Honey bee colonies, parasite source, and common rearing conditions

Newly emerged A. mellifera workers were obtained from three independent colonies maintained at the experimental apiary of Fujian Agriculture and Forestry University, Fuzhou, China. Workers originating from different colonies were kept in separate cages throughout the experiments. Unless otherwise stated, each treatment comprised three independent cages, with one cage established from each source colony. The cage/colony combination was therefore considered one biological replicate. Functional experiments used 35 workers per cage, whereas the initial infection time-course experiment used 50 workers per cage. Bees were maintained at 34 ± 0.5°C and 75% relative humidity and were provided with sterile 50% (w/v) sucrose solution.

N. ceranae spores were obtained from a laboratory stock established in previous work and deposited at the China General Microbiological Culture Collection Center (CGMCC accession no. 28110). Midguts were dissected from donor workers, homogenized in sterile water, and examined by light microscopy at 400 × magnification. Only samples containing abundant N. ceranae spores were selected for subsequent spore purification.

Purification of N. ceranae spores and oral inoculation

Midguts containing abundant spores were pooled in RNase-free microcentrifuge tubes containing 500 μL sterile water and homogenized using a high-throughput tissue grinder (Meibi, China). Homogenates were centrifuged at 13,000 × g for 10 min at 4°C. The resulting pellets were purified using a discontinuous Percoll gradient composed of 25%, 50%, 75%, and 100% Percoll. Following centrifugation at 15,000 × g for 40 min at 4°C, the spore-containing material was recovered, and the gradient-purification step was repeated once. Purified spores were washed in 500 μL sterile water by centrifugation at 5,000 × g for 5 min at 4°C, resuspended in 500 μL sterile water, and stored at 4°C. Spore concentration was determined using a hemocytometer.

For infection, capped brood combs from the three source colonies were transferred to the controlled incubator described above. Newly emerged workers were collected separately from each colony and fasted for 2 h. Each worker assigned to an infected group was then fed individually with 5 μL of 50% (w/v) sucrose solution containing 1 × 10⁶ N. ceranae spores. Workers in the uninfected control group received an equal volume of spore-free 50% sucrose solution. Two hours after inoculation, each cage was supplied with 5 mL sterile 50% sucrose solution, which was replaced every 24 h.

Infection time-course experiment

To define the temporal expression pattern of ATP-A during infection, workers were assigned to an uninfected control group or an N. ceranae-infected group. Each treatment included three colony-derived cages containing 50 workers per cage. Midguts were collected daily from 1 to 12 dpi. At each time point, three midguts from one cage were pooled in a single RNase-free 1.5-mL microcentrifuge tube and constituted one biological sample. Samples were immediately frozen in liquid nitrogen and stored at −80°C until RNA extraction.

Target-gene RNAi experiment

To determine the roles of ATP-A and γ-tubulin during infection, gene-specific siRNAs targeting ATP-A (GenBank accession XM_024475731.1) or γ-tubulin (GenBank accession XM_024474433.1), together with a non-targeting siRNA control, were synthesized by Genechem Biotech Co., Ltd. (Shanghai, China). These siRNAs were designated si-ATP-A, si-γ-tubulin, and si-scramble, respectively. After the 2-h fasting period, each worker received 5 μL of 50% sucrose solution containing 1 × 10⁶ N. ceranae spores and 1.5 μg of si-ATP-A, si-γ-tubulin, or si-scramble. These groups were designated N. ceranae + si-ATP-A, N. ceranae + si-γ-tubulin, and N. ceranae + si-scramble, respectively. For molecular analyses, the expression levels of ATP-A, γ-tubulin, and NcRBL (GenBank accession NW_020169305.1) were measured from 1 to 6 dpi in the three RNAi groups. At each time point, three midguts from each cage were pooled as one biological replicate. For the physiological analyses, an uninfected group and an N. ceranae-only group were included in addition to the three RNAi groups. Sucrose consumption and worker survival were monitored from 1 to 12 dpi. Each treatment comprised three independent cages containing 35 workers per cage.

nce-miR-12220 gain- and loss-of-function experiment

The raw miRNA-seq data were deposited in the NCBI Sequence Read Archive (SRA) under BioProject accession PRJNA395137. nce-miR-12220 was one of the ten miRNAs identified in this dataset and was mapped to scaffold NW_003312212 at positions 610–859. Its mature sequence was 5′-UUUAAUUGUGAAACAUGUCUUUAGG-3′. A synthetic nce-miR-12220 mimic (M-miR-12220), an antisense inhibitor (I-miR-12220), and the corresponding negative controls (M-NC and I-NC) were synthesized by Shanghai Jima Pharmaceutical Technology Co., Ltd. (Shanghai, China). Overexpression and inhibition were performed using the previously established feeding procedures described in references [25,26]. Infected workers were assigned to four groups: N. ceranae + M-miR-12220, N. ceranae + M-NC, N. ceranae + I-miR-12220, and N. ceranae + I-NC.

Each group comprised three independent colony-derived cages containing 35 workers per cage. For molecular analyses, the expression levels of nce-miR-12220, ATP-A, γ-tubulin, Cactus, dorsal, Defensin, and Hymenoptaecin were measured from 1 to 6 dpi. At each time point, three midguts from each cage were pooled as one biological replicate. For phenotypic analyses, worker survival, N. ceranae spore load, sucrose consumption, and midgut ATP content were evaluated from 1 to 12 dpi.

RNA extraction, reverse transcription, and RT-qPCR

Total RNA was extracted from pooled midgut samples using a commercial RNA extraction kit. For nce-miR-12220 quantification, reverse transcription was performed using a sequence-specific stem-loop primer. For protein-coding transcripts, first-strand cDNA was synthesized using oligo (dT) and random primers. Quantitative PCR was performed on a QuantStudio 3 Real-Time PCR System (Applied Biosystems, USA) using the reaction system and cycling conditions described previously [27]. U6 was used as the internal control for nce-miR-12220, whereas β-actin was used as the internal control for mRNA targets. Relative expression was calculated using the 2 − ΔΔCt method. Each biological sample was analyzed in three technical qPCR replicates, and the mean Ct value of the technical replicates was used for statistical analysis. Primer sequences and oligonucleotide information were provided in S1 Table.

Dual-luciferase reporter assay

Approximately 300-bp regions containing the predicted nce-miR-12220-binding sites in ATP-A and γ-tubulin were amplified and verified by Sanger sequencing. Complementary oligonucleotides carrying either the wild-type target sequence or a mutated binding sequence were synthesized. For annealing, 2 μL aliquots of the forward and reverse oligonucleotides (1 μg/μL) were combined with 46 μL of Oligo Annealing Buffer, heated at 90°C for 3 min, and incubated at 37°C for 15 min. A 1-μL aliquot of the annealed product was ligated overnight into 50 ng of the pmirGLO vector, which had been digested with ScaI and XhoI, using T4 DNA ligase (Promega, China). The ligation products were transformed into competent Escherichia coli cells (Qingke Biology, China). Transformants were selected on ampicillin-containing medium and verified by Sanger sequencing using the pCMV5-R primer. The resulting plasmids were designated pmirGLO-ATP-A-miR-12220-wt, pmirGLO-ATP-A-miR-12220-mut, pmirGLO-γ-tubulin-miR-12220-wt, and pmirGLO-γ-tubulin-miR-12220-mut.

HEK-293T cells were cultured in high-glucose Dulbecco’s modified Eagle medium supplemented with fetal bovine serum and antibiotics at 37°C in 5% CO2. Cells were seeded in 48-well plates and transfected at 90–95% confluence using a lipofection reagent (Yisheng, China). For ATP-A, the four transfection conditions were mimic-miR-12220 + wild-type reporter, M-NC + wild-type reporter, mimic-miR-12220 + mutant reporter, and M-NC + mutant reporter. The same four combinations were established for γ-tubulin. At 24 h after transfection, cells were lysed on ice for 5 min in 100 μL of lysis buffer, and the lysates were centrifuged at 10,000 × g for 1 min. Firefly and Renilla luciferase activities were measured sequentially using a dual-luciferase reporter assay kit (Yisheng, China) and a chemiluminescence detector (Promega, China). Relative reporter activity was calculated as the firefly-to-Renilla luciferase ratio. Three independent transfection experiments were performed.

Fluorescence in situ hybridization of nce-miR-12220

At 6 dpi, midguts were dissected from N. ceranae-infected and uninfected A. mellifera workers (n = 3) and submitted to Wuhan Servicebio Technology Co., Ltd. (Wuhan, China) for paraffin embedding, sectioning, and slide preparation. An antisense oligonucleotide probe complementary to nce-miR-12220 (5′-CCTAAAGACATGTTTCACAATTAAA-3′) was synthesized by Fuzhou Shangya Biotechnology Co., Ltd. (Fuzhou, China), purified by high-performance liquid chromatography, and labeled with Cy3 at the 5′ end. RNA fluorescence in situ hybridization was performed using the Fluorescence in Situ Hybridization Kit for RNA (Beyotime Biotechnology, Shanghai, China) according to the manufacturer’s instructions. Briefly, paraffin sections were deparaffinized, rehydrated through a graded ethanol series, permeabilized with proteinase K, post-fixed with 4% paraformaldehyde, treated with hydrochloric acid, and acetylated to reduce nonspecific background. After prehybridization, the sections were incubated with the Cy3-labeled probe in hybridization solution under RNase-free and light-protected conditions. Following stringent washing with the supplied wash buffers, nuclei were counterstained with DAPI. Sections were mounted with antifade mounting medium and examined by fluorescence microscopy. The Cy3 signal was used to indicate nce-miR-12220 localization, and infected and uninfected sections were imaged using identical acquisition settings.

Spore load quantification

At each time point, one worker was sampled from each cage. The dissected midgut was homogenized in 200 μL of sterile water using a high-throughput tissue grinder, after which 800 μL of sterile water was added. A 10 μL aliquot of the suspension was loaded into a hemocytometer, and spores were counted by light microscopy at 400 × magnification. One midgut from each of the three independent cages was analyzed per treatment and time point. Spore counts were log2-transformed before statistical analysis.

Sucrose solution consumption

Sucrose consumption was measured at the cage level. A feeder containing 5 mL sterile 50% (w/v) sucrose solution was weighed at the start of each 24-h interval and reweighed immediately before replacement. Daily consumption was corrected for the number of workers alive in the corresponding cage and expressed as milligrams of 50% (w/v) sucrose solution consumed per worker per day. The cage was the experimental unit, with three cages analyzed per treatment.

Survival analysis

Survival cohorts contained 35 workers per cage and three cages per treatment, corresponding to 105 workers per treatment at the start of follow-up. The number of living workers was recorded every 24 h from 1 to 12 dpi immediately before sucrose-solution replacement, and dead workers were removed after each observation. Survival was expressed as the proportion of workers remaining alive at each observation time and displayed as Kaplan–Meier survival curves.

Measurement of ATP content in worker midguts

Midgut ATP content was quantified using the Enhanced ATP Assay Kit (Beyotime Biotechnology, Shanghai, China) according to the manufacturer’s instructions. Sample preparation was performed on ice. Midgut tissue was homogenized in ATP detection lysis buffer at a ratio of 100–200 μL lysis buffer per 20 mg tissue. Homogenates were centrifuged at 12,000 × g for 5 min at 4°C, and the supernatants were collected for ATP measurement. ATP standards were prepared in ATP detection lysis buffer over a concentration range of 0.01–10 μM. The ATP detection reagent was diluted 1:4 with the supplied diluent. For each determination, 100 μL of working solution was added to a detection well and equilibrated at room temperature for 3–5 min to reduce background luminescence. A 20 μL aliquot of each sample or ATP standard was then added to the well and mixed immediately, after which luminescence was recorded using a luminometer. ATP concentrations were calculated from the standard curve and expressed as μM. Three independent biological replicates were analyzed for each treatment and time point.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA). Unless otherwise stated, data were presented as the mean ± standard deviation (SD) of three biological replicates. ATP-A infection time-course data were analyzed by one-way analysis of variance followed by Tukey’s multiple-comparisons test. Dual-luciferase reporter data were analyzed using two-tailed, unpaired Student’s t-tests. Longitudinal gene-expression, sucrose-consumption, spore-load, and midgut ATP data were analyzed using mixed-effects models fitted by restricted maximum likelihood. Treatment, dpi, and the treatment × dpi interaction were included as fixed effects, and repeated observations from the same cage/colony replicate were accounted for in the model. Šídák’s multiple-comparisons test was used for comparisons between treatment groups at corresponding time points. In the figures, bracketed P values were used to represent the overall difference between the indicated groups across the examined time course, whereas asterisks above individual time points were used to indicate the corresponding Šídák-adjusted pairwise comparisons. Survival distributions were estimated using the Kaplan–Meier method and compared using the log-rank (Mantel–Cox) test. All tests were two-sided, and P < 0.05 was considered statistically significant.

nce-miR-12220 interacted with target genes ATP-A and γ-tubulin

The expression pattern of ATP-A was examined in A. mellifera workers during N. ceranae infection. ATP-A expression increased at 2 dpi, reaching its highest level among the examined time points, and then declined markedly from 3 to 12 dpi. Its expression remained relatively low from 6 to 12 dpi (Fig 1a). Therefore, we considered 1–6 dpi to be the critical regulatory window in which nce-miR-12220 is most likely to exert its direct effects on target genes and host immune responses.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. ATP-A expression pattern, validation of nce-miR-12220 interactions with ATP-A and γ-tubulin, and localization of nce-miR-12220 in the honeybee midgut during N. ceranae infection.

(a) Relative expression levels of ATP-A in A. mellifera workers during N. ceranae infection from 1 to 12 days post-infection (dpi). Different letters indicate significant differences among time points, as determined by one-way ANOVA followed by Tukey’s multiple-comparison test. (b) Predicted base-pairing between nce-miR-12220 and the target sequences in ATP-A and γ-tubulin. (c) Schematic representation of the pmirGLO-ATP-A reporter constructs and Sanger sequencing confirmation of the wild-type and mutant ATP-A binding regions. (d) Dual-luciferase reporter assay validating the interaction between nce-miR-12220 and the predicted binding site in ATP-A. (e) Schematic representation of the pmirGLO-γ-tubulin reporter constructs and Sanger sequencing confirmation of the wild-type and mutant γ-tubulin binding regions. (f) Dual-luciferase reporter assay validating the interaction between nce-miR-12220 and the predicted binding site in γ-tubulin. (g) Representative RNA fluorescence in situ hybridization images of midgut sections from uninfected and N. ceranae-infected A. mellifera workers at 6 dpi. For panels (d) and (f), statistical significance was assessed using Student’s t-test. Data are presented as the mean ± SD of three biological replicates. ns, P > 0.05; ****, P < 0.0001.

https://doi.org/10.1371/journal.ppat.1014521.g001

Target prediction identified complementary binding sites for nce-miR-12220 within ATP-A and γ-tubulin. The predicted binding site was located at nucleotides 1712–1735 in ATP-A and 885–912 in γ-tubulin (Fig 1b). To validate these predicted interactions, wild-type and mutant reporter constructs containing the corresponding binding regions were generated. Sanger sequencing confirmed the correct construction of recombinant plasmids carrying either the wild-type or mutated binding regions of ATP-A and γ-tubulin (Fig 1c, 1e). Dual-luciferase reporter assays further validated these interactions. Co-transfection with mimic-miR-12220 significantly reduced the relative luciferase activity of the wild-type ATP-A reporter compared with the M-NC group, while mutation of the predicted binding site abolished this effect (Fig 1d). A similar result was obtained for γ-tubulin: mimic-miR-12220 significantly decreased the luciferase activity of the wild-type reporter, whereas no significant difference was detected for the mutant reporter (Fig 1f). These findings demonstrate that nce-miR-12220 interacts with ATP-A and γ-tubulin through the predicted binding sites in a sequence-dependent manner.

RNA fluorescence in situ hybridization was then used to determine whether nce-miR-12220 could be detected within host midgut cells during N. ceranae infection. At 6 dpi, Cy3-positive signals corresponding to nce-miR-12220 were observed within the midgut epithelial cells of infected A. mellifera workers, whereas no obvious signal was detected in the uninfected control group (Fig 1g). These findings provide spatial evidence that the N. ceranae-derived nce-miR-12220 is present within honeybee midgut cells during infection.

Target gene silencing significantly affected the expression of N. ceranae RBL gene

To assess the functional roles of the predicted target genes during N. ceranae infection, ATP-A and γ-tubulin were individually silenced in infected A. mellifera workers. ATP-A expression was significantly reduced in the si-ATP-A group compared with the si-scramble group across the examined period, confirming effective RNAi-mediated knockdown (P = 0.0031; Fig 2a). Likewise, γ-tubulin expression was significantly lower in the si-γ-tubulin group than in the si-scramble group (P = 0.011; Fig 2b).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Effects of ATP-A and γ-tubulin silencing on NcRBL expression in A. mellifera workers during N. ceranae infection.

(a) Relative expression level of ATP-A following RNAi-mediated ATP-A silencing from 1 to 6 days post-infection (dpi). (b) Relative expression level of γ-tubulin following RNAi-mediated γ-tubulin silencing from 1 to 6 dpi. (c) Relative expression level of NcRBL after ATP-A silencing. (d) Relative expression level of NcRBL after γ-tubulin silencing. Statistical significance was assessed using a mixed-effects model for repeated-measures data, followed by Šídák’s multiple-comparisons test. Data are presented as the mean ± SD of three biological replicates. ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

https://doi.org/10.1371/journal.ppat.1014521.g002

The effect of target-gene silencing on NcRBL expression was then examined. Silencing ATP-A significantly reduced NcRBL expression throughout the infection period relative to the si-scramble group (P < 0.0001; Fig 2c). A similar reduction was observed after γ-tubulin silencing, with NcRBL expression remaining significantly lower than that in the si-scramble group across the examined period (P < 0.0001; Fig 2d). These results suggest that ATP-A and γ-tubulin positively contribute to NcRBL expression during N. ceranae infection.

Target gene interference influenced host survival

To further assess the effects of target-gene silencing during N. ceranae infection, sucrose consumption and survival of infected A. mellifera workers were monitored from 1 to 12 dpi. Sucrose consumption in the N. ceranae-infected group generally increased over time. No significant overall difference was detected between the N. ceranae group and the N. ceranae + siRNA-scramble group (P = 0.1209; Fig 3a), indicating that the scramble siRNA treatment did not significantly affect sucrose intake. Likewise, γ-tubulin silencing did not significantly alter overall sucrose consumption compared with the siRNA-scramble control (P = 0.7169; Fig 3a). In contrast, ATP-A silencing significantly altered sucrose consumption across the examined period (P < 0.0001; Fig 3a).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Effects of ATP-A and γ-tubulin silencing on sucrose consumption and survival of A. mellifera workers during N. ceranae infection.

(a) Consumption of 50% (w/v) sucrose solution, expressed as milligrams per worker per day, by infected workers from 1 to 12 dpi following ATP-A or γ-tubulin silencing. Statistical significance was assessed using a mixed-effects model for repeated-measures data, followed by Dunnett’s multiple comparisons test. Data are presented as the mean ± SD of three biological replicates. (b) Kaplan–Meier survival curves of infected workers following ATP-A or γ-tubulin silencing. Survival distributions were compared using the log-rank (Mantel–Cox) test.

https://doi.org/10.1371/journal.ppat.1014521.g003

Survival analysis showed no significant difference between the N. ceranae group and the N. ceranae + si-scramble group (P = 0.9811; Fig 3b). Compared with the si-scramble control, ATP-A silencing significantly improved worker survival during N. ceranae infection (P = 0.0184; Fig 3b). Similarly, γ-tubulin silencing significantly increased survival probability relative to the si-scramble group (P = 0.0188; Fig 3b). These results demonstrate that ATP-A and γ-tubulin contribute to host physiological responses during N. ceranae infection, with silencing of either gene improving worker survival.

nce-miR-12220 modulated the expression of ATP-A and γ-tubulin during N. ceranae infection

To determine whether nce-miR-12220 affects the expression of its predicted target genes during N. ceranae infection, nce-miR-12220 was manipulated in infected A. mellifera workers using a mimic or inhibitor. Compared with the M-NC group, nce-miR-12220 expression was significantly increased overall in the M-miR-12220 group (P < 0.0001), with significant increases detected at all examined time points (Fig 4a). Conversely, nce-miR-12220 expression was significantly reduced overall in the I-miR-12220 group relative to the I-NC group (P = 0.0008), and significant decreases were observed from 1 to 6 dpi (Fig 4b). These results effective overexpression and knockdown of nce-miR-12220, respectively.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Effects of nce-miR-12220 overexpression and knockdown on ATP-A and γ-tubulin expression in A. mellifera workers during N. ceranae infection.

(a-b) Relative expression level of nce-miR-12220 following M-miR-12220 and I-miR-12220treatment from 1 to 6 dpi. (c-d) Relative expression levels of ATP-A following nce-miR-12220 overexpression and knockdown. (e-f) Relative expression level of γ-tubulin following nce-miR-12220 overexpression and knockdown. Statistical significance was assessed using a mixed-effects model for repeated-measures data, followed by Šídák’s multiple-comparisons test. Data are presented as the mean ± SD of three biological replicates. ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

https://doi.org/10.1371/journal.ppat.1014521.g004

The effects of nce-miR-12220 manipulation on ATP-A and γ-tubulin expression were then evaluated. Overall, overexpression of nce-miR-12220 significantly increased ATP-A expression compared with the M-NC group (P < 0.0001). Pairwise comparisons showed significant increases from 3 to 6 dpi, whereas no significant differences were detected at 1 or 2 dpi (Fig 4c). Inhibition of nce-miR-12220 significantly reduced ATP-A expression (P < 0.0001), with significant decreases observed at all examined time points (Fig 4d).

A similar overall pattern was observed for γ-tubulin. nce-miR-12220 overexpression significantly increased γ-tubulin expression (P < 0.0001), although significant pairwise increases were detected only at 5 and 6 dpi (Fig 4e). In contrast, inhibition of nce-miR-12220 significantly decreased γ-tubulin expression (P < 0.0001), with significant reductions observed from 1 to 6 dpi (Fig 4f). These findings indicate that nce-miR-12220 positively influences ATP-A and γ-tubulin expression during N. ceranae infection, with the effects of overexpression becoming more pronounced at later infection stages.

nce-miR-12220 participated in modulating host response to N. ceranae invasion through regulating the Toll-like signaling pathway and antimicrobial peptide gene expression

To determine whether nce-miR-12220 affects host immune responses during N. ceranae infection, the expression levels of genes associated with the Toll signaling pathway and AMPs production were examined after nce-miR-12220 overexpression or knockdown. Compared with the M-NC group, nce-miR-12220 overexpression significantly reduced the overall expression level of Cactus across the examined period (P < 0.0001; Fig 5a). Conversely, inhibition of nce-miR-12220 significantly increased Cactus expression compared with the I-NC group (P < 0.0001; Fig 5b).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Effects of nce-miR-12220 overexpression and knockdown on immune-related gene expression in A. mellifera workers during N. ceranae infection.

(a-b) Relative expression level of Cactus following nce-miR-12220 overexpression and knockdown from 1 to 6 dpi. (c-d) Relative expression level of dorsal following nce-miR-12220 overexpression and knockdown. (e-f) Relative expression level of Defensin following nce-miR-12220 overexpression and knockdown. (g-h) Relative expression level of Hymenoptaecin following nce-miR-12220 overexpression and knockdown. Statistical significance was assessed using a mixed-effects model for repeated-measures data, followed by Šídák’s multiple-comparisons test. Data are presented as the mean ± SD of three biological replicates. ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

https://doi.org/10.1371/journal.ppat.1014521.g005

A similar reciprocal pattern was observed for dorsal. Overexpression of nce-miR-12220 significantly decreased dorsal expression relative to the M-NC group (P < 0.0001; Fig 5c), whereas knockdown of nce-miR-12220 significantly elevated dorsal expression compared with the I-NC group (P = 0.0006; Fig 5d).

The expression of AMP genes was also altered by nce-miR-12220 manipulation. Compared with the M-NC group, nce-miR-12220 overexpression significantly reduced the overall expression of Defensin (P = 0.0301; Fig 5e), whereas nce-miR-12220 knockdown significantly increased Defensin expression relative to the I-NC group (P = 0.0062; Fig 5f). In addition, nce-miR-12220 overexpression significantly decreased Hymenoptaecin expression (P < 0.0001; Fig 5g), while nce-miR-12220 knockdown markedly increased Hymenoptaecin expression (P < 0.0001; Fig 5h). These results indicate that nce-miR-12220 manipulation is associated with coordinated changes in the expression of Toll pathway-related genes and AMP genes during N. ceranae infection.

nce-miR-12220 modulated N. ceranae spore load, host survival rate and sucrose solution consumption during the infection process

To further assess the biological effects of nce-miR-12220 during N. ceranae infection, worker survival, N. ceranae spore load, sucrose consumption, and midgut ATP content were evaluated following nce-miR-12220 overexpression or knockdown. Compared with the M-NC group, nce-miR-12220 overexpression was associated with a lower survival probability, although the difference did not reach statistical significance (P = 0.0576; Fig 6a). Conversely, workers in the I-miR-12220 group tended to show improved survival relative to the I-NC group, but the difference was also not significant (P = 0.0907; Fig 6b).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Effects of nce-miR-12220 overexpression and knockdown on survival, N. ceranae spore load, sucrose consumption, and midgut ATP content in infected A. mellifera workers.

(a-b) Kaplan–Meier survival curve of infected workers following nce-miR-12220 overexpression and knockdown. (c-d) N. ceranae spore load from 1 to 12 dpi following nce-miR-12220 overexpression and knockdown. (e-f) Consumption of 50% (w/v) sucrose solution, expressed as milligrams per worker per day, by infected workers following nce-miR-12220 overexpression and knockdown. (g-h) Midgut ATP content following nce-miR-12220 overexpression and knockdown. Survival distributions were compared using the log-rank (Mantel–Cox) test. For panels (c–h), statistical significance was assessed using a mixed-effects model for repeated-measures data. Data are presented as the mean ± SD of three biological replicates.

https://doi.org/10.1371/journal.ppat.1014521.g006

nce-miR-12220 overexpression significantly increased N. ceranae spore load compared with the M-NC group (P < 0.0001; Fig 6c), whereas nce-miR-12220 knockdown significantly reduced spore load relative to the I-NC group (P < 0.0001; Fig 6d). These findings support a role for nce-miR-12220 in promoting N. ceranae proliferation during infection.

Sucrose consumption also changed following nce-miR-12220 manipulation. Workers treated with mimic-miR-12220 consumed significantly more sucrose than those in the M-NC group across the examined period (P = 0.0123; Fig 6e). In contrast, nce-miR-12220 knockdown significantly reduced sucrose consumption compared with theI-NC group (P < 0.0001; Fig 6f).

Midgut ATP content was then measured to assess host energy status. nce-miR-12220 overexpression significantly decreased ATP content relative to the M-NC group (P = 0.0008; Fig 6g), whereas nce-miR-12220 knockdown significantly increased ATP content compared with the I-NC group (P < 0.0001; Fig 6h). Together, these results suggest that nce-miR-12220 promotes N. ceranae proliferation and is associated with altered feeding behavior and reduced host midgut energy reserves during infection.

Discussion

Currently, study on miRNAs in microsporidian, especially Nosema species, is very limited. Following deep sequencing and bioinformatics, Huang’s group previously identified 9 miRNAs from N. ceranae [18]. A total of 10 miRNAs, including nce-miR-12220, were recently discovered in the N. ceranae spores by our team [23]. However, no intersection was observed in the documented miRNAs between Huang’s study and our work. Here, following expression pattern detection, target prediction and functional annotation of nce-miR-12220 [28], we for the first time conducted deep investigation of the nce-miR-12220-mediated regulatory mechanism underlying the N. ceranae infection of A. mellifera worker, providing a solid experimental evidence for the bee microsporidian infection regulated by miRNA. Microsporidia within the genus Nosema are characterized by their ability to produce spores with a distinct layered wall structure [29]. Notably, no mitochondrial structures were detected in the N. ceranae spores, the absence of mitochondria implied that N. ceranae cannot generate ATP via traditional methods such as oxidative phosphorylation or the tricarboxylic acid (TCA) cycle. It has been reported that microsporidia can directly import host ATP through ATP/ADP transport proteins and secure their energy supply by manipulating host metabolism like activation of the glycolysis, TCA cycle, and key enzyme expression [30]. Such a strategy exemplifies the adaptive evolution of parasitic organisms under extreme genomic reduction [31]. However, the details of the interaction between microsporidia and host mitochondria are not well understood. The mechanism by which N. ceranae hijacks the host energy metabolism presents an intriguing scientific question.

N. ceranae spores exist as dormant entities in the environment, invading hosts through polar filament ejection to infect midgut epithelial cells and injecting sporoplasm for proliferation [32]. nce-miR-12220 was significantly upregulated at 2 dpi (224.58%) followed by marked downregulation from 3 to 12 dpi [24]. This expression pattern suggests that nce-miR-12220 may play a role not only in the initiation of spore polar filament ejection and early infection but also in the modulation of fungal proliferation. MiRNAs typically exert biological effects on the host by negatively regulating target genes [33,34]. Recent evidence has suggested that some miRNAs could also positively regulate target gene expression and some crucial processes [35]. In a previous study, we found that nce-miR-12220 formed a regulatory network with 15 target genes, encompassing various physiological processes, such as energy metabolism and cellular structure [24]. Both ATP-A and γ-tubulin are critical genes associated with energy metabolism. ATP-A, as a component of vacuolar ATP synthase, is involved in energy transport across membranes and regulates numerous physiological processes in host cells [36,37]. γ-tubulin is a pivotal protein in microtubule organizing centers, playing essential roles in microtubule formation and cell cycle regulation [38]. Here, we detected that during the infection process of N. ceranae, the ATP-A expression was significantly upregulated at 2 dpi, followed by downregulation from 3 to 12 dpi (Fig 1a), consistent with the expression pattern of nce-miR-12220. Meanwhile, γ-tubulin was continuously downregulated from 2 to 12 dpi, demonstrating a similar trend to that of nce-miR-12220 and ATP-A [24]. These demonstrated that nce-miR-12220 was likely to positively regulate both ATP-A and γ-tubulin. In addition, the binding relationship between nce-miR-12220 and ATP-A (γ-tubulin) was verified by dual-luciferase reporter assays (Fig 1), offering strong evidence for the hypothesis that nce-miR-12220 participated in the N. ceranae infection process through regulating these target genes. In this work, it’s observed that the nce-miR-12220 overexpression significantly upregulated the expression of ATP-A and γ-tubulin (Fig 4c and 4e). Conversely, the knockdown of nce-miR-12220 significantly reduced the expression of both genes (Fig 4d and 4f). Together, these results further validated the regulatory role of nce-miR-12220 in the N. ceranae infection.

RBL, serving as a virulence factor of microsporidia, plays a essential part in the fungal infection [39]. Studies have shown that the RBL domain, acting as an adhesion protein, is an indispensable component in microsporidia’s infection process and a pivotal mediator in their binding to host cell surfaces [40,41]. This process facilitates the initial invasion of the microsporidia, representing a key step in the establishment of infection [41]. Our study demonstrated that silencing ATP-A or γ-tubulin led to a significant reduction in the expression level of NcRBL during 1–6 dpi (Fig 2), indicating that targeting these genes could interfere with microsporidian invasion by suppressing the NcRBL expression. Further analysis revealed that interference with ATP-A or γ-tubulin significantly increased survival rates (Fig 3b). In summary, these results demonstrated that silencing ATP-A or γ-tubulin could suppress the microsporidian proliferation and subsequently improve host survival, indicative of their vital functions in the N. ceranae infection.

The Toll signaling pathway plays a critical role in insect immune responses [42], serving as a key immune barrier particularly for insects against infections by pathogenic microorganisms (including bacteria, viruses, and fungi) [42]. Serpin-1a and serpin-6 regulate the Toll pathway immune homeostasis by synergistically inhibiting the Spätzle-processing enzyme CLIP2 in silkworm, Bombyx mori. For instance, infection with Rice stripe virus (RSV) can activate the Toll pathway in insect vectors, upregulating the expression of core pathway genes (e.g., Toll, Dorsal), thereby inhibiting viral proliferation [43,44]. In shrimp and Drosophila, the Toll pathway regulates the expression of immune effector molecules such as anti-lipopolysaccharide factors (ALFs) through the activation of transcription factor Dorsal [45]. In the present study, nce-miR-12220 overexpression coordinately decreased the transcript levels of Cactus and dorsal, whereas nce-miR-12220 knockdown produced the opposite expression pattern (Fig 5a-5d). The simultaneous downregulation of Cactus and Dorsal transcripts after nce-miR-12220 overexpression likely indicates broad transcriptional suppression or dysregulation of Toll-related immune signaling during infection, rather than canonical activation of the pathway.

Insect antimicrobial peptides (AMPs) are core components of the insect innate immune system, playing a pivotal role in resisting the invasion of pathogenic microorganisms and maintaining host health [46]. They are primarily produced by humoral immune responses; for example, after pathogenic infection, they are released from the fat body into the hemolymph, exerting direct killing effects on microorganisms such as bacteria, fungi, and viruses [47]. Defensin is an important class of cationic peptides in the insect antimicrobial peptide family, renowned for its broad-spectrum antimicrobial activity and regulatory role in innate immunity [48]. For instance, in honeybees (Apis cerana cerana), the expression of Defensin genes (e.g., Defensin-1) serves as a significant marker of immune responses, which undergoes significant changes under the influence of environmental factors such as insecticides (e.g., azadirachtin). Hymenoptaecin is a specific insect antimicrobial peptide, whose immune functions have been mainly studied in honeybees [49]. Similarly, the present study also found that overexpression of nce-miR-12220 could significantly lead to the downregulation of antimicrobial peptide genes such as Hymenoptaecin and Defensin (Fig 5).

The results indicated that nce-miR-12220 may promote the N. ceranae proliferation by controlling host immune response via regulation of these four critical immune genes. A similar strategy exists in the infection process of other microsporidia against their hosts. For instance, microsporidia utilize the nucleus-targeted effector EnP1, which translocates into the host nucleus via nuclear localization signals, to interact with host histone H2B. This interaction disrupts H2B monoubiquitination, thereby inhibiting the expression of p53 and impairing its regulation of downstream SLC7A11. Consequently, the ferroptosis resistance of host cells is enhanced, creating favorable conditions for the proliferation of microsporidia themselves [49].

N. ceranae lacks the classical mitochondrial structure and thus cannot generate ATP through mitochondria like other eukaryotes; instead, it must obtain energy from the host to sustain its own metabolic requirements [30]. Studies have reported a significant increase in sugar intake in bees infected with N. ceranae [50]. In the present study, nce-miR-12220 overexpression increased sucrose consumption and decreased midgut ATP content, while its knockdown produced the opposite effects (Fig 6e- 6h). These coordinated changes suggest that nce-miR-12220 is associated with altered host energy expenditure and ATP homeostasis during N. ceranae infection. Consistently, nce-miR-12220 overexpression increased N. ceranae spore load, whereas its knockdown reduced spore abundance (Fig 6c- 6d), supporting a role for this miRNA in promoting parasite proliferation. Although whole-midgut ATP measurements do not resolve the specific metabolic pathways involved, future parasite-specific ATP/ADP analyses and glycolytic enzyme assays may further clarify the underlying mechanism.