Microb Ecol (2014) 67:671–678 DOI 10.1007/s00248-013-0359-0
INVERTEBRATE MICROBIOLOGY
Factors Affecting the Strength of Cardinium-Induced Cytoplasmic Incompatibility in the Parasitic Wasp Encarsia pergandiella (Hymenoptera: Aphelinidae) Steve J. Perlman & Nicolas J. Dowdy & Leanne R. Harris & Mahwish Khalid & Suzanne E. Kelly & Martha S. Hunter
Received: 13 August 2013 / Accepted: 19 December 2013 / Published online: 9 January 2014 # Springer Science+Business Media New York 2014
Abstract Bacteria that cause cytoplasmic incompatibility (CI) are among the most common maternally transmitted parasites of insects. In CI, uninfected females produce few or no offspring when they mate with infected males and, as a result, are often at a reproductive disadvantage relative to infected females. Two different bacteria are known to cause CI, Wolbachia and Cardinium. CI Cardinium was discovered more recently and has been little studied. Here, factors that could influence the reduction in reproductive output in a CI cross, or CI “strength,” were explored in the parasitic wasp Encarsia pergandiella. Cardinium in this wasp exhibits variable CI strength. Experiments tested the effect of male age, male size, male host species, Cardinium density, and male development time on CI strength. We found a striking effect of male development time, with males that took longer to develop exhibiting stronger CI when mated to uninfected females. Male age had little effect; although in one experiment, the oldest males exhibited stronger CI. Male size, host species, and bacterial density had no effect on the strength of CI. Identifying the factors that control CI are crucial for understanding the dynamics of infection, as well as the success of strategies that aim to use CI microbes to control insect pests and disease vectors. S. J. Perlman (*) : L. R. Harris Department of Biology, University of Victoria, Victoria, British Columbia V8P 5C2, Canada e-mail:
[email protected] N. J. Dowdy : M. Khalid : S. E. Kelly : M. S. Hunter Department of Entomology, The University of Arizona, 410 Forbes Building, Tucson, AZ 85721, USA N. J. Dowdy Department of Biology, Wake Forest University, Winston-Salem, NC 27106, USA
Introduction Bacteria that induce cytoplasmic incompatibility (CI) are among the most widespread parasites of insects and other terrestrial arthropods [45]. They are transmitted from mothers to their offspring, and consequently, their fitness is linked directly to the number of infected daughters produced by an infected female. In its simplest form, CI results when uninfected females mate with infected males, with this cross producing few or no viable offspring. Infected females, on the other hand, do not experience this reproductive loss and, as a result, are often able to rapidly replace uninfected females. For example, females infected with the CI-inducing bacterium Wolbachia were documented to spread in California Drosophila simulans populations at an astounding rate of 12 km per generation [41]. The population dynamics of CI microbes are expected to depend on three key factors: the strength of incompatibility, maternal transmission efficiency, and the fitness cost of the infection [9, 40]. Because of the often severe reduction in fitness of uninfected females, and the speed with which CI-inducing bacteria can drive through populations, there has been much recent interest in using CI to control arthropod pests and disease vectors. Control strategies have taken two forms. First, labreared males infected with an incompatible strain can be mass released in the wild to suppress wild populations [1, 30]; this is analogous to sterile male releases. Second, lab-reared females infected with an incompatible strain can be released in the wild, with the ultimate goal of replacing native uninfected hosts. For example, Hoffmann et al. [21] established a new Wolbachia infection in Aedes aegypti mosquitoes that causes strong CI and also interferes with the transmission of dengue virus. Infected mosquitoes were released in two Australian towns, where they spread and the infection reached near fixation. Understanding both the mechanism of CI, as well
672
as what factors control variability in CI strength (i.e., the reduction in offspring in the incompatible mating), will clearly impact the success of CI control initiatives. Thus far, only two bacteria have been found to be able to induce CI. Wolbachia, in the α proteobacteria, is the best known reproductive manipulator, estimated to infect ~40 % of all insect species [54], as well as other terrestrial arthropods and filarial nematodes. While the effects of Wolbachia infection are unknown for many hosts, this bacterium has been found to induce a wide variety of reproductive phenotypes, including CI [46]. The more recently discovered Cardinium is found in the phylum Bacteroidetes and infects approximately 6–7 % of arthropods [14, 44, 51]. Cardinium has been found in eight arthropod orders, as well as in plant-parasitic nematodes [8, 13, 15, 16, 31]. Like Wolbachia, the host effects of Cardinium infection are largely unknown; although it, too, has been found to induce several reproductive phenotypes. Cardinium has been shown to induce CI in the parasitoid wasp Encarsia pergandiella [25], in the spider mites Eotetranychus suginamensis [19], Bryobia sarothamni [39], Tetranychus cinnabarus [49], T. piercei [53], the predatory mite Metaseiulus occidentalis [48], and most recently, in the planthopper Sogatella furcifera [32, 52]. While the mechanism of CI is not fully understood, it is thought that the sperm of infected males is somehow modified within the testes, rendering it incompatible with eggs from uninfected females (or those that are not infected with the same CI strain); this modification is somehow rescued in infected females [45]. In Wolbachia, this incompatibility is thought to be caused by disruption of key cell cycle events affecting the male pronucleus [29]. No work has been done on the mechanism of CI in Cardinium. However, the genome of the CI-inducing strain of Cardinium in the parasitoid wasp E. pergandiella was recently sequenced [33]. It does not share genes with Wolbachia, demonstrating that CI evolved independently in these two bacteria. Many studies have documented extensive variation in the strength of CI Wolbachia, important in determining whether infections, including pest and vector control strategies that use CI, will invade and spread [40]. Many factors have been found to influence CI strength, including host [5] and bacterial genotype [43], temperature [10], development time [50], and bacterial density [6]. Because CI expression is intimately involved with sperm, much work has focused on male traits. Many studies have shown that CI is strongest in younger males [38, 42], although some studies have found no effect of age [37]. For example, Hoffmann et al. [22] found that 1day old male D. simulans infected with Wolbachia produced 0.2 to 1.4 % of the number of offspring of the compatible cross; this number increased to 74 to 77 % when males were 21–24 days old. It is thought that CI strength declines in old males because of decreased bacterial density in the testes [7]. Only one study has examined factors affecting CI strength in
S.J. Perlman et al.
Cardinium [49]. In Tetranychus cinnabarinus spider mites, both CI strength and bacterial density rapidly declined with increasing male age. We investigated factors affecting CI strength in the parasitic wasp E. pergandiella, the first organism shown to carry a CI strain of Cardinium [25]. In previous studies, we observed a wide range in the strength of CI induced by wasps reared in the same lab conditions and collected from the same locations, with females in the incompatible cross producing from 7 to 38 % the number of offspring in compatible crosses [20, 25, 34]. In this study, we tested the effects of male age, male body size, the host species that male wasps develop in, bacterial density, and male development time on CI strength and found a striking effect of male development time, with males that took longer to develop expressing stronger CI. This relationship has not been previously reported and is in fact the opposite of that found in an earlier study of Wolbachia in Drosophila melanogaster, where stronger CI was found in the males of a cohort that developed more rapidly [50]. We found no effect of body size, male host species or bacterial density on CI strength and a small effect of male age, with CI increasing in very old males in one experiment.
Methods Encarsia pergandiella Cultures E. pergandiella (Hymenoptera: Aphelinidae) is a solitary parasitoid wasp with an unusual “autoparasitoid” biology [26]; female eggs are deposited into whitefly nymphs, while male eggs are deposited into developing wasp larvae or pupae, either of their own species, or of other whitefly parasitoids. Infected lines were established from wasps collected from the same locations in the Rio Grande Valley in Texas in 2003 and 2006 and were both fixed for Cardinium (and no other maternally transmitted bacteria). The 2003 line was used in previously published studies of CI [20, 34]. Male age experiments were performed on both 2003 and 2006 laboratory lines, while body size and development time experiments were performed using only the more recently established (2006) line. A Cardinium-free line was obtained by feeding adult wasps with 50 mg/ml rifampicin in honey for three generations. All cultures (cured and uncured) were maintained in the laboratory on Bemisia tabaci whiteflies reared on cowpea plants (Vigna unguiculata) for several generations prior to the study. Males were generally produced by exposing pupal wasps of another species, Eretmocerus eremicus, to E. pergandiella females. All experiments were conducted at the University of Arizona in Tucson, Arizona, during the period 2007–2009. Cardinium infection status was monitored throughout the study, using PCR, with Cardinium-specific primers [20].
Factors Affecting the Strength of CI
Testing for Cytoplasmic Incompatibility We performed separate experiments testing for an effect of male age, male body size and host species, and male development time on the strength of CI. In all experiments, Cardinium-infected males were mated to young (i.e., less than 3 days old) virgin uninfected females. Each female was then transferred to a 35-mm petri dish with a cowpea leaf disk bearing whitefly (B. tabaci) nymphs on 1 % agar. Following the oviposition period, infested leaves were incubated at 27 °C until E. pergandiella pupae could be counted and removed (8– 12 days). When an E. pergandiella egg fails to develop due to cytoplasmic incompatibility, the parasitized whitefly is developmentally arrested [25]. Whiteflies were considered to be developmentally arrested if, by the time that wasp pupae were present, the whitefly nymphs had not developed eye spots or wing buds. Female wasps that did not produce any pupal offspring and had less than two developmentally arrested whiteflies on their leaf disk were considered to be unmated and were removed from the data analysis. Data were analyzed in R [36] or JMP 7.0 (SAS Institute Inc.). Proportional data representing the strength of CI (i.e., the number of pupae/all pupae+developmentally arrested hosts) were analyzed with generalized linear models for binomial data, and quasibinomial models fitted for overdispersed data.
Effect of Male Age Freshly emerged cured virgin females were mated individually to young, middle-aged, or old virgin males of known infection status. Young, middle-aged, and old males were 0– 2, 6–9, and 11–13 days post-emergence for the first trial using the 2003 line and 1–2, 4–5, and 9–11 days post-emergence for the second trial, using the 2006 line. In the first trial, females were held in mating vials until visual confirmation of mating, which normally occurs within the first 10 s. Because this resulted in some females being unmated even after appearing to have mated, mating pairs were held in vials for 24 h for the second trial. Each female was then transferred to a 35-mm petri dish with a cowpea leaf disk on 1 % agar. Each leaf disk was infested with 50 to 100 third to early fourth instar whitefly nymphs. Females were allowed to oviposit for 24 h before being removed from leaf disks.
Effect of Body Size and Male Host Species To test the effect of male body size on the expression of CI, we reared the hyperparasitic E. pergandiella males on different wasp hosts in order to produce large and small males. Larger males emerged from E. eremicus pupae, while smaller males emerged from (female) E. pergandiella pupae. Both species
673
are common hosts for E. pergandiella males. The largest and smallest males were selected from the “large” and “small” host treatments, respectively, for use in CI experiments. Young (0–2 days old) male wasps were individually paired with 0- to 1-day-old virgin females for 6 hours to allow mating. Both hind tibiae of each male wasp were removed and measured under×400 magnification. The remaining male carcass was then transferred to 95 % ethanol until DNA extraction for Cardinium density estimation. Females were allowed to oviposit for 24 h on a cowpea leaf disk infested with 35 to 75 mostly second instar whitefly nymphs. In one analysis, we examined the effects of male size (mean tibia length) and Cardinium density on CI strength, and in another, we examined the effects of the host for males (E. pergandiella or E. eremicus) and Cardinium density on CI strength.
Effect of Development Time To produce experimental males, five to ten female E. pergandiella wasps were placed in arenas with approximately 50 E. eremicus pupae for 3 h. Males started to emerge from these arenas 9 days after maternal females were removed (“day 1”) and were collected daily from this point. Emerged males were then isolated and mated to cured females for 16– 24 h, and the emergence date of each male was recorded. Females were allowed to oviposit for 24 h on a leaf disk infested with 35–75 third-early fourth instar whitefly nymphs.
Quantifying Cardinium Densities In the body size CI experiment, quantitative PCR was used to quantify Cardinium densities in individual E. pergandiella males to determine if bacterial density was correlated with CI strength or male body size. DNA was extracted using the DNeasy tissue extraction kit (Qiagen) according to the manufacturer’s protocol for tissue samples. Cardinium was quantified by amplifying a ~200 bp region of gyrase b using primers gyrb737F (AAGTTATTGTAGCCGCTCAAG) and gyrb911R (GCAGTACCACCAGCAGAG, Beacon Design Pro 7.0). The quantity of Cardinium in each sample was standardized by amplifying a ~200 bp fragment of the insect housekeeping gene opsin using primers ops937F (TGGCTC TCCCGATCTTAC) and ops1117R (CTACGCTCTGGGCA TCTC, Beacon Design Pro 7.0). This effectively measured the relative abundance of Cardinium in each insect cell. For each PCR reaction, 2 μL of DNA extract was added to 13 μL of PCR master mix containing 1X Platinum SYBR Green qPCR SuperMix-UDG (Invitrogen), 385 nM of forward and reverse primer, 100 nM ROX reference dye. Samples were run in quadruplicate, and the average Ct value was used for all calculations. If the standard deviation of the Ct value exceeded
674
0.5, the sample was re-run or a single outlying Ct value was dropped from the analysis. Primer efficiencies were calculated using a minimum of three dilution curves run in quadruplicates. To account for inter-run variation, the same calibrator sample was included in each PCR. All calculations of the relative abundance of Cardinium in each sample were performed according to Pfaffl [35].
Results Effect of Male Age on Cytoplasmic Incompatibility In the first male age experiment using the 2003 line, CI strength (pupae divided by the total of pupae and arrested whiteflies, with lower values representing stronger CI) averaged 16 % (n=34). A large proportion of parental males in the 2003 experiment died before they were old enough to use in the old male treatment. Additionally, all but three of the females mated to old males produced no offspring and no arrested whiteflies and were therefore presumed to be unmated. This 2003 experiment showed no effect of male age on CI strength (Fig. 1a, generalized linear model (GLM); age, Χ2df= 2 =3.75, p=0.15; male infection status, Χ2df=1 =1355.37, p
