That’s why I liked using an integrating virus as a vector, and specified a super “tumor gene” attached to a strong constituitive enhancer. I saw lethality as separate from infection. The virus might infect a woman (a good way to help spread the virus), but wouldn’t kill her, because there was no suitable place in her genome for it to splice in and become tumorigenic. If didn’t want to infect women at all, you might choose a male receptor as the binding point for transfection (entry into the cell) and the DNA splicing would be your backup, to assure minimal lethality to women.
Though I agree that such a virus is unlikely to arise naturally, there is absolutely no reason that a natural mutant like this couldn’t cause a world-wide epidemic.
Chronos said “A germ might not care if it kills its hosts, but it will care if it kills them too quickly. Kill quickly, and you run out of hosts, and without hosts, the germ dies, too.” A virus doesn’t “care” if it dies or not, because it was never alive to begin with. It spreads if it can, like a biochemical machine. Chronos is thinking of strains that have been in the environment for a long time. This argument wouldn’t apply to new strains – how would it “know” that it would kill off all its hosts? It wouldn’t. An “anti-redhead virus” could sweep through the world population, wiping out all its hosts and be heard no more, but “being an evolutionary failure” like this wouldn’t keep it from existing the first time around. All species die out eventually. That doesn’t mean they couldn’t have existed -or flourished- in the first place.
I had, nontheless, envisioned a man-made plague. When the OP said “Barring some massive human conspiracy to intentionally infect only men” I assumed they meant infection specifically – a James Bondian “League of Man-Hating Women” – and didn’t mean to rule out genetically engineered viruses
Perhaps the most hypothetical part of the virus (the hardest to engineer) would be the “super tumor gene”. While it’s possible that such a thing exists or will be discovered by accident, I’d like to suggest something more within reach: a pakage of genes for proteins that bind or block several anti-tumor pathways.
Tumorigenesis is a complex process. It’s not enough to just go rogue, you have to grow and find some way to sustain yourself (e.g Dr. Judah Folkman’s work on angiogenesis in tumors - or “how tumors grow themselves a blood supply”) One advantage of using substances that block the central human anti-tumor pathways, you wouldn’t need to let the tumors grow much. When the infected cells overgrew and died, they would release their content of “anti-tumor blockers” alond with their load of replicating viruses, and trigger small tumors all around the body.
I should also explain the “enhancer”. Enhancers are a type of control element in DNA that cause nearby genes to be expressed more - up to hundreds of times as much as an identical gene that isn’t near an enhancer. They are sequences that don’t actually code for anything, but might provide (e.g.) a optimal site for RNA polymerase (the enzyme that copies the DNA blueprint into the RNA working copies that the ribosomes use to create proteins) to attach to the DNA. One the RNA polymerase attaches, it is “highly processive” – it chugs along, doing its thing like a train on a track or a paving machine.
