This virus is whack. It exists as a set of genes inside the genome of parasitic wasps. At some point in the past, there was some sort of endosymbiotic event that incorporated the viral genome into the wasp genome. Along the way, they lost the genes necessary for viral transcription and replication. Okay, so far so normal, it happens, whatever.
But wait, here’s where it gets whack.
Inside the ovaries of the wasps, the viral genes are expressed, and are used to produce virus particles. The wasp reproduces by stinging a caterpillar to paralyze it, then injecting its eggs into the still-living paralyzed caterpillar, where they hatch and eat the caterpillar while it’s still alive, because nature is horrifying.
But wait! The polydnaviruses are produced in the wasp’s ovaries too, and the viruses are injected into a paralyzed caterpillar along with the eggs, where…
…the viruses destroy the caterpillar’s immune system, which prevents the immune system from destroying the eggs.
I told you it was whack.
This symbiosis is so weird and so complex that for a long time biologists didn’t believe the polydnaviruses were really viruses at all, but rather some kind of trick of the wasp’s own genome, some odd quirk that caused the wasp to make virus-like particles that acted like viruses in the caterpillar but really contained only wasp genes.
It wasn’t until relatively recently that the polydnavirus genome was sequenced and researchers were all like, “nope, those are viral genes, not wasp genes.” Specifically, polydnaviruses seem to have originated as nudiviruses, which were involved in an endosymbiosis event that saw the wasp hijack the virus instead of the virus hijacking the wasp as is more common.