|Home | About | Journals | Submit | Contact Us | Français|
We investigated the vertical transmission, reproductive phenotype, and infection density of a male-killing Spiroplasma symbiont in two Drosophila species under physiological high and low temperatures through successive host generations. In both the native host Drosophila nebulosa and the nonnative host Drosophila melanogaster, the symbiont infection and the male-killing phenotype were stably maintained at 25°C, rapidly lost at 18°C, and gradually lost at 28°C. In the nonnative host, both the high and low temperatures significantly suppressed the infection density of the spiroplasma. In the native host, by contrast, the low temperature suppressed the infection density of the spiroplasma whereas the high temperature had little effect on the infection density. These results suggested that the low temperature suppresses both the infection density and the vertical transmission of the spiroplasma whereas the high temperature suppresses the vertical transmission preferentially. The spiroplasma density was consistently higher in the native host than in the nonnative host, suggesting that the host genotype may affect the infection density of the symbiont. The temperature- and genotype-dependent instability of the symbiont infection highlights a complex genotype-by-genotype-by-environment interaction and may be relevant to the low infection frequencies of the male-killing spiroplasmas in natural Drosophila populations.
Endosymbiotic bacteria are commonly found in a variety of insects (4). Some symbionts are mutualistic and contribute to the fitness of their hosts, other symbionts are rather parasitic and tend to affect their hosts negatively, and the majority of the others are of unknown nature and are recognized by means of microscopy, PCR detection, and/or 16S rRNA gene sequencing. Such unseen microbial associates often play important roles in various biological aspects of diverse insects (2, 3).
For maintenance in host populations, a maternally inherited symbiont must be capable of proliferating in the host and also be consistently transmitted through host generations. Hence, infection density and fidelity of vertical transmission are important parameters that govern the infection dynamics of a symbiont. Effects on the host fitness also contribute to the prevalence of the symbiont infection: positive and negative fitness effects may result in fixation and disappearance of the infection, respectively. In addition, symbiont-induced alteration of host reproduction, such as parthenogenesis, feminization, cytoplasmic incompatibility, and male-killing, can promote prevalence of the infection (35, 43). It should be noted that these parameters, infection density, vertical transmission, fitness effect, and reproductive phenotype, are often interrelated. For example, reduced infection density may result in imperfect vertical transmission and attenuated phenotypic effects, which can lead to loss of the infection. Excessive infection density may lead to enhanced phenotypic effects on the host, causing pathology at an extreme, which can significantly influence the host fitness (28, 31).
Environmental factors such as ambient temperature affect host-symbiont interactions in a variety of ways. High temperatures have been reported to disrupt insect-microbe endosymbiosis and result in loss of symbiont infection in aphids (34), mealybugs (20), planthoppers (40), bedbugs (5), and other insects. Wolbachia infections in drosophilid flies are also known to be vulnerable to high temperatures (6, 16). In most of these studies, acute effects of heat stress, at 30°C or higher, were examined in the treated adult insects or in their immediate offspring. Few studies have systematically investigated the effects of different physiological temperatures on symbiont infection through successive host generations.
Members of the genus Spiroplasma are wall-less, helical, and actively motile bacteria of the class Mollicutes and are associated with a variety of arthropods and plants (7, 14). It is notable that some spiroplasmas cause female-biased sex ratios of their host insects, such as drosophilid flies, ladybird beetles, and butterflies, as a result of selective death of male offspring during embryogenesis (15, 17). Among drosophilid flies, male-killing spiroplasmas have been identified from Drosophila willistoni, Drosophila nebulosa, Drosophila equinoxialis, Drosophila paulistorum, and Drosophila melanogaster, and are often called by the acronyms WSRO (from willistoni sex ratio organism), NSRO, ESRO, PSRO, and MSRO, respectively (30, 44).
Thus far, only a few studies have reported loss of symbiont infection at low temperatures, most of which are from Drosophila-Spiroplasma associations. In D. melanogaster, a low temperature at 16.5°C cured the infection with male-killing spiroplasma MSRO, while higher temperatures at 20°C, 24°C, and 28°C maintained extremely female-biased sex ratios due to the symbiont infection over two generations of the host (29). Male-killing spiroplasma NSRO in D. melanogaster was stable at 25°C but easily lost at 18°C (13). When male-killing spiroplasma WSRO was transfected from D. willistoni to D. melanogaster, the expression of the male-killing phenotype was stronger at 25°C than at 20°C (8).
Meanwhile, it has been reported that high temperatures also destabilize the Drosophila-Spiroplasma associations. In D. equinoxialis, heat treatment of embryos cured female-biased sex ratios caused by spiroplasma strain ESRO, whereas the same heat treatment failed to disrupt the male-killing caused by WSRO in D. willistoni (25). Hence, it was expected that the Drosophila-Spiroplasma associations might provide a unique system wherein both high and low temperatures affect the insect-microbe symbiosis. It is of great interest whether high and low temperatures differently affect the symbiont infection, and if so, how.
In this study, we investigated the vertical transmission, reproductive phenotype, and infection density of the spiroplasma strain NSRO in two Drosophila species under physiological high- and low-temperature conditions through successive host generations. The results unveiled previously unknown aspects of temperature effects on proliferation and transmission of the bacterial symbiont in the insect host: low temperature suppressed both infection density and vertical transmission whereas high temperature suppressed vertical transmission preferentially.
In this study, two strains of drosophilid flies were used: strain ORNSRO of D. melanogaster and strain G87 of D. nebulosa, both of which are infected with a male-killing spiroplasma designated as NSRO and exhibit an almost complete male-killing phenotype. The flies were reared with a standard cornmeal medium in plastic bottles at 25°C under a long-day regimen (16 h of light, 8 h of dark) unless otherwise described. For maintenance of the fly strains that produce females only, males were supplied from corresponding uninfected fly stocks that had been generated by a low-temperature treatment.
In the 1960s, an Oregon-R strain of D. melanogaster, which had initially been free of spiroplasma infection, was injected with hemolymph of the tropical drosophilid fly D. nebulosa that harbored the male-killing spiroplasma NSRO. The transfected fly strain exhibited a strong male-killing phenotype and has been maintained in the laboratory for over 40 years (36, 37). In 2000, hemolymph of adult females of the fly strain was injected into a fresh, uninfected Oregon-R strain of D. melanogaster, whereby an isofemale strain, named ORNSRO, was established and has been maintained in our laboratory (1).
A fly stock of D. nebulosa, which had been collected at Guadeloupe, Brazil, in 2001, was kindly provided by Gregory D. D. Hurst (University of Liverpool). An isofemale line, named G87, was established from the stock, which was naturally infected with the spiroplasma NSRO and produced all-female offspring. Note that the spiroplasma in the D. melanogaster strain ORNSRO is certainly derived from D. nebulosa but that the source population is different from that for the D. nebulosa strain G87.
Total DNA samples were prepared from the hemolymph of flies of the strains ORNSRO and G87. The following bacterial genes were amplified by PCR: the 16S rRNA gene, encoding a small subunit of rRNA, with the primers 16SA1 (5′-AGA GTT TGA TCM TGG CTC AG-3′) and 16SB1 (5′-TAC GGY TAC CTT GTT ACG ACT T-3′) (12); p58, encoding an adhesin-like protein, with the primers p58-f (5′-GTT GGT TGA ATA ATA TCT GTT G-3′) and p58-r (5′-GAT GGT GCT AAA TTA TAT TGA C-3′) (30); spoT, encoding a guanosine-3′,5′-bis(diphosphate)3′-pyrophosphohydrolase, with the primers spoT-f (5′-CAA ACA AAA GGA CAA ATT GAA G-3′) and spoT-r (5′-CAC TGA AGC GTT TAA ATG AC-3′) (30); and dnaA, encoding a chromosomal replication initiator protein, with the primers SRdnaAF1 (5′-GGA GAY TCT GGA YTA GGA AA-3′) and SRdnaAR1 (5′-CCY TCT AWY TTT CTR ACA TCA-3′) (1). The PCR products were electrophoresed in agarose gels, excised and purified by using a Geneclean Spin kit (Qbiogene), and cloned with a TA cloning vector (pT7Blue [Novagen, Darmstadt, Germany]) and Escherichia coli DH5α competent cells (Takara, Japan). Cloned plasmids were purified by using a QIAprep Spin Miniprep kit (Qiagen, Japan) and subjected to DNA sequencing with a BigDye terminator v.3.1 cycle sequencing kit and an ABI-3130xl sequencer (Applied Biosystems, Japan) as previously described (1).
Nucleotide sequences were aligned by using the program package Clustal X version 2.0 (24) and were then realigned manually. Neighbor-joining trees were constructed by using Clustal X with Kimura's two-parameter model (19). Bootstrap resampling was performed with 1,000 replicates. Maximum-parsimony trees and maximum-likelihood trees were constructed by using the program PAUP* version 4.0b10 (41).
In order to establish experimental fly lines, 7-day-old females were collected from the spiroplasma-infected fly strains ORNSRO and G87, which had been maintained at 25°C. Of the infected virgin females, 15 insects were placed in each of the rearing bottles with seven males from the uninfected fly stocks and were maintained at either 18°C, 25°C, or 28°C. Initially, the following numbers of fly bottles were generated: for ORNSRO, 12 lines at 18°C, 10 lines at 25°C, and 11 lines at 28°C; for G87, 27 lines at 18°C, 15 lines at 25°C, and 20 lines at 28°C. From each of the bottles, adult insects were harvested for 1 to 3 days after the onset of adult emergence. All harvested adult insects were counted and sexed under a binocular microscope for monitoring of male-killing expression. From the collected flies, 15 females were taken for initiating the next generation, and the remaining flies were preserved in acetone for subsequent molecular analyses (11). At 3 to 4 days after the collection, the 15 females were crossed with seven males. Seven days after the collection, these flies were transferred to a new bottle and allowed to lay eggs for 4 days. The intervals between the start of oviposition in the new bottle and the collection of adult flies of the next generation were 23 to 26 days for ORNSRO and 25 to 30 days for G87 at 18°C, 11 to 14 days for both ORNSRO and G87 at 25°C, and 11 to 13 days for both ORNSRO and G87 at 28°C. In this way, these insect lines were maintained at different temperatures through six to eight generations. The fly lines that died out during the experiment were discarded. Consequently, the following numbers of insect lines persisted throughout the experimental period and were subjected to subsequent diagnostic PCR and sex ratio analyses: for ORNSRO, six lines at 18°C, nine lines at 25°C, and five lines at 28°C; for G87, 17 lines at 18°C, 13 lines at 25°C, and 8 lines at 28°C.
For DNA extraction, 10 acetone-preserved adult females per line per generation were individually homogenized in 100 μl of a squishing buffer (10 mM Tris-HCl [pH 8.0], 1 mM EDTA, 25 mM NaCl). After 200 μg of proteinase K was added, the samples were incubated at 55°C for 60 min, heated at 95°C for 10 min, and centrifuged at 13,000 rpm for 1 min. The supernatants were subjected to diagnostic PCR detection of the spiroplasma infection by using the Sybr green and Mx3000P QPCR system (Stratagene, La Jolla, CA) under the “allele discrimination/SNP's real-time mode.” The primers dnaA109F and dnaA246R (1) were used for detection of dnaA of the spiroplasma. The reaction mixture consisted of 1× AmpliTaq Gold buffer; 1.5 mM MgCl2; 0.2 mM each of dATP, dGTP, dCTP, and dUTP; 0.3 μM each of the forward and reverse primers; 1/100,000 Sybr green; and 0.02 U/μl AmpliTaq Gold DNA polymerase (Applied Biosystems). PCR was performed under a temperature profile of 95°C for 10 min followed by 35 cycles of 95°C for 30 s, 58°C for 20 s, and 72°C for 20 s and one cycle of 95°C for 1 min, 55°C for 30 s, and 95°C for 30 s. The infection rate per line per generation was calculated as 100 × (number of infected females)/(number of all females examined) (%).
All collected adult insects per line per generation were counted and sexed as mentioned above. When the number of males was less than or the same as that of females, the intensity of male-killing per line per generation was calculated as 100 × [(number of females) − (number of males)]/(number of females) (%), which produces values ranging from 0% for a 1:1 sex ratio to 100% for all females. When the number of males was more than that of females, the intensity of male killing was regarded as 0%.
The insect samples preserved in acetone were air dried, weighed, and subjected to DNA extraction by using a QIAamp DNA mini kit (Qiagen). The purified DNA from a sample was eluted with 200 μl of AE buffer supplied in the kit. Real-time fluorescence detection quantitative PCR was performed by using the TaqMan PCR and Mx3000P QPCR system. The dnaA gene copies of the spiroplasma were quantified by using the probe dnaA180T and the primers dnaA109F and dnaA246R as previously described (1). The reaction mixture consisted of 1× ABsolute QPCR ROX mix (ABgene, Surrey, United Kingdom), 0.4 μM each of the forward and reverse primers, and 0.2 μM of the probe. PCR was performed under a temperature profile of 95°C for 15 min followed by 40 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 30 s. Standard curves were drawn by using standard PCR product samples of the dnaA gene at concentrations of 102, 103, 104, 105, 106, 107, and 108 copies/4 μl. Relative symbiont density was evaluated in terms of dnaA gene copies per μg of insect dry weight.
All statistical analyses were conducted by using the software R version 2.2.0 (38). Multiple comparisons of spiroplasma densities were performed with Bonferroni corrections. Since some of the data sets did not satisfy normal distribution and/or homogeneous variance, we adopted the generalized linear model (27) with a normal, gamma, or negative binominal distribution, which was selected according to the Akaike information criterion.
The nucleotide sequences reported in this paper have been deposited in the DDBJ, EMBL, and GenBank nucleotide sequence databases under accession numbers AB089822, AB253400, and AB434486 to AB434491.
The nucleotide sequences of the bacterial 16S rRNA, p58, spoT, and dnaA genes were completely identical between the fly strains ORNSRO and G87. Molecular phylogenetic analyses unequivocally showed that the spiroplasma sequences from ORNSRO and G87 are the most closely related among the spiroplasma strains ever reported (Fig. (Fig.11).
At 25°C, the normal rearing temperature for the host insects, spiroplasma infections were stably maintained through successive host generations. In the eighth generation, all nine strains maintained 100% infection frequencies in ORNSRO and 11 of 13 strains retained over 80% infection frequencies while one strain lost infection in G87 (Fig. 2C and D). At a lower temperature (18°C), strikingly, all strains of both ORNSRO and G87 lost infection in the second generation (Fig. 2A and B). At a higher temperature (28°C), the spiroplasma infections were gradually lost through successive host generations. In the eighth generation, only one of five strains of ORNSRO and one of eight strains of G87 retained high infection frequencies (Fig. 2E and F).
At 25°C, production of almost all-female offspring due to symbiont-induced male killing was consistently observed through successive host generations. In the eighth generation, all nine strains produced all-female broods in ORNSRO and 9 of 13 strains exhibited highly female-biased sex ratios while one strain restored the normal 1:1 sex ratio in G87 (Fig. 3C and D). At 18°C, in both ORNSRO and G87, production of nearly all-female offspring was observed only in the first generation and disappeared from the second generation and onward (Fig. 3A and B). At 28°C, the female-biased sex ratios gradually diminished through successive host generations. In the eighth generation, only one of five strains of ORNSRO and one of eight strains of G87 exhibited strongly female-biased sex ratios (Fig. 3E and F).
In the first generation of ORNSRO, infection densities at 18°C were significantly lower than those at 25°C. In the first and successive generations, infection densities at 28°C were generally lower than those at 25°C (Fig. (Fig.4A).4A). In the first generation of G87, infection density at 18°C was significantly lower than that at 25°C, while infection density at 28°C was almost the same as that at 25°C. In successive generations, infection densities at 28°C were consistently at levels similar to those at 25°C (Fig. (Fig.4B4B).
Across almost all temperature conditions and host generations (except for the eighth generation at 25°C), infection densities were significantly higher in G87 than in ORNSRO (see asterisks below the columns of Fig. Fig.4A4A).
Figure Figure55 shows the relationships between infection densities and infection frequencies in ORNSRO and G87 under different temperature conditions. At 25°C, infection densities were almost constant in both ORNSRO and G87 irrespective of infection frequencies at the sampling points (Fig. 5C and D). At 18°C, infection densities were consistently low in both ORNSRO and G87 irrespective of infection frequencies at the sampling points (Fig. 5A and B). At 28°C, infection densities were, although showing large variances, largely constant in both ORNSRO and G87 irrespective of infection frequencies (Fig. 5E and F). Regression analyses revealed that the infection densities were not significantly correlated with the infection frequencies in both insect strains maintained under different temperature conditions (linear models; P > 0.05). Neither infection densities nor infection frequencies exhibited an obvious relationship to host generations G1, G2, G3, and G8 (Fig. (Fig.55).
At a lower temperature (18°C), infection densities were significantly lowered (Fig. (Fig.44 and 5A and B). At a higher temperature (28°C), meanwhile, infection densities were either moderately lowered (in ORNSRO) or not affected (in G87) (Fig. (Fig.44 and 5E and F). These results suggest that the low- and high-temperature conditions interfered with the symbiont infection and transmission in different ways. Namely, in both ORNSRO and G87, the low-temperature condition suppressed the symbiont density in the host body, possibly by deterring the symbiont proliferation, and the lowered infection density probably resulted in loss of the infection. Meanwhile, at least in G87 and possibly also in ORNSRO, the high-temperature condition had little effect on the symbiont density in the host body (Fig. (Fig.4),4), but the symbiont infection was lost in the subsequent generations (Fig. (Fig.2).2). Therefore, it is conjectured that the high-temperature condition preferentially affected the symbiont transmission through host generations, with lesser effects on the symbiont proliferation.
How the high-temperature condition selectively suppresses the vertical transmission of the symbiont is currently unknown. High temperature might alter either tissue tropism or motility of the spiroplasmas, thereby resulting in less-efficient vertical transmission. In eukaryotes, high temperature facilitates depolymerization of some motor proteins such as tubulin and actin (10). In spiroplasmas, high temperature was reported to suppress the cell motility (22). Alternatively, high temperature might directly interfere with the mechanism or process involved in the vertical transmission of the symbiont. The temperature-dependent differential effects on proliferation and transmission of the spiroplasmas are of great interest because they would provide clues to understanding mechanisms of control over the symbiotic system.
Previous studies have demonstrated that, in various insect-microbe symbiotic systems, infection densities of the symbionts are affected by symbiont genotype, host genotype, coinfection by other symbionts, environmental factors, etc. (13, 23, 32, 33, 42). In this study, infection densities of the spiroplasma were consistently higher in G87 than in ORNSRO (Fig. (Fig.4).4). Moreover, responses of the infection densities to higher temperature also differed between G87 and ORNSRO: no effect in the former but significant suppression in the latter (Fig. (Fig.4).4). Considering that the spiroplasmas in G87 and ORNSRO are genetically very close to each other (Fig. (Fig.1),1), the differences in spiroplasma densities between the fly strains may be attributable to the host genotype rather than to the symbiont genotype, although the possibility cannot be ruled out that slight genetic differences between the symbionts somehow affected the observed differences in bacterial density. Note that the G87 insect is the native host for the spiroplasma, while the ORNSRO insect is the nonnative host generated by transfection with the spiroplasma from a different source (36, 37). Thus, it is conceivable that the infection densities of the spiroplasma might be negatively affected in the nonnative host ORNSRO. It is also notable that D. nebulosa is distributed in South American tropics whereas D. melanogaster is a cosmopolitan species found not only in tropical countries but also in temperate regions worldwide. Hence, it appears plausible that higher temperature little affected the infection density of the spiroplasma in the D. nebulosa strain G87, due to adaptation of the tropical species to high-temperature conditions.
In this study, infection densities of the spiroplasmas were evaluated by a quantitative PCR technique in terms of bacterial dnaA gene copies. It should be noted, however, that bacterial gene copy number might not always be in agreement with bacterial cell number. As known from insect endosymbionts like Buchnera in aphids, genomic DNA copies in a bacterial cell may be highly polyploid (21). Quantitative PCR detects not only DNA from live bacterial cells but also DNA from dead bacterial cells and cell-free bacterial DNA present in the host insect (18). Since conventional quantification methods like colony formation assay and direct cell counting are difficult to apply to the fastidious, tiny, and actively motile spiroplasma cells, we adopted the quantitative PCR approach as a practical and reliable means for quantification of the spiroplasmas.
The temperature conditions examined in this study, 18°C, 25°C, and 28°C, are within the range of natural conditions wherein both D. melanogaster and D. nebulosa are able to grow and reproduce. Notwithstanding this, the spiroplasma infection and transmission were significantly suppressed under the higher and lower, but ecologically realistic, temperatures. In natural Drosophila populations, infection frequencies of male-killing spiroplasmas were reported to be generally low (44), although the male-killing phenotype has been regarded as an evolutionarily adaptive trait for maternally transmitted symbionts on the grounds that infected females can gain an extra fitness due to reallocated resources from their killed brothers (15, 17). For example, in a natural population of D. nebulosa in Campo Grande, Brazil, frequencies of the male-killing phenotype due to the spiroplasma NSRO were from 3 to 6% depending on the season (26). In Brazilian natural populations of D. melanogaster, the frequency of infection with the male-killing spiroplasma MSRO was 2.3% (30). The temperature-dependent instability of vertical transmission may be relevant to the low infection frequencies of the male-killing spiroplasmas observed in natural Drosophila populations. In this study, a series of experiments were conducted under constant-temperature conditions. In an attempt to gain further insights into the infection dynamics of the symbiont in nature, experiments under fluctuating-temperature conditions will be of interest.
Recent studies have revealed that insect-microbe symbiotic systems often respond to environmental conditions like ambient temperature in an unpredictable manner, which results from complex interactions between host genotype, symbiont genotype, and environmental factors (9, 33, 39, 42). The Drosophila-Spiroplasma symbiosis, wherein the host insect is genetically manipulatable, the host-symbiont combinations are experimentally exchangeable, and the symbiont infection is easily recognizable by the remarkable male-killing phenotype, would provide an ideal model system for experimentally dissecting the relevant factors underpinning the complex nature of the symbiotic interactions.
We thank A. Sugimura, S. Tatsuno, S. Suo, H. Ohuchi, N. Totsuka, K. Nomura, and W. Kikuchi for technical and secretarial assistance and G. D. D. Hurst, T. Koana, and T. Murata for providing fly strains.
This research was supported by the Program for Promotion of Basic Research Activities for Innovation Biosciences (ProBRAIN) of the Bio-Oriented Technology Research Advancement Institution.
Published ahead of print on 15 August 2008.