Origin and effect of Alpha 2.2 Acetobacteraceae in honey bee larvae and description of Parasaccharibacter apium gen. nov., sp. nov

Abstract

The honey bee hive environment contains a rich microbial community that differs according to niche. Acetobacteraceae Alpha 2.2 (Alpha 2.2) bacteria are present in the food stores, the forager crop, and larvae but at negligible levels in the nurse and forager midgut and hindgut. We first sought to determine the source of Alpha 2.2 in young larvae by assaying the diversity of microbes in nurse crops, hypopharyngeal glands (HGs), and royal jelly (RJ). Amplicon-based pyrosequencing showed that Alpha 2.2 bacteria occupy each of these environments along with a variety of other bacteria, including Lactobacillus kunkeei. RJ and the crop contained fewer bacteria than the HGs, suggesting that these tissues are rather selective environments. Phylogenetic analyses showed that honey bee-derived Alpha 2.2 bacteria are specific to bees that "nurse" the hive's developing brood with HG secretions and are distinct from the Saccharibacter-type bacteria found in bees that provision their young differently, such as with a pollen ball coated in crop-derived contents. Acetobacteraceae can form symbiotic relationships with insects, so we next tested whether Alpha 2.2 increased larval fitness. We cultured 44 Alpha 2.2 strains from young larvae that grouped into nine distinct clades. Three isolates from these nine clades flourished in royal jelly, and one isolate increased larval survival in vitro. We conclude that Alpha 2.2 bacteria are not gut bacteria but are prolific in the crop-HG-RJ-larva niche, passed to the developing brood through nurse worker feeding behavior. We propose the name Parasaccharibacter apium for this bacterial symbiont of bees in the genus Apis.

FIG 1

The distribution of bacterial taxa in royal jelly (RJ), crops, and hypopharyngeal glands (HGs) of nurse workers. The proportion of sequences belonging to each bacterial taxon was determined relative to the number of sequences in each individual sequencing library. The number of sequences from each library is given at the right of each group. Bacterial taxa boxed in black are members of the core gut microbiome. The Lactobacillales and Acetobacteraceae clades marked with asterisks represent the remaining sequences within these clades after Lactobacillus kunkeei and Alpha 2.2 bacteria were accounted for and indicated elsewhere in the graph. L. Firm5, Lactobacillus sp. Firm 5; L. Firm4, Lactobacillus sp. Firm 5; Bifido, Bifidobacterium sp.

FIG 2

The percentage of Alpha 2.2 and Lactobacillus kunkeei sequences in nurse worker crops, hypopharyngeal glands (HGs), and royal jelly (RJ). The average percentages of sequences in the sequence libraries from each sample type are shown for Alpha 2.2 and L. kunkeei. Alpha 2.2 bacteria were represented equally in all sample types. L. kunkeei was more prevalent in the RJ than in HGs (indicated with a line connecting the significant comparison). Although the levels of L. kunkeei bacteria appeared higher in the RJ than in the crops, the difference was not statistically significant.

FIG 3

The diversity and number of bacterial OTUs in the hypopharyngeal glands (HG), royal jelly (RJ), and crop samples. (A) The mean inverse Simpson index ± SE is plotted for each sample type (HGs, RJ, or crop). (B) The mean number of 97% OTUs ± SE found in each sample type. Post hoc analyses yielding significant differences among sample types are indicated with a line.

FIG 4

The number of taxa and sequences shared among hypopharyngeal glands (HG), royal jelly (RJ), and nurse crops (C). OTUs are defined based on 97% sequence similarity. The numbers of OTUs are indicated, and below them are the percentages of sequences of each sample type comprising the respective OTU group.

FIG 5

Neighbor-joining phylogeny of larval isolates relative to the Alpha 2.2 OTUs identified via high-throughput 16S rRNA gene sequencing. The V1 region of the 16S rRNA gene sequence was used to compare the representative sequences from the predominant Alpha 2.2 OTUs identified using high-throughput sequencing with the three isolates cultured from 1st-instar larvae that were used in the in vitro rearing experiments. Each Alpha 2.2 isolate is labeled with its OTU number (see Table S2 in the supplemental material) and its representative sequence title. Isolates found in larvae only are members of groups A, B, and C as described in the legend of Fig. S3 in the supplemental material. The yellow star represents the Alpha 2.2 strain (strain C6) that increased larval survival. Evolutionary distances were computed using the maximum composite likelihood method and are represented as the number of base substitutions per site. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) is shown next to the branches. Only bootstrap values of ≥60% are shown. Black asterisks (*) indicate reference sequences that are the best BLAST hits to OTUs associated with increased Crithidia infection in bumblebees (26). HG, hypopharyngeal gland; RJ, royal jelly.

FIG 6

Acetobacteraceae Alpha 2.2 bacteria flourish in the presence of royal jelly compared to nonhive bacteria. (A) Three Alpha 2.2 strains and E. coli were plated onto SDA and LB medium, respectively. A royal jelly disk was added to the newly plated culture. (B) CFU of Alpha 2.2 strain C6 and E. coli per 100 μl of sample from culture (bacterial growth media) and from 100 ml of the larval diet incubated overnight at 34°C with 300 CFU of bacteria (larval diet). Error bars represent the standard error around the mean number of CFU for five replicate samples.

FIG 7

Acetobacteraceae Alpha 2.2 bacteria increase larval survival. The percentage of larvae surviving to the last larval instar when the larval diet was supplemented with Alpha 2.2 strain A29, B8, or C6, E. coli, or no bacteria (negative) is shown (n = 48 for each treatment per trial). Black lines indicate significantly different planned comparisons between the Alpha 2.2 treatments and the E. coli and negative-control treatments.

FIG 8

Taxonomy of Alpha 2.2 (Parasaccharibacter apium) isolates based on a longer 16S rRNA gene sequence. The regions of V1 through V8 of the 16S rRNA gene sequence were used to infer the taxonomy of the three Alpha 2.2 isolates cultured from 1st-instar larvae that were used in the in vitro rearing experiments. Larval isolates are members of groups A, B, and C as described in the legend of Fig. S3 in the supplemental material. A total of 1,306 bp was used to construct the original alignment. Evolutionary relationships were inferred using the neighbor-joining method. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) is shown next to the branches. Only bootstrap values of ≥60% are shown. The type strain for the newly named Parasaccharibacter apium, Alpha 2.2 isolate A29, is boxed in red.