Mammals repeatedly borrowed viruses to build placentas
Several mammal lineages use proteins called syncytins to fuse placental cells. The genes encoding them descend from retroviral envelope genes that entered ancestral germ lines. Different lineages captured different viruses, so evolution appears to have recruited the same viral cell-fusion tool repeatedly rather than inheriting one universal syncytin.
A retrovirus must fuse its membrane with a host cell, and its envelope protein helps perform that task. If viral DNA becomes embedded in an egg or sperm ancestor, descendants can inherit it. Most such sequences decay, but some are co-opted. Human syncytin-1 is expressed in placental tissue and can drive cell fusion; blocking it in cultured trophoblast cells inhibits fusion. In mice, deleting particular syncytins disrupts placental layers and can prevent embryo survival. Comparative genomics reveals separate captures in primates, rodents, rabbits, carnivores and ruminants.
A boundary between ‘virus’ and ‘host’ can become historical rather than functional. Evolution does not only defend against invaders; it can retain a useful molecular operation and make it developmentally necessary. The repeated captures also show convergent evolution at the level of acquired genes, not just body shape.
Once a captured viral gene improves reproduction, selection can preserve and integrate it into regulatory networks. Later mutations elsewhere may evolve around its presence, making removal harmful. What began as an external parasite’s machinery can become part of normal development. ‘Foreign’ therefore describes origin, not permanent status. Genomes are records of incorporation as well as descent. This separation lets later evidence revise one part of the account without making the whole observation disappear.
Placental mammals share deep ancestry, but the named syncytins in different orders often come from unrelated retroviral insertions acquired at different times. The repeated solution is functional: an envelope protein already capable of membrane fusion is useful for creating multinucleated placental layers. This is stronger than noticing superficial similarity, yet it does not imply that every mammal uses syncytins in exactly the same tissue architecture. The important constraint is reproducibility of the relation, not exact repetition of every historical or material circumstance.
Does this mean mammals could not have placentas until a particular virus infected them?
No single infection explains all placentas. Placental structures pre-date many lineage-specific syncytin captures, and other fusion mechanisms may have preceded or accompanied them. The supported claim is narrower: several mammal lineages independently domesticated retroviral envelope genes and made them important—sometimes essential—for later placental development.
Demonstrating that the candidate genes lack retroviral ancestry or that their loss never affects placental fusion or development would remove the central evidence.