30,000 Parasitoids Can't Be Wrong

by Merry

Doomed. The tomato hornworm, the larva of the
hawk moth, Manduca quinquemaculata, is parasi-
tized by the braconid wasp Cotesia congregata
that lays dozens of eggs within each larva. When
fully mature, each wasp larva tunnels to the
outside of the caterpillar's body and spins a
whitish cocoon. Source.

Over 30,000 species of parasitoid wasps can't be wrong. Au contraire, their reproductive strategy has been highly successful. They deposit their eggs inside the larvae (or eggs) of other insects, typically lepidopterans (butterflies and moths). Although commonly called parasitic wasps, they qualify as “parasitoids” because parasitization is always deadly for the host. Normally, a caterpillar would recognize a foreign invader, such as a wasp egg, and activate its immune defenses to encapsulate the foreign object. The wasp prevents this by depositing venom and what looks to be viruses in the caterpillar along with her eggs. The venoms are chemically complex and act in a variety of ways to convert the caterpillar into a nursery for wasp larvae. In some species, the venom paralyzes the caterpillar and/or arrests its development prior to pupation. For the moment, we'll call those "viruses" virus-like particles (VLPs), a term used for particles that look like a virus but whose infectivity and components have not yet been characterized. The VLPs suppress the host immune response and are required for successful parasitization.

Electron micrograph of VLPs deposited along with
their eggs by a braconid wasp, Glyptapanteles
indiensis. Each membrane envelope enclose 5-10
nucleocapsids. Source.

The two lineages of wasp parasitoids (the braconids and ichneumonids) independently evolved similar reproductive strategies—an elegant example of convergent evolution. More is known about the braconids, so we'll focus on their story, especially on the essential role of their VLPs. VLPs are produced only by females and only in a small number of specialized ovarian cells called calyx cells. VLP production begins when the level of molting hormones, called ecdysteroids, have peaked and initiated ovarian development early in pupation. VLPs are then produced throughout the female's adult life. The capsids of the VLPs are assembled in specialized regions of calyx cell nuclei. The VLPs are bounded by an envelope derived from the nuclear membrane and released into the oviduct by cell lysis. The capsids themselves are rod-shaped, 35-40 nm in diameter, varying between 30 and 100 nm in length—even in the same wasp. Structurally they resemble virions of the baculoviruses and their sister group, the nudiviruses. Both of those groups infect arthropods, particularly insects.

What are these particular VLPs? Well, viruses they’re not, at least not in the usual sense. Each particle contains no viral nucleic acid, but rather 10 to 30 circles of double-stranded DNA derived from the wasp genome, the number depending on the wasp species. Total packaged DNA ranges from ~150 to over 275 kb. When the VLPs are injected into a host larva along with the wasp eggs, the VLPs are seen to enter host cells very quickly—in as little as 2 hours, with the phagocytic hemocytes being most heavily infected. They enter the nucleus where their DNA is released and transcribed, but never replicated. The gene products play a vital role in the takeover of the host, including suppression of the immune response. (For a review of this topic, click here.)

The braconid wasp Cotesia congregata. Source.

This is a pretty unusual and highly sophisticated arrangement for the delivery of genes from a parasite to its host. So what genes are packaged and delivered to the caterpillar? Some answers are known for one of the braconid wasps, Cotesia congregata. You might be familiar with its host of choice, the caterpillar of the sphinx or hummingbird moth (Manduca sexta). These huge caterpillars, known as tobacco hornworms, often make their presence known in vegetable gardens where they noticeably defoliate prized tomato and pepper plants. The VLPs delivered with the eggs by this wasp contain 30 DNA circles. The DNA has been sequenced and 156 genes identified. The sequences of the DNA circles are found as linear genome segments dispersed throughout the wasp genome. In the ovarian calyx cells, those segments are selectively and intensively amplified, then excised and circularized for packaging in the VLPs.

None of the packaged sequences show similarity to any major virus gene. Two-thirds of them contain introns, a quarter of them are potential virulence factors. Strikingly, 66 genes are members of 9 gene families. The proteins encoded by 4 of these families contain domains found in toxins utilized by pathogenic bacteria or parasitic worms. The largest gene family (27 genes) encodes protein tyrosine phosphatases, key players in signal transduction cascades. The proteins encoded by the second largest family have akyrin repeat motifs that are typically found in eukaryotes where they participate in varied protein-protein interactions.

The abundance of gene families with members of a family distributed on several of the DNA circles suggests a history of gene duplication. Having multiple copies of these delivered genes may have been a key factor allowing the wasp to rapidly adapt to new hosts or to changes in host immune defenses. Indeed, although the VLP packaging mechanism is highly conserved among the braconid wasps, the contents vary greatly. Cross-infection experiments with three co-generic species demonstrated that the effectiveness of the VLPs is highly specific for particular host species.

Negatively stained VLPs from the ichneumonid
parasitoid Glypta fumiferanae. The ichneumonid
lineage evolved a similar strategy independently
of the braconids, deriving their capsid compo-
nents from a baculovirus. Several nucleocapsids
(N) are visible within each of two disrupted
membrane envelopes (E). Source.

If there is a capsid, then isn't there a virus lurking somewhere? Well, yes—long, long ago. Some of its genes remain in the wasp genome, to tell the tale. Recently, researchers identified 22 actively-transcribed virus-related genes in that same braconid wasp (C. congregata). These genes are most closely related to nudivirus genes. Some of them are in the core group that is shared by nudiviruses and baculoviruses. The researchers investigated a group of those genes further and found that they are integrated in the wasp genome, that they are expressed only in ovarian calyx cells that are producing VLPS, and that some of their protein products are present in the capsids. No viral genes are packaged in the VLPs. These viral-derived genes are highly conserved among the 17,000 species of this lineage of braconid wasps (up to 80% identity). In contrast, their similarity to the nudivirus genes is less than 60%—not too surprising given that this wasp-nudivirus association is estimated to have occurred 100 MYA ago.

Early on the VLPs were dubbed polydnaviruses. Knowing what we know now, would you call them viruses? Or parasitoid-wasp-gene-delivery devices?

Comments

[Nathan Myers]

Are we sure that wasps didn't themselves invent nudiviruses, which then, eventually, broke loose and achieved a more or less independent existence?

If so, can we say the same about bacteriophages?

Merry replies:

You have touched on a longstanding and much debated question: which came first, the virus or the cell? But in this case, you are asking about the origin of one particular family of viruses, not all viruses. The nudiviruses are clearly evolutionarily related to the rest of the viral world, not an independent lineage created in the wasps.

[Britt]

"Endless forms most beautiful and most wonderful!" I thought of these words (of Charles Darwin) as I read this article. This is a fabulous story.

One commentor suggested, "Maybe the wasps are just the viruses' way of reproducing themselves?"...I'm reminded of Dawkins' "The Selfish Gene". This wasp story is really amazing...parasite/parasitoid evolution is so fascinating. People that argue against evolutionary stories like this (or the evolution of a sophisticated eye, etc.) are really cheating themselves out of enjoying some real wonders of nature, real wonders of nature. (I said it twice because it was the subsject of convergent evolution.) Not only did/does it happen, but it happened more than once!

Another commentor posited, "I don't think we'd be here if not for them" but we could just as easily (more easily?) say the converse! Which came first, the virus or the host genome?

[Mark O. Martin]

Oh, and I loved this comment from Merry:

"Knowing what we know now, would you call them viruses? Or parasitoid-wasp-gene-delivery devices?"

Hmmm. Maybe the wasps are just the viruses' way of reproducing themselves? My head hurts.

[Mark O. Martin]

Merry, I know you are trying to stay away from the "definition of a virus" tar pit, but this sort of observation is relevant.

Me, I think that there is a broad spectrum of what we call "viruses." Personally (and without a jot of hard evidence), I think that in the "Pre-Darwinian" period posited by Carl Woese, these kinds of "packets" of information were shared very, very widely. Some "selfish nucleic acids" have learned to hijack the system and go into business for itself.

I keep thinking about gene transfer agent, of which STC has written before.

Great post that will make students scratch their heads---as it should be.

[Welkin]

Hi Merry, very nice article. Here's the pubmed ID for a highly speculative piece that Luis Villareal wrote about a decade ago for Journal of Virology, proposing that ancient retroviral particles expressed in placental tissues play a role in preventing maternal immunological rejection of the fetus. PMID: 8995601

Merry replies:
Thanks, Welkin. I nabbed a copy of the placental article to investigate. I am fascinated by the stories pointing to the roles played by viruses and kin in the evolution of diverse organisms. I don't think we'd be here if not for them. :-)

[Christopher Taylor]

The two lineages of wasp parasitoids (the braconids and ichneumonids) independently evolved similar reproductive strategies—an elegant example of convergent evolution.

Braconids and ichneumonids are sister taxa, so how do we know that polydnaviruses didn't evolve just once in their common ancestor? Do not all braconids and/or ichneumonids possess polydnaviruses?

(Merry replies)
Good point, Chris. All braconids and all ichneumonids so far do produce VLPs ("polydnaviruses"), but their VLPs have markedly different capsid structures and originated from associations with different viruses (a nudivirus in the case of the braconids, one or more baculoviruses for the ichneumonids). Phylogenetic studies suggest that these viral associations occurred after the two lineages diverged.

Although VLPs are produced in the nuclei of ovarian calyx cells in both genera, the mode of particle release differs: in ichneumonids, the VLPs emerge by budding; in braconids, VLPs are released by cell lysis. Other differences found involve the distribution of the encapsidated segments within the wasp genome, their mode of excision and replication, and the gene families they encode. Less than a handful of the 30,000+ wasps have been investigated, and the nudivirus origin of the braconid VLP capsid genes was only recently demonstrated. More surprises are bound to be in store in the future.