Previous Article | Next Article ![]()
Applied and Environmental Microbiology, November 2009, p. 6757-6763, Vol. 75, No. 21
0099-2240/09/$08.00+0 doi:10.1128/AEM.01583-09
Copyright © 2009, American Society for Microbiology. All Rights Reserved.
,
,
National Institute of Agrobiological Sciences, Owashi, Tsukuba, Ibaraki 305-8634, Japan,1 Department of Integrated Biosciences, Graduate School of Frontier Sciences, The University of Tokyo, Kashiwanoha, Kashiwa, Chiba 277-8562, Japan,2 Faculty of Agriculture, Ibaraki University, Ami, Ibaraki 300-0393, Japan,3 Kohinata, 1-7-10, Satte, Saitama 340-0164, Japan,4 and Kyushu Research Station, National Institute of Animal Health, Kagoshima, Japan5
Received 5 July 2009/ Accepted 28 August 2009
|
|
|---|
|
|
|---|
Three independent studies have shown that rates of Cardinium infection were consistently low in wide samplings of arthropods, i.e., 7.2% of 223 species (46), 6% of 99 species (51), and 4.4% of 136 species (11). However, the infection frequencies in mites and spiders were 31.6% (46) and 22% (12), respectively. Cardinium has previously been detected only in hymenopteran insects (20, 25, 46, 50, 51), hemipteran insects (6, 24, 37, 46, 51), mites (13, 14, 15, 19, 45, 46), and spiders (11, 12). Infection by Wolbachia, another group of bacteria belonging to the Alphaproteobacteria that are capable of manipulating arthropod reproduction, is more widespread among arthropods. A recent meta-analysis of published data on Wolbachia infection surveys demonstrated that the proportion of insect species with at least one infected individual is around 66% (16). Other arthropods, such as wood lice, spiders, and mites, are also infected with Wolbachia. Outside of arthropods, Wolbachia infection has been detected in filarial nematodes (2, 23). Compared to Wolbachia, Cardinium organisms have been found in more restricted taxonomic groups (11, 46, 51).
In this study, we performed PCR-based screening of various species of planthoppers (Hemiptera: Fulgoroidea), spider mites (Acari: Tetranychidae), and Culicoides biting midges (Diptera: Ceratopogonidae) for Cardinium infection by using primers that detect bacteria closely related to Cardinium. The frequencies of Cardinium infection were considerably higher in planthoppers and spider mites. In Culicoides biting midges, which are important vectors of arthropod-borne viruses pathogenic to livestock (27), some species were infected with Cardinium-like bacteria that had lower nucleotide sequence similarity to other Cardinium species, including those previously found in arthropods. Morphological characteristics and molecular phylogenetic analyses of these bacteria are reported, and their taxonomic classification is reconsidered.
|
|
|---|
PCR screening for Cardinium and Wolbachia infection.
Based on the alignment of 16S rRNA gene sequences in the GenBank database of Cardinium obtained from 35 arthropod species, a set of primers, i.e., Car-sp-F (5'-CGG CTT ATT AAG TCA GTT GTG AAA TCC TAG-3') and Car-sp-R (5'-TCC TTC CTC CCG CTT ACA CG-3'), which amplify a 544-bp product was designed for highly conserved regions. To detect Wolbachia infection, a Wolbachia-specific primer pair that amplifies an approximately 900-bp DNA fragment of the 16S rRNA gene (32) was used.
PCR amplification was performed in 20 µl of reaction buffer containing 1x buffer, 0.16 mM of each deoxynucleoside triphosphate, 0.5 mM of each primer, 0.5 U of Taq DNA polymerase (Takara Bio Inc., Tokyo, Japan), and 2 µl of template DNA. DNA of Cardinium from Ixodes scapularis (22) was used as a positive control. The PCR conditions were as follows: 1 cycle of 95°C for 1 min; 35 cycles of 95°C for 30 s, 57°C (except 52°C for Culicoides) for 30 s, and 72°C for 1 min; and final extension at 72°C for 5 min. In order to validate the prepared templates, the mitochondrial cytochrome oxidase 1 (cox1) gene region was amplified in planthoppers (using primers COI-F [5'-ATA GCN TTY CCW CGA ATA AAT AAY ATA AGA TTY TG-3'] and COI-R [5'-GTT GRG GAA ARA AKG TTA AAT TTA CNC CNR NRA ATA-3']) and in spider mites (29). The cox1 and cox2 regions were amplified in biting midges (26). All PCR products of the 16S rRNA gene from Cardinium-positive samples were cloned, and at least three clones were sequenced.
Sequencing.
Typically, the full-length bacterial 16S rRNA gene (approximately 1,444 to 1,448 bp) was amplified from the DNA in Cardinium-positive samples by performing PCR using the fD1 and rP2 primers (47), and the PCR products were cloned and sequenced (26). A fragment of the Cardinium gyrase B subunit (gyrB) gene (1,267 bp) from planthoppers was amplified by using the primer pair Car gyrB2F (5'-GGK GTY TCB TGT GTA AAT GC-3') and Car gyrB2R (5'-TAS TGY TCT TCT TTR TCT CG-3'). A fragment of the Cardinium gyrB gene (1,389 bp) from spider mites was amplified using the primer pair Car gyrB1F (5'-CAA AGA YAC CTA TAA RAT TTC TG-3') and Car gyrB1R (5'-GTA ACG TTG TAC ARA KAC RGC AT-3'). Fragments of the Cardinium gyrB genes from biting midges and A. asiaticus (1,432 bp and 1,432 bp, respectively) were amplified using the primer pair UP-21 and 2Tr-SR1 (41). The temperature profile of the PCR was 1 cycle of 95°C for 1 min; 35 cycles of 95°C for 30 s, 52°C for 30 s (or 1 min), and 72°C for 1.5 min; and a final extension at 72°C for 5 min. Five clones per sample were usually sequenced, using the 3700 DNA analyzer (Applied Biosystems, Foster City, CA).
Phylogenetic analyses.
The phylogenetic relationship of Cardinium bacteria from planthoppers, spider mites, and biting midges to related bacteria was analyzed using the nucleotide sequences of the 16S rRNA and gyrB genes. A. asiaticus was used as an outgroup. The sequences were aligned using Clustal X software (44). Gaps were treated as a fifth state. Neighbor-joining (NJ) analyses for 16S rRNA and gyrB genes were conducted using PAUP* version 4.0b10 software (42) under the Hasegawa-Kishino-Yano 85 distance model. Bootstrap replicates were calculated 1,000 times. Maximum-likelihood (ML) analyses were also conducted for the 16S rRNA gene. MODELTEST 3.7 software (34) was used to estimate the most appropriate substitution model for ML analyses. The Akaike information criterion model chosen for the 16S rRNA gene was general time reversible (GTR) + G (gamma distribution shape parameter = 0.6597) + I (including invariable site), Pinvr = 0.6157. Bootstrap pseudoreplications were performed 500 times.
Analysis of nucleotide variation.
Numbers of polymorphic sites, nucleotide differences (k) (43), and nucleotide diversity (Pi) (30) within and among Cardinium groups were examined for a 1,442-bp sequence of the 16S rRNA gene and a 858-bp sequence of the gyrB gene by using DnaSP 4.5 software (36). The sequences were identical to those used in phylogenetic analyses, except for Ixodes Cardinium, which was excluded from the analysis of nucleotide variation because the sequence contained an unknown base. For the analysis of the gyrB gene, only one of the three sequences of Tetranychus urticae (red form A) was used.
Electron microscopy.
The morphology of Cardinium was observed in the planthopper Sogatella furcifera; three species of spider mites, Eotetranychus suginamensis, Panonychus mori, and Tetranychus pueraricola; and the biting midge Culicoides lungchiensis by transmission electron microscopy. Ovaries of the planthoppers and whole bodies of the female spider mites and biting midges were fixed with a solution of 0.8% glutaraldehyde (Nakalai Tesque, Kyoto, Japan) and 1% paraformaldehyde in 0.06 M phosphate buffer for 1 to 2 h on ice. After brief rinsing with the same buffer, postfixation was performed with 2% osmium tetroxide for 1 h at room temperature, followed by dehydration with a graded ethanol series. The samples were replaced into propylene oxide and embedded in Spurr resin (Polysciences, Warrington, PA). Semithin sections were stained with 2% uranyl acetate and Sato's lead solution and observed under a JEM-1010 transmission electron microscope (JEOL, Tokyo).
|
|
|---|
Cardinium infections were also widespread in spider mites. Out of 22 species (30 strains), 9 (11 strains) were positive for Cardinium infection (40.9% of the species and 36.7% of the strains) (see Table S2 in the supplemental material). In contrast, Wolbachia bacteria were detected in seven species (nine strains) (31.8% of the species and 30.0% of the strains). Three species (three strains) of spider mites were doubly infected with Cardinium and Wolbachia (13.6% of the species and 10.0% of the strains).
In Culicoides biting midges, 4 out of 25 species were positive (16.0%) (see Table S3 in the supplemental material). On the other hand, only C. paraflavescens out of 25 species was infected with Wolbachia (4.0%), but this species was negative for Cardinium. Among four Cardinium-positive species, three (C. arakawae, C. ohmorii, and C. peregrinus) had high infection frequencies, since all the examined individuals were positive (16/16, 4/4, and 3/3, respectively). In contrast, in the case of C. lungchiensis, only 2 out of 10 individuals were positive for Cardinium infection.
16S rRNA gene.
The bacterial 16S rRNA genes derived from three planthoppers (Euides speciosa, S. furcifera, and Harmalia sirokata) and four biting midges (C. arakawae, C. lungchiensis, C. ohmorii, and C. peregrinus) were sequenced. These sequences were analyzed together with the previously published sequences of Cardinium 16S rRNA genes derived from five spider mites (Amphitetranychus quercivorus, E. suginamensis, Oligonychus ilicis, T. pueraricola, and T. urticae) (14). Four features, namely, the size (between fD1 and rP2 primers), G+C content, nucleotide similarity to the type strain C. hertigii, and identity to unique Cardinium sequences previously published by Zchori-Fein et al. (52), were examined for these sequences. First, the size of the 16S rRNA gene was 1,447 bp in planthoppers and spider mites, which was the same as that of 16S rRNA genes of Cardinium bacteria derived from other arthropods. Conversely, all the sequences of the bacteria derived from Culicoides were 1 bp larger (1,448 bp) than those from other Cardinium bacteria. The 16S rRNA gene of A. asiaticus was 1,444 bp, which was 4 and 3 bp shorter, respectively, than those of the Cardinium bacteria derived from Culicoides and from other arthropods. Second, bacterial 16S rRNA genes derived from planthoppers, spider mites, and biting midges consistently exhibited a G+C content of 49 mol%, which was exactly the same level for all Cardinium bacteria previously found in arthropods. Third, the nucleotide similarity shared with C. hertigii was 96% in planthoppers, 96% in spider mites (except for 97% in T. urticae), and 94% in biting midges. Fourth, these sequences were examined for identity with two unique Cardinium sequences, GCG GTG TAA AAT GAG CGTG and GGT CTT TAA CTG ACG CT, which are shared by Cardinium bacteria derived from Encarsia wasps, Brevipalpus mites, and Ixodes ticks (52). The bacteria detected from planthoppers and spider mites had identities in both sequences, but bacteria detected from Culicoides had nucleotide substitutions in both sequences; they exhibited 1- or 2-bp substitutions in the former sequence and 1-bp substitutions in the latter sequence. Moreover, bacteria found from Culicoides did not have either of the two sequences (CGC GAT ACA ATC GTG TGT GTCC and CAT CCC TAG AAA TAG GGA GTT CCG AAA) that are unique to Paenicardinium (31). They exhibited 9- or 10-bp substitutions in the former sequence and 12-bp substitutions in the latter sequence. Based on the differences in 16S rRNA gene sequence, Cardinium-like bacteria can be categorized into three groups: group A (previously named Cardinium), which has been found widely in arthropods; group B (previously named Paenicardinium), which has been found in plant-parasitic nematodes; and group C, which has been found in Culicoides biting midges.
gyrB gene.
The fragments of the bacterial gyrB gene (858 bp) derived from planthoppers (E. speciosa and Indozuriel dantur), spider mites (A. quercivorus, E. suginamensis, O. ilicis, T. pueraricola, and T. urticae), and biting midges (C. arakawae and C. ohmorii) were aligned with that of C. hertigii. The G+C content was 41 mol% in both planthoppers, 40 to 42 mol% in the five spider mites and 40 mol% in the two biting midges. Nucleotide similarities shared with C. hertigii were 85% in planthoppers, 84% in spider mites (except for 87% in T. urticae), and 79% in biting midges.
Nucleotide variations.
Molecular variation within sets of Cardinium-like bacteria containing group A bacteria, group B bacteria, and group C bacteria in various combinations was examined for 16S rRNA (1,442 bp) and gyrB (858 bp) gene sequences (see Table S4 in the supplemental material). Nucleotide variation in the 16S rRNA gene was highest for both nucleotide differences (k, 43.93) and nucleotide diversity (Pi, 0.03046) in the set that contained group A and C bacteria. In the gyrB gene, k was highest in the set that contained group A and B bacteria (216.70), whereas Pi was highest in the set that contained group B and C bacteria (0.19075).
Phylogenetic analyses based on molecular data.
In order to elucidate the phylogenetic relationships of group A, B, and C bacteria, phylogenetic analyses were performed based on the nucleotide sequences of 16S rRNA genes (1,447 or 1,448 bp) (see Fig. S1 in the supplemental material) and the amino acid sequence of GyrB (286 amino acids [aa]). In the NJ tree based on 16S rRNA gene sequences, group C bacteria formed a monophyletic group (100% bootstrap support) (Fig. 1a). In the group A clade, bacteria derived from planthoppers formed a monophyletic group (100% bootstrap support), and bacteria derived from spider mites (except for T. urticae) formed a monophyletic group (80% bootstrap support). The ML tree of the 16S rRNA gene also supported the notion that Cardinium-like bacteria can be divided into three groups (Fig. 1b). Phylogenetic relationships within group A were similar to those for the NJ tree.
![]() View larger version (46K): [in a new window] |
FIG. 1. Phylogenetic analyses based on nucleotide sequences of Cardinium 16S rRNA genes (1,444 to 1,448 bp). (a) NJ tree. (b) ML tree. Scientific names of host species are shown at the terminal nodes. A. asiaticus is defined as an outgroup. The numbers above the branches indicate bootstrap probabilities (percent). GenBank accession numbers are in parentheses. The examined species are denoted by asterisks.
|
![]() View larger version (23K): [in a new window] |
FIG. 2. NJ tree based on the amino acid sequences of Cardinium GyrB (286 aa). Scientific names of host species are shown at the terminal nodes. A. asiaticus is defined as an outgroup. The numbers above the branches indicate bootstrap probabilities (percent). GenBank accession numbers are in parentheses. The sequences of A. quercivorus and three forms of T. uriticae were determined from the samples in a previous study (14). The examined species are denoted by asterisks.
|
Morphological characterization of Cardinium.
Cardinium-like bacteria (n = 32) were observed in the planthopper S. furcifera; the three spider mites E. suginamensis (n = 24), P. mori (n = 7), and T. pueraricola (n = 67); and the biting midge C. lungchiensis (n = 106). The bacteria were circular or rod shaped (major axis, 0.3 to 2.1 µm [n = 236; average ± standard deviation = 0.64 ± 0.32 µm]; minor axis, 0.2 to 0.87 µm [n = 236; average ± standard deviation = 0.37 ± 0.08 µm]). Microfilament-like structures, previously found in Cardinium and Paenicardinium (6, 22, 31, 37, 38, 52), were observed consistently in Cardinium-like bacteria inhabiting the five arthropod species (Fig. 3). Microfilament-like structures are considered to be a characteristic of Cardinium-like bacteria that are unlikely to be present in the close relative "Candidatus Amoebophilus asiaticus" (17). Microfilament-like structures of these bacteria were attached to the plasma membrane and extended vertically from an electron-dense plaque.
![]() View larger version (101K): [in a new window] |
FIG. 3. Electron micrographs of Cardinium-like bacteria. (a) Follicle cell of the planthopper S. furcifera. (b) Oocyte of the spider mite E. suginamensis. (c) Follicle cell of the spider mite P. mori. (d) Epidermal cell of the spider mite T. pueraricola. (e and f) Oocyte and sheath, respectively, of the biting midge C. lungchiensis. Cardinium cells in anaplastic sheaths are indicated by arrows. Bars, 0.20 µm.
|
|
|
|---|
![]() View larger version (25K): [in a new window] |
FIG. 4. Relationship of Cardinium groups A, B, and C. Each rectangular box represents a Cardinium group. Identities of 16S rRNA gene nucleotide sequences (1,447 or 1,448 bp) and in GyrB amino acid sequences (286 aa) are indicated on the lines which connect the groups. The nucleotide or amino acid identities were calculated with Genetyx version 9.1 software (Genetyx Corporation, Tokyo).
|
Planthoppers and spider mites were also infected with Wolbachia at high frequencies, with 52.7% and 31.8% of the examined species, respectively, being infected. Double infections with Cardinium and Wolbachia in single individuals were found to be common; 32.7% of 55 planthopper species and 13.6% of 22 spider mite species were doubly infected. In planthoppers, 18 strains were doubly infected with Cardinium and Wolbachia, 7 strains were singly infected with Cardinium, 12 strains were singly infected with Wolbachia, and 20 strains were uninfected. Fisher's exact probability test showed that significantly higher numbers of strains were doubly infected (P < 0.05), which suggests mutualistic dependency of the two bacterial species in planthoppers. In spider mites, on the other hand, the numbers of strains doubly infected with Cardinium and Wolbachia were not significantly large (Fisher's exact probability test).
In contrast to the case for planthoppers and biting midges, Cardinium sequences from several spider mites were poorly amplified by PCR even after 35 cycles. Additionally, different PCR primers showed different positive and negative results in some samples, for example, between the primer pairs CFBf1/CFBr1 (28) and Car-sp-F/Car-sp-R in this study. The inconsistent results between primers probably reflect low infection titers of Cardinium. Extremely low infection titers of Wolbachia were recently reported for natural populations of the bark beetle Pityogenes chalcographus (49). In Drosophila, the severity of CI is associated with the density of Wolbachia bacteria (7). It is possible that, in some species of spider mites, Cardinium bacteria have the potential to cause CI but cannot do so since they are unable to proliferate to titers sufficient to cause strong CI.
In biting midges, a novel group of bacteria that we described as members of the genus Cardinium was found. The frequency of Cardinium infection among biting midges was 16.0% (n = 25). All of the PCR primers used for screening of Cardinium infection in earlier studies, i.e., CLOf1 and CLOr1 (46), ChF and ChR (51), and CFBf1 and CFBr1 (28), could amplify PCR products from the samples, which were proven positive for Cardinium infection (data not shown). The fact that this group of Cardinium bacteria had not been found in previous screenings strongly suggests that screening of a greater number of arthropod species will be required for a more complete understanding of the phylogenetic diversity of Cardinium.
Morphological observation.
Cardinium bacteria derived from C. lungchiensis had a common type of microfilament-like structure, i.e., straight, parallel, and perpendicular tubules with tips attached to the long dimension of the bacterial cell wall. Although the function of the microfilament-like structure remains unknown, this Cardinium-specific structure in bacterial cells is likely to play some important role in the survival of the bacteria. Cardinium was not observed in immature ovaries but was present in anaplastic sheaths adjacent to immature ovaries in young adult females of the biting midge C. lungchiensis. Cardinium bacteria have also been observed in terminal filaments and follicle cells. In every generation, Cardinium may be vertically transmitted during oogenesis by migrating from surrounding tissues into immature oocytes. Detailed systematic observations would be necessary to elucidate the mode of vertical transmission of Cardinium in future studies.
Nomenclature of Cardinium.
We proposed a single species name, "Candidatus Cardinium hertigii," for three groups of bacteria, namely, the previously described "Candidatus Cardinium hertigii" and "Candidatus Paenicardinium endonii" and the bacteria derived from biting midges found in the present study (Fig. 4). It is generally assumed that a 3% substitution in the nucleotide sequence of the 16S rRNA is necessary to assign bacteria to a different species (1, 39). Particularly in the case of insect endosymbionts, bacteria having even more substitutions are conventionally assigned to a single species. For instance, Wolbachia pipientis, derived from Trichogramma cordubensis (group B) (40) and Difilaria repens (group C) (9), exhibited 7.2% substitution in 16S rRNA gene (1,395 bp). Likewise, "Buchnera aphidicola," a mutualistic endosymbiont of aphids, exhibits 9.3% substitution in the 16S-23S rRNA gene region, and Carsonella, Portiera, and Termblaya exhibit 8% to 10% differences in the 16S-23S rRNA gene region (5). At present, Wolbachia bacteria with more than 3% substitution in the 16S rRNA gene are assigned to different phylogenetic supergroups (23).
In Cardinium, the largest substitution in the 16S rRNA gene (1,447 bp) was 7.2% between Cardinium bacteria of E. pergandiella (50) and Paenicardinium (31), which is a value comparable to that for other insect symbionts. We therefore consider our taxonomic revision quite reasonable. Taking the data together, practical conditions for identifying Cardinium could be (i) the presence of microfilament-like structures and (ii) less than 7% to 10% substitution in 16S rRNA gene. Since every Cardinium organism found so far was present in ovaries (6, 31, 37, 38, 52), transovarial transmission could be included as a third condition.
Nomenclature of the symbiotic bacteria is still controversial. Considering the complex relationship and extensive recombination between various strains of Wolbachia, Lo et al. (23) have argued for the single species Wolbachia pipientis. Conversely, Pfarr et al. (33) claimed that Wolbachia should be divided into two species according to the host (arthropods versus filarial nematodes). In any case, a better understanding of their nature derived from further studies will allow more convincing classification of these symbionts.
This work is supported in part by Grants-in-Aid for Scientific Research (no. 19380038 and 19380032) from the Japan Society for the Promotion of Science.
Published ahead of print on 4 September 2009. ![]()
Supplemental material for this article may be found at http://aem.asm.org/. ![]()
The authors have paid a fee to allow immediate free access to this article. ![]()
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»