Systemic infection generates a local-like immune response of the bacteriome organ in insect symbiosis

Abstract

Endosymbiosis is common in insects thriving in nutritionally unbalanced habitats. The cereal weevil, Sitophilus oryzae, houses Sodalis pierantonius, a Gram-negative intracellular symbiotic bacterium (endosymbiont), within a dedicated organ called a bacteriome. Recent data have shown that the bacteriome expresses certain immune genes that result in local symbiont tolerance and control. Here, we address the question of whether and how the bacteriome responds to insect infections involving exogenous bacteria. We have established an infection model by challenging weevil larvae with the Gram-negative bacterium Dickeya dadantii. We showed that D. dadantii infects host tissues and triggers a systemic immune response. Gene transcript analysis indicated that the bacteriome is also immune responsive, but it expresses immune effector genes to a lesser extent than the systemic and intestinal responses. Most genes putatively involved in immune pathways remain weakly expressed in the bacteriome following D. dadantii infection. Moreover, quantitative PCR experiments showed that the endosymbiont load is not affected by insect infection or the resulting bacteriome immune activation. Thus, the contained immune effector gene expression in the bacteriome may prevent potentially harmful effects of the immune response on endosymbionts, whilst efficiently protecting them from bacterial intruders.

Fig. 1

Kinetics of AMP gene expression in whole aposymbiotic larvae injected with M. luteus or D. dadantii. The effect of the type of injection and the effect of time are statistically significant for all genes. For detailed statistical results, see online supplementary tables 2 and 3. Error bars represent the standard error of four independent measurements.

Fig. 2

Tissue distribution of D. dadantii in infected symbiotic larvae. a Kinetics of D. dadantii counts in whole infected larvae. Bacteria count increases from 2 to 12 h after the injection. At 24 h postinjection, the count falls under the detection limit, indicating bacterial clearance in surviving larvae. Error bars represent the standard error of five independent measurements. At 6 h postinfection, D. dadantii could be seen in most tissues of symbiotic larvae, including the fat body (b), muscles (c) and bacteriome (d). It had also accumulated along the gut epithelium (e), sometimes causing damage (f). Bacteria were, however, never found inside the host cells. At 12 h postinfection, D. dadantii was found mostly on the borders of the organs (g). In surviving larvae, bacteria were seen as small stacks along the bacteriome (h) and the gut epithelium (i) only. In larvae dying from infection, the bacteria had spread to all the tissues, including the fat body (j), the bacteriome (k) and the gut (l).

Fig. 3

Bacteriome immune response in symbiotic larvae injected with M. luteus or D. dadantii. AMP gene expression was quantified at 2, 6, 12 and 24 h after the bacterial challenge. The effect of the type of injection, the effect of time and their interaction are statistically significant for all genes. For detailed statistical results, see online supplementary tables 2 and 3. Error bars represent the standard error of four independent measurements.

Fig. 4

Expression kinetics of immune-related genes observed for humoral (H) and bacteriome (B) responses in larvae injected with M. luteus or D. dadantii. Whole aposymbiotic larvae were used to assess the systemic humoral immune response; symbiotic larvae were used to measure the bacteriome-specific response. Statistical results from the global two-way ANOVA analysis are displayed on each graph with inf the infection type, t the time postinfection and int their interaction. ⁻ p > 0.05; * p < 0.05; ** p < 0.01; *** p < 0.001. Detailed p values and full contrast analysis results are displayed in online supplementary tables 2 and 3. Error bars represent the standard error of four independent measurements.

Fig. 5

Expression of immune-related genes in the bacteriome (B) and gut (G) of symbiotic larvae 6 h after D. dadantii infection. Asterisks attached to gene names indicate a significant difference in global expression levels between B and G from a two-way ANOVA analysis. Asterisks on the bar plots indicate a significant difference between infected and control conditions within one organ from a Welch two-sample t test (* p < 0.05; ** p < 0.01; *** p < 0.001). Error bars represent the standard error of three independent measurements. PBS = PBS injected; Dd = D. dadantii injected.

Fig. 6

S. pierantonius quantification in whole symbiotic larvae from 2 to 24 h after D. dadantii or M. luteus infection. The solid line represents the mock infection, the dashed line indicates infection with D. dadantii and the dotted line shows infection with M. luteus. No significant difference was found between infected and control larvae. Error bars represent the standard error of five independent measurements.