Honeybee gut microbiota profile
In this study, the gut bacterial community of honeybees was characterized via 16S rRNA amplicon Illumina sequencing. A quality control procedure was performed on the paired-end reads, including trimming the barcodes and primers and filtering low-quality reads and chimeras, according to QIIME to yield the results for processed data (Table S1). As shown in Fig. S1, the rarefaction curves for all samples approached a saturation plateau, indicating that the current analysis contained adequate depth to capture most microbial diversity information.
All valid reads were classified taxonomically (phylum, class, order, family, and genus levels) using QIIME. The top 10 most abundant microbes in honeybee guts are presented in Fig. S2. The top 10 most abundant phyla in the honeybee gut were Proteobacteria, Firmicutes, Actinobacteria, Bacteroidetes, Cyanobacteria, Verrucomicrobia, Acidobacteria, Chloroflexi, Spirochaetes, and Planctomycetes (Fig. S2A). The top 10 most abundant genera of bacteria in the honeybee gut were Lactobacillus, Bifidobacterium, Bacteroides, Roseburia, Xanthobacter, Marivita, Faecalibacterium, Ruminococcus, Dialister, and Pseudanabaena (Fig. S2E). When the data from the midgut and hindgut were combined, the dominant populations were Proteobacteria (63.17%), Firmicutes (17.61%), Actinobacteria (4.06%), and Bacteroidetes (1.72%) at the phylum level (Fig. 1A) and Lactobacillus (15.91%) and Bifidobacterium (3.24%) at the genus level (Fig. 1B), regardless of intestinal segment or sugar type.
Based on further analysis of the dominant microbiota at the phylum and genus levels in the honeybee gut, the dominant taxa were not different between the midgut and hindgut. However, the relative abundance of these groups was different between midgut and hindgut and was affected by sugar type. Before overwinter feeding, the abundance of Actinobacteria in the hindgut was significantly higher than that in the midgut (P< 0.05, Fig. 2A), but this distinction disappeared after the bees were fed honey and sucrose. The difference in abundance of Bacteroidetes between midgut and hindgut was also affected by sugar type. Before overwinter feeding, the abundance of Bacteroidetes was not significantly different between midgut and hindgut (P > 0.05). However, the relative abundance of Bacteroidetes in the hindgut was significantly higher than that in the midgut after the bees were fed honey (P < 0.05, Fig. 2B) and high-fructose syrup (P < 0.01, Fig. 2D) for overwintering. In general, the abundance of most dominant bacterial taxa was higher in the hindgut than in the midgut. The sucrose group was an exception, and the relative abundance of Bacteroidetes in the hindgut was significantly lower than that in the midgut (P < 0.01, Fig. 2C).
For the dominant genera of bacteria, the relative abundance of Lactobacillus and Bifidobacterium was significantly higher in the hindgut than in the midgut during prewinter and in the high-fructose syrup group (P < 0.05, Fig. 3AD). However, the difference in abundance of Lactobacillus and Bifidobacterium between the midgut and hindgut disappeared after the bees were fed honey and sucrose for overwintering (P > 0.05, Fig. 3BC). In general, the abundance of most dominant microbiota was higher in the hindgut than in the midgut.
The differences in gut microbiota of honeybees fed honey, sucrose, and high-fructose syrup
The microbial diversity in the midgut and hindgut of honeybees fed different types of sugars as winter food was analyzed using beta-diversity analysis. The microbial diversity of honeybee midguts was not significantly different among the honey, sucrose, and high-fructose syrup groups (P > 0.05, Fig. 4; Table S2). However, the microbial diversity of honeybee hindguts was affected by sugar type, with higher diversity in the hindguts of honeybees fed honey and high-fructose syrup than in those fed sucrose (P < 0.05, Fig. 4; Table S2).
From the clustering analysis of OTUs, Venn diagrams were created to analyze the common and unique OTUs among the three sugar groups. Among the three groups, 396 OTUs were common to the midguts (Fig. 5), and 870 OTUs were common to the hindguts (Fig. 6). As shown in Fig. 5, 878 OTUs were unique to the midgut of the sucrose group, which was more than in the honey (92 OTUs) and the high-fructose syrup (77 OTUs) groups. However, in the hindgut, only 89 OTUs were unique in the sucrose group, far less than in the honey (375 OTUs) and high-fructose syrup (230 OTUs) groups (Fig. 6). Thus, in the midgut, the bacterial diversity of the sucrose group was higher than that in the honey and high-fructose syrup groups, whereas in the hindgut, the bacterial diversity of the honey and high-fructose groups was higher than that in the sucrose group.
The detailed changes in the gut microorganisms of bees fed the different sugar types were analyzed using LEfSe. The contributions of the microorganisms to the differences among the groups were evaluated using the LDA score. The results for the midgut comparisons (HMG vs. FMG vs. SMG) are shown in Fig. 7, and the main taxa that were different were Betaproteobacteria (Neisseriales: Neisseriaceae), Actinobacteria (Bifidobacteriales: Bifidobacteriaceae), and Alphaproteobacteria (Rhizobiales and Mitochondria). The relative abundance of Betaproteobacteria (Neisseriales: Neisseriaceae) was the highest in the high-fructose syrup group, whereas in the bees fed sucrose, the relative abundance of Actinobacteria (Bifidobacteriales Bifidobacteriaceae) and Alphaproteobacteria (Rhizobiales and Mitochondria) increased. In the hindgut comparisons, the main taxa that were different among the three groups were the Bacteroidetes and Gammaproteobacteria (Pasteurellales) (Fig. 8), which increased in the honey group.
The differences in honeybee gut microbiota between prewinter and postfeeding for overwintering
The differences in gut microbes were analyzed between preoverwintering and postfeeding overwintering bees. The primary taxa that were different in the midguts of honeybees between prewinter and postfeeding for overwintering included Bacteroidetes, Clostridia (Lachnospiraceae and Ruminococcaeae), and Pasteurellales, independent of sugar type (Figs. 9–11). However, some sugar-specific bacterial taxa were identified in midguts between prewinter and postfeeding for overwintering bees (e.g., MG vs. HMG: Rhodospirillales Acetobacteraceae of Alphaproteobacteria was more abundant in MG; MG vs. SMG/FMG: Neisseriaceae of Betaproteobacteria was more prominent in MG). In addition, compared with MG, sucrose as the winter food increased the colonization of Actinobacteria, Lactobacillales (Lactobacillaceae), and Alphaproteobacteria (Xanthobacteraceae, Rickettsiales and Mitochondria) in the midgut of honeybees (Fig. 10).
The primary microbes in hindguts that were different between prewinter and postfeeding for overwintering included Actinobacteria (Bifidobacteriaceae), Bacteroidia (Bacteroidales: Bacteroidaceae), Clostridia (Clostridiales), and Alphaproteobacteria (Rhodospirillales: Acetobacteraceae). The relative abundance of these taxa was higher in HG than in HHG/SHG/FHG and did not vary among the different sugars. However, there were some sugar-special effects on some microbial taxa depending on sugar type. For example, honey fed as winter food increased the relative abundance of the Pasteurellales of Gammaproteobacteria in the hindgut (Fig. 12). High-fructose syrup as winter food increased the relative abundance of Gammaproteobacteria and Lactobacillus in the hindgut (Fig. 14). With the exception of the above different microbes between HG and SHG, the other taxa that were different between HG and SHG included the Bacilli (Lactobacillales), Betaproteobacteria (Neisseriaceae), and Alphaproteobacteria (Rhizobiales). The relative abundance of Alphaproteobacteria (Rhizobiales) was higher in SHG, whereas that of the others was higher in HG (Fig. 13).