Functional convergence in reduced genomes of bacterial symbionts spanning 200 My of evolution

Abstract

The main genomic changes in the evolution of host-restricted microbial symbionts are ongoing inactivation and loss of genes combined with rapid sequence evolution and extreme structural stability; these changes reflect high levels of genetic drift due to small population sizes and strict clonality. This genomic erosion includes irreversible loss of genes in many functional categories and can include genes that underlie the nutritional contributions to hosts that are the basis of the symbiotic association. Candidatus Sulcia muelleri is an ancient symbiont of sap-feeding insects and is typically coresident with another bacterial symbiont that varies among host subclades. Previously sequenced Sulcia genomes retain pathways for the same eight essential amino acids, whereas coresident symbionts synthesize the remaining two. Here, we describe a dual symbiotic system consisting of Sulcia and a novel species of Betaproteobacteria, Candidatus Zinderia insecticola, both living in the spittlebug Clastoptera arizonana. This Sulcia has completely lost the pathway for the biosynthesis of tryptophan and, therefore, retains the ability to make only 7 of the 10 essential amino acids. Zinderia has a tiny genome (208 kb) and the most extreme nucleotide base composition (13.5% G + C) reported to date, yet retains the ability to make the remaining three essential amino acids, perfectly complementing capabilities of the coresident Sulcia. Combined with the results from related symbiotic systems with complete genomes, these data demonstrate the critical role that bacterial symbionts play in the host insect's biology and reveal one outcome following the loss of a critical metabolic activity through genome reduction.

FIG. 1.—

Transmission electron microscopy of a Zinderia-containing bacteriocyte cell from Clastoptera arizonana. Three insect nuclei are indicated with white asterisks. The scale bar is 10 μm.

FIG. 2.—

Whole-genome alignment for three Sulcia species. Each complete Sulcia genome is shown in a linear representation where each gene is represented by a box. Boxes for genes involved in the synthesis of essential amino acids are colored red; all others are represented by white boxes except for selected genes flanking the tryptophan biosynthetic pathway region. (a) The histogram above the linear genome schematic indicates the level of conservation, where a higher bar represents greater sequence identity. Regions that are shared between all three genomes are colored orange; those that are shared between Sulcia-CARI and Sulcia-DSEM are green; those shared between Sulcia-CARI and Sulcia-GWSS are blue; and those shared between Sulcia-DSEM and Sulcia-GWSS are purple. The region flanking the tryptophan biosynthetic pathway is shaded in gray. (b) Zooming in on the tryptophan biosynthesis region shows the precise nature of genome reduction, as no fragment of any gene in the tryptophan pathway remains.

FIG. 3.—

Phylogenetic analysis indicates that Zinderia is a Betaproteobacteria in the family Oxalobacteraceae. Select bootstrap values greater than 70 are shown on each maximum likelihood tree. The tree in (a) was calculated from a concatenated alignment of several protein sequences and that in (b) was calculated from the 16S rDNA sequence alone.

FIG. 4.—

Alignment of conserved regions of proteins suggests that UGA codes for tryptophan in the Zinderia genome. The numbers indicate the position in the Zinderia protein. Zinderia, Janthinobacterium (GenBank accession: NC_009659), Herminiimonas (NC_009138), Burkholderia (NC_008784, NC_008785), and Neisseria (NC_003112) are all Betaproteobacteria; Sinorhizobium (NC_003047) is an Alphaproteobacteria and Escherichia (NC_000913) is a Gammaproteobacteria.

FIG. 5.—

Amino acid frequency distributions for six bacterial genomes reveal extreme biases in GC-poor symbiont genomes. The asterisk on TGA indicates that this codon has been reassigned to tryptophan in the Hodgkinia and Zinderia genomes. The most GC-poor codons (e.g., phenylalanine [F], isoleucine [I], lysine [K], aparagine [N], and tyrosine [Y]) are all overrepresented in reduced symbiont genomes that are GC-poor such as Zinderia, Carsonella, and Buchnera. The opposite pattern of GC-rich codons being overrepresented in tiny GC-rich symbiont genomes such as Hodgkinia is apparent in some (e.g., alanine [A]) but not all (e.g., proline [P] and glutamine [Q]) codon families.

FIG. 6.—

The essential amino acid metabolisms of three Sulcia-containing dual symbiont systems. Complete pathways for the production of essential amino acids (red font) or related cofactor compounds (blue font; Hodgkinia uses the B12-dependent version of methionine synthase in the last step of methioine production [McCutcheon et al. 2009a]) are shown. Missing genes are represented in a light gray font. Note the three different pathways taken in the production of methionine in Sulcia’s cosymbionts.