Analysis of virion structural components reveals vestiges of the ancestral ichnovirus genome

Abstract

Many thousands of endoparasitic wasp species are known to inject polydnavirus (PDV) particles into their caterpillar host during oviposition, causing immune and developmental dysfunctions that benefit the wasp larva. PDVs associated with braconid and ichneumonid wasps, bracoviruses and ichnoviruses respectively, both deliver multiple circular dsDNA molecules to the caterpillar. These molecules contain virulence genes but lack core genes typically involved in particle production. This is not completely unexpected given that no PDV replication takes place in the caterpillar. Particle production is confined to the wasp ovary where viral DNAs are generated from proviral copies maintained within the wasp genome. We recently showed that the genes involved in bracovirus particle production reside within the wasp genome and are related to nudiviruses. In the present work we characterized genes involved in ichnovirus particle production by analyzing the components of purified Hyposoter didymator Ichnovirus particles by LC-MS/MS and studying their organization in the wasp genome. Their products are conserved among ichnovirus-associated wasps and constitute a specific set of proteins in the virosphere. Strikingly, these genes are clustered in specialized regions of the wasp genome which are amplified along with proviral DNA during virus particle replication, but are not packaged in the particles. Clearly our results show that ichnoviruses and bracoviruses particles originated from different viral entities, thus providing an example of convergent evolution where two groups of wasps have independently domesticated viruses to deliver genes into their hosts.

Figure 1. Map of the three IVSPERs identified in H. didymator genome and of their flanking sequences.

GenBank accession numbers for BAC clones BQ, BR and BT are GQ923581, GQ923582 and GQ923583 respectively. The IVSPERs were defined as starting at the first ATG and ending at the last stop codon of the coding sequences cluster. Genes of a given family are represented by arrows of the same color. U = Unknown protein encoding gene, IVSP = gene family encoding IV Structural Protein (from IVSP1 to IVSP4). Dark blue arrows correspond to wasp genes: HYP PROTEIN = Hypothetical protein (gi|158297979| in BQ and gi|156554181 in BT); XRCC1 = DNA repair protein (gi|156554771|); Myosin (gi|156553326|); Prohibitin (gi|156538068|); PPI = phosphoglucose isomerase (gi|156554183|); albumin (gi|156543451|). The unknown proteins in clone BQ are sequences with no significant similarity to NCBI nr database entries. Proviral HdIV segments SH-BQ and SH-BR are indicated by red lines.

Figure 2. HdIV particle proteins identified by mass spectrometry and sequence searches using the translated sequences of the three IVSPER genes (16 bands were analyzed).

Names and predicted molecular masses (see Table S3) of IVSPER proteins are shown beside the bands. Upper right panel shows a negative staining of HdIV virions.

Figure 3. Specific IVPER gene expression in H. didymator calyx cells.

The fold changes in transcript levels (i.e. N0 values) between calyx tissue and ovarioles are given for the IVPER genes. The calyx tissues were separated from the ovarioles under microscope as shown in the picture in the upper right panel. Amongst the genes shown by LC MS/MS to encode proteins (see Figure 2), some were not analyzed (Not Determined, ND). The wasp gene XRCC1, 1.2 kb upstream of IVSPER-1, was used as a negative control for transcriptional analyses. Gene names are indicated in Figure 1 (I1 to I4 correspond to IVSP1 to IVSP4).

Figure 4. Fold change in genomic DNA amount between calyx tissue and H. didymator larvae for a subset of IVSPERs sequences.

Results are given for calyx cells dissected 2 and 24 h after emergence of the adult wasps from the cocoons, as indicated. Viral sequences packaged in the particles (N-gene, SH-BQ; vinnexin, Vx) and wasp genes (XRCC1 located 1.2 kb upstream of IVSPER-1, ribosomal protein L55 and elongation factor ELF1) were used as positive and negative controls of DNA amplification during particle production, respectively.

Figure 5. Schematic representation of the region in H. didymator IVSPER-2 displaying similarity to Campoletis sonorensis IV segment SH-C (NCBI Reference: NC_007986.1).

Five related sequences, showing more than 65% nucleotide identity, are highlighted. Arrows represent the predicted coding sequences in IVSPER-2 (names as per Figure 1) and CsIV SH-C (as determined using FGENESV0 at http://linux1.softberry.com/berry.phtml). The p12 genes are located in a less conserved region that has undergone duplication in IVSPER-2 (repeats 1 and 2).

Figure 6. Schematic representation of a hypothetical scenario leading to the current organization of IV sequences from an unknown virus that integrated its own genome into the DNA of an ancestor wasp (ancestral virus).

The loss of ability of ancestral IVSPER genes to be encapsidated may result from a shift in the position of recognition signals involved in viral DNA encapsidation (ellipses in brown red) or from other evolutive events, followed by subsequent duplications and diversification in the gene content of these DNA regions, which would have led to the modern IVs.