Gut pathology and responses to the microsporidium Nosema ceranae in the honey bee Apis mellifera

Abstract

The microsporidium Nosema ceranae is a newly prevalent parasite of the European honey bee (Apis mellifera). Although this parasite is presently spreading across the world into its novel host, the mechanisms by it which affects the bees and how bees respond are not well understood. We therefore performed an extensive characterization of the parasite effects at the molecular level by using genetic and biochemical tools. The transcriptome modifications at the midgut level were characterized seven days post-infection with tiling microarrays. Then we tested the bee midgut response to infection by measuring activity of antioxidant and detoxification enzymes (superoxide dismutases, glutathione peroxidases, glutathione reductase, and glutathione-S-transferase). At the gene-expression level, the bee midgut responded to N. ceranae infection by an increase in oxidative stress concurrent with the generation of antioxidant enzymes, defense and protective response specifically observed in the gut of mammals and insects. However, at the enzymatic level, the protective response was not confirmed, with only glutathione-S-transferase exhibiting a higher activity in infected bees. The oxidative stress was associated with a higher transcription of sugar transporter in the gut. Finally, a dramatic effect of the microsporidia infection was the inhibition of genes involved in the homeostasis and renewal of intestinal tissues (Wnt signaling pathway), a phenomenon that was confirmed at the histological level. This tissue degeneration and prevention of gut epithelium renewal may explain early bee death. In conclusion, our integrated approach not only gives new insights into the pathological effects of N. ceranae and the bee gut response, but also demonstrate that the honey bee gut is an interesting model system for studying host defense responses.

Figure 1. Mortality induced by N. ceranae.

Data show the percentages of surviving bees per replicate (n = 3) and per day in cages composed of 30 bees each (90 bees total/treatment). Cages with N. ceranae infected bees achieved one-hundred percent of mortality at day 14 post infection, while in control groups mortality remained low.

Figure 2. Activity of antioxidant enzymes in the midguts of bees challenged by N. ceranae.

Differences in enzymatic activity of A) superoxide dismutase (SOD), B) glutathione reductase (GR), C) glutathione peroxydase (GP) and D) glutathione-S-transferase (GST) were estimated by a Mann-Whitney U test. Means±SE are shown for 4 pools of 3 midguts per replicate (n = 3 replicates, 36 bees total/treatment). * and ** denote significant differences at p<0.05 and p<0.01, respectively.

Figure 3. Histology of honey bee midguts 7 days post-infection.

Light microscopy analysis of control (A, B) and N. ceranae infected midguts (C, D) stained with Hematoxylin-Eosin. In control guts, the peritrophic membrane (pm) and epithelial cells (ec) are homogenous, while in parasitized the guts peritrophic membrane and epithelial cells show signs of degeneration and lyses, respectively. Similar lesions were observed in each infected bees (n = 2 bees per replicate and treatment, giving n = 6 bees per treatment). A) and C) x100, B) and D) x400. Scale bar: 10 µm.

Figure 4. Network of genes downregulated by Nosema in the bee gut.

The composite functional association network derived from different genomic and proteomic data sources was generated with GeneMania using the Drosophila orthologs of bee genes. Physical and genetic interactions between genes are indicated by dot and solid lines, respectively. Grey and white circles represent implemented genes (known genes affected by Nosema) and new genes predicted to be functionally associated to the known genes, respectively. The size of the predicted gene circle provides an indication of its interaction score. Except ETS-domain lacking (edl), the predicted genes roundabout (robo), longitudinal (lola), Cyclin D (CycD), Optix, kekkon-1 (kek1) had bee orthologs: GB17658, GB12094, GB14028, GB16761 and GB17490.

Figure 5. Activity of alkaline phosphatase (AP) in the midguts of bees challenged by N. ceranae.

Means±SE are shown for 4 pools of 3 midguts per replicate (n = 3 replicates, 36 bees total/treatment). ** denotes significant differences at p<0.01 using a Mann-Whitney U test.