Should Symbionts Be Nice or Selfish? Antiviral Effects of Wolbachia Are Costly but Reproductive Parasitism Is Not - PubMed (original) (raw)
Should Symbionts Be Nice or Selfish? Antiviral Effects of Wolbachia Are Costly but Reproductive Parasitism Is Not
Julien Martinez et al. PLoS Pathog. 2015.
Abstract
Symbionts can have mutualistic effects that increase their host's fitness and/or parasitic effects that reduce it. Which of these strategies evolves depends in part on the balance of their costs and benefits to the symbiont. We have examined these questions in Wolbachia, a vertically transmitted endosymbiont of insects that can provide protection against viral infection and/or parasitically manipulate its hosts' reproduction. Across multiple symbiont strains we find that the parasitic phenotype of cytoplasmic incompatibility and antiviral protection are uncorrelated. Strong antiviral protection is associated with substantial reductions in other fitness-related traits, whereas no such trade-off was detected for cytoplasmic incompatibility. The reason for this difference is likely that antiviral protection requires high symbiont densities but cytoplasmic incompatibility does not. These results are important for the use of Wolbachia to block dengue virus transmission by mosquitoes, as natural selection to reduce these costs may lead to reduced symbiont density and the loss of antiviral protection.
Conflict of interest statement
The authors have declared that no competing interests exist.
Figures
Fig 1. Phylogenetic distribution of CI levels and Wolbachia effects on egg hatch rates, fecundity and lifespan.
(A) The phylogeny based on the MLST genes 16S rRNA, aspC, atpD, ftsZ, sucB, groEL, coxA and fbpA was inferred using ClonalFrame v1.2 [43] as in [14]. Strains in bold conferred significant antiviral protection [14]. Branch labels represent posterior support values. Nodes with less than 50% support were collapsed. Branch lengths indicate relative time. (B) CI measured as egg hatch rates in crosses between uninfected females and _Wolbachia_-infected males. (C) Egg hatch rates in crosses between _Wolbachia_-infected females and _Wolbachia_-infected males (blue bars) or uninfected males (grey bars). (D) Fecundity of _Wolbachia_-infected females. (E) Lifespan of _Wolbachia_-infected females. Error bars are standard errors. *: significance relative to the _Wolbachia_-free line (Dunnett’s test; *: P < 0.05; **: P < 0.01; ***: P < 0.001). The dotted line indicates for each trait the mean value in the _Wolbachia_-free controls. (F) Original host species of the Wolbachia strains.
Fig 2. Correlation between CI and antiviral protection.
Levels of CI estimated as the percentage of unhatched eggs relative to the mean hatch rate in crosses between uninfected females and uninfected males. Level of protection measured as survival in [14] upon infection with (A) DCV and (B) FHV (0 and positive values mean no difference and increase in survival compared to _Wolbachia_-free control respectively). Means and standard errors are shown. Solid lines show predicted values from linear regressions using all strains (black) or only CI-inducing strains (red). r is the Pearson’s correlation coefficient between traits.
Fig 3. Correlations between antiviral protection and other host life-history traits.
A and B: correlation between survival after viral infection and egg hatch rates in crosses between _Wolbachia_-free males and _Wolbachia_-infected females. Virus infections used (A) DCV and (B) FHV [14] (0 and positive values mean no difference and increase in survival compared to _Wolbachia_-free control respectively). C and D: correlation between decrease in male fertility in crosses between _Wolbachia_-infected parents and survival after infection with (C) DCV and (D) FHV. E and F: correlation between egg number and survival after infection with (E) DCV and (F) FHV. Means and standard errors are shown. Solid lines show predicted values from linear regressions. r is the Pearson’s correlation coefficient between traits.
Fig 4. Correlations between CI and other host life-history traits.
The level of CI is correlated with (A) the egg hatch rates in crosses with _Wolbachia_-free males, (B) the decrease in male fertility and (C) the egg number. Means and standard errors are shown. Solid lines show predicted values from linear regressions using all strains (black) or only CI-inducing strains (red). r is the Pearson’s or Spearman’s (*) correlation coefficient between traits.
Fig 5. Wolbachia tissue tropism.
Mean Wolbachia density in (A) head and thorax of females, (B) testes and (C) freshly laid eggs. Error bars are standard errors. Letters indicate significant differences based on a Tukey’s honest significance test on ln-transformed data. All tissues were analyzed in a single linear model to test for difference in tissue tropism: strain effect: F15,427 = 131. 1; P < 0.0001; tissue effect: F2,427 = 4448. 8; P < 0.0001; strain × tissue effect: F30,427 = 11.5; P < 0.0001.
Fig 6. Correlations between Wolbachia density in somatic tissues and antiviral protection, CI or other host life-history traits.
The relative Wolbachia density in head and thorax of females is correlated with survival [14] upon infection with (A) DCV or (B) FHV (0 and positive values mean no difference and increase in survival compared to _Wolbachia_-free control respectively), (C) the level of CI, (D) the egg hatch rate in crosses with _Wolbachia_-free males, (E) the decrease in male fertility and (F) the egg number. Means and standard errors are shown. Solid lines show predicted values from linear regressions. r is the Pearson’s correlation coefficient between traits.
References
- Douglas AE. The microbial dimension in insect nutritional ecology. Funct Ecol. 2009;23: 38–47.
Publication types
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources
Molecular Biology Databases