Virus-like Particles from Wolbachia-Infected Cells May Include a Gene Transfer Agent

Abstract

Wolbachia are obligate intracellular bacteria that occur in insects and filarial worms. Strains that infect insects have genomes that encode mobile genetic elements, including diverse lambda-like prophages called Phage WO. Phage WO packages an approximately 65 kb viral genome that includes a unique eukaryotic association module, or EAM, that encodes unusually large proteins thought to mediate interactions between the bacterium, its virus, and the eukaryotic host cell. The Wolbachia supergroup B strain, wStri from the planthopper Laodelphax striatellus, produces phage-like particles that can be recovered from persistently infected mosquito cells by ultracentrifugation. Illumina sequencing, assembly, and manual curation of DNA from two independent preparations converged on an identical 15,638 bp sequence that encoded packaging, assembly, and structural proteins. The absence of an EAM and regulatory genes defined for Phage WO from the wasp, Nasonia vitripennis, was consistent with the possibility that the 15,638 bp sequence represents an element related to a gene transfer agent (GTA), characterized by a signature head-tail region encoding structural proteins that package host chromosomal DNA. Future investigation of GTA function will be supported by the improved recovery of physical particles, electron microscopic examination of potential diversity among particles, and rigorous examination of DNA content by methods independent of sequence assembly.

Keywords: +1 ribosomal frameshift; GH_25 hydrolase; GTA; alpha-proteobacteria; bacteriophage; horizontal gene transfer; mosquito cell culture.

Figure 1

Combined SPAdes 2016 and 2020 reads mapped to Phage WO (KX522565). Panel A. Geneious Prime “Map to reference” mapped 523 of 3499 contigs generated without a coverage cutoff to Phage WO (KX522565). Introduction of gaps to maximize the alignment lengthens the consensus from 65,653 nt to 115,077 nt. The open arrow indicates the coverage profile (Log scale, shaded in gray) of mapped reads, and black bars beneath the profile indicate coverage with 2 or more reads. Annotation of Phage WO is indicated by gray arrows beneath the black bar, with proximal, EAM, and distal regions indicated below the alignment. Note that short areas of two-fold coverage (circled) occur within (A) and just downstream (B) of the KX522565 sequence.

Figure 2

Structure of WOStri. Open reading frames were produced in Geneious Prime. Genes clearly identifiable as phage-related are listed above, and unknown genes are labeled below the alignment.

Figure 3

Overall sequence comparisons. (A) Pairwise alignment of WOStri 11.1_1 with residues 1–14,268 of KX522565. Green/yellow-green and red shading indicate Geneious designations of high, medium and low sequence identity. Gray (WOStri) and white (WOVitA1) horizontal arrows represent orfs and their direction of transcription. Triangles below the orfs indicate exact repeats of 50 nt or greater common to WOStri and the proximal region of KX522565. The double black arrow connecting Panels A and B indicates the variable region separating terminase/gpw and portal genes in Wolbachia prophages. (B) Sequence alignment of WOCauB2 (AB478515, bases 29,351–43,133); WOCauB3 (AB478516, bases 26,079–40,231); WOStri and WOKue (AB036666, bases 10,706–29,095); WOVitA1 (HQ906662, reversed, bases 49,358–64,760); and WOVitA4 (HQ906664; reversed, bases 3035–21,167). The region common to all six genomes was extracted and aligned in Geneious Prime.

Figure 4

Sequence support for a programmed + 1 ribosomal frameshift in hydrolase-encoding orfs. Panel (A) shows the sequence of the hydrolase Orf6 and Orf7 in the region of the potential frameshift. Rectangular boxes show initiation and termination codons of Orf6, and oval boxes show the start and stop codons of Orf7. The downward pointing arrow indicates the potential + 1 frameshift. Codons in the Orf6 reading frame are separated by underscores, and codons in the Orf7 reading frame are separated by periods. The 3′-end of Orf7 is not shown. Panel (B) shows a translation of GH_25 Orf6/Orf7 using Geneious Prime. Colors are the Geneious designations for residues and are not relevant here. Large solid black bars indicate the two translation products: N-terminal MREY… and MEYW…, with WP_ accessions. Short arrows below indicate positions of 9 and 10 bp repeats in the Orf6/Orf7 region, suggesting possible secondary structure in the mRNA. Horizontal lines flanked by asterisks show sequence flanked by repeat pairs 2, 3, 7, and 9, oriented in opposite directions. The white vertical arrow indicates a 10 nt palindrome. Panel (C) shows alignment of hydrolase reading frames in WOStri and WOVitA4. Dots indicate identities. Small rectangular boxes with only three nucleotides represent ATG start and TAA stop codons in Orf6. The region of exact identity immediately surrounding the out-of-frame Orf7 ATG (downward pointing arrow and white text on a black background) is shown in the long horizontal rectangle that terminates with the Orf6 TAA stop codon. Within the box, a slippery sequence AAATTTT upstream of the start codon is shown in italics and underlined. The oval represents the Orf7 stop codon.

Figure 5

Mapping of combined SPAdes contigs without coverage cutoff to MUIX_01.1 (Panel (A)) and MUIX_02.1 (Panel (B)) contigs representing an incomplete wStri genome. Green/yellow-green and red shading in consensus panels indicate Geneious designations of high, medium and low sequence identity. Of 3499 total reads, 1193 mapped to MUIX_01.1, and 515 mapped to MUIX_02.1. Profiles show consensus identities generated with Geneious Prime, with highlighted coverages set to >0 (highest possible representation of matches between contigs and wStri DNA) and to >5, represented by black bars below the top and lower consensus bars, respectively. Distribution of coverage bars emphasizes potential coverage of most of the wStri genome with the complete set of SPAdes contigs, and loss of matches as Geneious coverage criteria increase.