Toxoplasma gondii, an obligate intracellular parasite of the phylum Apicomplexa, has the unusual ability to infect virtually any warm-blooded animal. It is an extraordinarily successful parasite, infecting an estimated 30% of humans worldwide. The outcome of Toxoplasma infection is highly dependent on allelic differences in the large number of effectors that the parasite secretes into the host cell. Here, we show that the largest determinant of the virulence difference between two of the most common strains of Toxoplasma is the ROP5 locus. This is an unusual segment of the Toxoplasma genome consisting of a family of 4-10 tandem, highly divergent genes encoding pseudokinases that are injected directly into host cells. Given their hypothesized catalytic inactivity, it is striking that deletion of the ROP5 cluster in a highly virulent strain caused a complete loss of virulence, showing that ROP5 proteins are, in fact, indispensable for Toxoplasma to cause disease in mice. We find that copy number at this locus varies among the three major Toxoplasma lineages and that extensive polymorphism is clustered into hotspots within the ROP5 pseudokinase domain. We propose that the ROP5 locus represents an unusual evolutionary strategy for sampling of sequence space in which the gene encoding an important enzyme has been (i) catalytically inactivated, (ii) expanded in number, and (iii) subject to strong positive selection. Such a strategy likely contributes to Toxoplasma's successful adaptation to a wide host range and has resulted in dramatic differences in virulence.copy number variation | host-pathogen interaction T he evolutionary pressure on the molecules that form the interface between a pathogenic organism and its host has been likened to an arms race (1, 2) and has been proposed to give rise to highly polymorphic genes such as those that encode host MHC and viral capsid proteins. Whereas most pathogens have a single or restricted range of hosts with which they coevolve, the obligate intracellular parasite Toxoplasma gondii can infect virtually any cell of almost any warm-blooded animal, a remarkable feat from an evolutionary perspective.Different strains of Toxoplasma cause significantly different disease in mice (3) and humans as well (4, 5). In North America and Europe, three strains have grown to dominate Toxoplasma's genetic landscape, and they are characterized by distinct disease outcomes in mice: although a single Type I parasite is uniformly lethal to a mouse, Type II and III parasites have a lethal dose (LD 50 ) ranging from 10 2 to 10 5 (3). Sexual crosses between members of these lineages (Type I × Type III and Type II × Type III) (6, 7) have shown that virulence in mice segregates among F1 progeny (8). This has allowed for the discovery of a number of virulence loci using quantitative trait loci (QTL) mapping (9, 10).Toxoplasma secretes an arsenal of effector proteins into its host cell from specialized secretory organelles termed rhoptries and dense granules (11)(12)(13)(14). Given the selective pressure...