Isolation and characterization of mollicute symbionts from a fungus-growing ant reveals high niche overlap leading to co-exclusion

Abstract

Two mollicute species belonging to the Mesoplasma and Spiroplasma genera have been detected in several species of fungus-growing ants using molecular methods. However, their ecological roles remain largely inferred from metagenomic data. To better understand their diversity and specialization, we cultured both of these Mesoplasma and Spiroplasma symbionts from the fungus-growing ant Trachymyrmex septentrionalis, providing the first isolated mollicutes from any fungus-growing ant species. The genomes of our isolates and related metagenome-assembled genomes (MAGs) from T. septentrionalis fungus gardens comprise two unique phylogenetic lineages compared to previously described Mesoplasma and Spiroplasma species, and from related MAGs previously sequenced from the leaf-cutting ant Acromyrmex echinatior. This suggests that the T. septentrionalis symbionts comprise undescribed species that can exclude each other from a niche that is largely shared between them. Mesoplasma genomes and MAGs also demonstrate regional specificity with their T. septentrionalis ant hosts. Both Mesoplasma and Spiroplasma strains from T. septentrionalis can catabolize glucose and fructose; both sugars are common in the ant's diet. Similarly, both these Mesoplasma and Spiroplasma can catabolize arginine, but only Mesoplasma can catabolize N-acetylglucosamine; both could produce ammonia for the ants or fungus garden. Based on our genomic and phenotypic analyses, we describe these T. septentrionalis symbionts as Mesoplasma whartonense sp. nov. and Spiroplasma attinicola sp. nov., providing insight into their genomic and phenotypic diversity and cultures to facilitate future studies of how these common but poorly understood members of the fungus-growing ant symbiosis separately colonize different ant colonies despite having highly overlapping niches.

Importance: Fungus-growing ants partner with multiple microbial symbionts to obtain food and remain free from disease. Of these symbionts, those inhabiting the ant gut remain the least understood and are known only from environmental surveys. Such surveys can infer potential functions of gut symbionts, but cultures are required to experimentally validate these hypotheses. Here, we describe the first cultures of the ant gut symbionts of the fungus-growing ant Trachymyrmex septentrionalis, using comparative genomics and phenotypic experiments to describe them as two novel species: Mesoplasma whartonense sp. nov. and Spiroplasma attinicola sp. nov. This genomic analysis suggests that these species are highly specialized to T. septentrionalis and are distinct from related environmental data generated from the related ant species Acromyrmex echinatior, implying substantial host specificity. Our phenotypic experiments and genomic reconstructions highlight the highly overlapping niches and likely costs and benefits of these symbionts to their ant host, setting the stage for further experimentation.

Keywords: Mesoplasma; Spiroplasma; comparative genomics; fungus-growing ants; host specialization; symbiosis; taxonomy.

Fig 1

Anvi’o pangenome analysis of (A) Mesoplasma whartonense genomes (bold) and MAGs (*), the EntoAcro1 MAG (*), and the M. lactucae genomes as a reference and (B) Spiroplasma attinicola genomes (bold), the EntAcro10 MAG (*), and the S. platyhelix genome as a reference. Text colors indicate strains or MAGs from the same state. Two M. lactucae genomes are shown, with the ^ indicating the genome downloaded from NCBI (SRR6082029) vs the other that we re-sequenced as a control (JKS002656). In the central heatmap, black and gray bars indicate homologous genes that are present or absent in each genome. The dendrogram clusters groups of homologous genes (anvi’o gene clusters) using Euclidean distances.

Fig 2

(A) Mesoplasma nucleotide phylogeny created using 550 core gene clusters as input data and the FastTree JC + CAT substitution model. The tree was rooted using the midpoint of the branch leading to the M. lactucae genomes. The ^ indicates the NCBI (SRR6082029) genome. (B) Spiroplasma nucleotide phylogeny created using 684 core gene clusters as input data and the FastTree JC + CAT substitution model. The tree was rooted using the midpoint of the branch leading to the S. platyhelix genome. Asterisks indicate MAGs and bolded names indicate T. septentrionalis isolate genomes. Colors group strains by state. Local bootstrap values of 60–79 and 80–100 are indicated by * and **, respectively.

Fig 3

Average nucleotide identity heatmap of (A) M. whartonense genomes and MAGs, MAG EntAcro1, and Mesoplasma reference genomes, and (B) S. attinicola genomes, MAG EntAcro10, and Spiroplasma reference genomes, produced by anvi’o. Asterisks indicate MAGs, and bold names indicate our isolate genomes. The ^ indicates the NCBI (SRR6082029) genome, and colors indicate the state from which strains or MAGs were collected.

Fig 4

Representative gene function predictions for ant-associated Spiroplasma and Mesoplasma, as predicted using RAST. Orange colors indicate genes present for Spiroplasma, and blue indicates genes present for Mesoplasma. The lighter of each color denotes genes present in MAGs EntAcro1 or EntAcro10. White indicates absence.

Fig 5

Representative catabolic pathways in M. whartonense (left) and S. attinicola (right). Anabolic reactions are not shown, and reactions are not balanced. Major predicted catabolic substrates and products are bolded.

Fig 6

Results of phenotypic tests and growth at varying temperatures. Orange colors indicate phenotypes observed for Spiroplasma, and blue indicates phenotypes observed for Mesoplasma. The lighter of each color denotes phenotypes observed for the control strain, S. platyhelix or M. lactucae. White indicates no phenotype was observed.

Fig 7

Transmission electron microscope imaging of M. whartonense strain JKS002660 (A) and S. attinicola strain JKS002670 (B).