Evolutionary trends in Bombella apis CRISPR-Cas systems

Abstract

Bacteria and archaea employ a rudimentary immune system, CRISPR-Cas, to protect against foreign genetic elements such as bacteriophage. CRISPR-Cas systems are found in Bombella apis. B. apis is an important honey bee symbiont, found primarily in larvae, queens, and hive compartments. B. apis is found in the worker bee gut but is not considered a core member of the bee microbiome and has therefore been understudied with regard to its importance in the honey bee colony. However, B. apis appears to play beneficial roles in the colony, by protecting developing brood from fungal pathogens and by bolstering their development under nutritional stress. Previously, we identified CRISPR-Cas systems as being acquired by B. apis in its transition to bee association, as they are absent in a sister clade. Here, we assess the variation and distribution of CRISPR-Cas types across B. apis strains. We found multiple CRISPR-Cas types, some of which have multiple arrays, within the same B. apis genomes and also in the honey bee queen gut metagenomes. We analyzed the spacers between strains to identify the history of mobile element interaction for each B. apis strain. Finally, we predict interactions between viral sequences and CRISPR systems from different honey bee microbiome members. Our analyses show that the B. apis CRISPR-Cas systems are dynamic; that microbes in the same niche have unique spacers, which supports the functionality of these CRISPR-Cas systems; and that acquisition of new spacers may be occurring in multiple locations in the genome, allowing for a flexible antiviral arsenal for the microbe.

Importance: Honey bee worker gut microbes have been implicated in everything from protection from pathogens to breakdown of complex polysaccharides in the diet. However, there are multiple niches within a honey bee colony that host different groups of microbes, including the acetic acid bacterium Bombella apis. B. apis is found in the colony food stores, in association with brood, in worker hypopharyngeal glands, and in the queen's digestive tract. The roles that B. apis may serve in these environments are just beginning to be discovered and include the production of a potent antifungal that protects developing bees and supplementation of dietary lysine to young larvae, bolstering their nutrition. Niche specificity in B. apis may be affected by the pressures of bacteriophage and other mobile elements, which may target different strains in each specific bee environment. Studying the interplay between B. apis and its mobile genetic elements (MGEs) may help us better understand microbial community dynamics within the colony and the potential ramifications for the honey bee host.

Keywords: Bombella; CRISPR; Cas; defense systems; honey bee; microbiome; phage; royal jelly.

Fig 1

CRISPR systems found across Acetobacteraceae. Cladogram depicting Bombella apis and related Bombella species with reference to other non-insect-associated Acetobacteraceae such as Swingsia, Neokomagataea, and Gluconobacter oxydans. The tree was generated using the Gloome software with orthologs shared among these genomes. Columns indicate the presence of different types of CRISPR arrays, with the number of arrays indicated by shades of gray, with darker colors meaning more arrays. CRISPR arrays are further broken down for type I-E by the specific tetraloops in the repeats, as well as whether the array occurs as an orphan (no nearby Cas genes). The “Other” column can include an array of any type other than those specified by other columns.

Fig 2

Schematic model for CRISPR systems found in B. apis genomes. Type I-E and type II-C CRISPR systems are both found in most B. apis genomes, with two arrays for type I-E. (A) Typical layout for genes in the CRISPR system arranged as an operon, followed by a leader sequence and then the repeat-spacer array. (B) Secondary type I-E array prevalent in most B. apis genomes, called “orphan” here as no Cas genes are present nearby. (C) Typical type II-C CRISPR system, including the tracrRNA and operon, followed by the array. (D) Structure of the direct repeat found in each array. Nucleotides highlighted in yellow differ between canonical and orphan arrays. The sample type II-C repeat shown is from the B. apis strain SME1.

Fig 3

Array sizes differ among Bombella genomes. (A) Canonical CRISPR arrays immediately follow their associated Cas operon and tend to have fewer spacers than orphan arrays (no nearby Cas genes). (B) There is a difference in the average size of type I-E CRISPR arrays and the smaller type II-C CRISPR arrays in Bombella genomes.

Fig 4

Bombella strains do not share many CRISPR spacers, despite genome-wide identity. Numbers shown in the heatmap indicate the number of spacers with at least 80% sequence homology. Color of the map indicates the average nucleotide identity (ANI) between genomes. Note that the color scale is skewed to better visually resolve B. apis strains, which all share over 99% identity. Cladogram representation of ANI is displayed on top. To the left of the heatmap is a histogram depicting the total number of spacers in each genome.

Fig 5

Matches between spacer sequences and prophage regions. Arrows point toward the genome containing the predicted prophage. The thickness of the lines depicts how many spacer matches were found, while the color of the lines shows the average similarity in %matching nucleotides. (A) Intact prophage regions predicted by Phaster. (B) Incomplete/questionable prophage regions predicted by Phaster.

Fig 6

Bioinformatics workflow schematic. Genomes were sourced from public repositories and analyzed using CRISPROne to predict CRISPR regions. Metagenomic reads were analyzed with crass to pull out spacers and repeats. These spacers were mapped to known viruses from previous bee studies and to the genomes and other spacers; after cleaning up, the matches were visualized in Cytoscape.