Cardinium prevalence in mites

This paper includes of review of 34 papers reporting Cardinium infections in mites. Cardinium infections were reported in 45 mite species belonging to 3 orders and 14 families. Detailed information was present in Table 1. The associations of Cardinium and family Tetranychidae were relatively well documented. Additionally, Cardinium was found in more than one species of family Phytoseiidae, Acaridae, Glycyphagidae, Pyroglyphidae and Tenuipalpidae.

Cardinium is an intracellular symbiotic bacterium that cannot be cultured outside of host cells. Thus, the detection of its presence in mites is done mainly using PCR-based techniques, sometimes aided by bioassays (crossing breeding studies). Available molecular markers for its detection in mites include, but not limited to, the 16S ribosomal RNA (16S rRNA) and the Gyrase B (gyrB) (Hoy & Jeyaprakash 2005; Chigira & Miura 2005; Liu et al. 2006; Novelli et al. 2007; Enigl & Schausberger 2007; Novelli et al. 2008; Chen et al. 2009; Moro et al. 2009b; Jeong et al. 2009; Ros et al. 2012; Zhu et al. 2012b; Zhao et al. 2013a; Zhao et al. 2013b; Hubert et al. 2012; Famah Sourasson et al. 2014). The other techniques used to determine the presence and distribution of Cardinium in mites are transmission electron microscopy (Kitajima et al. 2007) and fluorescence in situ hybridization (Zhu et. al. 2012a). Kitajima et. al. (2007) presented a visual evidence for the presence of Cardinium in three Brevipalpus species, while Zhu et al. (2012a) demonstrated that Cardinium mainly enters the oocytes with trophic flow from nurse cells, midgut and oviduct in Tetranychus piercei.

Cardinium seems to be restricted to four arthropod lineages: Hymenoptera, Hemiptera, Diptera and Acari (Weeks et al. 2003; Zchori-Fein & Perlman 2004; Duron et al. 2008; Nakamura et al. 2009). Among these, mites probably represent Cardinium hotspots. For example, Weeks et al. (2003) had estimated that about 31.6% mite species should to be infected by Cardinium, which seems higher compared with infection rate in others (6.1% in Hemiptera, 10.4% in Hymenoptera). Nine out of 22 spider mite species and 13 astigmatid mite species were reported to harbor Cardinium (Nakamura et al. 2009; Kopecky et al. 2013). Zindel et al. (2012) screening for bacterial symbionts in 21 mite species, including phytophagous, predatory or parasitic species, had discussed their possible importance in biological control. To explain high prevalence of Cardinium in mites, Weinert et al. (2015) speculated that this might reflect evolutionary changes in arthropod immunity, as many of the groups with high Cardinium incidences lack components of immune deficiency (IMD) signal pathway, and IMD is activated by diaminopimelic acid-type peptidoglycan, which is produced by Cardinium (Penz et al. 2012).

Given high prevalence of Cardinium and mite species diversity, the number of infected mite species was obviously underestimated. In addition, the infection in mite species of economic importance appears to receive much attention, as indicated by well-documented data for these species. The screened mite species are only a small proportion, and the bacterial community of mites is influenced by environmental factors (Novelli et al. 2008; Wu & Hoy 2012; Kopecky et al. 2014). It is predicated that more infected species exist on the basis of extensive sampling and sensitive detection approaches. Meanwhile, extensive sampling of mites that have genetic divergence and live in various habitats would allow us to infer how host phylogeny and ecology affect their endosymbiotic bacterial infections.

TABLE 1.

Documented Cardinium infections in mite species.

Continued.

Cardinium's reproductive phenotypes in mites

Cardinium is known to affect their hosts in a variety of ways in order to favor their own transmission. These include CI and feminization. For example, Cardinium was found associated with nonreciprocal reproductive incompatibility in M. occidentalis, a phytoseiid mite that is an important predator of spider mite pests of agricultural crops (Wu & Hoy 2012). Cardinium-induced CI was also revealed in six phytophagous mite species, Eotetranychus suginamensis (Gotoh et al. 2007), Bryobia sarothamni (Ros & Breeuwer 2009), Tetranychus cinnabarinus (Xie et al. 2010), T. piercei (Zhu et al. 2012b), Tetranychus phaselus (Zhao et al. 2013a) and Tetranychus truncatus (Zhao et al. 2013b). All these CI types were expressed as female mortality, which is the only CI type found for Cardinium so far. Moreover, Groot & Breeuwer (2006) found Cardinium induce thelytoky by feminizing unfertilized haploid eggs in the related species Brevipalpus californicus and B. phoenicis. Through reproductive manipulations, Cardinium could strongly impact mite population dynamics and demography as well as the mite genetic diversity (Groot 2007; Ros 2010).

As mentioned previously, factors such as bacterial strains and densities, host genetic background or interactions between bacteria and host species can influence Cardinium-induced phenotypes. Gotoh et al. (2007) revealed that phylogenetic similarity of Cardinium was not related to their phenotypes, as CI-inducing Cardinium strain in E. suginamensis is closely related to the non-CI strains found in T. pueraricola and Amphitetranychus quercivorus, based on the 16S rRNA tree. Furthermore, some studies revealed that Cardinium genotype, host genotype and Cardinium density may account for variations in CI expression (Xie et al. 2010; Zhu et al. 2012b; Zhao et al. 2013b). To date, studies on Cardinium-mediated reproductive alterations are still limited to a few phytophagous, predatory or parasitic species and more related studies especially in synanthropic mites infesting agricultural crops, stored products or acting as allergen are needed.

Potential roles of Cardinium in mite biological control

Predatory mites are widely used for biological control of phytophagous mites. Cardinium-induced CI in predatory mites M. occidentalis has been reported (Johanowicz & Hoy 1998; Wu & Hoy 2012). Weeks & Stouthamer (2004) also reported that three inbred lines of M. occidentalis showed a clear and significant increase in fecundity associated with infection by Cardinium. In these cases. Cardinium-mediated reproductive incompatibility may have significant implications for biological control programs using M. occidentalis because the release of M. occidentalis strains that are reproductively incompatible with the wild populations (caused by a different status of Cardinium infection) may reduce the effectiveness of this predatory mite as a biological control agent.

Phylogenetic relationship of Cardinium and their mite hosts

Phylogenetic analyses indicated that the evolutionary relationship among Cardinium infecting mites was well resolved and the observed tree showed congruence with the classical taxonomy based on morphology, except for the family Chortoglyphidae and Carpoglyphidae (Fig. 1), suggesting 18S rRNA's better discriminating ability. Although the resolution of the phylogenetic tree of Cardinium using 16S rRNA was low, it can provide information to identify two clusters of closely related Cardinium strains. As shown in Fig. 2, cluster I consisted of Cardinium from T. pueraricola, T. phaselus, T. kanzawai, E. suginamensis, A. quercivorus, A. Viennensis, Oligonychus illcis, B. obovatus, Petrobia hard and Balaustium sp. And the remaining strains formed the cluster II.

Comparisons between the evolutionary relationships among mites and associated Cardinium symbionts showed that there is no complete association between the two groups, as several Cardinium strains infecting different species were monophyletic. We cannot completely exclude the possibility that there has been repeated loss of infection post-speciation from an infected ancestor. This difference between the two phylogenies indicated that Cardinium was not solely acquired vertically, and points to the likelihood of horizontal transmission. Meanwhile, a degree of Cardinium specialization was observed, for example, closely related Cardinium strains clustered among three closely related Brevipalpus species and Cardinium infecting family Tetranychidae appeared to be closely related. These results were not surprising and similar trends have been observed in other Cardinium-arthropod associations (Weeks et al. 2003; Zchori-Fein & Perlman 2004; Duron et al. 2008; Kopecky et al. 2013) and in other endosymbionts (Weinert et al. 2009; Lachowska et al. 2010; Gerth et al. 2015). In addition, the single molecular marker will fail to assess the accurate phylogeny of Cardinium (Weeks et al. 2003; Duron et al. 2008). Thus more rapidly evolving and phylogenetically informative Cardinium genes are required in further study.

Correlation between Cardinium infections and mite hosts ecology

Mites exhibit a breadth of ecological interactions unmatched in any other arthropod group (Krantz & Walter 2009). Diversity of habitats allowed us to test whether host ecology affects Cardinium strain infection patterns. Infected mite species can be spitted into seven types according to their habitats or economic importance. Species of the same type may share common environment or have ecological interactions. In general, mite ecology appeared to shape Cardinium strain similarity (Fig. 2). For example, the phylogenetic analysis revealed that Cardinium from phytophagous mites were more similar to one another. In addition, similar Cardinium strains were observed in mites that inhabiting house environment. These observations suggest that mite species sharing close ecological conditions easily transmit Cardinium among them. Ecologically mediated transmission of bacterial symbionts had also been reported in fruit feeding drosophilids (Haine et al. 2005), mycophagous flies (Stahlhut et al. 2010) and Megalomyrmex social parasites and their fungus-growing ant hosts (Liberti et al. 2015).

FIGURE 1.

Bayesian phylogenetic tree of mite species infected with Cardinium symbionts based on 18S rRNA genes. The topology resulting from Maximum Likelihood (ML) method was similar. ML bootstrap values (left numbers) and Bayesian probabilities (right numbers) were presented above the branches. Words on the tree indicated the family in which mite species is placed.

FIGURE 2.

Maximum Likelihood phylogenetic tree of Cardinium symbionts associated with several mite species based on 16S rRNA genes. Strains were characterized by the names of their host species. The topology resulting from Bayesian method was similar. ML bootstrap values (right numbers) and Bayesian posterior (left numbers) were presented above the branches. Different shapes on the right side indicated mites' economic importance. Strains with the ability of manipulating reproduction were marked with bold letters.

Multiple infections of Cardinium and other endosymbionts in mites

Apart from Cardinium, mites are hosts to other endosymbionts. For instance, Chaisiri et al. (2015) have revealed that mite species were associated with 85 bacterial genera (approximately 150 identified species) belonging to 7 phyla and 25 orders. Among these bacteria, Cardinium, Wolbachia and Orientia are the most prevalent bacteria. Wolbachia is another known reproductive parasite (Werren et al. 2008). Many mite species showed co-infected with Cardinium and Wolbachia (Table 1) and sometimes, the two endosymbionts were found co-infecting within the same individuals of particular species. Population differentiation itself may facilitate the persistence of individual endosymbionts within different populations, whereas the acquisition of superinfections through horizontal transmission may contribute to intra-individual co-infection.

The presence of competing endosymbiont strains is expected to influence the dynamics of the population as a whole, for example, if neighbouring populations are differently infected with one or more bacterial strains with different phenotypes or if a new strain invades one population. An example of this is observed in Bryobia spider mites where Cardinium, which causes CI, sometimes co-occurs with Wolbachia (Ros & Breeuwer 2009). Similarly, either or both Wolbachia and Cardinium induce CI, as Wolbachia induces weak CI, whereas Cardinium induces strong CI and doubly-infected ones showed the strongest CI in T. piercei (Zhu et al. 2012), T. phaselus (Zhao et al. 2013a) and T. truncates (Zhao et al. 2013b). In addition, Weeks et al. (2003) and Zchori-Fein & Perlman (2004) reported mixed infections with Wolbachia and Cardinium in the predatory mite M. occidentalis. The poultry red mite Dermanyssus gallinae was found to be infected with both Cardinium and Spiroplasma (De Luna et al. 2009). These observations indicated that Cardinium might shape its hosts' ecology and evolution in cooperation with other endosymbionts.