Global discovery, expression pattern, and regulatory role of miRNA-like RNAs in Ascosphaera apis infecting the Asian honeybee larvae

Abstract

Ascosphaera apis, a specialized fungal pathogen, causes lethal infection in honeybee larvae. miRNA-like small RNAs (milRNAs) are fungal small non-coding RNAs similar to miRNAs, which have been shown to regulate fungal hyphal growth, spore formation, and pathogenesis. Based on the transcriptome data, differentially expressed miRNA-like RNAs (DEmilRNAs) in A. apis infecting the Apis cerana cerana worker 4-, 5-, and 6-day-old larvae (Aa-4, Aa-5, and Aa-6) were screened and subjected to trend analysis, followed by target prediction and annotation as well as investigation of regulatory networks, with a focus on sub-networks relative to MAPK signaling pathway, glycerolipid metabolism, superoxide dismutase, and enzymes related to chitin synthesis and degradation. A total of 606 milRNAs, with a length distribution ranging from 18 nt to 25 nt, were identified. The first nucleotide of these milRNAs presented a bias toward U, and the bias patterns across bases of milRNAs were similar in the aforementioned three groups. There were 253 milRNAs, of which 68 up-and 54 down-regulated milRNAs shared by these groups. Additionally, the expression and sequences of three milRNAs were validated by stem-loop RT-PCR and Sanger sequencing. Trend analysis indicated that 79 DEmilRNAs were classified into three significant profiles (Profile4, Profile6, and Profile7). Target mRNAs of DEmilRNAs in these three significant profiles were engaged in 42 GO terms such as localization, antioxidant activity, and nucleoid. These targets were also involved in 120 KEGG pathways including lysine biosynthesis, pyruvate metabolism, and biosynthesis of antibiotics. Further investigation suggested that DEmilRNA-targeted mRNAs were associated with the MAPK signaling pathway, glycerolipid metabolism, superoxide dismutase, and enzymes related to chitin synthesis and degradation. Moreover, the binding relationships between aap-milR10516-x and ChsD as well as between aap-milR-2478-y and mkh1 were confirmed utilizing a combination of dual-luciferase reporter gene assay and RT-qPCR. Our data not only provide new insights into the A. apis proliferation and invasion, but also lay a basis for illustrating the DEmilRNA-modulated mechanisms underlying the A. apis infection.

Keywords: Apis cerana; Ascosphaera apis; chalkbrood; milRNA; regulatory network; target mRNA.

Figure 1

Number statistics, Venn diagram, and expression clustering of A. apis milRNAs. (A) Statistics of known and novel milRNAs in three different groups; (B) Venn diagram of milRNAs in three different groups; (C) Heat map of expression clustering of shared milRNAs by different groups.

Figure 2

Molecular validation of expression and sequences of A. apis milRNAs. (A) Agarose gel electrophoresis for the amplification products from Stem-loop RT-PCR of aap-milR-31-x, aap-milR-7977-x, and aap-milR-9993-y; (B) Peak diagrams of Sanger sequencing of the amplified fragments from aap-milR-31-x, aap-milR-7977-x, and aap-milR-9993-y.

Figure 3

The structural property of milRNAs identified in 4-, 5-, and 6-day-old larval guts infected by A. apis. (A) Length distribution of milRNAs; (B) First base bias of milRNAs; (C) Base bias of milRNAs at each base.

Figure 4

Trend analysis of A. apis milRNAs. Colorful squares represent significant trends, while colorless squares represent non-significant trends; the numbers located in the upper left of each square indicate different trends, whereas those located in the lower left of each square indicate the p value of each trend.

Figure 5

Loop diagrams of GO terms enriched by target mRNAs of A. apis milRNAs within 3 significant trends including profile4 (A), profile6 (B), and profile7 (C).

Figure 6

Chord diagrams of KEGG pathways enriched by A. apis milRNA-targeted mRNAs in 3 significant trends. The scale value represents the proportion of the corresponding color label. The line in the middle indicated the presence of pathways in a Profile, and the thicker the line segment, the more genes could be annotated to the pathway.

Figure 7

Regulatory networks between A. apis DEmilRNAs in significant trends and target mRNAs associated with virulence factors. (A) Regulatory network of milRNAs-targeted superoxide dismutase-encoding mRNAs; (B) Regulatory network of milRNAs-targeted enzymes related to chitin synthesis and degradation-encoding mRNAs.

Figure 8

Validation of binding relationships between ChsD and aap-milR10516-x and between mkh1 and aap-milR-2478-y. (A) Peak diagram of Sanger sequencing of the amplified binding sites. (B) Dual-luciferase reporter gene assay of the binding relationship between aap-milR10516-x and ChsD. (C) RT-qPCR detection of aap-milR10516-x and ChsD. (D) Peak diagram of Sanger sequencing of the mutated binding sites. (E) Dual-luciferase reporter gene assay binding relationship between aap-milR-2478-y and mkh1. (F) RT-qPCR detection of aap-milR-2478-y and mkh1. The dual-luciferase assay data were presented as mean ± standard deviation (SD) and analyzed by two-sided Student’s t-test; ns, p > 0.05; **p < 0.01; ***p < 0.001.

Figure 9

A hypothetical working model of milRNA regulation in A. apis infecting the A. c. cerana larvae.