Numerous cultivated and uncultivated viruses encode ribosomal proteins

Abstract

Viruses modulate ecosystems by directly altering host metabolisms through auxiliary metabolic genes. However, viral genomes are not known to encode the core components of translation machinery, such as ribosomal proteins (RPs). Here, using reference genomes and global-scale viral metagenomic datasets, we identify 14 different RPs across viral genomes arising from cultivated viral isolates and metagenome-assembled viruses. Viruses tend to encode dynamic RPs, easily exchangeable between ribosomes, suggesting these proteins can replace cellular versions in host ribosomes. Functional assays confirm that the two most common virus-encoded RPs, bS21 and bL12, are incorporated into 70S ribosomes when expressed in Escherichia coli. Ecological distribution of virus-encoded RPs suggests some level of ecosystem adaptations as aquatic viruses and viruses of animal-associated bacteria are enriched for different subsets of RPs. Finally, RP genes are under purifying selection and thus likely retained an important function after being horizontally transferred into virus genomes.

Fig. 1

Virus-encoded ribosomal protein bS21. a Structure of the Escherichia coli 30S ribosomal subunit (PDB id: 4ADV). 16S ribosomal RNA is shown as blue ribbon. bS21 ribosomal protein is highlighted in pink. b Alignment of the ribosomal protein bS21 encoded by pelagiphage HTVC008M with homologs from representatives of distinct bacterial taxa and environmental sequences obtained from the Global Ocean Sampling (GOS) dataset. c Phylogenetic tree of ribosomal protein bS21. Taxonomic affiliations are represented by colored circles (see panel (b) legend)

Fig. 2

Detection of ribosomal proteins in uncultivated viral genomes. For each ribosomal protein detected, the total number of detection is shown on the y-axis (log₁₀ scale), and the bar is colored according to the type of samples in which this protein was detected (the sizes of the colored parts are proportional to the number of detections made in each type of samples). Ribosomal proteins also identified in cultivated viruses are identified with stars

Fig. 3

Ribosomal proteins bS21 identified in uncultivated viral genomes. Top: Phylogenetic tree of ribosomal protein bS21. Viral sequences are highlighted with blue branches. Bottom: Genome comparison of viral contigs encoding ribosomal protein bS21. Comparisons were done at the amino acid level, with the % identity displayed with a color scale. The predicted bS21-encoding genes are indicated in red. The position of these sequences in the tree (top panel) is indicated with colored circles

Fig. 4

Ribosome analysis of extracts from NM522 Escherichia coli cells. a Sedimentation profiles of NM522 E. coli ribosomes. Wild-type NM522 E. coli cells (black curve) and cells expressing viral bS21 (red), bL12 (green) or HPF (blue) were lysed and their ribosomes were purified using a 10–50% sucrose gradient (see experimental section). The dotted lines indicate the fractions that were pooled and further analyzed by mass spectrometry. b Quantification of in vitro translation of GFP by E. coli 70S ribosomes carrying either E. coli wt bS21 (control), E. coli streptavidin-tagged bS21 (70S-eS21) or viral streptavidin-tagged bS21 (70S-pS21). Translation assay was performed using PURExpress® ΔRibosome Kit, complemented with 10 pmol of purified ribosomes and 250 ng of a PCR product encoding for GFP under control of T7 promoter. Fluorescence signal was detected by spectrofluorimetry at 510 nm with an excitation at 485 nm. The percentage of fluorescence was measured with respect to the translation control. The error bar represents the standard deviation measured over three independent experiments. The variance was analyzed using Kruskal–Wallis test followed by a Dunn’s multiple comparison test (p-value for 70S-pS21 = 0.0219)