha giant scales are indicated by bold font; collection localities are in parentheses; sequence accession numbers are in brackets; and the AT content (%) of each sequence is given. E-symbiont, enterobacterial symbiont.

Fig. 5.

Molecular phylogenetic placement of the Wolbachia symbiont from Drosicha pinicola in Wolbachia supergroup A, on the basis of wsp gene sequences. In all, 504 aligned nucleotide sites were analyzed. The BA tree is shown, but the MP and ML analyses gave substantially the same results (data not shown). Support values are indicated near nodes, with the posterior probability from the BA analysis shown above the bootstrap values from the MP/ML analyses. Asterisks indicate support values lower than 50%; the sequence from D. pinicola is indicated by bold font; sequence accession numbers are in brackets.

Fine structure of the flavobacterial, enterobacterial, and Wolbachia symbionts

Fig. 6 shows transmission electron micrographs of the bacteriomes of Drosicha giant scales. In D. corpulenta, the flavobacterial symbiont was found in the peripheral bacteriocytes as large, pleomorphic bacterial cells with reduced cell wall, whereas the enterobacterial symbiont densely populated the central bacteriocytes as rod-shaped bacterial cells (Fig. 6A–C). In D. pinicola, the pleomorphic flavobacterial symbiont and the rod-shaped enterobacterial symbiont were similarly localized in the peripheral and central bacteriocytes, respectively (Fig. 6D–F). In D. pinicola, moreover, small rod-shaped bacterial cells of a different type densely populated the sheath cells (Fig. 6D, G) and were also observed in the bacteriocytes at lower densities (Fig. 6D, F). The small bacterial rods exhibited a well-developed cell wall (Fig. 6H). Judging from the molecular phylogenetic and diagnostic PCR data (Table 1; Fig. 5), the small bacteria were probably the Wolbachia symbiont.

Molecular evolution of the flavobacterial and enterobacterial symbionts

Relative rate tests detected no significant acceleration in molecular evolutionary rates in the 16S rRNA gene in the lineages of the flavobacterial and enterobacterial symbionts of Drosicha giant scales (Table 2). The 16S sequences from the symbionts exhibited no remarkable AT bias in nucleotide composition compared to the 16S sequences from their free-living relatives (Figs. 3, 4).

Fig. 6.

Transmission electron micrographs of the bacterial symbionts of Drosicha giant scales. (A) Pleomorphic flavobacterial symbiont cells in a peripheral bacteriocyte (left) and rod-shaped enterobacterial symbiont cells in a central bacteriocyte (right) in D. corpulenta. (B) A central bacteriocyte in D. corpulenta filled with enterobacterial symbiont cells (right) and an adjacent sheath cell harboring few symbionts (left). (C) Enlarged image of enterobacterial symbiont cells in D.corpulenta. (D) Pleomorphic flavobacterial symbiont cells in a peripheral bacteriocyte (left), rod-shaped enterobacterial symbiont cells in a central bacteriocyte (right), and presumed Wolbachia symbiont cells in a sheath cell (center) of D. pinicola. Some Wolbachia cells are also seen in the central bacteriocyte. (E) Enlarged image of flavobacterial symbiont cells in D. pinicola. A crystalline body (arrow) is evident in a symbiont cell. (F) Enlarged image of enterobacterial symbiont cells in D.pinicola. (G) Presumed Wolbachia symbiont cells in sheath cells of D. pinicola. (H) Enlarged image of presumed Wolbachia cells. Abbreviations: cbc, central bacteriocyte; mt, mitochondrion; n, nucleus; pbc, peripheral bacteriocyte; sc, sheath cell. Arrows, crystalline bodies observed in flavobacterial symbiont cells. Bars, 2µm.