Search

Here we report the discovery of Yaravirus, a new lineage of amoebal virus with a puzzling origin and phylogeny. Yaravirus presents 80 nm-sized particles and a 44,924 bp dsDNA genome encoding for 74 predicted proteins. More than 90% (68) of Yaravirus predicted genes have never been described before, representing ORFans. Only six genes had distant homologs in public databases: an exonuclease/recombinase, a packaging-ATPase, a bifunctional DNA primase/polymerase and three hypothetical proteins. Furthermore, we were not able to retrieve viral genomes closely related to Yaravirus in 8,535 publicly available metagenomes spanning diverse habitats around the globe. The Yaravirus genome also contained six types of tRNAs that did not match commonly used codons. Proteomics revealed that Yaravirus particles contain 26 viral proteins, one of which potentially representing a novel capsid protein with no significant homology with NCLDV major capsid proteins but with a predicted double-jelly roll domain. Yaravirus expands our knowledge of the diversity of DNA viruses. The phylogenetic distance between Yaravirus and all other viruses highlights our still preliminary assessment of the genomic diversity of eukaryotic viruses, reinforcing the need for the isolation of new viruses of protists.Significance statementMost of the known viruses of amoeba have been seen to share many features that eventually prompted authors to classify them into common evolutionary groups. Here we describe Yaravirus, an entity that could represent either the first isolated virus of Acanthamoeba spp. out of the group of NCLDVs or, in alternative evolutive scenario, it is a distant and extremely reduced virus of this group. Contrary to what is observed in other isolated viruses of amoeba, Yaravirus is not represented by a large/giant particle and a complex genome, but at the same time carries an important number of previously undescribed genes, including one encoding a novel major capsid protein. Metagenomic approaches also testified for the rarity of Yaravirus in the environment.

Enzymatic proteins with a metallo-beta-lactamase (MBL) fold have been essentially studied in bacteria for their activity on beta-lactam antibiotics. However, the MBL fold is ancient and highly conserved, and these proteins are capable of cleaving a broad range of substrates. It has recently been shown that MBLs are present in a wide array of cellular organisms, including eukaryotes and archaea. We show here that Tupanvirus deep ocean, a giant virus, also encodes a protein with a MBL fold. Phylogeny showed its clustering with transfer ribonucleases (RNases) and the presence of orthologs in other giant viruses, mainly those harboring the largest sets of translation components. In addition, it suggests an ancient origin for these genes and a transfer between giant viruses and Acanthamoeba spp., a host of many giant viruses. Biologically, after its expression in Escherichia coli, the tupanvirus protein was found to hydrolyse nitrocefin, a chromogenic beta-lactam. We also observed an hydrolysis of penicillin G (10 μg/mL) and detected the metabolite of penicillin G hydrolysis, benzylpenilloic acid. This was inhibited by sulbactam, a beta-lactamase inhibitor. In addition, we tested the degradation of single-stranded DNA, double-stranded DNA, and RNAs, and observed a strong activity on RNAs from seven bacteria with G+C varying from 42% to 67%, and from Acanthamoeba castellanii, the tupanvirus host. This was not inhibited by sulbactam or ceftriaxone. RNase activity was estimated to be 0.45±0.15 mU/mg using a fluorescence-based assay. Our results still broaden the range of hosts of MBL fold proteins and demonstrate that such protein can have dual beta-lactamase/nuclease activities. We suggest that they should be annotated according to this finding to avoid further confusion.

β-lactams targeting the bacterial cell wall are not active on archaea. Here, we figure out that annotation of genes as β-lactamase in Archeae on the basis of homologous genes, initially annotated β-lactamases, is a remnant of the identification of the original activities of this group of enzymes, which in fact, have multiple functions including nuclease, ribonuclease, β-lactamase, or glyoxalase; which may specialized over time. We expressed a class B β-lactamase enzyme from Methanosarcina barkeri that digest penicillin G. Moreover, while a weak glyoxalase activity was detected, a significant ribonuclease activity on bacterial and synthetic RNAs was demonstrated. The β-lactamase activity was inhibited by a β-lactamase inhibitor (sulbactam), but its RNAse activity was not. This gene appears to has been transferred to the Flavobacteriaceae group including Elizabethkingia genus in which the expressed gene shows a more specialized activity toward resistance to tienanmicin but no glyoxalase activity. The expressed class C-like β-lactamase gene, also from Methanosarcina sp., shows also hydrolysis activity and was more closely related to DD-peptidase enzymes than known bacterial class C β-lactamases. Our findings highlight the requalification needness of annotated enzymes as β-lactamases and the specification overtime of multipotent enzymes in different ways in Archaea and bacteria.