Previous studies have recognized a highly clonal population structure of T. gondii composed of three predominant lineages in NA and E, with rare variants being considered "exotic" (2, 3). Other studies have indicated that strains from SA were more genetically diverse, although their relationship to the clonal and exotic groups was unclear (11, 12). Here, we reconcile previous uncertainties about the population structure of T. gondii by demonstrating extreme geographic separation between strains that are predominant in NA and E vs. those that are common in several regions of SA. Overall, the population structure of T. gondii consists of discrete lineages that tend to be highly clonal and that show strong geographic restriction. Many of these clonal lineages occupy broad geographic ranges where they are found in multiple hosts. These data are most consistent with a model that strains in NA, E, and SA have a common ancestry, but that they have diverged and evolved in two distinct phases: (i) differentiation of northern and southern haplogroups occurred among geographically isolated populations, and (ii) a recent global sweep has homogenized many regions, driven by inheritance of a monomorphic version of ChrIa (Fig. 4).
Our analyses confirm that SA strains embody genetic diversity absent in NA and E (11) and extend this insight by recognizing 11 distinct haplogroups of T. gondii, four of which are almost exclusively found in SA (haplogroups 4, 5, 8, and 9), and one of which is globally distributed (haplogroup 6). Previously characterized "exotic" strains are now recognized as belonging to one of the common haplogroups, albeit existing outside their normal range. T. gondii infects a wide range of warm-blooded vertebrates, including birds, many of which are migratory. Hence it is unexpected that greater mixing has not occurred between continents. Examination of T. gondii strains from other regions of South and Central America may reveal evidence of such exchange, but are unlikely to alter the basic North vs. South dichotomy revealed here. Recognizing that most SA strains are demarcated by an entirely distinct array of polymorphic haplotypes, it is no longer appropriate to type SA strains using those markers that were designed to discriminate among strains in NA and E.
The finding of increased genetic diversity among SA strains using intron variation mirrors similar conclusions derived from microsatelite variation (11); however, we find little evidence for the suggestion that strains from SA are older. Rather, coalescent analysis indicates that variation in NA/E vs. SA appears to have accumulated over approximately equivalent time. Interestingly, the initial divergence between NA and SA populations of T. gondii coincides with the reconnection of the Panamanian land bridge (1–2 Mya during the upper Pliocene) after 
50 Mya of separation (dating back to the late Eocene), an event that facilitated the southerly migration (and subsequent speciation) of its definitive hosts, members of the cat family (Felidae) (18). Comigration of the parasite with its definitive host may therefore have led to the divergence of southern strains of T. gondii from preexisting northern ones (Fig. 4). Consistent with this, the genetic diversity of T. gondii is greater in the South, similar to the situation observed in felids (18).
Although recombination has generated a greater diversity of T. gondii genotypes in SA than elsewhere, clonal propagation also characterizes some SA haplogroups. Certain genotypes are repeatedly recovered from epidemiologically unrelated SA samples, and many others manifest strong LD in physically unlinked locus pairs and are tightly clustered in star-shaped phylogenies. Our analyses suggest that ancient population patterns in the North, and to a lesser extent in the South, have been substantially obscured by the comparatively recent emergence of a small number of clonal strains. The greater genetic diversity in the South, which has led others to conclude that lineages from this region are older (11), may instead reflect the incomplete penetrance of Mono-ChrIa in this region (Fig. 4).
Here, we extend to additional variants the earlier observation that genetic complexity among clonal lineages of T. gondii in NA and E can be explained by just a few genetic crosses between similar strains in the wild (6). Two unique parental strains, when separately crossed with a type II progenitor, likely gave rise to types I and III, respectively (6). The unique parental strain for type III was recognized to strongly resemble P89, then considered an "exotic strain" (6). Here, we demonstrate that P89 belongs to a prevalent SA haplogroup (group 9), which is highly similar to one of the four ancestral groups (Fig. 2B). The parental strain for type I was previously unknown, but our studies suggest it was derived from a progenitor that resembled haplogroup 6, another major ancestral group (Fig. 2B). Our findings indicate that limited genetic exchange among four ancestral populations may have given rise to all of the currently recognized extant haplogroups. The ancestral population was likely composed of lineages resembling haplogroups endemic to the North (haplogroup 2), the South (haplogroups 4, 9), and a pandemic group (haplogroup 6). Albeit infrequent, sexual recombination appears to have played an extremely important role in shaping the population structure of T. gondii.
Remarkably, clonally perpetuating strains in SA harbor the same invariant version of Chr1a (Mono-ChrIa) that characterizes the three clonal lineages in the North (8). It is striking that an entire chromosomal variant, estimated to have originated within the last 10,000 yr (8), has become established in genomes otherwise estimated to have last shared common ancestry 1 Mya. Collectively, these findings indicate a recent origin and rapid spread of Mono-ChrIa associated with the expansion of highly clonal lineages that have emerged over the last 10,000 yr (Fig. 4). The penetrance of Mono-ChrIa into other regions of the world (i.e., other regions of SA, Asia, and Africa) is presently unknown, because relatively few studies have explored the genetic diversity of T. gondii strains from these regions. Future studies need to adequately address genetic diversity by direct sequencing of selectively neutral loci and in particular assess the genetic homogeneity of ChrIa from broader regions of the world.
The extreme success of lineages that harbor Mono-ChrIa, which has rapidly spread through a wide range of hosts across much of three major continents, suggests it carries some major selective advantage. Several traits could potentially explain this, including: (i) enhanced oral transmission between intermediate hosts, thus bypassing the need for sexual reproduction; (ii) enhanced virulence leading to higher tissue burdens and hence greater transmission; and (iii) especially high reproductive fecundity in the cat, such that other stains are effectively outcompeted. Although Mono-ChrIa does not show a strong association with enhanced oral transmission or virulence in the laboratory mouse model, this highly permissive host may not best predict those phenotypes most important for transmission in the wild. Oral transmissibility of tissue cysts to intermediate hosts differentiates T. gondii from many, if not all, of its closest coccidian relatives, and its highly clonal population structure argues that such strictly asexual transmission contributes markedly to its exceptionally widespread dissemination.
Strains of T. gondii differ phenotypically despite their generally limited genetic variability, and certain genotypes cause especially severe human disease. Type I strains are especially prevalent in some studies of AIDS patients (19), highly divergent strains have been shown to present with atypical clinical severity (20), and certain SA strains have been associated with severe ocular disease (12). Understanding why Mono-ChrIa has swept disparate T. gondii populations may elucidate how genes responsible for virulence have spread through this parasite's population. Comparing the content and expression of its 120 genes (8) in parasites that harbor the Mono-ChrIa to those containing divergent versions of Chr1a may identify the selective basis for its recent sweep of T. gondii populations.
Answers to all your Biology Questions
